Merge PR #273

# Conflicts: # docs/parity.md
Merge PR #464
2026-10-09 12:04:03 -04:00 · 2026-10-09 12:03:42 -04:00 · 2026-10-09 12:02:17 -04:00 · 2026-10-09 12:02:10 -04:00 · 2026-10-09 11:56:38 -04:00 · 2026-10-09 11:52:23 -04:00
496 changed files with 120205 additions and 4444 deletions
@@ -0,0 +1,55 @@
+# Builds and tests the workspace on FreeBSD, inside a FreeBSD VM on a Linux runner
+# (GitHub has no FreeBSD runners). GPU tests skip themselves when the VM has no adapter.
+name: FreeBSD
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    paths:
+      - Cargo.toml
+      - Cargo.lock
+      - "**/Cargo.toml"
+      - .github/workflows/freebsd.yml
+  workflow_dispatch:
+
+concurrency:
+  group: freebsd-${{ github.ref }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read
+
+jobs:
+  freebsd:
+    name: FreeBSD 14 (build + test)
+    runs-on: ubuntu-latest
+    timeout-minutes: 300
+    steps:
+      - uses: actions/checkout@v4
+      # Optional font build input, so the Japanese font tests run (they skip without it).
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4
+        with:
+          repository: storytold/craft-fonts
+          ref: 8dcdacd5153e64560d109541a47d806f26f048c0
+          path: craft-fonts
+      - name: Build and test in a FreeBSD VM
+        uses: vmactions/freebsd-vm@c46abacb49f09938ca4e1702d15d836285d694cc # v1
+        with:
+          release: "14.3"
+          usesh: true
+          cpu: 4
+          mem: 12288
+          # Nothing to bring back; copying the target dir out of the VM fails on large builds.
+          copyback: false
+          prepare: |
+            pkg install -y git curl pkgconf cmake python3 libxkbcommon wayland libX11 libXcursor libXrandr libXi libxcb mesa-libs vulkan-loader gtk3 fontconfig freetype2 alsa-lib
+            curl -fsSL https://sh.rustup.rs | sh -s -- -y --profile minimal
+          run: |
+            . "$HOME/.cargo/env"
+            # The checkout is synced in under another owner; let tests that call git read it.
+            git config --global --add safe.directory "*"
+            rustc --version
+            export CRAFT_FONTS_DIR="$PWD/craft-fonts" CRAFT_FONTS_REQUIRED=1
+            CARGO_PROFILE_DEV_DEBUG=0 cargo build --workspace --all-targets --locked
+            CARGO_PROFILE_DEV_DEBUG=0 cargo test --workspace --locked
@@ -47,7 +47,22 @@ jobs:
      - name: XML, desktop entry, AppStream, Flatpak manifest
        run: |
          xmllint --noout packaging/windows/lightcraft.wxs packaging/macos/*.plist* packaging/linux/*.xml*
+          # Windows Installer needs Icon ids ending in .ico or .exe, or shortcuts show a blank icon.
+          if grep -oE '<Icon Id="[^"]*"' packaging/windows/lightcraft.wxs | grep -vE '\.(ico|exe)"$'; then
+            echo "::error::Icon Id must end in .ico or .exe"; exit 1
+          fi
          desktop-file-validate packaging/linux/ai.storyteller.lightcraft.desktop
          sed -e 's/@VERSION@/0.0.0/' -e 's/@DATE@/2026-01-01/' packaging/linux/ai.storyteller.lightcraft.metainfo.xml.in >"$RUNNER_TEMP/ai.storyteller.lightcraft.metainfo.xml"
          appstreamcli validate --no-net --explain "$RUNNER_TEMP/ai.storyteller.lightcraft.metainfo.xml"
-          python3 -c "import yaml; m = yaml.safe_load(open('packaging/linux/flatpak/ai.storyteller.lightcraft.yml')); assert m['id'] == 'ai.storyteller.lightcraft' and m['command'] == 'lightcraft'"
+          python3 - <<'EOF'
+          import yaml
+          src = yaml.safe_load(open("packaging/linux/flatpak/ai.storyteller.lightcraft.yml"))
+          bundle = yaml.safe_load(open("packaging/linux/flatpak/ai.storyteller.lightcraft.bundle.yml"))
+          for m in (src, bundle):
+              assert m["id"] == "ai.storyteller.lightcraft" and m["command"] == "lightcraft", m["id"]
+          for key in ("runtime", "runtime-version", "sdk", "finish-args"):
+              assert src[key] == bundle[key], f"Flatpak manifests disagree on {key}: {src[key]} vs {bundle[key]}"
+          print("Flatpak manifests ok")
+          EOF
+      - name: FreeBSD tarball layout (stub binaries)
+        run: bash packaging/freebsd/package.sh --dry-run
@@ -1,9 +1,12 @@
-# Release pipeline: every push to `release` builds signed installers for every platform and
-# creates or updates the draft GitHub Release "LightCraft v<version>". See docs/releasing.md.
+# Release pipeline: every push to `release` builds platform artifacts (signed where credentials
+# are configured) and creates or updates the draft GitHub Release "LightCraft v<version>".
+# See docs/releasing.md.
 #
-# All jobs run in the `release` environment (restricted to the `release` branch), which holds
-# the signing secrets. Every secret is optional: missing ones produce unsigned artifacts and a
-# warning, never a failed build.
+# The signing jobs (macOS, Windows) and the draft-release job run in the `release` environment
+# (restricted to the `release` branch), which holds the signing secrets. Every secret is optional:
+# missing ones produce unsigned artifacts and a warning, never a failed build. The jobs that sign
+# nothing (version, Linux, Flatpak, FreeBSD, web) run without it, so a dispatch on any branch is
+# a dry run of them; the release job needs the signed jobs, so a branch run never drafts a release.
 name: Release

 on:
@@ -35,7 +38,7 @@ jobs:
  version:
    name: version
    runs-on: ubuntu-latest
-    environment: release
+    # No `environment: release` here or on the other jobs that sign nothing: no release secrets.
    outputs:
      version: ${{ steps.v.outputs.version }}
      date: ${{ steps.v.outputs.date }}
@@ -63,13 +66,19 @@ jobs:
  macos:
    name: macOS universal
    needs: version
-    runs-on: macos-15
+    runs-on: macos-15-xlarge  # M2 Pro larger runner; use only in release.yml (actions_macos_xl budget caps spend)
    environment: release
    env:
      LIGHTCRAFT_VERSION: ${{ needs.version.outputs.version }}
      LIGHTCRAFT_BUILD_DATE: ${{ needs.version.outputs.date }}
    steps:
      - uses: actions/checkout@v4
+      # Fonts (optional build input; required for releases): see craftrules standards/fonts.md.
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4
+        with:
+          repository: storytold/craft-fonts
+          ref: 8dcdacd5153e64560d109541a47d806f26f048c0 # bump deliberately
+          path: craft-fonts
      - uses: dtolnay/rust-toolchain@stable
        with:
          targets: aarch64-apple-darwin, x86_64-apple-darwin
@@ -87,6 +96,8 @@ jobs:
        run: packaging/macos/import-cert.sh
      - name: Build, sign, notarize
        env:
+          CRAFT_FONTS_DIR: ${{ github.workspace }}/craft-fonts
+          CRAFT_FONTS_REQUIRED: "1"
          APPLE_ID: ${{ secrets.APPLE_ID }}
          APPLE_PASSWORD: ${{ secrets.APPLE_PASSWORD }}
          APPLE_TEAM_ID: ${{ secrets.APPLE_TEAM_ID }}
@@ -112,11 +123,20 @@ jobs:
        include:
          - { arch: x64, target: x86_64-pc-windows-msvc }
          - { arch: x86, target: i686-pc-windows-msvc }
+          # Windows on ARM: cross-compiled on the x64 runner, so signing and WiX run as above.
+          # .github/workflows/windows-arm64.yml installs and runs it on ARM64 hardware.
+          - { arch: arm64, target: aarch64-pc-windows-msvc }
    env:
      LIGHTCRAFT_VERSION: ${{ needs.version.outputs.version }}
      LIGHTCRAFT_BUILD_DATE: ${{ needs.version.outputs.date }}
    steps:
      - uses: actions/checkout@v4
+      # Fonts (optional build input; required for releases): see craftrules standards/fonts.md.
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4
+        with:
+          repository: storytold/craft-fonts
+          ref: 8dcdacd5153e64560d109541a47d806f26f048c0 # bump deliberately
+          path: craft-fonts
      - uses: dtolnay/rust-toolchain@stable
        with:
          targets: ${{ matrix.target }}
@@ -133,6 +153,8 @@ jobs:
      - name: Build, sign, package
        shell: pwsh
        env:
+          CRAFT_FONTS_DIR: ${{ github.workspace }}/craft-fonts
+          CRAFT_FONTS_REQUIRED: "1"
          WINDOWS_CERTIFICATE: ${{ secrets.WINDOWS_CERTIFICATE }}
          WINDOWS_CERTIFICATE_PASSWORD: ${{ secrets.WINDOWS_CERTIFICATE_PASSWORD }}
          AZURE_TENANT_ID: ${{ secrets.AZURE_TENANT_ID }}
@@ -154,7 +176,6 @@ jobs:
    # The oldest GitHub-hosted image, so the binaries only need glibc >= 2.35 (Ubuntu 22.04+,
    # Debian 12+, Fedora 36+, RHEL 10). See docs/releasing.md.
    runs-on: ${{ matrix.runner }}
-    environment: release
    strategy:
      fail-fast: false
      matrix:
@@ -167,12 +188,18 @@ jobs:
      NFPM_VERSION: 2.47.0
    steps:
      - uses: actions/checkout@v4
+      # Fonts (optional build input; required for releases): see craftrules standards/fonts.md.
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4
+        with:
+          repository: storytold/craft-fonts
+          ref: 8dcdacd5153e64560d109541a47d806f26f048c0 # bump deliberately
+          path: craft-fonts
      - name: Install build and packaging tools
        run: |
          sudo apt-get update
          sudo apt-get install -y --no-install-recommends \
            libxkbcommon-dev libwayland-dev libx11-dev libxrandr-dev libxi-dev libgl1-mesa-dev \
-            desktop-file-utils appstream python3-yaml
+            desktop-file-utils appstream python3-yaml zsync
          curl -fsSL -o "$RUNNER_TEMP/nfpm.deb" \
            "https://github.com/goreleaser/nfpm/releases/download/v${NFPM_VERSION}/nfpm_${NFPM_VERSION}_${{ matrix.deb }}.deb"
          sudo dpkg -i "$RUNNER_TEMP/nfpm.deb"
@@ -184,6 +211,9 @@ jobs:
        if: inputs.version != ''
        run: cargo xtask version set "$LIGHTCRAFT_VERSION"
      - name: Build and package (AppImage, deb, rpm, tar.gz)
+        env:
+          CRAFT_FONTS_DIR: ${{ github.workspace }}/craft-fonts
+          CRAFT_FONTS_REQUIRED: "1"
        run: packaging/linux/package.sh
      - name: Check the packages
        run: |
@@ -194,6 +224,9 @@ jobs:
          for appimage in dist/release/*.AppImage; do
            chmod +x "$appimage"
            APPIMAGE_EXTRACT_AND_RUN=1 "$appimage" --version
+            # Update information is embedded, and the .zsync for delta updates sits beside it.
+            "$appimage" --appimage-updateinformation | grep -F 'gh-releases-zsync|'
+            test -s "$appimage.zsync"
          done
      - uses: actions/upload-artifact@v4
        with:
@@ -201,17 +234,110 @@ jobs:
          path: dist/release/*
          if-no-files-found: error

+  flatpak:
+    name: Flatpak ${{ matrix.arch }}
+    needs: [version, linux]
+    # Repackages the Linux job's tarball (the same binaries as the AppImage/deb/rpm); no Rust
+    # build here. 24.04 for a newer flatpak-builder (1.4) than 22.04 ships.
+    runs-on: ${{ matrix.runner }}
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - { arch: x86_64, runner: ubuntu-24.04 }
+          - { arch: aarch64, runner: ubuntu-24.04-arm }
+    env:
+      LIGHTCRAFT_VERSION: ${{ needs.version.outputs.version }}
+      LIGHTCRAFT_BUILD_DATE: ${{ needs.version.outputs.date }}
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/download-artifact@v4
+        with:
+          name: release-linux-${{ matrix.arch }}
+          path: dist/linux
+      - name: Install flatpak-builder
+        run: |
+          sudo apt-get update
+          # librsvg2-common: the SVG pixbuf loader. flatpak-builder runs the host's
+          # `appstreamcli compose`, which renders the scalable icon and fails without it.
+          sudo apt-get install -y --no-install-recommends flatpak flatpak-builder librsvg2-common
+          # Ubuntu 24.04's AppArmor blocks the unprivileged user namespaces bubblewrap needs.
+          sudo sysctl -w kernel.apparmor_restrict_unprivileged_userns=0
+      - name: Build the Flatpak bundle
+        run: packaging/linux/flatpak-bundle.sh "dist/linux/lightcraft-$LIGHTCRAFT_VERSION-linux-${{ matrix.arch }}.tar.gz"
+      - uses: actions/upload-artifact@v4
+        with:
+          name: release-flatpak-${{ matrix.arch }}
+          path: dist/release/*.flatpak
+          if-no-files-found: error
+
+  freebsd:
+    name: FreeBSD x86_64
+    needs: version
+    # GitHub has no FreeBSD runners: build in a FreeBSD 14.3 VM, the same image and packages as
+    # .github/workflows/freebsd.yml.
+    runs-on: ubuntu-latest
+    timeout-minutes: 180
+    env:
+      LIGHTCRAFT_VERSION: ${{ needs.version.outputs.version }}
+      LIGHTCRAFT_BUILD_DATE: ${{ needs.version.outputs.date }}
+    steps:
+      - uses: actions/checkout@v4
+      # Fonts (optional build input; required for releases): see craftrules standards/fonts.md.
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4
+        with:
+          repository: storytold/craft-fonts
+          ref: 8dcdacd5153e64560d109541a47d806f26f048c0 # bump deliberately
+          path: craft-fonts
+      - name: Build and package in a FreeBSD VM (tar.gz)
+        # vmactions/freebsd-vm v1.5.9, pinned by commit (it runs our build).
+        uses: vmactions/freebsd-vm@c46abacb49f09938ca4e1702d15d836285d694cc
+        with:
+          release: "14.3"
+          usesh: true
+          cpu: 4
+          mem: 12288
+          # Passed into the VM by name; never templated into the script below.
+          envs: "LIGHTCRAFT_VERSION LIGHTCRAFT_BUILD_DATE LIGHTCRAFT_BUILD_SHA INPUT_VERSION"
+          # The FreeBSD CI packages, plus bash for the packaging script.
+          prepare: |
+            pkg install -y bash git curl pkgconf cmake python3 libxkbcommon wayland libX11 libXcursor libXrandr libXi libxcb mesa-libs vulkan-loader gtk3 fontconfig freetype2 alsa-lib
+            curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y --profile minimal --default-toolchain stable
+          run: |
+            set -e
+            . "$HOME/.cargo/env"
+            git config --global --add safe.directory "$PWD"
+            # Build outside the checkout so only dist/ is copied back out of the VM (copying the
+            # target dir back fails on large builds). 4 jobs: more runs the VM out of memory.
+            export CARGO_TARGET_DIR=/var/tmp/lightcraft-target CARGO_BUILD_JOBS=4
+            if [ -n "$INPUT_VERSION" ]; then cargo xtask version set "$LIGHTCRAFT_VERSION"; fi
+            # craft-fonts was checked out next to the sources and synced into the VM with them.
+            export CRAFT_FONTS_DIR="$PWD/craft-fonts" CRAFT_FONTS_REQUIRED=1
+            bash packaging/freebsd/package.sh
+            tar -tzf dist/release/lightcraft-*-freebsd-x86_64.tar.gz \
+              | grep -E '/bin/lightcraft$|/bin/lightcraft-cli$|/share/applications/ai\.storyteller\.lightcraft\.desktop$|/share/doc/lightcraft/LICENSE-MIT$'
+      - uses: actions/upload-artifact@v4
+        with:
+          name: release-freebsd-x86_64
+          path: dist/release/*
+          if-no-files-found: error
+
  web:
    name: Web (wasm)
    needs: version
    runs-on: ubuntu-latest
-    environment: release
    env:
      LIGHTCRAFT_VERSION: ${{ needs.version.outputs.version }}
      LIGHTCRAFT_BUILD_DATE: ${{ needs.version.outputs.date }}
      WASM_BINDGEN_VERSION: 0.2.129
    steps:
      - uses: actions/checkout@v4
+      # Fonts (optional build input; required for releases): see craftrules standards/fonts.md.
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4
+        with:
+          repository: storytold/craft-fonts
+          ref: 8dcdacd5153e64560d109541a47d806f26f048c0 # bump deliberately
+          path: craft-fonts
      - uses: dtolnay/rust-toolchain@stable
        with:
          targets: wasm32-unknown-unknown
@@ -221,13 +347,17 @@ jobs:
      - name: Install wasm-bindgen CLI
        run: |
          # Must match the wasm-bindgen version in Cargo.lock exactly (cargo xtask web checks it).
-          command -v wasm-bindgen >/dev/null && wasm-bindgen --version | grep -q "$WASM_BINDGEN_VERSION" \
-            || cargo install wasm-bindgen-cli --version "$WASM_BINDGEN_VERSION" --locked
+          if ! { command -v wasm-bindgen >/dev/null && wasm-bindgen --version | grep -q "$WASM_BINDGEN_VERSION"; }; then
+            cargo install wasm-bindgen-cli --version "$WASM_BINDGEN_VERSION" --locked
+          fi
          wasm-bindgen --version
      - name: Apply version override
        if: inputs.version != ''
        run: cargo xtask version set "$LIGHTCRAFT_VERSION"
      - name: Build and zip
+        env:
+          CRAFT_FONTS_DIR: ${{ github.workspace }}/craft-fonts
+          CRAFT_FONTS_REQUIRED: "1"
        run: packaging/web/package.sh
      - uses: actions/upload-artifact@v4
        with:
@@ -237,7 +367,7 @@ jobs:

  release:
    name: Draft GitHub Release
-    needs: [version, macos, windows, linux, web]
+    needs: [version, macos, windows, linux, flatpak, freebsd, web]
    runs-on: ubuntu-latest
    environment: release
    permissions:
@@ -0,0 +1,95 @@
+# Windows on ARM: package the ARM64 release the way release.yml does (cross-compiled on an x64
+# runner, unsigned here), then install that MSI on a Windows 11 ARM64 runner, check the installed
+# binaries are ARM64, run the CLI natively, and uninstall. (A native ARM64 `cargo test` job is left
+# out for now: PdfCraft saw an intermittent access violation in egui UI tests there.)
+name: Windows ARM64
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    paths:
+      - Cargo.toml
+      - Cargo.lock
+      - "**/Cargo.toml"
+      - "**/build.rs"
+      - .github/workflows/windows-arm64.yml
+      - packaging/windows/**
+  schedule:
+    - cron: "41 4 * * 1" # weekly, so drift in the runner images shows up
+  workflow_dispatch:
+
+concurrency:
+  group: windows-arm64-${{ github.ref }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read
+
+env:
+  CARGO_TERM_COLOR: always
+  CARGO_INCREMENTAL: 0
+
+jobs:
+  package:
+    name: package (cross-compiled on x64)
+    runs-on: windows-latest
+    timeout-minutes: 90
+    steps:
+      - uses: actions/checkout@v4
+      - uses: dtolnay/rust-toolchain@89b12181fb390509a0842a86cc55eeb8eb928c1d # stable
+        with:
+          targets: aarch64-pc-windows-msvc
+      - uses: Swatinem/rust-cache@6323deb102c322ba6fcbdcafc7e3dddab59af2b6 # v2
+        with:
+          key: windows-arm64-package
+      - name: Install WiX
+        shell: pwsh
+        run: dotnet tool install --global wix --version 5.0.2
+      - name: Build and package (unsigned)
+        shell: pwsh
+        env:
+          DIST: ${{ github.workspace }}\dist\arm64
+        run: ./packaging/windows/package.ps1 -Arch arm64
+      - uses: actions/upload-artifact@v4
+        with:
+          name: lightcraft-windows-arm64
+          path: dist/arm64/*
+          if-no-files-found: error
+          retention-days: 7
+
+  install:
+    name: install and run on ARM64
+    needs: package
+    runs-on: windows-11-arm
+    timeout-minutes: 30
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: lightcraft-windows-arm64
+          path: dist
+      - name: Install the MSI, check it, run it, uninstall it
+        shell: pwsh
+        run: |
+          $ErrorActionPreference = 'Stop'
+          $msi = (Get-ChildItem dist/*.msi | Select-Object -First 1).FullName
+          $p = Start-Process msiexec.exe -ArgumentList '/i', "`"$msi`"", '/qn', '/l*v', 'install.log' -Wait -PassThru
+          if ($p.ExitCode -ne 0) { Get-Content install.log -Tail 60; throw "msiexec /i exited $($p.ExitCode)" }
+          $dir = Join-Path $env:ProgramFiles 'LightCraft'
+          foreach ($exe in 'lightcraft.exe', 'lightcraft-cli.exe') {
+            $path = Join-Path $dir $exe
+            if (-not (Test-Path $path)) { throw "$exe not installed in $dir" }
+            $bytes = [System.IO.File]::ReadAllBytes($path)
+            $machine = [BitConverter]::ToUInt16($bytes, [BitConverter]::ToInt32($bytes, 0x3C) + 4)
+            if ($machine -ne 0xAA64) { throw "$exe is for machine 0x$('{0:X}' -f $machine), not ARM64" }
+            Write-Output "ok $path (ARM64)"
+          }
+          # The installed CLI runs natively on ARM64.
+          $cli = Join-Path $dir 'lightcraft-cli.exe'
+          $out = & $cli --version 2>&1 | Out-String
+          if ($LASTEXITCODE -ne 0) { Write-Output $out; throw "lightcraft-cli --version exited $LASTEXITCODE" }
+          Write-Output "ok lightcraft-cli --version: $($out.Trim())"
+          $p = Start-Process msiexec.exe -ArgumentList '/x', "`"$msi`"", '/qn' -Wait -PassThru
+          if ($p.ExitCode -ne 0) { throw "msiexec /x exited $($p.ExitCode)" }
+          if (Test-Path (Join-Path $dir 'lightcraft.exe')) { throw 'uninstall left lightcraft.exe behind' }
+          Write-Output 'ok uninstalled'
@@ -1,8 +1,16 @@
 *~
 plan/
 /target
+/target-*/
 /dist
 /corpus
 /log
 .DS_Store
 .claude/
+/craft-fonts
+# nix build / nix develop
+/result
+/result-*
+# model weights are never committed (SAM 3: docs/ai-masks.md)
+*.safetensors
+*.safetensors.part
@@ -1,12 +1,17 @@
 # LightCraft — instructions for agents

-LightCraft is a clean-room, open-source, pure-Rust photo library + non-destructive raw developer targeting Adobe Lightroom parity (and beyond). Native on macOS, Windows, Linux; web via WASM. Sibling of `../printcraft` (Acrobat), `../photocraft` (Photoshop), `../drawcraft` (Illustrator) and `../filmcraft` (Premiere), with the same conventions.
+LightCraft is a clean-room, open-source, pure-Rust photo library + non-destructive raw developer targeting Adobe Lightroom parity (and beyond). Native on macOS, Windows, Linux; web via WASM. Sibling of `../pdfcraft` (Acrobat), `../photocraft` (Photoshop), `../vectorcraft` (Illustrator) and `../filmcraft` (Premiere), with the same conventions.

 ## Start every session here
 1. Read `plan/STATUS.md` (current milestone, next unchecked task, blockers).
 2. Read the task in `plan/execution-plan.md` §3, the relevant section of `plan/architecture.md`, and the README/docs of the crate you touch. Behaviour/visual reference: `plan/lightroom/` (incl. `10-observed-ui.md` + screenshots).
-3. **Pick work** from [`docs/parity.md`](docs/parity.md) → *Top gaps* (Lightroom parity tracker: one row per feature, menu item and shortcut). When you land a feature, update its row(s) and the gap list in the same commit; `cargo xtask parity` (part of `ci`) checks that every `cmd:`/`ctl:` id and path the tracker cites still exists, and `cargo xtask parity --write` refreshes its summary.
-4. Follow the autonomous operation protocol (`plan/execution-plan.md` §7). Don't stop to ask unless §7 lists the decision as the user's.
+3. **Know where we stand:** read [`ROADMAP.md`](ROADMAP.md) → *Where we stand* (honest assessment by dimension and
+   by kind of user) and *Where we're going* (ordered priorities). The checklist counts features that *exist*; the real
+   gaps are quality and coverage: camera colour calibration, CR3 / compressed raws, per-model verification, render
+   fidelity against Lightroom, AI models, HDR / video / Classic modules. A ✅ row is not proof of parity — if you find a
+   ✅ feature that is wrong or incomplete, downgrade it to 🟡 with a note.
+4. **Pick work** from [`docs/parity.md`](docs/parity.md) → *Top gaps* (ordered by user impact; the tracker has one row per feature, menu item and shortcut). When you land a feature, update its row(s) and the gap list in the same commit, and the *Where we stand* / *Where we're going* sections of `ROADMAP.md` when a listed gap closes; `cargo xtask parity` (part of `ci`) checks that every `cmd:`/`ctl:` id and path the tracker cites still exists, and `cargo xtask parity --write` refreshes its summary.
+5. Follow the autonomous operation protocol (`plan/execution-plan.md` §7). Don't stop to ask unless §7 lists the decision as the user's.

 `plan/` is gitignored (local-only).

@@ -17,6 +22,34 @@ branches can still break each other (e.g. a new struct field vs. a new construct
 `git log main..HEAD` + `git status` for unmerged commits or uncommitted work before starting anything new. Commit small
 and often so a crash loses minutes, not hours.

+## Never crash (outranks feature work)
+People trust LightCraft with their photo libraries and edits; a crash loses their work. A malformed raw/JPEG/XMP, a bad
+command, control or MCP argument, a corrupt catalog or settings file, or a full disk must produce an error the user (or
+agent) can act on, never a panic. Don't ship a feature by adding a panic path, and fix a crash before building on top
+of it. Full standard: `../craftrules/standards/never-crash.md`
+([storytold/craftrules](https://github.com/storytold/craftrules/blob/main/standards/never-crash.md)).
+- **Non-test code never panics:** no `unwrap()`, `expect()`, `panic!`, `unreachable!`, `todo!`, `unimplemented!`, and
+  no `unsafe`. Return the crate's error type through `Result` and `?`; use `ok_or(..)?`, `let … else`, `if let`, or a
+  fallback (`unwrap_or…`) only where it can't silently corrupt a document. An unfinished feature returns an
+  "unsupported" error or is disabled. Sole exception: a provably infallible literal, as `#[allow(clippy::expect_used)]`
+  + `.expect("why it can't fail")`.
+- **`unsafe` lives only in `crates/sysmem`** (one FFI call, `malloc_zone_pressure_relief`, that returns freed
+  allocator pages to macOS after raw decodes). Every other production crate root has `#![forbid(unsafe_code)]`. A new
+  unsafe need goes in an isolated, well-tested helper crate like it: `// SAFETY:` on every block, a safe API, a safe
+  fallback where possible, and a line here naming it.
+- **Input-derived numbers are hostile:** `get()` instead of `[i]`/`[a..b]` for offsets from files, users, agents or
+  arithmetic on them; checked/saturating math for lengths, offsets and counts; no division by zero, NaN/inf or negative
+  casts to `usize`; cap allocations sized by input; slice strings only at char boundaries.
+- **Bound recursion** (depth limits or seen-sets: IFD chains, nested metadata, collections). **Don't cascade:**
+  `lock().unwrap_or_else(PoisonError::into_inner)` or an error; worker-thread joins are `Result`s.
+- **Last-resort guard:** a panic hook plus `catch_unwind` around command dispatch and import/export turns an escaped
+  panic into an error dialog and keeps the document. It's a safety net, not a licence; keep `panic = "unwind"` on native.
+- **Prove it:** every crash fix lands with a small synthetic regression test that panicked before the fix.
+- Enforced by clippy: root `clippy.toml` allows unwrap/expect/panic/indexing in tests only, and every production crate
+  root (`lib.rs`, each binary's `main.rs`) carries `#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic,
+  clippy::unimplemented, clippy::todo, clippy::unreachable)]`. Not `[workspace.lints]`: those would also hit
+  integration tests, examples and benches. New crates start with the attribute.
+
 ## Non-negotiables
 - **Clean-room.** Never read/disassemble anything inside Adobe app bundles (names/listings only). Never copy Adobe icons, presets, profiles (DCP), lens profiles (LCP), camera matrices, fonts. Observation of the installed Lightroom is read-only (it syncs the user's personal library: never import/edit/rate/delete there). Never copy GPL/LGPL/AGPL code (darktable, RawTherapee, ART, LibRaw, rawspeed, rawloader, rawler, lensfun, dcraw-derived GPL code…).
 - **Pure Rust** in the product. No C/C++ dependencies.
@@ -29,9 +62,25 @@ and often so a crash loses minutes, not hours.
 ## Assets: icons, images, fonts (ABSOLUTE RULE — never violate)
 - **Never use any iconography, image, artwork, font, sound or other asset from Adobe products** (no Lightroom/Creative Cloud icons, no screenshots, no presets/profiles/LUTs, no UI bitmaps — not even as a temporary placeholder or "reference copy"). Observing Adobe's UI to imitate *layout and behaviour* is allowed; copying or tracing its assets is not.
 - **This includes Adobe's open-licensed assets**: no Source Sans/Serif/Code or Source Han fonts, no Adobe Fonts, no
-  Adobe-published icon sets, sample photos, colour profiles or LUTs — even when OFL/MIT. The UI font is Inter (OFL).
+  Adobe-published icon sets, sample photos, colour profiles or LUTs — even when OFL/MIT. The UI font is Inter (OFL);
+  Japanese and Chinese fonts come from craft-fonts (below). **One exception (maintainer decision, 2026-10-07): Noto
+  Sans/Serif CJK** (Google-branded, OFL, co-developed with Adobe as Source Han) is allowed via craft-fonts for Chinese
+  text, because nearly every OFL Chinese face derives from it. Use it unmodified under its OFL; this does not open the
+  door to Source Han under Adobe's name or to any other Adobe asset.
 - Every asset in the repository must be one of: **our own original work** (e.g. icons drawn in code as vectors, procedurally generated demo photos), **public domain / CC0**, **Creative Commons** (CC-BY / CC-BY-SA with attribution honoured), **OFL** (fonts), or **permissive open-source** (MIT/Apache-2.0/BSD/ISC) — or contributed by a person who created the asset and licenses it openly.
+- **Exception: `docs/brand/`.** The ArtCraft name, wordmark and logos there are ArtCraft Team trademarks, not open source and not covered by LightCraft's MIT OR Apache-2.0 licence (`LICENSE-MIT`, `LICENSE-APACHE`, `NOTICE`); their terms are in `docs/brand/LICENSE-brand.txt`. Use them only unmodified and never redraw, recolour or derive from them.
 - **Every asset must have an entry in `assets/ATTRIBUTION.md`** (path, title, author/creator, source URL or "original work", licence, date added, modifications) and its licence text when required (e.g. `assets/fonts/OFL-*.txt`). Add the entry in the same commit as the asset. Assets without an attribution entry must not be committed.
+- **Fonts live in [storytold/craft-fonts](https://github.com/storytold/craft-fonts), never in this repo.** Don't commit
+  font files here (Inter, already in `assets/fonts/`, is the one exception); add new fonts to craft-fonts. LightCraft
+  uses it as the optional build input `CRAFT_FONTS_DIR`: `git clone https://github.com/storytold/craft-fonts ../craft-fonts`
+  then `CRAFT_FONTS_DIR=../craft-fonts cargo run -p lightcraft` (or any cargo/xtask command). `crates/engine/build.rs`
+  embeds the manifest's fonts as `lightcraft_engine::CRAFT_FONTS` (wasm32: BIZ UDPGothic Regular only); `cargo xtask web`
+  ships Noto Sans CJK SC Regular as a separate asset loaded by the UI before startup. The UI
+  (`theme::font_definitions`) and the export watermark renderer use CJK faces (picked by script) as fallbacks after Inter. Unset,
+  `CRAFT_FONTS` is empty: everything builds, tests and runs, but Chinese and Japanese text lack glyphs. Releases always build
+  with it (`release.yml`, `CRAFT_FONTS_REQUIRED=1`) and ship the fonts' OFL licences. Tests that need these fonts skip
+  without it; the FreeBSD CI job runs them with it. Rules: craftrules
+  [`standards/fonts.md`](https://github.com/storytold/craftrules/blob/main/standards/fonts.md).
 - Icons drawn in code (e.g. `crates/ui-egui/src/icons.rs`) are original work and are recorded in `assets/ATTRIBUTION.md` as such; do not trace them from Adobe icons.
 - Demo/test images: generated procedurally by `lightcraft-scenes`, or CC0 downloads kept in the gitignored `corpus/` with their source recorded. Screenshots of Adobe apps live only in the gitignored `plan/` and are never committed or published.
 - **Enforced:** `cargo xtask assets` (in `ci`) fails when an image/icon/font/sound/video/raw/ICC/XMP file is not matched
@@ -57,6 +106,7 @@ and often so a crash loses minutes, not hours.
 - Shell gotcha: `mv`/`cp` are aliased interactive here — use `/bin/mv -f` / `/bin/cp -f`.
 - Parallel agents: separate git worktrees and `CARGO_TARGET_DIR=target/agent-<name>`; each agent uses its **own control port** (pick one in 18000–19999, never the default 7980) and its own scratch subfolder (`<scratch>/<agent-name>/`) — never `rm -rf` shared paths; delete your target dir when done (disk is shared); keep every `Cargo.toml` valid at all times (the `crates/*` glob means one broken manifest breaks everyone).
 - Test corpora: `cargo xtask corpus --download` into `corpus/` (gitignored, CC0 only). Never commit media.
+- Shared real-file test corpora (Photoshop-authored PSDs, etc.) live in [`storytold/photocraft-corpus`](https://github.com/storytold/photocraft-corpus), explained in [craftrules `standards/test-corpora.md`](https://github.com/storytold/craftrules/blob/main/standards/test-corpora.md). Never commit large binary fixtures to this repo; fetch them pinned by commit and sha256-verified, as PhotoCraft does with `cargo xtask corpus`.

 ## Testing & performance (do this often)
 - Unit/property tests next to the code; end-to-end tests drive real binaries (`crates/mcp/tests/e2e.rs`,
@@ -73,4 +123,25 @@ and often so a crash loses minutes, not hours.
 ## Map of the code
 `geom`, `color`, `raster`, `tiff` (L0) → `raw`, `codecs`, `meta`, `develop` (L1) → `pipeline` → `catalog` → `engine`
 → `ui-egui`, `mcp` (L5) → apps `lightcraft` (desktop), `lightcraft-cli` (render/commands/MCP). `scenes` generates demo
-photos. `xtask` = tooling (`ci`, `layers`, `assets`, `parity`, `wasm`, `corpus`, `stats`).
+photos. `xtask` = tooling (`ci`, `layers`, `assets`, `parity`, `wasm`, `corpus`, `stats`). `flake.nix` +
+`nix/package.nix` = the Nix package (`nix build` builds both binaries with the craft-fonts input, installs the
+desktop file/icons/AppStream metadata and runs `cargo test --workspace`; `nix develop` = dev shell). Community-maintained and not
+in CI: it may lag behind the workspace; see README → Quick start.
+
+## Contributor credits (About window)
+
+- About ▸ Contributors/Models are compiled into the binary from `contributors/contributors.json`
+  (commit stats; generated, never hand-edit) and `contributors/people.toml` (names people chose for
+  themselves). See `docs/contributors.md`.
+- **Agents working for a contributor:** when you prepare a PR, check whether your human's GitHub
+  username has a `[people.<username>]` entry in `contributors/people.toml`. If not, ask them once
+  whether they want to be credited by more than their username: a real name, a display name, and/or
+  their public GitHub profile name (`sync_github_name = true`). If yes, add **only their own** entry
+  (copy the template at the top of the file, or run
+  `python3 ../../craftrules/scripts/contributors.py --add-me . --real-name "…" --sync-github-name`)
+  and include it in their PR, committed as them. If no, change nothing: they are credited as
+  `@username` anyway.
+- Never add, edit, guess or copy anyone else's entry or name (not from git config, commit authors or
+  GitHub profiles). Never hand-edit `contributors.json`.
+- Maintainers refresh the stats with `python3 ../../craftrules/scripts/contributors.py .` (it also
+  re-verifies who wrote each `people.toml` entry; `--check` only verifies).
@@ -1,12 +1,17 @@
 # LightCraft — instructions for agents

-LightCraft is a clean-room, open-source, pure-Rust photo library + non-destructive raw developer targeting Adobe Lightroom parity (and beyond). Native on macOS, Windows, Linux; web via WASM. Sibling of `../printcraft` (Acrobat), `../photocraft` (Photoshop), `../drawcraft` (Illustrator) and `../filmcraft` (Premiere), with the same conventions.
+LightCraft is a clean-room, open-source, pure-Rust photo library + non-destructive raw developer targeting Adobe Lightroom parity (and beyond). Native on macOS, Windows, Linux; web via WASM. Sibling of `../pdfcraft` (Acrobat), `../photocraft` (Photoshop), `../vectorcraft` (Illustrator) and `../filmcraft` (Premiere), with the same conventions.

 ## Start every session here
 1. Read `plan/STATUS.md` (current milestone, next unchecked task, blockers).
 2. Read the task in `plan/execution-plan.md` §3, the relevant section of `plan/architecture.md`, and the README/docs of the crate you touch. Behaviour/visual reference: `plan/lightroom/` (incl. `10-observed-ui.md` + screenshots).
-3. **Pick work** from [`docs/parity.md`](docs/parity.md) → *Top gaps* (Lightroom parity tracker: one row per feature, menu item and shortcut). When you land a feature, update its row(s) and the gap list in the same commit; `cargo xtask parity` (part of `ci`) checks that every `cmd:`/`ctl:` id and path the tracker cites still exists, and `cargo xtask parity --write` refreshes its summary.
-4. Follow the autonomous operation protocol (`plan/execution-plan.md` §7). Don't stop to ask unless §7 lists the decision as the user's.
+3. **Know where we stand:** read [`ROADMAP.md`](ROADMAP.md) → *Where we stand* (honest assessment by dimension and
+   by kind of user) and *Where we're going* (ordered priorities). The checklist counts features that *exist*; the real
+   gaps are quality and coverage: camera colour calibration, CR3 / compressed raws, per-model verification, render
+   fidelity against Lightroom, AI models, HDR / video / Classic modules. A ✅ row is not proof of parity — if you find a
+   ✅ feature that is wrong or incomplete, downgrade it to 🟡 with a note.
+4. **Pick work** from [`docs/parity.md`](docs/parity.md) → *Top gaps* (ordered by user impact; the tracker has one row per feature, menu item and shortcut). When you land a feature, update its row(s) and the gap list in the same commit, and the *Where we stand* / *Where we're going* sections of `ROADMAP.md` when a listed gap closes; `cargo xtask parity` (part of `ci`) checks that every `cmd:`/`ctl:` id and path the tracker cites still exists, and `cargo xtask parity --write` refreshes its summary.
+5. Follow the autonomous operation protocol (`plan/execution-plan.md` §7). Don't stop to ask unless §7 lists the decision as the user's.

 `plan/` is gitignored (local-only).

@@ -17,6 +22,34 @@ branches can still break each other (e.g. a new struct field vs. a new construct
 `git log main..HEAD` + `git status` for unmerged commits or uncommitted work before starting anything new. Commit small
 and often so a crash loses minutes, not hours.

+## Never crash (outranks feature work)
+People trust LightCraft with their photo libraries and edits; a crash loses their work. A malformed raw/JPEG/XMP, a bad
+command, control or MCP argument, a corrupt catalog or settings file, or a full disk must produce an error the user (or
+agent) can act on, never a panic. Don't ship a feature by adding a panic path, and fix a crash before building on top
+of it. Full standard: `../craftrules/standards/never-crash.md`
+([storytold/craftrules](https://github.com/storytold/craftrules/blob/main/standards/never-crash.md)).
+- **Non-test code never panics:** no `unwrap()`, `expect()`, `panic!`, `unreachable!`, `todo!`, `unimplemented!`, and
+  no `unsafe`. Return the crate's error type through `Result` and `?`; use `ok_or(..)?`, `let … else`, `if let`, or a
+  fallback (`unwrap_or…`) only where it can't silently corrupt a document. An unfinished feature returns an
+  "unsupported" error or is disabled. Sole exception: a provably infallible literal, as `#[allow(clippy::expect_used)]`
+  + `.expect("why it can't fail")`.
+- **`unsafe` lives only in `crates/sysmem`** (one FFI call, `malloc_zone_pressure_relief`, that returns freed
+  allocator pages to macOS after raw decodes). Every other production crate root has `#![forbid(unsafe_code)]`. A new
+  unsafe need goes in an isolated, well-tested helper crate like it: `// SAFETY:` on every block, a safe API, a safe
+  fallback where possible, and a line here naming it.
+- **Input-derived numbers are hostile:** `get()` instead of `[i]`/`[a..b]` for offsets from files, users, agents or
+  arithmetic on them; checked/saturating math for lengths, offsets and counts; no division by zero, NaN/inf or negative
+  casts to `usize`; cap allocations sized by input; slice strings only at char boundaries.
+- **Bound recursion** (depth limits or seen-sets: IFD chains, nested metadata, collections). **Don't cascade:**
+  `lock().unwrap_or_else(PoisonError::into_inner)` or an error; worker-thread joins are `Result`s.
+- **Last-resort guard:** a panic hook plus `catch_unwind` around command dispatch and import/export turns an escaped
+  panic into an error dialog and keeps the document. It's a safety net, not a licence; keep `panic = "unwind"` on native.
+- **Prove it:** every crash fix lands with a small synthetic regression test that panicked before the fix.
+- Enforced by clippy: root `clippy.toml` allows unwrap/expect/panic/indexing in tests only, and every production crate
+  root (`lib.rs`, each binary's `main.rs`) carries `#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic,
+  clippy::unimplemented, clippy::todo, clippy::unreachable)]`. Not `[workspace.lints]`: those would also hit
+  integration tests, examples and benches. New crates start with the attribute.
+
 ## Non-negotiables
 - **Clean-room.** Never read/disassemble anything inside Adobe app bundles (names/listings only). Never copy Adobe icons, presets, profiles (DCP), lens profiles (LCP), camera matrices, fonts. Observation of the installed Lightroom is read-only (it syncs the user's personal library: never import/edit/rate/delete there). Never copy GPL/LGPL/AGPL code (darktable, RawTherapee, ART, LibRaw, rawspeed, rawloader, rawler, lensfun, dcraw-derived GPL code…).
 - **Pure Rust** in the product. No C/C++ dependencies.
@@ -29,9 +62,25 @@ and often so a crash loses minutes, not hours.
 ## Assets: icons, images, fonts (ABSOLUTE RULE — never violate)
 - **Never use any iconography, image, artwork, font, sound or other asset from Adobe products** (no Lightroom/Creative Cloud icons, no screenshots, no presets/profiles/LUTs, no UI bitmaps — not even as a temporary placeholder or "reference copy"). Observing Adobe's UI to imitate *layout and behaviour* is allowed; copying or tracing its assets is not.
 - **This includes Adobe's open-licensed assets**: no Source Sans/Serif/Code or Source Han fonts, no Adobe Fonts, no
-  Adobe-published icon sets, sample photos, colour profiles or LUTs — even when OFL/MIT. The UI font is Inter (OFL).
+  Adobe-published icon sets, sample photos, colour profiles or LUTs — even when OFL/MIT. The UI font is Inter (OFL);
+  Japanese and Chinese fonts come from craft-fonts (below). **One exception (maintainer decision, 2026-10-07): Noto
+  Sans/Serif CJK** (Google-branded, OFL, co-developed with Adobe as Source Han) is allowed via craft-fonts for Chinese
+  text, because nearly every OFL Chinese face derives from it. Use it unmodified under its OFL; this does not open the
+  door to Source Han under Adobe's name or to any other Adobe asset.
 - Every asset in the repository must be one of: **our own original work** (e.g. icons drawn in code as vectors, procedurally generated demo photos), **public domain / CC0**, **Creative Commons** (CC-BY / CC-BY-SA with attribution honoured), **OFL** (fonts), or **permissive open-source** (MIT/Apache-2.0/BSD/ISC) — or contributed by a person who created the asset and licenses it openly.
+- **Exception: `docs/brand/`.** The ArtCraft name, wordmark and logos there are ArtCraft Team trademarks, not open source and not covered by LightCraft's MIT OR Apache-2.0 licence (`LICENSE-MIT`, `LICENSE-APACHE`, `NOTICE`); their terms are in `docs/brand/LICENSE-brand.txt`. Use them only unmodified and never redraw, recolour or derive from them.
 - **Every asset must have an entry in `assets/ATTRIBUTION.md`** (path, title, author/creator, source URL or "original work", licence, date added, modifications) and its licence text when required (e.g. `assets/fonts/OFL-*.txt`). Add the entry in the same commit as the asset. Assets without an attribution entry must not be committed.
+- **Fonts live in [storytold/craft-fonts](https://github.com/storytold/craft-fonts), never in this repo.** Don't commit
+  font files here (Inter, already in `assets/fonts/`, is the one exception); add new fonts to craft-fonts. LightCraft
+  uses it as the optional build input `CRAFT_FONTS_DIR`: `git clone https://github.com/storytold/craft-fonts ../craft-fonts`
+  then `CRAFT_FONTS_DIR=../craft-fonts cargo run -p lightcraft` (or any cargo/xtask command). `crates/engine/build.rs`
+  embeds the manifest's fonts as `lightcraft_engine::CRAFT_FONTS` (wasm32: BIZ UDPGothic Regular only); `cargo xtask web`
+  ships Noto Sans CJK SC Regular as a separate asset loaded by the UI before startup. The UI
+  (`theme::font_definitions`) and the export watermark renderer use CJK faces (picked by script) as fallbacks after Inter. Unset,
+  `CRAFT_FONTS` is empty: everything builds, tests and runs, but Chinese and Japanese text lack glyphs. Releases always build
+  with it (`release.yml`, `CRAFT_FONTS_REQUIRED=1`) and ship the fonts' OFL licences. Tests that need these fonts skip
+  without it; the FreeBSD CI job runs them with it. Rules: craftrules
+  [`standards/fonts.md`](https://github.com/storytold/craftrules/blob/main/standards/fonts.md).
 - Icons drawn in code (e.g. `crates/ui-egui/src/icons.rs`) are original work and are recorded in `assets/ATTRIBUTION.md` as such; do not trace them from Adobe icons.
 - Demo/test images: generated procedurally by `lightcraft-scenes`, or CC0 downloads kept in the gitignored `corpus/` with their source recorded. Screenshots of Adobe apps live only in the gitignored `plan/` and are never committed or published.
 - **Enforced:** `cargo xtask assets` (in `ci`) fails when an image/icon/font/sound/video/raw/ICC/XMP file is not matched
@@ -57,6 +106,7 @@ and often so a crash loses minutes, not hours.
 - Shell gotcha: `mv`/`cp` are aliased interactive here — use `/bin/mv -f` / `/bin/cp -f`.
 - Parallel agents: separate git worktrees and `CARGO_TARGET_DIR=target/agent-<name>`; each agent uses its **own control port** (pick one in 18000–19999, never the default 7980) and its own scratch subfolder (`<scratch>/<agent-name>/`) — never `rm -rf` shared paths; delete your target dir when done (disk is shared); keep every `Cargo.toml` valid at all times (the `crates/*` glob means one broken manifest breaks everyone).
 - Test corpora: `cargo xtask corpus --download` into `corpus/` (gitignored, CC0 only). Never commit media.
+- Shared real-file test corpora (Photoshop-authored PSDs, etc.) live in [`storytold/photocraft-corpus`](https://github.com/storytold/photocraft-corpus), explained in [craftrules `standards/test-corpora.md`](https://github.com/storytold/craftrules/blob/main/standards/test-corpora.md). Never commit large binary fixtures to this repo; fetch them pinned by commit and sha256-verified, as PhotoCraft does with `cargo xtask corpus`.

 ## Testing & performance (do this often)
 - Unit/property tests next to the code; end-to-end tests drive real binaries (`crates/mcp/tests/e2e.rs`,
@@ -73,4 +123,7 @@ and often so a crash loses minutes, not hours.
 ## Map of the code
 `geom`, `color`, `raster`, `tiff` (L0) → `raw`, `codecs`, `meta`, `develop` (L1) → `pipeline` → `catalog` → `engine`
 → `ui-egui`, `mcp` (L5) → apps `lightcraft` (desktop), `lightcraft-cli` (render/commands/MCP). `scenes` generates demo
-photos. `xtask` = tooling (`ci`, `layers`, `assets`, `parity`, `wasm`, `corpus`, `stats`).
+photos. `xtask` = tooling (`ci`, `layers`, `assets`, `parity`, `wasm`, `corpus`, `stats`). `flake.nix` +
+`nix/package.nix` = the Nix package (`nix build` builds both binaries with the craft-fonts input, installs the
+desktop file/icons/AppStream metadata and runs `cargo test --workspace`; `nix develop` = dev shell). Community-maintained and not
+in CI: it may lag behind the workspace; see README → Quick start.
@@ -3,7 +3,7 @@ resolver = "3"
 members = ["crates/*", "apps/*", "xtask"]

 [workspace.package]
-version = "0.1.1"
+version = "0.4.0"
 edition = "2024"
 license = "MIT OR Apache-2.0"
 rust-version = "1.90"
@@ -11,7 +11,11 @@ repository = "https://github.com/storytold/lightcraft"

 [workspace.dependencies]
 # internal
+lightcraft-denoise-core = { path = "crates/denoise-core" }
+lightcraft-denoise = { path = "crates/denoise", default-features = false }
 lightcraft-geom = { path = "crates/geom" }
+lightcraft-faces = { path = "crates/faces" }
+lightcraft-sysmem = { path = "crates/sysmem" }
 lightcraft-color = { path = "crates/color" }
 lightcraft-raster = { path = "crates/raster" }
 lightcraft-tiff = { path = "crates/tiff" }
@@ -28,9 +32,13 @@ lightcraft-merge = { path = "crates/merge" }
 lightcraft-engine = { path = "crates/engine" }
 lightcraft-ui-egui = { path = "crates/ui-egui" }
 lightcraft-mcp = { path = "crates/mcp" }
+lightcraft-fetch = { path = "crates/fetch" }
+lightcraft-heif = { path = "crates/heif" }
+lightcraft-segment = { path = "crates/segment" }

 # foundation
 thiserror = "2"
+sha2 = "0.10"
 serde = { version = "1", features = ["derive", "rc"] }
 serde_json = { version = "1", features = ["float_roundtrip"] }
 log = "0.4"
@@ -58,9 +66,14 @@ tiff = "0.11"
 image-webp = "0.2"
 moxcms = { version = "0.9", default-features = false, features = ["lut"] }
 jxl-oxide = { version = "0.12", default-features = false }
+zune-jpegxl = { version = "0.5", default-features = false, features = ["std"] } # MIT/Apache-2.0/Zlib lossless JXL encoder (test fixtures only)
 ravif = { version = "0.13", default-features = false, features = ["threading"] } # no "asm": it needs nasm (non-Rust)

 # gpu (crates/gpu; native backends only, no GL: everything stays pure Rust)
+# SAM 3 inference (lightcraft-segment). 0.9.2: later releases pull `tokenizers` with the
+# Oniguruma C library into candle-core, and the product is pure Rust.
+candle-core = { version = "=0.9.2", default-features = false }
+candle-nn = { version = "=0.9.2", default-features = false }
 wgpu = { version = "30", default-features = false, features = ["std", "parking_lot", "metal", "vulkan", "dx12", "wgsl"] }
 pollster = "1"

@@ -0,0 +1,177 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 ArtCraft Team and the LightCraft contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
@@ -1,12 +1,43 @@
 LightCraft
-Copyright the LightCraft contributors. Licensed under MIT OR Apache-2.0.
+Copyright (c) 2026 ArtCraft Team and the LightCraft contributors.
+Licensed under the MIT License or the Apache License, Version 2.0, at your option
+(LICENSE-MIT, LICENSE-APACHE).

-This product bundles third-party material under the following licences:
+The ArtCraft name, wordmark and logos in docs/brand/ are trademarks of the ArtCraft Team. They are
+not open source and are not covered by the license above; their terms are in
+docs/brand/LICENSE-brand.txt.

- Source Sans 3 (assets/fonts/SourceSans3-*) — SIL Open Font License 1.1. See assets/fonts/OFL-SourceSans3.txt.
- Inter (assets/fonts/Inter-*.ttf) — SIL Open Font License 1.1. See assets/fonts/OFL-Inter.txt.
+Third-party material

-Every bundled asset (fonts, icons, images) and its licence is listed in assets/ATTRIBUTION.md.
+This product bundles third-party material under its own licenses, listed below. The per-file
+list (author, source, license) is in assets/ATTRIBUTION.md.

-Algorithms are implemented from public specifications and papers (see crate READMEs). Rust dependencies
-(egui, eframe, image, rayon, serde, …) are permissively licensed (MIT/Apache-2.0/BSD/ISC/Zlib) and fetched through crates.io.
+Inter (assets/fonts/Inter-*.ttf) — SIL Open Font License 1.1. See assets/fonts/OFL-Inter.txt.
+Showcase photographs by Ansel Adams, Dorothea Lange and NASA (docs/images/) — public domain. See assets/ATTRIBUTION.md.
+
+crates/segment (lightcraft-segment) ports the SAM 3 model code of Hugging Face Transformers
+(https://github.com/huggingface/transformers, src/transformers/models/sam3 and sam3_tracker),
+Copyright The HuggingFace Team and Meta Platforms, Inc., licensed under the Apache License 2.0.
+The port was modified by the LightCraft contributors (2026): translated from Python/PyTorch to
+Rust on candle and restructured; each ported file says so. Because of this, crates/segment is
+licensed under the Apache License 2.0 only (LICENSE-APACHE), not under the MIT option.
+
+The SAM 3 weights are not part of LightCraft and are never distributed with it. A user downloads
+the facebook/sam3 checkpoint only when they choose to (in the app, or by hand), under Meta's SAM
+License (https://github.com/facebookresearch/sam3/blob/main/LICENSE), a separate, non-OSI
+licence. See docs/ai-masks.md.
+
+AI denoise weights are not part of LightCraft and are never distributed with it.
+The default build offers user-supplied models only. The optional rawnind-model
+feature offers a separately downloaded RawNIND UtNet2 Bayer checkpoint under
+GPL-3.0 (https://www.gnu.org/licenses/gpl-3.0.html), after the user accepts its terms.
+Commercial use is subject to those terms, including applicable source, licence and
+notice obligations on redistribution. Enabling that offer in releases remains a
+maintainer decision. The ONNX reader and CPU/GPU executor are our Rust implementation;
+no darktable, RawSpeed, LibRaw or other GPL decoder/runtime code is included.
+
+Builds made with craft-fonts (the optional CRAFT_FONTS_DIR build input; all official releases)
+also embed the fonts listed in https://github.com/storytold/craft-fonts/blob/main/ATTRIBUTION.md
+(the Japanese fonts BIZ UDPGothic, BIZ UDMincho and Shippori Mincho; the web build embeds only BIZ
+UDPGothic Regular), under the SIL Open Font License 1.1. Release packages include each font's
+licence as OFL-<family>.txt.
@@ -59,6 +59,7 @@
  <a href="#feature-status">Status</a> ·
  <a href="#quick-start">Quick start</a> ·
  <a href="ROADMAP.md">Roadmap</a> ·
+  <a href="#downloads">Downloads</a> ·
  <a href="#the-crafting-apps">Crafting Apps</a>
 </p>

@@ -200,7 +201,7 @@ A library that stays out of your way: **All Photos**, **Recently Added**, **Pick
 **Folders**, and **Recently Deleted**. Rate with <kbd>0</kbd>–<kbd>5</kbd>, flag with <kbd>P</kbd> / <kbd>X</kbd> /
 <kbd>U</kbd>, colour-label with <kbd>6</kbd>–<kbd>9</kbd>. Search understands fields:
 `rating:>3 flag:pick iso:>800 camera:x2 date:2026-04 keyword:mountains`. Every view sorts by capture date,
-import date, edit date, name, rating or size. The justified **Photo Grid** and **Square Grid** views are virtualized,
+import date, edit date, name, rating, size or at random (a stable shuffle; View → Sort → Reshuffle for a new one). The justified **Photo Grid** and **Square Grid** views are virtualized,
 so they stay smooth whether you have forty photos or forty thousand.

 <table>
@@ -233,11 +234,11 @@ lightcraft --control 7980 ~/Pictures/trip
 {"method": "ui.screenshot",    "params": {"path": "after.png"}}
 ```

- **91 engine commands** and **48 UI commands.** List them all with `engine.commands`; read every slider's range,
-  default and current value with `develop.controls`.
- **MCP server.** `lightcraft-cli mcp` gives Claude (or any MCP client) ~100 tools: import, query, develop, mask,
-  render (returned as an image), export. It runs headless, or attached to the running app with screenshots, clicks
-  and gestures. See [docs/mcp.md](docs/mcp.md).
+- **Command registry.** `engine.commands` lists the available commands; `develop.controls` lists every slider's
+  range, default and current value.
+- **MCP server.** `lightcraft-cli mcp` exposes the command registry to Claude (or any MCP client), alongside
+  helpers for import, query, develop, mask, render (returned as an image) and export. It runs headless, or attached
+  to the running app with screenshots, clicks and gestures. See [docs/mcp.md](docs/mcp.md).

  ```sh
  cargo build --release -p lightcraft-cli
@@ -257,9 +258,10 @@ lightcraft --control 7980 ~/Pictures/trip

 ## Fast, native, private

- **Pure Rust, no C.** Our own RAW decoders (DNG, Canon CR2, Sony ARW, Nikon NEF, Fujifilm RAF incl. X-Trans,
-  Panasonic RW2, Pentax PEF, Olympus ORF), our own colour science, our own pipeline. JPEG, PNG, TIFF, WebP, PSD
-  composites and JPEG XL open today.
+- **Pure Rust, no C.** Our own RAW decoders (DNG, Canon CR2/CR3, Sony ARW, Nikon NEF, Fujifilm RAF incl. X-Trans,
+  Panasonic RW2 / Leica RWL, Pentax PEF, Olympus ORF), our own colour science, our own pipeline. JPEG, PNG, TIFF, WebP,
+  PSD composites and JPEG XL open today; HEIC/HEIF photos too with `--features heif` (HEVC is a build-time choice,
+  as in PhotoCraft).
 - **Scene-referred & wide-gamut.** Linear Rec.2020 float internally, Bradford-adapted white balance, gamut mapping
  instead of clipping, a filmic shoulder for raw and pixel-exact pass-through for JPEGs you haven't touched.
 - **Resolution-independent edits.** Radii and brush sizes are relative to the image, so a 400 px preview, your
@@ -280,9 +282,18 @@ lightcraft --control 7980 ~/Pictures/trip

 ## Feature status

-LightCraft is young and moving fast; see the [roadmap](ROADMAP.md) for estimates. The detailed, row-by-row
-Lightroom parity tracker (every feature, menu item and shortcut with its status, evidence and the top gaps) is
-[docs/parity.md](docs/parity.md).
+LightCraft is young and moving fast. **Where we honestly stand** (details in the [roadmap](ROADMAP.md#where-we-stand)):
+
+- **By feature count we're at ~79%** of Lightroom (core features 98%), tracked row by row in
+  [docs/parity.md](docs/parity.md).
+- **As a day-to-day Lightroom replacement we're nearer 60–70%.** It's great for JPEG/DNG and most Nikon / Sony /
+  older-Canon raws on one machine.
+- **The biggest gaps:**
+  - **camera colour calibration:** Sony, Nikon, Panasonic, Fujifilm and Canon CR3 raws have guarded estimates from their camera JPEGs, with built-in ILCE-7M4, X-H2S and X-T4 profiles; measured calibration is missing, and other raws or rejected fits retain a neutral matrix;
+  - **compressed Olympus raws and unsupported CR3 variants:** these use embedded JPEG previews when present. Fujifilm lossless/lossy compressed RAF now decodes sensor data; [verification and existing-library reload instructions](docs/raf-compression.md);
+  - **AI masks and denoise:** subject and sky selection are classical heuristics;
+  - **HDR, video and the Classic Print / Book / Map modules.**
+- **What's next:** see [where we're going](ROADMAP.md#where-were-going).

 | Area | Status |
 |---|---|
@@ -292,17 +303,20 @@ Lightroom parity tracker (every feature, menu item and shortcut with its status,
 | Masking: brush, linear, radial, luminance/colour range, add/subtract/intersect | ✅ (AI subject/sky use classical heuristics for now) |
 | Crop, straighten tool + auto straighten, flip, rotate, aspect ratios, overlays | ✅ |
 | Profiles (Color, Neutral, Vivid, Landscape, Portrait, Monochrome: our own looks), presets, versions, history, copy/paste/sync settings | ✅ |
+| Camera colour: DNG files use their own matrices | ✅ DNG · 🟡 own Sony/Fujifilm profiles; measured calibration database missing |
 | Native macOS menu bar (generated from the command registry), control channel + every widget addressable, headless UI snapshots | ✅ |
-| RAW: DNG, CR2, ARW, NEF (uncompressed), Fujifilm RAF (Bayer + X-Trans), Panasonic RW2, Pentax PEF, Olympus ORF; embedded previews for every format incl. CR3 | ✅ · compressed NEF/RAF/ORF, CR3 decode 🚧 |
+| RAW: DNG, CR2, CR3 (lossless CRX Bayer and version 0x100/0x200 C-RAW), ARW, NEF (uncompressed + lossless/lossy compressed), Fujifilm RAF (uncompressed + lossless/lossy compressed, Bayer + X-Trans), Panasonic RW2 / Leica RWL / Panasonic RAW (every raw format, DMC-LX1 to DC-S1RM2), Pentax PEF, Olympus ORF (uncompressed), Samsung SRW (uncompressed); embedded previews for every format incl. CR3 | 🟡 · CR3, compressed ORF and compressed SRW decode ⬜ |
 | Detail: sharpening, luminance + colour noise reduction | ✅ · AI Denoise, Super Resolution ⬜ |
 | Remove / Heal / Clone spots (auto source), Visualize Spots, Red Eye and Pet Eye (auto pupil detection, catchlight) | ✅ · content-aware fill, spot pin editing 🚧 |
-| Export: JPEG / PNG / TIFF / WebP / AVIF, sizing, file-size limit, output sharpening, naming, batch, metadata policy, text watermark | ✅ · DNG export, image watermark ⬜ |
-| Library persistence (crash-safe op log + snapshots), disk thumbnail cache, import with duplicate detection | ✅ |
+| Export: JPEG / PNG / TIFF / WebP / AVIF / DNG / original, sizing, file-size limit, output sharpening, naming templates, batch, metadata policy, text or image watermark | ✅ · HDR export ⬜ |
+| Library persistence (crash-safe op log + snapshots, background compaction, failed saves reported), disk thumbnail cache | ✅ |
+| Import: Add in place / Copy / Move, rename and folder templates, devices, duplicate detection, watched folders; Local folder browsing | ✅ |
 | MCP server (headless or live app, persistent libraries), CLI, control channel | ✅ |
 | XMP sidecars (read/write, auto-write), reading `crs:` develop settings, preset files (`.lcpreset`, XMP presets) | ✅ |
-| Optics (distortion, vignetting, auto + manual CA, defringe, DNG-embedded lens corrections), Geometry (transforms, Constrain Crop), Upright (Auto/Level/Vertical/Full/Guided) | ✅ · camera lens profiles (our own) ⬜ |
+| Optics (distortion, vignetting, auto + manual CA, defringe, lens corrections embedded in DNG files and Panasonic / Leica RW2 / RWL distortion data), Geometry (transforms, Constrain Crop), Upright (Auto/Level/Vertical/Full/Guided) | ✅ · camera lens profiles (our own) ⬜ |
 | Photo Merge: HDR (auto-align, deghost), Panorama (spherical/cylindrical/perspective, boundary warp, auto crop), HDR Panorama → DNG | ✅ |
-| GPU pipeline (wgpu compute, CPU-exact within 1/255) with CPU fallback | ✅ · WebGPU in the browser 🚧 |
+| GPU pipeline (wgpu compute, CPU-exact within 1/255), CPU fallback on device limits / errors | ✅ · WebGPU in the browser 🚧 |
+| AI: segmentation masks, AI denoise, super resolution, faces; HDR editing; video | ⬜ (see [roadmap](ROADMAP.md#where-were-going)) |
 | Web build (same UI in the browser via WASM): persistent library in OPFS/IndexedDB, Web Worker rendering, export downloads | ✅ · WebGPU, Safari/Firefox testing 🚧 |

 <sub>✅ works today · 🚧 in progress · ⬜ not started</sub>
@@ -322,9 +336,69 @@ cargo run --release -p lightcraft-cli -- render photo.jpg -o out.jpg --set light
 cargo xtask ci                                          # fmt, clippy, tests, layering, wasm checks
 ```

+CI defaults to line-table debug information and one build job per 1.5 GB of RAM
+available when it starts (at most one per CPU; 4 if memory can't be read), so an
+8 GB machine with 6 GB free builds with 4, and a busy machine with little free
+builds with fewer. Test threads follow, capped at 4. Full-debuginfo linkers, one
+per CPU, otherwise exhaust RAM before any test runs. GPU coverage is unchanged. Explicit
+`CARGO_PROFILE_DEV_DEBUG`, `CARGO_BUILD_JOBS` and `RUST_TEST_THREADS` settings
+override these defaults. Ordinary development commands keep their own settings.
+
+**Chinese and Japanese text** need the shared font repo, an optional build input (official releases always include it):
+
+```sh
+git clone https://github.com/storytold/craft-fonts ../craft-fonts
+CRAFT_FONTS_DIR=../craft-fonts cargo run --release -p lightcraft
+```
+
+Without it LightCraft builds and runs the same, but Chinese and Japanese text have no glyphs. Fonts are never committed to this
+repo; see [craftrules `standards/fonts.md`](https://github.com/storytold/craftrules/blob/main/standards/fonts.md).
+
+**Interface language:** **Edit → Language** (English, 简体中文, 繁體中文（台灣）, 日本語, Português (Brasil), Deutsch) or **Settings → General →
+Language**; the choice applies immediately and persists. See [docs/localization.md](docs/localization.md).
+
+**Logs:** the desktop app writes its log to standard error and to `logs/lightcraft.log` in its settings folder
+(Linux `$XDG_CONFIG_HOME/lightcraft/logs/`, by default `~/.config/lightcraft/logs/`; macOS
+`~/Library/Application Support/LightCraft/logs/`; Windows `%APPDATA%\LightCraft\logs\`), never in the library. A
+launch from a desktop menu or the Dock has no terminal, so attach this file to a bug report (**Help ▸ Open Log
+Folder** shows it in the file manager). Each start moves the
+previous log to `lightcraft.1.log` and that one to `lightcraft.2.log`, so the log of a run that crashed survives the
+next start; the file stops growing at 16 MiB, `--version` and `--help` write none, and runs with
+`LIGHTCRAFT_NO_PREFS` log to standard error only. By default LightCraft's own crates log at `info` and everything else
+at `warn`. `LIGHTCRAFT_LOG=info` or `debug` works as before (that level for LightCraft's own crates, warnings and
+errors from the rest; any other value: warnings and errors only) and wins over `RUST_LOG`, which otherwise replaces
+the default with env_logger-style directives such as `RUST_LOG=debug` or `RUST_LOG=warn,lightcraft_pipeline=trace` (a
+directive ending in `*` covers every target starting with it, as in `lightcraft*=debug`). Panics are recorded there
+too, and still in `lightcraft-panics.log` in the temp folder. The logger is `apps/lightcraft/src/logging.rs`.
+
 The web build needs the `wasm32-unknown-unknown` target and the matching `wasm-bindgen` CLI
 (`cargo xtask web` prints the exact install command); see [docs/web.md](docs/web.md).

+**Nix** builds the desktop app and `lightcraft-cli` (the Nix build always includes the craft-fonts input, so Japanese
+text has glyphs):
+
+```sh
+nix run github:storytold/lightcraft                    # the desktop app
+nix build github:storytold/lightcraft                  # → ./result/bin/{lightcraft,lightcraft-cli}
+nix develop github:storytold/lightcraft                # rust toolchain + native deps + fonts
+```
+
+In a flake configuration (NixOS, home-manager, nix-darwin):
+
+```nix
+# flake.nix
+inputs.lightcraft.url = "github:storytold/lightcraft";
+# optional: build against your own nixpkgs instead of the one LightCraft pins
+# inputs.lightcraft.inputs.nixpkgs.follows = "nixpkgs";
+
+# then, in a NixOS or home-manager module (where `inputs` is in scope):
+nixpkgs.overlays = [ inputs.lightcraft.overlays.default ];   # makes `pkgs.lightcraft` available
+environment.systemPackages = [ pkgs.lightcraft ];           # home-manager: home.packages = [ pkgs.lightcraft ];
+```
+
+`nix build` installs the same desktop file, hicolor icons and AppStream metadata as the .deb/.rpm, and runs
+`cargo test --workspace` as its check phase (skip it with `pkgs.lightcraft.overrideAttrs { doCheck = false; }`).
+
 **Keyboard:** <kbd>G</kbd> grid · <kbd>D</kbd> detail · <kbd>E</kbd> edit · <kbd>C</kbd> crop · <kbd>M</kbd> masking ·
 <kbd>Shift</kbd>+<kbd>P</kbd> presets · <kbd>\\</kbd> original · <kbd>Y</kbd> before/after · <kbd>Z</kbd> zoom ·
 <kbd>J</kbd> clipping · <kbd>⌘Z</kbd> undo · <kbd>⌘/</kbd> all shortcuts.
@@ -344,14 +418,65 @@ Humans and agents follow the same rules, so read [AGENTS.md](AGENTS.md) first. T
  work from public specs and black-box observation.
 - **No Adobe assets, ever:** no icons, screenshots, presets, profiles, LUTs or fonts from Adobe products. Every
  image, icon and font in the repo is original, public domain, Creative Commons, OFL or permissively licensed, and has
-  an entry in [assets/ATTRIBUTION.md](assets/ATTRIBUTION.md) added in the same commit.
+  an entry in [assets/ATTRIBUTION.md](assets/ATTRIBUTION.md) added in the same commit. New fonts go to
+  [storytold/craft-fonts](https://github.com/storytold/craft-fonts), not here.
 - **Pure Rust**, enforced crate layering, everything is a command, and `cargo xtask ci` green before every commit
  (one task id per commit).
+- **Never crash.** Non-test code returns errors instead of panicking: no `unwrap()`, `expect()`, `panic!` or
+  `unsafe`, checked indexing on anything derived from input, and a regression test with every crash fix. Details in
+  [AGENTS.md](AGENTS.md#never-crash-outranks-feature-work).

 Questions, ideas or a bug you'd like to talk through first? Bring them to [Discord](https://discord.gg/artcraft).

 <br>

+## Downloads
+
+**New to LightCraft?** Download it from the [LightCraft page on getartcraft.com](https://getartcraft.com/apps/lightcraft). That's the easiest way to install it.
+
+**Want a specific build or format?** On GitHub, the [latest release](https://github.com/storytold/lightcraft/releases/latest) has every build listed below, and [all releases](https://github.com/storytold/lightcraft/releases) has earlier versions and their notes. `<ver>` in the file names is the version number, and `SHA256SUMS.txt` lists a checksum for every file.
+
+### Windows
+
+| Build | Installer | Portable |
+|---|---|---|
+| x64 (64-bit Intel/AMD) | `lightcraft-<ver>-windows-x64.msi` | `lightcraft-<ver>-windows-x64-portable.zip` |
+| arm64 (Snapdragon and other ARM PCs) | `lightcraft-<ver>-windows-arm64.msi` | `lightcraft-<ver>-windows-arm64-portable.zip` |
+| x86 (32-bit) | `lightcraft-<ver>-windows-x86.msi` | `lightcraft-<ver>-windows-x86-portable.zip` |
+
+Installers and executables are code-signed.
+
+### macOS
+
+| Build | File | Notes |
+|---|---|---|
+| App, universal (Apple silicon + Intel) | `lightcraft-<ver>-macos-universal.dmg` | Signed and notarized |
+| Command-line tool, universal | `lightcraft-cli-<ver>-macos-universal.zip` | Signed and notarized |
+
+### Linux
+
+| Format | x86_64 | aarch64 (ARM64) | Notes |
+|---|---|---|---|
+| AppImage | `lightcraft-<ver>-linux-x86_64.AppImage` | `lightcraft-<ver>-linux-aarch64.AppImage` | Runs anywhere; updates itself with [AppImageUpdate](https://github.com/AppImageCommunity/AppImageUpdate) (`.zsync` files) |
+| Flatpak | `lightcraft-<ver>-linux-x86_64.flatpak` | `lightcraft-<ver>-linux-aarch64.flatpak` | Sandboxed; `flatpak install --user <file>` |
+| Debian/Ubuntu | `lightcraft-<ver>-linux-x86_64.deb` | `lightcraft-<ver>-linux-aarch64.deb` | |
+| Fedora/RHEL/openSUSE | `lightcraft-<ver>-linux-x86_64.rpm` | `lightcraft-<ver>-linux-aarch64.rpm` | |
+| Tarball | `lightcraft-<ver>-linux-x86_64.tar.gz` | `lightcraft-<ver>-linux-aarch64.tar.gz` | Unpack anywhere |
+
+### FreeBSD
+
+| Build | File |
+|---|---|
+| x86_64 | `lightcraft-<ver>-freebsd-x86_64.tar.gz` |
+
+### Web (WebAssembly)
+
+| Build | File | Notes |
+|---|---|---|
+| Static site | `lightcraft-web-<ver>.zip` | Runs in a modern browser; host it on any static server |
+
+<br>
+
 ## The Crafting Apps

 LightCraft is one of the **Crafting Apps**: free, open-source creative tools from the
@@ -364,7 +489,7 @@ stand on its own.
 | <img src="https://raw.githubusercontent.com/storytold/vectorcraft/main/assets/app-icon/hicolor/64x64/apps/ai.storyteller.vectorcraft.png" alt="" width="32" height="32"> | **VectorCraft** | Vector illustration | [GitHub](https://github.com/storytold/vectorcraft) | [Website](https://getartcraft.com/apps/vectorcraft) |
 | <img src="https://raw.githubusercontent.com/storytold/filmcraft/main/assets/app-icon/hicolor/64x64/apps/ai.storyteller.filmcraft.png" alt="" width="32" height="32"> | **FilmCraft** | Video editing, color and sound | [GitHub](https://github.com/storytold/filmcraft) | [Website](https://getartcraft.com/apps/filmcraft) |
 | <img src="https://raw.githubusercontent.com/storytold/lightcraft/main/assets/app-icon/hicolor/64x64/apps/ai.storyteller.lightcraft.png" alt="" width="32" height="32"> | **LightCraft** | **Photo library and raw development · you are here** | [GitHub](https://github.com/storytold/lightcraft) | [Website](https://getartcraft.com/apps/lightcraft) |
-| <img src="https://raw.githubusercontent.com/storytold/printcraft/main/assets/app-icon/hicolor/64x64/apps/ai.storyteller.printcraft.png" alt="" width="32" height="32"> | **PrintCraft** | Reading, organizing and protecting PDFs | [GitHub](https://github.com/storytold/printcraft) | [Website](https://getartcraft.com/apps/printcraft) |
+| <img src="https://raw.githubusercontent.com/storytold/pdfcraft/main/assets/app-icon/hicolor/64x64/apps/ai.storyteller.pdfcraft.png" alt="" width="32" height="32"> | **PdfCraft** | Reading, organizing and protecting PDFs | [GitHub](https://github.com/storytold/pdfcraft) | [Website](https://getartcraft.com/apps/pdfcraft) |
 | <img src="https://raw.githubusercontent.com/storytold/effectcraft/main/assets/app-icon/hicolor/64x64/apps/ai.storyteller.effectcraft.png" alt="" width="32" height="32"> | **EffectCraft** | Motion graphics and visual effects | [GitHub](https://github.com/storytold/effectcraft) | [Website](https://getartcraft.com/apps/effectcraft) |
 | <img src="https://raw.githubusercontent.com/storytold/designcraft/main/assets/app-icon/hicolor/64x64/apps/ai.storyteller.designcraft.png" alt="" width="32" height="32"> | **DesignCraft** | Page layout and publishing | [GitHub](https://github.com/storytold/designcraft) | [Website](https://getartcraft.com/apps/designcraft) |

@@ -396,12 +521,23 @@ And [**ArtCraft**](https://getartcraft.com/) itself, our AI image and video stud

 ## License and credits

-LightCraft is licensed under **MIT OR Apache-2.0**.
+LightCraft is dual-licensed under [MIT](LICENSE-MIT) or [Apache-2.0](LICENSE-APACHE), at your option.
+Copyright (c) 2026 ArtCraft Team and the LightCraft contributors. Required notices are in [NOTICE](NOTICE).
+
+Bundled fonts, icons, images and other assets keep their own open licenses; each one is listed
+with its author, source and license in [assets/ATTRIBUTION.md](assets/ATTRIBUTION.md).

 Showcase photographs are public-domain works, used via Wikimedia Commons: Ansel Adams, *The Tetons and the Snake River*
 (1942, U.S. National Archives); Dorothea Lange, *Migrant Mother* (1936, Library of Congress); Bill Anders / NASA,
 *Earthrise* (1968); NASA, *The Blue Marble* (1972). The demo library is procedurally generated by LightCraft. UI font:
-Inter (SIL OFL). All icons are original. See [assets/ATTRIBUTION.md](assets/ATTRIBUTION.md).
+Inter (SIL OFL 1.1). Builds made with [craft-fonts](https://github.com/storytold/craft-fonts) (all official releases)
+also embed its Chinese and Japanese fonts (Noto Sans CJK SC, BIZ UDPGothic, BIZ UDMincho, Shippori Mincho; SIL OFL 1.1), listed in its
+[ATTRIBUTION.md](https://github.com/storytold/craft-fonts/blob/main/ATTRIBUTION.md). All icons are original.
+
+The ArtCraft name, wordmark and logos in [`docs/brand/`](docs/brand/) are trademarks of the
+ArtCraft Team and are not covered by this license. They may be used only unmodified, and only as
+part of this repository and LightCraft, under [`docs/brand/LICENSE-brand.txt`](docs/brand/LICENSE-brand.txt).
+Forks and modified versions must remove them.

 <sub>Adobe, Photoshop, Illustrator, Premiere Pro, Lightroom, Acrobat, After Effects and InDesign are trademarks or registered trademarks of Adobe Inc. in the United States and/or other countries. LightCraft is an independent, open-source project and is not affiliated with, sponsored by or endorsed by Adobe Inc.; these names are used only to describe the workflows it is compatible with.</sub>

@@ -409,3 +545,7 @@ Inter (SIL OFL). All icons are original. See [assets/ATTRIBUTION.md](assets/ATTR
  <a href="https://getartcraft.com/"><img alt="ArtCraft" src="docs/brand/artcraft-mark.svg" width="28"></a><br>
  <sub>Made by the <a href="https://getartcraft.com/">ArtCraft</a> team and community.</sub>
 </p>
+
+## Star history
+
+[![Star History Chart](https://api.star-history.com/svg?repos=storytold/lightcraft&type=Date&legend=top-left)](https://www.star-history.com/?repos=storytold%2Flightcraft&type=date&legend=top-left)
@@ -4,6 +4,77 @@ Milestones toward full Adobe Lightroom parity (cloud Lightroom first, then every
 estimates for continuous (24/7) agent-driven development with 4–6 parallel agents. Estimates are calibrated on the sibling
 projects (DrawCraft reached its first four milestones in ≈ 4½ h) and are revised as milestones land.

+## Where we stand
+
+*Honest assessment, 2026-10-08. Agents: read this before picking work. The checklist in
+[`docs/parity.md`](docs/parity.md) counts features that **exist**; this section is about whether a photographer can
+**switch** from Lightroom. Update it when a gap below closes.*
+
+**In one line:** the checklist says **79.7%** (P0 98.2%, P1 95.6%, P2 41.6%), but measured by whether a working
+photographer could replace Lightroom without noticing, we are at roughly **60–70%**. The remaining gap is mostly
+**quality of results and camera coverage**, not missing buttons.
+
+Caveat on the checklist: ✅ is set by whoever lands a feature, and nobody has systematically checked rows against
+Lightroom's behaviour or output. Bugs keep turning up in ✅ areas (CR2 colour-filter phase on some Canon models #85,
+duplicate Local entries #22, black GPU exports on an Intel iGPU #78).
+
+### By dimension
+
+| Dimension | Estimate | What's true today | Biggest gaps |
+|---|---:|---|---|
+| **Feature checklist** | 79.7% | P0 core and P1 nearly complete: import (Add / Copy / Move, templates, devices), library, grid/loupe/compare/survey, every Edit slider, curves, colour grading, masking tools, crop/Upright, heal/clone, presets/profiles, versions/history, sync, export, menus, shortcuts | P0: full-resolution zoom detail; P1: lens-profile database, content-aware fill (patch synthesis), video playback/trim |
+| **RAW coverage** (formats people shoot) | ~55% | DNG (all kinds), CR2, CR3 lossless Bayer and version 0x100/0x200 C-RAW (M50/R100/R8 pixel-exact corpus), ARW, NEF (uncompressed + Huffman lossless/lossy), RAF (uncompressed, lossless/lossy compressed Bayer and X-Trans), uncompressed ORF, uncompressed Samsung SRW (NX5 to NX1100, EX1, WB2000), RW2 / RWL / Panasonic RAW (every raw format, checked on 178 files from 118 Panasonic and Leica bodies), PEF; every container's embedded preview (incl. CR3) | **CR3** (every Canon since ~2018), compressed ORF, compressed Samsung SRW (NX1, NX30, NX300, NX500, NX2000, NX3000, NX3300, NX mini), Nikon lossy-after-split, Canon sRAW, HEIC/AVIF. NEFs labelled compressed but stored uncompressed (Z 6 packed 14-bit, D850 12-bit uncompressed). Per-model verification is thin (~55 corpus files vs >1,000 models) |
+| **Colour & image quality** | ~55–65% | Pipeline is complete and fast; GPU path CPU-exact within 1/255 | **No measured camera calibration database**: ARW, NEF, RW2, RAF, CR3, CR2 and PEF have guarded camera-JPEG colour estimates, with bundled ILCE-7M4, X-H2S and X-T4 profiles (docs/camera-preview-colour.md); other non-DNG raws and rejected estimates use a neutral matrix. Colour fidelity remains incomplete. No lens-profile database. No measured fidelity against Lightroom (tone, highlights, texture/clarity, NR, sharpening are tuned by eye) |
+| **AI & computational** | ~15–20% | Assisted culling (focus, bursts), auto tone, HDR/panorama merge; subject/sky/background masks as classical heuristics, or the photo's own segmentation mattes when it ships them (DNG semantic masks, e.g. iPhone ProRAW's sky matte); non-destructive Bayer RAW denoise with opt-in weights on pure-Rust CPU/GPU; **faces (MVP, [docs/faces.md](docs/faces.md))**: a pure-Rust detector (a 232 KB opt-in download), opt-in recognition models (one-click download or add a file; they run on the shared pure-Rust CPU interpreter; pinned SFace/AuraFace embeddings agree with an independent reference runtime on 40 public crops; one-thread medians 34–36 / 233–252 ms (AuraFace remains expensive)), a background scan that finds and embeds faces, name suggestions, a person's page with look-alike faces to confirm, an Unnamed faces section (look-alikes together, select a group and name it at once) | Real segmentation masks (subject, sky, people, objects, landscape, depth), denoise model policy / linear RGB / quality verification, super resolution, lens blur, generative remove, natural-language search; for faces: clusters with separators between groups, small faces in large photos (the detector looks at 640 px), thresholds calibrated on live faces, writing face regions back to XMP. **Blocked on a model strategy** (licensable weights or our own training; pure-Rust inference is feasible) |
+| **Workflow & library** | ~85% (single machine) | Robust catalog (journal + snapshots, background compaction, crash-tested), 85k-photo libraries stay responsive, Local browsing with automatic cleanup, XMP interop, keywords, smart albums, Move import | Opening an 85k library takes 1.7–4.7 s; no cloud sync (out of scope), no tablet companion (#74, roadmap), shared albums, publish services, tethering |
+| **Classic modules** | ~30% | Geotagging from GPX track logs, soft proofing (partial), slideshow (basic) | **Map view, Book, Print, Slideshow module, Web, publish services**: ~40 tracker rows ⬜ |
+| **HDR & video** | 0% | | HDR edit/display/export; video play/trim/edit/export |
+| **Platform & robustness** | ~70% | macOS native; Windows/Linux builds; web via WASM; no-panic lints workspace-wide, `unsafe` confined to `crates/sysmem`; failed saves are reported; GPU errors fall back to CPU | Windows installer UI unverified on Windows (PR #79); GPU path proven only on Apple + user reports; English, Simplified/Traditional Chinese, Japanese, Brazilian Portuguese and German UI (see docs/localization.md); remaining technical errors; Traditional Chinese uses the Simplified Chinese font until craft-fonts has a TC face; accessibility partial; headless UI tests time out under machine load |
+
+### By kind of user
+
+| User | Readiness | What blocks them |
+|---|---:|---|
+| JPEG / DNG shooter, single machine | ~85% | Fidelity polish, AI masks |
+| Nikon / Sony / Panasonic / older-Canon raw shooter | ~65% | Camera colour fidelity and coverage (ARW, NEF and RW2 preview estimates are only a starting point); Panasonic compacts and kit zooms also miss their embedded distortion correction |
+| Fujifilm raw shooter | ~65% | Uncompressed and lossless/lossy compressed RAF decoded; camera colour calibration remains missing. Verified on 13 bodies, including 14/16-bit GFX; see `docs/raf-compression.md` |
+| Canon CR3 / Olympus compressed-raw shooter | ~45% | CR3 lossless Bayer and version 0x100/0x200 C-RAW develop from sensor data (M50, R100, R8 verified pixel for pixel); other CRX variants and compressed ORF open as embedded JPEG previews. Camera colour and broader model verification remain |
+| Lightroom Classic power user | ~45% | Print, Book, Map, publish, tethering |
+| Relies on AI (masks, denoise) | ~25% | Object / Describe masks via optional SAM 3 (download not yet hosted); Bayer AI denoise needs separately installed weights; X-Trans/linear RGB, model policy and quality parity remain open |
+
+## Where we're going
+
+Priorities, in order. Each points at tracker rows in [`docs/parity.md`](docs/parity.md) → *Top gaps*.
+
+1. **Camera colour calibration of our own** (LR-PROF-CAMERACOLOR, P0): fit each camera to its own embedded JPEG, use
+   matrices the files carry themselves, then chart shots. Sony ARW, Nikon NEF, Panasonic RW2, Fujifilm RAF, Canon CR3/CR2,
+   Pentax PEF and Samsung SRW have the file-local fit (matrix + tone curve from their own JPEG; CR2 and PEF since issue #310, not yet
+   checked on the corpus); then ORF, then validate fidelity.
+2. **Raw formats, clean-room** (LR-IMP-FORMATS, P0): Remaining **unverified CR3 CRX** variants, compressed ORF, NEF
+   lossy-after-split, sRAW. Decided 2026-10-05: write our own decoders from prose descriptions (never decoder source,
+   no LGPL dependency); compressed NEF (#86) is the template.
+3. **Verified camera coverage** (LR-IMP-CAMERA-COVERAGE, P0): a CC0 sample per model in the corpus, each decoded and
+   checked for plausible colour; fix per-model bugs (#85).
+4. **Render fidelity suite** (LR-BEHAV-RENDER-FIDELITY, P1): measure our output against Lightroom on the same CC0 raws
+   (references stay in the local `plan/`), then tune against the numbers.
+5. **Lens profiles of our own** (LR-EDIT-OPTICS-PROFILE, P1).
+6. **AI model strategy** (maintainer decision): which permissively licensed models (or our own training) for
+   segmentation masks and denoise. Bayer denoise inference is implemented in pure Rust with an opt-in GPL model; release policy and independently licensed weights remain open. Unblocks M12 and Enhance. Face detection and recognition run on our own checked Rust engine. Object / Describe masks now
+   run Meta's SAM 3 in pure Rust (optional, user-consented download under the non-OSI SAM License; CDN mirrors still
+   to be set up, docs/ai-masks.md); Subject / Sky / People, Raw Details and Super Resolution are still open.
+7. **Then:** HDR (Q), the Classic output modules (Print first, then Map view, Book, Slideshow), video (R), localisation
+   and accessibility.
+
+Already closed in the week of 2026-10-03: compressed NEF (#10), Move import (#29), per-library smart-preview folder,
+Local roots / cleanup, grid and catalog performance at 85k photos (#35, #37), failed-save reporting, GPU export
+hardening (#78), copyright metadata (#51), GPX geotagging (#60), import tag help and folder templates (#31, #32).
+
+Image navigation now supports native pinch zoom and two-finger pan in Detail, Compare and Reference views.
+The whole-image preview still has a configurable size cap (2,560 px default), so 100% on larger photos does not
+yet guarantee original-pixel detail; visible-region rendering remains a gap (LR-VIEW-ZOOM).
+
+## Milestones
+
 **Status legend:** ✅ done · 🚧 in progress · ⬜ not started

 | # | Milestone | Scope (summary) | Estimate (h) | Status |
@@ -11,32 +82,35 @@ projects (DrawCraft reached its first four milestones in ≈ 4½ h) and are revi
 | M0 | Skeleton + visual shell | workspace, xtask CI + layering, geom/color/raster, develop model, pipeline v0, catalog v0, engine commands, Lightroom-look UI (grid, loupe, filmstrip, Edit panel), control channel, MCP, web build | 3–5 | ✅ |
 | M1 | Library core | import (JPEG/PNG/TIFF/WebP), EXIF/XMP, persistent catalog (op log + snapshots), albums, ratings/flags/labels, filter/search/sort, thumbnail cache, 100k-photo grid | 6–10 | ✅ (100k-photo grid scale test pending) |
 | M2 | Pipeline v1 (quality) | WB temp/tint, profiles, local tone mapping (highlights/shadows), curves, HSL, point colour, colour grading, texture/clarity/dehaze, vignette, grain, B&W, auto tone/WB, histogram, before/after | 10–15 | ✅ (look tuning vs our references ongoing) |
-| M3 | RAW I | TIFF/DNG (LJ92, deflate, tiles, opcodes), demosaic (AHD/PPG/bilinear), highlight recovery, DNG colour model, CR2, NEF, ARW, embedded previews | 10–15 | 🚧 (DNG, CR2, ARW, NEF uncompressed, embedded previews ✅; NEF Huffman ⬜) |
+| M3 | RAW I | TIFF/DNG (LJ92, deflate, tiles, opcodes), demosaic (AHD/PPG/bilinear), highlight recovery, DNG colour model, CR2, NEF, ARW, embedded previews | 10–15 | 🚧 (DNG, CR2, ARW, NEF uncompressed + Huffman lossless/lossy, embedded previews ✅; NEF lossy-after-split ⬜) |
 | M4 | Crop, geometry, optics | crop tool + overlays, straighten, Upright (auto/level/vertical/full/guided), manual transforms, CA, defringe, manual lens corrections | 6–10 | ✅ |
 | M5 | Performance | source pyramids, wgpu compute pipeline (CPU oracle), draft/full renders, prefetch, budgets (16 ms slider updates on 24 MP) | 10–15 | 🚧 (stage cache, source pyramid, wgpu pipeline, prefetch, memory budget ✅; colour NR at half resolution, GPU histogram ⬜) |
 | M6 | Masking | brush, linear/radial gradients, colour/luminance/depth range, add/subtract/intersect/invert, all local adjustments, masks panel | 8–12 | 🚧 (brush/linear/radial/colour/luminance range, add/subtract/intersect, masks panel ✅; depth range, AI masks ⬜) |
-| M7 | Detail | sharpening + masking preview, luminance/colour NR, Denoise, Raw Details, Super Resolution | 6–10 | 🚧 (sharpening, luminance/colour NR ✅; AI Denoise, Raw Details, Super Resolution ⬜) |
+| M7 | Detail | sharpening + masking preview, luminance/colour NR, Denoise, Raw Details, Super Resolution | 6–10 | 🚧 (sharpening, luminance/colour NR ✅; AI Denoise 🟡 (Bayer, optional weights; policy/fidelity open); Raw Details, Super Resolution ⬜) |
 | M8 | Heal / Remove | content-aware remove (PatchMatch), heal, clone, brush spots, visualize spots, red/pet eye | 6–10 | 🚧 (heal, clone, auto source, visualize spots, red/pet eye ✅; PatchMatch remove ⬜) |
 | M9 | Presets, profiles, versions, sync | preset browser + amount, create/import presets, profile browser, versions, history, copy/paste/sync settings | 5–8 | ✅ |
 | M10 | Export & share | export dialog (JPEG/PNG/TIFF/DNG/AVIF/JXL/original), sizing, sharpening, metadata, watermark, naming, batch jobs, XMP sidecars, HDR export | 6–10 | 🚧 (all formats incl. DNG/original, sizing, presets, background jobs ✅; JXL encode, HDR export ⬜) |
-| M11 | RAW II | CR3, RAF (X-Trans), ORF, RW2, PEF, SRW, 3FR, IIQ + long tail; camera calibration DB; HEIC/AVIF/JXL import | 20–35 | 🚧 (RAF uncompressed, RW2 packed, PEF, ORF uncompressed ✅; CR3, compressed NEF/ORF/RAF ⬜) |
+| M11 | RAW II | CR3, RAF (X-Trans), ORF, RW2, PEF, SRW, 3FR, IIQ + long tail; camera calibration DB; HEIC/AVIF/JXL import | 20–35 | 🚧 (CR3 lossless CRX and C-RAW, RAF uncompressed and lossless/lossy compressed, RW2 (every raw format), PEF, ORF uncompressed ✅; **camera colour calibration** 🚧 (guarded ARW, NEF, RW2 and RAF preview fitting; measured database still missing), CR3, compressed ORF ⬜) |
 | M12 | AI & smart features | subject/sky/background/people/object masks, semantic search, faces/People (permissively licensed models, pure-Rust inference) | 20–40 | ⬜ |
 | M13 | Merge | HDR merge (deghost), panorama (projections, boundary warp, fill edges), HDR panorama | 10–15 | ✅ |
 | M14 | Video | import/playback/trim via FilmCraft crates, global edits + presets on video, video export | 6–10 | ⬜ |
 | M15 | Classic modules | Map, Book, Slideshow, Print, Web; smart collections, stacks, virtual copies, publish services, tethering | 25–40 | 🚧 (smart albums, stacks, virtual copies, compare/survey ✅; Map/Book/Slideshow/Print/Web ⬜) |
-| M16 | 1.0 polish | preferences, shortcut editor, accessibility, localization, packaging (dmg/msi/AppImage/web), hardening | 10–20 | 🚧 (settings, keyboard shortcuts sheet, packaging basics ✅; accessibility, localisation ⬜) |
+| M16 | 1.0 polish | preferences, shortcut editor, accessibility, localization, packaging (dmg/msi/AppImage/web), hardening | 10–20 | 🚧 (settings, keyboard shortcuts sheet, packaging basics ✅; English/Chinese (Simplified, Traditional)/Japanese/Brazilian Portuguese/German localisation 🟡; accessibility and further locales ⬜) |

-## Parity estimate (2026-10-02, evening update)
+## Parity estimate (feature count updated 2026-10-08; effort estimate from 2026-10-02)

 **By feature count** — from `docs/parity.md` (one row per Lightroom feature, menu item and shortcut; `cargo xtask
 parity` prints this line on every run, so it stays current):

-| Scope | Weighted completion | Rows | Morning |
+| Scope | Weighted completion | Rows | 2026-10-02 |
 |---|---:|---:|---:|
-| P0 (core) | **98.7%** | 198 | 98.5% |
-| P1 (important parity) | **94.4%** | 144 | 79.9% |
-| P2 (later / AI / niche, incl. Classic modules) | **29.4%** | 158 | 11.7% |
-| **All in-scope rows** | **75.6%** | 500 | 65.7% |
+| P0 (core) | **98.2%** | 200 | 98.7% |
+| P1 (important parity) | **95.6%** | 149 | 94.4% |
+| P2 (later / AI / niche, incl. Classic modules) | **41.6%** | 160 | 29.4% |
+| **All in-scope rows** | **79.7%** | 509 | 75.6% |
+
+The 2026-10-05 count includes three new 🟡 rows that make quality gaps visible (camera colour calibration, verified
+camera coverage, render fidelity), which is why P0 dipped slightly.

 ✅ counts 1, 🟡 ½, ⬜ 0; out-of-scope rows (cloud sharing, Adobe accounts…) are left out.

@@ -50,7 +124,7 @@ trip, slideshow, auto import…) in ≈ 12 h on 2026-10-02; four to six parallel
 | Work package | Tracker rows | Agent-hours | Risk |
 |---|---|---:|---|
 | Remaining P0/P1 UI and library features (folder rename/move, keyword painter, people view…) | ≈ 8 | 5–10 | low |
-| Raw codecs: CR3 (CRX), compressed NEF / ORF / RAF, RW2 v4, HEIC/AVIF decode, JPEG XL DNG | LR-IMP-FORMATS | 40–80 | **high** — clean-room black-box analysis, no permissive specs |
+| Raw codecs: CR3 (CRX), compressed ORF, NEF lossy-after-split, HEIC/AVIF decode | LR-IMP-FORMATS | 40–80 | **high** — clean-room black-box analysis, no permissive specs |
 | Lens-profile database of our own (calibration targets, fitting, data) | LR-EDIT-OPTICS-PROFILE | 15–30 | data collection |
 | Video: playback, trim, edits, export (pure-Rust decode, ideally shared with FilmCraft) | R. Video | 20–40 | medium |
 | AI: subject / sky / background / people / object masks, object-aware remove, AI denoise, super resolution, lens blur, people & faces, natural-language search, culling | ≈ 30 | 80–150 | **high** — permissively licensed weights, pure-Rust inference, maybe training |
@@ -86,36 +160,76 @@ The milestone estimates in the table above were made before work started and are
  fallbacks first. No permissively licensed sky-segmentation or raw-denoise model was found — we may need to train our own.
 - **Camera colour and lens data** is a data problem: we never use Adobe's matrices, DCPs or LCPs. DNG-embedded data first,
  then our own calibration; long-tail camera/lens coverage grows over time.
- **Legal decisions pending:** whether GPL-licensed *prose* format descriptions (e.g. the public CR3 write-up) may be read
-  by a designated engineer to produce an internal spec; freedom-to-operate review for local Laplacian filters, PatchMatch and
-  HEVC (HEIC).
- **Look parity** with Adobe's default rendering is subjective tuning against our own reference targets.
+- **Raw-format sources:** decided 2026-10-05: decoders are written from *prose* format descriptions (even ones
+  published alongside GPL code); decoder source is never read. Still open: freedom-to-operate review for local
+  Laplacian filters, PatchMatch and HEVC (HEIC).
+- **Look parity** with Adobe's default rendering is tuned by eye today; the planned fidelity suite (LR-BEHAV-RENDER-FIDELITY)
+  turns it into measured comparisons against local-only Lightroom references.

 ## Raw format coverage and known gaps

+Sony ILCE-7M4 downsized lossless ARW (subsampled YCbCr 4:2:0 / 4:2:2 tiles) now decodes to linear RGB;
+private 3:2 & 4:3 examples verified through native rendering & full-resolution 16-bit export.
+
+CR3 lossless Bayer and version 0x100/0x200 C-RAW sensor decoding is verified pixel-exactly on Canon M50, R100 and R8; any other CR3, or one that fails to decode, opens from its embedded JPEG as before. See [`docs/cr3.md`](docs/cr3.md) for coding limits, reference provenance and colour limitations.
+
 Decoded (CC0 corpus from raw.pixls.us, `cargo xtask corpus --download`, `crates/raw/tests/corpus.rs`): DNG (uncompressed,
-LJ92, lossy JPEG / Smart Previews, Deflate, float, linear), CR2, ARW (uncompressed, ARW2, LJ92), NEF/NRW uncompressed, RAF uncompressed (Bayer and
-X-Trans), RW2 packed 12/14-bit, PEF (uncompressed and Huffman), ORF uncompressed (16-bit and 12-bit packed). Every
-supported container also yields its embedded JPEG preview (CR3 too), and the engine shows that preview for raw variants
-it can't decode yet.
+LJ92, lossy JPEG / Smart Previews, Deflate, float, linear, DNG 1.7 JPEG XL tiles: lossless ones bit-exact on synthetic files,
+lossy ones unclipped in the declared linear primaries, checked on one real file), CR2, ARW (uncompressed, ARW2, LJ92; as-shot white balance and black level of pre-2017 bodies from the enciphered
+maker-note `Tag2010` and the encrypted `SR2SubIFD`, both recovered by black-box analysis, `crates/raw/src/vendor/arw.rs`), NEF/NRW uncompressed and Huffman-compressed (lossless, lossy type 1/2, 12/14-bit), RAF uncompressed and lossless/lossy compressed (Bayer and
+X-Trans, 14/16-bit real-file verification), RW2 / Leica RWL / Panasonic RAW in every raw format (compressed formats 4 and 6, the prefix-coded strips of format 8,
+packed formats 2, 5 and 7, the 16-bit words of the 2005–2007 bodies; all recovered by black-box analysis of 178 CC0 files from
+118 bodies, `crates/raw/src/vendor/rw2.rs`), PEF (uncompressed and Huffman), ORF uncompressed (16-bit and 12-bit packed). Supported containers yield their embedded JPEG preview when present (including CR3). The engine uses it for RAW variants it cannot decode yet; HEVC-only CR3 previews are not decoded.

 Not decoded yet — preview only (no permissively licensed description; black-box analysis incomplete):
- **Nikon Huffman NEF** (lossless / lossy). Black-box findings so far: maker note `0x0096` holds per-parity predictor
-  seeds (2048 for 14-bit lossless); ≈ 9 bits per pixel; the category-0 code word is a 6-bit rotation of `011111`. The
-  Huffman tables are not stored in the files, and an automatic table search (beam search over prefix codes scored by
-  smoothness and bit rate, validated on Pentax files whose tables *are* stored) did not converge.
- **Panasonic RW2 raw format 4** (quantised): the block layout is known (0x4000-byte chunks rotated by 0x1ff8; 128-bit
-  blocks of two 12-bit seeds plus four groups of a 2-bit scale and three 8-bit codes; scales 0/1 are ×1/×2 differences
-  from the same-colour pixel two to the left), the reconstruction rule for scales 2/3 is not established.
- **Olympus compressed ORF**, **Fujifilm compressed RAF**, **Canon CR3/CRX** (M11.1), **Canon sRAW/mRAW**, lossy DNG.
+- **Nikon "lossy after split" NEF** (non-zero split row in maker note `0x0096`, e.g. some D3400/D5000/D5200/D5500/D5600
+  files): rows above the split decode with the regular lossy table; from the split row on a different code is used
+  that our black-box analysis has not recovered yet (none of the four regular tables fits, also not with byte
+  alignment or reset predictors at the split row). The other Nikon Huffman variants are decoded (`crates/raw/src/vendor/nefc.rs` documents the analysis).
+- **Olympus compressed ORF**, **Canon CR3 unverified CRX variants** (M11.1), **Canon sRAW/mRAW**.

-**Camera colour matrices:** non-DNG raws use the documented neutral fallback (camera RGB ≈ linear sRGB, flagged
+**Camera colour matrices:** ARW, NEF, RW2, RAF, CR3, CR2, PEF and Samsung SRW files can use guarded, separate chromaticity and tone estimates from their own embedded JPEG (see `docs/camera-preview-colour.md`); this is a per-file camera-look estimate with relative WB, not measured calibration or absolute-Kelvin WB. Other non-DNG raws and rejected fits use the documented neutral fallback (camera RGB ≈ linear sRGB, flagged
 `matrix_is_fallback`) with the file's as-shot white-balance multipliers. Clean sources to evaluate next: manufacturer
-matrices stored in the files themselves (Olympus ImageProcessing `ColorMatrix`, Pentax/Panasonic equivalents) and our
+matrices stored in the files themselves (Olympus ImageProcessing `ColorMatrix`, Pentax equivalents; Panasonic RW2 files carry none) and our
 own chart-based calibration (M11.4). Adobe matrices are never used.

+## Lightroom Classic catalog migration
+
+File → Import Lightroom Catalog… reads `.lrcat` plus committed WAL pages directly through a pure-Rust
+SQLite reader. Originals stay in place; ratings, flags, labels, XMP metadata, hierarchical keywords,
+collections/sets, virtual copies and supported develop settings migrate into LightCraft. Existing edits
+are preserved by default, reimport identities are persistent, and source settings/history/snapshots are
+archived before catalog mutation. Missing originals remain available for relinking.
+
+Rendering is approximate: unsupported Adobe profiles/AI/process settings are reported and archived;
+history/snapshots remain source data, and smart collections import current membership. The Lightroom
+database is read-only; LightCraft owns subsequent edits. See `docs/lightroom-catalog-import.md`.
+
 ## Log
 - 2026-09-30: roadmap created; M0 in progress; research docs (Lightroom reference, Rust imaging ecosystem) complete.
 - 2026-09-30 (later): app running with the full Lightroom-style UI; pipeline v0; DNG/CR2/ARW; README showcase. ≈ 8 h elapsed.
 - 2026-10-02: parity estimate added (xtask parity prints weighted completion); milestone statuses refreshed.
+- 2026-10-08: Fujifilm camera colour: guarded JPEG matching and relative WB, plus bundled X-H2S/X-T4 profiles from 201/545 photos. On 150 withheld files the pooled profiles improve mean ΔE slightly over per-file fitting (X-H2S 1.70 → 1.62, X-T4 2.43 → 2.34); dark X-H2S samples 8.74 → 2.25. Measured calibration remains open.
+- 2026-10-08: Fujifilm lossless/lossy compressed RAF: pure-Rust striped decoder; full sensor arrays verified against a black-box reference on 17 CC0 files and 7 supplied X-H2S / X-T4 files (13 camera bodies, 14/16-bit Bayer and X-Trans). Camera colour calibration remains a separate gap.
 - 2026-10-01: RAW II formats: RAF (uncompressed Bayer + X-Trans), RW2 (packed), PEF (incl. Huffman), ORF (uncompressed); embedded previews for every container incl. CR3; raw corpus test with 37 CC0 samples.
+- 2026-10-03 – 10-05: community PRs (#42–#51, #60) and 25+ issues landed: compressed NEF, Move import, folder templates,
+  Local roots and cleanup, 85k-photo grid and catalog performance, background compaction, failed-save errors, GPU
+  export hardening, clippy 1.99. Added the honest *Where we stand* assessment and *Where we're going* priorities; added
+  tracker rows for camera colour, camera coverage and render fidelity.
+- 2026-10-07: Panasonic RW2 / Leica RWL / Panasonic RAW decode in every raw format (compressed 4/6/8, packed 2/5/7,
+  16-bit words), black levels, in-camera aspect crops and mapped-out defects established on 178 CC0 files from 118
+  bodies; RW2 gets the file-local camera look (accepted on 140 of 174 files with a preview).
+- 2026-10-08: independent Rust CR3 lossless and version 0x100/0x200 C-RAW decoding, six exact full-sensor regressions on M50/R100/R8; any CR3 decode error falls back to the embedded JPEG; guarded CR3 camera-JPEG colour fitting with independent framing validation (`docs/cr3.md`). Unverified CRX variants and broader model verification remain.
+
+## Chinese and Japanese interfaces, and text watermarks
+
+English, Simplified Chinese, Traditional Chinese (Taiwan) and Japanese interface languages are persisted in UI state, and the
+language table (`crates/ui-egui/src/i18n.rs`) drives the menus, settings and fonts, so another
+language is a table entry plus two catalogs (docs/localization.md). Core menus and panels are
+translated; untranslated text retains English. CJK glyphs (Chinese: Noto Sans CJK SC; Japanese UI:
+BIZ UDPGothic; watermarks: BIZ UDMincho) come from storytold/craft-fonts, embedded by builds made
+with the optional `CRAFT_FONTS_DIR` input (all releases), so no system fonts are needed. Text watermarks now accept
+`vertical: true` in export JSON/presets and expose an orientation selector. Japanese
+characters stay upright in top-to-bottom columns, with newlines starting columns to the
+left. This is basic lettering, without tate-chu-yoko, ruby, kinsoku, or general vertical
+OpenType shaping. The same coverage renderer serves 8/16/32-bit exports.
@@ -15,6 +15,7 @@ workspace = true

 [dependencies]
 lightcraft-engine = { workspace = true }
+lightcraft-sysmem = { workspace = true }
 lightcraft-mcp = { workspace = true }
 lightcraft-ui-egui = { workspace = true }
 lightcraft-codecs = { workspace = true }
@@ -24,7 +25,14 @@ serde_json = { workspace = true }
 dhat = { workspace = true, optional = true }

 [features]
+default = ["denoise"]
+denoise = ["lightcraft-engine/denoise"]
+rawnind-model = ["denoise", "lightcraft-engine/rawnind-model"]
 # Heap profiling: `cargo run --release -p lightcraft-cli --features dhat-heap -- snapshot …` counts
 # every allocation (live/peak bytes in `library.memory` / `ui.inspect`) and writes dhat-heap.json
 # (open it in DHAT's viewer: allocation sites by bytes at the peak). Slows allocation-heavy work.
 dhat-heap = ["dep:dhat"]
+
+# Open HEIC/HEIF photos (the optional lightcraft-heif crate). Not default so distributors choose
+# it (HEVC patents); build with `--features heif` to include it.
+heif = ["lightcraft-codecs/heif"]
@@ -1,32 +1,12 @@
 //! Give memory the system allocator keeps after frees back to the system
-//! ([`lightcraft_engine::memory::set_release_hook`]).
-//!
-//! macOS's allocator keeps freed large blocks mapped as "reusable": they no longer count in the
-//! process's memory footprint but stay in its resident size until the system runs short. After a
-//! raw decode (hundreds of MB of temporaries) that made the resident size climb towards the sum of
-//! everything ever decoded. `malloc_zone_pressure_relief` (libSystem) returns those pages now.
-//! Elsewhere (glibc, Windows) blocks this large are unmapped when freed: nothing to do.
+//! ([`lightcraft_engine::memory::set_release_hook`]). The platform call lives in
+//! `lightcraft-sysmem`, the workspace's only crate allowed `unsafe`.

-/// Install the hook for this platform.
+/// Install the hook.
 pub fn install() {
-    #[cfg(target_os = "macos")]
-    lightcraft_engine::memory::set_release_hook(macos::release);
+    lightcraft_engine::memory::set_release_hook(release);
 }

-#[cfg(target_os = "macos")]
-#[allow(unsafe_code)]
-mod macos {
-    unsafe extern "C" {
-        /// libSystem malloc: release free memory of `zone` (all zones when null) back to the
-        /// system, up to `goal` bytes (0 = as much as possible). Returns the bytes released.
-        fn malloc_zone_pressure_relief(zone: *mut std::ffi::c_void, goal: usize) -> usize;
-    }
-
-    pub fn release() {
-        // SAFETY: a documented libSystem function taking a null zone (= every zone) and a size;
-        // it only returns free pages to the system and touches no memory we own.
-        unsafe {
-            malloc_zone_pressure_relief(std::ptr::null_mut(), 0);
-        }
-    }
+fn release() {
+    let _ = lightcraft_sysmem::release_free_memory();
 }
@@ -9,7 +9,10 @@
 //! lightcraft-cli synth-merge hdr|panorama -o DIR
 //! lightcraft-cli commands [--json]
 //! lightcraft-cli controls [--json]
+//! lightcraft-cli calibrate [--max N] [--out DIR] FOLDERS/FILES…
 //! ```
+#![forbid(unsafe_code)]
+#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)]

 mod alloc_release;

@@ -68,6 +71,8 @@ USAGE:
                          per line: {\"method\": \"ui.set\", \"params\": {\"view\": \"detail\"}}.
                          Replies go to stdout. `ui.screenshot` without a path writes -o (then
                          OUT-2.png, OUT-3.png…); `ui.settle {timeoutMs?}` waits for renders.
+                          A failed request (\"ok\": false) does not stop the script, but the exit
+                          status is non-zero when any request failed.
        -o, --output OUT  PNG path (a final screenshot is written here if the script took none)
        --size WxH        window size in points (default 1600x1000)
        --scale S         pixels per point (default 1)
@@ -82,6 +87,12 @@ USAGE:
      Write synthetic merge inputs (procedural scene; bracketed DNGs or overlapping PNG views).
  lightcraft-cli commands [--json]   list every command id with its parameters
  lightcraft-cli controls [--json]   list every develop control id with its range
+  lightcraft-cli calibrate [--max N] [--out DIR] FOLDERS/FILES…
+      Fit a colour profile per camera model from raw files and their embedded camera JPEGs
+      (Sony ARW, Nikon NEF, Fujifilm RAF): up to N files spread over the folders (default 300; 0 = all), pooled per
+      model, written as <model>.json to DIR (default: the profiles folder LightCraft reads,
+      <config>/camera-profiles, or $LIGHTCRAFT_CAMERA_PROFILES). Raws of a profiled model then
+      take their colour from the profile and only their tone from their own JPEG.
  lightcraft-cli --version | --help
 ";

@@ -89,6 +100,25 @@ USAGE:
 #[global_allocator]
 static ALLOC: dhat::Alloc = dhat::Alloc;

+/// Why `--library DIR` can't be opened; for a library open in another program, how to work with
+/// that one instead (issue #99).
+/// Print the open library's warnings (an unlockable library, damaged settings files) on stderr:
+/// stdout carries command results / MCP protocol messages only.
+fn library_warnings(session: &mut Session, who: &str) {
+    for w in session.take_library_warnings() {
+        eprintln!("{who}: warning: {w}");
+    }
+}
+
+fn library_error(dir: &str, e: lightcraft_engine::EngineError) -> String {
+    match e {
+        lightcraft_engine::EngineError::LibraryInUse(why) => format!(
+            "{dir}: {why}\nTo work with the library while the app has it open, start the app with `--control PORT` and use `lightcraft-cli mcp --connect 127.0.0.1:PORT`."
+        ),
+        e => format!("{dir}: {e}"),
+    }
+}
+
 fn main() -> ExitCode {
    // `--features dhat-heap`: count allocations; the profile is written when `_heap` drops
    // (LIGHTCRAFT_DHAT_FILE, default dhat-heap.json).
@@ -102,6 +132,9 @@ fn main() -> ExitCode {
        dhat::Profiler::builder().file_name(file).build()
    };
    alloc_release::install();
+    // Warnings (a GPU render redone on the CPU, an unknown backend name) on stderr; LIGHTCRAFT_LOG
+    // or RUST_LOG picks another level (#168).
+    lightcraft_engine::logging::install("lightcraft-cli");
    let args: Vec<String> = std::env::args().skip(1).collect();
    let r = match args.first().map(String::as_str) {
        Some("run") => run(&args[1..]),
@@ -112,6 +145,7 @@ fn main() -> ExitCode {
        Some("merge") => merge(&args[1..]),
        Some("synth-merge") => synth_merge(&args[1..]),
        Some("controls") => controls(&args[1..]),
+        Some("calibrate") => calibrate(&args[1..]),
        Some("--version" | "-V" | "version") => {
            println!("lightcraft-cli {}", env!("CARGO_PKG_VERSION"));
            Ok(())
@@ -131,6 +165,115 @@ fn main() -> ExitCode {
    }
 }

+/// Raw files below `path` (or `path` itself), skipping hidden and NAS metadata folders.
+fn raw_files(path: &Path, out: &mut Vec<std::path::PathBuf>, depth: usize) {
+    let is_raw =
+        |p: &Path| p.extension().and_then(|e| e.to_str()).is_some_and(|e| ["arw", "nef", "nrw", "raf"].iter().any(|x| e.eq_ignore_ascii_case(x)));
+    if path.is_file() {
+        if is_raw(path) {
+            out.push(path.to_path_buf());
+        }
+        return;
+    }
+    let Ok(entries) = std::fs::read_dir(path) else { return };
+    if depth > 32 {
+        return;
+    }
+    for entry in entries.flatten() {
+        let name = entry.file_name();
+        if name.to_string_lossy().starts_with(['.', '@']) {
+            continue;
+        }
+        raw_files(&entry.path(), out, depth + 1);
+    }
+}
+
+fn calibrate(args: &[String]) -> Result<(), String> {
+    let mut max = 300usize;
+    let mut out: Option<std::path::PathBuf> = None;
+    let mut inputs = Vec::new();
+    let mut i = 0;
+    while i < args.len() {
+        match args[i].as_str() {
+            "--max" => max = take_value(args, &mut i, "--max")?.parse().map_err(|_| "--max needs a number".to_string())?,
+            "--out" => out = Some(take_value(args, &mut i, "--out")?.into()),
+            a if a.starts_with("--") => return Err(format!("unknown option `{a}`")),
+            f => inputs.push(f.to_string()),
+        }
+        i += 1;
+    }
+    if inputs.is_empty() {
+        return Err("calibrate needs folders or raw files".into());
+    }
+    let mut files = Vec::new();
+    for input in &inputs {
+        raw_files(Path::new(input), &mut files, 0);
+    }
+    files.sort();
+    files.dedup();
+    let found = files.len();
+    if max > 0 && files.len() > max {
+        // spread over all folders (dates, scenes) rather than the first N
+        files = (0..max).filter_map(|k| files.get(k * found / max).cloned()).collect();
+    }
+    eprintln!("calibrate: {} of {found} raw files", files.len());
+    let dir = out.or_else(lightcraft_engine::camera_profiles::dir).ok_or("no profiles folder: pass --out DIR")?;
+    let workers = std::thread::available_parallelism().map_or(4, |n| n.get()).clamp(1, 6);
+    let next = std::sync::atomic::AtomicUsize::new(0);
+    let done = std::sync::atomic::AtomicUsize::new(0);
+    let pools: Vec<lightcraft_engine::camera_profiles::Pool> = std::thread::scope(|scope| {
+        let handles: Vec<_> = (0..workers)
+            .map(|_| {
+                scope.spawn(|| {
+                    let mut pool = lightcraft_engine::camera_profiles::Pool::default();
+                    while let Some(path) = files.get(next.fetch_add(1, std::sync::atomic::Ordering::Relaxed)) {
+                        let result = std::fs::read(path).map_err(|e| e.to_string()).and_then(|bytes| pool.add(&bytes));
+                        let n = done.fetch_add(1, std::sync::atomic::Ordering::Relaxed) + 1;
+                        match result {
+                            Ok(Some(model)) => eprintln!("[{n}/{}] {} ({model})", files.len(), path.display()),
+                            Ok(None) => eprintln!("[{n}/{}] {}: skipped (no usable camera JPEG or colour)", files.len(), path.display()),
+                            Err(e) => eprintln!("[{n}/{}] {}: {e}", files.len(), path.display()),
+                        }
+                    }
+                    pool
+                })
+            })
+            .collect();
+        handles.into_iter().filter_map(|h| h.join().ok()).collect()
+    });
+    let mut pool = lightcraft_engine::camera_profiles::Pool::default();
+    for p in pools {
+        pool.merge(p);
+    }
+    let mut written = 0;
+    for result in pool.fit(5) {
+        match result {
+            Ok(profile) => {
+                let path = lightcraft_engine::camera_profiles::save(&profile, &dir)?;
+                println!(
+                    "{}: {} photos, {} colour pairs, hue/saturation table {} → {}",
+                    profile.model,
+                    profile.files,
+                    profile.samples,
+                    profile.hue_sat.is_some(),
+                    path.display()
+                );
+                written += 1;
+            }
+            Err(e) => eprintln!("calibrate: {e}"),
+        }
+    }
+    for (model, n) in pool.files() {
+        if n < 5 {
+            eprintln!("calibrate: {model}: only {n} usable photo(s), at least 5 needed");
+        }
+    }
+    if written == 0 {
+        return Err("no profile written".into());
+    }
+    Ok(())
+}
+
 fn mcp(args: &[String]) -> Result<(), String> {
    let mut connect: Option<String> = None;
    let mut demo = false;
@@ -180,8 +323,9 @@ fn mcp(args: &[String]) -> Result<(), String> {
            let mut h = match &library {
                Some(dir) => {
                    let mut h = Headless::default();
-                    let r = h.session.open_library(dir, demo).map_err(|e| e.to_string())?;
+                    let r = h.session.open_library(dir, demo).map_err(|e| library_error(dir, e))?;
                    eprintln!("lightcraft-cli mcp: opened library {dir} ({r:?})");
+                    library_warnings(&mut h.session, "lightcraft-cli mcp");
                    h
                }
                None if demo => Headless::demo(),
@@ -198,7 +342,7 @@ fn mcp(args: &[String]) -> Result<(), String> {
    eprintln!("lightcraft-cli mcp: serving MCP on stdio ({})", backend.describe());
    let mut server = Server::new(backend).with_command_tools(!compact);
    let stdin = std::io::stdin();
-    server.serve(BufReader::new(stdin.lock()), std::io::stdout().lock()).map_err(|e| e.to_string())
+    server.serve(BufReader::new(stdin), std::io::stdout()).map_err(|e| e.to_string())
 }

 fn merge(args: &[String]) -> Result<(), String> {
@@ -247,7 +391,7 @@ fn merge(args: &[String]) -> Result<(), String> {
        }
        i += 1;
    }
-    let mut s = Session::new().with_fs();
+    let mut s = Session::new().with_fs().with_default_denoise_models();
    let paths = expand_paths(&files);
    let r = s.execute("library.import", &json!({"paths": paths})).map_err(|e| e.to_string())?;
    let mut ids: Vec<u64> = r["imported"].as_array().map(|a| a.iter().filter_map(Value::as_u64).collect()).unwrap_or_default();
@@ -283,14 +427,18 @@ fn synth_merge(args: &[String]) -> Result<(), String> {
    let mut written = Vec::new();
    match kind {
        "hdr" => {
-            for (k, b) in lightcraft_merge::synth::bracket_dngs(1800, 1200, &[-2.0, 0.0, 2.0]).into_iter().enumerate() {
+            for (k, b) in lightcraft_merge::synth::bracket_dngs(1800, 1200, &[-2.0, 0.0, 2.0]).map_err(|e| e.to_string())?.into_iter().enumerate() {
                let p = Path::new(&dir).join(format!("bracket-{k}.dng"));
                std::fs::write(&p, b).map_err(|e| e.to_string())?;
                written.push(p);
            }
        }
        "panorama" | "pano" => {
-            for (k, v) in lightcraft_merge::synth::pano_views(1200, 900, 1000.0, &[-50.0, -25.0, 0.0, 25.0, 50.0]).into_iter().enumerate() {
+            for (k, v) in lightcraft_merge::synth::pano_views(1200, 900, 1000.0, &[-50.0, -25.0, 0.0, 25.0, 50.0])
+                .map_err(|e| e.to_string())?
+                .into_iter()
+                .enumerate()
+            {
                let p = Path::new(&dir).join(format!("view-{k}.png"));
                let img = v.to_srgb8();
                let png = lightcraft_codecs::encode_png(&lightcraft_codecs::EncodeImage::rgba8(&img), &lightcraft_codecs::EncodeMeta::default())
@@ -399,39 +547,40 @@ fn run(args: &[String]) -> Result<(), String> {
    if steps.is_empty() {
        return Err("run: no command given (see `lightcraft-cli commands`)".into());
    }
-    let mut backend: Box<dyn Backend> = match connect {
-        Some(addr) => {
-            if demo || library.is_some() || !imports.is_empty() {
-                return Err("--demo, --library and --import apply to headless mode only".into());
-            }
-            Box::new(
-                Remote::connect(&addr)
-                    .map_err(|e| format!("LightCraft is not reachable at {addr} ({e}); start it with `lightcraft --control 7980`"))?,
-            )
-        }
-        None => {
-            let mut h = match &library {
-                Some(dir) => {
-                    let mut h = Headless::default();
-                    h.session.open_library(dir, demo).map_err(|e| e.to_string())?;
-                    h
-                }
-                None if demo => Headless::demo(),
-                None => Headless::default(),
-            };
-            if !imports.is_empty() {
-                let paths = expand_paths(&imports);
-                let r = h.session.execute("library.import", &json!({"paths": paths})).map_err(|e| e.to_string())?;
-                // what follows acts on the imported photos (already-known files are reported as duplicates)
-                let mut ids: Vec<Value> = r["imported"].as_array().cloned().unwrap_or_default();
-                ids.extend(r["duplicates"].as_array().into_iter().flatten().filter_map(|d| d.get("existing").filter(|v| v.is_u64()).cloned()));
-                if let Some(first) = ids.first().cloned() {
-                    h.session.execute("library.select", &json!({"ids": ids, "active": first})).map_err(|e| e.to_string())?;
+    let mut backend: Box<dyn Backend> =
+        match connect {
+            Some(addr) => {
+                if demo || library.is_some() || !imports.is_empty() {
+                    return Err("--demo, --library and --import apply to headless mode only".into());
                }
+                Box::new(Remote::connect(&addr).map_err(|e| {
+                    format!("LightCraft is not reachable at {addr} ({e}); start it with `lightcraft --control {}`", connect_port(&addr))
+                })?)
            }
-            Box::new(h)
-        }
-    };
+            None => {
+                let mut h = match &library {
+                    Some(dir) => {
+                        let mut h = Headless::default();
+                        h.session.open_library(dir, demo).map_err(|e| library_error(dir, e))?;
+                        library_warnings(&mut h.session, "lightcraft-cli");
+                        h
+                    }
+                    None if demo => Headless::demo(),
+                    None => Headless::default(),
+                };
+                if !imports.is_empty() {
+                    let paths = expand_paths(&imports);
+                    let r = h.session.execute("library.import", &json!({"paths": paths})).map_err(|e| e.to_string())?;
+                    // what follows acts on the imported photos (already-known files are reported as duplicates)
+                    let mut ids: Vec<Value> = r["imported"].as_array().cloned().unwrap_or_default();
+                    ids.extend(r["duplicates"].as_array().into_iter().flatten().filter_map(|d| d.get("existing").filter(|v| v.is_u64()).cloned()));
+                    if let Some(first) = ids.first().cloned() {
+                        h.session.execute("library.select", &json!({"ids": ids, "active": first})).map_err(|e| e.to_string())?;
+                    }
+                }
+                Box::new(h)
+            }
+        };
    let mut out = std::io::stdout().lock();
    let mut failed = 0;
    for s in &steps {
@@ -457,6 +606,12 @@ fn take_value<'a>(args: &'a [String], i: &mut usize, flag: &str) -> Result<&'a s
    args.get(*i).map(String::as_str).ok_or_else(|| format!("{flag} needs a value"))
 }

+/// The port an `--connect` address names, for the recovery hint. The whole address when it has
+/// no port, so a hint is never built from a guess.
+fn connect_port(addr: &str) -> &str {
+    addr.rsplit_once(':').map_or(addr, |(_, port)| port)
+}
+
 fn render(args: &[String]) -> Result<(), String> {
    let mut input = None;
    let mut output = None;
@@ -497,7 +652,7 @@ fn render(args: &[String]) -> Result<(), String> {
    }
    let input = input.ok_or("render: missing input file")?;
    let output = output.ok_or("render: missing -o OUTPUT")?;
-    let mut s = Session::new().with_fs();
+    let mut s = Session::new().with_fs().with_default_denoise_models();
    let abs = expand_paths(std::slice::from_ref(&input));
    let r = s.execute("library.import", &json!({"paths": abs})).map_err(|e| e.to_string())?;
    let id = r["imported"][0].as_u64().ok_or_else(|| format!("{input}: not a readable photo"))?;
@@ -520,18 +675,24 @@ fn render(args: &[String]) -> Result<(), String> {
    for (k, v) in opts {
        p[k] = v;
    }
-    let mut o = ExportOptions::from_json(&p);
+    let mut o = ExportOptions::from_params(&p).map_err(|e| e.to_string())?;
    if p.get("format").is_none() {
        let ext = Path::new(&output).extension().map(|e| e.to_string_lossy().to_string()).unwrap_or_default();
        o.format = ExportFormat::parse(&ext)
            .ok_or_else(|| format!("{output}: unknown extension (use .jpg .png .tif .webp .avif .dng or --opt format=…)"))?;
    }
    let e = export_photo(&mut s, lightcraft_engine::catalog::PhotoId(id), &o, 1)?;
-    std::fs::write(&output, &e.bytes).map_err(|err| format!("{output}: {err}"))?;
-    for (ext, bytes) in &e.sidecars {
-        let sc = Path::new(&output).with_extension(ext);
-        std::fs::write(&sc, bytes).map_err(|err| format!("{}: {err}", sc.display()))?;
-        eprintln!("lightcraft-cli: wrote {}", sc.display());
+    // never over the input (or its sidecar), however it is spelled: `render IMG.jpg -o IMG.jpg`
+    let sidecars: Vec<(String, &Vec<u8>)> =
+        e.sidecars.iter().map(|(ext, bytes)| (Path::new(&output).with_extension(ext).to_string_lossy().to_string(), bytes)).collect();
+    let guard = s.original_guard();
+    for p in std::iter::once(&output).chain(sidecars.iter().map(|(p, _)| p)) {
+        guard.check(Path::new(p)).map_err(|err| format!("render: {err}"))?;
+    }
+    lightcraft_engine::export::write_file(&output, &e.bytes)?;
+    for (sc, bytes) in &sidecars {
+        lightcraft_engine::export::write_file(sc, bytes)?;
+        eprintln!("lightcraft-cli: wrote {sc}");
    }
    eprintln!("lightcraft-cli: wrote {output} ({}×{})", e.width, e.height);
    Ok(())
@@ -567,19 +728,19 @@ fn snapshot(args: &[String]) -> Result<(), String> {
    if script.is_none() && output.is_none() {
        return Err("snapshot: give -o OUT.png and/or --script FILE".into());
    }
-    if !(scale > 0.0 && size[0] >= 1.0 && size[1] >= 1.0) {
-        return Err("snapshot: bad --size/--scale".into());
-    }
+    lightcraft_ui_egui::headless::viewport_pixels(size, scale).map_err(|e| format!("snapshot: bad --size/--scale: {e}"))?;
    let t0 = Instant::now();
    let mut session = match &library {
        Some(dir) => {
-            let mut s = Session::new().with_fs();
-            s.open_library(dir, false).map_err(|e| format!("{dir}: {e}"))?;
+            let mut s = Session::new().with_fs().with_default_denoise_models().with_default_face_models();
+            s.open_library(dir, false).map_err(|e| library_error(dir, e))?;
            s
        }
-        None if files.is_empty() => Session::with_demo().with_fs(),
-        None => Session::new().with_fs(),
+        None if files.is_empty() => Session::with_demo().with_fs().with_default_denoise_models().with_default_face_models(),
+        None => Session::new().with_fs().with_default_denoise_models().with_default_face_models(),
    };
+    // Apply compute policy before imports or the first scripted query can discover an adapter.
+    session.execute("app.gpu", &json!({"enabled": false})).map_err(|e| e.to_string())?;
    if !files.is_empty() {
        let ti = Instant::now();
        let r = session.execute("library.import", &json!({"paths": expand_paths(&files)})).map_err(|e| e.to_string())?;
@@ -587,24 +748,18 @@ fn snapshot(args: &[String]) -> Result<(), String> {
        eprintln!("lightcraft-cli snapshot: imported {n} files in {:.0} ms", ti.elapsed().as_secs_f64() * 1e3);
    }
    let services = lightcraft_ui_egui::Services {
-        write_shared: Some(std::sync::Arc::new(|p: &str, b: &[u8]| {
-            if let Some(dir) = Path::new(p).parent().filter(|d| !d.as_os_str().is_empty()) {
-                std::fs::create_dir_all(dir).map_err(|e| e.to_string())?;
-            }
-            std::fs::write(p, b).map_err(|e| format!("{p}: {e}"))
-        })),
-        write: Some(Box::new(|p: &str, b: &[u8]| {
-            if let Some(dir) = Path::new(p).parent().filter(|d| !d.as_os_str().is_empty()) {
-                std::fs::create_dir_all(dir).map_err(|e| e.to_string())?;
-            }
-            std::fs::write(p, b).map_err(|e| format!("{p}: {e}"))
-        })),
+        write_shared: Some(std::sync::Arc::new(lightcraft_engine::export::write_file)),
+        write: Some(Box::new(lightcraft_engine::export::write_file)),
        png: Some(Box::new(|img: &lightcraft_raster::Rgba8| {
            lightcraft_codecs::encode_png(&lightcraft_codecs::EncodeImage::rgba8(img), &lightcraft_codecs::EncodeMeta::default()).unwrap_or_default()
        })),
        ..Default::default()
    };
-    let app = lightcraft_ui_egui::LightcraftApp::new(session, services);
+    let mut app = lightcraft_ui_egui::LightcraftApp::new(session, services);
+    // Snapshot sessions promise no GPU: disable photo compute as well as the compositor
+    // before the first frame can start background adapter discovery.
+    // This UI state is session-local: the CLI neither loads nor saves desktop ui.json.
+    app.ui.settings.gpu = false;
    let mut h = Headless::new(app, size, scale);
    let timeout = Duration::from_secs(60);
    let mut shots = 0usize;
@@ -620,6 +775,9 @@ fn snapshot(args: &[String]) -> Result<(), String> {
        Some(p.with_file_name(format!("{stem}-{shots}.{ext}")).to_string_lossy().to_string())
    };
    let mut wrote_output = false;
+    // Issue #166: a scripted request that fails still gets its reply printed and the run carries
+    // on (like `run --keep-going`), but the exit status reports it — CI judges by exit status.
+    let mut failed = 0usize;
    if let Some(script) = &script {
        let text = std::fs::read_to_string(script).map_err(|e| format!("{script}: {e}"))?;
        let mut out = std::io::stdout().lock();
@@ -652,6 +810,9 @@ fn snapshot(args: &[String]) -> Result<(), String> {
                }
                _ => h.request(&method, params, timeout),
            };
+            if reply["ok"] != true {
+                failed += 1;
+            }
            if let Some(o) = reply.as_object_mut() {
                o.insert("id".into(), id);
                // wall time of the request (incl. the frames it ran), and since the start
@@ -688,7 +849,7 @@ fn snapshot(args: &[String]) -> Result<(), String> {
        }
    }
    eprintln!("lightcraft-cli snapshot: done in {:.2} s ({} frames)", t0.elapsed().as_secs_f64(), h.frames());
-    Ok(())
+    if failed > 0 { Err(format!("{failed} scripted request(s) failed")) } else { Ok(()) }
 }

 fn commands(args: &[String]) -> Result<(), String> {
@@ -19,6 +19,19 @@ fn gradient_png(path: &std::path::Path) {
    std::fs::write(path, png).unwrap();
 }

+#[test]
+fn calibrate_reads_raf_inputs_and_reports_bad_files() {
+    let input = tmp("invalid.RAF");
+    std::fs::write(&input, b"FUJIFILMCCD-RAW").unwrap();
+    let out = tmp("raf-profiles");
+    let result = Command::new(BIN).args(["calibrate", "--out"]).arg(&out).arg(&input).output().unwrap();
+    assert!(!result.status.success());
+    let stderr = String::from_utf8_lossy(&result.stderr);
+    assert!(stderr.contains("1 of 1 raw files"), "RAF was silently ignored: {stderr}");
+    assert!(stderr.contains("invalid.RAF") && stderr.contains("no profile written"), "{stderr}");
+    assert!(!out.join("X-T4.json").exists(), "bad input must not produce a profile");
+}
+
 fn mean(path: &std::path::Path) -> f64 {
    let d = lightcraft_codecs::decode(&std::fs::read(path).unwrap(), Default::default()).unwrap();
    let img = d.to_srgb8();
@@ -86,6 +99,24 @@ fn render_subcommand() {
    assert!(String::from_utf8_lossy(&o.stderr).contains("unknown control"));
 }

+/// Issue #93: `render IMG -o IMG` replaced its own input with the render.
+#[test]
+fn render_refuses_to_overwrite_its_input() {
+    let input = tmp("self-in.png");
+    gradient_png(&input);
+    let before = std::fs::read(&input).unwrap();
+    let dir = input.parent().unwrap();
+    std::fs::create_dir_all(dir.join("sub")).unwrap();
+    let other_spelling = dir.join("sub/../self-in.png");
+    for out in [&input, &other_spelling] {
+        let o =
+            Command::new(BIN).args(["render", input.to_str().unwrap(), "-o", out.to_str().unwrap(), "--set", "light.exposure=1"]).output().unwrap();
+        assert!(!o.status.success());
+        assert!(String::from_utf8_lossy(&o.stderr).contains("never writes over an original"), "{}", String::from_utf8_lossy(&o.stderr));
+    }
+    assert_eq!(std::fs::read(&input).unwrap(), before, "the input is untouched");
+}
+
 #[test]
 fn render_export_options() {
    let input = tmp("o-in.png");
@@ -136,10 +167,14 @@ fn snapshot_subcommand_renders_the_ui_headlessly() {
    std::fs::write(
        &script,
        format!(
-            "# comment\n{}\n{}\n{}\n{}\n",
+            "# comment\n{}\n{}\n{}\n{}\n{}\n{}\n",
            json!({"method": "ui.set", "params": {"view": "photoGrid"}}),
+            // (settled: under load a grid shot taken before its thumbnails is darker than the
+            // dimmed export dialog over them, and the comparison below fails)
+            json!({"method": "ui.settle", "params": {"timeoutMs": 20000}}),
            json!({"method": "ui.screenshot"}),
            json!({"method": "engine.execute", "params": {"command": "dialog.export"}}),
+            json!({"method": "ui.settle", "params": {"timeoutMs": 20000}}),
            json!({"method": "ui.screenshot", "params": {"path": b.to_str().unwrap()}}),
        ),
    )
@@ -150,7 +185,7 @@ fn snapshot_subcommand_renders_the_ui_headlessly() {
        .unwrap();
    assert!(o.status.success(), "{}", String::from_utf8_lossy(&o.stderr));
    let replies: Vec<Value> = String::from_utf8_lossy(&o.stdout).lines().map(|l| serde_json::from_str(l).unwrap()).collect();
-    assert_eq!(replies.len(), 4);
+    assert_eq!(replies.len(), 6);
    assert!(replies.iter().all(|r| r["ok"] == true), "{replies:?}");
    for (p, dimmed) in [(&a, false), (&b, true)] {
        let d = lightcraft_codecs::decode(&std::fs::read(p).unwrap(), Default::default()).unwrap();
@@ -165,6 +200,127 @@ fn snapshot_subcommand_renders_the_ui_headlessly() {
    assert_eq!((d.width, d.height), (640, 480));
 }

+/// Issue #166: a scripted request that fails makes `snapshot` exit non-zero. The failed reply is
+/// still printed, the later lines still run, and the final `-o` screenshot is still written.
+#[test]
+fn snapshot_script_failure_exits_non_zero() {
+    let script = tmp("snap-fail.jsonl");
+    let out = tmp("snap-fail.png");
+    std::fs::write(
+        &script,
+        format!(
+            "{}\n{}\n",
+            json!({"method": "ui.set", "params": {"view": "grid"}}), // not a view: `photoGrid`, `detail`, …
+            json!({"method": "ui.set", "params": {"view": "photoGrid"}}),
+        ),
+    )
+    .unwrap();
+    let _ = std::fs::remove_file(&out);
+    let o = Command::new(BIN)
+        .args(["snapshot", "--demo", "--script", script.to_str().unwrap(), "-o", out.to_str().unwrap(), "--size", "320x200"])
+        .output()
+        .unwrap();
+    let stderr = String::from_utf8_lossy(&o.stderr);
+    assert!(!o.status.success(), "a failed scripted request must fail the run\n{stderr}");
+    assert!(stderr.contains("1 scripted request(s) failed"), "{stderr}");
+    let replies: Vec<Value> = String::from_utf8_lossy(&o.stdout).lines().map(|l| serde_json::from_str(l).unwrap()).collect();
+    assert_eq!(replies.len(), 2, "{replies:?}");
+    assert_eq!(replies[0]["ok"], false, "{replies:?}");
+    assert!(replies[0]["error"].as_str().unwrap_or("").contains("grid"), "{replies:?}");
+    assert_eq!(replies[1]["ok"], true, "{replies:?}");
+    // continue-and-report: the final screenshot is still written
+    let d = lightcraft_codecs::decode(&std::fs::read(&out).unwrap(), Default::default()).unwrap();
+    assert_eq!((d.width, d.height), (320, 200));
+}
+
+#[test]
+fn snapshot_ui_zoom_keeps_requested_pixel_dimensions() {
+    let script = tmp("snap-ui-zoom.jsonl");
+    std::fs::write(
+        &script,
+        format!("{}\n{}\n", json!({"method": "ui.key", "params": {"key": "Plus", "cmd": true}}), json!({"method": "ui.inspect"})),
+    )
+    .unwrap();
+    for (size, scale, expected) in [("400x240", "1", (400, 240)), ("345x200", "0.9", (311, 180))] {
+        let out = tmp(&format!("snap-ui-zoom-{scale}.png"));
+        let o = Command::new(BIN)
+            .args(["snapshot", "--script", script.to_str().unwrap(), "-o", out.to_str().unwrap(), "--size", size, "--scale", scale])
+            .output()
+            .unwrap();
+        assert!(o.status.success(), "{}", String::from_utf8_lossy(&o.stderr));
+        let replies: Vec<Value> = String::from_utf8_lossy(&o.stdout).lines().map(|line| serde_json::from_str(line).unwrap()).collect();
+        assert!(replies.iter().all(|reply| reply["ok"] == true), "{replies:?}");
+        assert!(replies[1]["result"]["pixelsPerPoint"].as_f64().unwrap() > scale.parse::<f64>().unwrap());
+        let decoded = lightcraft_codecs::decode(&std::fs::read(out).unwrap(), Default::default()).unwrap();
+        assert_eq!((decoded.width, decoded.height), expected);
+    }
+}
+
+/// Headless snapshots default to CPU photo rendering as well as CPU UI rasterization.
+#[test]
+fn snapshot_defaults_to_cpu_without_gpu_environment_overrides() {
+    let script = tmp("default-cpu.jsonl");
+    std::fs::write(
+        &script,
+        format!("{}\n{}\n", json!({"method": "engine.execute", "params": {"command": "app.gpu"}}), json!({"method": "ui.screenshot"})),
+    )
+    .unwrap();
+    let image = tmp("default-cpu.png");
+    let output = Command::new(BIN)
+        .env_remove("LIGHTCRAFT_GPU")
+        .env_remove("LIGHTCRAFT_GPU_BACKEND")
+        .env_remove("WGPU_BACKEND")
+        .args(["snapshot", "--demo", "--script", script.to_str().unwrap(), "-o", image.to_str().unwrap(), "--size", "480x320"])
+        .output()
+        .unwrap();
+    assert!(output.status.success(), "status {}: {}", output.status, String::from_utf8_lossy(&output.stderr));
+    let replies: Vec<Value> = String::from_utf8_lossy(&output.stdout).lines().map(|line| serde_json::from_str(line).unwrap()).collect();
+    assert!(replies.iter().all(|reply| reply["ok"] == true), "{replies:?}");
+    let gpu = &replies[0]["result"];
+    assert_eq!(gpu["enabled"], false, "{gpu}");
+    assert_eq!(gpu["available"], false, "{gpu}");
+    assert_eq!(gpu["adapter"], Value::Null, "no device is created: {gpu}");
+    assert!(gpu["reason"].as_str().is_some_and(|reason| reason.contains("preference")), "{gpu}");
+    let decoded = lightcraft_codecs::decode(&std::fs::read(image).unwrap(), Default::default()).unwrap();
+    assert_eq!((decoded.width, decoded.height), (480, 320));
+}
+
+/// Issue #136: with the GPU switched off from the environment the UI starts and renders on the CPU,
+/// without creating a GPU device (no driver is loaded), and says why.
+#[test]
+fn snapshot_starts_without_a_gpu() {
+    let script = tmp("nogpu.jsonl");
+    std::fs::write(
+        &script,
+        format!(
+            "{}\n{}\n{}\n",
+            json!({"method": "ui.set", "params": {"view": "photoGrid"}}),
+            json!({"method": "engine.execute", "params": {"command": "app.gpu"}}),
+            json!({"method": "ui.screenshot"}),
+        ),
+    )
+    .unwrap();
+    for (var, value, reason) in [("LIGHTCRAFT_GPU", "0", "LIGHTCRAFT_GPU=0"), ("LIGHTCRAFT_GPU_BACKEND", "off", "LIGHTCRAFT_GPU_BACKEND=off")] {
+        let out = tmp(&format!("nogpu-{var}.png"));
+        let o = Command::new(BIN)
+            .env_remove("LIGHTCRAFT_GPU")
+            .env_remove("LIGHTCRAFT_GPU_BACKEND")
+            .env(var, value)
+            .args(["snapshot", "--demo", "--script", script.to_str().unwrap(), "-o", out.to_str().unwrap(), "--size", "480x320"])
+            .output()
+            .unwrap();
+        assert!(o.status.success(), "{var}: {}", String::from_utf8_lossy(&o.stderr));
+        let replies: Vec<Value> = String::from_utf8_lossy(&o.stdout).lines().map(|l| serde_json::from_str(l).unwrap()).collect();
+        assert!(replies.iter().all(|r| r["ok"] == true), "{replies:?}");
+        let gpu = &replies[1]["result"];
+        assert_eq!(gpu["available"], false, "{var}: {gpu}");
+        assert_eq!(gpu["adapter"], Value::Null, "{var}: no device was created: {gpu}");
+        assert!(gpu["reason"].as_str().is_some_and(|r| r.contains(reason)), "{var}: {gpu}");
+        let d = lightcraft_codecs::decode(&std::fs::read(&out).unwrap(), Default::default()).unwrap();
+        assert_eq!((d.width, d.height), (480, 320));
+    }
+}
+
 fn run_cli(args: &[&str], stdin: Option<&str>) -> (bool, Vec<Value>, String) {
    let mut child = Command::new(BIN).arg("run").args(args).stdin(Stdio::piped()).stdout(Stdio::piped()).stderr(Stdio::piped()).spawn().unwrap();
    if let Some(text) = stdin {
@@ -238,7 +394,149 @@ fn devices_are_listed_and_imported_from() {
    assert_eq!(dev["name"], "CARD", "{line}");
    // the import review of its DCIM folder finds the photo
    let dcim = dev["path"].as_str().unwrap().to_string();
-    let o = Command::new(BIN).args(["run", "library.importPreview", &format!("paths=[\"{dcim}\"]")]).output().unwrap();
+    // JSON-escaped: a Windows path has backslashes
+    let paths = serde_json::to_string(&[&dcim]).unwrap();
+    let o = Command::new(BIN).args(["run", "library.importPreview", &format!("paths={paths}")]).output().unwrap();
    let line: Value = serde_json::from_slice(o.stdout.split(|b| *b == b'\n').next().unwrap()).unwrap();
    assert_eq!(line["result"]["candidates"].as_array().map(Vec::len), Some(1), "{line}");
 }
+
+/// Issue #99: a library open in one process (here `mcp --library`) is refused by a second one,
+/// with who has it and how to drive the running app instead; free again once the first exits.
+#[test]
+fn a_library_open_in_another_process_is_refused() {
+    let lib = tmp("locked-lib");
+    let _ = std::fs::remove_dir_all(&lib);
+    let lib_s = lib.to_str().unwrap();
+    let mut holder =
+        Command::new(BIN).args(["mcp", "--library", lib_s]).stdin(Stdio::piped()).stdout(Stdio::null()).stderr(Stdio::piped()).spawn().unwrap();
+    let mut err = BufReader::new(holder.stderr.take().unwrap());
+    let mut line = String::new();
+    while !line.contains("opened library") {
+        line.clear();
+        assert!(err.read_line(&mut line).unwrap() > 0, "mcp exited before opening the library");
+    }
+    let (ok, _, stderr) = run_cli(&["--library", lib_s, "library.info"], None);
+    assert!(!ok);
+    assert!(stderr.contains("already open in lightcraft-cli") && stderr.contains(&format!("process {}", holder.id())), "{stderr}");
+    assert!(stderr.contains("mcp --connect"), "{stderr}");
+
+    drop(holder.stdin.take()); // EOF: the server exits and lets go of the library
+    assert!(holder.wait().unwrap().success());
+    let (ok, lines, stderr) = run_cli(&["--library", lib_s, "library.info"], None);
+    assert!(ok, "{stderr}");
+    assert_eq!(lines[0]["ok"], true);
+    let _ = std::fs::remove_dir_all(&lib);
+}
+
+/// Issue #167: when `--connect ADDR` cannot reach the app, the recovery hint must name the port
+/// ADDR actually names — following the old fixed "7980" left the retry connecting elsewhere.
+#[test]
+fn connect_failure_hint_names_the_attempted_port() {
+    let (ok, _lines, err) = run_cli(&["--connect", "127.0.0.1:18437", "ui.inspect"], None);
+    assert!(!ok);
+    assert!(err.contains("--control 18437"), "the hint must name the port it tried: {err}");
+    assert!(!err.contains("--control 7980"), "{err}");
+}
+
+/// The CLI logs warnings on stderr (#168): with `RUST_LOG=warn` an unknown GPU backend name is
+/// reported, and `RUST_LOG=off` silences it. Before, no logger was installed and nothing appeared.
+#[test]
+fn warnings_are_logged_on_stderr() {
+    let input = tmp("log-in.png");
+    gradient_png(&input);
+    let out = tmp("log-out.jpg");
+    let run = |level: &str| {
+        let o = Command::new(BIN)
+            .args(["render", input.to_str().unwrap(), "-o", out.to_str().unwrap()])
+            .env("LIGHTCRAFT_GPU_BACKEND", "bogus")
+            .env("RUST_LOG", level)
+            .env_remove("LIGHTCRAFT_LOG")
+            .env_remove("LIGHTCRAFT_GPU")
+            .output()
+            .unwrap();
+        let stderr = String::from_utf8_lossy(&o.stderr).into_owned();
+        assert!(o.status.success(), "{stderr}");
+        stderr
+    };
+    let warn = run("warn");
+    assert!(warn.contains("LIGHTCRAFT_GPU_BACKEND=bogus names no known backend"), "{warn}");
+    let off = run("off");
+    assert!(!off.contains("names no known backend"), "{off}");
+}
+
+#[test]
+fn snapshot_rejects_invalid_dimensions_before_creating_an_image() {
+    for (index, (size, scale)) in
+        [("NaNx100", "1"), ("100x100", "inf"), ("9000x9000", "1"), ("9000x100", "2"), ("100x100", "0.001")].into_iter().enumerate()
+    {
+        let image = tmp(&format!("invalid-viewport-{index}.png"));
+        let output = Command::new(BIN).args(["snapshot", "--size", size, "--scale", scale, "-o"]).arg(&image).output().unwrap();
+        assert!(!output.status.success());
+        assert!(String::from_utf8_lossy(&output.stderr).contains("snapshot: bad --size/--scale"), "{}", String::from_utf8_lossy(&output.stderr));
+        assert!(!image.exists());
+    }
+}
+
+#[test]
+fn snapshot_preserves_fractional_scale_rounding() {
+    let image = tmp("fractional-scale-rounding.png");
+    let output =
+        Command::new(BIN).env("LIGHTCRAFT_GPU", "0").args(["snapshot", "--size", "345x200", "--scale", "0.9", "-o"]).arg(&image).output().unwrap();
+    assert!(output.status.success(), "{}", String::from_utf8_lossy(&output.stderr));
+    let decoded = lightcraft_codecs::decode(&std::fs::read(image).unwrap(), Default::default()).unwrap();
+    assert_eq!((decoded.width, decoded.height), (311, 180));
+}
+
+/// Snapshot's CPU setting must never replace the desktop preference, including with --library.
+#[cfg(target_os = "linux")]
+#[test]
+fn snapshot_leaves_desktop_gpu_preferences_untouched() {
+    let dir = tmp("snapshot-desktop-preferences");
+    let library = dir.join("library");
+    let mut session = lightcraft_engine::Session::new().with_fs();
+    session.open_library(&library, true).unwrap();
+    drop(session);
+    let config = dir.join("config");
+    let app_config = config.join("lightcraft");
+    std::fs::create_dir_all(&app_config).unwrap();
+    let ui_path = app_config.join("ui.json");
+    let mut ui = lightcraft_ui_egui::UiState::default();
+    ui.settings.gpu = true;
+    let saved = serde_json::to_vec_pretty(&ui).unwrap();
+    std::fs::write(&ui_path, &saved).unwrap();
+    let script = dir.join("preferences.jsonl");
+    std::fs::write(
+        &script,
+        format!("{}\n{}\n", json!({"method": "engine.execute", "params": {"command": "app.gpu"}}), json!({"method": "ui.inspect"})),
+    )
+    .unwrap();
+    for gpu_environment in [None, Some("1")] {
+        let image = dir.join(format!("{}.png", gpu_environment.unwrap_or("default")));
+        let mut cmd = Command::new(BIN);
+        cmd.env("XDG_CONFIG_HOME", &config).env_remove("LIGHTCRAFT_GPU").env_remove("LIGHTCRAFT_GPU_BACKEND").env_remove("WGPU_BACKEND");
+        if let Some(value) = gpu_environment {
+            cmd.env("LIGHTCRAFT_GPU", value);
+        }
+        let output = cmd
+            .args(["snapshot", "--library"])
+            .arg(&library)
+            .args(["--script"])
+            .arg(&script)
+            .args(["-o"])
+            .arg(&image)
+            .args(["--size", "480x320"])
+            .output()
+            .unwrap();
+        assert!(output.status.success(), "{}", String::from_utf8_lossy(&output.stderr));
+        let replies: Vec<Value> = String::from_utf8_lossy(&output.stdout).lines().map(|line| serde_json::from_str(line).unwrap()).collect();
+        assert!(replies.iter().all(|reply| reply["ok"] == true), "{replies:?}");
+        assert_eq!(replies[0]["result"]["enabled"], false);
+        assert_eq!(replies[0]["result"]["available"], false);
+        assert_eq!(replies[0]["result"]["adapter"], Value::Null);
+        assert_eq!(replies[1]["result"]["ui"]["settings"]["gpu"], false);
+        assert_eq!(std::fs::read(&ui_path).unwrap(), saved, "snapshot changed the desktop preferences");
+        assert!(!app_config.join("ui.json.tmp").exists());
+        assert!(image.exists());
+    }
+}
@@ -21,6 +21,7 @@ lightcraft-preview = { workspace = true }
 serde = { workspace = true }
 serde_json = { workspace = true }
 log = { workspace = true }
+crc32fast = { workspace = true }

 [target.'cfg(target_arch = "wasm32")'.dependencies]
 eframe = { workspace = true, features = ["default_fonts", "glow"] }
@@ -70,3 +71,9 @@ web-sys = { workspace = true, features = [
    "WorkerType",
    "WritableStream",
 ] }
+
+[features]
+# Open HEIC/HEIF photos (the optional lightcraft-heif crate). Not default so distributors choose
+# it (HEVC patents); build with `--features heif` to include it. On wasm32 panics abort, so
+# heic-rs's `catch_unwind` guard does not apply there.
+heif = ["lightcraft-codecs/heif"]
@@ -24,13 +24,21 @@
  <script type="module">
    import init, { start, command } from "./lightcraft_web.js";
    (async () => {
+      // The Chinese font is a separate file: embedding it would exceed the host's WASM size limit.
+      // Font-free development builds can still run, with the same missing-glyph warning as before.
+      const chineseFont = fetch("./lightcraft_zh_hans.otf")
+        .then(async response => response.ok ? new Uint8Array(await response.arrayBuffer()) : new Uint8Array())
+        .catch(() => new Uint8Array());
      // Compile once; the render workers (worker.js) instantiate the same compiled module.
-      const module = await WebAssembly.compileStreaming(fetch("./lightcraft_web_bg.wasm"));
+      const [module, font] = await Promise.all([
+        WebAssembly.compileStreaming(fetch("./lightcraft_web_bg.wasm")),
+        chineseFont,
+      ]);
      globalThis.lightcraftModule = module;
      await init({ module_or_path: module });
      // automation: `await lightcraft.command("library.info", "{}")` → JSON text
      globalThis.lightcraft = { command };
-      start();
+      start(font);
    })().catch((e) => {
      console.error(e);
      const el = document.getElementById("lightcraft_loading");
@@ -232,6 +232,18 @@ impl Backend {
        }
    }

+    /// Is there a file at `path`?
+    pub async fn exists(&self, path: &str) -> Result<bool, String> {
+        match self {
+            Backend::Opfs(root) => Ok(Self::opfs_file(root, path, false).await?.is_some()),
+            Backend::Idb(db) => {
+                let (_tx, store) = Self::idb_store(db, false)?;
+                let n = idb_req(&store.count_with_key(&path.into()).map_err(err)?).await?;
+                Ok(n.as_f64().is_some_and(|n| n > 0.0))
+            }
+        }
+    }
+
    pub async fn remove(&self, path: &str) -> Result<(), String> {
        match self {
            Backend::Opfs(root) => {
@@ -309,14 +321,25 @@ async fn dir_keys(dir: &FileSystemDirectoryHandle) -> Result<Vec<String>, String
    Ok(out)
 }

-/// Ask the browser not to evict our storage under pressure (best effort; fire and forget).
-pub fn request_persistence() {
-    if let Some(sm) = storage_manager()
-        && let Ok(p) = sm.persist()
-    {
-        wasm_bindgen_futures::spawn_local(async move {
-            let granted = JsFuture::from(p).await.ok().is_some_and(|v| v.is_truthy());
-            log::info!("lightcraft: persistent storage {}", if granted { "granted" } else { "not granted (best effort)" });
-        });
-    }
+/// Ask the browser not to evict our storage under pressure; `on_result(granted)` once known
+/// (`false` also when the browser has no storage manager).
+pub fn request_persistence(on_result: impl FnOnce(bool) + 'static) {
+    let Some(sm) = storage_manager() else {
+        on_result(false);
+        return;
+    };
+    wasm_bindgen_futures::spawn_local(async move {
+        // already persistent: nothing to ask
+        let already = match sm.persisted() {
+            Ok(p) => JsFuture::from(p).await.ok().is_some_and(|v| v.is_truthy()),
+            Err(_) => false,
+        };
+        let granted = already
+            || match sm.persist() {
+                Ok(p) => JsFuture::from(p).await.ok().is_some_and(|v| v.is_truthy()),
+                Err(_) => false,
+            };
+        log::info!("lightcraft: persistent storage {}", if granted { "granted" } else { "not granted" });
+        on_result(granted);
+    });
 }
@@ -0,0 +1,338 @@
+//! Library backups for the browser build: a plain zip (entries stored, not compressed — photos
+//! don't shrink) of the library files and every stored original, built one entry at a time so
+//! only one file is in wasm memory at once, and read back the same way for a restore.
+//!
+//! Layout of a backup:
+//! - `library/catalog.snap`, `library/catalog.log`, `library/presets.json`, … (see
+//!   [`crate::files::LIBRARY_FILES`]);
+//! - `originals/<content hash>/<file name>`: each imported photo's original bytes, under its own
+//!   name so an unzipped backup is browsable;
+//! - `README.txt`.
+//!
+//! Restoring never deletes anything: the library files go to a new library folder in browser
+//! storage, originals that aren't stored yet are added (they're named by content, so existing
+//! ones are identical; a restored one must match its content hash, see [`verify_original`]),
+//! and only then does a small pointer file ([`ACTIVE_LIBRARY`]) switch to the restored library.
+//! The library that was there before stays in storage.
+//!
+//! No zip64: a backup holds at most [`MAX_BYTES`] and 65 535 entries.
+
+/// Storage key of the file naming the library folder in use (absent: `library`).
+pub const ACTIVE_LIBRARY: &str = "active-library";
+
+/// The default library folder.
+pub const DEFAULT_LIBRARY_DIR: &str = "library";
+
+/// Largest backup (offsets are 32-bit without zip64).
+pub const MAX_BYTES: u64 = 0xFFFF_FFFF - (64 << 20);
+
+const LOCAL_SIG: u32 = 0x0403_4b50;
+const CENTRAL_SIG: u32 = 0x0201_4b50;
+const END_SIG: u32 = 0x0605_4b50;
+/// General-purpose flag bit 11: names are UTF-8.
+const UTF8: u16 = 0x0800;
+/// 1980-01-01 00:00 in MS-DOS format (date, time).
+const DOS_DATE: u16 = (1 << 5) | 1;
+
+/// The readme inside every backup.
+pub const README: &str = "LightCraft library backup (browser version)\n\n\
+library/      the catalog (catalog.snap + catalog.log), presets, view state and preferences\n\
+originals/    every imported photo, as originals/<content hash>/<file name>\n\n\
+Restore it in LightCraft for the web with File > Restore Library from Backup...\n\
+The photos in originals/ are the untouched files you imported; edits live in the catalog.\n";
+
+/// Is `name` a safe library folder name (`library` or `library-<letters, digits, ->`)?
+pub fn valid_library_dir(name: &str) -> bool {
+    name == DEFAULT_LIBRARY_DIR
+        || name.strip_prefix("library-").is_some_and(|s| !s.is_empty() && s.len() <= 64 && s.chars().all(|c| c.is_ascii_alphanumeric() || c == '-'))
+}
+
+/// Is `h` a plausible content hash (used as a storage key segment)?
+pub fn valid_hash(h: &str) -> bool {
+    !h.is_empty() && h.len() <= 128 && h.chars().all(|c| c.is_ascii_alphanumeric())
+}
+
+/// Zip entry name for an original.
+pub fn original_entry(hash: &str, file_name: &str) -> String {
+    let name: String = file_name.chars().map(|c| if c == '/' || c == '\\' || c.is_control() { '_' } else { c }).collect();
+    let name = if name.is_empty() || name == "." || name == ".." { "original".to_string() } else { name };
+    format!("originals/{hash}/{name}")
+}
+
+/// Where a backup entry goes when restoring into library folder `dir`: `Some(storage key)`, or
+/// `None` to skip it (README, thumbnails, anything unknown or unsafe).
+pub fn restore_key(entry: &str, dir: &str) -> Option<String> {
+    if let Some(name) = entry.strip_prefix("library/") {
+        return crate::files::LIBRARY_FILES.contains(&name).then(|| format!("{dir}/{name}"));
+    }
+    let rest = entry.strip_prefix("originals/")?;
+    let (hash, _) = rest.split_once('/').unwrap_or((rest, ""));
+    valid_hash(hash).then(|| format!("originals/{hash}"))
+}
+
+fn u16le(v: &mut Vec<u8>, x: u16) {
+    v.extend_from_slice(&x.to_le_bytes());
+}
+fn u32le(v: &mut Vec<u8>, x: u32) {
+    v.extend_from_slice(&x.to_le_bytes());
+}
+
+struct Central {
+    name: String,
+    crc: u32,
+    size: u32,
+    offset: u32,
+}
+
+/// Writes a stored (uncompressed) zip piece by piece: [`ZipWriter::entry`] returns the local
+/// header to emit before the entry's bytes, [`ZipWriter::finish`] the central directory.
+#[derive(Default)]
+pub struct ZipWriter {
+    entries: Vec<Central>,
+    offset: u64,
+}
+
+impl ZipWriter {
+    /// The local header for `name` holding `data` (emit it, then `data`).
+    pub fn entry(&mut self, name: &str, data: &[u8]) -> Result<Vec<u8>, String> {
+        let size = u32::try_from(data.len()).map_err(|_| format!("{name}: larger than 4 GB"))?;
+        let name_len = u16::try_from(name.len()).map_err(|_| format!("{name}: name too long"))?;
+        if self.entries.len() >= usize::from(u16::MAX) {
+            return Err("too many files for one backup".into());
+        }
+        let header_len = 30 + u64::from(name_len);
+        if self.offset + header_len + u64::from(size) > MAX_BYTES {
+            return Err("the library is too large for one backup file (4 GB)".into());
+        }
+        let offset = u32::try_from(self.offset).map_err(|_| "backup too large".to_string())?;
+        let crc = crc32fast::hash(data);
+        let mut h = Vec::with_capacity(header_len as usize);
+        u32le(&mut h, LOCAL_SIG);
+        u16le(&mut h, 20); // version needed
+        u16le(&mut h, UTF8);
+        u16le(&mut h, 0); // stored
+        u16le(&mut h, 0); // time
+        u16le(&mut h, DOS_DATE);
+        u32le(&mut h, crc);
+        u32le(&mut h, size);
+        u32le(&mut h, size);
+        u16le(&mut h, name_len);
+        u16le(&mut h, 0); // extra
+        h.extend_from_slice(name.as_bytes());
+        self.offset += header_len + u64::from(size);
+        self.entries.push(Central { name: name.to_string(), crc, size, offset });
+        Ok(h)
+    }
+
+    /// The central directory and end record (emit last).
+    pub fn finish(self) -> Result<Vec<u8>, String> {
+        let mut cd = Vec::new();
+        for e in &self.entries {
+            u32le(&mut cd, CENTRAL_SIG);
+            u16le(&mut cd, 20); // made by
+            u16le(&mut cd, 20); // needed
+            u16le(&mut cd, UTF8);
+            u16le(&mut cd, 0);
+            u16le(&mut cd, 0);
+            u16le(&mut cd, DOS_DATE);
+            u32le(&mut cd, e.crc);
+            u32le(&mut cd, e.size);
+            u32le(&mut cd, e.size);
+            u16le(&mut cd, e.name.len() as u16);
+            u16le(&mut cd, 0); // extra
+            u16le(&mut cd, 0); // comment
+            u16le(&mut cd, 0); // disk
+            u16le(&mut cd, 0); // internal attrs
+            u32le(&mut cd, 0); // external attrs
+            u32le(&mut cd, e.offset);
+            cd.extend_from_slice(e.name.as_bytes());
+        }
+        let cd_offset = u32::try_from(self.offset).map_err(|_| "backup too large".to_string())?;
+        let cd_size = u32::try_from(cd.len()).map_err(|_| "backup too large".to_string())?;
+        let n = self.entries.len() as u16;
+        let mut out = cd;
+        u32le(&mut out, END_SIG);
+        u16le(&mut out, 0);
+        u16le(&mut out, 0);
+        u16le(&mut out, n);
+        u16le(&mut out, n);
+        u32le(&mut out, cd_size);
+        u32le(&mut out, cd_offset);
+        u16le(&mut out, 0); // comment
+        Ok(out)
+    }
+}
+
+/// One entry of a zip's central directory.
+#[derive(Clone, Debug, PartialEq, Eq)]
+pub struct ZipEntry {
+    pub name: String,
+    pub crc: u32,
+    pub size: u64,
+    /// Offset of the entry's local header.
+    pub header: u64,
+    /// 0 = stored (the only method a restore reads).
+    pub method: u16,
+}
+
+fn rd16(b: &[u8], at: usize) -> Option<u16> {
+    Some(u16::from_le_bytes(b.get(at..at + 2)?.try_into().ok()?))
+}
+fn rd32(b: &[u8], at: usize) -> Option<u32> {
+    Some(u32::from_le_bytes(b.get(at..at + 4)?.try_into().ok()?))
+}
+
+/// How many bytes from the end of a zip to read to find its end record.
+pub fn tail_len(file_len: u64) -> u64 {
+    file_len.min(22 + 65_535)
+}
+
+/// The central directory's (offset, size) from the last [`tail_len`] bytes of a zip.
+pub fn find_central(tail: &[u8]) -> Result<(u64, u64), String> {
+    let at = (0..tail.len().saturating_sub(21)).rev().find(|&i| rd32(tail, i) == Some(END_SIG)).ok_or("not a zip file (no end record)")?;
+    let size = rd32(tail, at + 12).ok_or("truncated zip")?;
+    let offset = rd32(tail, at + 16).ok_or("truncated zip")?;
+    if size == u32::MAX || offset == u32::MAX {
+        return Err("zip64 backups are not supported".into());
+    }
+    Ok((u64::from(offset), u64::from(size)))
+}
+
+/// Parse a central directory.
+pub fn parse_central(cd: &[u8]) -> Result<Vec<ZipEntry>, String> {
+    let mut out = Vec::new();
+    let mut i = 0;
+    while i + 46 <= cd.len() {
+        if rd32(cd, i) != Some(CENTRAL_SIG) {
+            return Err("damaged zip directory".into());
+        }
+        let bad = || "damaged zip directory".to_string();
+        let method = rd16(cd, i + 10).ok_or_else(bad)?;
+        let crc = rd32(cd, i + 16).ok_or_else(bad)?;
+        let size = rd32(cd, i + 24).ok_or_else(bad)?;
+        let n = usize::from(rd16(cd, i + 28).ok_or_else(bad)?);
+        let extra = usize::from(rd16(cd, i + 30).ok_or_else(bad)?);
+        let comment = usize::from(rd16(cd, i + 32).ok_or_else(bad)?);
+        let header = rd32(cd, i + 42).ok_or_else(bad)?;
+        let name = cd.get(i + 46..i + 46 + n).ok_or_else(bad)?;
+        out.push(ZipEntry { name: String::from_utf8_lossy(name).into_owned(), crc, size: u64::from(size), header: u64::from(header), method });
+        i += 46 + n + extra + comment;
+    }
+    Ok(out)
+}
+
+/// Offset of an entry's data from its local header's first 30 bytes.
+pub fn data_offset(entry: &ZipEntry, local: &[u8]) -> Result<u64, String> {
+    if rd32(local, 0) != Some(LOCAL_SIG) {
+        return Err(format!("{}: damaged entry", entry.name));
+    }
+    let n = rd16(local, 26).ok_or("truncated entry")?;
+    let extra = rd16(local, 28).ok_or("truncated entry")?;
+    Ok(entry.header + 30 + u64::from(n) + u64::from(extra))
+}
+
+/// Check an entry's bytes.
+pub fn verify(entry: &ZipEntry, data: &[u8]) -> Result<(), String> {
+    if entry.method != 0 {
+        return Err(format!("{}: compressed entries are not supported (not a LightCraft backup?)", entry.name));
+    }
+    if data.len() as u64 != entry.size || crc32fast::hash(data) != entry.crc {
+        return Err(format!("{}: damaged (checksum mismatch)", entry.name));
+    }
+    Ok(())
+}
+
+/// Check that an original's bytes are the photo its storage key names: originals are stored
+/// under their content hash, and a restore that kept bytes under another photo's hash would
+/// give that photo's catalog record a different image (the zip checksum can't tell).
+pub fn verify_original(entry: &ZipEntry, hash: &str, data: &[u8]) -> Result<(), String> {
+    if crate::store::content_hash(data) != hash {
+        return Err(format!("{}: the file doesn't match its content hash (damaged or altered backup)", entry.name));
+    }
+    Ok(())
+}
+
+/// A name for the restored library's folder.
+pub fn restored_dir_name(now_ms: f64) -> String {
+    format!("library-restored-{}", now_ms.max(0.0) as u64)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    fn build(files: &[(&str, &[u8])]) -> Vec<u8> {
+        let mut w = ZipWriter::default();
+        let mut out = Vec::new();
+        for (n, d) in files {
+            out.extend(w.entry(n, d).unwrap());
+            out.extend_from_slice(d);
+        }
+        out.extend(w.finish().unwrap());
+        out
+    }
+
+    #[test]
+    fn written_backups_read_back() {
+        let files: [(&str, &[u8]); 3] =
+            [("README.txt", README.as_bytes()), ("library/catalog.log", b"{\"seq\":1}\n"), ("originals/abc123/IMG 1.jpg", &[0, 1, 2, 3, 255])];
+        let zip = build(&files);
+        let tail = &zip[zip.len() - tail_len(zip.len() as u64) as usize..];
+        let (off, size) = find_central(tail).unwrap();
+        let entries = parse_central(&zip[off as usize..(off + size) as usize]).unwrap();
+        assert_eq!(entries.len(), 3);
+        for (e, (name, data)) in entries.iter().zip(files) {
+            assert_eq!(e.name, name);
+            let start = data_offset(e, &zip[e.header as usize..e.header as usize + 30]).unwrap() as usize;
+            let got = &zip[start..start + e.size as usize];
+            assert_eq!(got, data);
+            verify(e, got).unwrap();
+        }
+        // damage is caught
+        let e = &entries[2];
+        assert!(verify(e, &[0, 1, 2, 3, 254]).is_err());
+        assert!(find_central(b"not a zip at all, just some text").is_err());
+    }
+
+    #[test]
+    fn originals_must_match_their_content_hash() {
+        let red: &[u8] = b"red corner photo";
+        let blue: &[u8] = b"blue corner photo";
+        let hash = crate::store::content_hash(red);
+        let name = original_entry(&hash, "IMG 1.png");
+        let key = restore_key(&name, "d").unwrap();
+        let key_hash = key.strip_prefix("originals/").unwrap();
+        assert_eq!(key_hash, hash);
+        // the entry's own bytes pass; other bytes under the same name (valid zip CRC) don't
+        let zip = build(&[(name.as_str(), blue)]);
+        let tail = &zip[zip.len() - tail_len(zip.len() as u64) as usize..];
+        let (off, size) = find_central(tail).unwrap();
+        let entries = parse_central(&zip[off as usize..(off + size) as usize]).unwrap();
+        let entry = &entries[0];
+        verify(entry, blue).unwrap();
+        let err = verify_original(entry, key_hash, blue).unwrap_err();
+        assert!(err.contains(&name), "{err}");
+        verify_original(entry, key_hash, red).unwrap();
+    }
+
+    #[test]
+    fn restore_keys_are_safe() {
+        assert_eq!(restore_key("library/catalog.snap", "library-restored-1").as_deref(), Some("library-restored-1/catalog.snap"));
+        assert_eq!(restore_key("library/../../x", "library-r"), None);
+        assert_eq!(restore_key("library/evil.bin", "library-r"), None);
+        assert_eq!(restore_key("originals/abc123/IMG.jpg", "d").as_deref(), Some("originals/abc123"));
+        assert_eq!(restore_key("originals/../library/catalog.log", "d"), None);
+        assert_eq!(restore_key("README.txt", "d"), None);
+        assert_eq!(restore_key("thumbs/x.jpg", "d"), None);
+        assert_eq!(original_entry("abc", "a/b\\c.jpg"), "originals/abc/a_b_c.jpg");
+        assert_eq!(original_entry("abc", ".."), "originals/abc/original");
+        assert!(valid_library_dir("library") && valid_library_dir("library-restored-17"));
+        assert!(!valid_library_dir("library-../x") && !valid_library_dir("originals") && !valid_library_dir("library-"));
+    }
+
+    #[test]
+    fn size_limits_are_errors_not_panics() {
+        let mut w = ZipWriter { offset: MAX_BYTES - 10, ..Default::default() };
+        assert!(w.entry("x", &[0; 64]).is_err());
+    }
+}
@@ -49,6 +49,8 @@ struct Mirror {
    files: BTreeMap<String, Vec<u8>>,
    /// Dirty files, least recently modified first.
    dirty: Vec<(String, Dirty)>,
+    /// Why the last flush to browser storage failed (quota, eviction…), until one succeeds.
+    error: Option<String>,
 }

 impl Mirror {
@@ -138,6 +140,18 @@ impl Files {
        }
    }

+    /// Record the outcome of a flush to browser storage: `Some(why)` when it failed, `None` once
+    /// one succeeds. While a flush is failing, the catalog [`Store`] refuses new writes, so the
+    /// engine keeps the changes queued and reports them unsaved (`library.info` → `unsavedOps`,
+    /// the top bar's warning) instead of pretending they were saved.
+    pub fn set_error(&self, e: Option<String>) {
+        self.lock().error = e;
+    }
+
+    pub fn error(&self) -> Option<String> {
+        self.lock().error.clone()
+    }
+
    /// A catalog [`Store`] over these files.
    pub fn store(&self) -> FilesStore {
        FilesStore(self.clone())
@@ -147,19 +161,32 @@ impl Files {
 /// [`Store`] implementation over [`Files`].
 pub struct FilesStore(Files);

+impl FilesStore {
+    /// Writes fail while browser storage does (see [`Files::set_error`]).
+    fn writable(&self) -> io::Result<()> {
+        match self.0.error() {
+            Some(e) => Err(io::Error::other(format!("browser storage: {e}"))),
+            None => Ok(()),
+        }
+    }
+}
+
 impl Store for FilesStore {
    fn read(&mut self, name: &str) -> io::Result<Option<Vec<u8>>> {
        Ok(self.0.get(name))
    }
    fn write_atomic(&mut self, name: &str, data: &[u8]) -> io::Result<()> {
+        self.writable()?;
        self.0.write(name, data);
        Ok(())
    }
    fn append(&mut self, name: &str, data: &[u8]) -> io::Result<()> {
+        self.writable()?;
        self.0.append(name, data);
        Ok(())
    }
    fn truncate(&mut self, name: &str, len: u64) -> io::Result<()> {
+        self.writable()?;
        self.0.truncate(name, len as usize);
        Ok(())
    }
@@ -224,6 +251,37 @@ mod tests {
        assert_eq!(names, ["catalog.snap", "catalog.log"]);
    }

+    /// A failing flush (quota exceeded, storage evicted) is not silent: commands report the change
+    /// as unsaved and the engine keeps it queued; once storage works again it is written.
+    #[test]
+    fn flush_failures_show_as_unsaved_and_are_retried() {
+        let mut disk = BTreeMap::new();
+        let files = Files::default();
+        let mut s = Session::new();
+        let stores = LibraryStores { dir: "browser".into(), catalog: Box::new(files.store()), files: Box::new(files.store()), on_disk: false };
+        s.open_library_in(stores, true).unwrap();
+        flush(&files, &mut disk);
+        files.set_error(Some("QuotaExceededError".into()));
+        let r = s.execute("photo.rate", &json!({"rating": 5}));
+        assert!(matches!(r, Err(lightcraft_engine::EngineError::NotSaved(_))), "{r:?}");
+        let (ops, why) = s.unsaved().unwrap();
+        assert!(ops > 0 && why.contains("QuotaExceeded"), "{why}");
+        files.set_error(None);
+        s.persist_if_dirty();
+        s.persist().unwrap();
+        assert!(s.unsaved().is_none(), "written once storage works again");
+        flush(&files, &mut disk);
+        let id = s.selection.active.unwrap();
+        let files2 = Files::default();
+        for (k, v) in &disk {
+            files2.preload(k, v.clone());
+        }
+        let mut s2 = Session::new();
+        let stores = LibraryStores { dir: "browser".into(), catalog: Box::new(files2.store()), files: Box::new(files2.store()), on_disk: false };
+        s2.open_library_in(stores, true).unwrap();
+        assert_eq!(s2.catalog.photo(id).unwrap().rating, 5, "the rating survived");
+    }
+
    #[test]
    fn library_survives_reload_through_flushes() {
        let mut disk = BTreeMap::new();
@@ -14,7 +14,9 @@
 //! console (see [`bench`]); `?store=idb` forces the IndexedDB backend, `?store=memory` disables
 //! persistence and `?workers=0` renders on the main thread.
 #![forbid(unsafe_code)]
+#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)]

+pub mod backup;
 pub mod bench;
 pub mod files;
 pub mod store;
@@ -23,6 +25,8 @@ pub mod wire;
 #[cfg(target_arch = "wasm32")]
 mod backend;
 #[cfg(target_arch = "wasm32")]
+mod safety;
+#[cfg(target_arch = "wasm32")]
 mod web;
 #[cfg(target_arch = "wasm32")]
 mod workers;
@@ -0,0 +1,275 @@
+//! Keeping a library in the browser safe: notices the user must see (a failed save, storage the
+//! browser may evict, a file that couldn't be stored), one tab at a time per library, a message
+//! instead of a dead page after a panic, and backup / restore of the whole library
+//! ([`crate::backup`]).
+
+use std::cell::RefCell;
+use std::collections::HashMap;
+
+use js_sys::{Reflect, Uint8Array};
+use wasm_bindgen::JsCast;
+use wasm_bindgen::prelude::*;
+use wasm_bindgen_futures::JsFuture;
+
+use crate::backend::Backend;
+use crate::backup::{
+    ACTIVE_LIBRARY, README, ZipWriter, data_offset, find_central, original_entry, parse_central, restore_key, restored_dir_name, tail_len,
+    valid_hash, verify, verify_original,
+};
+use crate::files::{Files, LIBRARY_FILES};
+use crate::store::storage_key;
+
+thread_local! {
+    /// Messages for the user, shown by the app on its next frame.
+    static NOTICES: RefCell<Vec<String>> = const { RefCell::new(Vec::new()) };
+}
+
+/// Queue a message for the user (shown for several seconds; also logged).
+pub fn notice(text: impl Into<String>) {
+    let text = text.into();
+    log::warn!("{text}");
+    NOTICES.with(|n| n.borrow_mut().push(text));
+    crate::web::request_repaint();
+}
+
+/// Messages queued since the last call.
+pub fn take_notices() -> Vec<String> {
+    NOTICES.with(|n| std::mem::take(&mut *n.borrow_mut()))
+}
+
+fn document() -> Option<web_sys::Document> {
+    web_sys::window()?.document()
+}
+
+/// Ask the user (the browser's own confirm box). `false` when it can't be shown.
+pub fn confirm(text: &str) -> bool {
+    web_sys::window().and_then(|w| w.confirm_with_message(text).ok()).unwrap_or(false)
+}
+
+/// Cover the page with a message (the app can't continue, or must not start).
+pub fn show_blocking(text: &str) {
+    let Some(doc) = document() else { return };
+    let el = match doc.get_element_by_id("lightcraft_blocking") {
+        Some(el) => el,
+        None => {
+            let Ok(el) = doc.create_element("div") else { return };
+            el.set_id("lightcraft_blocking");
+            let _ = el.set_attribute(
+                "style",
+                "position:fixed;inset:0;z-index:10;display:flex;align-items:center;justify-content:center;padding:32px;\
+                 background:#1c1c1c;color:#e0e0e0;font:15px/1.5 system-ui,-apple-system,'Segoe UI',sans-serif;text-align:center;white-space:pre-line",
+            );
+            if let Some(body) = doc.body() {
+                let _ = body.append_child(&el);
+            }
+            el
+        }
+    };
+    el.set_text_content(Some(text));
+    if let Some(loading) = doc.get_element_by_id("lightcraft_loading") {
+        loading.remove();
+    }
+}
+
+/// `panic = "abort"` on wasm: a panic ends the module. Say so on the page instead of leaving a
+/// frozen canvas, and say that the stored library is intact.
+pub fn install_panic_hook() {
+    std::panic::set_hook(Box::new(|info| {
+        let msg = match (info.payload().downcast_ref::<&str>(), info.payload().downcast_ref::<String>()) {
+            (Some(s), _) => (*s).to_string(),
+            (_, Some(s)) => s.clone(),
+            _ => "unknown error".into(),
+        };
+        let at = info.location().map(|l| format!(" at {}:{}", l.file(), l.line())).unwrap_or_default();
+        log::error!("LightCraft panicked{at}: {msg}");
+        show_blocking(&format!(
+            "LightCraft stopped after an internal error:\n{msg}\n\nThe library saved in this browser is kept. Reload the page to continue.\n\
+             If it happens again right away, a photo may be the cause: reload with ?workers=0 or report the error."
+        ));
+    }));
+}
+
+/// The name of the lock that keeps a library to one tab.
+const TAB_LOCK: &str = "lightcraft-library";
+
+/// Hold the library lock for as long as this page lives (Web Locks API). `Some(true)`: ours;
+/// `Some(false)`: another tab of this site has it; `None`: the browser can't tell (no Web Locks, or
+/// it refused the request).
+pub async fn acquire_tab_lock() -> Option<bool> {
+    let nav = Reflect::get(&js_sys::global(), &"navigator".into()).ok()?;
+    let locks = Reflect::get(&nav, &"locks".into()).ok().filter(|l| !l.is_undefined() && !l.is_null())?;
+    let request: js_sys::Function = Reflect::get(&locks, &"request".into()).ok()?.dyn_into().ok()?;
+    let opts = js_sys::Object::new();
+    Reflect::set(&opts, &"ifAvailable".into(), &true.into()).ok()?;
+    let granted = js_sys::Promise::new(&mut |resolve, reject| {
+        let cb = Closure::once_into_js(move |lock: JsValue| -> JsValue {
+            let ours = !lock.is_null() && !lock.is_undefined();
+            let _ = resolve.call1(&JsValue::NULL, &ours.into());
+            // keep the lock until the page goes away: a promise that never settles
+            if ours { js_sys::Promise::new(&mut |_, _| {}).into() } else { JsValue::NULL }
+        });
+        // a refused request never calls `cb`: settle with its error instead of waiting forever
+        match request.call3(&locks, &TAB_LOCK.into(), &opts, &cb) {
+            Ok(pending) => {
+                let catch = Reflect::get(&pending, &"catch".into()).ok().and_then(|c| c.dyn_into::<js_sys::Function>().ok());
+                if let Some(catch) = catch {
+                    let _ = catch.call1(&pending, &reject);
+                }
+            }
+            Err(e) => {
+                let _ = reject.call1(&JsValue::NULL, &e);
+            }
+        }
+    });
+    match JsFuture::from(granted).await {
+        Ok(v) => Some(v.is_truthy()),
+        Err(e) => {
+            log::warn!("Web Locks request failed, starting without the one-tab guard: {e:?}");
+            None
+        }
+    }
+}
+
+/// Offer a blob as a download named `name`.
+pub fn download_blob(name: &str, blob: &web_sys::Blob) -> Result<(), String> {
+    let e = |e: JsValue| format!("download failed: {e:?}");
+    let url = web_sys::Url::create_object_url_with_blob(blob).map_err(e)?;
+    let doc = document().ok_or("no document")?;
+    let a: web_sys::HtmlAnchorElement = doc.create_element("a").map_err(e)?.unchecked_into();
+    a.set_href(&url);
+    a.set_download(name);
+    a.click();
+    // revoke once the download has started
+    let revoke = Closure::once_into_js(move || {
+        let _ = web_sys::Url::revoke_object_url(&url);
+    });
+    if let Some(w) = web_sys::window() {
+        let _ = w.set_timeout_with_callback_and_timeout_and_arguments_0(revoke.unchecked_ref(), 60_000);
+    }
+    Ok(())
+}
+
+/// File ▸ Back Up Library…: a zip of the library files (as the app has them, saved or not) and
+/// every original in storage, built entry by entry into the browser's memory (outside the wasm
+/// heap) and downloaded. `names`: content hash → the photo's file name.
+pub async fn backup(files: Files, backend: Backend, names: HashMap<String, String>) -> Result<(usize, f64), String> {
+    let parts = js_sys::Array::new();
+    let mut zip = ZipWriter::default();
+    let mut bytes = 0f64;
+    let mut add = |zip: &mut ZipWriter, name: &str, data: &[u8]| -> Result<(), String> {
+        let header = zip.entry(name, data)?;
+        parts.push(&Uint8Array::from(&header[..]));
+        parts.push(&Uint8Array::from(data));
+        bytes += (header.len() + data.len()) as f64;
+        Ok(())
+    };
+    add(&mut zip, "README.txt", README.as_bytes())?;
+    for name in LIBRARY_FILES {
+        if let Some(d) = files.get(name) {
+            add(&mut zip, &format!("library/{name}"), &d)?;
+        }
+    }
+    let mut originals = 0;
+    for hash in backend.list("originals").await? {
+        if !valid_hash(&hash) {
+            continue;
+        }
+        match backend.read(&storage_key(&hash)).await {
+            Ok(Some(data)) => {
+                let name = names.get(&hash).map_or("original", String::as_str);
+                add(&mut zip, &original_entry(&hash, name), &data)?;
+                originals += 1;
+            }
+            Ok(None) => {}
+            Err(e) => return Err(format!("reading an original from browser storage: {e}")),
+        }
+    }
+    let tail = zip.finish()?;
+    parts.push(&Uint8Array::from(&tail[..]));
+    let opts = web_sys::BlobPropertyBag::new();
+    opts.set_type("application/zip");
+    let blob = web_sys::Blob::new_with_u8_array_sequence_and_options(&parts, &opts).map_err(|e| format!("{e:?}"))?;
+    let day: String = String::from(js_sys::Date::new_0().to_iso_string()).chars().take(10).collect();
+    download_blob(&format!("lightcraft-library-backup-{day}.zip"), &blob)?;
+    Ok((originals, bytes + tail.len() as f64))
+}
+
+/// Bytes `start..end` of a picked file.
+async fn read_range(file: &web_sys::File, start: u64, end: u64) -> Result<Vec<u8>, String> {
+    let blob = file.slice_with_f64_and_f64(start as f64, end as f64).map_err(|e| format!("{e:?}"))?;
+    let buf = JsFuture::from(blob.array_buffer()).await.map_err(|e| format!("{e:?}"))?;
+    Ok(Uint8Array::new(&buf).to_vec())
+}
+
+/// Restore a backup into a new library folder (nothing in storage is deleted or overwritten),
+/// then point [`ACTIVE_LIBRARY`] at it. Returns the folder's name; the caller reloads the page.
+pub async fn restore(file: web_sys::File, backend: Backend) -> Result<String, String> {
+    let len = file.size() as u64;
+    let tail = read_range(&file, len - tail_len(len), len).await?;
+    let (cd_offset, cd_size) = find_central(&tail)?;
+    let cd = read_range(&file, cd_offset, cd_offset + cd_size).await?;
+    let entries = parse_central(&cd)?;
+    if !entries.iter().any(|e| e.name == "library/catalog.snap" || e.name == "library/catalog.log") {
+        return Err("this isn't a LightCraft library backup (no library/catalog files)".into());
+    }
+    let dir = restored_dir_name(js_sys::Date::now());
+    let mut written = 0usize;
+    for e in &entries {
+        let Some(key) = restore_key(&e.name, &dir) else { continue };
+        // originals are named by their content: one already stored is the same file
+        if key.starts_with("originals/") && backend.exists(&key).await? {
+            continue;
+        }
+        let local = read_range(&file, e.header, e.header + 30).await?;
+        let start = data_offset(e, &local)?;
+        let data = read_range(&file, start, start + e.size).await?;
+        verify(e, &data)?;
+        if let Some(hash) = key.strip_prefix("originals/") {
+            verify_original(e, hash, &data)?;
+        }
+        backend.write(&key, &data).await.map_err(|err| format!("storing {}: {err}", e.name))?;
+        written += 1;
+    }
+    backend.write(ACTIVE_LIBRARY, dir.as_bytes()).await?;
+    log::info!("lightcraft: restored {written} files into {dir}");
+    Ok(dir)
+}
+
+/// File ▸ Restore Library from Backup…: confirm, pick the zip, restore, reload.
+pub fn pick_and_restore(backend: Backend, before_reload: impl FnOnce() + 'static) {
+    let ok = confirm(
+        "Restore a LightCraft library backup?\n\nThe backup becomes the library in this browser and the page reloads. \
+         The current library is not deleted: it stays in browser storage.",
+    );
+    if !ok {
+        return;
+    }
+    let Some(doc) = document() else { return };
+    let Ok(input) = doc.create_element("input").map(|e| e.unchecked_into::<web_sys::HtmlInputElement>()) else { return };
+    input.set_type("file");
+    input.set_accept(".zip,application/zip");
+    let inp = input.clone();
+    let mut before_reload = Some(before_reload);
+    let on_change = Closure::<dyn FnMut()>::new(move || {
+        let Some(file) = inp.files().and_then(|f| f.get(0)) else { return };
+        let backend = backend.clone();
+        let before = before_reload.take();
+        notice(format!("Restoring {}…", file.name()));
+        wasm_bindgen_futures::spawn_local(async move {
+            match restore(file, backend).await {
+                Ok(_) => {
+                    if let Some(f) = before {
+                        f();
+                    }
+                    if let Some(w) = web_sys::window() {
+                        let _ = w.location().reload();
+                    }
+                }
+                Err(e) => notice(format!("Restore failed: {e}. Nothing was changed: the current library is still in use.")),
+            }
+        });
+    });
+    input.set_onchange(Some(on_change.as_ref().unchecked_ref()));
+    on_change.forget();
+    input.click();
+}
@@ -15,37 +15,53 @@ use wasm_bindgen::JsCast;
 use wasm_bindgen::prelude::*;

 use crate::backend::Backend;
+use crate::backup::{ACTIVE_LIBRARY, DEFAULT_LIBRARY_DIR, valid_library_dir};
 use crate::bench::Bench;
 use crate::files::{Files, FlushOp, LIBRARY_FILES};
+use crate::safety::{self, notice};
 use crate::store::{Originals, content_hash, download_name, storage_key};
 use crate::wire::ThumbIndex;
 use crate::workers::{THUMB_INDEX, Workers};

 const ACCEPT: &str = ".jpg,.jpeg,.png,.tif,.tiff,.webp,.dng,.cr2,.nef,.arw,.psd,.jxl,.gif,.bmp";

-/// Where the library files live in browser storage.
-const LIBRARY_DIR: &str = "library";
+/// After a failed save, wait this long before trying again (ms).
+const SAVE_RETRY_MS: f64 = 2000.0;
+
+/// How long a notice stays on screen (s).
+const NOTICE_SECS: f64 = 8.0;

 /// How often view state, UI prefs and the thumbnail index are saved (ms).
 const SAVE_EVERY_MS: f64 = 1000.0;

-fn window() -> web_sys::Window {
-    web_sys::window().expect("no window")
+/// The page's window (`None` only outside a browser page, e.g. in a worker).
+fn window() -> Option<web_sys::Window> {
+    web_sys::window()
 }

 /// Milliseconds since navigation start.
 fn perf_now() -> f64 {
-    window().performance().map(|p| p.now()).unwrap_or(0.0)
+    window().and_then(|w| w.performance()).map(|p| p.now()).unwrap_or(0.0)
 }

 fn set_status(text: &str) {
-    if let Some(el) = window().document().and_then(|d| d.get_element_by_id("lightcraft_status")) {
+    if let Some(el) = window().and_then(|w| w.document()).and_then(|d| d.get_element_by_id("lightcraft_status")) {
        el.set_text_content(Some(text));
    }
 }

-fn library_key(name: &str) -> String {
-    format!("{LIBRARY_DIR}/{name}")
+/// Storage key of library file `name` in library folder `dir`.
+fn library_key(dir: &str, name: &str) -> String {
+    format!("{dir}/{name}")
+}
+
+/// Repaint the app (notices from async tasks).
+pub(crate) fn request_repaint() {
+    CTX.with(|c| {
+        if let Some(c) = c.borrow().as_ref() {
+            c.request_repaint();
+        }
+    });
 }

 /// URL options (`?bench&store=idb&workers=2&reset`).
@@ -59,7 +75,7 @@ struct Options {

 impl Options {
    fn from_url() -> Options {
-        let q = window().location().search().unwrap_or_default();
+        let q = window().and_then(|w| w.location().search().ok()).unwrap_or_default();
        let params: Vec<(String, String)> = q
            .trim_start_matches('?')
            .split('&')
@@ -85,8 +101,9 @@ async fn store_file(originals: Originals, backend: Option<Backend>, name: String
    if let Some(b) = &backend
        && let Err(e) = b.write(&storage_key(&hash), &bytes).await
    {
-        // still importable for this session; it won't survive a reload
-        log::error!("storing {name} in browser storage failed: {e}");
+        // a photo whose bytes aren't stored would be gone after a reload: don't add it
+        notice(format!("{name} was not added: browser storage refused it ({e}). Free up space or back up and clear old photos."));
+        return;
    }
    originals.added(&name, &hash, Arc::from(bytes));
    ctx.request_repaint();
@@ -102,7 +119,7 @@ async fn read_file(originals: Originals, backend: Option<Backend>, file: web_sys

 /// Show the browser's open dialog; picked files are imported asynchronously.
 fn open_picker(originals: Originals, backend: Option<Backend>, ctx: egui::Context) {
-    let Some(doc) = window().document() else { return };
+    let Some(doc) = window().and_then(|w| w.document()) else { return };
    let Ok(input) = doc.create_element("input").map(|e| e.unchecked_into::<web_sys::HtmlInputElement>()) else { return };
    input.set_type("file");
    input.set_multiple(true);
@@ -131,7 +148,7 @@ fn download(path: &str, bytes: &[u8]) -> Result<(), String> {
    opts.set_type(crate::store::mime_for(name));
    let blob = web_sys::Blob::new_with_u8_array_sequence_and_options(&arr, &opts).map_err(e)?;
    let url = web_sys::Url::create_object_url_with_blob(&blob).map_err(e)?;
-    let doc = window().document().ok_or("no document")?;
+    let doc = window().and_then(|w| w.document()).ok_or("no document")?;
    let a: web_sys::HtmlAnchorElement = doc.create_element("a").map_err(e)?.unchecked_into();
    a.set_href(&url);
    a.set_download(name);
@@ -140,20 +157,64 @@ fn download(path: &str, bytes: &[u8]) -> Result<(), String> {
    let revoke = Closure::once_into_js(move || {
        let _ = web_sys::Url::revoke_object_url(&url);
    });
-    let _ = window().set_timeout_with_callback_and_timeout_and_arguments_0(revoke.unchecked_ref(), 5000);
+    if let Some(w) = window() {
+        let _ = w.set_timeout_with_callback_and_timeout_and_arguments_0(revoke.unchecked_ref(), 5000);
+    }
    log::info!("exported {name} ({} bytes)", bytes.len());
    Ok(())
 }

-fn services(originals: Originals, backend: Option<Backend>, ctx: egui::Context) -> Services {
+fn services(originals: Originals, backend: Option<Backend>, files: Files, frozen: Rc<Cell<bool>>, ctx: egui::Context) -> Services {
+    let (backup_backend, backup_files) = (backend.clone(), files);
+    let restore_backend = backend.clone();
    Services {
+        picker: None,
+        backup_library: Some(Box::new(move |session: &mut Session| {
+            let Some(b) = backup_backend.clone() else { return Err("nothing is stored in this browser session (?store=memory)".into()) };
+            // the photos' own names for the originals in the zip
+            let names: std::collections::HashMap<String, String> = session
+                .catalog
+                .photos()
+                .filter_map(|p| match &p.source {
+                    Source::File { path } => crate::store::hash_of_path(path).map(|h| (h.to_string(), p.file_name.clone())),
+                    _ => None,
+                })
+                .collect();
+            let files = backup_files.clone();
+            notice("Backing up the library… (the download starts when it's ready)");
+            wasm_bindgen_futures::spawn_local(async move {
+                match safety::backup(files, b, names).await {
+                    Ok((n, bytes)) => notice(format!(
+                        "Library backed up: {n} original{} ({:.1} MB). Keep the file somewhere safe.",
+                        if n == 1 { "" } else { "s" },
+                        bytes / 1e6
+                    )),
+                    Err(e) => notice(format!("Backup failed: {e}")),
+                }
+            });
+            Ok(json!({"started": true}))
+        })),
+        restore_library: Some(Box::new(move |_session: &mut Session| {
+            let Some(b) = restore_backend.clone() else { return Err("nothing is stored in this browser session (?store=memory)".into()) };
+            let frozen = frozen.clone();
+            // once the backup is in place, stop saving the old library (the page reloads)
+            safety::pick_and_restore(b, move || frozen.set(true));
+            Ok(json!({"started": true}))
+        })),
        pick_files: Some(Box::new(move || {
            open_picker(originals.clone(), backend.clone(), ctx.clone());
            Vec::new() // files arrive asynchronously and are imported on a later frame
        })),
+        pick_denoise_model: None,
        // Preset files: browser pickers are asynchronous; not wired on the web yet.
        pick_preset_files: None,
+        pick_tracklog: None,
+        pick_lightroom_catalog: None,
+        // Face models need a folder to live in; the web has none.
+        pick_model_file: None,
        save_preset_file: None,
+        pick_curve_preset_files: None,
+        save_curve_preset_file: None,
        write: Some(Box::new(download)),
        // downloads happen on the main thread: exports run in the foreground on the web
        write_shared: None,
@@ -161,6 +222,7 @@ fn services(originals: Originals, backend: Option<Backend>, ctx: egui::Context)
            lightcraft_codecs::encode_png(&lightcraft_codecs::EncodeImage::rgba8(img), &lightcraft_codecs::EncodeMeta::default()).unwrap_or_default()
        })),
        reveal: None,
+        log_file: None,
        open_with: None,
        open_url: Some(Box::new(|url: &str| {
            let w = web_sys::window().ok_or("no window")?;
@@ -170,21 +232,29 @@ fn services(originals: Originals, backend: Option<Backend>, ctx: egui::Context)
    }
 }

-/// Write dirty library files to storage, in order, until none are left.
-async fn flush(files: Files, backend: Backend, flushing: Rc<Cell<bool>>) {
+/// Write dirty library files to storage, in order, until none are left. A failure (quota,
+/// storage cleared) is recorded in `files` — the catalog then reports changes as unsaved instead
+/// of accepting them silently — and retried after [`SAVE_RETRY_MS`].
+async fn flush(files: Files, backend: Backend, dir: String, flushing: Rc<Cell<bool>>, failed_at: Rc<Cell<f64>>) {
    'outer: loop {
        let ops = files.take_dirty();
        if ops.is_empty() {
+            if files.error().is_some() {
+                log::info!("lightcraft: saving works again");
+            }
+            files.set_error(None);
            break;
        }
        for (i, op) in ops.iter().enumerate() {
            let r = match op {
-                FlushOp::Write { name, data } => backend.write(&library_key(name), data).await,
-                FlushOp::Append { name, offset, data } => backend.write_at(&library_key(name), Some(*offset), data).await,
+                FlushOp::Write { name, data } => backend.write(&library_key(&dir, name), data).await,
+                FlushOp::Append { name, offset, data } => backend.write_at(&library_key(&dir, name), Some(*offset), data).await,
            };
            if let Err(e) = r {
                log::error!("saving {} failed: {e}", op.name());
                files.failed(&ops[i..]);
+                files.set_error(Some(e));
+                failed_at.set(perf_now());
                break 'outer;
            }
        }
@@ -198,6 +268,8 @@ struct Boot {
    backend: Option<Backend>,
    files: Files,
    index: ThumbIndex,
+    /// The library folder in storage (`library`, or a restored backup's).
+    lib_dir: String,
 }

 async fn boot(opts: Options) -> Boot {
@@ -215,15 +287,32 @@ async fn boot(opts: Options) -> Boot {
    };
    let files = Files::default();
    let mut index = ThumbIndex::default();
+    let mut lib_dir = DEFAULT_LIBRARY_DIR.to_string();
    if let Some(b) = &backend {
        if opts.reset {
-            match b.clear().await {
-                Ok(()) => log::info!("lightcraft: storage cleared (?reset)"),
-                Err(e) => log::error!("clearing storage: {e}"),
+            // everything this site keeps in the browser, photos included: never without asking
+            let ok = safety::confirm(
+                "?reset deletes the LightCraft library stored in this browser, including every imported photo. \
+                 This can't be undone. Delete it?",
+            );
+            if ok {
+                match b.clear().await {
+                    Ok(()) => log::info!("lightcraft: storage cleared (?reset)"),
+                    Err(e) => log::error!("clearing storage: {e}"),
+                }
+            } else {
+                log::info!("lightcraft: ?reset declined; storage kept");
+            }
+        }
+        // a restored backup lives in its own folder (see `backup`)
+        if let Ok(Some(d)) = b.read(ACTIVE_LIBRARY).await {
+            let d = String::from_utf8_lossy(&d).trim().to_string();
+            if valid_library_dir(&d) {
+                lib_dir = d;
            }
        }
        for name in LIBRARY_FILES {
-            match b.read(&library_key(name)).await {
+            match b.read(&library_key(&lib_dir, name)).await {
                Ok(Some(data)) => files.preload(name, data),
                Ok(None) => {}
                Err(e) => log::error!("reading {name}: {e}"),
@@ -232,7 +321,14 @@ async fn boot(opts: Options) -> Boot {
        if let Ok(Some(j)) = b.read(THUMB_INDEX).await {
            index = ThumbIndex::from_json(&j);
        }
-        crate::backend::request_persistence();
+        crate::backend::request_persistence(|granted| {
+            if !granted {
+                notice(
+                    "This browser may delete LightCraft's library when it runs short of space (persistent storage wasn't granted). \
+                     Keep your original photos elsewhere and use File ▸ Back Up Library… now and then.",
+                );
+            }
+        });
        // thumbnails stored without an index entry (index saved late, or lost): delete them
        let b2 = b.clone();
        let known: std::collections::HashSet<String> = b
@@ -260,7 +356,7 @@ async fn boot(opts: Options) -> Boot {
        index.len(),
        index.total() as f64 / 1e6
    );
-    Boot { opts, backend, files, index }
+    Boot { opts, backend, files, index, lib_dir }
 }

 struct WebApp {
@@ -270,6 +366,11 @@ struct WebApp {
    backend: Option<Backend>,
    workers: Option<Workers>,
    flushing: Rc<Cell<bool>>,
+    /// When the last save to storage failed (ms; 0 = never).
+    flush_failed_at: Rc<Cell<f64>>,
+    /// Saving stopped (a restored backup is about to replace this library: the page reloads).
+    frozen: Rc<Cell<bool>>,
+    lib_dir: String,
    bench: Option<Bench>,
    first_frame_logged: bool,
    last_save: f64,
@@ -277,18 +378,19 @@ struct WebApp {
 }

 impl WebApp {
-    fn new(cc: &eframe::CreationContext<'_>, boot: Boot) -> Self {
-        let Boot { opts, backend, files, index } = boot;
+    fn new(cc: &eframe::CreationContext<'_>, boot: Boot, chinese_font: Vec<u8>) -> Self {
+        let Boot { opts, backend, files, index, lib_dir } = boot;
        let originals = Originals::default();
        let t = perf_now();
        let mut session = Session::new();
        originals.install(&mut session);
        let stores = LibraryStores {
-            dir: format!("browser:{}/{LIBRARY_DIR}", backend.as_ref().map_or("memory", |b| b.kind())).into(),
+            dir: format!("browser:{}/{lib_dir}", backend.as_ref().map_or("memory", |b| b.kind())).into(),
            catalog: Box::new(files.store()),
            files: Box::new(files.store()),
            on_disk: false,
        };
+        let mut problem = None;
        match session.open_library_in(stores, true).cloned() {
            Ok(r) => log::info!(
                "lightcraft: library {} in {:.0} ms ({} photos; snapshot seq {}, {} ops replayed)",
@@ -299,25 +401,52 @@ impl WebApp {
                r.replayed
            ),
            Err(e) => {
-                log::error!("opening the library failed: {e}; starting a temporary one");
-                session = Session::with_demo();
+                // never a silent demo session (issue #100): an empty one, and the window says why
+                log::error!("opening the library failed: {e}");
+                session = Session::new();
                originals.install(&mut session);
+                problem = Some(lightcraft_ui_egui::panels::library_problem::LibraryProblem {
+                    can_retry: false,
+                    ..lightcraft_ui_egui::panels::library_problem::LibraryProblem::new(
+                        "the browser's storage for this page",
+                        format!("{e}. Reload the page to try again."),
+                    )
+                });
            }
        }
+        if backend.is_some() {
+            notice(
+                "LightCraft in the browser is experimental. Your library is kept in this browser's storage: keep your original photos \
+                 elsewhere and back up with File ▸ Back Up Library….",
+            );
+        }
        // previews are large (≤ 2560 px, f32): keep few in a 32-bit address space
        session.media.preview_capacity = 3;
-        let mut app = LightcraftApp::new(session, services(originals.clone(), backend.clone(), cc.egui_ctx.clone()));
+        let frozen = Rc::new(Cell::new(false));
+        let mut app = LightcraftApp::new(session, services(originals.clone(), backend.clone(), files.clone(), frozen.clone(), cc.egui_ctx.clone()))
+            .with_chinese_font(chinese_font);
        let ui_written = files.get("ui.json").unwrap_or_default();
        if let Ok(ui) = serde_json::from_slice::<UiState>(&ui_written) {
            app.ui = ui;
        }
        app.ui = app.ui.sanitized();
+        app.library_problem = problem;
        let n = opts.workers.unwrap_or_else(|| {
-            let cores = window().navigator().hardware_concurrency() as usize;
+            let cores = window().map_or(1, |w| w.navigator().hardware_concurrency() as usize);
            cores.saturating_sub(1).clamp(1, 4)
        });
-        let workers =
-            (n > 0).then(|| Workers::start(n, backend.as_ref().map_or("memory", |b| b.kind()), backend.clone(), index, cc.egui_ctx.clone()));
+        let cache = app.session.media.rendered.clone();
+        let workers = (n > 0).then(|| {
+            Workers::start(
+                n,
+                backend.as_ref().map_or("memory", |b| b.kind()),
+                backend.clone(),
+                index,
+                &cache,
+                app.session.cache_bytes(),
+                cc.egui_ctx.clone(),
+            )
+        });
        if let Some(w) = &workers {
            app.renderer.set_offload(Box::new(w.clone()));
        }
@@ -331,6 +460,9 @@ impl WebApp {
            backend,
            workers,
            flushing: Rc::new(Cell::new(false)),
+            flush_failed_at: Rc::new(Cell::new(0.0)),
+            frozen,
+            lib_dir,
            bench: opts.bench.then(|| Bench::new(origin)),
            first_frame_logged: false,
            last_save: 0.0,
@@ -392,6 +524,9 @@ impl WebApp {

    /// Save view state, UI prefs and the thumbnail index now and then; flush dirty files.
    fn save(&mut self) {
+        if self.frozen.get() {
+            return;
+        }
        let now = perf_now();
        if now - self.last_save >= SAVE_EVERY_MS {
            self.last_save = now;
@@ -413,12 +548,16 @@ impl WebApp {
                });
            }
        }
+        // after a failure, retry every couple of seconds rather than every frame
+        let backing_off = self.files.error().is_some() && now - self.flush_failed_at.get() < SAVE_RETRY_MS;
        if let Some(b) = &self.backend
            && self.files.is_dirty()
            && !self.flushing.get()
+            && !backing_off
        {
            self.flushing.set(true);
-            wasm_bindgen_futures::spawn_local(flush(self.files.clone(), b.clone(), self.flushing.clone()));
+            let (files, dir, flushing, failed_at) = (self.files.clone(), self.lib_dir.clone(), self.flushing.clone(), self.flush_failed_at.clone());
+            wasm_bindgen_futures::spawn_local(flush(files, b.clone(), dir, flushing, failed_at));
        }
    }
 }
@@ -448,8 +587,10 @@ impl WebApp {
        let (orig_n, orig_bytes) = self.originals.usage();
        json!({
            "storage": self.backend.as_ref().map_or("memory", |b| b.kind()),
+            "library": self.lib_dir,
            "dirty": self.files.is_dirty(),
            "flushing": self.flushing.get(),
+            "saveError": self.files.error(),
            "workers": {"alive": alive, "ready": ready, "remoteJobs": remote, "inlineJobs": inline},
            "originalsInMemory": orig_n,
            "originalBytesInMemory": orig_bytes,
@@ -474,9 +615,18 @@ impl WebApp {

 impl eframe::App for WebApp {
    fn logic(&mut self, ctx: &egui::Context, _frame: &mut eframe::Frame) {
+        for text in safety::take_notices() {
+            let now = ctx.input(|i| i.time);
+            self.app.ui.toast = Some((text, now + NOTICE_SECS, None));
+        }
        self.run_inbox();
        self.import_dropped(ctx);
        self.load_originals(ctx);
+        if let Some(w) = &self.workers {
+            // `library.preferences ▸ cacheMb` is read by the session; the worker index prunes on
+            // its own thread, so it gets the budget every frame and applies it on the next store.
+            w.set_budget(self.app.session.cache_bytes());
+        }
        self.app.logic(ctx);
        self.save();
        if let Some(b) = self.bench.as_mut()
@@ -507,25 +657,39 @@ impl eframe::App for WebApp {
    }
 }

-/// Start the app (called by `index.html` after instantiating the module).
+/// Start the app with the separately fetched Simplified Chinese font (empty in font-free builds).
 #[wasm_bindgen]
-pub fn start() {
+pub fn start(chinese_font: Vec<u8>) {
    eframe::WebLogger::init(log::LevelFilter::Info).ok();
+    safety::install_panic_hook();
    log::info!("lightcraft: wasm instantiated at {:.0} ms", perf_now());
+    if chinese_font.is_empty() {
+        log::warn!("lightcraft: no Simplified Chinese font supplied; Chinese-only glyphs will show as boxes");
+    }
    let opts = Options::from_url();
    wasm_bindgen_futures::spawn_local(async move {
-        let Some(canvas) =
-            window().document().and_then(|d| d.get_element_by_id("lightcraft_canvas")).and_then(|e| e.dyn_into::<web_sys::HtmlCanvasElement>().ok())
+        // one tab per library: two would each keep their own copy and overwrite each other's saves
+        if opts.store != "memory" && safety::acquire_tab_lock().await == Some(false) {
+            safety::show_blocking(
+                "LightCraft is already open in another tab or window of this browser.\n\n\
+                 Switch to that tab, or close it and reload this page. (Two tabs would overwrite each other's changes.)",
+            );
+            return;
+        }
+        let Some(canvas) = window()
+            .and_then(|w| w.document())
+            .and_then(|d| d.get_element_by_id("lightcraft_canvas"))
+            .and_then(|e| e.dyn_into::<web_sys::HtmlCanvasElement>().ok())
        else {
            log::error!("missing <canvas id=\"lightcraft_canvas\">");
            return;
        };
        let boot = boot(opts).await;
        let runner = eframe::WebRunner::new();
-        let r = runner.start(canvas, eframe::WebOptions::default(), Box::new(move |cc| Ok(Box::new(WebApp::new(cc, boot))))).await;
+        let r = runner.start(canvas, eframe::WebOptions::default(), Box::new(move |cc| Ok(Box::new(WebApp::new(cc, boot, chinese_font))))).await;
        match r {
            Ok(()) => {
-                if let Some(el) = window().document().and_then(|d| d.get_element_by_id("lightcraft_loading")) {
+                if let Some(el) = window().and_then(|w| w.document()).and_then(|d| d.get_element_by_id("lightcraft_loading")) {
                    el.remove();
                }
            }
@@ -11,14 +11,13 @@
 //! ([`ThumbIndex`]).

 use std::collections::HashMap;
-use std::sync::Arc;
+use std::sync::{Arc, Weak};

 use lightcraft_engine::catalog::Source;
 use lightcraft_engine::develop::{DevelopSettings, EmbeddedLens};
-use lightcraft_engine::media::{MediaCache, RenderJob, SourceLevel, SourceRef};
+use lightcraft_engine::media::{DecodedSource, MediaCache, RenderJob, SourceLevel, SourceRef};
 use lightcraft_engine::pipeline::{Quality, RenderRequest, Rendered, SourceInfo, StageCache};
-use lightcraft_preview::Lru;
-use lightcraft_raster::Rgb32f;
+use lightcraft_preview::{Hash128, Lru, PreviewCache};
 use serde::{Deserialize, Serialize};

 use crate::store::hash_of_path;
@@ -26,11 +25,17 @@ use crate::store::hash_of_path;
 /// One render request for a worker.
 #[derive(Clone, Debug, Serialize, Deserialize)]
 pub struct WireJob {
+    #[serde(default)]
+    pub cache_generation: u64,
+    #[serde(default)]
+    pub source_identity: Option<String>,
    pub level: SourceLevel,
    pub origin: Source,
    /// Decode the source at most this long.
    pub max_edge: usize,
    pub raw: bool,
+    #[serde(default)]
+    pub relative_wb: bool,
    pub as_shot_temp: f64,
    pub as_shot_tint: f64,
    pub lens: Option<EmbeddedLens>,
@@ -48,6 +53,9 @@ pub struct WireJob {
    /// The request's diagnostic overlay (`Overlay::to_parts`).
    #[serde(default)]
    pub overlay: (u8, f64),
+    /// Render only this window `[x, y, w, h]` of the output `max_w × max_h` describes (a zoomed view).
+    #[serde(default)]
+    pub window: Option<[usize; 4]>,
 }

 impl WireJob {
@@ -55,14 +63,17 @@ impl WireJob {
    pub fn from_job(job: &RenderJob, thumb_cached: bool, view: &str) -> WireJob {
        let max_edge = match &job.source {
            SourceRef::Demo { max_edge, .. } | SourceRef::File { max_edge, .. } => *max_edge,
-            SourceRef::Loaded(img) => img.width.max(img.height),
+            SourceRef::Loaded(img) => img.image.width.max(img.image.height),
            SourceRef::Smart { .. } => 2560,
        };
        WireJob {
+            cache_generation: job.cache_generation,
+            source_identity: job.source_key.map(|key| key.to_string()),
            level: job.level,
            origin: job.origin.clone(),
            max_edge,
            raw: job.info.raw,
+            relative_wb: job.info.relative_wb,
            as_shot_temp: job.info.as_shot_temp,
            as_shot_tint: job.info.as_shot_tint,
            lens: job.info.lens,
@@ -75,11 +86,19 @@ impl WireJob {
            thumb_cached,
            stages: job.stages.is_some().then(|| view.to_string()),
            overlay: job.request.overlay.to_parts(),
+            window: job.request.window.map(|w| [w.x, w.y, w.w, w.h]),
        }
    }

    pub fn info(&self) -> SourceInfo {
-        SourceInfo { raw: self.raw, as_shot_temp: self.as_shot_temp, as_shot_tint: self.as_shot_tint, lens: self.lens }
+        SourceInfo {
+            raw: self.raw,
+            as_shot_temp: self.as_shot_temp,
+            as_shot_tint: self.as_shot_tint,
+            lens: self.lens,
+            relative_wb: self.relative_wb,
+            ..Default::default()
+        }
    }

    pub fn request(&self) -> RenderRequest {
@@ -93,14 +112,15 @@ impl WireJob {
            space: lightcraft_engine::pipeline::OutputSpace::Srgb,
            depth: lightcraft_engine::pipeline::OutputDepth::U8,
            proof: None,
+            window: self.window.map(|[x, y, w, h]| lightcraft_engine::pipeline::PixelWindow { x, y, w, h }),
        }
    }

    /// Identifies the decoded source this job needs.
    fn source_key(&self) -> String {
        match &self.origin {
-            Source::File { path } => format!("{}:{path}", self.max_edge),
-            Source::Demo { scene } => format!("{}:demo:{scene}", self.max_edge),
+            Source::File { path } => format!("{}:{:?}:{}:{path}", self.cache_generation, self.source_identity, self.max_edge),
+            Source::Demo { scene } => format!("{}:{:?}:{}:demo:{scene}", self.cache_generation, self.source_identity, self.max_edge),
        }
    }

@@ -119,7 +139,7 @@ const WORKER_SOURCE_BYTES: usize = 192 << 20;

 /// The synchronous part of a worker: decoded-source cache + rendering.
 pub struct WorkerCore {
-    sources: Lru<String, Arc<Rgb32f>>,
+    sources: Lru<String, DecodedSource>,
    media: MediaCache,
    stages: HashMap<String, Arc<StageCache>>,
 }
@@ -150,21 +170,25 @@ impl WorkerCore {
            Some(s) => s,
            None => {
                let src = match (&job.origin, original) {
-                    (Source::File { .. }, Some(bytes)) => Arc::new(lightcraft_engine::files::load_bytes(bytes, job.max_edge)?.0),
+                    (Source::File { .. }, Some(bytes)) => {
+                        let (image, info) = lightcraft_engine::files::load_bytes(bytes, job.max_edge)?;
+                        DecodedSource::new(Arc::new(image), Some(info))
+                    }
                    (Source::File { path }, None) => return Err(format!("{path}: original not found in browser storage")),
-                    (origin @ Source::Demo { .. }, _) => self.media.origin_ref(origin, job.max_edge).load()?,
+                    (origin @ Source::Demo { .. }, _) => self.media.origin_ref(origin, job.max_edge).load_source()?,
                };
-                let cost = src.width * src.height * 12 + 64;
+                let cost = src.image.width * src.image.height * 12 + 64;
                self.sources.insert(key, src.clone(), cost);
                src
            }
        };
+        let info = src.info_or(job.info());
        Ok(match &job.stages {
            Some(view) => {
                let st = self.stages.entry(view.clone()).or_default().clone();
-                lightcraft_engine::pipeline::render_cached(&src, &job.info(), &job.settings, &job.request(), &st)
+                lightcraft_engine::pipeline::render_cached(&src.image, &info, &job.settings, &job.request(), &st)
            }
-            None => lightcraft_engine::pipeline::render(&src, &job.info(), &job.settings, &job.request()),
+            None => lightcraft_engine::pipeline::render(&src.image, &info, &job.settings, &job.request()),
        })
    }
 }
@@ -173,6 +197,8 @@ impl WorkerCore {
 /// pruned like the native disk cache (least recently used first, down to 80 % of the budget).
 #[derive(Default, Serialize, Deserialize)]
 pub struct ThumbIndex {
+    #[serde(default)]
+    namespace: u64,
    /// hex key → (bytes, last use).
    entries: HashMap<String, (u64, u64)>,
    clock: u64,
@@ -183,6 +209,24 @@ pub struct ThumbIndex {
 }

 impl ThumbIndex {
+    pub fn namespace(&self) -> u64 {
+        self.namespace
+    }
+
+    /// A persistent new namespace prevents late old workers restoring an invalidated disk hit.
+    pub fn cache_key(&self, key: Hash128) -> String {
+        if self.namespace == 0 {
+            return key.to_string();
+        }
+        lightcraft_preview::Hasher128::new().str(&key.to_string()).str("cleared").u64(self.namespace).finish().to_string()
+    }
+
+    pub fn invalidate(&mut self) -> Vec<String> {
+        self.namespace = self.namespace.wrapping_add(1);
+        self.total = 0;
+        self.dirty = true;
+        self.entries.drain().map(|(key, _)| key).collect()
+    }
    pub fn from_json(bytes: &[u8]) -> ThumbIndex {
        let mut i: ThumbIndex = serde_json::from_slice(bytes).unwrap_or_default();
        i.total = i.entries.values().map(|e| e.0).sum();
@@ -260,11 +304,89 @@ impl ThumbIndex {
    }
 }

+/// Which rendered-thumbnail cache, at which generation, the stored thumbnails ([`ThumbIndex`])
+/// belong to (main thread). Clearing the previews bumps the cache's generation, and opening
+/// another library replaces the cache: either makes every stored thumbnail obsolete.
+pub struct CacheWatch(Option<(Weak<PreviewCache>, u64)>);
+
+impl CacheWatch {
+    /// Watch `active`, the session's cache when the workers start: the persisted index belongs
+    /// to it at its current generation. (Without this baseline, the first observation after a
+    /// clear would only record the cleared generation and keep the obsolete thumbnails.)
+    pub fn new(active: &Arc<PreviewCache>) -> CacheWatch {
+        CacheWatch(Some((Arc::downgrade(active), active.generation())))
+    }
+
+    /// The watched cache, while it's alive.
+    pub fn cache(&self) -> Option<Arc<PreviewCache>> {
+        self.0.as_ref().and_then(|(cache, _)| cache.upgrade())
+    }
+
+    /// `cache` at `generation` is the active one now. If it was cleared or replaced since the
+    /// last observation, invalidate `index` and return the keys whose files should be deleted.
+    pub fn observe(&mut self, cache: &Arc<PreviewCache>, generation: u64, index: &mut ThumbIndex) -> Vec<String> {
+        let changed = self.0.as_ref().is_some_and(|(old, g)| old.as_ptr() != Arc::as_ptr(cache) || *g != generation);
+        self.0 = Some((Arc::downgrade(cache), generation));
+        if changed { index.invalidate() } else { Vec::new() }
+    }
+}
+
 #[cfg(test)]
 mod tests {
    use super::*;
    use lightcraft_engine::Session;

+    #[test]
+    fn clearing_index_changes_disk_namespace_and_survives_restart() {
+        let key = lightcraft_preview::hash_bytes(b"thumbnail");
+        let mut index = ThumbIndex::default();
+        let old = index.cache_key(key);
+        assert_eq!(old, key.to_string());
+        index.insert(&old, 100);
+        assert_eq!(index.invalidate(), vec![old.clone()]);
+        let new = index.cache_key(key);
+        assert_ne!(old, new);
+        index.insert(&new, 100);
+        let reopened = ThumbIndex::from_json(&index.to_json());
+        assert_eq!(reopened.cache_key(key), new);
+        assert!(!reopened.contains(&old));
+        assert!(reopened.contains(&new));
+    }
+
+    // Issue #323: a worker given a zoomed view's window renders that window, not the whole frame
+    #[test]
+    fn a_window_job_crosses_the_wire_and_renders_the_same_window() {
+        use lightcraft_engine::pipeline::PixelWindow;
+        let mut s = Session::with_demo();
+        let id = s.visible_cloned()[0];
+        let win = PixelWindow { x: 300, y: 200, w: 160, h: 120 };
+        let job = s.region_job(id, 1920, 1280, win, true).unwrap();
+        let wire = WireJob::from_job(&job, false, "Region");
+        let wire: WireJob = serde_json::from_str(&serde_json::to_string(&wire).unwrap()).unwrap();
+        assert_eq!(wire.request().window, Some(win));
+        let mut core = WorkerCore::default();
+        let a = core.render(&wire, None).unwrap();
+        let b = job.run().rendered.unwrap();
+        assert_eq!((a.image.width, a.image.height), (160, 120));
+        assert_eq!((b.image.width, b.image.height), (160, 120));
+        // the worker renders on the CPU; the engine job uses a GPU when the machine has one, which
+        // stays within docs/gpu-pipeline.md's bound (≤ 3 LSB per channel) rather than bit-exact
+        let max = a.image.data.iter().zip(&b.image.data).flat_map(|(p, q)| (0..3).map(move |c| p[c].abs_diff(q[c]))).max();
+        assert!(max.is_some_and(|m| m <= 3), "max channel difference {max:?}");
+    }
+
+    // a job from before windows existed has none
+    #[test]
+    fn a_job_without_a_window_field_has_no_window() {
+        let mut s = Session::with_demo();
+        let id = s.visible_cloned()[0];
+        let wire = WireJob::from_job(&s.thumb_job(id, 128).unwrap(), false, "Thumb");
+        let mut v = serde_json::to_value(&wire).unwrap();
+        v.as_object_mut().unwrap().remove("window");
+        let old: WireJob = serde_json::from_value(v).unwrap();
+        assert_eq!(old.request().window, None);
+    }
+
    #[test]
    fn wire_job_renders_like_the_engine() {
        let mut s = Session::with_demo();
@@ -291,10 +413,13 @@ mod tests {
        let hash = crate::store::content_hash(&bytes);
        let origin = Source::File { path: crate::store::original_path(&hash, "a.png") };
        let job = WireJob {
+            cache_generation: 0,
+            source_identity: None,
            level: SourceLevel::Thumb,
            origin,
            max_edge: 512,
            raw: false,
+            relative_wb: false,
            as_shot_temp: 6500.0,
            as_shot_tint: 0.0,
            lens: None,
@@ -307,6 +432,7 @@ mod tests {
            thumb_cached: false,
            stages: Some("Main".into()),
            overlay: (0, 0.0),
+            window: None,
        };
        let mut core = WorkerCore::default();
        assert_eq!(core.needs_original(&job).as_deref(), Some(hash.as_str()));
@@ -314,6 +440,68 @@ mod tests {
        let r = core.render(&job, Some(&bytes)).unwrap();
        assert_eq!(r.image.width, 32);
        assert_eq!(core.needs_original(&job), None, "decoded source is kept");
+        let mut reloaded = job.clone();
+        reloaded.source_identity = Some("changed content".into());
+        assert_eq!(core.needs_original(&reloaded).as_deref(), Some(hash.as_str()), "content reload must not reuse an old decoded source");
+        let mut refreshed = job.clone();
+        refreshed.cache_generation = 1;
+        assert_eq!(core.needs_original(&refreshed).as_deref(), Some(hash.as_str()), "explicit refresh must not reuse the old decoded source");
+    }
+
+    /// The persisted index as a restarted page loads it: one stored thumbnail.
+    fn persisted_index(key: Hash128) -> (ThumbIndex, String) {
+        let mut index = ThumbIndex::default();
+        let hex = index.cache_key(key);
+        index.insert(&hex, 100);
+        (ThumbIndex::from_json(&index.to_json()), hex)
+    }
+
+    #[test]
+    fn clearing_previews_before_the_first_request_invalidates_stored_thumbnails() {
+        let key = lightcraft_preview::hash_bytes(b"thumbnail");
+        let (mut index, old) = persisted_index(key);
+        let cache = Arc::new(PreviewCache::memory(1 << 20));
+        let mut watch = CacheWatch::new(&cache);
+        let mut gone = Vec::new();
+        // each frame (`Workers::finished`) observes the watched cache, even with no requests
+        let mut frame = |watch: &mut CacheWatch, index: &mut ThumbIndex| {
+            if let Some(c) = watch.cache() {
+                gone.extend(watch.observe(&c, c.generation(), index));
+            }
+        };
+        frame(&mut watch, &mut index); // an empty filtered view: no render request yet
+        cache.clear(); // File ▸ Clear Preview Cache
+        frame(&mut watch, &mut index);
+        // the filter is removed: the first thumbnail request (`Workers::try_start`)
+        gone.extend(watch.observe(&cache, cache.generation(), &mut index));
+        let disk_key = index.cache_key(key);
+        assert!(!index.touch(&disk_key), "the pre-clear thumbnail must not be read back");
+        assert_eq!(gone, vec![old.clone()], "the stored file is deleted");
+        assert_ne!(disk_key, old, "a fresh namespace: a late old write can't come back");
+        assert!(index.dirty, "the invalidated index is saved");
+    }
+
+    #[test]
+    fn replacing_the_cache_before_the_first_request_invalidates_stored_thumbnails() {
+        let key = lightcraft_preview::hash_bytes(b"thumbnail");
+        let (mut index, old) = persisted_index(key);
+        let first = Arc::new(PreviewCache::memory(1 << 20));
+        let mut watch = CacheWatch::new(&first);
+        let other = Arc::new(PreviewCache::memory(1 << 20)); // another library, generation 0 again
+        assert_eq!(watch.observe(&other, other.generation(), &mut index), vec![old]);
+    }
+
+    #[test]
+    fn a_normal_start_keeps_stored_thumbnails() {
+        let key = lightcraft_preview::hash_bytes(b"thumbnail");
+        let (mut index, old) = persisted_index(key);
+        let cache = Arc::new(PreviewCache::memory(1 << 20));
+        let mut watch = CacheWatch::new(&cache);
+        for _ in 0..3 {
+            assert!(watch.observe(&cache, cache.generation(), &mut index).is_empty());
+        }
+        assert_eq!(index.cache_key(key), old);
+        assert!(index.touch(&old), "persisted thumbnails survive a restart");
    }

    #[test]
@@ -18,7 +18,7 @@ use std::sync::Arc;

 use js_sys::{Array, Object, Reflect, Uint8Array, Uint32Array};
 use lightcraft_engine::catalog::PhotoId;
-use lightcraft_engine::media::{DISK_CACHE_BYTES, RenderJob, RenderResult, SourceLevel};
+use lightcraft_engine::media::{RenderJob, RenderResult, SourceLevel};
 use lightcraft_engine::pipeline::Rendered;
 use lightcraft_preview::{Hash128, PreviewCache};
 use lightcraft_raster::{Histogram, Rgba8};
@@ -27,7 +27,7 @@ use wasm_bindgen::JsCast;
 use wasm_bindgen::prelude::*;

 use crate::backend::Backend;
-use crate::wire::{ThumbIndex, WireJob, WorkerCore, thumb_storage_key};
+use crate::wire::{CacheWatch, ThumbIndex, WireJob, WorkerCore, thumb_storage_key};

 /// Storage key of the thumbnail index.
 pub const THUMB_INDEX: &str = "thumbs/index.json";
@@ -155,6 +155,10 @@ async fn handle(scope: &web_sys::DedicatedWorkerGlobalScope, backend: Option<Bac
 // main side

 struct Busy {
+    disk_key: Option<String>,
+    namespace: u64,
+    request_id: u64,
+    cache_generation: u64,
    id: u32,
    slot: Slot,
    photo: PhotoId,
@@ -171,12 +175,17 @@ struct W {
 }

 struct Inner {
+    /// The rendered-thumbnail cache the stored thumbnails belong to.
+    preview_generation: CacheWatch,
    workers: Vec<W>,
    done: Vec<(Slot, RenderResult, f64)>,
    next_id: u32,
    /// (photo, level) → worker that last decoded it.
    affinity: HashMap<(PhotoId, SourceLevel), usize>,
    index: ThumbIndex,
+    /// The session's thumbnail cache budget in bytes. Kept here because the prune happens on a
+    /// worker message, where the session is out of reach; the app refreshes it every frame.
+    budget: u64,
    backend: Option<Backend>,
    ctx: egui::Context,
    /// Jobs finished by workers / inline (for diagnostics).
@@ -185,19 +194,46 @@ struct Inner {
    inline_this_frame: bool,
 }

+impl Inner {
+    fn observe_cache(&mut self, cache: &Arc<PreviewCache>, generation: u64) {
+        let gone = self.preview_generation.observe(cache, generation, &mut self.index);
+        if !gone.is_empty()
+            && let Some(backend) = self.backend.clone()
+        {
+            wasm_bindgen_futures::spawn_local(async move {
+                for key in gone {
+                    let _ = backend.remove(&thumb_storage_key(&key)).await;
+                }
+            });
+        }
+    }
+}
+
 /// The main thread's handle on the render workers.
 #[derive(Clone)]
 pub struct Workers(Rc<RefCell<Inner>>);

 impl Workers {
    /// Start `n` workers (none: every job runs inline). `store` is the backend kind for workers.
-    pub fn start(n: usize, store: &str, backend: Option<Backend>, index: ThumbIndex, ctx: egui::Context) -> Workers {
+    /// `index` lists the stored thumbnails of `cache`, the session's rendered-thumbnail cache
+    /// now: watching it from the start, a clear before the first render request still counts.
+    pub fn start(
+        n: usize,
+        store: &str,
+        backend: Option<Backend>,
+        index: ThumbIndex,
+        cache: &Arc<PreviewCache>,
+        budget: u64,
+        ctx: egui::Context,
+    ) -> Workers {
        let w = Workers(Rc::new(RefCell::new(Inner {
+            preview_generation: CacheWatch::new(cache),
            workers: Vec::new(),
            done: Vec::new(),
            next_id: 0,
            affinity: HashMap::new(),
            index,
+            budget,
            backend,
            ctx,
            remote_done: 0,
@@ -248,7 +284,16 @@ impl Workers {
            w.dead = true;
            w.worker.terminate();
            if let Some(b) = w.busy.take() {
-                let r = RenderResult { photo: b.photo, level: b.level, key: b.key, rendered: Err(msg), loaded: None, quick: None };
+                let r = RenderResult {
+                    request_id: b.request_id,
+                    source_key: None,
+                    photo: b.photo,
+                    level: b.level,
+                    key: b.key,
+                    rendered: Err(msg),
+                    loaded: None,
+                    quick: None,
+                };
                g.done.push((b.slot, r, 0.0));
            }
        }
@@ -284,16 +329,21 @@ impl Workers {
        };
        if let Ok(r) = &rendered
            && let Some((cache, key)) = &b.cache
+            && cache.generation() == b.cache_generation
+            && b.namespace == g.index.namespace()
        {
-            cache.put(*key, Arc::new(r.image.clone()));
-            let hex = key.to_string();
+            cache.put_at(b.cache_generation, *key, Arc::new(r.image.clone()));
+            let hex = b.disk_key.clone().unwrap_or_else(|| key.to_string());
            let stored = get(&data, "stored").as_f64().unwrap_or(0.0) as u64;
            if get(&data, "cacheMiss").is_truthy() {
                g.index.remove(&hex);
            }
            if stored > 0 {
                g.index.insert(&hex, stored);
-                let gone = g.index.prune(DISK_CACHE_BYTES);
+                // The session's own budget (`library.preferences ▸ cacheMb`), not the native
+                // default: the browser accepted the setting, so it has to be the one enforced
+                // (#245).
+                let gone = g.index.prune(g.budget);
                if !gone.is_empty()
                    && let Some(be) = g.backend.clone()
                {
@@ -307,8 +357,31 @@ impl Workers {
            }
        }
        g.affinity.insert((b.photo, b.level), i);
+        // A cleared worker may finish writing after the initial deletion. Its old namespace
+        // is never reused, so removing this file cannot delete a replacement's cache entry.
+        if let Some((cache, _)) = &b.cache
+            && (cache.generation() != b.cache_generation || b.namespace != g.index.namespace())
+            && let (Some(backend), Some(key)) = (g.backend.clone(), b.disk_key.clone())
+        {
+            wasm_bindgen_futures::spawn_local(async move {
+                let _ = backend.remove(&thumb_storage_key(&key)).await;
+            });
+        }
        g.remote_done += 1;
-        g.done.push((b.slot, RenderResult { photo: b.photo, level: b.level, key: b.key, rendered, loaded: None, quick: None }, ms));
+        g.done.push((
+            b.slot,
+            RenderResult {
+                request_id: b.request_id,
+                source_key: None,
+                photo: b.photo,
+                level: b.level,
+                key: b.key,
+                rendered,
+                loaded: None,
+                quick: None,
+            },
+            ms,
+        ));
    }

    /// The thumbnail index as JSON if it changed since the last call (the host saves it).
@@ -328,19 +401,30 @@ impl Workers {
        let ready = g.workers.iter().filter(|w| w.ready && !w.dead).count();
        (alive, ready, g.remote_done, g.inline_done)
    }
+
+    /// The budget the thumbnail index is pruned to. The app refreshes it from the session every
+    /// frame, so `library.preferences ▸ cacheMb` applies to the next stored thumbnail.
+    pub fn set_budget(&self, bytes: u64) {
+        self.0.borrow_mut().budget = bytes;
+    }
 }

 impl RenderOffload for Workers {
    fn try_start(&mut self, slot: Slot, job: RenderJob) -> Option<RenderJob> {
        let mut g = self.0.borrow_mut();
        let g = &mut *g;
+        if let Some((cache, _)) = job.cache.as_ref().or(job.view_cache.as_ref()) {
+            g.observe_cache(cache, job.cache_generation);
+        }
        // rendered thumbnail in memory: no worker needed
        if let Some((cache, key)) = &job.cache
-            && let Some(img) = cache.get(*key)
+            && let Some(img) = cache.get_at(job.cache_generation, *key)
        {
            let image = Arc::unwrap_or_clone(img);
            let histogram = Histogram::of_srgb8(&image);
            let r = RenderResult {
+                request_id: job.request_id,
+                source_key: job.source_key,
                photo: job.photo,
                level: job.level,
                key: job.key,
@@ -374,25 +458,45 @@ impl RenderOffload for Workers {
        let Some(i) = pick else { return Some(job) };
        // decided here, so the next job for this photo follows even before this one finishes
        g.affinity.insert((job.photo, job.level), i);
-        let thumb_cached = job.cache.as_ref().is_some_and(|(_, k)| g.index.touch(&k.to_string()));
-        let wire = WireJob::from_job(&job, thumb_cached, &format!("{slot:?}"));
+        let disk_key = job.cache.as_ref().map(|(_, key)| g.index.cache_key(*key));
+        let thumb_cached = disk_key.as_ref().is_some_and(|key| g.index.touch(key));
+        let mut wire = WireJob::from_job(&job, thumb_cached, &format!("{slot:?}"));
+        // The namespace also changes when a new library's cache starts again at generation zero.
+        wire.cache_generation = g.index.namespace();
+        wire.thumb = disk_key.clone();
        let Ok(text) = serde_json::to_string(&wire) else { return Some(job) };
        g.next_id = g.next_id.wrapping_add(1);
        let id = g.next_id;
        let msg = Object::new();
        set(&msg, "id", &id.into());
        set(&msg, "job", &text.into());
+        let namespace = g.index.namespace();
        let w = &mut g.workers[i];
        if let Err(e) = w.worker.post_message(&msg) {
            log::warn!("render worker {i}: {e:?}");
            return Some(job);
        }
-        w.busy = Some(Busy { id, slot, photo: job.photo, level: job.level, key: job.key, cache: job.cache });
+        w.busy = Some(Busy {
+            disk_key,
+            namespace,
+            request_id: job.request_id,
+            cache_generation: job.cache_generation,
+            id,
+            slot,
+            photo: job.photo,
+            level: job.level,
+            key: job.key,
+            cache: job.cache,
+        });
        None
    }

    fn finished(&mut self) -> Vec<(Slot, RenderResult, f64)> {
        let mut g = self.0.borrow_mut();
+        // Persist invalidation even when the cleared library has no visible render requests.
+        if let Some(cache) = g.preview_generation.cache() {
+            g.observe_cache(&cache, cache.generation());
+        }
        g.inline_this_frame = false;
        std::mem::take(&mut g.done)
    }
@@ -9,8 +9,17 @@ default-run = "lightcraft"
 [lints]
 workspace = true

+[features]
+default = ["denoise"]
+denoise = ["lightcraft-engine/denoise"]
+rawnind-model = ["denoise", "lightcraft-engine/rawnind-model"]
+# Open HEIC/HEIF photos (the optional lightcraft-heif crate). Not default so distributors choose
+# it (HEVC patents); build with `--features heif` to include it.
+heif = ["lightcraft-codecs/heif"]
+
 [dependencies]
-lightcraft-engine = { workspace = true }
+lightcraft-engine = { workspace = true, features = ["sam"] }
+lightcraft-sysmem = { workspace = true }
 lightcraft-ui-egui = { workspace = true }
 lightcraft-codecs = { workspace = true }
 lightcraft-raster = { workspace = true }
@@ -1,32 +1,12 @@
 //! Give memory the system allocator keeps after frees back to the system
-//! ([`lightcraft_engine::memory::set_release_hook`]).
-//!
-//! macOS's allocator keeps freed large blocks mapped as "reusable": they no longer count in the
-//! process's memory footprint but stay in its resident size until the system runs short. After a
-//! raw decode (hundreds of MB of temporaries) that made the resident size climb towards the sum of
-//! everything ever decoded. `malloc_zone_pressure_relief` (libSystem) returns those pages now.
-//! Elsewhere (glibc, Windows) blocks this large are unmapped when freed: nothing to do.
+//! ([`lightcraft_engine::memory::set_release_hook`]). The platform call lives in
+//! `lightcraft-sysmem`, the workspace's only crate allowed `unsafe`.

-/// Install the hook for this platform.
+/// Install the hook.
 pub fn install() {
-    #[cfg(target_os = "macos")]
-    lightcraft_engine::memory::set_release_hook(macos::release);
+    lightcraft_engine::memory::set_release_hook(release);
 }

-#[cfg(target_os = "macos")]
-#[allow(unsafe_code)]
-mod macos {
-    unsafe extern "C" {
-        /// libSystem malloc: release free memory of `zone` (all zones when null) back to the
-        /// system, up to `goal` bytes (0 = as much as possible). Returns the bytes released.
-        fn malloc_zone_pressure_relief(zone: *mut std::ffi::c_void, goal: usize) -> usize;
-    }
-
-    pub fn release() {
-        // SAFETY: a documented libSystem function taking a null zone (= every zone) and a size;
-        // it only returns free pages to the system and touches no memory we own.
-        unsafe {
-            malloc_zone_pressure_relief(std::ptr::null_mut(), 0);
-        }
-    }
+fn release() {
+    let _ = lightcraft_sysmem::release_free_memory();
 }
@@ -1,62 +1,267 @@
 //! Loopback JSON-lines control server: one request per line, one reply per line.
 //! This is the transport the MCP server (`lightcraft-cli mcp --connect`) wraps.
+//!
+//! The server only ever reads requests: a line that is not a JSON object with a string `method`
+//! (an HTTP request line from a browser's cross-origin `fetch`, a stray `nc`, binary junk) gets
+//! one error reply and the connection is closed, so nothing after it is executed. Lines are
+//! bounded ([`MAX_LINE`]) and so is the number of open connections ([`MAX_CONNECTIONS`]); junk
+//! never reaches the UI thread. See `docs/control-protocol.md`.

-use std::io::{BufRead, BufReader, Write};
+use std::io::{BufRead, BufReader, Read, Write};
 use std::net::{TcpListener, TcpStream};
+use std::sync::Arc;
+use std::sync::atomic::{AtomicUsize, Ordering};
 use std::sync::mpsc::{Receiver, Sender, channel};
 use std::time::Duration;

 use lightcraft_ui_egui::ControlRequest;
 use serde_json::{Value, json};

+/// Longest accepted request line (bytes, without the newline). Requests are small JSON objects;
+/// anything longer is refused and the connection closed.
+pub const MAX_LINE: usize = 4 * 1024 * 1024;
+/// Connections served at once; further ones get an error line and are closed.
+pub const MAX_CONNECTIONS: usize = 16;
+
 pub fn start(port: u16, ctx: egui::Context) -> Receiver<ControlRequest> {
    let (tx, rx) = channel::<ControlRequest>();
    let listener = match TcpListener::bind(("127.0.0.1", port)) {
        Ok(l) => l,
        Err(e) => {
-            eprintln!("lightcraft: control server failed to bind 127.0.0.1:{port}: {e}");
+            log::error!("control server failed to bind 127.0.0.1:{port}: {e}");
            return rx;
        }
    };
-    eprintln!("lightcraft: control server listening on 127.0.0.1:{port}");
-    std::thread::spawn(move || {
-        for stream in listener.incoming().flatten() {
-            let tx = tx.clone();
-            let ctx = ctx.clone();
-            std::thread::spawn(move || serve(stream, tx, ctx));
-        }
-    });
+    log::info!("control server listening on 127.0.0.1:{port}");
+    std::thread::spawn(move || accept_loop(listener, tx, ctx));
    rx
 }

-fn serve(stream: TcpStream, tx: Sender<ControlRequest>, ctx: egui::Context) {
-    let Ok(read) = stream.try_clone() else { return };
-    let mut out = stream;
-    for line in BufReader::new(read).lines() {
-        let Ok(line) = line else { break };
-        if line.trim().is_empty() {
+fn accept_loop(listener: TcpListener, tx: Sender<ControlRequest>, ctx: egui::Context) {
+    let open = Arc::new(AtomicUsize::new(0));
+    for mut stream in listener.incoming().flatten() {
+        if open.fetch_add(1, Ordering::SeqCst) >= MAX_CONNECTIONS {
+            open.fetch_sub(1, Ordering::SeqCst);
+            let _ = writeln!(stream, "{}", json!({"ok": false, "error": "too many control connections"}));
            continue;
        }
-        let reply = match serde_json::from_str::<Value>(&line) {
-            Ok(msg) => {
-                let id = msg.get("id").cloned().unwrap_or(Value::Null);
-                let method = msg.get("method").and_then(Value::as_str).unwrap_or("").to_string();
-                let params = msg.get("params").cloned().unwrap_or(json!({}));
-                let (req, rrx) = ControlRequest::new(method, params);
-                if tx.send(req).is_err() {
-                    break;
-                }
-                ctx.request_repaint();
-                let mut r = rrx.recv_timeout(Duration::from_secs(60)).unwrap_or_else(|_| json!({"ok": false, "error": "timeout"}));
-                if let Some(o) = r.as_object_mut() {
-                    o.insert("id".into(), id);
+        let tx = tx.clone();
+        let ctx = ctx.clone();
+        let open = Arc::clone(&open);
+        std::thread::spawn(move || {
+            serve(stream, &tx, &ctx);
+            open.fetch_sub(1, Ordering::SeqCst);
+        });
+    }
+}
+
+fn serve(stream: TcpStream, tx: &Sender<ControlRequest>, ctx: &egui::Context) {
+    let Ok(read) = stream.try_clone() else { return };
+    serve_lines(read, stream, |method, params| {
+        let (req, rrx) = ControlRequest::new(method, params);
+        tx.send(req).ok()?;
+        ctx.request_repaint();
+        Some(rrx.recv_timeout(Duration::from_secs(60)).unwrap_or_else(|_| json!({"ok": false, "error": "timeout"})))
+    });
+}
+
+/// One parsed line.
+enum Line {
+    Request {
+        id: Value,
+        method: String,
+        params: Value,
+    },
+    Blank,
+    /// Not a request: reply with this error and close the connection.
+    Reject(String),
+}
+
+fn parse_line(bytes: &[u8]) -> Line {
+    let Ok(text) = std::str::from_utf8(bytes) else {
+        return Line::Reject("not a control request (invalid UTF-8); closing the connection".into());
+    };
+    let text = text.trim();
+    if text.is_empty() {
+        return Line::Blank;
+    }
+    let msg = match serde_json::from_str::<Value>(text) {
+        Ok(v) => v,
+        Err(e) => return Line::Reject(format!("bad JSON ({e}); expected one JSON request object per line; closing the connection")),
+    };
+    let Some(obj) = msg.as_object() else {
+        return Line::Reject("not a control request (expected a JSON object); closing the connection".into());
+    };
+    let Some(method) = obj.get("method").and_then(Value::as_str) else {
+        return Line::Reject("not a control request (missing string `method`); closing the connection".into());
+    };
+    Line::Request {
+        id: obj.get("id").cloned().unwrap_or(Value::Null),
+        method: method.to_string(),
+        params: obj.get("params").cloned().unwrap_or(json!({})),
+    }
+}
+
+/// Read one `\n`-terminated line of at most `max` bytes into `buf` (newline stripped).
+/// `Ok(false)` at end of stream; `Err` when the line is too long or the read fails.
+fn read_bounded_line(reader: &mut impl BufRead, buf: &mut Vec<u8>, max: usize) -> Result<bool, String> {
+    buf.clear();
+    // `take` bounds the bytes buffered for one line: a peer can't make us allocate more than `max`.
+    let limit = u64::try_from(max).unwrap_or(u64::MAX).saturating_add(1);
+    let n = reader.take(limit).read_until(b'\n', buf).map_err(|e| e.to_string())?;
+    if n == 0 {
+        return Ok(false);
+    }
+    if buf.last() == Some(&b'\n') {
+        buf.pop();
+        if buf.last() == Some(&b'\r') {
+            buf.pop();
+        }
+        return Ok(true);
+    }
+    if buf.len() > max {
+        return Err(format!("request line longer than {max} bytes; closing the connection"));
+    }
+    // Last line without a newline, then end of stream.
+    Ok(true)
+}
+
+/// Serve requests from `read`, answering on `out`, until the peer closes, sends something that is
+/// not a request, or `handle` returns `None` (app gone). `handle` gets `(method, params)` and
+/// returns the reply object (the request's `id` is added here).
+fn serve_lines<R: Read, W: Write>(read: R, mut out: W, mut handle: impl FnMut(String, Value) -> Option<Value>) {
+    let mut reader = BufReader::new(read);
+    let mut buf = Vec::new();
+    loop {
+        let reject = match read_bounded_line(&mut reader, &mut buf, MAX_LINE) {
+            Ok(false) => return,
+            Err(e) => e,
+            Ok(true) => match parse_line(&buf) {
+                Line::Blank => continue,
+                Line::Reject(e) => e,
+                Line::Request { id, method, params } => {
+                    let Some(mut r) = handle(method, params) else { return };
+                    if let Some(o) = r.as_object_mut() {
+                        o.insert("id".into(), id);
+                    }
+                    if writeln!(out, "{r}").and_then(|()| out.flush()).is_err() {
+                        return;
+                    }
+                    continue;
                }
-                r
-            }
-            Err(e) => json!({"ok": false, "error": format!("bad JSON: {e}")}),
+            },
        };
-        if writeln!(out, "{reply}").is_err() {
-            break;
+        let _ = writeln!(out, "{}", json!({"ok": false, "error": reject})).and_then(|()| out.flush());
+        return;
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    fn run(input: &[u8]) -> (Vec<Value>, Vec<String>) {
+        let mut out = Vec::new();
+        let mut calls = Vec::new();
+        serve_lines(input, &mut out, |m, _p| {
+            calls.push(m);
+            Some(json!({"ok": true, "result": null}))
+        });
+        let replies = String::from_utf8(out).unwrap().lines().map(|l| serde_json::from_str::<Value>(l).unwrap()).collect();
+        (replies, calls)
+    }
+
+    #[test]
+    fn requests_are_answered_in_order_with_ids() {
+        let (r, calls) = run(b"{\"id\":1,\"method\":\"a\"}\n\n{\"id\":2,\"method\":\"b\",\"params\":{}}\r\n{\"method\":\"c\"}");
+        assert_eq!(calls, ["a", "b", "c"]);
+        assert_eq!(r.len(), 3);
+        assert_eq!(r[0]["id"], 1);
+        assert_eq!(r[1]["id"], 2);
+        assert_eq!(r[2]["id"], Value::Null);
+    }
+
+    #[test]
+    fn http_request_closes_before_the_body_runs() {
+        // A browser's cross-origin `fetch` POST: the request line is rejected and the connection
+        // closed, so the JSON body is never executed.
+        let http = b"POST / HTTP/1.1\r\nHost: 127.0.0.1:7980\r\nContent-Type: text/plain\r\n\r\n\
+            {\"id\":1,\"method\":\"engine.execute\",\"params\":{\"command\":\"folder.move\"}}\n";
+        let (r, calls) = run(http);
+        assert!(calls.is_empty(), "nothing may run: {calls:?}");
+        assert_eq!(r.len(), 1);
+        assert_eq!(r[0]["ok"], false);
+        assert!(r[0]["error"].as_str().unwrap().contains("closing the connection"));
+    }
+
+    #[test]
+    fn non_request_json_closes_the_connection() {
+        for junk in [
+            &b"[1,2]\n{\"method\":\"a\"}\n"[..],
+            b"{\"id\":1}\n{\"method\":\"a\"}\n",
+            b"{\"method\":5}\n{\"method\":\"a\"}\n",
+            b"\xff\xfe\n{\"method\":\"a\"}\n",
+        ] {
+            let (r, calls) = run(junk);
+            assert!(calls.is_empty());
+            assert_eq!(r.len(), 1);
+            assert_eq!(r[0]["ok"], false);
        }
+    }
+
+    #[test]
+    fn a_request_before_junk_still_runs() {
+        let (r, calls) = run(b"{\"method\":\"a\"}\nGET / HTTP/1.1\n{\"method\":\"b\"}\n");
+        assert_eq!(calls, ["a"]);
+        assert_eq!(r.len(), 2);
+        assert_eq!(r[1]["ok"], false);
+    }
+
+    #[test]
+    fn overlong_line_is_refused_without_buffering_it_all() {
+        // An endless line without a newline: reading stops after MAX_LINE + 1 bytes.
+        let endless = std::io::repeat(b'x');
+        let mut out = Vec::new();
+        let mut ran = false;
+        serve_lines(endless, &mut out, |_, _| {
+            ran = true;
+            None
+        });
+        assert!(!ran);
+        let reply: Value = serde_json::from_slice(out.strip_suffix(b"\n").unwrap()).unwrap();
+        assert!(reply["error"].as_str().unwrap().contains("longer than"));
+    }
+
+    #[test]
+    fn line_at_the_limit_is_accepted() {
+        let mut buf = Vec::new();
+        let line = vec![b'a'; 10];
+        let mut input = line.clone();
+        input.push(b'\n');
+        assert_eq!(read_bounded_line(&mut &input[..], &mut buf, 10), Ok(true));
+        assert_eq!(buf, line);
+        assert!(read_bounded_line(&mut &[b'a'; 11][..], &mut buf, 10).is_err());
+    }
+
+    #[test]
+    fn tcp_connection_is_closed_after_junk() {
+        // End to end over a real loopback socket (the transport the app uses).
+        let listener = TcpListener::bind(("127.0.0.1", 0)).unwrap();
+        let addr = listener.local_addr().unwrap();
+        let server = std::thread::spawn(move || {
+            let (s, _) = listener.accept().unwrap();
+            let read = s.try_clone().unwrap();
+            serve_lines(read, s, |_, _| Some(json!({"ok": true})));
+        });
+        let mut c = TcpStream::connect(addr).unwrap();
+        c.set_read_timeout(Some(Duration::from_secs(10))).unwrap();
+        c.write_all(b"{\"id\":7,\"method\":\"x\"}\nhello\n{\"method\":\"y\"}\n").unwrap();
+        let mut text = String::new();
+        c.read_to_string(&mut text).unwrap(); // returns at EOF: the server closed the connection
+        server.join().unwrap();
+        let lines: Vec<&str> = text.lines().collect();
+        assert_eq!(lines.len(), 2, "{text}");
+        assert!(lines[0].contains("\"id\":7"));
+        assert!(lines[1].contains("closing the connection"));
    }
 }
@@ -0,0 +1,145 @@
+//! File-dialog filters that match extensions whatever their letter case.
+//!
+//! On Linux, `rfd` turns each extension into a `*.ext` glob for the GTK dialog and for the
+//! desktop portal (`xdg-desktop-portal` FileChooser), and both match globs case-sensitively, so
+//! a filter of `nef` hid `DSC_1234.NEF` (issue #342). Windows and macOS ignore case. Every
+//! dialog filter of the app goes through [`FileDialogExt::add_filter_nocase`] (a test checks
+//! that nothing else calls `add_filter`), so no extension can hit this.
+
+use rfd::FileDialog;
+
+/// The patterns to hand to `rfd` for `exts`, which are plain extensions without the dot.
+///
+/// Windows and macOS take the extension literally and ignore case: lower case, de-duplicated.
+/// Linux (`linux`): `rfd` formats each entry as `*.{entry}`, which GTK 3 and the portal match
+/// as a glob, so an entry may carry glob syntax. Each extension gives the lower-case form, the
+/// UPPER-case form (kept in case a portal backend does not do `[...]` classes) and a bracket
+/// form such as `[nN][eE][fF]` that matches every mix (`.Nef`, `.nEf`, ...).
+pub fn dialog_extensions_for(exts: &[&str], linux: bool) -> Vec<String> {
+    let mut out: Vec<String> = Vec::new();
+    let mut push = |s: String| {
+        if !s.is_empty() && !out.contains(&s) {
+            out.push(s);
+        }
+    };
+    for ext in exts {
+        let ext = ext.trim_start_matches('.');
+        let lower = ext.to_lowercase();
+        push(lower.clone());
+        if !linux {
+            continue;
+        }
+        push(ext.to_uppercase());
+        // `[` `]` `*` `?` in an extension would already be glob syntax: leave those alone
+        if lower.chars().any(|c| c.is_alphabetic()) && !lower.contains(['[', ']', '*', '?', '\\']) {
+            push(
+                lower
+                    .chars()
+                    .map(|c| {
+                        let up: String = c.to_uppercase().collect();
+                        if up.chars().count() == 1 && up != c.to_string() { format!("[{c}{up}]") } else { c.to_string() }
+                    })
+                    .collect(),
+            );
+        }
+    }
+    out
+}
+
+/// [`dialog_extensions_for`] for the platform this is built for.
+pub fn dialog_extensions(exts: &[&str]) -> Vec<String> {
+    dialog_extensions_for(exts, cfg!(all(unix, not(target_os = "macos"))))
+}
+
+/// `add_filter` that matches the extensions in any letter case.
+pub trait FileDialogExt {
+    fn add_filter_nocase(self, name: impl Into<String>, exts: &[&str]) -> Self;
+}
+
+impl FileDialogExt for FileDialog {
+    fn add_filter_nocase(self, name: impl Into<String>, exts: &[&str]) -> Self {
+        self.add_filter(name, &dialog_extensions(exts))
+    }
+}
+
+#[cfg(test)]
+#[allow(clippy::unwrap_used)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn other_platforms_get_lower_case_once() {
+        assert_eq!(dialog_extensions_for(&["NEF", "nef", ".Jpg", "jpg", ""], false), ["nef", "jpg"]);
+    }
+
+    #[test]
+    fn linux_gets_lower_upper_and_any_mix() {
+        assert_eq!(dialog_extensions_for(&["nef"], true), ["nef", "NEF", "[nN][eE][fF]"]);
+        // digits have no case; a duplicate adds nothing
+        assert_eq!(dialog_extensions_for(&["mp4", "MP4"], true), ["mp4", "MP4", "[mM][pP]4"]);
+        assert_eq!(dialog_extensions_for(&["7z"], true), ["7z", "7Z", "7[zZ]"]);
+        assert_eq!(dialog_extensions_for(&["123"], true), ["123"]);
+    }
+
+    #[test]
+    fn linux_glob_syntax_is_not_wrapped_again() {
+        assert_eq!(dialog_extensions_for(&["[nN]ef"], true), ["[nn]ef", "[NN]EF"]);
+    }
+
+    /// The Linux patterns match like GTK's glob (`*.{entry}`): checked against a tiny matcher of
+    /// the subset used here (verified against real GTK 3.24 `GtkFileFilter` in the PR notes).
+    fn glob_ext(entry: &str, ext: &str) -> bool {
+        let mut pat = entry.chars().peekable();
+        let mut s = ext.chars();
+        while let Some(p) = pat.next() {
+            let Some(c) = s.next() else { return false };
+            if p == '[' {
+                let mut hit = false;
+                for q in pat.by_ref() {
+                    if q == ']' {
+                        break;
+                    }
+                    hit |= q == c;
+                }
+                if !hit {
+                    return false;
+                }
+            } else if p != c {
+                return false;
+            }
+        }
+        s.next().is_none()
+    }
+
+    #[test]
+    fn linux_patterns_match_every_case_mix() {
+        let pats = dialog_extensions_for(&["nef", "cr2", "xmp"], true);
+        let matches = |name: &str| pats.iter().any(|p| glob_ext(p, name));
+        for ok in ["nef", "NEF", "Nef", "nEF", "CR2", "Cr2", "XMP", "xMp"] {
+            assert!(matches(ok), "{ok}");
+        }
+        for bad in ["nex", "ne", "nefs", "jpg", "cr3"] {
+            assert!(!matches(bad), "{bad}");
+        }
+    }
+
+    /// The app's own extension list (the one the folder scan uses) is covered whole.
+    #[test]
+    fn import_extensions_are_all_covered_in_both_cases() {
+        let pats = dialog_extensions_for(lightcraft_engine::import::EXTENSIONS, true);
+        for e in lightcraft_engine::import::EXTENSIONS {
+            assert!(pats.iter().any(|p| glob_ext(p, &e.to_uppercase())), "{e}");
+            assert!(pats.iter().any(|p| p == e), "{e}");
+        }
+    }
+
+    /// Every dialog filter goes through the helper: `main.rs` must not call rfd's own
+    /// `add_filter`.
+    #[test]
+    fn no_dialog_bypasses_the_helper() {
+        for (file, src) in [("main.rs", include_str!("main.rs")), ("control_server.rs", include_str!("control_server.rs"))] {
+            assert!(!src.contains(".add_filter("), "{file} calls rfd's add_filter; use add_filter_nocase");
+        }
+        assert!(include_str!("main.rs").contains(".add_filter_nocase("));
+    }
+}
@@ -0,0 +1,689 @@
+//! The desktop app's logger: `log::` records go to standard error and to a log file.
+//!
+//! It replaces the stderr-only `StderrLog` that `main.rs` used to install. A launch from a
+//! desktop menu or the Dock has no terminal, so the file is what a bug report can attach:
+//! `<settings folder>/logs/lightcraft.log`, next to `ui.json` (`config_dir` in `main.rs`; Linux
+//! `$XDG_CONFIG_HOME/lightcraft/logs`, by default `~/.config/lightcraft/logs`). Never in the
+//! library folder. Each start moves the previous log to `lightcraft.1.log` (and that one to `.2`),
+//! so the log of a run that crashed survives the next launch. Runs with `LIGHTCRAFT_NO_PREFS`
+//! (tests, scripts) log to standard error only, so they don't rotate away the user's own logs.
+//!
+//! Levels ([`filter_spec`]): `info` for LightCraft's own crates, `warn` for everything else (wgpu
+//! and naga are chatty). `LIGHTCRAFT_LOG` keeps the meaning it had with the old logger: `info` or
+//! `debug` lowers LightCraft's own crates to that level, any other value means warnings and errors
+//! only. Without it, `RUST_LOG` replaces the default with env_logger-style directives: `debug`,
+//! `warn,lightcraft_pipeline=trace`, `wgpu_core=info`. A directive ending in `*` matches every
+//! target that starts with it (`lightcraft*=debug`).
+//!
+//! Records logged before the settings folder is known are kept (up to [`MAX_PENDING`]) and
+//! written once the file is attached. Writing never panics: a file that can't be created or
+//! written leaves standard error as the only sink. Panics that escape everything are added to the
+//! file by [`AppLogger::record_panic`] (the default hook already printed them on standard error).
+//!
+//! This is the desktop app's logger only; `lightcraft-cli` doesn't share a process with it.
+
+use std::fs::File;
+use std::io::Write;
+use std::path::{Path, PathBuf};
+use std::sync::{Mutex, OnceLock, PoisonError, TryLockError};
+use std::time::SystemTime;
+
+use log::LevelFilter;
+
+/// The current log file's name inside the log folder.
+pub const LOG_FILE: &str = "lightcraft.log";
+/// How many previous logs are kept (`lightcraft.1.log` … `lightcraft.<KEEP>.log`).
+pub const KEEP: usize = 2;
+/// The log file stops growing past this size (a runaway warning can't fill the disk).
+pub const MAX_FILE_BYTES: u64 = 16 * 1024 * 1024;
+/// Records kept in memory until the log file is attached.
+pub const MAX_PENDING: usize = 512;
+/// The built-in filter when neither `LIGHTCRAFT_LOG` nor `RUST_LOG` says otherwise.
+pub const DEFAULT_FILTER: &str = "warn,lightcraft*=info";
+
+/// The filter directives for this run from `LIGHTCRAFT_LOG` and `RUST_LOG` (blank = unset).
+///
+/// `LIGHTCRAFT_LOG` wins when set and means what it did for the old stderr logger: `info` or
+/// `debug` for LightCraft's own crates (warnings and errors from everything else), any other value
+/// warnings and errors only. Else `RUST_LOG` as env_logger directives, else [`DEFAULT_FILTER`].
+pub fn filter_spec(lightcraft_log: Option<&str>, rust_log: Option<&str>) -> String {
+    let set = |v: Option<&str>| v.map(str::trim).filter(|v| !v.is_empty()).map(str::to_owned);
+    if let Some(level) = set(lightcraft_log) {
+        return match level.as_str() {
+            "debug" => "warn,lightcraft*=debug".to_owned(),
+            "info" => "warn,lightcraft*=info".to_owned(),
+            _ => "warn".to_owned(),
+        };
+    }
+    set(rust_log).unwrap_or_else(|| DEFAULT_FILTER.to_owned())
+}
+
+/// Per-target level filter parsed from env_logger-style directives.
+#[derive(Clone, Debug, PartialEq, Eq)]
+pub struct Filter {
+    default: LevelFilter,
+    directives: Vec<(String, LevelFilter)>,
+}
+
+impl Filter {
+    /// Parse `spec`; unknown levels and empty parts are skipped. Without a bare level, targets no
+    /// directive names log errors only.
+    pub fn parse(spec: &str) -> Filter {
+        let mut f = Filter { default: LevelFilter::Error, directives: Vec::new() };
+        for part in spec.split(',').map(str::trim).filter(|p| !p.is_empty()) {
+            match part.split_once('=') {
+                Some((name, level)) => {
+                    let name = name.trim();
+                    if let (false, Some(level)) = (name.is_empty(), parse_level(level)) {
+                        f.directives.push((name.to_owned(), level));
+                    }
+                }
+                None => match parse_level(part) {
+                    Some(level) => f.default = level,
+                    // A bare target name: everything from it (env_logger does the same).
+                    None => f.directives.push((part.to_owned(), LevelFilter::Trace)),
+                },
+            }
+        }
+        f
+    }
+
+    /// The level that applies to `target` (the most specific matching directive wins).
+    pub fn level_for(&self, target: &str) -> LevelFilter {
+        self.directives.iter().filter(|(name, _)| matches(name, target)).max_by_key(|(name, _)| name.len()).map_or(self.default, |(_, level)| *level)
+    }
+
+    /// The most verbose level any target can reach (for `log::set_max_level`).
+    pub fn max(&self) -> LevelFilter {
+        self.directives.iter().map(|(_, level)| *level).fold(self.default, Ord::max)
+    }
+}
+
+fn parse_level(s: &str) -> Option<LevelFilter> {
+    s.trim().parse().ok()
+}
+
+/// `name` matches `target` itself and its submodules (`a` matches `a` and `a::b`, not `ab`);
+/// `name*` matches every target starting with `name`.
+fn matches(name: &str, target: &str) -> bool {
+    match name.strip_suffix('*') {
+        Some(prefix) => target.starts_with(prefix),
+        None => target.strip_prefix(name).is_some_and(|rest| rest.is_empty() || rest.starts_with("::")),
+    }
+}
+
+/// `2026-10-08T07:59:17.728Z` (UTC, milliseconds). A clock before 1970 reads as the epoch.
+pub fn timestamp(t: SystemTime) -> String {
+    let since = t.duration_since(std::time::UNIX_EPOCH).unwrap_or_default();
+    let secs = since.as_secs();
+    let (days, rem) = (secs / 86_400, secs % 86_400);
+    let (y, m, d) = civil_from_days(days);
+    format!("{y:04}-{m:02}-{d:02}T{:02}:{:02}:{:02}.{:03}Z", rem / 3600, rem % 3600 / 60, rem % 60, since.subsec_millis())
+}
+
+/// Days since 1970-01-01 to a proleptic Gregorian (year, month, day) (Howard Hinnant's
+/// `civil_from_days`, unsigned because `timestamp` clamps at the epoch). Saturating, so no clock
+/// value can overflow it.
+fn civil_from_days(days: u64) -> (u64, u64, u64) {
+    let z = days.saturating_add(719_468);
+    let era = z / 146_097;
+    let doe = z % 146_097;
+    let yoe = (doe - doe / 1460 + doe / 36_524 - doe / 146_096) / 365;
+    let doy = doe - (365 * yoe + yoe / 4 - yoe / 100);
+    let mp = (5 * doy + 2) / 153;
+    let d = doy - (153 * mp + 2) / 5 + 1;
+    let m = if mp < 10 { mp + 3 } else { mp - 9 };
+    let y = era.saturating_mul(400).saturating_add(yoe).saturating_add(u64::from(m <= 2));
+    (y, m, d)
+}
+
+fn numbered(dir: &Path, n: usize) -> PathBuf {
+    if n == 0 { dir.join(LOG_FILE) } else { dir.join(format!("lightcraft.{n}.log")) }
+}
+
+/// Shift the previous logs up one (`.log` → `.1.log` → … → `.<KEEP>.log`, the oldest dropped)
+/// and create a fresh, empty log file in `dir` (created if missing).
+pub fn rotate(dir: &Path) -> std::io::Result<(PathBuf, File)> {
+    std::fs::create_dir_all(dir)?;
+    for n in (1..=KEEP).rev() {
+        match std::fs::rename(numbered(dir, n - 1), numbered(dir, n)) {
+            Err(e) if e.kind() != std::io::ErrorKind::NotFound => return Err(e),
+            _ => {}
+        }
+    }
+    let path = numbered(dir, 0);
+    let file = File::create(&path)?;
+    Ok((path, file))
+}
+
+/// One formatted log line: `<timestamp> <LEVEL> [<thread>] <target>: <message>\n`.
+pub fn format_line(ts: &str, level: log::Level, thread: &str, target: &str, message: &str) -> String {
+    let message = message.strip_suffix('\n').unwrap_or(message);
+    format!("{ts} {level:<5} [{thread}] {target}: {message}\n")
+}
+
+/// Where formatted lines go besides standard error: kept in memory until a file is attached,
+/// then appended to it until it reaches its size cap.
+pub struct Sink {
+    file: Option<File>,
+    pending: Vec<String>,
+    dropped: usize,
+    written: u64,
+    max_bytes: u64,
+    capped: bool,
+}
+
+impl Sink {
+    pub fn new(max_bytes: u64) -> Sink {
+        Sink { file: None, pending: Vec::new(), dropped: 0, written: 0, max_bytes, capped: false }
+    }
+
+    /// Append `line` to the file, or keep it until [`Sink::attach`].
+    pub fn write(&mut self, line: &str) {
+        if self.file.is_none() {
+            if self.capped {
+                // No file is coming ([`Sink::no_file`]): standard error had the line.
+            } else if self.pending.len() < MAX_PENDING {
+                self.pending.push(line.to_owned());
+            } else {
+                self.dropped = self.dropped.saturating_add(1);
+            }
+            return;
+        }
+        self.append(line);
+    }
+
+    /// Start writing to `file` (a fresh one: the size cap starts over), first the lines kept so far.
+    pub fn attach(&mut self, file: File) {
+        self.file = Some(file);
+        self.written = 0;
+        self.capped = false;
+        for line in std::mem::take(&mut self.pending) {
+            self.append(&line);
+        }
+        if self.dropped > 0 {
+            let note = format!("{} earlier log lines were dropped before the log file was opened\n", self.dropped);
+            self.dropped = 0;
+            self.append(&note);
+        }
+    }
+
+    /// There will be no log file: forget the lines kept for it and keep no more.
+    pub fn no_file(&mut self) {
+        self.file = None;
+        self.pending = Vec::new();
+        self.dropped = 0;
+        self.capped = true;
+    }
+
+    fn append(&mut self, line: &str) {
+        if self.capped {
+            return;
+        }
+        let Some(file) = self.file.as_mut() else { return };
+        let len = u64::try_from(line.len()).unwrap_or(u64::MAX);
+        let text = if self.written.saturating_add(len) > self.max_bytes {
+            self.capped = true;
+            format!("log file reached {} bytes; later records go to standard error only\n", self.max_bytes)
+        } else {
+            self.written = self.written.saturating_add(len);
+            line.to_owned()
+        };
+        // A full disk or a vanished file must not take the app down: drop the file sink.
+        if file.write_all(text.as_bytes()).is_err() {
+            self.file = None;
+            self.capped = true;
+        }
+    }
+}
+
+/// The installed logger: a [`Filter`], standard error and a [`Sink`].
+pub struct AppLogger {
+    filter: Filter,
+    stderr: bool,
+    sink: Mutex<Sink>,
+}
+
+impl AppLogger {
+    pub fn new(filter: Filter, stderr: bool) -> AppLogger {
+        AppLogger { filter, stderr, sink: Mutex::new(Sink::new(MAX_FILE_BYTES)) }
+    }
+
+    /// Rotate the logs in `dir` and send records to the fresh file; returns its path. On an error
+    /// standard error stays the only sink.
+    pub fn attach_dir(&self, dir: &Path) -> Result<PathBuf, String> {
+        // A thread that panicked while logging leaves the sink usable: take it back.
+        let mut sink = self.sink.lock().unwrap_or_else(PoisonError::into_inner);
+        match rotate(dir) {
+            Ok((path, file)) => {
+                sink.attach(file);
+                Ok(path)
+            }
+            Err(e) => {
+                sink.no_file();
+                Err(format!("{}: {e}", dir.display()))
+            }
+        }
+    }
+
+    /// No log file for this run (`LIGHTCRAFT_NO_PREFS`): log to standard error only.
+    pub fn no_file(&self) {
+        self.sink.lock().unwrap_or_else(PoisonError::into_inner).no_file();
+    }
+
+    /// Add a panic report to the log file only (the default panic hook has printed it on standard
+    /// error), whatever the filter. Skipped when the sink is busy: the panic may have happened on
+    /// this very thread while it held the lock, and waiting would hang the app.
+    pub fn record_panic(&self, thread: &str, report: &str) {
+        let line = format_line(&timestamp(SystemTime::now()), log::Level::Error, thread, "lightcraft", report);
+        let mut sink = match self.sink.try_lock() {
+            Ok(sink) => sink,
+            Err(TryLockError::Poisoned(p)) => p.into_inner(),
+            Err(TryLockError::WouldBlock) => return,
+        };
+        sink.write(&line);
+    }
+}
+
+impl log::Log for AppLogger {
+    fn enabled(&self, metadata: &log::Metadata<'_>) -> bool {
+        metadata.level() <= self.filter.level_for(metadata.target())
+    }
+
+    fn log(&self, record: &log::Record<'_>) {
+        if !self.enabled(record.metadata()) {
+            return;
+        }
+        // Formatted before the sink is locked, so a message that logs while it is formatted can't deadlock.
+        let thread = std::thread::current();
+        let line =
+            format_line(&timestamp(SystemTime::now()), record.level(), thread.name().unwrap_or("?"), record.target(), &record.args().to_string());
+        if self.stderr {
+            // No terminal (a Windows GUI build, a closed pipe) is not an error worth reporting.
+            let _ = std::io::stderr().write_all(line.as_bytes());
+        }
+        self.sink.lock().unwrap_or_else(PoisonError::into_inner).write(&line);
+    }
+
+    fn flush(&self) {}
+}
+
+/// Install the logger (filter from [`filter_spec`]); `None` when another logger was installed
+/// first. Call [`AppLogger::attach_dir`] (or [`AppLogger::no_file`]) once the arguments are parsed.
+pub fn install() -> Option<&'static AppLogger> {
+    static LOGGER: OnceLock<AppLogger> = OnceLock::new();
+    let var = |name| std::env::var(name).ok();
+    let filter = Filter::parse(&filter_spec(var("LIGHTCRAFT_LOG").as_deref(), var("RUST_LOG").as_deref()));
+    let max = filter.max();
+    let logger = LOGGER.get_or_init(|| AppLogger::new(filter, true));
+    log::set_logger(logger).ok()?;
+    log::set_max_level(max);
+    Some(logger)
+}
+
+/// Chain a panic hook after the current one (the engine guard's: standard error plus
+/// `lightcraft-panics.log` in the temp folder) that also puts the panic in the log file.
+pub fn record_panics(logger: &'static AppLogger) {
+    let previous = std::panic::take_hook();
+    std::panic::set_hook(Box::new(move |info| {
+        previous(info);
+        let thread = std::thread::current();
+        let at = info.location().map(|l| format!(" at {}:{}", l.file(), l.line())).unwrap_or_default();
+        let report = format!("panic{at}: {}", lightcraft_engine::guard::panic_message(info.payload()));
+        logger.record_panic(thread.name().unwrap_or("?"), &report);
+    }));
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use std::sync::atomic::{AtomicUsize, Ordering};
+    use std::time::{Duration, UNIX_EPOCH};
+
+    fn temp_dir(tag: &str) -> PathBuf {
+        static N: AtomicUsize = AtomicUsize::new(0);
+        let d = std::env::temp_dir().join(format!("lightcraft-logging-{tag}-{}-{}", std::process::id(), N.fetch_add(1, Ordering::Relaxed)));
+        let _ = std::fs::remove_dir_all(&d);
+        d
+    }
+
+    fn read(path: &Path) -> String {
+        std::fs::read_to_string(path).expect("read log")
+    }
+
+    fn record(logger: &AppLogger, level: log::Level, target: &str, msg: &str) {
+        use log::Log;
+        logger.log(&log::Record::builder().level(level).target(target).args(format_args!("{msg}")).build());
+    }
+
+    /// The logger this replaced (`StderrLog`): warnings and errors from every target, `info` and
+    /// `debug` only from LightCraft's own crates, and only when `LIGHTCRAFT_LOG` asked for them.
+    fn old_stderr_log_enabled(lightcraft_log: &str, level: log::Level, target: &str) -> bool {
+        let max = match lightcraft_log {
+            "debug" => LevelFilter::Debug,
+            "info" => LevelFilter::Info,
+            _ => LevelFilter::Warn,
+        };
+        level <= max && (level <= log::Level::Warn || target.starts_with("lightcraft"))
+    }
+
+    #[test]
+    fn the_default_filter_shows_lightcraft_info_and_other_crates_warnings() {
+        let f = Filter::parse(DEFAULT_FILTER);
+        assert_eq!(f.level_for("lightcraft"), LevelFilter::Info);
+        assert_eq!(f.level_for("lightcraft::control_server"), LevelFilter::Info);
+        assert_eq!(f.level_for("lightcraft_engine::guard"), LevelFilter::Info);
+        assert_eq!(f.level_for("lightcraft_ui_egui::render"), LevelFilter::Info);
+        assert_eq!(f.level_for("wgpu_core::device"), LevelFilter::Warn);
+        assert_eq!(f.level_for("naga"), LevelFilter::Warn);
+        assert_eq!(f.max(), LevelFilter::Info);
+        assert_eq!(filter_spec(None, None), DEFAULT_FILTER);
+        // An empty variable is the same as an unset one.
+        assert_eq!(filter_spec(Some(""), Some("  ")), DEFAULT_FILTER);
+    }
+
+    /// `LIGHTCRAFT_LOG` keeps working exactly as it did with the old stderr-only logger.
+    #[test]
+    fn lightcraft_log_keeps_its_meaning() {
+        let targets =
+            ["lightcraft", "lightcraft::control_server", "lightcraft_engine::segment", "lightcraft_gpu", "wgpu_core::device", "naga", "eframe"];
+        let levels = [log::Level::Error, log::Level::Warn, log::Level::Info, log::Level::Debug, log::Level::Trace];
+        for value in ["info", "debug", "warn", "error", "verbose", "INFO"] {
+            let f = Filter::parse(&filter_spec(Some(value), None));
+            for target in targets {
+                for level in levels {
+                    assert_eq!(level <= f.level_for(target), old_stderr_log_enabled(value, level, target), "LIGHTCRAFT_LOG={value} {level} {target}");
+                }
+            }
+        }
+        assert_eq!(Filter::parse(&filter_spec(Some("debug"), None)).max(), LevelFilter::Debug);
+        assert_eq!(Filter::parse(&filter_spec(Some("warn"), None)).max(), LevelFilter::Warn);
+    }
+
+    #[test]
+    fn rust_log_replaces_the_default_and_lightcraft_log_wins_over_it() {
+        assert_eq!(filter_spec(None, Some("warn,wgpu_core=info")), "warn,wgpu_core=info");
+        let f = Filter::parse(&filter_spec(None, Some("debug")));
+        assert_eq!(f.level_for("eframe"), LevelFilter::Debug);
+        // The app's own variable is the more specific choice.
+        let f = Filter::parse(&filter_spec(Some("info"), Some("trace")));
+        assert_eq!(f.level_for("eframe"), LevelFilter::Warn);
+        assert_eq!(f.level_for("lightcraft_engine"), LevelFilter::Info);
+    }
+
+    #[test]
+    fn directives_follow_env_logger_and_the_most_specific_one_wins() {
+        let f = Filter::parse("info,wgpu_core=error,lightcraft_pipeline=trace,lightcraft_pipeline::tiles=off");
+        assert_eq!(f.level_for("eframe"), LevelFilter::Info);
+        assert_eq!(f.level_for("wgpu_core::instance"), LevelFilter::Error);
+        assert_eq!(f.level_for("lightcraft_pipeline::stage"), LevelFilter::Trace);
+        assert_eq!(f.level_for("lightcraft_pipeline::tiles"), LevelFilter::Off);
+        assert_eq!(f.max(), LevelFilter::Trace);
+        // A module name is matched at `::` boundaries, not as a bare prefix.
+        assert_eq!(f.level_for("wgpu_core_extra"), LevelFilter::Info);
+        // A trailing `*` is a prefix.
+        assert_eq!(Filter::parse("lightcraft*=debug").level_for("lightcraft_raw::cr2"), LevelFilter::Debug);
+        assert_eq!(Filter::parse("lightcraft*=debug").level_for("eframe"), LevelFilter::Error);
+        // A bare target name sets that target to the most verbose level, as env_logger does.
+        assert_eq!(Filter::parse("naga").level_for("naga::front"), LevelFilter::Trace);
+        assert_eq!(Filter::parse("naga").level_for("eframe"), LevelFilter::Error);
+        // Levels are case-insensitive and surrounding blanks are ignored.
+        assert_eq!(Filter::parse(" WARN , lightcraft = Debug ").level_for("lightcraft"), LevelFilter::Debug);
+        assert_eq!(Filter::parse(" WARN , lightcraft = Debug ").level_for("eframe"), LevelFilter::Warn);
+    }
+
+    #[test]
+    fn hostile_specs_never_panic_and_fall_back_sensibly() {
+        let long = "x".repeat(100_000);
+        let many = "a=info,".repeat(10_000);
+        for spec in [
+            "",
+            ",,,",
+            "=",
+            "==",
+            "=debug",
+            "lightcraft=",
+            "nonsense=loud",
+            "🦀=info",
+            "*",
+            "*=",
+            "=*",
+            "a=b=c",
+            "warn,,lightcraft=DEBUG",
+            "\0",
+            "é::ü=trace",
+            &long,
+            &many,
+        ] {
+            let f = Filter::parse(spec);
+            for target in ["", "lightcraft", "🦀", "a::b", "*"] {
+                let _ = f.level_for(target);
+            }
+            let _ = f.max();
+            let _ = filter_spec(Some(spec), Some(spec));
+        }
+        assert_eq!(Filter::parse("warn,,lightcraft=DEBUG").level_for("lightcraft"), LevelFilter::Debug);
+        // An unknown level is skipped: that target keeps the default.
+        assert_eq!(Filter::parse("nonsense=loud").level_for("nonsense"), LevelFilter::Error);
+        assert_eq!(Filter::parse("lightcraft=").level_for("lightcraft"), LevelFilter::Error);
+        assert_eq!(Filter::parse("=debug").max(), LevelFilter::Error);
+        assert_eq!(Filter::parse("🦀=info").level_for("🦀::claw"), LevelFilter::Info);
+    }
+
+    #[test]
+    fn timestamps_are_utc_with_milliseconds() {
+        assert_eq!(timestamp(UNIX_EPOCH), "1970-01-01T00:00:00.000Z");
+        assert_eq!(timestamp(UNIX_EPOCH + Duration::from_millis(1_791_446_357_728)), "2026-10-08T07:59:17.728Z");
+        // Leap day, and the day after it.
+        assert_eq!(timestamp(UNIX_EPOCH + Duration::from_secs(1_709_164_800)), "2024-02-29T00:00:00.000Z");
+        assert_eq!(timestamp(UNIX_EPOCH + Duration::from_secs(1_709_251_199)), "2024-02-29T23:59:59.000Z");
+        assert_eq!(timestamp(UNIX_EPOCH + Duration::from_secs(1_709_251_200)), "2024-03-01T00:00:00.000Z");
+        // 2000 is a leap year (divisible by 400), 2100 is not.
+        assert_eq!(timestamp(UNIX_EPOCH + Duration::from_secs(951_782_400)), "2000-02-29T00:00:00.000Z");
+        assert_eq!(timestamp(UNIX_EPOCH + Duration::from_secs(4_107_542_400)), "2100-03-01T00:00:00.000Z");
+        // Before the epoch (a clock set wrong) clamps instead of panicking.
+        assert_eq!(timestamp(UNIX_EPOCH - Duration::from_secs(5)), "1970-01-01T00:00:00.000Z");
+    }
+
+    #[test]
+    fn a_line_carries_time_level_thread_target_and_message() {
+        assert_eq!(
+            format_line("2026-10-08T07:59:17.728Z", log::Level::Info, "main", "lightcraft", "library /photos: 12 photos"),
+            "2026-10-08T07:59:17.728Z INFO  [main] lightcraft: library /photos: 12 photos\n"
+        );
+        // A multi-line message keeps its lines; the record still ends in one newline.
+        assert!(format_line("t", log::Level::Error, "w", "x", "a\nb\n").ends_with("x: a\nb\n"));
+    }
+
+    #[test]
+    fn rotation_keeps_the_previous_logs_and_starts_an_empty_file() {
+        let dir = temp_dir("rotate");
+        for run in 1..=4 {
+            let (path, _file) = rotate(&dir).expect("rotate");
+            assert_eq!(path, dir.join(LOG_FILE));
+            assert_eq!(read(&path), "");
+            std::fs::write(&path, format!("run {run}")).expect("write");
+        }
+        assert_eq!(read(&dir.join(LOG_FILE)), "run 4");
+        assert_eq!(read(&dir.join("lightcraft.1.log")), "run 3");
+        assert_eq!(read(&dir.join("lightcraft.2.log")), "run 2");
+        assert!(!dir.join("lightcraft.3.log").exists(), "only {KEEP} old logs are kept");
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    #[test]
+    fn rotation_into_an_unusable_directory_is_an_error_not_a_panic() {
+        let dir = temp_dir("blocked");
+        std::fs::create_dir_all(dir.parent().expect("parent")).expect("tmp");
+        std::fs::write(&dir, "a file where the directory should be").expect("block");
+        assert!(rotate(&dir).is_err());
+        let _ = std::fs::remove_file(&dir);
+    }
+
+    #[test]
+    fn lines_logged_before_the_file_exists_are_written_when_it_is_attached() {
+        let dir = temp_dir("pending");
+        let (path, file) = rotate(&dir).expect("rotate");
+        let mut sink = Sink::new(MAX_FILE_BYTES);
+        sink.write("early 1\n");
+        sink.write("early 2\n");
+        sink.attach(file);
+        sink.write("late\n");
+        assert_eq!(read(&path), "early 1\nearly 2\nlate\n");
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    #[test]
+    fn the_pending_buffer_is_bounded_and_says_how_much_it_dropped() {
+        let dir = temp_dir("bounded");
+        let (path, file) = rotate(&dir).expect("rotate");
+        let mut sink = Sink::new(MAX_FILE_BYTES);
+        for i in 0..MAX_PENDING + 5 {
+            sink.write(&format!("line {i}\n"));
+        }
+        sink.attach(file);
+        let text = read(&path);
+        assert_eq!(text.lines().filter(|l| l.starts_with("line ")).count(), MAX_PENDING);
+        assert!(text.contains("5 earlier log lines were dropped"), "{text}");
+        assert!(text.contains("line 0\n") && !text.contains(&format!("line {MAX_PENDING}\n")));
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    #[test]
+    fn exactly_a_full_pending_buffer_drops_nothing() {
+        let dir = temp_dir("full");
+        let (path, file) = rotate(&dir).expect("rotate");
+        let mut sink = Sink::new(MAX_FILE_BYTES);
+        for i in 0..MAX_PENDING {
+            sink.write(&format!("line {i}\n"));
+        }
+        sink.attach(file);
+        let text = read(&path);
+        assert_eq!(text.lines().count(), MAX_PENDING);
+        assert!(!text.contains("dropped"), "{text}");
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    #[test]
+    fn without_a_file_nothing_is_kept_in_memory() {
+        let mut sink = Sink::new(MAX_FILE_BYTES);
+        sink.write("early\n");
+        sink.no_file();
+        sink.write("late\n");
+        assert!(sink.pending.is_empty() && sink.dropped == 0);
+    }
+
+    #[test]
+    fn the_file_stops_at_its_size_cap_with_one_note() {
+        let dir = temp_dir("cap");
+        let (path, file) = rotate(&dir).expect("rotate");
+        let mut sink = Sink::new(100);
+        sink.attach(file);
+        for i in 0..50 {
+            sink.write(&format!("line {i:02} with some padding\n"));
+        }
+        let text = read(&path);
+        assert_eq!(text.lines().filter(|l| l.starts_with("line ")).count(), 3, "{text}");
+        assert_eq!(text.matches("log file reached").count(), 1, "{text}");
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    #[test]
+    fn the_logger_filters_records_and_writes_them_to_the_attached_file() {
+        use log::Log;
+        let dir = temp_dir("logger");
+        let logger = AppLogger::new(Filter::parse(DEFAULT_FILTER), false);
+        record(&logger, log::Level::Info, "lightcraft", "before the settings directory");
+        record(&logger, log::Level::Info, "wgpu_core::device", "too chatty");
+        let path = logger.attach_dir(&dir).expect("attach");
+        assert_eq!(path, dir.join(LOG_FILE));
+        record(&logger, log::Level::Warn, "wgpu_hal::vulkan", "a real warning");
+        record(&logger, log::Level::Error, "lightcraft_engine::guard", "`develop.reset` failed unexpectedly: boom");
+        record(&logger, log::Level::Debug, "lightcraft", "below info");
+        let text = read(&path);
+        let lines: Vec<&str> = text.lines().collect();
+        assert_eq!(lines.len(), 3, "{text}");
+        assert!(lines[0].contains(" INFO  [") && lines[0].ends_with("lightcraft: before the settings directory"), "{text}");
+        assert!(lines[1].ends_with("wgpu_hal::vulkan: a real warning"), "{text}");
+        assert!(lines[2].contains(" ERROR [") && lines[2].ends_with("lightcraft_engine::guard: `develop.reset` failed unexpectedly: boom"), "{text}");
+        assert!(logger.enabled(&log::Metadata::builder().level(log::Level::Info).target("lightcraft_ui_egui").build()));
+        assert!(!logger.enabled(&log::Metadata::builder().level(log::Level::Info).target("naga").build()));
+        // Attaching again rotates: the first log becomes `.1`.
+        logger.attach_dir(&dir).expect("attach again");
+        assert_eq!(read(&dir.join("lightcraft.1.log")), text);
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    #[test]
+    fn records_from_many_threads_are_whole_lines() {
+        let dir = temp_dir("threads");
+        let logger = std::sync::Arc::new(AppLogger::new(Filter::parse("info"), false));
+        let path = logger.attach_dir(&dir).expect("attach");
+        let workers: Vec<_> = (0..8)
+            .map(|t| {
+                let l = std::sync::Arc::clone(&logger);
+                std::thread::spawn(move || {
+                    for i in 0..100 {
+                        record(&l, log::Level::Info, "lightcraft", &format!("thread {t} record {i}"));
+                    }
+                })
+            })
+            .collect();
+        for w in workers {
+            w.join().expect("worker");
+        }
+        let text = read(&path);
+        assert_eq!(text.lines().count(), 800);
+        assert!(text.lines().all(|l| l.contains(" lightcraft: thread ")), "{text}");
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    #[test]
+    fn attaching_to_an_unusable_directory_reports_an_error_and_keeps_logging() {
+        let dir = temp_dir("unusable");
+        std::fs::create_dir_all(dir.parent().expect("parent")).expect("tmp");
+        std::fs::write(&dir, "not a directory").expect("block");
+        let logger = AppLogger::new(Filter::parse("info"), false);
+        record(&logger, log::Level::Info, "lightcraft", "early");
+        assert!(logger.attach_dir(&dir).is_err());
+        record(&logger, log::Level::Error, "x", "still fine");
+        // Nothing waits in memory for a file that won't come.
+        assert!(logger.sink.lock().expect("sink").pending.is_empty());
+        let _ = std::fs::remove_file(&dir);
+    }
+
+    #[test]
+    fn a_poisoned_sink_still_logs() {
+        let dir = temp_dir("poison");
+        let logger = std::sync::Arc::new(AppLogger::new(Filter::parse("info"), false));
+        let l = std::sync::Arc::clone(&logger);
+        let _ = std::thread::spawn(move || {
+            let _guard = l.sink.lock().expect("lock");
+            panic!("poison the sink");
+        })
+        .join();
+        assert!(logger.sink.is_poisoned());
+        let path = logger.attach_dir(&dir).expect("attach");
+        record(&logger, log::Level::Warn, "lightcraft", "after the poison");
+        assert!(read(&path).contains("lightcraft: after the poison"));
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+
+    /// The panic hook's report lands in the file, and never deadlocks when the panic happened
+    /// while this thread held the sink.
+    #[test]
+    fn a_panic_report_goes_to_the_file_and_never_deadlocks() {
+        let dir = temp_dir("panic");
+        let logger = AppLogger::new(Filter::parse("off"), false);
+        let path = logger.attach_dir(&dir).expect("attach");
+        logger.record_panic("render", "panicked at crates/x/src/lib.rs:3:5:\nboom");
+        {
+            let _held = logger.sink.lock().expect("lock");
+            logger.record_panic("main", "while the sink is held");
+        }
+        let text = read(&path);
+        assert!(text.contains(" ERROR [render] lightcraft: panicked at crates/x/src/lib.rs:3:5:\nboom\n"), "{text}");
+        assert!(!text.contains("while the sink is held"), "{text}");
+        let _ = std::fs::remove_dir_all(&dir);
+    }
+}
@@ -2,9 +2,12 @@
 //! the in-window menus ([`lightcraft_ui_egui::menubar`]): the registry's menu paths, shortcuts,
 //! enabled and checked state, and live labels ("Undo Exposure").
 //!
-//! Key handling: a native key equivalent consumes the key press before egui sees it, so
-//! - shortcuts installed natively are listed in `LightcraftApp::native_shortcuts` and skipped by
-//!   the egui shortcut handler (nothing fires twice);
+//! Key handling: AppKit offers a key press to the menu bar only when it carries ⌘ or ⌃, or is a
+//! function key (F1…). Every other key (`E`, `1`, `⇧P`, `⌥Y`…) goes straight to the window, whose
+//! winit view always takes it, so a menu item never fires from it. Hence
+//! - shortcuts the menu bar really receives ([`menu_delivers`]) are listed in
+//!   `LightcraftApp::native_shortcuts` and skipped by the egui shortcut handler (nothing fires
+//!   twice); the others stay on their menu items for display and egui runs them;
 //! - while a text field has keyboard focus, accelerators without ⌘ (`G`, `1`, `Delete`…) and the
 //!   text-editing ones (⌘A/⌘C/⌘V/⌘X/⌘Z) are removed, so typing and text editing work; they come
 //!   back when the field loses focus.
@@ -51,6 +54,9 @@ pub struct NativeMenu {
    /// Structure of the installed menu (ids, kinds, submenu names); a change rebuilds it.
    structure: String,
    text_focus: bool,
+    /// The keyboard shortcuts editor is recording a key: every accelerator is off so the key
+    /// press reaches it.
+    capturing: bool,
 }

 /// `Cmd+Shift+Z` → a muda accelerator (`None` for keys we leave to egui, e.g. Escape).
@@ -146,9 +152,21 @@ fn yields_to_text(sc: &str) -> bool {
    !sc.contains("Cmd") && !sc.contains("Ctrl") || TEXT_EDIT.contains(&sc)
 }

+/// Whether AppKit hands this shortcut to the menu bar: only key presses with ⌘ or ⌃, and function
+/// keys. A plain `C` or `⇧P` never reaches it (the window's view takes the key), so egui has to run
+/// those even though their menu item shows the key.
+fn menu_delivers(sc: &str) -> bool {
+    sc.split('+').any(|part| matches!(part, "Cmd" | "Ctrl") || part.strip_prefix('F').is_some_and(|n| n.parse::<u8>().is_ok()))
+}
+
+/// The shortcuts the menu bar runs itself, for `LightcraftApp::native_shortcuts`.
+fn owned_by_menu<'a>(installed: impl Iterator<Item = &'a str>, text_focus: bool) -> HashSet<String> {
+    installed.filter(|sc| menu_delivers(sc) && !(text_focus && yields_to_text(sc))).map(str::to_string).collect()
+}
+
 /// `&` marks a mnemonic in muda labels.
 fn label_text(s: &str) -> String {
-    s.replace('&', "&&")
+    lightcraft_ui_egui::i18n::tr(s).replace('&', "&&")
 }

 fn structure_of(bar: &[(String, Vec<MenuNode>)]) -> String {
@@ -171,7 +189,7 @@ fn structure_of(bar: &[(String, Vec<MenuNode>)]) -> String {
            out.push(';');
        }
    }
-    let mut s = String::new();
+    let mut s = format!("{};", lightcraft_ui_egui::i18n::language().code());
    for (t, items) in bar {
        s.push_str(t);
        walk(items, &mut s);
@@ -196,7 +214,7 @@ impl NativeMenu {
            let _ = tx.send(e.id.0);
            repaint.request_repaint();
        }));
-        let mut m = NativeMenu { menu: Menu::new(), items: HashMap::new(), rx, structure: String::new(), text_focus: false };
+        let mut m = NativeMenu { menu: Menu::new(), items: HashMap::new(), rx, structure: String::new(), text_focus: false, capturing: false };
        m.rebuild(app);
        app.native_menu = true;
        m
@@ -210,30 +228,30 @@ impl NativeMenu {

        // the application menu
        let app_menu = Submenu::new("LightCraft", true);
-        let about = MenuItem::with_id("app.about", "About LightCraft", true, None);
-        let settings = MenuItem::with_id(SETTINGS, "Settings…", true, accelerator(SETTINGS_KEY));
-        let quit = MenuItem::with_id(QUIT, "Quit LightCraft", true, accelerator("Cmd+Q"));
+        let about = MenuItem::with_id("app.about", lightcraft_ui_egui::i18n::tr("About LightCraft"), true, None);
+        let settings = MenuItem::with_id(SETTINGS, lightcraft_ui_egui::i18n::tr("Settings…"), true, accelerator(SETTINGS_KEY));
+        let quit = MenuItem::with_id(QUIT, lightcraft_ui_egui::i18n::tr("Quit LightCraft"), true, accelerator("Cmd+Q"));
        let _ = app_menu.append_items(&[
            &about,
            &PredefinedMenuItem::separator(),
            &settings,
            &PredefinedMenuItem::separator(),
-            &PredefinedMenuItem::services(None),
+            &PredefinedMenuItem::services(Some(lightcraft_ui_egui::i18n::tr("Services"))),
            &PredefinedMenuItem::separator(),
-            &PredefinedMenuItem::hide(Some("Hide LightCraft")),
-            &PredefinedMenuItem::hide_others(None),
-            &PredefinedMenuItem::show_all(None),
+            &PredefinedMenuItem::hide(Some(lightcraft_ui_egui::i18n::tr("Hide LightCraft"))),
+            &PredefinedMenuItem::hide_others(Some(lightcraft_ui_egui::i18n::tr("Hide Others"))),
+            &PredefinedMenuItem::show_all(Some(lightcraft_ui_egui::i18n::tr("Show All"))),
            &PredefinedMenuItem::separator(),
            &quit,
        ]);
        let _ = self.menu.append(&app_menu);

        for (title, nodes) in &bar {
-            let sub = Submenu::new(title, true);
+            let sub = Submenu::new(label_text(title), true);
            if title == "Window" {
                let _ = sub.append_items(&[
-                    &PredefinedMenuItem::minimize(None),
-                    &PredefinedMenuItem::maximize(Some("Zoom")),
+                    &PredefinedMenuItem::minimize(Some(lightcraft_ui_egui::i18n::tr("Minimize"))),
+                    &PredefinedMenuItem::maximize(Some(lightcraft_ui_egui::i18n::tr("Zoom"))),
                    &PredefinedMenuItem::separator(),
                ]);
            }
@@ -247,10 +265,16 @@ impl NativeMenu {
            self.append_nodes(&sub, &nodes);
            if title == "File" {
                // ⌘W, the system's own item
-                let _ = sub.append_items(&[&PredefinedMenuItem::separator(), &PredefinedMenuItem::close_window(None)]);
+                let _ = sub.append_items(&[
+                    &PredefinedMenuItem::separator(),
+                    &PredefinedMenuItem::close_window(Some(lightcraft_ui_egui::i18n::tr("Close Window"))),
+                ]);
            }
            if title == "Window" {
-                let _ = sub.append_items(&[&PredefinedMenuItem::separator(), &PredefinedMenuItem::bring_all_to_front(None)]);
+                let _ = sub.append_items(&[
+                    &PredefinedMenuItem::separator(),
+                    &PredefinedMenuItem::bring_all_to_front(Some(lightcraft_ui_egui::i18n::tr("Bring All to Front"))),
+                ]);
                sub.set_as_windows_menu_for_nsapp();
            }
            if title == "Help" {
@@ -260,6 +284,7 @@ impl NativeMenu {
        }
        self.menu.init_for_nsapp();
        self.text_focus = false;
+        self.capturing = false;
        self.publish_shortcuts(app);
    }

@@ -282,12 +307,23 @@ impl NativeMenu {
                    let accel = shortcut.as_deref().filter(|s| !CONTEXTUAL.contains(s)).and_then(accelerator);
                    let handle = match checked {
                        Some(c) => {
-                            let it = CheckMenuItem::with_id(key.clone(), label_text(label), *enabled, *c, accel);
+                            let it = CheckMenuItem::with_id(
+                                key.clone(),
+                                lightcraft_ui_egui::menubar::display_item_label(id, params, label).replace('&', "&&"),
+                                *enabled,
+                                *c,
+                                accel,
+                            );
                            let _ = sub.append(&it);
                            Handle::Check(it)
                        }
                        None => {
-                            let it = MenuItem::with_id(key.clone(), label_text(label), *enabled, accel);
+                            let it = MenuItem::with_id(
+                                key.clone(),
+                                lightcraft_ui_egui::menubar::display_item_label(id, params, label).replace('&', "&&"),
+                                *enabled,
+                                accel,
+                            );
                            let _ = sub.append(&it);
                            Handle::Plain(it)
                        }
@@ -312,22 +348,22 @@ impl NativeMenu {

    /// Tell the egui shortcut handler which shortcuts the menu bar currently owns.
    fn publish_shortcuts(&self, app: &mut LightcraftApp) {
-        app.native_shortcuts = self
-            .items
-            .values()
-            .filter(|i| i.accel.is_some())
-            .filter_map(|i| i.shortcut.clone())
-            .filter(|sc| !(self.text_focus && yields_to_text(sc)))
-            .collect::<HashSet<_>>();
+        let installed = self.items.values().filter(|i| i.accel.is_some()).filter_map(|i| i.shortcut.as_deref());
+        // (none while the keymap editor records a shortcut: every key goes to it)
+        app.native_shortcuts = if self.capturing { HashSet::new() } else { owned_by_menu(installed, self.text_focus) };
        app.native_shortcuts.insert(SETTINGS_KEY.to_string());
    }

    /// Per frame: run chosen items, then sync labels / enabled / checked and the text-focus
-    /// accelerators with the app state.
-    pub fn update(&mut self, app: &mut LightcraftApp, ctx: &egui::Context) {
+    /// accelerators with the app state. Returns whether Quit was chosen: the caller saves
+    /// everything and exits the process itself (closing the viewport while AppKit terminates the
+    /// app can crash in winit's teardown, PR #444).
+    pub fn update(&mut self, app: &mut LightcraftApp, ctx: &egui::Context) -> bool {
+        lightcraft_ui_egui::i18n::set_language(app.ui.language);
+        let mut quit = false;
        while let Ok(key) = self.rx.try_recv() {
            if key == QUIT {
-                ctx.send_viewport_cmd(egui::ViewportCommand::Close);
+                quit = true;
                continue;
            }
            if key == "app.about" || key == SETTINGS {
@@ -342,7 +378,7 @@ impl NativeMenu {
        let bar = menu_bar(app);
        if structure_of(&bar) != self.structure {
            self.rebuild(app);
-            return;
+            return quit;
        }
        fn walk(nodes: &[MenuNode], items: &mut HashMap<String, Item>) {
            for n in nodes {
@@ -352,8 +388,8 @@ impl NativeMenu {
                        let Some(it) = items.get_mut(&MenuNode::key(id, params)) else { continue };
                        if it.label != *label {
                            match &it.handle {
-                                Handle::Plain(h) => h.set_text(label_text(label)),
-                                Handle::Check(h) => h.set_text(label_text(label)),
+                                Handle::Plain(h) => h.set_text(lightcraft_ui_egui::menubar::display_item_label(id, params, label).replace('&', "&&")),
+                                Handle::Check(h) => h.set_text(lightcraft_ui_egui::menubar::display_item_label(id, params, label).replace('&', "&&")),
                            }
                            it.label = label.clone();
                        }
@@ -379,14 +415,16 @@ impl NativeMenu {
        walk(&bar.iter().flat_map(|(_, v)| v.clone()).collect::<Vec<_>>(), &mut self.items);

        let focus = ctx.egui_wants_keyboard_input();
-        if focus != self.text_focus {
+        let capturing = app.recording_shortcut.is_some();
+        if focus != self.text_focus || capturing != self.capturing {
            self.text_focus = focus;
+            self.capturing = capturing;
            for it in self.items.values() {
                let Some(sc) = it.shortcut.as_deref() else { continue };
-                if it.accel.is_none() || !yields_to_text(sc) {
+                if it.accel.is_none() {
                    continue;
                }
-                let accel = if focus { None } else { it.accel };
+                let accel = if capturing || (focus && yields_to_text(sc)) { None } else { it.accel };
                let _ = match &it.handle {
                    Handle::Plain(h) => h.set_accelerator(accel),
                    Handle::Check(h) => h.set_accelerator(accel),
@@ -394,6 +432,7 @@ impl NativeMenu {
            }
            self.publish_shortcuts(app);
        }
+        quit
    }
 }

@@ -401,6 +440,27 @@ impl NativeMenu {
 mod tests {
    use super::*;

+    /// The menu model's titles stay English, so only the language can tell the native menu to
+    /// rebuild its titles (and the macOS app menu): every language switch — also between two CJK
+    /// languages — must change the structure key.
+    #[test]
+    fn switching_language_rebuilds_native_menu_structure() {
+        use lightcraft_ui_egui::i18n::{Locale, set_language};
+        let app = LightcraftApp::new(lightcraft_engine::Session::with_demo(), Default::default());
+        set_language(Locale::En);
+        let bar = menu_bar(&app);
+        let mut seen = std::collections::BTreeSet::new();
+        for language in Locale::ALL {
+            set_language(*language);
+            assert!(seen.insert(structure_of(&bar)), "{language:?} shares another language's menu structure key");
+        }
+        set_language(Locale::ZhHant);
+        assert_eq!(label_text("File"), "檔案");
+        assert_eq!(label_text("Settings…"), "設定…");
+        assert_eq!(label_text("Quit LightCraft"), "結束 LightCraft");
+        set_language(Locale::En);
+    }
+
    #[test]
    fn shortcuts_map_to_accelerators() {
        assert_eq!(accelerator("Cmd+Shift+Z"), Some(Accelerator::new(Modifiers::META | Modifiers::SHIFT, Code::KeyZ)));
@@ -428,4 +488,39 @@ mod tests {
            assert!(accelerator(sc).is_some(), "{sc}");
        }
    }
+
+    /// Single keys shown in the menu bar (E, C, ⇧P, ratings…) never reach it on macOS, so egui must
+    /// keep handling them; ⌘ / ⌃ combinations and function keys are the menu bar's own.
+    #[test]
+    fn single_key_shortcuts_stay_with_egui() {
+        assert!(menu_delivers("Cmd+Shift+H") && menu_delivers("Ctrl+H") && menu_delivers("F2") && menu_delivers("Cmd+F11"));
+        assert!(!menu_delivers("E") && !menu_delivers("Shift+P") && !menu_delivers("Alt+Y") && !menu_delivers("1") && !menu_delivers("F"));
+        let app = LightcraftApp::new(lightcraft_engine::Session::with_demo(), Default::default());
+        fn all(n: &[MenuNode], out: &mut Vec<String>) {
+            for x in n {
+                match x {
+                    MenuNode::Item { shortcut: Some(s), .. } if accelerator(s).is_some() => out.push(s.clone()),
+                    MenuNode::Submenu { children, .. } => all(children, out),
+                    _ => {}
+                }
+            }
+        }
+        let mut installed = Vec::new();
+        for (_, v) in menu_bar(&app) {
+            all(&v, &mut installed);
+        }
+        for sc in ["E", "C", "Shift+P", "Cmd+Shift+H", "F2"] {
+            assert!(installed.iter().any(|s| s == sc), "{sc} is in the menu bar");
+        }
+        let owned = owned_by_menu(installed.iter().map(String::as_str), false);
+        for sc in ["E", "C", "Shift+P", "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "P", "U"] {
+            assert!(installed.iter().any(|s| s == sc), "{sc} is displayed in the menu");
+            assert!(!owned.contains(sc), "{sc} must be left to egui");
+        }
+        for sc in ["Cmd+Shift+H", "F2"] {
+            assert!(owned.contains(sc), "{sc} is run by the menu bar");
+        }
+        // while typing, the menu bar gives up F2 too (the text field has it)
+        assert!(!owned_by_menu(installed.iter().map(String::as_str), true).contains("F2"));
+    }
 }
@@ -10,6 +10,9 @@ covers a whole directory), or if a licence file referenced in the Licence column
 |---|---|---|---|---|---|---|
 | `assets/fonts/Inter-*.ttf` (Regular, Medium, SemiBold) | Inter | Rasmus Andersson | https://github.com/rsms/inter | SIL Open Font License 1.1 (`assets/fonts/OFL-Inter.txt`) | 2026-09-30 | none |
 | `assets/app-icon/` (`lightcraft.svg` and the PNG/ICNS/ICO/SVG files derived from it) | LightCraft app icon (lynx) | LightCraft project owner (drawn in ArtCraft) | original work (vectorised from the owner's drawing; see `assets/app-icon/README.md`) | MIT OR Apache-2.0 (`assets/app-icon/LICENSE.txt`) | 2026-10-02 | vectorised to SVG; rendered to PNG/ICNS/ICO by `packaging/icons.sh` |
+| `assets/camera-profiles/*.json` | Camera colour profiles (matrix + hue/saturation/value table), built into LightCraft | ILCE-7M4: Harald Hoyer | original work: fitted by `lightcraft-cli calibrate` to the contributor's own raw photos and their in-camera JPEGs (597 ILCE-7M4 photos, 2026); aggregate colour statistics only, no image content; no Adobe or third-party profile data | MIT OR Apache-2.0 | 2026-10-07 | none |
+| `packaging/macos/dmg/` (`background.svg`, the `background.tiff` rendered from it by `generate.py`, and the `dmg-layout.DS_Store` layout file) | LightCraft macOS DMG window background | @XusBadia | original work for this repository, derived from the LightCraft app icon (`assets/app-icon/lightcraft.svg`, referenced via `<image>`, not copied); text is outlined Inter / JetBrains Mono glyph paths | MIT OR Apache-2.0 | 2026-10-08 | none (original); used for the Finder window of the macOS DMG, `packaging/macos/dmg/generate.py` renders it to `background.tiff` |
+| `assets/camera-profiles/X-H2S.json`, `assets/camera-profiles/X-T4.json` | Fujifilm camera colour profiles (matrix + hue/saturation/value table), built into LightCraft | LightCraft contributors; source photographs supplied by Sebastian Pfluegelmeier | original work: fitted by `lightcraft-cli calibrate` to 201 X-H2S and 545 X-T4 RAFs and their in-camera JPEGs; aggregate colour statistics only, no image content; no Adobe or third-party profile data | MIT OR Apache-2.0 | 2026-10-08 | none |
 | `crates/scenes/` (generated images) | Procedural demo photographs | LightCraft contributors | original work (generated at runtime, no source imagery) | MIT OR Apache-2.0 | 2026-09-30 | n/a |
 | `crates/ui-egui/src/icons.rs` | UI icons drawn as vector paths in code | LightCraft contributors | original work | MIT OR Apache-2.0 | 2026-09-30 | n/a |
 | `docs/images/*-tetons.jpg`, `docs/images/grid-pd.jpg` (screenshots containing it) | "The Tetons and the Snake River" | Ansel Adams (1942), U.S. National Archives | https://commons.wikimedia.org/wiki/File:Adams_The_Tetons_and_the_Snake_River.jpg | Public domain (U.S. federal government work) | 2026-09-30 | developed in LightCraft; shown inside app screenshots |
@@ -17,7 +20,13 @@ covers a whole directory), or if a licence file referenced in the Licence column
 | `docs/images/*-earthrise.jpg`, `docs/images/grid-pd.jpg` | "Earthrise" (AS08-14-2383) | NASA / Bill Anders, Apollo 8 (1968) | https://commons.wikimedia.org/wiki/File:NASA-Apollo8-Dec24-Earthrise.jpg | Public domain (NASA) | 2026-09-30 | developed in LightCraft; shown inside app screenshots |
 | `docs/images/*-blue-marble.jpg`, `docs/images/grid-pd.jpg` | "The Blue Marble" (AS17-148-22727) | NASA, Apollo 17 crew (1972) | https://commons.wikimedia.org/wiki/File:The_Earth_seen_from_Apollo_17.jpg | Public domain (NASA) | 2026-09-30 | developed in LightCraft; shown inside app screenshots |
 | `docs/images/*.jpg` (UI) | LightCraft application screenshots | LightCraft contributors | original work (captured with `docs/showcase/`) | MIT OR Apache-2.0 | 2026-09-30 | n/a |
+| `docs/brand/` (artcraft-logo and artcraft-mark, SVG and PNG) | ArtCraft wordmark and mark | ArtCraft Team | original work (https://getartcraft.com/) | ArtCraft trademark, see `docs/brand/LICENSE-brand.txt` (not open source) | 2026-10-01 | none |
+| craft-fonts: optional build input, not files in this repository | BIZ UDPGothic (Regular, Bold), BIZ UDMincho Regular, Shippori Mincho Regular, Noto Sans CJK SC Regular | Morisawa Inc. / The BIZ UDGothic and BIZ UDMincho Project Authors; The Shippori Mincho Project Authors; Google and Adobe (Noto Sans CJK, maintainer-approved exception; see AGENTS.md) (per craft-fonts' attribution) | https://github.com/storytold/craft-fonts/blob/main/ATTRIBUTION.md | SIL Open Font License 1.1 (each font's OFL.txt in craft-fonts; release packages ship it as OFL-<family>.txt) | 2026-10-06 | none; native builds made with `CRAFT_FONTS_DIR` embed the manifest fonts (all official releases); the web build embeds BIZ UDPGothic Regular and ships Noto Sans CJK SC Regular as a separate font file |

 Notes
 - Fonts authored or published by Adobe (Source Sans/Serif/Code, Source Han, …) are not used, even though some are
  OFL-licensed: the asset rule excludes anything from Adobe. Source Sans 3 was removed on 2026-09-30 and replaced by Inter.
+- CJK fonts (BIZ UDPGothic, BIZ UDMincho, Shippori Mincho, Noto Sans CJK SC) are not in this repository. They live in
+  [storytold/craft-fonts](https://github.com/storytold/craft-fonts), with their licences and attribution in its
+  `ATTRIBUTION.md`, and are embedded only in builds made with the optional `CRAFT_FONTS_DIR` build input (all official
+  releases). Never commit font files here; add new fonts to craft-fonts (craftrules `standards/fonts.md`).
@@ -0,0 +1 @@
+{"version":1,"model":"ILCE-7M4","files":597,"samples":2182876,"matrix":[[1.06975,-0.0709886,-0.000770444],[-0.0169822,1.08012,-0.0651276],[-0.114835,-0.601525,1.74804]],"hue_sat":{"hue_divisions":72,"sat_divisions":5,"val_divisions":5,"data":[[0.0,1.0,1.0],[-0.819008,1.14375,1.0],[-6.07446,1.10974,1.0],[-4.0478,1.04285,1.0],[-0.546764,1.00329,1.0],[0.0,1.0,1.0],[-1.0412,1.10597,1.0],[-7.22645,1.05878,1.0],[-7.30635,0.989387,1.0],[-5.18589,0.974522,1.0],[0.0,1.0,1.0],[-1.02364,1.06303,1.0],[-6.94669,1.01579,1.0],[-7.47614,0.955435,1.0],[-7.02579,0.940854,1.0],[0.0,1.0,1.0],[-0.29852,1.04349,1.0],[-6.37202,1.00021,1.0],[-7.52098,0.935457,1.0],[-7.45702,0.922239,1.0],[0.0,1.0,1.0],[0.367464,1.03266,1.0],[-4.96951,0.996822,1.0],[-7.41706,0.948709,1.0],[-8.04735,0.91672,1.0],[0.0,1.0,1.0],[0.0388419,1.02587,1.0],[-4.47266,1.01282,1.0],[-7.26834,0.987279,1.0],[-9.02713,0.921453,1.0],[0.0,1.0,1.0],[-0.297476,1.02333,1.0],[-4.6323,1.00993,1.0],[-8.57248,0.962973,1.0],[-12.2375,0.872088,1.0],[0.0,1.0,1.0],[-0.785867,1.02786,1.0],[-4.73124,1.00876,1.0],[-9.09856,0.955343,1.0],[-12.4324,0.875472,1.0],[0.0,1.0,1.0],[-0.444072,1.03863,1.0],[-3.94773,1.01784,1.0],[-7.97996,0.975996,1.0],[-10.673,0.917401,1.0],[0.0,1.0,1.0],[1.22651,1.0559,1.0],[-1.89459,1.02934,1.0],[-5.54545,0.993046,1.0],[-8.48877,0.941628,1.0],[0.0,1.0,1.0],[3.24516,1.08292,1.0],[1.20522,1.0577,1.0],[-0.826224,1.01668,1.0],[-3.71884,0.946733,1.0],[0.0,1.0,1.0],[4.68802,1.12039,1.0],[3.23147,1.08619,1.0],[1.84476,1.04,1.0],[0.0970375,0.962298,1.0],[0.0,1.0,1.0],[5.36041,1.16313,1.0],[4.5937,1.10933,1.0],[3.83314,1.05056,1.0],[3.122,0.959876,1.0],[0.0,1.0,1.0],[5.22222,1.17824,1.0],[5.09129,1.1142,1.0],[4.99294,1.04813,1.0],[5.01485,0.957528,1.0],[0.0,1.0,1.0],[4.34382,1.18446,1.0],[4.9457,1.11351,1.0],[5.56715,1.04462,1.0],[6.27579,0.961892,1.0],[0.0,1.0,1.0],[3.5428,1.18985,1.0],[4.9116,1.11394,1.0],[6.11507,1.04098,1.0],[7.03638,0.964331,1.0],[0.0,1.0,1.0],[2.92641,1.20663,1.0],[4.57132,1.1281,1.0],[5.96964,1.04895,1.0],[6.98579,0.96261,1.0],[0.0,1.0,1.0],[2.07123,1.23257,1.0],[3.51736,1.15454,1.0],[4.85051,1.06995,1.0],[5.97805,0.971597,1.0],[0.0,1.0,1.0],[0.869122,1.26762,1.0],[2.22009,1.19021,1.0],[3.54297,1.10536,1.0],[4.31783,1.00254,1.0],[0.0,1.0,1.0],[-0.409355,1.31042,1.0],[0.778039,1.23845,1.0],[1.87845,1.15455,1.0],[1.53759,1.02961,1.0],[0.0,1.0,1.0],[-1.82067,1.35265,1.0],[-0.910751,1.28551,1.0],[-0.100378,1.18585,1.0],[0.0425375,1.01463,1.0],[0.0,1.0,1.0],[-3.26659,1.391,1.0],[-2.50117,1.33263,1.0],[-1.55789,1.18102,1.0],[-0.0853484,1.00594,1.0],[0.0,1.0,1.0],[-4.92446,1.43632,1.0],[-3.92066,1.35798,1.0],[-1.66274,1.11605,1.0],[-0.0358972,1.00178,1.0],[0.0,1.0,1.0],[-5.82474,1.46137,1.0],[-4.22466,1.34716,1.0],[-0.64525,1.04813,1.0],[-0.00357121,1.00023,1.0],[0.0,1.0,1.0],[-7.0672,1.47242,1.0],[-5.38667,1.32663,1.0],[-0.553542,1.02668,1.0],[-0.00188478,1.00008,1.0],[0.0,1.0,1.0],[-8.37522,1.43136,1.0],[-6.13631,1.28174,1.0],[-0.515332,1.01808,1.0],[-0.0015762,1.00005,1.0],[0.0,1.0,1.0],[-10.5196,1.39935,1.0],[-7.49316,1.24442,1.0],[-0.517621,1.01282,1.0],[-0.001445,1.00003,1.0],[0.0,1.0,1.0],[-10.5952,1.3534,1.0],[-7.81548,1.20414,1.0],[-0.63473,1.0111,1.0],[-0.00208267,1.00003,1.0],[0.0,1.0,1.0],[-10.879,1.32051,1.0],[-8.3374,1.18517,1.0],[-0.91817,1.01139,1.0],[-0.00474972,1.00003,1.0],[0.0,1.0,1.0],[-11.8094,1.29525,1.0],[-9.18818,1.17522,1.0],[-0.968552,1.01157,1.0],[-0.00411555,1.00003,1.0],[0.0,1.0,1.0],[-12.2038,1.27777,1.0],[-8.97374,1.15856,1.0],[-0.751018,1.00872,1.0],[-0.00250012,1.00002,1.0],[0.0,1.0,1.0],[-12.0564,1.2709,1.0],[-8.38288,1.15926,1.0],[-0.643806,1.00878,1.0],[-0.00202184,1.00002,1.0],[0.0,1.0,1.0],[-12.8403,1.25618,1.0],[-8.85664,1.14517,1.0],[-0.646111,1.00756,1.0],[-0.00195373,1.00001,1.0],[0.0,1.0,1.0],[-13.2604,1.25094,1.0],[-8.74013,1.11989,1.0],[-0.556009,1.00457,1.0],[-0.00173404,1.00001,1.0],[0.0,1.0,1.0],[-13.3939,1.24388,1.0],[-8.23916,1.11659,1.0],[-0.376536,1.00355,1.0],[-0.000916476,1.00001,1.0],[0.0,1.0,1.0],[-13.6578,1.24212,1.0],[-7.97883,1.11035,1.0],[-0.409859,1.00287,1.0],[-0.0018203,1.0,1.0],[0.0,1.0,1.0],[-15.7225,1.22414,1.0],[-10.2345,1.10217,1.0],[-0.968177,1.00427,1.0],[-0.00650675,1.00001,1.0],[0.0,1.0,1.0],[-16.1557,1.16423,1.0],[-12.721,1.06603,1.0],[-2.02337,1.00334,1.0],[-0.0134173,1.00001,1.0],[0.0,1.0,1.0],[-17.253,1.09714,1.0],[-15.785,1.02184,1.0],[-4.02264,0.996649,1.0],[-0.0279292,0.999956,1.0],[0.0,1.0,1.0],[-17.6206,1.04331,1.0],[-16.2039,0.99401,1.0],[-4.09728,0.990246,1.0],[-0.0269801,0.999896,1.0],[0.0,1.0,1.0],[-17.2302,0.987569,1.0],[-15.1656,0.964114,1.0],[-3.21398,0.983725,1.0],[-0.0202065,0.999807,1.0],[0.0,1.0,1.0],[-16.2184,0.943678,1.0],[-13.4633,0.938141,1.0],[-3.40639,0.971571,1.0],[-0.00923182,0.999563,1.0],[0.0,1.0,1.0],[-14.5653,0.909699,1.0],[-11.6873,0.915733,1.0],[-5.4654,0.942337,1.0],[-0.124331,0.998194,1.0],[0.0,1.0,1.0],[-11.5122,0.906891,1.0],[-9.31547,0.923874,1.0],[-6.12222,0.939832,1.0],[-0.337538,0.995963,1.0],[0.0,1.0,1.0],[-7.77619,0.92294,1.0]Line truncated
@@ -0,0 +1,4 @@
+allow-unwrap-in-tests = true
+allow-expect-in-tests = true
+allow-panic-in-tests = true
+allow-indexing-slicing-in-tests = true
@@ -0,0 +1,466 @@
+{
+  "schema": 1,
+  "repo": "storytold/lightcraft",
+  "branch": "main",
+  "head": "2694d7089fbf4ada62c6738a7c62fecdfd5711fb",
+  "head_date": "2026-10-08T00:53:19Z",
+  "commits": 424,
+  "unattributed_commits": 0,
+  "contributors": [
+    {
+      "login": "echelon",
+      "people_entry": true,
+      "github_name": null,
+      "prs": 71,
+      "commits": 330,
+      "lines_added": 125623,
+      "lines_deleted": 8643,
+      "binary_added": 38,
+      "binary_deleted": 8,
+      "first_commit": "2026-09-30T14:04:12Z",
+      "last_commit": "2026-10-08T00:32:34Z"
+    },
+    {
+      "login": "leipsfur",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 131,
+      "lines_deleted": 5,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-02T15:51:58Z",
+      "last_commit": "2026-10-02T15:51:58Z"
+    },
+    {
+      "login": "GlemNL",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 9,
+      "commits": 13,
+      "lines_added": 1233,
+      "lines_deleted": 96,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-02T21:16:29Z",
+      "last_commit": "2026-10-04T09:42:39Z"
+    },
+    {
+      "login": "storybored379",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 7,
+      "commits": 7,
+      "lines_added": 1194,
+      "lines_deleted": 82,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-04T00:28:49Z",
+      "last_commit": "2026-10-06T06:51:43Z"
+    },
+    {
+      "login": "midasdf",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 424,
+      "lines_deleted": 25,
+      "binary_added": 1,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-05T16:07:26Z",
+      "last_commit": "2026-10-05T16:07:26Z"
+    },
+    {
+      "login": "voshart",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 3,
+      "commits": 12,
+      "lines_added": 2481,
+      "lines_deleted": 1389,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-05T16:23:27Z",
+      "last_commit": "2026-10-07T16:07:24Z"
+    },
+    {
+      "login": "arata-1972",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 2,
+      "commits": 4,
+      "lines_added": 2892,
+      "lines_deleted": 726,
+      "binary_added": 2,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-05T20:14:35Z",
+      "last_commit": "2026-10-05T22:16:15Z"
+    },
+    {
+      "login": "benlukka",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 49,
+      "lines_deleted": 4,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-06T11:18:03Z",
+      "last_commit": "2026-10-06T11:18:03Z"
+    },
+    {
+      "login": "kywk",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 72,
+      "lines_deleted": 8,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-06T20:21:14Z",
+      "last_commit": "2026-10-06T20:21:14Z"
+    },
+    {
+      "login": "treyreynolds",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 159,
+      "lines_deleted": 27,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-06T22:21:46Z",
+      "last_commit": "2026-10-06T22:21:46Z"
+    },
+    {
+      "login": "kovalskiguinther",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 0,
+      "commits": 1,
+      "lines_added": 4328,
+      "lines_deleted": 17,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-06T23:15:04Z",
+      "last_commit": "2026-10-06T23:15:04Z"
+    },
+    {
+      "login": "cruzer45",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 2,
+      "commits": 3,
+      "lines_added": 658,
+      "lines_deleted": 26,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T01:30:23Z",
+      "last_commit": "2026-10-07T03:53:09Z"
+    },
+    {
+      "login": "coygeek",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 4,
+      "lines_deleted": 0,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T02:27:15Z",
+      "last_commit": "2026-10-07T02:27:15Z"
+    },
+    {
+      "login": "lzjtugjc",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 1895,
+      "lines_deleted": 294,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T03:34:33Z",
+      "last_commit": "2026-10-07T03:34:33Z"
+    },
+    {
+      "login": "itsmunzir",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 46,
+      "lines_deleted": 30,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T07:40:55Z",
+      "last_commit": "2026-10-07T07:40:55Z"
+    },
+    {
+      "login": "haraldh",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 2,
+      "commits": 5,
+      "lines_added": 1163,
+      "lines_deleted": 141,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T07:54:27Z",
+      "last_commit": "2026-10-07T13:16:08Z"
+    },
+    {
+      "login": "sakamoto824",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 2,
+      "lines_added": 3649,
+      "lines_deleted": 167,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T09:32:12Z",
+      "last_commit": "2026-10-07T10:54:19Z"
+    },
+    {
+      "login": "anarion80",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 402,
+      "lines_deleted": 2,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T11:54:51Z",
+      "last_commit": "2026-10-07T11:54:51Z"
+    },
+    {
+      "login": "radialmonster",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 3,
+      "commits": 19,
+      "lines_added": 2630,
+      "lines_deleted": 259,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T12:33:51Z",
+      "last_commit": "2026-10-08T00:46:10Z"
+    },
+    {
+      "login": "bogusyogi",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 434,
+      "lines_deleted": 29,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T17:06:00Z",
+      "last_commit": "2026-10-07T17:06:00Z"
+    },
+    {
+      "login": "nledford",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 4,
+      "commits": 8,
+      "lines_added": 525,
+      "lines_deleted": 60,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T17:33:38Z",
+      "last_commit": "2026-10-07T22:49:00Z"
+    },
+    {
+      "login": "MusseJusse",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 34,
+      "lines_deleted": 3,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T18:38:23Z",
+      "last_commit": "2026-10-07T18:38:23Z"
+    },
+    {
+      "login": "Ndsanes",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 2,
+      "commits": 2,
+      "lines_added": 332,
+      "lines_deleted": 86,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T19:08:28Z",
+      "last_commit": "2026-10-07T19:08:33Z"
+    },
+    {
+      "login": "RicSchonfelder",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 1439,
+      "lines_deleted": 6,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T20:03:20Z",
+      "last_commit": "2026-10-07T20:03:20Z"
+    },
+    {
+      "login": "Gremiger",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 19,
+      "lines_deleted": 0,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T21:43:23Z",
+      "last_commit": "2026-10-07T21:43:23Z"
+    },
+    {
+      "login": "DWAA1660",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 54,
+      "lines_deleted": 12,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T22:08:01Z",
+      "last_commit": "2026-10-07T22:08:01Z"
+    },
+    {
+      "login": "agorevski",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 68,
+      "lines_deleted": 11,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-07T22:18:44Z",
+      "last_commit": "2026-10-07T22:18:44Z"
+    },
+    {
+      "login": "andrewyao",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 1,
+      "lines_added": 1,
+      "lines_deleted": 1,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": "2026-10-08T00:53:19Z",
+      "last_commit": "2026-10-08T00:53:19Z"
+    },
+    {
+      "login": "GuintherKovalski",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 0,
+      "lines_added": 0,
+      "lines_deleted": 0,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": null,
+      "last_commit": null
+    },
+    {
+      "login": "XRTce",
+      "people_entry": false,
+      "github_name": null,
+      "prs": 1,
+      "commits": 0,
+      "lines_added": 0,
+      "lines_deleted": 0,
+      "binary_added": 0,
+      "binary_deleted": 0,
+      "first_commit": null,
+      "last_commit": null
+    }
+  ],
+  "models": [
+    {
+      "id": "anthropic/claude-opus/5.5",
+      "company": "Anthropic",
+      "model": "Claude Opus",
+      "version": "5.5",
+      "commits": 364,
+      "lines_added": 131810,
+      "lines_deleted": 9367
+    },
+    {
+      "id": "anthropic/claude-sonnet/5.5",
+      "company": "Anthropic",
+      "model": "Claude Sonnet",
+      "version": "5.5",
+      "commits": 31,
+      "lines_added": 3941,
+      "lines_deleted": 1543
+    },
+    {
+      "id": "openai/codex",
+      "company": "OpenAI",
+      "model": "Codex",
+      "version": "",
+      "commits": 14,
+      "lines_added": 8341,
+      "lines_deleted": 995
+    },
+    {
+      "id": "anthropic/claude",
+      "company": "Anthropic",
+      "model": "Claude",
+      "version": "",
+      "commits": 2,
+      "lines_added": 1304,
+      "lines_deleted": 192
+    },
+    {
+      "id": "anthropic/claude-sonnet/5",
+      "company": "Anthropic",
+      "model": "Claude Sonnet",
+      "version": "5",
+      "commits": 2,
+      "lines_added": 298,
+      "lines_deleted": 2
+    },
+    {
+      "id": "anthropic/claude-opus/4.8",
+      "company": "Anthropic",
+      "model": "Claude Opus",
+      "version": "4.8",
+      "commits": 1,
+      "lines_added": 1633,
+      "lines_deleted": 3
+    },
+    {
+      "id": "deepseek/deepseek/v3.2",
+      "company": "DeepSeek",
+      "model": "DeepSeek",
+      "version": "V3.2",
+      "commits": 1,
+      "lines_added": 1895,
+      "lines_deleted": 294
+    }
+  ]
+}
@@ -0,0 +1,29 @@
+# Contributor names for the About window (craftrules standards/contributors.md).
+#
+# Everyone who contributes is credited by GitHub username automatically. Anything more is up to
+# you: add YOUR OWN entry below, in your own PR (commit it yourself). Never add someone else; a
+# maintainer may add a person only when that person asked, recording where in consent_source.
+#
+# Easiest: python3 ../../craftrules/scripts/contributors.py --add-me . \
+#            [--real-name "Ada Lovelace"] [--display-name "Ada"] [--sync-github-name]
+#
+# Or copy this block, replacing your-github-username (every field is optional):
+#
+#   [people.your-github-username]
+#   real_name = "Your Name"        # shown when the viewer picks "Real name"
+#   display_name = "Nickname"      # shown when the viewer picks "Display name"
+#   sync_github_name = true        # or: use the name on your public GitHub profile as display name
+#
+# To remove your names, delete your entry (or set fields empty) in a PR. The next
+# contributors.py run, and the next build, drop them.
+
+[people.echelon]
+real_name = "Brandon Thomas"
+display_name = "Brandon Thomas"
+sync_github_name = false
+
+[people.rock-meister]
+sync_github_name = true
+
+[people.sanyamsingh001]
+display_name = "sanyamsingh001"
@@ -10,11 +10,13 @@ workspace = true

 [dependencies]
 lightcraft-develop = { workspace = true }
+lightcraft-meta = { workspace = true }
 serde = { workspace = true }
 serde_json = { workspace = true }
 thiserror = { workspace = true }
 crc32fast = { workspace = true }
 log = { workspace = true }
+web-time = { workspace = true }

 [dev-dependencies]
 proptest = { workspace = true }
@@ -0,0 +1,216 @@
+//! Catalog persistence costs on large synthetic catalogs (issue #37): log append, compaction
+//! (snapshot serialise / write+fsync / log reset), synchronous and on the background worker, load,
+//! and a catalog clone.
+//!
+//! ```text
+//! cargo run --release -p lightcraft-catalog --example persist_bench -- <dir> [photos…]
+//! ```
+//!
+//! `<dir>` is a scratch directory (wiped per size). Default sizes: 1000 10000 85000. 90 % of the
+//! photos are Local browse records, the rest library photos with a few edits and history steps.
+//! Peak memory: run under `/usr/bin/time -l` (macOS) / `-v` (Linux) with one size;
+//! `PERSIST_BENCH_NO_LOAD=1` skips the final load, so the peak is building + compacting.
+
+use std::sync::Arc;
+use std::time::Instant;
+
+use lightcraft_catalog::*;
+use lightcraft_develop::DevelopSettings;
+
+fn ms(t: Instant) -> f64 {
+    t.elapsed().as_secs_f64() * 1e3
+}
+
+fn photo(c: &mut Catalog, i: u64) -> Photo {
+    let id = c.alloc_photo_id();
+    let mut p = Photo::new(
+        id,
+        Source::File { path: format!("/Users/someone/Pictures/2026/09/{:05}/IMG_{i:05}.CR3", i / 500) },
+        &format!("IMG_{i:05}.CR3"),
+        "CR3",
+        6000,
+        4000,
+        "2026-09-30T12:00:00",
+    );
+    p.kind = MediaKind::Raw;
+    p.file_size = 25_000_000 + i;
+    p.captured = Some(format!("2026-09-{:02}T10:{:02}:{:02}", 1 + i % 28, i % 60, (i / 60) % 60));
+    p.content_hash = Some(format!("{:032x}", i.wrapping_mul(0x9e37_79b9_7f4a_7c15)));
+    p.meta.camera = "Camera Model X".into();
+    p.meta.lens = "24-70mm f/2.8".into();
+    p.meta.focal_mm = Some(35.0);
+    p.meta.aperture = Some(4.0);
+    p.meta.shutter = "1/250".into();
+    p.meta.iso = Some(400);
+    p.as_shot_wb = Some((5200.0, 3.0));
+    p.local = !i.is_multiple_of(10);
+    if p.local {
+        // as a browse catalogues it: import defaults, a baseline fingerprint
+        p.develop = Arc::new(p.import_defaults());
+        p.set_local_baseline();
+    }
+    if !p.local {
+        let mut d = DevelopSettings::default();
+        d.light.exposure = (i % 7) as f64 * 0.1;
+        let s = Arc::new(d);
+        p.develop = s.clone();
+        p.edited = Some("2026-09-30T12:00:00".into());
+        p.history = vec![HistoryStep { label: "Exposure".into(), settings: s }];
+        p.rating = (i % 6) as u8;
+    }
+    p
+}
+
+fn stats(mut v: Vec<f64>) -> String {
+    v.sort_by(f64::total_cmp);
+    let pick = |q: f64| v.get(((v.len() as f64 - 1.0) * q).round() as usize).copied().unwrap_or(0.0);
+    format!("median {:.2} ms, p95 {:.2} ms, max {:.2} ms", pick(0.5), pick(0.95), pick(1.0))
+}
+
+fn main() -> std::result::Result<(), Box<dyn std::error::Error>> {
+    let mut args = std::env::args().skip(1);
+    let dir = std::path::PathBuf::from(args.next().ok_or("usage: persist_bench <scratch dir> [photos…]")?);
+    let mut sizes: Vec<u64> = args.map(|a| a.parse()).collect::<std::result::Result<_, _>>()?;
+    if sizes.is_empty() {
+        sizes = vec![1000, 10_000, 85_000];
+    }
+    for n in sizes {
+        let _ = std::fs::remove_dir_all(&dir);
+        let (mut j, _, _) = Journal::open(Box::new(FsStore::open(&dir)?))?;
+        let mut cat = Catalog::new();
+        // bulk-add (as one import would) and log it, so the first compaction is an import's
+        for chunk in (0..n).collect::<Vec<_>>().chunks(500) {
+            let ops: Vec<Op> = chunk.iter().map(|i| Op::AddPhoto { photo: Box::new(photo(&mut cat, *i)) }).collect();
+            for op in &ops {
+                cat.apply(op.clone())?;
+            }
+            j.append(&ops)?;
+        }
+        j.snapshot(&cat)?;
+        let first = j.stats().last_snapshot;
+        println!("== {n} photos ({} local): snapshot {:.1} MB", cat.photos().filter(|p| p.local).count(), first.bytes as f64 / 1e6);
+
+        let t = Instant::now();
+        let copy = cat.clone();
+        println!("  catalog clone: {:.2} ms", ms(t));
+        drop(copy);
+
+        // single-op appends (a rating change each), as commands do
+        let ids: Vec<PhotoId> = cat.photos().map(|p| p.id).take(1000).collect();
+        let mut appends = Vec::new();
+        for k in 0..300usize {
+            let id = ids.get(k % ids.len().max(1)).copied().unwrap_or(PhotoId(1));
+            let op = Op::SetRating { id, rating: (k % 6) as u8 };
+            cat.apply(op.clone())?;
+            j.append(std::slice::from_ref(&op))?;
+            appends.push(j.stats().last_append_ms);
+        }
+        println!("  append (1 op, fsync): {}", stats(appends));
+
+        // compactions at the default threshold crossing
+        let mut totals = Vec::new();
+        for _ in 0..3 {
+            while !j.wants_snapshot() {
+                let op = Op::SetRating { id: ids.first().copied().unwrap_or(PhotoId(1)), rating: (j.seq() % 6) as u8 };
+                cat.apply(op.clone())?;
+                j.append(&[op])?;
+            }
+            j.snapshot(&cat)?;
+            let s = j.stats().last_snapshot;
+            println!(
+                "  compaction ({} records): serialize {:.1} ms, write+sync {:.1} ms, log reset {:.1} ms, total {:.1} ms",
+                s.records, s.serialize_ms, s.write_sync_ms, s.reset_ms, s.total_ms
+            );
+            totals.push(s.total_ms);
+        }
+        println!("  compaction total: {}", stats(totals));
+
+        // the same at the threshold, written by the worker while single-op appends go on
+        let mut blocking = Vec::new();
+        for _ in 0..3 {
+            while !j.wants_snapshot() {
+                let op = Op::SetRating { id: ids.first().copied().unwrap_or(PhotoId(1)), rating: (j.seq() % 6) as u8 };
+                cat.apply(op.clone())?;
+                j.append(&[op])?;
+            }
+            let t = Instant::now();
+            j.snapshot_in_background(&cat)?;
+            let start_ms = ms(t);
+            let mut during = Vec::new();
+            while !j.snapshot_written() {
+                let op = Op::SetRating { id: ids.get(1).copied().unwrap_or(PhotoId(1)), rating: (j.seq() % 6) as u8 };
+                cat.apply(op.clone())?;
+                j.append(&[op])?;
+                during.push(j.stats().last_append_ms);
+            }
+            let t = Instant::now();
+            j.poll()?;
+            let finish_ms = ms(t);
+            let s = j.stats().last_snapshot;
+            println!(
+                "  background compaction ({} records): start {start_ms:.1} ms + finish {finish_ms:.1} ms blocking; worker {:.1} ms; {} appends during it: {}",
+                s.records,
+                s.total_ms,
+                during.len(),
+                stats(during)
+            );
+            blocking.push(s.blocking_ms);
+        }
+        println!("  background compaction blocking: {}", stats(blocking));
+
+        // forget the untouched Local records (folders last browsed long ago; every 7th folder
+        // recently; every 50th Local record rated), then compact
+        let folders: std::collections::BTreeSet<String> = cat.photos().filter(|p| p.local).filter_map(|p| folder_of(p)).collect();
+        let mut ops: Vec<Op> = folders
+            .iter()
+            .enumerate()
+            .map(|(k, f)| Op::SetBrowsed {
+                folder: f.clone(),
+                at: Some(if k % 7 == 0 { "2026-09-30T00:00:00" } else { "2026-06-01T00:00:00" }.into()),
+            })
+            .collect();
+        ops.extend(cat.photos().filter(|p| p.local && p.id.0.is_multiple_of(50)).map(|p| Op::SetRating { id: p.id, rating: 2 }).collect::<Vec<_>>());
+        for op in &ops {
+            cat.apply(op.clone())?;
+        }
+        j.append(&ops)?;
+        j.snapshot(&cat)?;
+        let before = j.stats().last_snapshot.bytes;
+        let t = Instant::now();
+        let plan = cat.forget_local_plan("2026-10-05T00:00:00", DEFAULT_FORGET_DAYS, &|_| false);
+        let plan_ms = ms(t);
+        let t = Instant::now();
+        let ops = plan.ops();
+        for op in &ops {
+            cat.apply(op.clone())?;
+        }
+        j.append(&ops)?;
+        let apply_ms = ms(t);
+        j.snapshot(&cat)?;
+        let after = j.stats().last_snapshot.bytes;
+        println!(
+            "  forget Local: {} of {} local records forgotten (kept: {} recent, {} touched, {} in use); plan {plan_ms:.1} ms, apply+append {apply_ms:.1} ms; snapshot {:.1} MB -> {:.1} MB",
+            plan.evict.len(),
+            plan.local,
+            plan.kept_recent,
+            plan.kept_touched,
+            plan.kept_in_use,
+            before as f64 / 1e6,
+            after as f64 / 1e6
+        );
+
+        drop(j);
+        if std::env::var_os("PERSIST_BENCH_NO_LOAD").is_some() {
+            // peak memory of building + compacting only
+            continue;
+        }
+        let t = Instant::now();
+        let (_, loaded, _) = Journal::open(Box::new(FsStore::open(&dir)?))?;
+        println!("  open (load snapshot + replay): {:.1} ms", ms(t));
+        if loaded.len() != cat.len() {
+            return Err("reloaded catalog differs".into());
+        }
+    }
+    let _ = std::fs::remove_dir_all(&dir);
+    Ok(())
+}
@@ -0,0 +1,8 @@
+# Seeds for failure cases proptest has generated in the past. It is
+# automatically read and these particular cases re-run before any
+# novel cases are generated.
+#
+# It is recommended to check this file in to source control so that
+# everyone who runs the test benefits from these saved cases.
+cc 0227748e171237a13833c9175fe02c7fe6ff332266798961a6a43ec37b81d1b4 # shrinks to actions = [(3, 0), (0, 0), (4, 0), (0, 0)]
+cc 0b0ab2531bd46fbe261b55ecffebea9d3e8949900ab7bc33fb27ca19f5cd7b8b # shrinks to actions = [(4, 443), (3, 474), (2, 27), (5, 858), (5, 466), (0, 901), (4, 907), (3, 839), (0, 630), (7, 578), (4, 13), (2, 361), (6, 136), (6, 12), (6, 223), (1, 394), (6, 78), (5, 837), (6, 765), (4, 207), (2, 185), (2, 583), (2, 56), (5, 430), (1, 702), (1, 547), (5, 552), (4, 762), (0, 443), (0, 851), (7, 240), (4, 438), (3, 599), (6, 70), (4, 694), (1, 327), (7, 859), (6, 814), (4, 336), (3, 438), (0, 131), (0, 831), (3, 18), (2, 963), (6, 10), (3, 114)]
@@ -128,13 +128,15 @@ pub fn display_time(iso: &str) -> String {

 /// Header text for a group key.
 pub fn group_label(key: &str) -> String {
-    let month = |k: &str| k.get(5..7).and_then(|m| m.parse::<usize>().ok()).filter(|m| (1..=12).contains(m)).map(|m| MONTHS[m - 1]);
+    let month = |k: &str| k.get(5..7).and_then(|m| m.parse::<usize>().ok()).and_then(|m| MONTHS.get(m.wrapping_sub(1)));
+    // `get` rather than `[..]`: a non-ASCII key must not split a character
+    let year = key.get(..4).unwrap_or(key);
    match key.len() {
-        10 => match (weekday(key), month(key), key[8..10].parse::<u32>().ok()) {
-            (Some(w), Some(m), Some(d)) => format!("{w}, {d} {m} {}", &key[..4]),
+        10 => match (weekday(key), month(key), key.get(8..10).and_then(|d| d.parse::<u32>().ok())) {
+            (Some(w), Some(m), Some(d)) => format!("{w}, {d} {m} {year}"),
            _ => key.to_string(),
        },
-        7 => month(key).map(|m| format!("{m} {}", &key[..4])).unwrap_or_else(|| key.to_string()),
+        7 => month(key).map(|m| format!("{m} {year}")).unwrap_or_else(|| key.to_string()),
        4 => key.to_string(),
        _ => "Unknown Date".into(),
    }
@@ -153,7 +155,7 @@ impl Catalog {
        let date_of = |id: &PhotoId| -> String {
            let Some(p) = self.photo(*id) else { return String::new() };
            let d = match key {
-                SortKey::CaptureDate => Some(p.date()),
+                SortKey::CaptureDate => p.captured.as_deref(),
                SortKey::ImportDate => Some(p.imported.as_str()),
                SortKey::EditDate => p.edited.as_deref(),
                _ => None,
@@ -196,6 +198,9 @@ mod tests {
        assert_eq!(group_label("2026-02"), "February 2026");
        assert_eq!(group_label("2026"), "2026");
        assert_eq!(group_label(""), "Unknown Date");
+        // non-ASCII keys of a date's length used to be sliced mid-character
+        assert_eq!(group_label("2024-01é0"), "2024-01é0");
+        assert_eq!(group_label("202é01"), "January 202é01");
        assert_eq!(shift_iso("2026-12-31T23:30:00", 3600).as_deref(), Some("2027-01-01T00:30:00"));
        assert_eq!(shift_iso("2026-03-01T00:00:00.25+02:00", -1).as_deref(), Some("2026-02-28T23:59:59.25+02:00"));
        assert_eq!(shift_iso("nope", 5), None);
@@ -217,7 +222,7 @@ mod tests {
        }
        let shape = |r: &[DateRun]| r.iter().map(|r| (r.key.clone(), r.start, r.count)).collect::<Vec<_>>();
        let day = c.date_runs(&ids, SortKey::CaptureDate, GroupBy::Day);
-        // the undated photo falls back to its import date
+        // the undated photo is grouped under Unknown Date
        assert_eq!(
            shape(&day),
            vec![
@@ -225,16 +230,20 @@ mod tests {
                ("2026-09-29".into(), 2, 1),
                ("2026-08-01".into(), 3, 1),
                ("2025-12-31".into(), 4, 1),
-                ("2026-10-01".into(), 5, 1)
+                (String::new(), 5, 1)
            ]
        );
        assert_eq!(day[0].label, "Wednesday, 30 September 2026");
+        assert_eq!(day[4].label, "Unknown Date");
        assert_eq!(c.date_runs(&ids, SortKey::CaptureDate, GroupBy::Auto), day);
        assert_eq!(
-            shape(&c.date_runs(&ids[..5], SortKey::CaptureDate, GroupBy::Month)),
-            vec![("2026-09".into(), 0, 3), ("2026-08".into(), 3, 1), ("2025-12".into(), 4, 1)]
+            shape(&c.date_runs(&ids, SortKey::CaptureDate, GroupBy::Month)),
+            vec![("2026-09".into(), 0, 3), ("2026-08".into(), 3, 1), ("2025-12".into(), 4, 1), (String::new(), 5, 1)]
+        );
+        assert_eq!(
+            shape(&c.date_runs(&ids, SortKey::CaptureDate, GroupBy::Year)),
+            vec![("2026".into(), 0, 4), ("2025".into(), 4, 1), (String::new(), 5, 1)]
        );
-        assert_eq!(shape(&c.date_runs(&ids[..5], SortKey::CaptureDate, GroupBy::Year)), vec![("2026".into(), 0, 4), ("2025".into(), 4, 1)]);
        assert!(c.date_runs(&ids, SortKey::FileName, GroupBy::Day).is_empty());
        assert!(c.date_runs(&ids, SortKey::CaptureDate, GroupBy::None).is_empty());
        // edit date: unedited photos have no date
@@ -0,0 +1,226 @@
+//! The library's folder tree: where on disk the imported photos live.
+//!
+//! A *folder* here is a folder holding at least one photo of the library (imported, not deleted,
+//! not merely browsed in Local; demo scenes have no folder). The tree lists those folders and the
+//! ones above them, never a folder that holds nothing of the library, with how many photos each
+//! holds — its subfolders included, like a year in By Date counts its months.
+//!
+//! Every folder is a row of its own, the ones above the imported photos included
+//! (`/Users` → `me` → `Pictures`), so a row is exactly the folder it names. A folder spelled two
+//! ways (`/` or `\`, repeated or trailing separators, `.`) is one folder (see
+//! [`crate::query::folder_key`]).
+//!
+//! Each disk is a *volume*, a top-level row of its own: a drive letter (`D:`), a Windows share
+//! (`//server/share`), a mounted disk or share (macOS `/Volumes/tokyo`, Linux `/media/me/usb`,
+//! `/mnt/disk`, `/run/user/1000/gvfs/smb-share:…`, `/net/host`) or, for everything else, the
+//! startup disk (`/`). Only the spelling of the stored paths tells them apart: no disk is
+//! touched, so a share that is asleep or gone lists and counts like one that is there, and the
+//! sidebar never waits on the network.
+//!
+//! Limits of reading only the stored spelling of a path: file names are compared as written
+//! (Windows folds case, other systems do not), so on a case-insensitive disk `/Volumes/tokyo` and
+//! `/Volumes/TOKYO` are two volumes here; and one place reached by two spellings
+//! (`/Volumes/Macintosh HD/Users/me` and `/Users/me`, `/media/me/usb` and `/run/media/me/usb`) is
+//! two rows. Each row is still consistent: choosing it shows exactly the photos it counts.
+//!
+//! This is the library's own view of its photos; browsing any folder on disk without importing
+//! is Local's job (see [`crate::local`]).
+
+use std::collections::{BTreeMap, HashMap};
+
+use serde::Serialize;
+
+use crate::query::{folder_key, folder_within, split_path};
+use crate::{Catalog, Source};
+
+/// Folder levels kept per photo. No real path comes near it; a deeper one is counted at its
+/// ancestor this many levels down, which bounds the tree's depth (and so every walk of it)
+/// whatever a catalog holds.
+pub const MAX_DEPTH: usize = 64;
+
+/// One row of the library's folder tree.
+#[derive(Clone, Debug, PartialEq, Serialize)]
+pub struct FolderNode {
+    /// What the row says: the folder's name; for a volume, its name (`nas`, `D:`,
+    /// `//server/share`, `/` for the startup disk).
+    pub name: String,
+    /// The folder the row stands for, written out from the names of the photos' paths (forward
+    /// slashes, no verbatim prefix); hand it back to [`crate::Filter::library_folder`] to see its
+    /// photos.
+    pub path: String,
+    /// Photos in this folder or any folder inside it.
+    pub count: usize,
+    /// Photos directly in this folder.
+    pub own: usize,
+    /// A disk rather than a folder in one; its `path` is where it is mounted.
+    pub volume: bool,
+    /// Whether choosing the row (`Filter::library_folder`) shows exactly its `count` photos. Not
+    /// so for the startup disk and for folders that hold other disks (`/Volumes`, `/mnt`…): their
+    /// path covers those disks' photos too, so choosing them is not offered.
+    pub selectable: bool,
+    pub children: Vec<FolderNode>,
+}
+
+/// A folder while the tree is being built (children by identity key).
+struct Level {
+    name: String,
+    path: String,
+    own: usize,
+    volume: bool,
+    children: Vec<String>,
+}
+
+/// Where a stored folder path lies: its volume and the folder names below it.
+struct Placed {
+    mount: String,
+    volume: String,
+    names: Vec<String>,
+}
+
+/// A gvfs mount's name made readable: `smb-share:server=tokyo,share=photos` → `photos on tokyo`;
+/// any other (`sftp:host=nas`) as it is.
+fn mount_name(n: &str) -> String {
+    let field = |key: &str| {
+        let rest = n.strip_prefix("smb-share:")?;
+        rest.split(',').find_map(|kv| kv.strip_prefix(key)?.strip_prefix('=')).filter(|v| !v.is_empty())
+    };
+    match (field("share"), field("server")) {
+        (Some(share), Some(server)) => format!("{share} on {server}"),
+        _ => n.to_string(),
+    }
+}
+
+/// The volume `dir` lies on and the folder names below it, read the way [`folder_key`] reads a
+/// path (so a row and the filter that chooses it never disagree). `None` for a path that is not
+/// absolute or names no share.
+fn place(dir: &str) -> Option<Placed> {
+    let sp = split_path(dir, false);
+    let names: Vec<&str> = sp.parts.iter().map(String::as_str).collect();
+    let named = |mount: String, volume: &str, used: usize| (mount, volume.to_string(), used);
+    let (mount, volume, used) = if sp.prefix.starts_with("//") {
+        // `//server/share`: a server alone (no share) is no place to keep photos
+        if sp.prefix.ends_with('/') {
+            return None;
+        }
+        named(sp.prefix.clone(), &sp.prefix, 0)
+    } else if !sp.prefix.is_empty() {
+        named(sp.prefix.clone(), &sp.prefix, 0)
+    } else if sp.absolute {
+        match names.as_slice() {
+            ["Volumes", n, ..] => named(format!("/Volumes/{n}"), n, 2),
+            ["run", "media", u, n, ..] => named(format!("/run/media/{u}/{n}"), n, 4),
+            ["media", u, n, ..] => named(format!("/media/{u}/{n}"), n, 3),
+            ["mnt", n, ..] => named(format!("/mnt/{n}"), n, 2),
+            // Linux desktops mount network places per user, naming server and share
+            ["run", "user", uid, "gvfs", n, ..] => named(format!("/run/user/{uid}/gvfs/{n}"), &mount_name(n), 5),
+            // NFS automounts: Linux `/net/<host>`, macOS `/Network/Servers/<host>`
+            ["net", host, ..] => named(format!("/net/{host}"), host, 2),
+            ["Network", "Servers", host, ..] => named(format!("/Network/Servers/{host}"), host, 3),
+            _ => named("/".to_string(), "/", 0),
+        }
+    } else {
+        return None;
+    };
+    Some(Placed { mount, volume, names: names.iter().skip(used).take(MAX_DEPTH).map(|n| n.to_string()).collect() })
+}
+
+/// The disk or share (`/Volumes/nas`, `D:`, `/` for the startup disk) a file-backed photo is on,
+/// from its stored path alone.
+pub fn volume_of(p: &crate::Photo) -> Option<String> {
+    let Source::File { path } = &p.source else { return None };
+    let dir = path.rfind(['/', '\\']).and_then(|i| path.get(..i.max(1)))?;
+    place(dir).map(|pl| folder_key(&pl.mount))
+}
+
+/// Whether `path` is a whole disk or share (`/`, `C:\`, `\\srv\share`, `/Volumes/nas`), not a
+/// folder in one.
+pub fn is_disk_root(path: &str) -> bool {
+    place(path).is_some_and(|p| p.names.is_empty())
+}
+
+/// Whether `path` is the startup disk's root (`/`), whose path also covers every other disk.
+pub fn is_startup_disk(path: &str) -> bool {
+    place(path).is_some_and(|p| p.names.is_empty() && p.mount == "/")
+}
+
+/// A folder named by its last two names (`photos/travel`), so two folders called `travel` are
+/// told apart; the path itself when it has no names (`/`).
+pub fn folder_label(path: &str) -> String {
+    let names: Vec<&str> = path.split(['/', '\\']).filter(|n| !n.is_empty()).collect();
+    if names.is_empty() { path.to_string() } else { names.iter().skip(names.len().saturating_sub(2)).copied().collect::<Vec<_>>().join("/") }
+}
+
+impl Catalog {
+    /// The library's volumes with their folders and photo counts: volumes by name
+    /// (case-insensitive), each folder with its subfolders the same way.
+    pub fn folder_tree(&self) -> Vec<FolderNode> {
+        // photos per containing folder, as spelled (sorted, so the spelling that names a folder
+        // never depends on hash order)
+        let mut dirs: BTreeMap<&str, usize> = BTreeMap::new();
+        for p in self.photos().filter(|p| p.in_library()) {
+            if let Source::File { path } = &p.source
+                && let Some(i) = path.rfind(['/', '\\'])
+                // a photo straight in the root has an empty folder part: the root itself
+                && let Some(dir) = path.get(..i.max(1))
+            {
+                *dirs.entry(dir).or_default() += 1;
+            }
+        }
+        let mut levels: HashMap<String, Level> = HashMap::new();
+        let mut roots: Vec<String> = Vec::new();
+        for (dir, n) in dirs {
+            let Some(placed) = place(dir) else { continue };
+            let vkey = folder_key(&placed.mount);
+            if !levels.contains_key(&vkey) {
+                levels.insert(
+                    vkey.clone(),
+                    Level { name: placed.volume.clone(), path: placed.mount.clone(), own: 0, volume: true, children: Vec::new() },
+                );
+                roots.push(vkey.clone());
+            }
+            let mut prefix = if placed.mount == "/" { String::new() } else { placed.mount.clone() };
+            let mut parent = vkey;
+            for name in &placed.names {
+                prefix.push('/');
+                prefix.push_str(name);
+                let key = folder_key(&prefix);
+                if !levels.contains_key(&key) {
+                    levels.insert(key.clone(), Level { name: name.clone(), path: prefix.clone(), own: 0, volume: false, children: Vec::new() });
+                    if let Some(l) = levels.get_mut(&parent) {
+                        l.children.push(key.clone());
+                    }
+                }
+                parent = key;
+            }
+            if let Some(l) = levels.get_mut(&parent) {
+                l.own += n;
+            }
+        }
+        let mut tree: Vec<FolderNode> = roots.iter().filter_map(|k| build(k, &levels)).collect();
+        sort(&mut tree);
+        // a folder that holds another disk's mount covers that disk's photos too
+        let mounts: Vec<String> = tree.iter().filter(|v| v.path != "/").map(|v| v.path.clone()).collect();
+        mark_selectable(&mut tree, &mounts);
+        tree
+    }
+}
+
+/// The node for `key` with everything below it (at most [`MAX_DEPTH`] deep).
+fn build(key: &str, levels: &HashMap<String, Level>) -> Option<FolderNode> {
+    let l = levels.get(key)?;
+    let mut children: Vec<FolderNode> = l.children.iter().filter_map(|k| build(k, levels)).collect();
+    sort(&mut children);
+    let count = l.own + children.iter().map(|c| c.count).sum::<usize>();
+    Some(FolderNode { name: l.name.clone(), path: l.path.clone(), count, own: l.own, volume: l.volume, selectable: true, children })
+}
+
+fn mark_selectable(nodes: &mut [FolderNode], mounts: &[String]) {
+    for n in nodes {
+        n.selectable = !(n.volume && n.path == "/") && (n.volume || !mounts.iter().any(|m| folder_within(m, &n.path)));
+        mark_selectable(&mut n.children, mounts);
+    }
+}
+
+fn sort(v: &mut [FolderNode]) {
+    v.sort_by(|a, b| a.name.to_lowercase().cmp(&b.name.to_lowercase()).then_with(|| a.name.cmp(&b.name)));
+}
@@ -1,8 +1,9 @@
 //! Crash-safe catalog persistence: an append-only op log plus periodic snapshots.
 //!
 //! Files (in a [`Store`]):
-//! - `catalog.snap` — `{"format":"lightcraft-catalog","version":1,"seq":N,"catalog":{…}}`, the
-//!   state after op `N`; replaced atomically (temp + fsync + rename).
+//! - `catalog.snap` — `{"format":"lightcraft-catalog","version":V,"seq":N,"catalog":{…}}`, the
+//!   state after op `N` in catalog format `V` ([`VERSION`]); replaced atomically (temp + fsync +
+//!   rename).
 //! - `catalog.log` — JSON lines, one per op applied after the snapshot:
 //!   `{"seq":N,"crc":C,"op":{…}}` where `C` is the CRC-32 of the op's JSON text. Appends are
 //!   fsynced before [`Journal::append`] returns.
@@ -10,10 +11,56 @@
 //! Loading = snapshot + replay of the records with `seq > snapshot seq`. Recovery rules:
 //! - a **torn final record** (crash mid-append: truncated line or bad CRC at the end) is dropped
 //!   and the file is cut back to the last good record, so later appends start on a clean line;
+//! - a **failed append** (disk full, a dropped network share) may leave part of its batch in the
+//!   file: [`Journal::append`] cuts the log back to the last whole record before returning the
+//!   error (or, if that fails too, before the next append), so the retry — the same ops with the
+//!   same `seq`s — starts on a clean line. Logs written by builds without that cut-back hold the
+//!   fragment glued to the first retried record on one line; the loader recognises the whole,
+//!   CRC-valid record at the end of such a line and skips only the fragment;
 //! - records already covered by the snapshot (crash between writing the snapshot and resetting
 //!   the log) are skipped;
 //! - a bad record **followed by good ones** is real damage: replay stops there, the log is kept
 //!   as `catalog.log.damaged-<seq>`, and a fresh snapshot is written so the library stays usable.
+//!
+//! **Background compaction** ([`Journal::snapshot_in_background`], when the store offers a
+//! [`Store::background_writer`]): a worker thread writes the snapshot of a copy of the catalog at
+//! op `N` while appends go on (durably, as always) to the same log. Once the snapshot is durable,
+//! [`Journal::poll`] atomically replaces the log with the records appended since the snapshot
+//! started (`seq > N`, kept in memory). Every crash point is covered by the rules above:
+//! - before the rename of `catalog.snap`: the old snapshot + the whole log (nothing was removed);
+//! - between the rename and the log rewrite: the new snapshot + the whole log, whose records
+//!   `<= N` are skipped as stale;
+//! - after the log rewrite: the new snapshot + exactly the records `> N`.
+//!
+//! At most one snapshot is in flight; a synchronous [`Journal::snapshot`] and dropping the journal
+//! first wait for it, so an older snapshot can never replace a newer one.
+//!
+//! # Format versions
+//!
+//! The snapshot's `version` is the catalog format, and it covers the log beside it: the log has
+//! no header of its own, and is only ever appended to under a snapshot of this build's format.
+//!
+//! | Version | Written by | Adds |
+//! |---|---|---|
+//! | (none) | early builds: a log without a snapshot | |
+//! | 1 | up to v0.2.0 | |
+//! | 2 | after v0.2.0 | `Op::SetBrowsed`, `Catalog.browsed`, `Photo.local_baseline` |
+//! | 3 | albums by hand | `Op::SetAlbumOrder`, `Album.order` |
+//!
+//! Rules:
+//! - **Bump [`VERSION`]** (and add a row above) in the change that adds an [`Op`] variant or a
+//!   serialized field to the catalog, a photo, an album, a stack or develop settings that an older
+//!   build would misread or silently drop. New fields must deserialize from older data
+//!   (`#[serde(default)]`), so newer builds keep reading every older format.
+//! - **Opening an older format upgrades it**: right after loading, [`Journal::open`] writes a
+//!   snapshot in this build's format (also for a new library), before anything is appended — so a
+//!   library is never a mix, and older builds refuse it from then on instead of reading part of it.
+//!   [`LoadReport::upgraded_from`] says when that happened.
+//! - **A newer format is refused, untouched**: a snapshot whose `version` is above [`VERSION`],
+//!   or a log record whose CRC matches (so it is exactly what was written) but whose op doesn't
+//!   parse, fails [`Journal::open`] with [`CatalogError::Newer`] before any file is modified — it
+//!   is never mistaken for a torn tail or damage. (Builds up to v0.2.0 refuse a v2 snapshot with
+//!   "unsupported format".)

 use crate::store::Store;
 use crate::{Catalog, CatalogError, Op, Result};
@@ -21,7 +68,9 @@ use crate::{Catalog, CatalogError, Op, Result};
 pub const SNAPSHOT: &str = "catalog.snap";
 pub const LOG: &str = "catalog.log";
 const FORMAT: &str = "lightcraft-catalog";
-const VERSION: u32 = 1;
+/// The catalog format this build writes (and the newest it reads). See the module docs →
+/// *Format versions*; bump it whenever an [`Op`] variant or a serialized field is added.
+pub const VERSION: u32 = 3;

 /// When [`Journal::wants_snapshot`] says it's time to compact the log.
 #[derive(Clone, Copy, Debug)]
@@ -49,10 +98,75 @@ pub struct LoadReport {
    pub failed: usize,
    /// Bytes of a torn final record that were dropped.
    pub torn_bytes: u64,
+    /// Partial records left mid-log by a failed append that was retried (written by builds that
+    /// didn't cut the log back after a failed write): skipped, the retried records were replayed.
+    pub torn_fragments: usize,
    /// The damaged log was preserved under this name.
    pub damaged: Option<String>,
    /// No catalog files existed (a new library).
    pub created: bool,
+    /// The library was in an older catalog format (`0` = before format versions: a log without a
+    /// snapshot) and was rewritten in this version's format ([`VERSION`]).
+    pub upgraded_from: Option<u32>,
+}
+
+/// Where persistence time goes (reported by `library.info` → `persistence`; printed to stderr
+/// per write under `LIGHTCRAFT_PROFILE`). Times are wall-clock milliseconds on the calling
+/// thread, i.e. how long the caller (the UI thread, for the app) was blocked.
+#[derive(Clone, Copy, Debug, Default, PartialEq, serde::Serialize)]
+#[serde(rename_all = "camelCase")]
+pub struct PersistStats {
+    /// [`Journal::append`] calls that wrote something.
+    pub appends: u64,
+    /// Encode + write + `sync_data` of the last / slowest append.
+    pub last_append_ms: f64,
+    pub max_append_ms: f64,
+    /// Snapshots written (compactions, plus the ones on close / repair).
+    pub snapshots: u64,
+    /// The last snapshot, by stage.
+    pub last_snapshot: SnapshotTiming,
+    /// Total time of the slowest snapshot.
+    pub max_snapshot_ms: f64,
+    /// Longest time a snapshot blocked the caller (for background snapshots: copying the catalog
+    /// and starting the worker, plus rewriting the log when it is done).
+    pub max_blocking_ms: f64,
+    /// A background snapshot is being written.
+    pub snapshot_running: bool,
+    /// Background snapshots that failed (the log was kept whole; retried later).
+    pub failed_snapshots: u64,
+}
+
+/// One snapshot (compaction), by stage.
+#[derive(Clone, Copy, Debug, Default, PartialEq, serde::Serialize)]
+#[serde(rename_all = "camelCase")]
+pub struct SnapshotTiming {
+    /// Serialising the catalog to JSON, streamed into the temp file (buffered writes included).
+    pub serialize_ms: f64,
+    /// Flushing, `sync_all` and the atomic rename of `catalog.snap`.
+    pub write_sync_ms: f64,
+    /// Resetting `catalog.log` (an atomic rewrite to empty, or to the records appended while a
+    /// background snapshot ran; fsynced).
+    pub reset_ms: f64,
+    /// Start to finish (for a background snapshot: until [`Journal::poll`] saw it done).
+    pub total_ms: f64,
+    /// How long the caller was blocked (equal to `total_ms` unless `background`).
+    pub blocking_ms: f64,
+    /// Written by a worker thread.
+    pub background: bool,
+    /// Size of `catalog.snap`.
+    pub bytes: u64,
+    /// Log records the snapshot compacted.
+    pub records: u64,
+}
+
+/// `LIGHTCRAFT_PROFILE` is set: print persistence timings to stderr.
+fn profiling() -> bool {
+    static ON: std::sync::OnceLock<bool> = std::sync::OnceLock::new();
+    *ON.get_or_init(|| std::env::var_os("LIGHTCRAFT_PROFILE").is_some())
+}
+
+fn ms_since(t: web_time::Instant) -> f64 {
+    t.elapsed().as_secs_f64() * 1e3
 }

 pub struct Journal {
@@ -63,6 +177,29 @@ pub struct Journal {
    log_records: u64,
    log_bytes: u64,
    pub policy: SnapshotPolicy,
+    stats: PersistStats,
+    /// The background snapshot in flight.
+    pending: Option<Pending>,
+    /// After a failed background snapshot: don't retry before the log has this many records.
+    retry_at_records: u64,
+    /// A failed append may have left part of its records after `log_bytes`, and cutting them off
+    /// failed too: the next append cuts the log back to `log_bytes` first.
+    tail_dirty: bool,
+}
+
+/// A background snapshot in flight.
+struct Pending {
+    /// The snapshot holds the state after this op.
+    seq: u64,
+    /// Log bytes / records appended since it started (all `> seq`): the log once it's done.
+    tail: Vec<u8>,
+    tail_records: u64,
+    /// Log records it compacts.
+    records: u64,
+    started: web_time::Instant,
+    /// Time the caller spent starting it.
+    start_ms: f64,
+    worker: std::thread::JoinHandle<std::io::Result<SnapshotTiming>>,
 }

 fn io(e: std::io::Error) -> CatalogError {
@@ -76,24 +213,93 @@ pub fn encode_record(seq: u64, op: &Op) -> String {
    format!("{{\"seq\":{seq},\"crc\":{crc},\"op\":{body}}}")
 }

-/// Parse one log line (without the newline). `None` if it's malformed or the CRC doesn't match.
+/// Parse one log line (without the newline). `None` if it's malformed, the CRC doesn't match, or
+/// the op is one this version doesn't know (see [`Journal::open`]).
 pub fn decode_record(line: &str) -> Option<(u64, Op)> {
-    let rest = line.strip_prefix("{\"seq\":")?;
-    let (seq, rest) = rest.split_once(",\"crc\":")?;
-    let (crc, rest) = rest.split_once(",\"op\":")?;
-    let body = rest.strip_suffix('}')?;
-    let seq: u64 = seq.parse().ok()?;
-    let crc: u32 = crc.parse().ok()?;
-    if crc32fast::hash(body.as_bytes()) != crc {
-        return None;
+    match decode_line(line) {
+        Line::Record(seq, op) => Some((seq, op)),
+        Line::Newer(_) | Line::Bad => None,
    }
-    serde_json::from_str(body).ok().map(|op| (seq, op))
 }

+/// One log line, decoded.
+enum Line {
+    Record(u64, Op),
+    /// The CRC matches — the line is exactly what was written — but the op doesn't parse: it
+    /// was written by a newer LightCraft (an op or field this version doesn't know), not damaged.
+    Newer(u64),
+    /// Malformed or a CRC mismatch: torn or damaged.
+    Bad,
+}
+
+fn decode_line(line: &str) -> Line {
+    let parts = || {
+        let rest = line.strip_prefix("{\"seq\":")?;
+        let (seq, rest) = rest.split_once(",\"crc\":")?;
+        let (crc, rest) = rest.split_once(",\"op\":")?;
+        let body = rest.strip_suffix('}')?;
+        Some((seq.parse::<u64>().ok()?, crc.parse::<u32>().ok()?, body))
+    };
+    let Some((seq, crc, body)) = parts() else { return Line::Bad };
+    if crc32fast::hash(body.as_bytes()) != crc {
+        return Line::Bad;
+    }
+    match serde_json::from_str(body) {
+        Ok(op) => Line::Record(seq, op),
+        Err(_) => Line::Newer(seq),
+    }
+}
+
+/// The error for a library written by a newer LightCraft.
+fn newer(what: String) -> CatalogError {
+    CatalogError::Newer(format!("{what}; this version reads catalog format v{VERSION} and older"))
+}
+
+/// A line that doesn't decode but ends with a whole record: a fragment of a failed append followed
+/// by the record that retried it (the retry re-encodes the same ops with the same `seq`s). Returns
+/// the fragment's length and the record, if that record continues the log (`seq <= last + 1`).
+/// The record's CRC must match, so damage is never mistaken for it.
+fn resync(line: &[u8], last: u64) -> Option<(usize, (u64, Op))> {
+    const START: &[u8] = b"{\"seq\":";
+    (1..line.len().saturating_sub(START.len() - 1)).filter(|&i| line[i..].starts_with(START)).find_map(|i| {
+        let rec = std::str::from_utf8(&line[i..]).ok().map(|l| l.trim_end_matches('\r')).and_then(decode_record)?;
+        (rec.0 <= last + 1).then_some((i, rec))
+    })
+}
+
+/// Write `catalog.snap` for the state after op `seq`, streaming the JSON into the store (no
+/// whole-file string). The bytes are exactly
+/// `{"format":"lightcraft-catalog","version":1,"seq":N,"catalog":<serde_json of the catalog>}\n`,
+/// as before streaming. Fills in the serialise / write+sync times and the size.
+fn write_snapshot(store: &mut dyn Store, seq: u64, catalog: &Catalog) -> std::io::Result<SnapshotTiming> {
+    let t0 = web_time::Instant::now();
+    let mut serialize_ms = 0.0;
+    let bytes = store.write_atomic_with(SNAPSHOT, &mut |w| {
+        use std::io::Write;
+        let t = web_time::Instant::now();
+        // serde_json emits many tiny writes: buffer them in a concrete writer (inlined), so only
+        // 1 MiB chunks go through the store's `dyn Write`
+        let mut b = std::io::BufWriter::with_capacity(1 << 20, w);
+        write!(b, "{{\"format\":\"{FORMAT}\",\"version\":{VERSION},\"seq\":{seq},\"catalog\":")?;
+        serde_json::to_writer(&mut b, catalog).map_err(std::io::Error::other)?;
+        b.write_all(b"}\n")?;
+        b.flush()?;
+        serialize_ms = ms_since(t);
+        Ok(())
+    })?;
+    Ok(SnapshotTiming { serialize_ms, write_sync_ms: (ms_since(t0) - serialize_ms).max(0.0), bytes, ..Default::default() })
+}
+
+/// Just the identification of `catalog.snap` (read before the catalog itself).
 #[derive(serde::Deserialize)]
-struct SnapFile {
+struct SnapHeader {
    format: String,
    version: u32,
+}
+
+/// The whole `catalog.snap` (after [`SnapHeader`] was checked; `format` and `version` are ignored).
+#[derive(serde::Deserialize)]
+struct SnapFile {
    seq: u64,
    catalog: Catalog,
 }
@@ -106,18 +312,34 @@ impl Journal {
        let snap = store.read(SNAPSHOT).map_err(io)?;
        let log = store.read(LOG).map_err(io)?;
        report.created = snap.is_none() && log.is_none();
-        let (mut catalog, snapshot_seq) = match snap {
+        let (mut catalog, snapshot_seq, snapshot_version) = match snap {
            Some(bytes) => {
-                let s: SnapFile = serde_json::from_slice(&bytes).map_err(|e| CatalogError::Corrupt(format!("{SNAPSHOT}: {e}")))?;
-                if s.format != FORMAT || s.version > VERSION {
-                    return Err(CatalogError::Corrupt(format!("{SNAPSHOT}: unsupported format {} v{}", s.format, s.version)));
+                // the header first: a newer snapshot may not parse as this version's catalog
+                let h: SnapHeader = serde_json::from_slice(&bytes).map_err(|e| CatalogError::Corrupt(format!("{SNAPSHOT}: {e}")))?;
+                if h.format != FORMAT {
+                    return Err(CatalogError::Corrupt(format!("{SNAPSHOT}: not a LightCraft catalog (format {:?})", h.format)));
                }
-                (s.catalog, s.seq)
+                if h.version > VERSION {
+                    return Err(newer(format!("{SNAPSHOT} is catalog format v{}", h.version)));
+                }
+                let s: SnapFile = serde_json::from_slice(&bytes).map_err(|e| CatalogError::Corrupt(format!("{SNAPSHOT}: {e}")))?;
+                (s.catalog, s.seq, Some(h.version))
            }
-            None => (Catalog::new(), 0),
+            None => (Catalog::new(), 0, None),
        };
        report.snapshot_seq = snapshot_seq;
-        let mut j = Journal { store, seq: snapshot_seq, snapshot_seq, log_records: 0, log_bytes: 0, policy: SnapshotPolicy::default() };
+        let mut j = Journal {
+            store,
+            seq: snapshot_seq,
+            snapshot_seq,
+            log_records: 0,
+            log_bytes: 0,
+            policy: SnapshotPolicy::default(),
+            stats: PersistStats::default(),
+            pending: None,
+            retry_at_records: 0,
+            tail_dirty: false,
+        };
        let log = log.unwrap_or_default();

        // Scan records; `good_end` is the byte offset just past the last good record.
@@ -136,7 +358,22 @@ impl Journal {
                good_end = next;
                continue;
            }
-            match line.and_then(decode_record) {
+            let mut record = line.and_then(decode_record);
+            if record.is_none()
+                && let Some((at, r)) = resync(&log[pos..line_end], j.seq)
+            {
+                // the torn start of a failed append, then the retried records (older builds
+                // appended them right after the fragment)
+                log::warn!("catalog log: skipped a partial record ({at} bytes) left by a failed write");
+                report.torn_fragments += 1;
+                record = Some(r);
+            }
+            if record.is_none()
+                && let Some(Line::Newer(seq)) = line.map(decode_line)
+            {
+                return Err(newer(format!("{LOG} record {seq} holds a change this version doesn't know")));
+            }
+            match record {
                Some((seq, op)) if seq <= j.seq => {
                    // already in the snapshot
                    let _ = op;
@@ -152,7 +389,15 @@ impl Journal {
                }
                _ => {
                    // bad record (or a gap): torn tail if nothing good follows, else damage
-                    let rest_has_good = log[next..].split(|b| *b == b'\n').any(|l| std::str::from_utf8(l).ok().and_then(decode_record).is_some());
+                    let mut rest_has_good = false;
+                    for l in log[next..].split(|b| *b == b'\n') {
+                        match std::str::from_utf8(l).map(|l| decode_line(l.trim_end_matches('\r'))) {
+                            Ok(Line::Record(..)) => rest_has_good = true,
+                            // refused before anything is modified
+                            Ok(Line::Newer(seq)) => return Err(newer(format!("{LOG} record {seq} holds a change this version doesn't know"))),
+                            Ok(Line::Bad) | Err(_) => {}
+                        }
+                    }
                    if rest_has_good {
                        damaged_at = Some(pos);
                    }
@@ -176,10 +421,34 @@ impl Journal {
                report.torn_bytes = (log.len() - good_end) as u64;
                log::warn!("catalog log: dropped a torn final record ({} bytes)", report.torn_bytes);
                j.store.truncate(LOG, good_end as u64).map_err(io)?;
+                j.log_bytes = good_end as u64;
            } else if needs_newline {
                j.store.append(LOG, b"\n").map_err(io)?;
+                j.log_bytes = good_end as u64 + 1;
+            } else {
+                j.log_bytes = good_end as u64;
+            }
+        }
+        // Bring an older (or new, or snapshot-less) library to this version's format before
+        // anything is appended: a snapshot in this format makes older builds refuse the library
+        // instead of misreading ops or dropping fields they don't know.
+        if report.created {
+            // tiny: written directly, not counted as a compaction
+            let empty = format!("{{\"format\":\"{FORMAT}\",\"version\":{VERSION},\"seq\":0,\"catalog\":{}}}\n", catalog.to_snapshot());
+            if let Err(e) = j.store.write_atomic(SNAPSHOT, empty.as_bytes()) {
+                log::warn!("catalog: can't write the new library's snapshot: {e}");
+            }
+        } else if report.damaged.is_none() && snapshot_version != Some(VERSION) {
+            report.upgraded_from = Some(snapshot_version.unwrap_or(0));
+            match j.snapshot(&catalog) {
+                Ok(()) => {
+                    if let Some(v) = report.upgraded_from {
+                        log::info!("catalog: upgraded from format v{v} to v{VERSION}");
+                    }
+                }
+                // nothing lost: the old files are intact and still load; retried on next open
+                Err(e) => log::warn!("catalog: can't write the format v{VERSION} snapshot: {e}"),
            }
-            j.log_bytes = good_end as u64;
        }
        catalog.revision = 0;
        Ok((j, catalog, report))
@@ -190,6 +459,7 @@ impl Journal {
        if ops.is_empty() {
            return Ok(());
        }
+        let t0 = web_time::Instant::now();
        let mut buf = String::new();
        let mut seq = self.seq;
        for op in ops {
@@ -197,29 +467,202 @@ impl Journal {
            buf.push_str(&encode_record(seq, op));
            buf.push('\n');
        }
-        self.store.append(LOG, buf.as_bytes()).map_err(io)?;
+        if self.tail_dirty {
+            self.store.truncate(LOG, self.log_bytes).map_err(io)?;
+            self.tail_dirty = false;
+        }
+        if let Err(e) = self.store.append(LOG, buf.as_bytes()) {
+            // Part of the batch may be in the file. Cut it back to the last whole record, so the
+            // retry (same seqs) starts on a clean line instead of after a fragment that would
+            // read back as damage.
+            if let Err(t) = self.store.truncate(LOG, self.log_bytes) {
+                log::error!("catalog: can't cut the log back after a failed append: {t}");
+                self.tail_dirty = true;
+            }
+            return Err(io(e));
+        }
+        if let Some(p) = self.pending.as_mut() {
+            p.tail.extend_from_slice(buf.as_bytes());
+            p.tail_records += ops.len() as u64;
+        }
        self.seq = seq;
        self.log_records += ops.len() as u64;
        self.log_bytes += buf.len() as u64;
+        let ms = ms_since(t0);
+        self.stats.appends += 1;
+        self.stats.last_append_ms = ms;
+        self.stats.max_append_ms = self.stats.max_append_ms.max(ms);
+        if profiling() {
+            eprintln!("catalog: append {} op(s), {} B: {ms:.2} ms", ops.len(), buf.len());
+        }
        Ok(())
    }

    /// Write a snapshot of `catalog` (which must reflect every appended op) and reset the log.
    pub fn snapshot(&mut self, catalog: &Catalog) -> Result<()> {
-        let body = serde_json::to_string(catalog).map_err(|e| CatalogError::Invalid(e.to_string()))?;
-        let file = format!("{{\"format\":\"{FORMAT}\",\"version\":{VERSION},\"seq\":{},\"catalog\":{body}}}\n", self.seq);
-        self.store.write_atomic(SNAPSHOT, file.as_bytes()).map_err(io)?;
+        self.snapshot_with_unlogged(catalog, 0)
+    }
+
+    /// [`Journal::snapshot`] of a catalog that also reflects `unlogged` ops applied after the
+    /// last appended one but never written to the log (their append failed). The snapshot counts
+    /// them (it holds the state after op `seq + unlogged`), so they are saved and must not be
+    /// appended afterwards. Once the snapshot is durable [`Journal::seq`] includes them, even if
+    /// resetting the log then fails (the error is still returned).
+    pub fn snapshot_with_unlogged(&mut self, catalog: &Catalog, unlogged: u64) -> Result<()> {
+        // never two snapshot writers; the one in flight is older, so it must land first
+        if let Err(e) = self.wait() {
+            log::warn!("catalog: background snapshot failed ({e}); writing one now");
+        }
+        let t0 = web_time::Instant::now();
+        let seq = self.seq + unlogged;
+        let mut timing = match write_snapshot(self.store.as_mut(), seq, catalog) {
+            Ok(t) => t,
+            Err(e) => {
+                // nothing lost (the log is whole); don't retry on every append
+                self.stats.failed_snapshots += 1;
+                self.retry_at_records = self.log_records + self.policy.max_records;
+                return Err(io(e));
+            }
+        };
+        // the snapshot is durable: it holds every op up to `seq`
+        self.seq = seq;
+        self.snapshot_seq = seq;
+        let t2 = web_time::Instant::now();
        // A crash here leaves old records in the log; they are skipped by seq on load.
        self.store.write_atomic(LOG, b"").map_err(io)?;
-        self.snapshot_seq = self.seq;
+        timing.reset_ms = ms_since(t2);
+        timing.total_ms = ms_since(t0);
+        timing.blocking_ms = timing.total_ms;
+        timing.records = self.log_records;
+        self.record_snapshot(timing);
        self.log_records = 0;
        self.log_bytes = 0;
+        self.retry_at_records = 0;
+        self.tail_dirty = false;
        Ok(())
    }

+    /// Compact like [`Journal::snapshot`], but write the snapshot on a worker thread from a copy
+    /// of `catalog` (cheap: photos are shared) when the store supports it (else synchronously).
+    /// `catalog` must reflect every appended op. Appends go on meanwhile; [`Journal::poll`]
+    /// finishes the compaction. No-op while a snapshot is already in flight.
+    pub fn snapshot_in_background(&mut self, catalog: &Catalog) -> Result<()> {
+        if self.pending.is_some() {
+            return Ok(());
+        }
+        let t0 = web_time::Instant::now();
+        let Some(mut writer) = self.store.background_writer() else { return self.snapshot(catalog) };
+        let copy = catalog.clone();
+        let seq = self.seq;
+        let spawned = std::thread::Builder::new().name("catalog-snapshot".into()).spawn(move || write_snapshot(writer.as_mut(), seq, &copy));
+        let worker = match spawned {
+            Ok(w) => w,
+            Err(e) => {
+                log::warn!("catalog: can't start a snapshot thread ({e}); writing it now");
+                return self.snapshot(catalog);
+            }
+        };
+        self.pending =
+            Some(Pending { seq, tail: Vec::new(), tail_records: 0, records: self.log_records, started: t0, start_ms: ms_since(t0), worker });
+        Ok(())
+    }
+
+    /// Finish a background snapshot if its worker is done (cheap otherwise). Returns whether one
+    /// finished. An error means the snapshot failed: nothing was lost (the log is whole), and
+    /// compaction is retried once the log has grown by another [`SnapshotPolicy::max_records`].
+    pub fn poll(&mut self) -> Result<bool> {
+        if !self.pending.as_ref().is_some_and(|p| p.worker.is_finished()) {
+            return Ok(false);
+        }
+        self.finish().map(|()| true)
+    }
+
+    /// Wait for the background snapshot in flight, if any, and finish it.
+    pub fn wait(&mut self) -> Result<()> {
+        if self.pending.is_none() {
+            return Ok(());
+        }
+        self.finish()
+    }
+
+    /// A background snapshot is in flight (until [`Journal::poll`] or [`Journal::wait`] finish it).
+    pub fn snapshot_running(&self) -> bool {
+        self.pending.is_some()
+    }
+
+    /// The background worker is done (its result not yet applied by [`Journal::poll`]).
+    pub fn snapshot_written(&self) -> bool {
+        self.pending.as_ref().is_some_and(|p| p.worker.is_finished())
+    }
+
+    /// Join the worker; on success replace the log with the records appended meanwhile.
+    fn finish(&mut self) -> Result<()> {
+        let Some(p) = self.pending.take() else { return Ok(()) };
+        let t0 = web_time::Instant::now();
+        let written = p.worker.join().unwrap_or_else(|_| Err(std::io::Error::other("snapshot thread panicked")));
+        let mut timing = match written {
+            Ok(t) => t,
+            Err(e) => {
+                // the log still holds every record: nothing lost, retry later
+                self.stats.failed_snapshots += 1;
+                self.retry_at_records = self.log_records + self.policy.max_records;
+                log::error!("catalog: background snapshot failed: {e}");
+                return Err(io(e));
+            }
+        };
+        // `catalog.snap` (op `p.seq`) is durable; the log's records `<= p.seq` are now stale.
+        // Keep exactly the ones appended since (a crash before this rewrite skips the stale ones).
+        let t1 = web_time::Instant::now();
+        self.snapshot_seq = p.seq;
+        if let Err(e) = self.store.write_atomic(LOG, &p.tail) {
+            // the log stays whole (stale records are skipped on load); compact again later
+            self.retry_at_records = self.log_records + self.policy.max_records;
+            log::error!("catalog: can't trim the log after a snapshot: {e}");
+            return Err(io(e));
+        }
+        timing.reset_ms = ms_since(t1);
+        timing.total_ms = ms_since(p.started);
+        timing.blocking_ms = p.start_ms + ms_since(t0);
+        timing.background = true;
+        timing.records = p.records;
+        self.record_snapshot(timing);
+        self.log_records = p.tail_records;
+        self.log_bytes = p.tail.len() as u64;
+        self.retry_at_records = 0;
+        self.tail_dirty = false;
+        Ok(())
+    }
+
+    fn record_snapshot(&mut self, t: SnapshotTiming) {
+        self.stats.snapshots += 1;
+        self.stats.last_snapshot = t;
+        self.stats.max_snapshot_ms = self.stats.max_snapshot_ms.max(t.total_ms);
+        self.stats.max_blocking_ms = self.stats.max_blocking_ms.max(t.blocking_ms);
+        if profiling() {
+            eprintln!(
+                "catalog: {}snapshot of {} records, {} B: serialize {:.1} ms, write+sync {:.1} ms, log reset {:.1} ms, total {:.1} ms, blocking {:.1} ms",
+                if t.background { "background " } else { "" },
+                t.records,
+                t.bytes,
+                t.serialize_ms,
+                t.write_sync_ms,
+                t.reset_ms,
+                t.total_ms,
+                t.blocking_ms
+            );
+        }
+    }
+
+    /// Where persistence time went so far.
+    pub fn stats(&self) -> PersistStats {
+        PersistStats { snapshot_running: self.pending.is_some(), ..self.stats }
+    }
+
    /// The log is long enough to be worth compacting.
    pub fn wants_snapshot(&self) -> bool {
-        self.log_records >= self.policy.max_records || self.log_bytes >= self.policy.max_bytes
+        self.pending.is_none()
+            && self.log_records >= self.retry_at_records
+            && (self.log_records >= self.policy.max_records || self.log_bytes >= self.policy.max_bytes)
    }

    pub fn seq(&self) -> u64 {
@@ -238,3 +681,16 @@ impl Journal {
        self.store.describe()
    }
 }
+
+impl Drop for Journal {
+    /// Wait for a background snapshot (never leave a writer behind that could replace a newer
+    /// snapshot of a journal reopened on the same files). The log is left whole: the snapshot's
+    /// records in it are skipped as stale on the next load.
+    fn drop(&mut self) {
+        if let Some(p) = self.pending.take()
+            && let Ok(Err(e)) = p.worker.join()
+        {
+            log::error!("catalog: background snapshot failed: {e}");
+        }
+    }
+}
@@ -5,14 +5,24 @@
 //!   snapshot — crash-safe and diff-friendly (see [`journal`]);
 //! - **undo/redo**: the engine keeps inverse ops;
 //! - **determinism**: replaying the log reproduces the state exactly (property-tested).
+//!
+//! **Catalog format version** ([`journal::VERSION`], see [`journal`] → *Format versions*): adding
+//! an [`Op`] variant or a serialized field means bumping it. Newer builds read every older format
+//! (and upgrade it on open); older builds refuse a newer library with [`CatalogError::Newer`]
+//! instead of reading part of it.
 #![forbid(unsafe_code)]
+#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)]

 pub mod dates;
+pub mod folders;
 pub mod journal;
 pub mod keywords;
+pub mod local;
+pub mod lock;
 pub mod model;
 pub mod query;
 pub mod rules;
+pub mod safe_file;
 pub mod stacks;
 pub mod store;

@@ -20,11 +30,14 @@ use std::collections::BTreeMap;
 use std::sync::Arc;

 pub use dates::{DateRun, GroupBy};
-pub use journal::{Journal, LoadReport, SnapshotPolicy};
+pub use folders::FolderNode;
+pub use journal::{Journal, LoadReport, PersistStats, SnapshotPolicy, SnapshotTiming};
 pub use keywords::KeywordNode;
 use lightcraft_develop::DevelopSettings;
+pub use local::{DEFAULT_FORGET_DAYS, ForgetPlan, folder_of};
+pub use lock::{LibraryLock, LockError, LockOwner};
 pub use model::*;
-pub use query::{DateGroup, Filter, RatingOp, Sort, SortKey};
+pub use query::{DateGroup, Filter, Person, RatingOp, Sort, SortKey, mix64};
 pub use rules::{Match, Rule, RuleSet};
 use serde::{Deserialize, Serialize};
 pub use store::{FsStore, MemStore, Store};
@@ -43,6 +56,10 @@ pub enum CatalogError {
    Corrupt(String),
    #[error("catalog storage: {0}")]
    Io(String),
+    /// The library was written by a newer LightCraft (a newer catalog format, or a change this
+    /// version doesn't know). Nothing was read into the session and nothing was modified.
+    #[error("this library was written by a newer version of LightCraft ({0}); update LightCraft to open it. The library was left unchanged.")]
+    Newer(String),
 }

 pub type Result<T> = std::result::Result<T, CatalogError>;
@@ -119,6 +136,11 @@ pub enum Op {
        id: AlbumId,
        parent: Option<AlbumId>,
    },
+    /// Where the album stands among its siblings (`None`: by name). Format version 3.
+    SetAlbumOrder {
+        id: AlbumId,
+        order: Option<u32>,
+    },
    SetAlbumPhotos {
        id: AlbumId,
        photos: Vec<PhotoId>,
@@ -179,12 +201,22 @@ pub enum Op {
        file_size: u64,
        #[serde(default)]
        content_hash: Option<String>,
+        /// Why the file can only be shown from its embedded preview (see [`Photo::preview_only`]);
+        /// `None` = its raw data decodes (or it isn't a raw).
+        #[serde(default, skip_serializing_if = "Option::is_none")]
+        preview_only: Option<String>,
    },
    /// The name shown for a colour label (`None` = its colour's name).
    SetLabelName {
        label: ColorLabel,
        name: Option<String>,
    },
+    /// When a Local folder was last browsed (ISO 8601; `None` = forget the time). Not an undo
+    /// step: it drives forgetting untouched Local records (see [`local`]).
+    SetBrowsed {
+        folder: String,
+        at: Option<String>,
+    },
    /// Several ops as one step (undo applies the inverses in reverse).
    Batch {
        ops: Vec<Op>,
@@ -204,6 +236,9 @@ pub struct Catalog {
    /// Custom colour label names.
    #[serde(default, skip_serializing_if = "BTreeMap::is_empty")]
    label_names: BTreeMap<ColorLabel, String>,
+    /// When each Local folder was last browsed (folder path → ISO 8601), see [`local`].
+    #[serde(default, skip_serializing_if = "BTreeMap::is_empty")]
+    browsed: BTreeMap<String, String>,
    /// Increments on every applied op.
    #[serde(skip)]
    pub revision: u64,
@@ -252,6 +287,28 @@ impl Catalog {
    pub fn albums(&self) -> impl Iterator<Item = &Album> {
        self.albums.values()
    }
+    /// The albums and folders directly inside `parent` (`None`: the top level), in the order the
+    /// sidebar lists them: folders first, then by their place if the user ordered them (those
+    /// without one after, by name), else by name.
+    pub fn album_children(&self, parent: Option<AlbumId>) -> Vec<&Album> {
+        let mut kids: Vec<&Album> = self.albums.values().filter(|a| a.parent == parent).collect();
+        kids.sort_by_cached_key(|a| album_order_key(a));
+        kids
+    }
+    /// [`Self::album_children`] of every folder (and of the top level, under `None`) in one pass:
+    /// what a tree view that draws all of them wants each frame.
+    pub fn album_children_by_parent(&self) -> std::collections::HashMap<Option<AlbumId>, Vec<&Album>> {
+        let mut map: std::collections::HashMap<Option<AlbumId>, Vec<&Album>> = std::collections::HashMap::new();
+        for a in self.albums.values() {
+            map.entry(a.parent).or_default().push(a);
+        }
+        map.values_mut().for_each(|kids| kids.sort_by_cached_key(|a| album_order_key(a)));
+        map
+    }
+    /// Whether anything inside `parent` has a place of its own (the folder is ordered by hand).
+    pub fn album_children_are_ordered(&self, parent: Option<AlbumId>) -> bool {
+        self.albums.values().any(|a| a.parent == parent && a.order.is_some())
+    }
    /// The Quick Collection, once something was added to it.
    pub fn quick_collection(&self) -> Option<AlbumId> {
        self.albums.values().find(|a| a.quick).map(|a| a.id)
@@ -285,7 +342,8 @@ impl Catalog {
    /// Number of photos shown for an album in the sources list (excludes deleted photos).
    pub fn album_count(&self, id: AlbumId) -> usize {
        match self.albums.get(&id) {
-            Some(Album { smart: Some(_), .. }) => self.album_photos(id).len(),
+            // counted in place: no id list is built just for its length
+            Some(Album { smart: Some(_), .. }) => self.photos.values().filter(|p| self.album_contains(id, p)).count(),
            Some(a) => a.photos.iter().filter(|p| self.photos.get(p).is_some_and(|p| !p.deleted)).count(),
            None => 0,
        }
@@ -448,7 +506,20 @@ impl Catalog {
                    }
                }
                let a = self.album_mut(id)?;
-                Op::MoveAlbum { id, parent: std::mem::replace(&mut a.parent, parent) }
+                if a.parent == parent {
+                    return Ok(Op::MoveAlbum { id, parent });
+                }
+                // its place belonged to the old neighbours: the undo gives it back with the folder
+                let old_order = a.order.take();
+                let old_parent = std::mem::replace(&mut a.parent, parent);
+                match old_order {
+                    Some(_) => Op::Batch { ops: vec![Op::MoveAlbum { id, parent: old_parent }, Op::SetAlbumOrder { id, order: old_order }] },
+                    None => Op::MoveAlbum { id, parent: old_parent },
+                }
+            }
+            Op::SetAlbumOrder { id, order } => {
+                let a = self.album_mut(id)?;
+                Op::SetAlbumOrder { id, order: std::mem::replace(&mut a.order, order) }
            }
            Op::SetAlbumPhotos { id, photos } => {
                let a = self.album_mut(id)?;
@@ -514,7 +585,7 @@ impl Catalog {
                let old_format = format.map(|f| std::mem::replace(&mut p.format, f));
                Op::Relink { id, file_name: old_name, source: std::mem::replace(&mut p.source, source), format: old_format }
            }
-            Op::SetContent { id, width, height, file_size, content_hash } => {
+            Op::SetContent { id, width, height, file_size, content_hash, preview_only } => {
                let p = self.photo_mut(id)?;
                Op::SetContent {
                    id,
@@ -522,6 +593,7 @@ impl Catalog {
                    height: std::mem::replace(&mut p.height, height),
                    file_size: std::mem::replace(&mut p.file_size, file_size),
                    content_hash: std::mem::replace(&mut p.content_hash, content_hash),
+                    preview_only: std::mem::replace(&mut p.preview_only, preview_only),
                }
            }
            Op::SetFile { id, file_name, source } => {
@@ -540,6 +612,14 @@ impl Catalog {
                };
                Op::SetLabelName { label, name: old }
            }
+            Op::SetBrowsed { folder, at } => {
+                let folder = crate::query::folder_key(&folder);
+                let old = match at {
+                    Some(t) => self.browsed.insert(folder.clone(), t),
+                    None => self.browsed.remove(&folder),
+                };
+                Op::SetBrowsed { folder, at: old }
+            }
            Op::Batch { ops } => {
                let mut inverses = Vec::with_capacity(ops.len());
                for op in ops {
@@ -574,6 +654,22 @@ impl Catalog {
        Op::Batch { ops }
    }

+    /// [`Self::delete_permanently_ops`] for several photos at once. One op list built against the
+    /// catalog as it is now: albums lose all of them in one write each, and stacks are shortened
+    /// (or dissolved) once, however many of their photos go.
+    pub fn delete_photos_permanently_ops(&self, ids: &[PhotoId]) -> Op {
+        let gone: std::collections::HashSet<PhotoId> = ids.iter().copied().collect();
+        let mut ops: Vec<Op> = self
+            .albums
+            .values()
+            .filter(|a| a.photos.iter().any(|p| gone.contains(p)))
+            .map(|a| Op::SetAlbumPhotos { id: a.id, photos: a.photos.iter().copied().filter(|p| !gone.contains(p)).collect() })
+            .collect();
+        ops.extend(self.remove_from_stacks_ops(ids));
+        ops.extend(gone.iter().map(|id| Op::RemovePhoto { id: *id }));
+        Op::Batch { ops }
+    }
+
    // ---- persistence

    /// Full snapshot as JSON.
@@ -612,7 +708,27 @@ impl Catalog {
    }
 }

+/// How siblings are listed: folders first, then the ones with a place of their own by it, then by
+/// name; the id settles any tie.
+fn album_order_key(a: &Album) -> (bool, bool, u32, String, AlbumId) {
+    (!a.folder, a.order.is_none(), a.order.unwrap_or(0), a.name.to_lowercase(), a.id)
+}
+
 #[cfg(test)]
 mod tests;
 #[cfg(test)]
+mod tests_album_order;
+#[cfg(test)]
+mod tests_background;
+#[cfg(test)]
+mod tests_folders;
+#[cfg(test)]
+mod tests_format_version;
+#[cfg(test)]
 mod tests_journal;
+#[cfg(test)]
+mod tests_local;
+#[cfg(test)]
+mod tests_lock;
+#[cfg(test)]
+mod tests_torn_append;
@@ -0,0 +1,305 @@
+//! Forgetting untouched Local browse records.
+//!
+//! Browsing a folder (Local) catalogues its photos as `local` records so they can be shown,
+//! rated and edited in place. Most are only ever looked at, and on a large disk they can be most
+//! of the catalog. A record is **forgotten** (removed from the catalog, never from the disk —
+//! files and XMP sidecars are not touched) when all of these hold:
+//!
+//! - it is a Local record ([`Photo::local`]) of a file, not deleted, not a virtual copy;
+//! - its folder was last browsed more than `days` ago ([`Catalog::last_browsed`]). A folder with
+//!   no recorded time (a catalog from before this existed) counts as browsed *now*: it is stamped
+//!   with the current time and nothing in it is forgotten this time;
+//! - it is **untouched** ([`Photo::untouched_local`]): nothing about it changed since it was
+//!   browsed — see below;
+//! - nothing refers to it: no album holds it or shows it as cover, no stack holds it, no virtual
+//!   copy was made of it; and the caller (the engine) says it isn't in use (visible, selected,
+//!   in the undo history, has a smart preview…).
+//!
+//! **Untouched.** A record catalogued by a browse carries a fingerprint of its whole state at
+//! that moment ([`Photo::local_baseline`]). It is untouched while its state still hashes to that
+//! fingerprint, i.e. *any* change since — develop settings or history, versions, rating, flag,
+//! label, any metadata (title, caption, keywords, location, copyright…), capture time, crop,
+//! rename or relink, culling scores, file content — keeps it. Records without a fingerprint
+//! (browsed before fingerprints existed) are judged by their data alone and kept unless every
+//! user-editable field is still empty / default: no rating, flag or label; develop settings equal
+//! to the import defaults with no history, versions or edit time; no title, caption, alt text,
+//! description, keywords, creator, copyright, rights, location or GPS; no culling scores.
+//! When in doubt a record is kept.
+//!
+//! Browsing the folder again catalogues its files afresh (new records, new fingerprints).
+
+use std::collections::{BTreeMap, BTreeSet, HashSet};
+
+use serde::Serialize;
+
+use crate::{Catalog, Op, Photo, PhotoId, Source};
+
+/// Default for how long a Local folder may go unbrowsed before its untouched records are
+/// forgotten (days).
+pub const DEFAULT_FORGET_DAYS: u32 = 30;
+
+/// Records removed per logged batch (keeps log lines a reasonable size).
+const BATCH: usize = 2000;
+
+/// 64-bit FNV-1a over everything written to it (stable across builds and platforms, unlike
+/// `std`'s hasher), so a photo is hashed as it serialises, without a buffer.
+struct Fnv(u64);
+
+impl std::io::Write for Fnv {
+    fn write(&mut self, buf: &[u8]) -> std::io::Result<usize> {
+        for b in buf {
+            self.0 ^= u64::from(*b);
+            self.0 = self.0.wrapping_mul(0x0100_0000_01b3);
+        }
+        Ok(buf.len())
+    }
+    fn flush(&mut self) -> std::io::Result<()> {
+        Ok(())
+    }
+}
+
+/// The folder a file-backed photo is in (`None` for demo photos).
+pub fn folder_of(p: &Photo) -> Option<String> {
+    match &p.source {
+        // `/` and `\` are both separators whatever the platform (a catalog can move between them)
+        Source::File { path } => {
+            let key = crate::query::folder_key(path);
+            key.rsplit_once('/').map(|(dir, _)| if dir.is_empty() { "/".to_string() } else { dir.to_string() })
+        }
+        Source::Demo { .. } => None,
+    }
+}
+
+impl Photo {
+    /// A hash of everything about the photo except [`Photo::local_baseline`] itself.
+    pub fn fingerprint(&self) -> u64 {
+        let mut h = Fnv(0xcbf2_9ce4_8422_2325);
+        let r = if self.local_baseline.is_some() {
+            let mut p = self.clone();
+            p.local_baseline = None;
+            serde_json::to_writer(&mut h, &p)
+        } else {
+            serde_json::to_writer(&mut h, self)
+        };
+        // (serialising a photo can't fail; if it ever did, no baseline would match: kept)
+        if r.is_err() { 0 } else { h.0 }
+    }
+
+    /// Record the current state as the browse baseline (done when a browse catalogues it).
+    pub fn set_local_baseline(&mut self) {
+        self.local_baseline = None;
+        self.local_baseline = Some(self.fingerprint());
+    }
+
+    /// A Local record nothing was changed on since it was browsed (see the module docs). Says
+    /// nothing about albums, stacks or virtual copies referring to it.
+    pub fn untouched_local(&self) -> bool {
+        if !self.local || self.deleted || self.copy_of.is_some() || self.copy_name.is_some() || folder_of(self).is_none() {
+            return false;
+        }
+        match self.local_baseline {
+            Some(b) => self.fingerprint() == b,
+            None => self.looks_unedited(),
+        }
+    }
+
+    /// Every user-editable field is still empty / default (records without a baseline).
+    fn looks_unedited(&self) -> bool {
+        let m = &self.meta;
+        self.rating == 0
+            && self.flag == crate::Flag::None
+            && self.label.is_none()
+            && self.edited.is_none()
+            && self.history.is_empty()
+            && self.versions.is_empty()
+            && *self.develop == self.import_defaults()
+            && self.analysis.is_none()
+            && m.title.is_empty()
+            && m.caption.is_empty()
+            && m.alt_text.is_empty()
+            && m.extended_description.is_empty()
+            && m.keywords.is_empty()
+            && m.creator.is_empty()
+            && m.copyright.is_empty()
+            && m.copyright_status.is_unknown()
+            && m.usage_terms.is_empty()
+            && m.copyright_url.is_empty()
+            && m.location.is_empty()
+            && m.city.is_empty()
+            && m.state.is_empty()
+            && m.country.is_empty()
+            && m.gps.is_none()
+    }
+}
+
+/// What forgetting Local records would do / did.
+#[derive(Clone, Debug, Default, PartialEq, Serialize)]
+#[serde(rename_all = "camelCase")]
+pub struct ForgetPlan {
+    /// Local records looked at.
+    pub local: usize,
+    /// Records to forget.
+    pub evict: Vec<PhotoId>,
+    /// Kept: their folder was browsed within the period (or has no time yet).
+    pub kept_recent: usize,
+    /// Kept: changed by the user since they were browsed.
+    pub kept_touched: usize,
+    /// Kept: in an album, a stack, the source of a virtual copy, or in use.
+    pub kept_in_use: usize,
+    /// Folders with Local records but no last-browsed time: stamped now.
+    pub stamp: Vec<String>,
+    /// Folders whose last Local record goes: their time is dropped.
+    pub unstamp: Vec<String>,
+    /// The time stamps are given.
+    #[serde(skip)]
+    pub now: String,
+}
+
+impl ForgetPlan {
+    /// The ops that carry the plan out (empty when there is nothing to do).
+    pub fn ops(&self) -> Vec<Op> {
+        let mut ops: Vec<Op> = self.stamp.iter().map(|f| Op::SetBrowsed { folder: f.clone(), at: Some(self.now.clone()) }).collect();
+        ops.extend(self.evict.chunks(BATCH).map(|c| Op::Batch { ops: c.iter().map(|id| Op::RemovePhoto { id: *id }).collect() }));
+        ops.extend(self.unstamp.iter().map(|f| Op::SetBrowsed { folder: f.clone(), at: None }));
+        ops
+    }
+}
+
+impl Catalog {
+    /// When `folder` was last browsed (ISO 8601), if known.
+    pub fn last_browsed(&self, folder: &str) -> Option<&str> {
+        self.browsed.get(&crate::query::folder_key(folder)).map(String::as_str)
+    }
+
+    /// Folders with a last-browsed time.
+    pub fn browsed_folders(&self) -> &BTreeMap<String, String> {
+        &self.browsed
+    }
+
+    /// Plan forgetting untouched Local records whose folder wasn't browsed for `days` days (as of
+    /// `now`, ISO 8601). `in_use` keeps records the caller still needs. `days == 0` (never) or an
+    /// unreadable `now` forgets nothing. See the module docs.
+    pub fn forget_local_plan(&self, now: &str, days: u32, in_use: &dyn Fn(&Photo) -> bool) -> ForgetPlan {
+        let mut plan = ForgetPlan { now: now.to_string(), ..Default::default() };
+        let now_s = crate::stacks::iso_seconds(now);
+        let max_age = i64::from(days) * 86_400;
+        let mut referenced: HashSet<PhotoId> = HashSet::new();
+        for a in self.albums.values() {
+            referenced.extend(a.photos.iter().copied());
+            referenced.extend(a.cover);
+        }
+        for s in self.stacks.values() {
+            referenced.extend(s.photos.iter().copied());
+        }
+        referenced.extend(self.photos.values().filter_map(|p| p.copy_of));
+        // the fingerprints are the costly part: work them out for the candidates in parallel
+        let candidates: Vec<&Photo> = self
+            .photos
+            .values()
+            .filter(|p| p.local && !referenced.contains(&p.id))
+            .filter(|p| folder_of(p).and_then(|f| self.browsed.get(&f)).is_some_and(|at| is_old(at, now_s, days, max_age)))
+            .map(|p| &**p)
+            .collect();
+        let untouched: HashSet<PhotoId> = untouched_of(&candidates);
+        let mut stamp = BTreeSet::new();
+        let mut remaining: BTreeSet<String> = BTreeSet::new();
+        for p in self.photos.values().filter(|p| p.local) {
+            plan.local += 1;
+            let Some(folder) = folder_of(p) else {
+                plan.kept_in_use += 1;
+                continue;
+            };
+            let recent = match (self.browsed.get(&folder), now_s) {
+                (None, _) => {
+                    stamp.insert(folder.clone());
+                    true
+                }
+                (Some(at), _) => !is_old(at, now_s, days, max_age),
+            };
+            let keep = if recent {
+                plan.kept_recent += 1;
+                true
+            } else if referenced.contains(&p.id) || in_use(p) {
+                plan.kept_in_use += 1;
+                true
+            } else if !untouched.contains(&p.id) {
+                plan.kept_touched += 1;
+                true
+            } else {
+                plan.evict.push(p.id);
+                false
+            };
+            if keep {
+                remaining.insert(folder);
+            }
+        }
+        plan.stamp = stamp.into_iter().collect();
+        plan.unstamp = self.browsed.keys().filter(|f| !remaining.contains(*f)).cloned().collect();
+        plan
+    }
+}
+
+/// A folder browsed at `at` is past the period (never when `days` is 0 or a time is unreadable).
+fn is_old(at: &str, now_s: Option<i64>, days: u32, max_age: i64) -> bool {
+    days > 0 && matches!((now_s, crate::stacks::iso_seconds(at)), (Some(n), Some(a)) if n - a >= max_age)
+}
+
+/// The untouched ones among `photos` ([`Photo::untouched_local`]), checked on several threads
+/// where there are many.
+fn untouched_of(photos: &[&Photo]) -> HashSet<PhotoId> {
+    let check = |chunk: &[&Photo]| chunk.iter().filter(|p| p.untouched_local()).map(|p| p.id).collect::<Vec<_>>();
+    #[cfg(not(target_arch = "wasm32"))]
+    {
+        let threads = std::thread::available_parallelism().map_or(1, |n| n.get()).min(8);
+        if threads > 1 && photos.len() >= 256 {
+            let size = photos.len().div_ceil(threads);
+            return std::thread::scope(|sc| {
+                let jobs: Vec<_> = photos.chunks(size).map(|c| sc.spawn(move || check(c))).collect();
+                // a job that panicked (it can't) leaves its photos out: they are kept
+                jobs.into_iter().filter_map(|j| j.join().ok()).flatten().collect()
+            });
+        }
+    }
+    check(photos).into_iter().collect()
+}
+
+impl Op {
+    /// Every photo the op refers to.
+    pub fn photo_ids(&self, out: &mut impl FnMut(PhotoId)) {
+        match self {
+            Op::AddPhoto { photo } => out(photo.id),
+            Op::RemovePhoto { id }
+            | Op::SetRating { id, .. }
+            | Op::SetFlag { id, .. }
+            | Op::SetLabel { id, .. }
+            | Op::SetDevelop { id, .. }
+            | Op::SetMeta { id, .. }
+            | Op::SetDeleted { id, .. }
+            | Op::SetLocal { id, .. }
+            | Op::SetVersions { id, .. }
+            | Op::SetHistory { id, .. }
+            | Op::PushHistory { id, .. }
+            | Op::SetCaptured { id, .. }
+            | Op::SetAnalysis { id, .. }
+            | Op::SetFile { id, .. }
+            | Op::Relink { id, .. }
+            | Op::SetContent { id, .. } => out(*id),
+            Op::AddAlbum { album } => {
+                album.photos.iter().for_each(|p| out(*p));
+                album.cover.into_iter().for_each(&mut *out);
+            }
+            Op::SetAlbumPhotos { photos, .. } | Op::SetStack { photos, .. } => photos.iter().for_each(|p| out(*p)),
+            Op::SetAlbumCover { cover, .. } => cover.iter().for_each(|p| out(*p)),
+            Op::AddStack { stack } => stack.photos.iter().for_each(|p| out(*p)),
+            Op::Batch { ops } => ops.iter().for_each(|o| o.photo_ids(out)),
+            Op::RemoveAlbum { .. }
+            | Op::RenameAlbum { .. }
+            | Op::MoveAlbum { .. }
+            | Op::SetAlbumOrder { .. }
+            | Op::SetAlbumRules { .. }
+            | Op::RemoveStack { .. }
+            | Op::SetLabelName { .. }
+            | Op::SetBrowsed { .. } => {}
+        }
+    }
+}
@@ -0,0 +1,140 @@
+//! One process per library: an exclusive OS lock on `<library>/catalog.lock`, held for as long as
+//! the library is open (issue #99).
+//!
+//! Two processes on the same library (the app and `lightcraft-cli --library`, two app instances,
+//! two computers on a shared folder) would each keep their own catalog and `seq` counter, and
+//! whichever wrote the last snapshot would silently drop the other's edits. So opening a library
+//! takes the lock first, and a second opener gets [`LockError::InUse`] instead.
+//!
+//! - The lock is the OS's (`flock` on Unix, `LockFileEx` on Windows, via [`std::fs::File::try_lock`]):
+//!   it is released when the [`LibraryLock`] is dropped and, after a crash or a kill, by the OS when
+//!   the process ends. A `catalog.lock` file left behind is therefore never "stale": only a lock
+//!   that's actually held refuses the open.
+//! - Who holds it (process id, computer, program, version) is written to `catalog.lock.owner`, for
+//!   the error message. It's a separate file because a Windows lock also blocks reading the
+//!   locked file. It's only a hint: it may be missing or left over from a crash.
+//! - File systems without lock support (some network shares): the library opens unlocked, with a
+//!   warning — refusing it would lock the user out of their own library. Locks over SMB/NFS are
+//!   advisory and depend on the server; the owner file is then the only hint.
+//! - Browser storage (OPFS) has no file locks: see the web host for its own guard.
+
+use std::fs::{File, OpenOptions, TryLockError};
+use std::io;
+use std::path::{Path, PathBuf};
+
+/// The lock file in the library directory.
+pub const LOCK: &str = "catalog.lock";
+/// Who holds the lock (JSON, see [`LockOwner`]).
+pub const OWNER: &str = "catalog.lock.owner";
+
+/// Who has the library open (as written in [`OWNER`]).
+#[derive(Clone, Debug, Default, PartialEq, serde::Serialize, serde::Deserialize)]
+#[serde(default)]
+pub struct LockOwner {
+    pub pid: u32,
+    /// Computer name, if known.
+    pub host: String,
+    /// The program (`LightCraft`, `lightcraft-cli`…) and its version.
+    pub program: String,
+    pub version: String,
+    /// When it opened the library (seconds since 1970).
+    pub since: u64,
+}
+
+impl LockOwner {
+    /// This process, as `program`.
+    pub fn this_process(program: &str) -> LockOwner {
+        LockOwner {
+            pid: std::process::id(),
+            host: host_name(),
+            program: program.to_string(),
+            version: env!("CARGO_PKG_VERSION").to_string(),
+            since: web_time::SystemTime::now().duration_since(web_time::UNIX_EPOCH).map(|d| d.as_secs()).unwrap_or(0),
+        }
+    }
+
+    /// "LightCraft 0.3.0 (process 123 on studio-mac)".
+    pub fn describe(&self) -> String {
+        let program = if self.program.is_empty() { "another LightCraft" } else { self.program.as_str() };
+        let version = if self.version.is_empty() { String::new() } else { format!(" {}", self.version) };
+        let host = if self.host.is_empty() {
+            String::new()
+        } else if self.host == host_name() {
+            " on this computer".to_string()
+        } else {
+            format!(" on {}", self.host)
+        };
+        format!("{program}{version} (process {}{host})", self.pid)
+    }
+}
+
+/// The computer's name, best effort (empty if unknown).
+fn host_name() -> String {
+    let env = ["COMPUTERNAME", "HOSTNAME"].iter().find_map(|k| std::env::var(k).ok().filter(|h| !h.trim().is_empty()));
+    env.or_else(|| std::fs::read_to_string("/etc/hostname").ok()).map(|h| h.trim().to_string()).filter(|h| !h.is_empty()).unwrap_or_default()
+}
+
+#[derive(Debug, thiserror::Error)]
+pub enum LockError {
+    /// Another process has the library open.
+    #[error("This library is already open in {}. Close it there first.", .0.as_ref().map(LockOwner::describe).unwrap_or_else(|| "another LightCraft program".into()))]
+    InUse(Option<LockOwner>),
+}
+
+/// The exclusive lock on a library directory; released on drop (and by the OS if the process
+/// dies).
+#[derive(Debug)]
+pub struct LibraryLock {
+    dir: PathBuf,
+    /// `None` when the file system can't lock (the library is open unlocked).
+    _file: Option<File>,
+    held: bool,
+}
+
+impl LibraryLock {
+    /// Lock `dir` (which must exist) for this process. Fails only when another process holds it.
+    pub fn acquire(dir: &Path, program: &str) -> Result<LibraryLock, LockError> {
+        let unlocked = |why: io::Error| {
+            log::warn!("library {}: can't lock {LOCK} ({why}); opening it without protection against a second program", dir.display());
+            LibraryLock { dir: dir.to_path_buf(), _file: None, held: false }
+        };
+        let file = match OpenOptions::new().read(true).write(true).create(true).truncate(false).open(dir.join(LOCK)) {
+            Ok(f) => f,
+            Err(e) => return Ok(unlocked(e)),
+        };
+        match file.try_lock() {
+            Ok(()) => {}
+            Err(TryLockError::WouldBlock) => {
+                let owner = std::fs::read(dir.join(OWNER)).ok().and_then(|b| serde_json::from_slice::<LockOwner>(&b).ok());
+                return Err(LockError::InUse(owner));
+            }
+            Err(TryLockError::Error(e)) => return Ok(unlocked(e)),
+        }
+        let owner = LockOwner::this_process(program);
+        if let Ok(json) = serde_json::to_vec_pretty(&owner)
+            && let Err(e) = std::fs::write(dir.join(OWNER), json)
+        {
+            log::warn!("library {}: can't write {OWNER}: {e}", dir.display());
+        }
+        Ok(LibraryLock { dir: dir.to_path_buf(), _file: Some(file), held: true })
+    }
+
+    /// The OS lock is held (`false`: the file system can't lock; opened unprotected).
+    pub fn held(&self) -> bool {
+        self.held
+    }
+
+    pub fn dir(&self) -> &Path {
+        &self.dir
+    }
+}
+
+impl Drop for LibraryLock {
+    /// Remove the owner note while still holding the lock (so it can't delete the next holder's),
+    /// then the file handle closes and the OS releases the lock.
+    fn drop(&mut self) {
+        if self.held {
+            let _ = std::fs::remove_file(self.dir.join(OWNER));
+        }
+    }
+}
@@ -2,7 +2,7 @@

 use std::sync::Arc;

-use lightcraft_develop::DevelopSettings;
+use lightcraft_develop::{DevelopSettings, WbMode};
 use serde::{Deserialize, Serialize};

 #[derive(Clone, Copy, Debug, PartialEq, Eq, PartialOrd, Ord, Hash, Serialize, Deserialize)]
@@ -93,6 +93,65 @@ pub enum Source {
    Demo { scene: u32 },
 }

+/// Copyright status (IPTC / XMP Rights Management `xmpRights:Marked`).
+#[derive(Clone, Copy, Debug, Default, PartialEq, Eq, Hash, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
+pub enum CopyrightStatus {
+    /// Not stated (no `xmpRights:Marked`).
+    #[default]
+    Unknown,
+    /// `xmpRights:Marked` = True.
+    Copyrighted,
+    /// `xmpRights:Marked` = False.
+    PublicDomain,
+}
+
+impl CopyrightStatus {
+    pub const ALL: [CopyrightStatus; 3] = [CopyrightStatus::Unknown, CopyrightStatus::Copyrighted, CopyrightStatus::PublicDomain];
+    /// `unknown` / `copyrighted` / `publicDomain` (also `public domain`, `public-domain`).
+    pub fn parse(s: &str) -> Option<CopyrightStatus> {
+        match s.trim().to_ascii_lowercase().replace([' ', '-', '_'], "").as_str() {
+            "unknown" | "" => Some(CopyrightStatus::Unknown),
+            "copyrighted" => Some(CopyrightStatus::Copyrighted),
+            "publicdomain" => Some(CopyrightStatus::PublicDomain),
+            _ => None,
+        }
+    }
+    /// The id used by commands (`photo.setMeta`'s `copyrightStatus`).
+    pub fn id(self) -> &'static str {
+        match self {
+            CopyrightStatus::Unknown => "unknown",
+            CopyrightStatus::Copyrighted => "copyrighted",
+            CopyrightStatus::PublicDomain => "publicDomain",
+        }
+    }
+    pub fn label(self) -> &'static str {
+        match self {
+            CopyrightStatus::Unknown => "Unknown",
+            CopyrightStatus::Copyrighted => "Copyrighted",
+            CopyrightStatus::PublicDomain => "Public Domain",
+        }
+    }
+    /// As `xmpRights:Marked`: `Some(true)` copyrighted, `Some(false)` public domain.
+    pub fn marked(self) -> Option<bool> {
+        match self {
+            CopyrightStatus::Unknown => None,
+            CopyrightStatus::Copyrighted => Some(true),
+            CopyrightStatus::PublicDomain => Some(false),
+        }
+    }
+    pub fn from_marked(m: Option<bool>) -> CopyrightStatus {
+        match m {
+            None => CopyrightStatus::Unknown,
+            Some(true) => CopyrightStatus::Copyrighted,
+            Some(false) => CopyrightStatus::PublicDomain,
+        }
+    }
+    pub fn is_unknown(&self) -> bool {
+        *self == CopyrightStatus::Unknown
+    }
+}
+
 /// Descriptive + capture metadata.
 #[derive(Clone, Debug, Default, PartialEq, Serialize, Deserialize)]
 #[serde(default)]
@@ -115,8 +174,20 @@ pub struct Meta {
    pub alt_text: String,
    pub extended_description: String,
    pub copyright: String,
+    /// Copyright status, rights usage terms and copyright info URL (IPTC Core rights fields).
+    #[serde(skip_serializing_if = "CopyrightStatus::is_unknown")]
+    pub copyright_status: CopyrightStatus,
+    #[serde(skip_serializing_if = "String::is_empty")]
+    pub usage_terms: String,
+    #[serde(skip_serializing_if = "String::is_empty")]
+    pub copyright_url: String,
    pub creator: String,
    pub keywords: Vec<String>,
+    /// Face/pet/focus regions read from XMP (MWG-RS), on the upright (EXIF-oriented) photo.
+    /// Removing or resizing one edits the catalog only; LightCraft never writes regions to XMP.
+    /// Left out of the catalog JSON when empty (most photos), so older catalogs read unchanged.
+    #[serde(default, skip_serializing_if = "Vec::is_empty")]
+    pub regions: Vec<lightcraft_meta::Region>,
 }

 #[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
@@ -200,6 +271,16 @@ pub struct Photo {
    /// Assisted culling scores (`None` until analysed).
    #[serde(default, skip_serializing_if = "Option::is_none")]
    pub analysis: Option<Analysis>,
+    /// A raw file whose sensor data can't be decoded yet (an unsupported raw variant): why (the
+    /// raw decoder's reason, e.g. "Nikon Huffman-compressed NEF …"). The photo is shown and
+    /// edited from the camera's embedded JPEG preview — a rendered image with the camera's
+    /// picture style baked in — so it is treated as a rendered (non-raw) source.
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub preview_only: Option<String>,
+    /// A Local record: the fingerprint of its state when the browse catalogued it, to tell
+    /// whether the user changed anything since (see [`crate::local`]).
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub local_baseline: Option<u64>,
 }

 /// What assisted culling measured on a photo.
@@ -246,8 +327,23 @@ impl Photo {
            copy_name: None,
            import_look: None,
            analysis: None,
+            preview_only: None,
+            local_baseline: None,
        }
    }
+    /// A raw file developed from its sensor data: not a rendered image, and not a raw shown from
+    /// its embedded preview ([`Photo::preview_only`]).
+    pub fn develops_raw(&self) -> bool {
+        self.kind == MediaKind::Raw && self.preview_only.is_none()
+    }
+    /// The current ARW, NEF, RW2, RAF, CR3, CR2, PEF and SRW readers have vendor WB multipliers but no measured camera
+    /// illuminant. Use adjustments relative to the camera's as-shot look, as for rendered
+    /// photographs (the engine's `camera_preview::file_local_look` covers the same formats; RWL and
+    /// RAW are Leica's and the oldest Panasonic bodies' names for RW2 files).
+    pub fn relative_wb(&self) -> bool {
+        self.develops_raw()
+            && ["ARW", "NEF", "NRW", "RW2", "RWL", "RAW", "RAF", "CR3", "CR2", "PEF", "SRW"].iter().any(|f| self.format.eq_ignore_ascii_case(f))
+    }
    /// The develop settings import gave this photo: [`Photo::camera_defaults`], or the user's
    /// default preset applied on top of them ([`Photo::import_look`]).
    pub fn import_defaults(&self) -> DevelopSettings {
@@ -259,7 +355,8 @@ impl Photo {
    /// The built-in defaults for this photo, before any user default preset: raws start from
    /// their as-shot white balance; embedded lens corrections on when the file has them.
    pub fn camera_defaults(&self) -> DevelopSettings {
-        let mut d = match self.as_shot_wb {
+        let wb = if self.relative_wb() { Some((6500.0, 0.0)) } else { self.as_shot_wb };
+        let mut d = match wb {
            Some((t, tint)) => DevelopSettings::for_raw(t, tint),
            None => DevelopSettings::default(),
        };
@@ -274,7 +371,12 @@ impl Photo {
    }
    /// Edited by the user: settings differ from what import gave the photo.
    pub fn is_edited(&self) -> bool {
-        *self.develop != self.import_defaults() && !self.develop.is_unedited()
+        let mut defaults = self.import_defaults();
+        // As Shot reads the source's current WB; stored numbers can predate a decoder update.
+        if self.develop.wb.mode == WbMode::AsShot && defaults.wb.mode == WbMode::AsShot {
+            defaults.wb = self.develop.wb;
+        }
+        *self.develop != defaults && !self.develop.is_unedited()
    }
    /// Capture time if known, else import time (sort key).
    pub fn date(&self) -> &str {
@@ -303,12 +405,16 @@ pub struct Album {
    /// The Quick Collection: a regular album that B adds to (one per library).
    #[serde(default, skip_serializing_if = "std::ops::Not::not")]
    pub quick: bool,
+    /// Its place among the albums of the same kind in its folder, once the user ordered them by
+    /// hand (see [`crate::Catalog::album_children`]); `None`: listed by name.
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub order: Option<u32>,
 }

 impl Album {
    /// A regular (manual) album.
    pub fn new(id: AlbumId, name: impl Into<String>) -> Album {
-        Album { id, name: name.into(), parent: None, folder: false, photos: Vec::new(), cover: None, smart: None, quick: false }
+        Album { id, name: name.into(), parent: None, folder: false, photos: Vec::new(), cover: None, smart: None, quick: false, order: None }
    }
    pub fn is_smart(&self) -> bool {
        self.smart.is_some()
@@ -332,4 +438,36 @@ mod edited_tests {
        p.develop = Arc::new(DevelopSettings::default());
        assert!(!p.is_edited(), "a full reset is unedited too");
    }
+
+    #[test]
+    fn supported_raws_use_relative_white_balance() {
+        for (name, format, relative) in [
+            ("a.arw", "ARW", true),
+            ("a.nef", "NEF", true),
+            ("a.nrw", "nrw", true),
+            ("a.rw2", "RW2", true),
+            ("a.rwl", "RWL", true),
+            ("a.raw", "RAW", true),
+            ("a.raf", "rAf", true),
+            ("a.cr3", "cr3", true),
+            ("a.srw", "SRW", true),
+            ("a.dng", "DNG", false),
+        ] {
+            let mut p = Photo::new(PhotoId(1), Source::Demo { scene: 0 }, name, format, 10, 10, "2026-10-01T00:00:00");
+            p.kind = MediaKind::Raw;
+            p.as_shot_wb = Some((5200.0, 4.0));
+            assert_eq!(p.relative_wb(), relative, "{format}");
+            let wb = p.camera_defaults().wb;
+            assert_eq!((wb.temp, wb.tint), if relative { (6500.0, 0.0) } else { (5200.0, 4.0) }, "{format}");
+            p.develop = Arc::new(DevelopSettings::for_raw(5200.0, 4.0));
+            assert!(!p.is_edited(), "{format}: old As Shot values are not an edit");
+            let mut custom = (*p.develop).clone();
+            custom.wb.mode = WbMode::Custom;
+            p.develop = Arc::new(custom);
+            assert!(p.is_edited(), "{format}: Custom WB is still an edit");
+            // shown from the embedded preview: a rendered image, not a relative-WB raw
+            p.preview_only = Some("unsupported".into());
+            assert!(!p.relative_wb());
+        }
+    }
 }
@@ -37,8 +37,14 @@ pub struct Filter {
    pub deleted: bool,
    /// Capture date prefix (`2026`, `2026-04`, `2026-04-12`).
    pub date: Option<String>,
+    /// A folder the library's photos were imported from: those photos and the ones imported from
+    /// the folders inside it (see [`crate::folders`]). Photos only browsed in Local are never part
+    /// of it, unlike [`Filter::folder`].
+    pub library_folder: Option<String>,
    /// A keyword; hierarchical keywords match their children too (`travel` finds `travel|italy`).
    pub keyword: Option<String>,
+    /// A person: photos with a named face region of this name (case-insensitive), as read from XMP.
+    pub person: Option<String>,
    pub camera: Option<String>,
    /// Lens (case-insensitive substring).
    pub lens: Option<String>,
@@ -70,6 +76,9 @@ pub enum SortKey {
    FileName,
    Rating,
    FileSize,
+    /// A shuffle fixed by [`Sort::seed`]: the same seed always gives the same order, and photos
+    /// added or edited later never reshuffle the others. Pick a new seed to reshuffle.
+    Random,
 }

 #[derive(Clone, Copy, Debug, PartialEq, Eq, Serialize, Deserialize)]
@@ -79,14 +88,30 @@ pub struct Sort {
    pub ascending: bool,
    /// Date headers in the grid (date sort keys only).
    pub group: crate::GroupBy,
+    /// Which shuffle [`SortKey::Random`] gives (ignored by the other keys).
+    pub seed: u64,
 }

 impl Default for Sort {
    fn default() -> Self {
-        Sort { key: SortKey::CaptureDate, ascending: false, group: crate::GroupBy::Auto }
+        Sort { key: SortKey::CaptureDate, ascending: false, group: crate::GroupBy::Auto, seed: 0 }
    }
 }

+/// splitmix64 finaliser: a stateless, platform-independent mix (no RNG state, no `rand` version
+/// to drift), so a seed names the same shuffle on every machine.
+pub fn mix64(x: u64) -> u64 {
+    let mut z = x.wrapping_add(0x9E37_79B9_7F4A_7C15);
+    z = (z ^ (z >> 30)).wrapping_mul(0xBF58_476D_1CE4_E5B9);
+    z = (z ^ (z >> 27)).wrapping_mul(0x94D0_49BB_1331_11EB);
+    z ^ (z >> 31)
+}
+
+/// A photo's place in the shuffle for `seed`.
+fn shuffle_rank(seed: u64, id: PhotoId) -> u64 {
+    mix64(mix64(seed) ^ id.0)
+}
+
 fn token_matches(p: &Photo, tok: &str) -> bool {
    let t = tok.to_lowercase();
    if let Some((field, val)) = t.split_once(':') {
@@ -113,9 +138,10 @@ fn token_matches(p: &Photo, tok: &str) -> bool {
            "camera" => p.meta.camera.to_lowercase().contains(val),
            "lens" => p.meta.lens.to_lowercase().contains(val),
            "keyword" | "kw" => p.meta.keywords.iter().any(|k| crate::keywords::is_under(k, val)),
+            "person" | "who" => has_person(p, val),
            "type" | "kind" => format!("{:?}", p.kind).eq_ignore_ascii_case(val),
            "edited" => (val == "true" || val == "yes") == p.is_edited(),
-            "date" => p.date().starts_with(val),
+            "date" => p.captured.as_deref().is_some_and(|c| c.starts_with(val)),
            "copy" | "virtual" => (val == "true" || val == "yes") == p.copy_of.is_some(),
            "name" | "file" => p.file_name.to_lowercase().contains(val),
            _ => false,
@@ -125,6 +151,12 @@ fn token_matches(p: &Photo, tok: &str) -> bool {
    hay.iter().any(|h| h.to_lowercase().contains(&t)) || p.meta.keywords.iter().any(|k| k.to_lowercase().contains(&t))
 }

+/// Whether `p` has a named face region called `name` (case-insensitive).
+fn has_person(p: &Photo, name: &str) -> bool {
+    let name = name.trim().to_lowercase();
+    p.meta.regions.iter().any(|r| r.kind == lightcraft_meta::RegionKind::Face && r.name.as_deref().is_some_and(|n| n.to_lowercase() == name))
+}
+
 impl Filter {
    /// Human-readable summary of the active rules (smart album tooltips, `album.list`).
    pub fn describe(&self) -> String {
@@ -152,9 +184,15 @@ impl Filter {
        if let Some(e) = self.edited {
            v.push(if e { "edited".into() } else { "unedited".into() });
        }
-        for (name, val) in
-            [("keyword", &self.keyword), ("camera", &self.camera), ("lens", &self.lens), ("date", &self.date), ("imported", &self.imported)]
-        {
+        for (name, val) in [
+            ("keyword", &self.keyword),
+            ("person", &self.person),
+            ("camera", &self.camera),
+            ("lens", &self.lens),
+            ("date", &self.date),
+            ("folder", &self.library_folder),
+            ("imported", &self.imported),
+        ] {
            if let Some(x) = val {
                v.push(format!("{name} {x}"));
            }
@@ -174,7 +212,25 @@ impl Filter {
        if v.is_empty() { "all photos".into() } else { v.join(", ") }
    }

+    /// Whether matches depend on the clock (only "in the last…" rules do).
+    pub fn depends_on_now(&self) -> bool {
+        self.rule_set.as_ref().is_some_and(crate::RuleSet::depends_on_now)
+    }
+
    pub fn matches(&self, p: &Photo, cat: &Catalog) -> bool {
+        self.matches_in(p, cat, self.library_root().as_deref())
+    }
+
+    /// [`Filter::library_folder`] as a folder identity ([`folder_key`]); `None` without a choice.
+    fn library_root(&self) -> Option<String> {
+        // as given: a folder may be named with spaces at its edge, and trimming would make it
+        // another folder
+        self.library_folder.as_deref().filter(|d| !d.trim().is_empty()).map(folder_key)
+    }
+
+    /// [`Filter::matches`] with the library folder already turned into its identity, so a query
+    /// over a big library does that once, not once per photo.
+    fn matches_in(&self, p: &Photo, cat: &Catalog, root: Option<&str>) -> bool {
        if p.deleted != self.deleted {
            return false;
        }
@@ -228,12 +284,18 @@ impl Filter {
            return false;
        }
        if let Some(from) = &self.date_from
-            && p.date() < from.as_str()
+            && match p.captured.as_deref() {
+                Some(c) => c < from.as_str(),
+                None => true,
+            }
        {
            return false;
        }
        if let Some(to) = &self.date_to
-            && p.date().get(..to.len()).unwrap_or(p.date()) > to.as_str()
+            && match p.captured.as_deref() {
+                Some(c) => c.get(..to.len()).unwrap_or(c) > to.as_str(),
+                None => true,
+            }
        {
            return false;
        }
@@ -243,7 +305,12 @@ impl Filter {
            return false;
        }
        if let Some(d) = &self.date
-            && !p.date().starts_with(d.as_str())
+            && !p.captured.as_deref().is_some_and(|c| c.starts_with(d.as_str()))
+        {
+            return false;
+        }
+        if let Some(root) = root
+            && !photo_in_root(p, root)
        {
            return false;
        }
@@ -262,6 +329,11 @@ impl Filter {
        {
            return false;
        }
+        if let Some(n) = &self.person
+            && !has_person(p, n)
+        {
+            return false;
+        }
        if let Some(c) = &self.camera
            && !p.meta.camera.eq_ignore_ascii_case(c)
        {
@@ -274,15 +346,17 @@ impl Filter {
 impl Catalog {
    /// Photos matching `filter`, in `sort` order (ties broken by id for stability).
    pub fn query(&self, filter: &Filter, sort: &Sort) -> Vec<PhotoId> {
-        let mut v: Vec<&Photo> = self.photos().map(|p| p.as_ref()).filter(|p| filter.matches(p, self)).collect();
+        let root = filter.library_root();
+        let mut v: Vec<&Photo> = self.photos().map(|p| p.as_ref()).filter(|p| filter.matches_in(p, self, root.as_deref())).collect();
        v.sort_by(|a, b| {
            let o = match sort.key {
-                SortKey::CaptureDate => a.date().cmp(b.date()),
+                SortKey::CaptureDate => a.captured.cmp(&b.captured).then_with(|| a.imported.cmp(&b.imported)),
                SortKey::ImportDate => a.imported.cmp(&b.imported),
                SortKey::EditDate => a.edited.cmp(&b.edited),
                SortKey::FileName => a.file_name.to_lowercase().cmp(&b.file_name.to_lowercase()),
                SortKey::Rating => a.rating.cmp(&b.rating),
                SortKey::FileSize => a.file_size.cmp(&b.file_size),
+                SortKey::Random => shuffle_rank(sort.seed, a.id).cmp(&shuffle_rank(sort.seed, b.id)),
            }
            .then(a.id.cmp(&b.id));
            if sort.ascending { o } else { o.reverse() }
@@ -295,7 +369,7 @@ impl Catalog {
        use std::collections::BTreeMap;
        let mut years: BTreeMap<&str, (BTreeMap<&str, usize>, BTreeMap<&str, usize>)> = BTreeMap::new();
        for p in self.photos().filter(|p| p.in_library()) {
-            let d = p.date();
+            let Some(d) = p.captured.as_deref() else { continue };
            if d.len() >= 10 {
                let (months, days) = years.entry(&d[..4]).or_default();
                *months.entry(&d[..7]).or_default() += 1;
@@ -325,6 +399,55 @@ impl Catalog {
        }
        m.into_iter().collect()
    }
+
+    /// The people named on faces in the library (MWG regions read from XMP): how many photos each
+    /// appears in and the photo showing their largest face (for a card's picture). Most photos first,
+    /// then by name. Names that differ only in case are one person, shown as first seen; a person
+    /// twice in one photo counts once.
+    pub fn people(&self) -> Vec<Person> {
+        self.people_in(&Filter::default())
+    }
+
+    /// [`Self::people`] among the photos `filter` lets through (its `person` is ignored): what the
+    /// People view offers while other filters (a date, a rating, an album…) are active, so picking
+    /// a person never ends in an empty grid.
+    pub fn people_in(&self, filter: &Filter) -> Vec<Person> {
+        let filter = Filter { person: None, ..filter.clone() };
+        let mut m: std::collections::HashMap<String, (Person, f64)> = Default::default();
+        let root = filter.library_root();
+        for p in self.photos().filter(|p| filter.matches_in(p, self, root.as_deref())) {
+            let mut seen: Vec<String> = Vec::new();
+            for r in p.meta.regions.iter().filter(|r| r.kind == lightcraft_meta::RegionKind::Face) {
+                let Some(name) = r.name.as_deref().map(str::trim).filter(|n| !n.is_empty()) else { continue };
+                let key = name.to_lowercase();
+                // the face's size in pixels of the photo
+                let area = (r.rect.x1 - r.rect.x0) * (r.rect.y1 - r.rect.y0) * p.width as f64 * p.height as f64;
+                let (person, best) =
+                    m.entry(key.clone()).or_insert_with(|| (Person { name: name.to_string(), count: 0, photo: p.id, face: r.rect }, area));
+                if !seen.contains(&key) {
+                    person.count += 1;
+                    seen.push(key);
+                }
+                if area > *best || (area == *best && p.id < person.photo) {
+                    (person.photo, person.face, *best) = (p.id, r.rect, area);
+                }
+            }
+        }
+        let mut v: Vec<Person> = m.into_values().map(|(p, _)| p).collect();
+        v.sort_by(|a, b| b.count.cmp(&a.count).then_with(|| a.name.to_lowercase().cmp(&b.name.to_lowercase())));
+        v
+    }
+}
+
+/// A person named on faces in the library ([`Catalog::people`]).
+#[derive(Clone, Debug, PartialEq)]
+pub struct Person {
+    pub name: String,
+    /// Photos they appear in.
+    pub count: usize,
+    /// The photo with their largest face, and that face (normalized, in the photo's upright frame).
+    pub photo: PhotoId,
+    pub face: lightcraft_meta::Rect,
 }

 #[derive(Clone, Debug, PartialEq, Serialize)]
@@ -346,6 +469,112 @@ pub fn in_folder(path: &str, dir: &str, deep: bool) -> bool {
    !rest.is_empty() && (deep || !rest.contains(['/', '\\']))
 }

+/// A path read the way [`folder_key`] reads it: the part `..` can't climb above (`//server/share`,
+/// `c:`; empty for `/` and relative paths), the names below it, and whether it is absolute.
+pub(crate) struct SplitPath {
+    pub prefix: String,
+    pub parts: Vec<String>,
+    pub absolute: bool,
+}
+
+/// Both separators, repeated separators (also between a server and its share), `.` and a
+/// trailing separator dropped, `..` resolved lexically, a Windows verbatim prefix removed;
+/// `lower_drive` lower-cases a drive letter (the identity does, a display name wants it as is).
+pub(crate) fn split_path(path: &str, lower_drive: bool) -> SplitPath {
+    let mut s = path.replace('\\', "/");
+    if let Some(rest) = s.strip_prefix("//?/").or_else(|| s.strip_prefix("//./")) {
+        s = match rest.strip_prefix("UNC/") {
+            Some(unc) => format!("//{unc}"),
+            None => rest.to_string(),
+        };
+    }
+    let (prefix, rest) = if let Some(unc) = s.strip_prefix("//").filter(|u| !u.is_empty() && !u.starts_with('/')) {
+        let mut segs = unc.split('/').filter(|x| !x.is_empty());
+        let (server, share) = (segs.next().unwrap_or(""), segs.next().unwrap_or(""));
+        (format!("//{server}/{share}"), segs.collect::<Vec<_>>().join("/"))
+    } else if let (true, Some(drive), Some(rest)) = (matches!(s.as_bytes(), [a, b':', ..] if a.is_ascii_alphabetic()), s.get(..2), s.get(2..)) {
+        (if lower_drive { drive.to_ascii_lowercase() } else { drive.to_string() }, rest.to_string())
+    } else {
+        (String::new(), s.clone())
+    };
+    let absolute = rest.starts_with('/') || !prefix.is_empty();
+    let mut parts: Vec<String> = Vec::new();
+    for c in rest.split('/') {
+        match c {
+            "" | "." => {}
+            ".." if parts.last().is_some_and(|p| p != "..") => {
+                parts.pop();
+            }
+            ".." if absolute => {}
+            c => parts.push(c.to_string()),
+        }
+    }
+    SplitPath { prefix, parts, absolute }
+}
+
+/// A folder path's identity, for telling whether two spellings name the same folder: `/` and
+/// `\\` are both separators, repeated separators, `.` and a trailing separator are dropped,
+/// `..` is resolved lexically, a Windows verbatim prefix (`\\?\`) is removed and the drive
+/// letter lower-cased; on Windows (case-insensitive file names) the whole path is lower-cased.
+/// File names are otherwise compared as written, so on a case-insensitive disk elsewhere
+/// (macOS, a Samba share) two spellings that differ in case are two folders here.
+/// No file-system access: the paths compared should already be absolute.
+pub fn folder_key(path: &str) -> String {
+    let SplitPath { prefix, parts, absolute } = split_path(path, true);
+    let key = if absolute { format!("{prefix}/{}", parts.join("/")) } else { parts.join("/") };
+    if cfg!(windows) { key.to_lowercase() } else { key }
+}
+
+/// Whether `path` is folder `root` or lies somewhere inside it, however either is spelled
+/// (see [`folder_key`]).
+pub fn folder_within(path: &str, root: &str) -> bool {
+    key_within(&folder_key(path), &folder_key(root))
+}
+
+/// The names below folder `root` that lead to `path`, as the caller spelled them (`None` when
+/// `path` is not at or inside `root`; empty when it is `root`). Compared by [`folder_key`], but
+/// the names keep their case, which the key lowers on Windows.
+pub fn folder_rest(path: &str, root: &str) -> Option<Vec<String>> {
+    if !key_within(&folder_key(path), &folder_key(root)) {
+        return None;
+    }
+    let depth = split_path(root, false).parts.len();
+    Some(split_path(path, false).parts.into_iter().skip(depth).collect())
+}
+
+/// Whether a library photo (not one only browsed in Local) lies in the folder whose
+/// [`folder_key`] is `root`. A plain POSIX path (no `\`, `.`, `..`, repeated or trailing
+/// separators) is its own identity, so it is read without building a key for it: the check runs
+/// for every photo of the library on each refresh, and this keeps it cheap.
+pub(crate) fn photo_in_root(p: &Photo, root: &str) -> bool {
+    if p.local {
+        return false;
+    }
+    let crate::Source::File { path } = &p.source else { return false };
+    let plain = path.starts_with('/')
+        && !path.starts_with("//")
+        && !path.ends_with('/')
+        && !path.contains(['\\'])
+        && !path.contains("//")
+        && !path.contains("/.");
+    if !cfg!(windows) && plain && root.starts_with('/') && !root.starts_with("//") {
+        let r = root.trim_end_matches('/');
+        return path.strip_prefix(r).is_some_and(|rest| rest.starts_with('/'));
+    }
+    crate::local::folder_of(p).is_some_and(|f| key_within(&f, root))
+}
+
+/// [`folder_within`] for two paths already turned into their [`folder_key`].
+pub(crate) fn key_within(p: &str, r: &str) -> bool {
+    // `.` or `a/..` name no folder: not "everything"
+    if r.is_empty() {
+        return false;
+    }
+    // a disk's root key ends in `/` (`c:/`, `//srv/share/`), a folder's does not: same folder
+    let (p, r) = (p.trim_end_matches('/'), r.trim_end_matches('/'));
+    p == r || p.strip_prefix(r).is_some_and(|rest| rest.starts_with('/'))
+}
+
 /// The kind of Photo Merge result a file is, from the name merges give it (`IMG_1-HDR.dng`,
 /// `IMG_1-Pano.dng`, `IMG_1-HDR-Pano.dng` — also the names other raw editors use):
 /// `hdr`, `panorama` or `hdrPanorama`.
@@ -7,7 +7,7 @@
 use serde::{Deserialize, Serialize};
 use serde_json::Value;

-use crate::{Catalog, Photo};
+use crate::{Catalog, Photo, Source};

 #[derive(Clone, Copy, Debug, Default, PartialEq, Eq, Serialize, Deserialize)]
 #[serde(rename_all = "lowercase")]
@@ -62,6 +62,7 @@ pub const FIELDS: &[(&str, &str, Kind)] = &[
    ("keywords", "Keywords", Kind::Keywords),
    ("text", "Any Searchable Text", Kind::Text),
    ("fileName", "Filename", Kind::Text),
+    ("filePath", "File Path", Kind::Text),
    ("format", "File Format", Kind::Text),
    ("title", "Title", Kind::Text),
    ("caption", "Caption", Kind::Text),
@@ -70,6 +71,7 @@ pub const FIELDS: &[(&str, &str, Kind)] = &[
    ("location", "Location", Kind::Text),
    ("creator", "Creator", Kind::Text),
    ("copyright", "Copyright", Kind::Text),
+    ("copyrightStatus", "Copyright Status", Kind::Choice(&["copyrighted", "publicDomain", "unknown"])),
    ("captureDate", "Capture Date", Kind::Date),
    ("importDate", "Import Date", Kind::Date),
    ("editDate", "Edit Date", Kind::Date),
@@ -303,6 +305,20 @@ impl Rule {
                text(&all)
            }
            "fileName" => text(&p.file_name),
+            "filePath" => {
+                // `/` and `\` both separate, whatever platform the catalog came from; demo photos have no path
+                let path = match &p.source {
+                    Source::File { path } => path.replace('\\', "/"),
+                    Source::Demo { .. } => String::new(),
+                };
+                let want = want.replace('\\', "/");
+                match op {
+                    // the whole string, spaces included (not word by word like the other text fields)
+                    "contains" => !want.trim().is_empty() && path.to_lowercase().contains(want.trim()),
+                    "notContains" => want.trim().is_empty() || !path.to_lowercase().contains(want.trim()),
+                    _ => text_op(op, &path, &want),
+                }
+            }
            "format" => text(&p.format),
            "title" => text(&m.title),
            "caption" => text(&m.caption),
@@ -311,6 +327,10 @@ impl Rule {
            "location" => text(&[m.location.as_str(), &m.city, &m.state, &m.country].join(" ")),
            "creator" => text(&m.creator),
            "copyright" => text(&m.copyright),
+            "copyrightStatus" => {
+                let want = crate::CopyrightStatus::parse(&want);
+                (want == Some(m.copyright_status)) == (op == "is")
+            }
            "captureDate" => date_op(op, p.captured.as_deref(), value),
            "importDate" => date_op(op, Some(&p.imported), value),
            "editDate" => date_op(op, p.edited.as_deref(), value),
@@ -341,6 +361,15 @@ impl RuleSet {
        }
    }

+    /// Whether matches depend on the clock ("in the last…" rules, nested groups included): the
+    /// same photos can enter or leave the set without any catalog change.
+    pub fn depends_on_now(&self) -> bool {
+        self.rules.iter().any(|r| match r {
+            Rule::Group { group } => group.depends_on_now(),
+            Rule::Field { op, .. } => op == "inLast" || op == "notInLast",
+        })
+    }
+
    /// Unknown fields or operators (for command validation), as readable messages.
    pub fn problems(&self) -> Vec<String> {
        let mut out = Vec::new();
@@ -431,6 +460,17 @@ mod tests {
        );
        no(json!({"rules": [{"field": "keywords", "op": "isEmpty"}]}));
        yes(json!({"rules": [{"field": "fileName", "op": "startsWith", "value": "img_"}, {"field": "fileName", "op": "endsWith", "value": ".cr2"}]}));
+        // file path: the whole string (spaces included), either separator, case-insensitive
+        let mut fp = photo(3);
+        fp.source = Source::File { path: "D:\\Photos\\Aliah Ira Polanco-Grylls\\2026\\IMG_0042.CR2".into() };
+        let fpm = |v: serde_json::Value| rs(v).matches(&fp, &cat);
+        assert!(fpm(json!({"rules": [{"field": "filePath", "op": "contains", "value": "/Aliah Ira Polanco-Grylls/"}]})));
+        assert!(!fpm(json!({"rules": [{"field": "filePath", "op": "contains", "value": "/Ira Aliah/"}]})));
+        assert!(!fpm(json!({"rules": [{"field": "filePath", "op": "contains", "value": "Polanco Grylls"}]})));
+        assert!(fpm(json!({"rules": [{"field": "filePath", "op": "notContains", "value": "/Other/"}]})));
+        assert!(fpm(json!({"rules": [{"field": "filePath", "op": "startsWith", "value": "d:/photos/"}]})));
+        no(json!({"rules": [{"field": "filePath", "op": "contains", "value": "/Aliah/"}]})); // demo photo has no path
+        yes(json!({"rules": [{"field": "filePath", "op": "isEmpty"}]}));
        yes(json!({"rules": [{"field": "camera", "op": "contains", "value": "x2"}, {"field": "title", "op": "isEmpty"}]}));
        yes(json!({"rules": [{"field": "iso", "op": "between", "value": [800, 3200]}, {"field": "aperture", "op": "lte", "value": "f/4"}]}));
        yes(json!({"rules": [{"field": "megapixels", "op": "gte", "value": 24}]}));
@@ -440,6 +480,13 @@ mod tests {
        yes(json!({"rules": [{"field": "captureDate", "op": "between", "value": ["2026-08-01", "2026-08"]}]}));
        no(json!({"rules": [{"field": "captureDate", "op": "after", "value": "2026-08"}]}));
        yes(json!({"rules": [{"field": "kind", "op": "isNot", "value": "video"}, {"field": "edited", "op": "is", "value": false}]}));
+        // copyright status (unknown until set)
+        yes(json!({"rules": [{"field": "copyrightStatus", "op": "is", "value": "unknown"}]}));
+        no(json!({"rules": [{"field": "copyrightStatus", "op": "is", "value": "copyrighted"}]}));
+        let mut pd = photo(2);
+        pd.meta.copyright_status = crate::CopyrightStatus::PublicDomain;
+        assert!(rs(json!({"rules": [{"field": "copyrightStatus", "op": "is", "value": "publicDomain"}]})).matches(&pd, &cat));
+        assert!(rs(json!({"rules": [{"field": "copyrightStatus", "op": "isNot", "value": "copyrighted"}]})).matches(&pd, &cat));
        // nested: rating ≥ 4 and (label is blue or keywords contain food)
        yes(json!({"rules": [{"field": "rating", "op": "gte", "value": 4}, {"group": {"match": "any", "rules": [
            {"field": "label", "op": "is", "value": "blue"}, {"field": "keywords", "op": "contains", "value": "food"}]}}]}));
@@ -0,0 +1,388 @@
+//! Crash-safe writes of user-visible files: exports, renders, DNG conversions, smart previews and
+//! XMP sidecars all go through here (the catalog's own files use [`crate::store::FsStore`]).
+//!
+//! **Write policy** (issue #134): what can't be recreated — originals, XMP sidecars, the catalog,
+//! settings, DNGs that replace raws, smart previews (the only stand-in while an original is
+//! offline), an edit copy that becomes a library photo — is written **durably**: synced to disk
+//! before it counts as written. What can always be made again from the originals — exports,
+//! renders, screenshots, merge previews — is written **atomically** only: a half-written file
+//! never replaces a good one, but there is no per-file sync (on a USB drive or a NAS that sync
+//! dominates a large export).
+//!
+//! - [`write_atomic`] replaces a file so that after a crash, a full disk or an unplugged drive
+//!   either the old or the complete new content is there — never a truncated mix. The content
+//!   goes to a new temp file in the same folder (unique name, `create_new`), is synced to disk and
+//!   renamed over the target; then the folder is synced (POSIX). Any failure removes the temp
+//!   file and leaves the target as it was.
+//! - [`write_atomic_nosync`] is the same temp file + rename without the two syncs, for
+//!   recreatable outputs only: a failed write still removes the temp file and leaves the target as
+//!   it was, but after a power cut the new file may be lost (on some file systems left empty) —
+//!   it is then simply exported again.
+//! - [`write_new`] / [`write_new_unique`] write a file that must not exist yet: the same temp file
+//!   and sync, plus a read-back check, and the file then appears under its final name **without
+//!   replacing anything** (a hard link, which fails on a taken name; a check-then-rename where the
+//!   volume has no hard links).
+//! - [`same_file`] tells whether two paths name the same file on disk (device + inode on Unix,
+//!   resolved paths elsewhere), for "is this target an original?" checks.
+
+use std::cell::Cell;
+use std::fs::{self, File, OpenOptions};
+use std::io::{self, Write};
+use std::path::{Path, PathBuf};
+use std::sync::atomic::{AtomicU64, Ordering};
+
+thread_local! {
+    static FAIL_AFTER: Cell<Option<u64>> = const { Cell::new(None) };
+    static SYNCS: Cell<u64> = const { Cell::new(0) };
+}
+
+/// How many file and folder syncs this module made **on the calling thread** (for tests: a
+/// durable write path syncs, a recreatable output doesn't).
+#[doc(hidden)]
+pub fn syncs_on_this_thread() -> u64 {
+    SYNCS.with(Cell::get)
+}
+
+fn sync_file(f: &File) -> io::Result<()> {
+    SYNCS.with(|c| c.set(c.get() + 1));
+    f.sync_all()
+}
+
+/// Fault injection for tests (in any crate): while the returned guard lives, every write made by
+/// this module **on the calling thread** fails once `bytes` bytes of a file were written — as a
+/// full disk or an unplugged drive would.
+#[doc(hidden)]
+pub fn fail_writes_after(bytes: u64) -> FaultGuard {
+    FAIL_AFTER.with(|c| c.set(Some(bytes)));
+    FaultGuard(())
+}
+
+/// Ends a [`fail_writes_after`] when dropped.
+#[doc(hidden)]
+pub struct FaultGuard(());
+
+impl Drop for FaultGuard {
+    fn drop(&mut self) {
+        FAIL_AFTER.with(|c| c.set(None));
+    }
+}
+
+/// Counts what is written; fails past an injected limit.
+struct Limited<W> {
+    inner: W,
+    n: u64,
+    limit: Option<u64>,
+}
+
+impl<W: Write> Write for Limited<W> {
+    fn write(&mut self, buf: &[u8]) -> io::Result<usize> {
+        let buf = match self.limit {
+            Some(l) => {
+                let room = usize::try_from(l.saturating_sub(self.n)).unwrap_or(usize::MAX);
+                if room == 0 && !buf.is_empty() {
+                    return Err(io::Error::other("write failed (injected fault)"));
+                }
+                buf.get(..room.min(buf.len())).unwrap_or(buf)
+            }
+            None => buf,
+        };
+        let k = self.inner.write(buf)?;
+        self.n += k as u64;
+        Ok(k)
+    }
+    fn flush(&mut self) -> io::Result<()> {
+        self.inner.flush()
+    }
+}
+
+fn parent_of(path: &Path) -> &Path {
+    path.parent().filter(|d| !d.as_os_str().is_empty()).unwrap_or(Path::new("."))
+}
+
+/// Make a rename or a new name in `dir` durable (POSIX). Opening a directory fails on Windows,
+/// where NTFS journals the rename itself, so errors are ignored.
+pub fn sync_dir(dir: &Path) {
+    if let Ok(d) = File::open(dir) {
+        let _ = sync_file(&d);
+    }
+}
+
+/// A new, empty temp file next to `path` (hidden, unique in this process and folder).
+fn create_temp(path: &Path) -> io::Result<(PathBuf, File)> {
+    static N: AtomicU64 = AtomicU64::new(0);
+    if path.file_name().is_none() {
+        return Err(io::Error::new(io::ErrorKind::InvalidInput, format!("{} is not a file path", path.display())));
+    }
+    let dir = parent_of(path);
+    for _ in 0..1000 {
+        let tmp = dir.join(format!(".lightcraft-{}-{}.tmp", std::process::id(), N.fetch_add(1, Ordering::Relaxed)));
+        match OpenOptions::new().write(true).create_new(true).open(&tmp) {
+            Ok(f) => return Ok((tmp, f)),
+            Err(e) if e.kind() == io::ErrorKind::AlreadyExists => continue,
+            Err(e) => return Err(e),
+        }
+    }
+    Err(io::Error::new(io::ErrorKind::AlreadyExists, format!("no free temp file name in {}", dir.display())))
+}
+
+/// Write a temp file next to `path` with what `fill` writes, synced to disk when `sync`. On
+/// failure the temp file is removed. Returns its path and length.
+fn write_temp(path: &Path, sync: bool, fill: &mut dyn FnMut(&mut dyn Write) -> io::Result<()>) -> io::Result<(PathBuf, u64)> {
+    let (tmp, file) = create_temp(path)?;
+    let r = (|| {
+        let limit = FAIL_AFTER.with(Cell::get);
+        let mut w = Limited { inner: io::BufWriter::with_capacity(1 << 20, file), n: 0, limit };
+        fill(&mut w)?;
+        w.flush()?;
+        let n = w.n;
+        let f = w.inner.into_inner().map_err(|e| e.into_error())?;
+        if sync {
+            sync_file(&f)?;
+        }
+        Ok(n)
+    })();
+    match r {
+        Ok(n) => Ok((tmp, n)),
+        Err(e) => {
+            let _ = fs::remove_file(&tmp);
+            Err(e)
+        }
+    }
+}
+
+/// Replace (or create) `path` with `data` atomically and durably; see the module docs. The
+/// folder must exist.
+pub fn write_atomic(path: &Path, data: &[u8]) -> io::Result<()> {
+    write_atomic_with(path, &mut |w| w.write_all(data)).map(|_| ())
+}
+
+/// [`write_atomic`] with streamed content: `fill` writes it (an error aborts and leaves the old
+/// file). Returns the length written.
+pub fn write_atomic_with(path: &Path, fill: &mut dyn FnMut(&mut dyn Write) -> io::Result<()>) -> io::Result<u64> {
+    replace_with(path, true, fill)
+}
+
+/// Replace (or create) `path` with `data` atomically but **without syncing** it to disk — only for
+/// outputs that can always be recreated (exports, renders, screenshots); see the module docs.
+/// Anything that can't be made again goes through [`write_atomic`].
+pub fn write_atomic_nosync(path: &Path, data: &[u8]) -> io::Result<()> {
+    replace_with(path, false, &mut |w| w.write_all(data)).map(|_| ())
+}
+
+/// Temp file (synced when `sync`) renamed over `path`, then the folder synced when `sync`.
+fn replace_with(path: &Path, sync: bool, fill: &mut dyn FnMut(&mut dyn Write) -> io::Result<()>) -> io::Result<u64> {
+    let (tmp, n) = write_temp(path, sync, fill)?;
+    if let Err(e) = fs::rename(&tmp, path) {
+        let _ = fs::remove_file(&tmp);
+        return Err(e);
+    }
+    if sync {
+        sync_dir(parent_of(path));
+    }
+    Ok(n)
+}
+
+/// Write `data` as the new file `path`: never replaces an existing file (`AlreadyExists`), and
+/// the file only appears once it is complete, synced and read back identical.
+pub fn write_new(path: &Path, data: &[u8]) -> io::Result<()> {
+    write_new_unique(&mut std::iter::once(path.to_path_buf()), data).map(|_| ())
+}
+
+/// [`write_new`] under the first free name of `candidates` (all in the same folder, e.g.
+/// `IMG_1.dng`, `IMG_1-2.dng`, …): the file is written once and then given the first name that
+/// is not taken. Returns that name; `AlreadyExists` when every candidate is taken.
+pub fn write_new_unique(candidates: &mut dyn Iterator<Item = PathBuf>, data: &[u8]) -> io::Result<PathBuf> {
+    let mut candidates = candidates.peekable();
+    let Some(first) = candidates.peek().cloned() else {
+        return Err(io::Error::new(io::ErrorKind::InvalidInput, "no file name to write"));
+    };
+    let (tmp, _) = write_temp(&first, true, &mut |w| w.write_all(data))?;
+    let r = (|| {
+        // Kept (issue #134 point 4): these files replace raws (Convert / Copy as DNG) or become
+        // library photos (merges). On a local disk the read usually comes from the cache, but on a
+        // network volume it can go back to the server, and next to encoding a DNG it costs little.
+        if !same_content(&tmp, data)? {
+            return Err(io::Error::other("the written file reads back different (failing drive or connection?)"));
+        }
+        for path in candidates.by_ref().take(100_000) {
+            match publish_new(&tmp, &path) {
+                Ok(()) => {
+                    sync_dir(parent_of(&path));
+                    return Ok(path);
+                }
+                Err(e) if e.kind() == io::ErrorKind::AlreadyExists => continue,
+                Err(e) => return Err(e),
+            }
+        }
+        Err(io::Error::new(io::ErrorKind::AlreadyExists, format!("{}: no free file name", first.display())))
+    })();
+    // after a hard link the temp name is a second name of the new file; after a rename it's gone
+    let _ = fs::remove_file(&tmp);
+    r
+}
+
+/// Give the complete temp file `tmp` the name `path`, unless that is taken.
+fn publish_new(tmp: &Path, path: &Path) -> io::Result<()> {
+    match fs::hard_link(tmp, path) {
+        Ok(()) => Ok(()),
+        Err(e) if e.kind() == io::ErrorKind::AlreadyExists => Err(e),
+        // no hard links on this volume (FAT/exFAT, some shares): check, then rename
+        Err(_) => {
+            if fs::symlink_metadata(path).is_ok() {
+                return Err(io::Error::new(io::ErrorKind::AlreadyExists, format!("{} exists", path.display())));
+            }
+            fs::rename(tmp, path)
+        }
+    }
+}
+
+/// Does the file at `path` hold exactly `data`?
+fn same_content(path: &Path, data: &[u8]) -> io::Result<bool> {
+    use std::io::Read;
+    let mut f = File::open(path)?;
+    if f.metadata()?.len() != data.len() as u64 {
+        return Ok(false);
+    }
+    let mut buf = vec![0u8; 1 << 20];
+    let mut at = 0usize;
+    loop {
+        let k = match f.read(&mut buf) {
+            Ok(0) => return Ok(at == data.len()),
+            Ok(k) => k,
+            Err(e) if e.kind() == io::ErrorKind::Interrupted => continue,
+            Err(e) => return Err(e),
+        };
+        if data.get(at..at + k) != buf.get(..k) {
+            return Ok(false);
+        }
+        at += k;
+    }
+}
+
+/// Do `a` and `b` name the same existing file (through links, `..`, case-insensitive names)?
+pub fn same_file(a: &Path, b: &Path) -> bool {
+    let (Ok(ma), Ok(mb)) = (fs::metadata(a), fs::metadata(b)) else { return false };
+    #[cfg(unix)]
+    {
+        use std::os::unix::fs::MetadataExt;
+        ma.dev() == mb.dev() && ma.ino() == mb.ino()
+    }
+    #[cfg(not(unix))]
+    {
+        if ma.len() != mb.len() {
+            return false;
+        }
+        match (fs::canonicalize(a), fs::canonicalize(b)) {
+            (Ok(x), Ok(y)) => x == y || x.to_string_lossy().to_lowercase() == y.to_string_lossy().to_lowercase(),
+            _ => false,
+        }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    fn temp(tag: &str) -> PathBuf {
+        let d = std::env::temp_dir().join(format!("lc-safe-file-{tag}-{}", std::process::id()));
+        let _ = fs::remove_dir_all(&d);
+        fs::create_dir_all(&d).unwrap();
+        d
+    }
+
+    fn names(dir: &Path) -> Vec<String> {
+        let mut v: Vec<String> = fs::read_dir(dir).unwrap().flatten().map(|e| e.file_name().to_string_lossy().to_string()).collect();
+        v.sort();
+        v
+    }
+
+    #[test]
+    fn atomic_write_replaces_and_leaves_no_temp() {
+        let d = temp("replace");
+        let p = d.join("out.jpg");
+        write_atomic(&p, b"one").unwrap();
+        write_atomic(&p, b"two two").unwrap();
+        assert_eq!(fs::read(&p).unwrap(), b"two two");
+        assert_eq!(names(&d), vec!["out.jpg"]);
+        let _ = fs::remove_dir_all(&d);
+    }
+
+    #[test]
+    fn a_failure_mid_write_keeps_the_old_file() {
+        let d = temp("midwrite");
+        let p = d.join("photo.jpg");
+        fs::write(&p, b"the original bytes").unwrap();
+        {
+            let _f = fail_writes_after(4);
+            let e = write_atomic(&p, &[7u8; 10_000]).unwrap_err();
+            assert!(e.to_string().contains("injected"), "{e}");
+            assert!(write_new(&d.join("new.dng"), &[1u8; 100]).is_err());
+        }
+        assert_eq!(fs::read(&p).unwrap(), b"the original bytes", "the target is untouched");
+        assert_eq!(names(&d), vec!["photo.jpg"], "no temp file and no partial new file");
+        // a fill error aborts the same way
+        let r = write_atomic_with(&p, &mut |w| {
+            w.write_all(b"half")?;
+            Err(io::Error::other("encoder failed"))
+        });
+        assert!(r.is_err());
+        assert_eq!(fs::read(&p).unwrap(), b"the original bytes");
+        assert_eq!(names(&d), vec!["photo.jpg"]);
+        let _ = fs::remove_dir_all(&d);
+    }
+
+    #[test]
+    fn recreatable_outputs_are_atomic_without_a_sync() {
+        let d = temp("nosync");
+        let p = d.join("export.jpg");
+        let before = syncs_on_this_thread();
+        write_atomic_nosync(&p, b"one").unwrap();
+        write_atomic_nosync(&p, b"two two").unwrap();
+        assert_eq!(syncs_on_this_thread(), before, "no file or folder sync");
+        assert_eq!(fs::read(&p).unwrap(), b"two two");
+        assert_eq!(names(&d), vec!["export.jpg"], "no temp file left");
+        // still atomic: a failure part-way leaves the previous file and no temp file
+        {
+            let _f = fail_writes_after(3);
+            assert!(write_atomic_nosync(&p, &[9u8; 10_000]).is_err());
+        }
+        assert_eq!(fs::read(&p).unwrap(), b"two two");
+        assert_eq!(names(&d), vec!["export.jpg"]);
+        // the durable writers do sync (the file; the folder too where it can be opened)
+        let before = syncs_on_this_thread();
+        write_atomic(&p, b"three").unwrap();
+        let mid = syncs_on_this_thread();
+        assert!(mid > before);
+        write_new(&d.join("new.dng"), b"dng").unwrap();
+        assert!(syncs_on_this_thread() > mid);
+        let _ = fs::remove_dir_all(&d);
+    }
+
+    #[test]
+    fn new_files_never_replace_anything() {
+        let d = temp("new");
+        let p = d.join("a.dng");
+        fs::write(&p, b"keep").unwrap();
+        assert_eq!(write_new(&p, b"other").unwrap_err().kind(), io::ErrorKind::AlreadyExists);
+        assert_eq!(fs::read(&p).unwrap(), b"keep");
+        let mut c = (1..).map(|k| if k == 1 { d.join("a.dng") } else { d.join(format!("a-{k}.dng")) });
+        let got = write_new_unique(&mut c, b"fresh").unwrap();
+        assert_eq!(got, d.join("a-2.dng"));
+        assert_eq!(fs::read(&got).unwrap(), b"fresh");
+        assert_eq!(names(&d), vec!["a-2.dng", "a.dng"], "the temp file is gone");
+        let _ = fs::remove_dir_all(&d);
+    }
+
+    #[test]
+    fn same_file_sees_through_other_spellings() {
+        let d = temp("same");
+        let p = d.join("x.jpg");
+        fs::write(&p, b"x").unwrap();
+        fs::write(d.join("y.jpg"), b"x").unwrap();
+        fs::create_dir_all(d.join("sub")).unwrap();
+        assert!(same_file(&p, &d.join("sub/../x.jpg")));
+        assert!(!same_file(&p, &d.join("y.jpg")));
+        assert!(!same_file(&p, &d.join("missing.jpg")));
+        let _ = fs::remove_dir_all(&d);
+    }
+}
@@ -277,7 +277,7 @@ mod tests {
    #[test]
    fn group_merge_set_top_remove_and_arrange() {
        let (mut c, ids) = cat(6, &[]);
-        let [a, b, x, d, e, f] = ids[..] else { unreachable!() };
+        let [a, b, x, d, e, f] = ids[..] else { panic!("six ids") };
        let op = c.group_ops(b, &[a, b, x], false).unwrap();
        let inv = c.apply(op).unwrap();
        let s = c.stack_of(a).unwrap().clone();
@@ -16,12 +16,33 @@ pub trait Store: Send {
    /// Replace a file atomically: after a crash either the old or the new content is visible,
    /// never a mix (native: write temp + fsync + rename + fsync directory).
    fn write_atomic(&mut self, name: &str, data: &[u8]) -> io::Result<()>;
-    /// Append and make durable (fsync) before returning.
+    /// [`Store::write_atomic`] with streamed content: `fill` writes the new content into the
+    /// given writer (an error aborts and leaves the old file). Same durability: the file is
+    /// replaced only once all of it is written (native: and fsynced). Returns its length.
+    ///
+    /// The default buffers the content in memory; [`FsStore`] streams it into the temp file, so
+    /// a large snapshot is never held in memory as a whole.
+    fn write_atomic_with(&mut self, name: &str, fill: &mut dyn FnMut(&mut dyn io::Write) -> io::Result<()>) -> io::Result<u64> {
+        let mut buf = Vec::new();
+        fill(&mut buf)?;
+        self.write_atomic(name, &buf)?;
+        Ok(buf.len() as u64)
+    }
+    /// Append and make durable (fsync) before returning. On an error, part of `data` may have
+    /// been written (a full disk, a dropped network share): the journal cuts the file back to its
+    /// previous length with [`Store::truncate`] before appending again.
    fn append(&mut self, name: &str, data: &[u8]) -> io::Result<()>;
-    /// Cut a file to `len` bytes (drop a torn tail before appending again).
+    /// Cut a file to `len` bytes (drop a torn tail before appending again). A missing file is
+    /// left missing.
    fn truncate(&mut self, name: &str, len: u64) -> io::Result<()>;
    /// Human-readable location (diagnostics).
    fn describe(&self) -> String;
+    /// Another handle on the same files that a worker thread can write `catalog.snap` through
+    /// while this one keeps appending to the log (background compaction). `None` (the default):
+    /// snapshots are written synchronously.
+    fn background_writer(&self) -> Option<Box<dyn Store>> {
+        None
+    }
 }

 /// Files in a directory.
@@ -77,6 +98,32 @@ impl Store for FsStore {
        Ok(())
    }

+    fn write_atomic_with(&mut self, name: &str, fill: &mut dyn FnMut(&mut dyn io::Write) -> io::Result<()>) -> io::Result<u64> {
+        use std::io::Write;
+        if self.appender.as_ref().is_some_and(|(n, _)| n == name) {
+            self.appender = None;
+        }
+        let tmp = self.path(&format!("{name}.tmp"));
+        let written = (|| {
+            let mut w = Counter { inner: io::BufWriter::with_capacity(1 << 20, std::fs::File::create(&tmp)?), n: 0 };
+            fill(&mut w)?;
+            w.flush()?;
+            let f = w.inner.into_inner().map_err(|e| e.into_error())?;
+            f.sync_all()?;
+            Ok(w.n)
+        })();
+        let n = match written {
+            Ok(n) => n,
+            Err(e) => {
+                let _ = std::fs::remove_file(&tmp);
+                return Err(e);
+            }
+        };
+        std::fs::rename(&tmp, self.path(name))?;
+        self.sync_dir();
+        Ok(n)
+    }
+
    fn append(&mut self, name: &str, data: &[u8]) -> io::Result<()> {
        use std::io::Write;
        if self.appender.as_ref().is_none_or(|(n, _)| n != name) {
@@ -87,16 +134,27 @@ impl Store for FsStore {
            }
            self.appender = Some((name.to_string(), f));
        }
-        let (_, f) = self.appender.as_mut().expect("appender just opened");
-        f.write_all(data)?;
-        f.sync_data()
+        let Some((_, f)) = self.appender.as_mut() else { return Err(io::Error::other("appender not open")) };
+        let written = f.write_all(data).and_then(|()| f.sync_data());
+        if written.is_err() {
+            // The handle may be dead (a network share that reconnected, a removed volume): the next
+            // attempt reopens the file. Part of `data` may be in the file; the caller cuts it back
+            // ([`Store::truncate`]) before appending again.
+            self.appender = None;
+        }
+        written
    }

    fn truncate(&mut self, name: &str, len: u64) -> io::Result<()> {
        if self.appender.as_ref().is_some_and(|(n, _)| n == name) {
            self.appender = None;
        }
-        let f = std::fs::OpenOptions::new().write(true).open(self.path(name))?;
+        let f = match std::fs::OpenOptions::new().write(true).open(self.path(name)) {
+            Ok(f) => f,
+            // nothing to cut (an append that failed before creating the file)
+            Err(e) if e.kind() == io::ErrorKind::NotFound => return Ok(()),
+            Err(e) => return Err(e),
+        };
        f.set_len(len)?;
        f.sync_all()
    }
@@ -104,12 +162,36 @@ impl Store for FsStore {
    fn describe(&self) -> String {
        self.dir.display().to_string()
    }
+
+    #[cfg(not(target_arch = "wasm32"))]
+    fn background_writer(&self) -> Option<Box<dyn Store>> {
+        Some(Box::new(FsStore { dir: self.dir.clone(), appender: None }))
+    }
+}
+
+/// Counts the bytes written through it.
+struct Counter<W> {
+    inner: W,
+    n: u64,
+}
+
+impl<W: io::Write> io::Write for Counter<W> {
+    fn write(&mut self, buf: &[u8]) -> io::Result<usize> {
+        let k = self.inner.write(buf)?;
+        self.n += k as u64;
+        Ok(k)
+    }
+    fn flush(&mut self) -> io::Result<()> {
+        self.inner.flush()
+    }
 }

 /// In-memory files. Clones share the same files (so a test can "reopen" or corrupt them).
 #[derive(Clone, Default)]
 pub struct MemStore {
    pub files: Arc<Mutex<BTreeMap<String, Vec<u8>>>>,
+    /// Offer a [`Store::background_writer`] (off by default: snapshots stay synchronous).
+    pub background: bool,
 }

 impl MemStore {
@@ -145,4 +227,7 @@ impl Store for MemStore {
    fn describe(&self) -> String {
        "memory".into()
    }
+    fn background_writer(&self) -> Option<Box<dyn Store>> {
+        self.background.then(|| Box::new(self.clone()) as Box<dyn Store>)
+    }
 }
@@ -58,12 +58,22 @@ fn albums_and_folders() {
    let a = photo(&mut c, "a.jpg", "2026-04-01");
    let f = c.alloc_album_id();
    c.apply(Op::AddAlbum {
-        album: Album { id: f, name: "Trips".into(), parent: None, folder: true, photos: vec![], cover: None, smart: None, quick: false },
+        album: Album { id: f, name: "Trips".into(), parent: None, folder: true, photos: vec![], cover: None, smart: None, quick: false, order: None },
    })
    .unwrap();
    let al = c.alloc_album_id();
    c.apply(Op::AddAlbum {
-        album: Album { id: al, name: "Alps".into(), parent: Some(f), folder: false, photos: vec![a], cover: None, smart: None, quick: false },
+        album: Album {
+            id: al,
+            name: "Alps".into(),
+            parent: Some(f),
+            folder: false,
+            photos: vec![a],
+            cover: None,
+            smart: None,
+            quick: false,
+            order: None,
+        },
    })
    .unwrap();
    assert_eq!(c.albums_of(a), vec![al]);
@@ -231,3 +241,263 @@ fn merge_results_are_recognised() {
    assert_eq!(merged_kind("hdr.jpg"), None);
    assert_eq!(merged_kind("panorama.jpg"), None);
 }
+
+/// Preview-only raws (an undecodable raw variant shown from its embedded JPEG): the reason is
+/// stored with the photo, old catalogs without it still load, and Reload's op sets / clears it
+/// with an exact inverse that survives serialisation (the journal).
+#[test]
+fn preview_only_survives_serde_and_set_content_undo() {
+    let mut c = Catalog::new();
+    let a = c.alloc_photo_id();
+    let mut p = Photo::new(a, Source::File { path: "/x/DSC_0001.NEF".into() }, "DSC_0001.NEF", "NEF", 6000, 4000, "2026-10-01T00:00:00");
+    p.kind = MediaKind::Raw;
+    p.preview_only = Some("Nikon Huffman-compressed NEF".into());
+    assert!(!p.develops_raw());
+    c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+    // the catalog round-trips it
+    let back: Catalog = serde_json::from_str(&serde_json::to_string(&c).unwrap()).unwrap();
+    assert_eq!(back.photo(a).unwrap().preview_only.as_deref(), Some("Nikon Huffman-compressed NEF"));
+    // an older catalog (no field) loads as decodable
+    let mut v = serde_json::to_value(&c).unwrap();
+    let photos = v["photos"].as_object_mut().unwrap();
+    let ph = photos.values_mut().next().unwrap();
+    ph.as_object_mut().unwrap().remove("preview_only").expect("serialised when set");
+    let old: Catalog = serde_json::from_value(v).unwrap();
+    assert_eq!(old.photo(a).unwrap().preview_only, None);
+    assert!(old.photo(a).unwrap().develops_raw());
+    // Reload: the file decodes now → cleared; undo (through the journal's JSON) restores it
+    let op = Op::SetContent { id: a, width: 6016, height: 4000, file_size: 1, content_hash: None, preview_only: None };
+    let inv = c.apply(op).unwrap();
+    assert!(c.photo(a).unwrap().develops_raw());
+    let inv: Op = serde_json::from_str(&serde_json::to_string(&inv).unwrap()).unwrap();
+    c.apply(inv).unwrap();
+    assert_eq!(c.photo(a).unwrap().preview_only.as_deref(), Some("Nikon Huffman-compressed NEF"));
+    assert_eq!(c.photo(a).unwrap().width, 6000);
+}
+
+/// Equivalent spellings of one folder (separators, trailing or doubled separators, `.`/`..`,
+/// drive-letter case, verbatim and UNC prefixes) share an identity; different folders don't.
+#[test]
+fn folder_identity_ignores_spelling() {
+    use crate::query::{folder_key, folder_within};
+    let same = |a: &str, b: &str| assert_eq!(folder_key(a), folder_key(b), "{a} vs {b}");
+    let differ = |a: &str, b: &str| assert_ne!(folder_key(a), folder_key(b), "{a} vs {b}");
+    for (a, b) in [
+        ("D:/Example/Photos", "D:\\Example\\Photos"),
+        ("D:/Example/Photos", "d:\\Example\\Photos\\"),
+        ("D:\\Example/Photos", "D:/Example//Photos/."),
+        ("D:/Example/Photos", "D:/Example/Other/../Photos"),
+        ("D:/Example/Photos", "\\\\?\\D:\\Example\\Photos"),
+        ("\\\\server\\share\\Photos", "//server/share/Photos/"),
+        ("\\\\?\\UNC\\server\\share\\Photos", "//server/share/Photos"),
+        ("/home/me/Photos", "/home/me/Photos/"),
+        ("/home/me/Photos", "/home//me/./Photos"),
+        ("/", "//"),
+        ("C:\\", "c:/"),
+    ] {
+        same(a, b);
+    }
+    differ("D:/Example/Photos", "D:/Example/Photos2");
+    differ("D:/Example/Photos", "E:/Example/Photos");
+    differ("/home/me/Photos", "/home/me");
+    assert_eq!(folder_key("/a/../../b"), "/b", "`..` stops at the root");
+    assert!(folder_within("D:\\Example\\Photos\\2026", "D:/Example/Photos/"));
+    assert!(folder_within("D:/Example/Photos", "d:\\Example\\Photos"));
+    assert!(!folder_within("D:/Example/Photos2", "D:/Example/Photos"));
+    assert!(folder_within("/a/b", "/"));
+    assert!(folder_within("C:\\x", "c:\\"));
+}
+
+/// People come from named *face* regions: counted once per photo, case-insensitive, most photos
+/// first; pets and unnamed faces are not people; the `person` filter and `person:` token match.
+#[test]
+fn people_from_named_face_regions() {
+    use lightcraft_meta::{Rect, Region, RegionKind};
+    let region = |name: Option<&str>, kind: RegionKind| Region {
+        rect: Rect { x0: 0.4, y0: 0.4, x1: 0.6, y1: 0.6 },
+        kind,
+        name: name.map(str::to_string),
+        description: None,
+    };
+    let mut c = Catalog::new();
+    let mut add = |name: &str, regions: Vec<Region>| {
+        let id = c.alloc_photo_id();
+        let mut p = Photo::new(id, Source::Demo { scene: 1 }, name, "JPEG", 6000, 4000, "2026-09-30T10:00:00");
+        p.meta.regions = regions;
+        c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+        id
+    };
+    let a = add(
+        "a.jpg",
+        vec![region(Some("Jane Doe"), RegionKind::Face), region(Some("jane doe"), RegionKind::Face), region(Some("Rex"), RegionKind::Pet)],
+    );
+    let b =
+        add("b.jpg", vec![region(Some("JANE DOE"), RegionKind::Face), region(Some("John Roe"), RegionKind::Face), region(None, RegionKind::Face)]);
+    let big = Rect { x0: 0.2, y0: 0.2, x1: 0.8, y1: 0.8 };
+    let d = add("d.jpg", vec![Region { rect: big, ..region(Some("John Roe"), RegionKind::Face) }]);
+    let e = add("e.jpg", vec![region(Some("Sam"), RegionKind::Face)]);
+    add("f.jpg", vec![]);
+    let people = c.people();
+    let summary: Vec<(&str, usize)> = people.iter().map(|p| (p.name.as_str(), p.count)).collect();
+    assert_eq!(summary, vec![("Jane Doe", 2), ("John Roe", 2), ("Sam", 1)], "once per photo, pets and unnamed faces left out");
+    // the picture is the person's largest face; ties go to the lower photo id
+    assert_eq!((people[0].photo, people[0].face), (a, Rect { x0: 0.4, y0: 0.4, x1: 0.6, y1: 0.6 }));
+    assert_eq!((people[1].photo, people[1].face), (d, big), "the larger face wins over an earlier, smaller one");
+
+    let q = |f: Filter| c.query(&f, &Sort::default());
+    assert_eq!(q(Filter { person: Some("jane doe".into()), ..Default::default() }).len(), 2);
+    let mut got = q(Filter { person: Some("John Roe".into()), ..Default::default() });
+    got.sort();
+    assert_eq!(got, vec![b, d]);
+    assert!(q(Filter { person: Some("Rex".into()), ..Default::default() }).is_empty(), "a pet is not a person");
+    assert_eq!(q(Filter { text: "person:SAM".into(), ..Default::default() }), vec![e], "the search token finds a person, any case");
+    assert!(q(Filter { person: Some("Jane Doe".into()), ..Default::default() }).contains(&a));
+
+    // other filters narrow who is offered (so picking a person never ends in an empty grid); the
+    // `person` filter itself does not
+    c.apply(Op::SetRating { id: b, rating: 3 }).unwrap();
+    let names = |f: Filter| c.people_in(&f).into_iter().map(|p| (p.name, p.count)).collect::<Vec<_>>();
+    let rated = Filter { rating: 1, ..Default::default() };
+    assert_eq!(
+        names(rated.clone()),
+        vec![("JANE DOE".to_string(), 1), ("John Roe".to_string(), 1)],
+        "only the rated photo's people, counted within it, spelled as first seen there"
+    );
+    assert_eq!(
+        names(Filter { person: Some("Sam".into()), ..rated }),
+        names(Filter { rating: 1, ..Default::default() }),
+        "the person filter is ignored"
+    );
+    assert_eq!(names(Filter { rating: 5, ..Default::default() }), vec![], "nobody in the filtered photos");
+}
+
+/// Most photos have no regions: the field is left out of the catalog JSON then, and a catalog written
+/// before regions existed (no `regions` key) reads with none.
+#[test]
+fn empty_regions_are_not_serialized_and_default_when_missing() {
+    let m = Meta { keywords: vec!["k".into()], ..Default::default() };
+    let v = serde_json::to_value(&m).unwrap();
+    assert!(v.get("regions").is_none(), "{v}");
+    let back: Meta = serde_json::from_value(v).unwrap();
+    assert!(back.regions.is_empty());
+    let mut with = m.clone();
+    with.regions.push(lightcraft_meta::Region {
+        rect: lightcraft_meta::Rect { x0: 0.1, y0: 0.1, x1: 0.2, y1: 0.2 },
+        kind: lightcraft_meta::RegionKind::Face,
+        name: Some("A".into()),
+        description: None,
+    });
+    let back: Meta = serde_json::from_str(&serde_json::to_string(&with).unwrap()).unwrap();
+    assert_eq!(back, with);
+}
+
+// ---- random sort: Given a seed, the order is a stable, reproducible shuffle of the matching photos
+
+fn many(c: &mut Catalog, n: usize) -> Vec<PhotoId> {
+    (0..n).map(|i| photo(c, &format!("p{i:03}.jpg"), &format!("2026-04-{:02}T10:00:00", i % 28 + 1))).collect()
+}
+
+fn random(seed: u64) -> Sort {
+    Sort { key: SortKey::Random, seed, ..Default::default() }
+}
+
+#[test]
+fn random_sort_is_a_reproducible_permutation() {
+    let mut c = Catalog::new();
+    let ids = many(&mut c, 50);
+    let a = c.query(&Filter::default(), &random(7));
+    assert_eq!(a, c.query(&Filter::default(), &random(7)), "same seed, same order");
+    let mut sorted = a.clone();
+    sorted.sort();
+    let mut all = ids.clone();
+    all.sort();
+    assert_eq!(sorted, all, "every photo exactly once");
+    assert_ne!(a, c.query(&Filter::default(), &random(8)), "another seed, another order");
+    assert_ne!(a, c.query(&Filter::default(), &Sort::default()), "not just the date order");
+}
+
+#[test]
+fn random_sort_keeps_the_relative_order_when_the_set_changes() {
+    let mut c = Catalog::new();
+    let ids = many(&mut c, 30);
+    let before = c.query(&Filter::default(), &random(3));
+    let added = photo(&mut c, "new.jpg", "2026-05-01T10:00:00");
+    c.apply(Op::SetRating { id: ids[0], rating: 5 }).unwrap();
+    let after: Vec<PhotoId> = c.query(&Filter::default(), &random(3)).into_iter().filter(|id| *id != added).collect();
+    assert_eq!(after, before, "adding a photo or editing metadata must not reshuffle the others");
+}
+
+#[test]
+fn random_sort_respects_the_filter_and_handles_tiny_sets() {
+    let mut c = Catalog::new();
+    assert!(c.query(&Filter::default(), &random(1)).is_empty());
+    let ids = many(&mut c, 10);
+    c.apply(Op::SetRating { id: ids[4], rating: 5 }).unwrap();
+    assert_eq!(c.query(&Filter { rating: 5, ..Default::default() }, &random(1)), vec![ids[4]]);
+}
+
+#[test]
+fn random_sort_direction_reverses_and_serde_defaults_hold() {
+    let mut c = Catalog::new();
+    many(&mut c, 20);
+    let desc = c.query(&Filter::default(), &Sort { ascending: false, ..random(5) });
+    let mut asc = c.query(&Filter::default(), &Sort { ascending: true, ..random(5) });
+    asc.reverse();
+    assert_eq!(desc, asc);
+    // saved sorts from before the seed existed still load
+    let old: Sort = serde_json::from_str(r#"{"key":"fileName","ascending":true}"#).unwrap();
+    assert_eq!(old.seed, 0);
+    assert_eq!(serde_json::from_str::<Sort>(r#"{"key":"random","seed":9}"#).unwrap(), Sort { key: SortKey::Random, seed: 9, ..Default::default() });
+}
+
+#[test]
+fn random_sort_has_no_date_headers() {
+    let mut c = Catalog::new();
+    let ids = many(&mut c, 5);
+    assert!(c.date_runs(&ids, SortKey::Random, GroupBy::Day).is_empty());
+}
+
+#[test]
+fn undated_photos_group_under_unknown_date_and_sort_together() {
+    let mut c = Catalog::new();
+    let add_photo = |c: &mut Catalog, name: &str, captured: Option<&str>, imported: &str| {
+        let id = c.alloc_photo_id();
+        let mut p = Photo::new(id, Source::Demo { scene: 0 }, name, "JPEG", 100, 100, imported);
+        p.captured = captured.map(str::to_string);
+        c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+        id
+    };
+    let p_dated1 = add_photo(&mut c, "dated1.jpg", Some("2026-05-10T12:00:00"), "2026-10-01T00:00:00");
+    let p_dated2 = add_photo(&mut c, "dated2.jpg", Some("2026-05-20T12:00:00"), "2026-10-01T00:00:00");
+    let p_undated1 = add_photo(&mut c, "undated1.jpg", None, "2026-10-05T00:00:00");
+    let p_undated2 = add_photo(&mut c, "undated2.jpg", None, "2026-10-02T00:00:00");
+
+    // Descending sort (default, newest first): dated photos first (newest to oldest), undated photos at the end
+    let desc = c.query(&Filter::default(), &Sort { key: SortKey::CaptureDate, ascending: false, ..Default::default() });
+    assert_eq!(desc, vec![p_dated2, p_dated1, p_undated1, p_undated2]);
+
+    // Ascending sort (oldest first): undated photos at the beginning, dated photos from oldest to newest
+    let asc = c.query(&Filter::default(), &Sort { key: SortKey::CaptureDate, ascending: true, ..Default::default() });
+    assert_eq!(asc, vec![p_undated2, p_undated1, p_dated1, p_dated2]);
+
+    // date_runs groups undated photos under Unknown Date
+    let runs = c.date_runs(&desc, SortKey::CaptureDate, GroupBy::Day);
+    assert_eq!(runs.len(), 3);
+    assert_eq!(runs[0].key, "2026-05-20");
+    assert_eq!(runs[1].key, "2026-05-10");
+    assert_eq!(runs[2].key, "");
+    assert_eq!(runs[2].label, "Unknown Date");
+    assert_eq!(runs[2].count, 2);
+
+    // date_groups sidebar tree only counts photos with capture dates
+    let groups = c.date_groups();
+    assert_eq!(groups.len(), 1);
+    assert_eq!(groups[0].year, "2026");
+    assert_eq!(groups[0].count, 2);
+
+    // Filter by capture date prefix does not match undated photos
+    let filter_day = Filter { date: Some("2026-10-05".into()), ..Default::default() };
+    assert_eq!(c.query(&filter_day, &Sort::default()), vec![]);
+    let filter_year = Filter { date: Some("2026".into()), ..Default::default() };
+    assert_eq!(c.query(&filter_year, &Sort { key: SortKey::CaptureDate, ascending: false, ..Default::default() }), vec![p_dated2, p_dated1]);
+}
@@ -0,0 +1,122 @@
+//! Album order: the children of a folder (or of the top level) are listed folders first, then
+//! alphabetically, until some of them are given a place of their own (`Album.order`); moving an
+//! album to another folder drops its place, and undoing the move brings it back.
+
+use crate::*;
+
+fn add(c: &mut Catalog, name: &str, parent: Option<AlbumId>, folder: bool) -> AlbumId {
+    let id = c.alloc_album_id();
+    c.apply(Op::AddAlbum { album: Album { parent, folder, ..Album::new(id, name) } }).unwrap();
+    id
+}
+
+fn names(c: &Catalog, parent: Option<AlbumId>) -> Vec<String> {
+    c.album_children(parent).iter().map(|a| a.name.clone()).collect()
+}
+
+fn place(c: &mut Catalog, id: AlbumId, order: Option<u32>) {
+    c.apply(Op::SetAlbumOrder { id, order }).unwrap();
+}
+
+/// Given albums that were never ordered, when they are listed, then folders come first and each
+/// group is alphabetical regardless of case.
+#[test]
+fn children_are_folders_first_then_alphabetical() {
+    let mut c = Catalog::new();
+    add(&mut c, "beta", None, false);
+    add(&mut c, "Alpha", None, false);
+    add(&mut c, "zoo", None, true);
+    add(&mut c, "Archive", None, true);
+    assert_eq!(names(&c, None), ["Archive", "zoo", "Alpha", "beta"]);
+    assert!(!c.album_children_are_ordered(None));
+}
+
+/// Given some children with a place of their own, then they follow it; the others come after, by
+/// name; folders stay ahead of albums.
+#[test]
+fn ordered_children_follow_their_order() {
+    let mut c = Catalog::new();
+    let a = add(&mut c, "A", None, false);
+    let b = add(&mut c, "B", None, false);
+    let cc = add(&mut c, "C", None, false);
+    add(&mut c, "D", None, false);
+    let f = add(&mut c, "F", None, true);
+    place(&mut c, cc, Some(0));
+    place(&mut c, a, Some(1));
+    assert_eq!(names(&c, None), ["F", "C", "A", "B", "D"]);
+    assert!(c.album_children_are_ordered(None));
+    // equal places fall back to the name, so a damaged catalog still lists deterministically
+    place(&mut c, b, Some(1));
+    assert_eq!(names(&c, None), ["F", "C", "A", "B", "D"]);
+    // another container is unaffected
+    let inner = add(&mut c, "Z", Some(f), false);
+    add(&mut c, "Y", Some(f), false);
+    assert_eq!(names(&c, Some(f)), ["Y", "Z"]);
+    assert!(!c.album_children_are_ordered(Some(f)));
+    place(&mut c, inner, Some(0));
+    assert_eq!(names(&c, Some(f)), ["Z", "Y"]);
+}
+
+/// The one-pass map lists every folder's children exactly as `album_children` does.
+#[test]
+fn the_children_map_matches_the_children_of_each_folder() {
+    let mut c = Catalog::new();
+    let f = add(&mut c, "F", None, true);
+    let g = add(&mut c, "G", Some(f), true);
+    for (name, parent) in [("b", None), ("a", None), ("y", Some(f)), ("x", Some(f)), ("q", Some(g))] {
+        add(&mut c, name, parent, false);
+    }
+    let x = c.albums().find(|a| a.name == "x").unwrap().id;
+    place(&mut c, x, Some(7));
+    let map = c.album_children_by_parent();
+    for parent in [None, Some(f), Some(g)] {
+        let from_map: Vec<AlbumId> = map.get(&parent).map(|v| v.iter().map(|a| a.id).collect()).unwrap_or_default();
+        let direct: Vec<AlbumId> = c.album_children(parent).iter().map(|a| a.id).collect();
+        assert_eq!(from_map, direct, "{parent:?}");
+    }
+    assert!(!map.contains_key(&Some(AlbumId(999))));
+}
+
+/// A place is one undo step, and `None` puts the album back among the alphabetical ones.
+#[test]
+fn a_place_can_be_set_cleared_and_undone() {
+    let mut c = Catalog::new();
+    let a = add(&mut c, "A", None, false);
+    add(&mut c, "B", None, false);
+    let inv = c.apply(Op::SetAlbumOrder { id: a, order: Some(5) }).unwrap();
+    assert_eq!(c.album(a).unwrap().order, Some(5));
+    c.apply(inv).unwrap();
+    assert_eq!(c.album(a).unwrap().order, None);
+    assert!(c.apply(Op::SetAlbumOrder { id: AlbumId(999), order: Some(1) }).is_err(), "no such album");
+}
+
+/// Moving to another folder drops the album's place there (it belonged to the old neighbours),
+/// and undoing the move gives it back.
+#[test]
+fn moving_drops_the_place_and_undo_restores_it() {
+    let mut c = Catalog::new();
+    let from = add(&mut c, "From", None, true);
+    let to = add(&mut c, "To", None, true);
+    let a = add(&mut c, "A", Some(from), false);
+    place(&mut c, a, Some(3));
+    let inv = c.apply(Op::MoveAlbum { id: a, parent: Some(to) }).unwrap();
+    assert_eq!((c.album(a).unwrap().parent, c.album(a).unwrap().order), (Some(to), None));
+    c.apply(inv).unwrap();
+    assert_eq!((c.album(a).unwrap().parent, c.album(a).unwrap().order), (Some(from), Some(3)));
+    // "moving" to where it already is changes nothing
+    c.apply(Op::MoveAlbum { id: a, parent: Some(from) }).unwrap();
+    assert_eq!(c.album(a).unwrap().order, Some(3));
+}
+
+/// Catalogs written before albums had an order load as they were; an unset order is not written.
+#[test]
+fn albums_without_an_order_load_and_save_unchanged() {
+    let old = r#"{"id":4,"name":"Old","parent":null,"folder":false,"photos":[],"cover":null}"#;
+    let a: Album = serde_json::from_str(old).unwrap();
+    assert_eq!(a.order, None);
+    assert!(!serde_json::to_string(&a).unwrap().contains("order"));
+    let mut a = a;
+    a.order = Some(2);
+    let back: Album = serde_json::from_str(&serde_json::to_string(&a).unwrap()).unwrap();
+    assert_eq!(back.order, Some(2));
+}
@@ -0,0 +1,339 @@
+//! Background compaction (issue #37): a snapshot written by a worker while appends go on.
+//! "Crashes" are simulated by copying the files at a given moment and loading the copy: every
+//! acknowledged (appended) op must be there, in order, whatever the snapshot was doing.
+
+use std::sync::atomic::{AtomicBool, Ordering};
+use std::sync::{Arc, Condvar, Mutex};
+use std::time::Duration;
+
+use proptest::prelude::*;
+
+use crate::journal::{LOG, SNAPSHOT};
+use crate::*;
+
+/// A MemStore whose snapshot writes wait for a gate (to hold a background snapshot mid-flight)
+/// and can be made to fail.
+#[derive(Clone)]
+struct Gated {
+    files: MemStore,
+    gate: Arc<(Mutex<bool>, Condvar)>,
+    fail: Arc<AtomicBool>,
+}
+
+impl Gated {
+    fn new() -> Gated {
+        Gated { files: MemStore::new(), gate: Arc::new((Mutex::new(false), Condvar::new())), fail: Arc::new(AtomicBool::new(false)) }
+    }
+    fn set_gate(&self, open: bool) {
+        *self.gate.0.lock().unwrap() = open;
+        self.gate.1.notify_all();
+    }
+    /// The files as a crash at this moment would leave them.
+    fn crash(&self) -> MemStore {
+        let m = MemStore::new();
+        *m.files.lock().unwrap() = self.files.files.lock().unwrap().clone();
+        m
+    }
+    fn snapshot_seq_on_disk(&self) -> Option<u64> {
+        let v: serde_json::Value = serde_json::from_slice(&self.files.get(SNAPSHOT)?).ok()?;
+        v["seq"].as_u64()
+    }
+}
+
+impl Store for Gated {
+    fn read(&mut self, name: &str) -> std::io::Result<Option<Vec<u8>>> {
+        self.files.read(name)
+    }
+    fn write_atomic(&mut self, name: &str, data: &[u8]) -> std::io::Result<()> {
+        self.files.write_atomic(name, data)
+    }
+    fn write_atomic_with(&mut self, name: &str, fill: &mut dyn FnMut(&mut dyn std::io::Write) -> std::io::Result<()>) -> std::io::Result<u64> {
+        if name == SNAPSHOT {
+            let (lock, cv) = &*self.gate;
+            let (open, timeout) = cv.wait_timeout_while(lock.lock().unwrap(), Duration::from_secs(20), |open| !*open).unwrap();
+            if !*open || timeout.timed_out() {
+                return Err(std::io::Error::other("gate never opened"));
+            }
+        }
+        let mut buf = Vec::new();
+        fill(&mut buf)?;
+        if self.fail.load(Ordering::SeqCst) {
+            return Err(std::io::Error::other("disk full"));
+        }
+        self.files.write_atomic(name, &buf)?;
+        Ok(buf.len() as u64)
+    }
+    fn append(&mut self, name: &str, data: &[u8]) -> std::io::Result<()> {
+        self.files.append(name, data)
+    }
+    fn truncate(&mut self, name: &str, len: u64) -> std::io::Result<()> {
+        self.files.truncate(name, len)
+    }
+    fn describe(&self) -> String {
+        "gated".into()
+    }
+    fn background_writer(&self) -> Option<Box<dyn Store>> {
+        Some(Box::new(self.clone()))
+    }
+}
+
+/// Apply and durably append one edit (two ops in one append every fourth time).
+fn edit(j: &mut Journal, live: &mut Catalog, k: u64) {
+    let one = |live: &mut Catalog, k: u64| {
+        let op = if live.len() < 4 || k.is_multiple_of(3) {
+            let id = live.alloc_photo_id();
+            Op::AddPhoto { photo: Box::new(Photo::new(id, Source::File { path: format!("/p/{}.jpg", id.0) }, "p.jpg", "JPEG", 4, 3, "t")) }
+        } else {
+            Op::SetRating { id: PhotoId(1 + k % live.len() as u64), rating: (k % 6) as u8 }
+        };
+        live.apply(op.clone()).unwrap();
+        op
+    };
+    let mut ops = vec![one(live, k)];
+    if k.is_multiple_of(4) {
+        ops.push(one(live, k / 4 + 1));
+    }
+    j.append(&ops).unwrap();
+}
+
+/// Load the files a crash now would leave: they must hold exactly the live state.
+fn assert_recovers(g: &Gated, live: &Catalog) -> LoadReport {
+    let (_, c, r) = Journal::open(Box::new(g.crash())).unwrap();
+    assert_eq!(c.to_snapshot(), live.to_snapshot(), "{r:?}");
+    assert_eq!((r.failed, r.torn_bytes, r.damaged.clone()), (0, 0, None));
+    r
+}
+
+fn wait_written(j: &Journal) {
+    for _ in 0..20_000 {
+        if j.snapshot_written() {
+            return;
+        }
+        std::thread::sleep(Duration::from_millis(1));
+    }
+    panic!("background snapshot never finished");
+}
+
+#[test]
+fn background_snapshot_keeps_edits_made_while_it_runs() {
+    let g = Gated::new();
+    let (mut j, mut live, _) = Journal::open(Box::new(g.clone())).unwrap();
+    for k in 0..30 {
+        edit(&mut j, &mut live, k);
+    }
+    let n = j.seq();
+    j.snapshot_in_background(&live).unwrap();
+    assert!(j.snapshot_running() && !j.wants_snapshot() && j.stats().snapshot_running);
+
+    // edits while the snapshot is being written are appended (durably) as usual
+    let mut during = 0;
+    for k in 30..50 {
+        edit(&mut j, &mut live, k);
+        during = j.seq() - n;
+    }
+    // crash before the snapshot landed: no snapshot yet, the whole log replays
+    let r = assert_recovers(&g, &live);
+    assert_eq!((r.snapshot_seq, r.replayed as u64), (0, j.seq()));
+
+    // the snapshot lands (op n); crash before the log was trimmed: its records are stale
+    g.set_gate(true);
+    wait_written(&j);
+    assert_eq!(g.snapshot_seq_on_disk(), Some(n));
+    for k in 50..55 {
+        edit(&mut j, &mut live, k);
+        during = j.seq() - n;
+    }
+    let r = assert_recovers(&g, &live);
+    assert_eq!((r.snapshot_seq, r.stale as u64, r.replayed as u64), (n, n, during));
+
+    // finishing trims the log to exactly the records after the snapshot, in order
+    assert!(j.poll().unwrap());
+    assert!(!j.snapshot_running());
+    assert_eq!((j.snapshot_seq(), j.log_records()), (n, during));
+    let log = String::from_utf8(g.files.get(LOG).unwrap()).unwrap();
+    let seqs: Vec<u64> = log.lines().map(|l| journal::decode_record(l).unwrap().0).collect();
+    assert_eq!(seqs, ((n + 1)..=j.seq()).collect::<Vec<_>>());
+    assert_eq!(j.log_bytes(), log.len() as u64);
+    let r = assert_recovers(&g, &live);
+    assert_eq!((r.snapshot_seq, r.stale, r.replayed as u64), (n, 0, during));
+    let s = j.stats();
+    assert!(s.last_snapshot.background && s.last_snapshot.records == n && !s.snapshot_running, "{s:?}");
+
+    // and appending goes on after it
+    edit(&mut j, &mut live, 99);
+    assert_recovers(&g, &live);
+}
+
+#[test]
+fn failed_background_snapshot_loses_nothing_and_retries_later() {
+    let g = Gated::new();
+    let (mut j, mut live, _) = Journal::open(Box::new(g.clone())).unwrap();
+    j.policy = SnapshotPolicy { max_records: 10, max_bytes: u64::MAX };
+    for k in 0..12 {
+        edit(&mut j, &mut live, k);
+    }
+    assert!(j.wants_snapshot());
+    g.fail.store(true, Ordering::SeqCst);
+    g.set_gate(true);
+    j.snapshot_in_background(&live).unwrap();
+    edit(&mut j, &mut live, 12);
+    wait_written(&j);
+    assert!(j.poll().is_err());
+    assert_eq!((j.snapshot_seq(), j.stats().failed_snapshots), (0, 1));
+    // still the (empty) snapshot written when the library was created
+    assert_eq!(g.snapshot_seq_on_disk(), Some(0));
+    assert_recovers(&g, &live);
+    // no retry storm: the next attempt waits for another `max_records` records
+    assert!(!j.wants_snapshot());
+    g.fail.store(false, Ordering::SeqCst);
+    while !j.wants_snapshot() {
+        let k = j.seq();
+        edit(&mut j, &mut live, k);
+    }
+    j.snapshot_in_background(&live).unwrap();
+    j.wait().unwrap();
+    assert_eq!((j.snapshot_seq(), j.log_records()), (j.seq(), 0));
+    let r = assert_recovers(&g, &live);
+    assert_eq!((r.snapshot_seq, r.replayed), (j.seq(), 0));
+}
+
+/// A synchronous snapshot (close, "Optimize Library") waits for the background one, so the older
+/// snapshot can't land after (and replace) the newer one.
+#[test]
+fn sync_snapshot_waits_for_the_background_one() {
+    let g = Gated::new();
+    let (mut j, mut live, _) = Journal::open(Box::new(g.clone())).unwrap();
+    for k in 0..10 {
+        edit(&mut j, &mut live, k);
+    }
+    j.snapshot_in_background(&live).unwrap();
+    for k in 10..20 {
+        edit(&mut j, &mut live, k);
+    }
+    let opener = {
+        let g = g.clone();
+        std::thread::spawn(move || {
+            std::thread::sleep(Duration::from_millis(50));
+            g.set_gate(true);
+        })
+    };
+    j.snapshot(&live).unwrap();
+    opener.join().unwrap();
+    assert!(!j.snapshot_running());
+    assert_eq!(g.snapshot_seq_on_disk(), Some(j.seq()));
+    assert_eq!((j.snapshot_seq(), j.log_records()), (j.seq(), 0));
+    assert!(g.files.get(LOG).unwrap().is_empty());
+    assert_eq!(j.stats().snapshots, 2);
+    assert_recovers(&g, &live);
+}
+
+/// Dropping the journal (e.g. reopening the library) waits for the snapshot in flight and leaves
+/// the log whole: reopening finds every op.
+#[test]
+fn dropping_the_journal_waits_for_its_snapshot() {
+    let g = Gated::new();
+    let (mut j, mut live, _) = Journal::open(Box::new(g.clone())).unwrap();
+    for k in 0..10 {
+        edit(&mut j, &mut live, k);
+    }
+    let n = j.seq();
+    j.snapshot_in_background(&live).unwrap();
+    for k in 10..15 {
+        edit(&mut j, &mut live, k);
+    }
+    let opener = {
+        let g = g.clone();
+        std::thread::spawn(move || {
+            std::thread::sleep(Duration::from_millis(50));
+            g.set_gate(true);
+        })
+    };
+    let total = j.seq();
+    drop(j);
+    opener.join().unwrap();
+    assert_eq!(g.snapshot_seq_on_disk(), Some(n));
+    let (_, c, r) = Journal::open(Box::new(g.files.clone())).unwrap();
+    assert_eq!(c.to_snapshot(), live.to_snapshot());
+    assert_eq!((r.snapshot_seq, r.stale as u64, r.replayed as u64), (n, n, total - n));
+}
+
+/// On disk, with a real worker thread: a crash mid-write leaves only a temp file, which loading
+/// ignores; edits during the snapshot survive; reopening resumes appends.
+#[test]
+fn fs_store_background_snapshot() {
+    let dir = std::env::temp_dir().join(format!("lc-journal-bg-{}", std::process::id()));
+    let _ = std::fs::remove_dir_all(&dir);
+    let (mut j, mut live, _) = Journal::open(Box::new(FsStore::open(&dir).unwrap())).unwrap();
+    for k in 0..200 {
+        edit(&mut j, &mut live, k);
+    }
+    // a previous crash mid-snapshot left a partial temp file
+    std::fs::write(dir.join(format!("{SNAPSHOT}.tmp")), b"{\"format\":\"lightcraft-catal").unwrap();
+    let (mut j2, c2, r) = Journal::open(Box::new(FsStore::open(&dir).unwrap())).unwrap();
+    assert_eq!((c2.to_snapshot(), r.snapshot_seq), (live.to_snapshot(), 0));
+    drop(j);
+    let n = j2.seq();
+    j2.snapshot_in_background(&live).unwrap();
+    for k in 200..260 {
+        edit(&mut j2, &mut live, k);
+    }
+    j2.wait().unwrap();
+    assert_eq!((j2.snapshot_seq(), j2.log_records()), (n, j2.seq() - n));
+    assert!(!dir.join(format!("{SNAPSHOT}.tmp")).exists());
+    edit(&mut j2, &mut live, 999);
+    drop(j2);
+    let (_, c3, r) = Journal::open(Box::new(FsStore::open(&dir).unwrap())).unwrap();
+    assert_eq!(c3.to_snapshot(), live.to_snapshot());
+    assert_eq!((r.snapshot_seq, r.stale), (n, 0));
+    let _ = std::fs::remove_dir_all(&dir);
+}
+
+proptest! {
+    #![proptest_config(ProptestConfig::with_cases(48))]
+
+    /// Any interleaving of edits, background snapshots (started, landed, finished), synchronous
+    /// snapshots and crashes: a crash at any point recovers exactly the acknowledged ops.
+    #[test]
+    fn background_compaction_never_loses_acknowledged_ops(actions in proptest::collection::vec((0u8..8, 0u64..1000), 1..60)) {
+        let g = Gated::new();
+        let (mut j, mut live, _) = Journal::open(Box::new(g.clone())).unwrap();
+        for (ctl, k) in actions {
+            match ctl {
+                0..=2 => edit(&mut j, &mut live, k),
+                3 => j.snapshot_in_background(&live).unwrap(),
+                4 | 5 => {
+                    // the worker writes its snapshot; finish it now or later
+                    if j.snapshot_running() {
+                        g.set_gate(true);
+                        wait_written(&j);
+                        g.set_gate(false);
+                        if ctl == 4 {
+                            prop_assert!(j.poll().unwrap());
+                        }
+                    }
+                }
+                6 => {
+                    g.set_gate(true);
+                    j.snapshot(&live).unwrap();
+                    g.set_gate(false);
+                    prop_assert_eq!(g.snapshot_seq_on_disk(), Some(j.seq()));
+                }
+                _ => {
+                    // a crash now
+                    let (_, c, r) = Journal::open(Box::new(g.crash())).unwrap();
+                    prop_assert_eq!(c.to_snapshot(), live.to_snapshot());
+                    prop_assert_eq!(r.failed, 0);
+                    prop_assert!(r.damaged.is_none());
+                }
+            }
+            // the snapshot on disk never covers more than what was appended
+            prop_assert!(g.snapshot_seq_on_disk().unwrap_or(0) <= j.seq());
+        }
+        g.set_gate(true);
+        let (_, c, _) = Journal::open(Box::new(g.crash())).unwrap();
+        prop_assert_eq!(c.to_snapshot(), live.to_snapshot());
+        drop(j);
+        let (_, c, _) = Journal::open(Box::new(g.files.clone())).unwrap();
+        prop_assert_eq!(c.to_snapshot(), live.to_snapshot());
+    }
+}
@@ -0,0 +1,404 @@
+//! The library's folder tree: where the imported photos live on disk (see [`crate::folders`]).
+//!
+//! Scenarios, in the words of someone browsing their library:
+//!
+//! * Given photos imported from several folders, the tree lists those folders and no others,
+//!   each with how many photos it holds, under the disk (volume) they are on.
+//! * A folder's count includes the folders inside it.
+//! * Every folder is a row of its own, the ones above the imported photos included.
+//! * Photos only browsed in Local, deleted photos and demo scenes are not "imported from" anywhere.
+//! * However a folder is spelled, it is one folder.
+//! * Choosing a folder shows what was imported from it and from the folders inside it — exactly
+//!   as many photos as the row says.
+
+use crate::*;
+
+fn add(c: &mut Catalog, path: &str) -> PhotoId {
+    add_with(c, path, |_| {})
+}
+
+fn add_with(c: &mut Catalog, path: &str, tweak: impl FnOnce(&mut Photo)) -> PhotoId {
+    let id = c.alloc_photo_id();
+    let name = path.rsplit(['/', '\\']).next().unwrap_or(path);
+    let mut p = Photo::new(id, Source::File { path: path.into() }, name, "JPEG", 60, 40, "2026-01-01T10:00:00");
+    tweak(&mut p);
+    c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+    id
+}
+
+fn library(paths: &[&str]) -> Catalog {
+    let mut c = Catalog::new();
+    paths.iter().for_each(|p| {
+        add(&mut c, p);
+    });
+    c
+}
+
+/// `(name, count, own)` of each node, depth first.
+fn outline(nodes: &[FolderNode]) -> Vec<(String, usize, usize)> {
+    fn walk(n: &FolderNode, out: &mut Vec<(String, usize, usize)>) {
+        out.push((n.name.clone(), n.count, n.own));
+        n.children.iter().for_each(|c| walk(c, out));
+    }
+    let mut out = Vec::new();
+    nodes.iter().for_each(|n| walk(n, &mut out));
+    out
+}
+
+fn row(name: &str, count: usize, own: usize) -> (String, usize, usize) {
+    (name.to_string(), count, own)
+}
+
+fn rows(paths: &[&str]) -> Vec<(String, usize, usize)> {
+    outline(&library(paths).folder_tree())
+}
+
+#[test]
+fn an_empty_library_has_no_folders() {
+    assert!(Catalog::new().folder_tree().is_empty());
+}
+
+#[test]
+fn only_folders_with_imported_photos_are_listed_with_their_counts() {
+    let c = library(&["/pics/trip/a.jpg", "/pics/trip/b.jpg", "/pics/home/c.jpg"]);
+    let tree = c.folder_tree();
+    assert_eq!(outline(&tree), vec![row("/", 3, 0), row("pics", 3, 0), row("home", 1, 1), row("trip", 2, 2)]);
+    assert!(tree[0].volume && !tree[0].children[0].volume);
+    assert_eq!(tree[0].children[0].children[1].path, "/pics/trip", "a row knows its folder, for choosing it");
+}
+
+#[test]
+fn a_folders_count_includes_the_folders_inside_it() {
+    let r = rows(&["/pics/trip/a.jpg", "/pics/trip/day1/b.jpg", "/pics/trip/day1/c.jpg", "/pics/trip/day2/d.jpg"]);
+    assert_eq!(r, vec![row("/", 4, 0), row("pics", 4, 0), row("trip", 4, 1), row("day1", 2, 2), row("day2", 1, 1)]);
+}
+
+#[test]
+fn every_folder_is_a_row_of_its_own_even_when_it_holds_nothing_itself() {
+    let c = library(&["/Users/me/Pictures/2024/a.jpg", "/Users/me/Pictures/2025/b.jpg", "/Volumes/nas/photos/c.jpg"]);
+    let tree = c.folder_tree();
+    assert_eq!(
+        outline(&tree),
+        vec![
+            row("/", 2, 0),
+            row("Users", 2, 0),
+            row("me", 2, 0),
+            row("Pictures", 2, 0),
+            row("2024", 1, 1),
+            row("2025", 1, 1),
+            row("nas", 1, 0),
+            row("photos", 1, 1)
+        ]
+    );
+    assert_eq!(tree[0].children[0].children[0].children[0].path, "/Users/me/Pictures", "a row is exactly the folder it names");
+}
+
+#[test]
+fn a_folder_with_photos_of_its_own_is_not_folded_into_its_child() {
+    assert_eq!(rows(&["/pics/a.jpg", "/pics/trip/b.jpg"]), vec![row("/", 2, 0), row("pics", 2, 1), row("trip", 1, 1)]);
+}
+
+#[test]
+fn browsed_deleted_and_demo_photos_are_not_imported_from_anywhere() {
+    let mut c = Catalog::new();
+    add(&mut c, "/pics/kept/a.jpg");
+    add_with(&mut c, "/pics/browsed/b.jpg", |p| p.local = true);
+    add_with(&mut c, "/pics/gone/c.jpg", |p| p.deleted = true);
+    let id = c.alloc_photo_id();
+    let demo = Photo::new(id, Source::Demo { scene: 1 }, "demo", "JPEG", 60, 40, "2026-01-01T10:00:00");
+    c.apply(Op::AddPhoto { photo: Box::new(demo) }).unwrap();
+    assert_eq!(outline(&c.folder_tree()), vec![row("/", 1, 0), row("pics", 1, 0), row("kept", 1, 1)]);
+}
+
+#[test]
+fn however_a_folder_is_spelled_it_is_one_folder() {
+    let r = rows(&["/pics/trip/a.jpg", "/pics//trip/./b.jpg", "/pics\\trip\\c.jpg", "/pics/x/../trip/d.jpg"]);
+    assert_eq!(r, vec![row("/", 4, 0), row("pics", 4, 0), row("trip", 4, 4)]);
+}
+
+#[test]
+fn each_disk_is_a_volume_of_its_own() {
+    // macOS, Linux and Windows habits for where disks appear
+    let r = rows(&["/Volumes/nas/p/1.jpg", "/media/me/usb/2.jpg", "/run/media/me/card/3.jpg", "/mnt/disk/4.jpg"]);
+    assert_eq!(r, vec![row("card", 1, 1), row("disk", 1, 1), row("nas", 1, 0), row("p", 1, 1), row("usb", 1, 1)]);
+    let r = rows(&["D:\\Photos\\2024\\a.jpg", "D:\\Photos\\2025\\b.jpg", "C:\\x\\c.jpg"]);
+    assert_eq!(r, vec![row("C:", 1, 0), row("x", 1, 1), row("D:", 2, 0), row("Photos", 2, 0), row("2024", 1, 1), row("2025", 1, 1)]);
+}
+
+#[test]
+fn a_network_share_is_a_volume_and_two_spellings_are_one() {
+    let r = rows(&["\\\\srv\\s1\\a\\1.jpg", "\\\\srv\\s2\\b\\2.jpg"]);
+    assert_eq!(r, vec![row("//srv/s1", 1, 0), row("a", 1, 1), row("//srv/s2", 1, 0), row("b", 1, 1)]);
+    // the verbatim form Windows APIs give is the same share
+    let r = rows(&["\\\\?\\UNC\\srv\\s1\\a\\1.jpg", "\\\\srv\\s1\\a\\2.jpg"]);
+    assert_eq!(r, vec![row("//srv/s1", 2, 0), row("a", 2, 2)]);
+}
+
+#[test]
+fn a_samba_share_mounted_on_a_mac_is_a_volume_named_after_the_share() {
+    // Finder mounts every share under /Volumes, and a second mount of the same share gets a suffix
+    let r = rows(&["/Volumes/tokyo/Photos/2024/a.jpg", "/Volumes/tokyo/Photos/2025/b.jpg", "/Volumes/tokyo-1/old/c.jpg", "/Volumes/東京 NAS/d.jpg"]);
+    assert_eq!(
+        r,
+        vec![
+            row("tokyo", 2, 0),
+            row("Photos", 2, 0),
+            row("2024", 1, 1),
+            row("2025", 1, 1),
+            row("tokyo-1", 1, 0),
+            row("old", 1, 1),
+            row("東京 NAS", 1, 1)
+        ]
+    );
+    let tree = library(&["/Volumes/tokyo/Photos/a.jpg"]).folder_tree();
+    assert_eq!((tree[0].path.as_str(), tree[0].volume), ("/Volumes/tokyo", true), "choosing the share shows everything on it");
+}
+
+#[test]
+fn shares_the_other_systems_mount_are_volumes_too() {
+    // Linux desktops: gvfs mounts per user, with the server and share in the mount's name
+    let gvfs = "/run/user/1000/gvfs/smb-share:server=tokyo,share=photos";
+    let r = rows(&[&format!("{gvfs}/2024/a.jpg"), "/run/user/1000/gvfs/sftp:host=nas/b.jpg"]);
+    assert_eq!(r, vec![row("photos on tokyo", 1, 0), row("2024", 1, 1), row("sftp:host=nas", 1, 1)]);
+    // NFS automounts: Linux /net/<host>, macOS /Network/Servers/<host>
+    let r = rows(&["/net/tokyo/export/a.jpg", "/Network/Servers/osaka/vol/b.jpg"]);
+    assert_eq!(r, vec![row("osaka", 1, 0), row("vol", 1, 1), row("tokyo", 1, 0), row("export", 1, 1)]);
+    // Windows: a share by name or by address, the WSL file system, a mapped drive
+    let r = rows(&[r"\\tokyo.local\photos\a.jpg", r"\\192.168.1.5\scans\b.jpg", r"\\wsl$\Ubuntu\home\me\c.jpg", r"Z:\d.jpg"]);
+    assert_eq!(
+        r,
+        vec![
+            row("//192.168.1.5/scans", 1, 1),
+            row("//tokyo.local/photos", 1, 1),
+            row("//wsl$/Ubuntu", 1, 0),
+            row("home", 1, 0),
+            row("me", 1, 1),
+            row("Z:", 1, 1)
+        ]
+    );
+}
+
+#[test]
+fn a_share_that_is_not_mounted_costs_nothing() {
+    // the tree is made from the catalog's own records: it never looks at a disk, so a share that
+    // is asleep or gone lists (and counts) exactly like one that is there
+    let c = library(&["/Volumes/definitely-not-mounted/a/b.jpg", r"\\unreachable.invalid\share\c.jpg"]);
+    let tree = c.folder_tree();
+    assert_eq!(tree.iter().map(|n| n.count).sum::<usize>(), 2);
+}
+
+#[test]
+fn a_photo_in_the_root_of_a_disk_counts_for_the_disk() {
+    assert_eq!(rows(&["/x.jpg", "/d/y.jpg"]), vec![row("/", 2, 1), row("d", 1, 1)]);
+    assert_eq!(rows(&["C:\\x.jpg"]), vec![row("C:", 1, 1)]);
+}
+
+#[test]
+fn photos_with_no_folder_are_left_out() {
+    assert!(rows(&["x.jpg", ""]).is_empty());
+}
+
+#[test]
+fn folders_are_listed_by_name_ignoring_case() {
+    let names: Vec<String> = rows(&["/p/b/1.jpg", "/p/A/2.jpg", "/p/c/3.jpg"]).into_iter().skip(2).map(|r| r.0).collect();
+    assert_eq!(names, ["A", "b", "c"]);
+}
+
+#[test]
+fn absurdly_deep_paths_neither_crash_nor_lose_photos() {
+    let deep = format!("/{}/a.jpg", vec!["d"; 5000].join("/"));
+    let tree = library(&[&deep]).folder_tree();
+    assert_eq!(tree.iter().map(|n| n.count).sum::<usize>(), 1);
+    assert!(outline(&tree).len() <= crate::folders::MAX_DEPTH + 2);
+}
+
+#[test]
+fn choosing_a_folder_shows_what_was_imported_from_it_and_below() {
+    let mut c = Catalog::new();
+    let trip = add(&mut c, "/pics/trip/a.jpg");
+    let day = add(&mut c, "/pics/trip/day1/b.jpg");
+    let sibling = add(&mut c, "/pics/trip2/c.jpg"); // shares the name's start, not the folder
+    let browsed = add_with(&mut c, "/pics/trip/d.jpg", |p| p.local = true);
+    let f = Filter { library_folder: Some("/pics//trip/".into()), ..Default::default() };
+    let shown = |id: PhotoId| f.matches(c.photo(id).unwrap(), &c);
+    assert!(shown(trip) && shown(day));
+    assert!(!shown(sibling), "/pics/trip2 is another folder");
+    assert!(!shown(browsed), "browsed photos are not part of the library's folders");
+    assert_eq!(f.describe(), "folder /pics//trip/");
+}
+
+#[test]
+fn recently_deleted_shows_the_deleted_photos_of_the_chosen_folder() {
+    let mut c = Catalog::new();
+    let gone = add_with(&mut c, "/pics/trip/a.jpg", |p| p.deleted = true);
+    let kept = add(&mut c, "/pics/trip/b.jpg");
+    let f = Filter { library_folder: Some("/pics/trip".into()), deleted: true, ..Default::default() };
+    assert!(f.matches(c.photo(gone).unwrap(), &c));
+    assert!(!f.matches(c.photo(kept).unwrap(), &c));
+}
+
+#[test]
+fn an_empty_folder_choice_is_no_choice() {
+    let c = library(&["/pics/a.jpg"]);
+    let f = Filter { library_folder: Some("  ".into()), ..Default::default() };
+    assert_eq!(c.query(&f, &Sort::default()).len(), 1);
+}
+
+#[test]
+fn every_row_shows_exactly_as_many_photos_as_it_says() {
+    let c = library(&[
+        "/Users/me/Pictures/2024/a.jpg",
+        "/Users/me/Pictures/2024/b.jpg",
+        "/Users/me/Pictures/2025/c.jpg",
+        "/Users/me/Pictures/d.jpg",
+        "/Volumes/nas/photos/e.jpg",
+        "/Volumes/nas/f.jpg",
+        "/x.jpg",
+        "/a/../b/g.jpg",
+        "D:\\Photos\\2024\\h.jpg",
+        "d:/Photos/i.jpg",
+        "\\\\srv\\s1\\a\\j.jpg",
+        "\\\\?\\UNC\\srv\\s2\\k.jpg",
+        "/pics/trip/l.jpg",
+        "/pics/trip2/m.jpg",
+        "/Volumes/tokyo/Photos/2024/n.jpg",
+        "/Volumes/tokyo/o.jpg",
+        "/run/user/1000/gvfs/smb-share:server=tokyo,share=photos/p.jpg",
+        "/net/osaka/export/q.jpg",
+        r"\\wsl$\Ubuntu\home\me\r.jpg",
+        // odd spellings the tree and the filter must read the same way
+        "//server/share/../other/1.jpg",
+        "//server/share/ok/2.jpg",
+        "//a b//A/f.jpg",
+        r"\\wsl$\Ubuntu\..\a\1.jpg",
+        // a folder that holds other disks as well as photos of its own
+        "/Volumes/1.jpg",
+        "/mnt/2.jpg",
+        "/mnt/disk/3.jpg",
+    ]);
+    fn check(c: &Catalog, nodes: &[FolderNode], seen: &mut usize) {
+        for n in nodes {
+            // the startup volume's row only opens: its path "/" would cover every other disk
+            if n.selectable {
+                let f = Filter { library_folder: Some(n.path.clone()), ..Default::default() };
+                assert_eq!(c.query(&f, &Sort::default()).len(), n.count, "{} ({})", n.name, n.path);
+                *seen += 1;
+            }
+            check(c, &n.children, seen);
+        }
+    }
+    let mut seen = 0;
+    check(&c, &c.folder_tree(), &mut seen);
+    assert!(seen > 10, "{seen} rows checked");
+}
+
+fn find<'a>(nodes: &'a [FolderNode], path: &str) -> Option<&'a FolderNode> {
+    nodes.iter().find_map(|n| if n.path == path { Some(n) } else { find(&n.children, path) })
+}
+
+#[test]
+fn a_row_that_would_cover_other_disks_is_not_offered_for_choosing() {
+    let tree = library(&["/Users/me/a.jpg", "/Volumes/1.jpg", "/Volumes/tokyo/x.jpg", "/mnt/2.jpg"]).folder_tree();
+    let sel = |p: &str| find(&tree, p).map(|n| n.selectable);
+    assert_eq!(sel("/"), Some(false), "the startup disk's path covers every disk");
+    assert_eq!(sel("/Volumes"), Some(false), "it holds the tokyo disk");
+    assert_eq!(sel("/Volumes/tokyo"), Some(true));
+    assert_eq!(sel("/Users/me"), Some(true));
+    assert_eq!(sel("/mnt"), Some(true), "no disk is mounted below it in this library");
+}
+
+#[test]
+fn the_tree_and_the_filter_read_odd_spellings_the_same_way() {
+    // `..` cannot climb above a share, and repeated separators between server and share are one
+    let tree = library(&["//server/share/../other/1.jpg", "//a b//A/f.jpg"]).folder_tree();
+    assert_eq!(outline(&tree), vec![row("//a b/A", 1, 1), row("//server/share", 1, 0), row("other", 1, 1)]);
+}
+
+#[test]
+fn a_path_that_names_no_folder_chooses_nothing() {
+    let c = library(&["/pics/a.jpg", "C:\\b\\c.jpg"]);
+    for dot in [".", "./", "a/..", "../.."] {
+        let f = Filter { library_folder: Some(dot.into()), ..Default::default() };
+        assert_eq!(c.query(&f, &Sort::default()).len(), 0, "{dot:?} is no folder, not everything");
+    }
+}
+
+#[test]
+fn a_folder_is_labelled_by_its_last_two_names() {
+    use crate::folders::folder_label;
+    assert_eq!(folder_label("/Volumes/tokyo/photos/travel"), "photos/travel");
+    assert_eq!(folder_label("/Volumes/tokyo/"), "Volumes/tokyo");
+    assert_eq!(folder_label(r"C:\Photos\2024"), "Photos/2024");
+    assert_eq!(folder_label("/"), "/");
+    assert_eq!(folder_label(""), "");
+}
+
+#[test]
+fn disk_roots_are_told_from_folders() {
+    use crate::folders::{is_disk_root, is_startup_disk};
+    for p in ["/", "C:", "C:\\", r"\\?\C:\", r"\\srv\share", "/Volumes/nas", "/mnt/disk", "/net/host"] {
+        assert!(is_disk_root(p), "{p}");
+    }
+    for p in ["/Volumes", "/Users", "C:\\Photos", r"\\srv\share\x", "/Volumes/nas/p", "."] {
+        assert!(!is_disk_root(p), "{p}");
+    }
+    assert!(is_startup_disk("/") && is_startup_disk("//") && !is_startup_disk("/Volumes/nas") && !is_startup_disk("C:"));
+}
+
+#[test]
+fn a_folder_name_with_spaces_at_its_edge_is_its_own_folder() {
+    // legal on macOS and Linux: `/p/shoot ` is not `/p/shoot`
+    let c = library(&["/p/shoot /a.jpg", "/p/shoot/b.jpg", "/p/other/c.jpg"]);
+    let tree = c.folder_tree();
+    let shoot_space = find(&tree, "/p/shoot ").expect("its own row");
+    let shoot = find(&tree, "/p/shoot").expect("and the other");
+    for (node, want) in [(shoot_space, "a.jpg"), (shoot, "b.jpg")] {
+        let f = Filter { library_folder: Some(node.path.clone()), ..Default::default() };
+        let ids = c.query(&f, &Sort::default());
+        let names: Vec<&str> = ids.iter().filter_map(|id| c.photo(*id)).map(|p| p.file_name.as_str()).collect();
+        assert_eq!(names, [want], "{:?} shows its own photo", node.path);
+    }
+}
+
+#[test]
+fn a_local_photo_never_counts_as_imported_even_when_a_disk_folder_is_also_shown() {
+    let mut c = Catalog::new();
+    let browsed = add_with(&mut c, "/pics/trip/b.jpg", |p| p.local = true);
+    let f = Filter { folder: Some("/pics/trip".into()), library_folder: Some("/pics/trip".into()), ..Default::default() };
+    assert!(!f.matches(c.photo(browsed).unwrap(), &c));
+}
+
+#[test]
+fn the_quick_check_agrees_with_the_careful_one() {
+    use crate::query::{folder_key, key_within, photo_in_root};
+    let paths = [
+        "/a/b/c.jpg",
+        "/a/b/d/e.jpg",
+        "/a/bc/x.jpg",
+        "/a/./b/x.jpg",
+        "/a/../a/b/x.jpg",
+        "/a//b/x.jpg",
+        "/a\\b\\x.jpg",
+        "/a/b ",
+        "/a/b /x.jpg",
+        "/.hidden/x.jpg",
+        "/a/.b/x.jpg",
+        "/x.jpg",
+        "//srv/share/x.jpg",
+        "C:\\a\\b\\x.jpg",
+        "relative/a/b/x.jpg",
+        "/Volumes/tokyo/a/x.jpg",
+        "/日本/写真/x.jpg",
+    ];
+    let roots = ["/", "/a", "/a/b", "/a/b/", "/a/bc", "/a/b ", "/Volumes/tokyo", "//srv/share", "C:\\a", "/日本", "relative", "/a/../a/b"];
+    for p in paths {
+        let mut c = Catalog::new();
+        let id = add(&mut c, p);
+        let photo = c.photo(id).unwrap();
+        for r in roots {
+            let root = folder_key(r);
+            let slow = crate::local::folder_of(photo).is_some_and(|f| key_within(&f, &root));
+            assert_eq!(photo_in_root(photo, &root), slow, "{p:?} in {r:?}");
+        }
+    }
+}
@@ -0,0 +1,200 @@
+//! Issue #102: catalog format versions. Older formats load and are upgraded; a newer format is
+//! refused with a clear error and left untouched (never read in part, truncated or set aside).
+
+use crate::journal::{LOG, SNAPSHOT, VERSION, encode_record};
+use crate::*;
+
+fn photo(c: &mut Catalog) -> Op {
+    let id = c.alloc_photo_id();
+    Op::AddPhoto { photo: Box::new(Photo::new(id, Source::File { path: format!("/p/{}.jpg", id.0) }, "p.jpg", "JPEG", 4, 3, "2026-01-01")) }
+}
+
+/// A catalog with two photos and the log line for a third op on top of it.
+fn legacy_parts() -> (Catalog, String, Catalog) {
+    let mut c = Catalog::new();
+    for _ in 0..2 {
+        let op = photo(&mut c);
+        c.apply(op).unwrap();
+    }
+    let base = c.clone();
+    let id = c.photos().next().unwrap().id;
+    let op = Op::SetRating { id, rating: 4 };
+    c.apply(op.clone()).unwrap();
+    (base, format!("{}\n", encode_record(3, &op)), c)
+}
+
+fn snapshot_version(m: &MemStore) -> u64 {
+    let v: serde_json::Value = serde_json::from_slice(&m.get(SNAPSHOT).unwrap()).unwrap();
+    v["version"].as_u64().unwrap()
+}
+
+#[test]
+fn v1_library_loads_and_is_upgraded() {
+    let (base, log, full) = legacy_parts();
+    let m = MemStore::new();
+    m.set(SNAPSHOT, format!("{{\"format\":\"lightcraft-catalog\",\"version\":1,\"seq\":2,\"catalog\":{}}}\n", base.to_snapshot()).into_bytes());
+    m.set(LOG, log.into_bytes());
+    let (mut j, c, r) = Journal::open(Box::new(m.clone())).unwrap();
+    assert_eq!(c.to_snapshot(), full.to_snapshot());
+    assert_eq!((r.replayed, r.upgraded_from), (1, Some(1)));
+    // rewritten in this format before anything is appended
+    assert_eq!(snapshot_version(&m), u64::from(VERSION));
+    assert!(m.get(LOG).unwrap().is_empty());
+    assert_eq!(j.seq(), 3);
+    j.append(&[Op::SetLabelName { label: ColorLabel::ALL[0], name: Some("x".into()) }]).unwrap();
+    drop(j);
+    let (_, _, r) = Journal::open(Box::new(m)).unwrap();
+    assert_eq!((r.replayed, r.upgraded_from), (1, None));
+}
+
+/// A library saved before albums could be ordered by hand opens as it was and is rewritten as v3.
+#[test]
+fn v2_library_loads_and_is_upgraded() {
+    let (base, log, full) = legacy_parts();
+    let m = MemStore::new();
+    m.set(SNAPSHOT, format!("{{\"format\":\"lightcraft-catalog\",\"version\":2,\"seq\":2,\"catalog\":{}}}\n", base.to_snapshot()).into_bytes());
+    m.set(LOG, log.into_bytes());
+    let (_, c, r) = Journal::open(Box::new(m.clone())).unwrap();
+    assert_eq!(c.to_snapshot(), full.to_snapshot());
+    assert_eq!((r.replayed, r.upgraded_from), (1, Some(2)));
+    assert_eq!(snapshot_version(&m), 3);
+}
+
+#[test]
+fn versionless_log_only_library_is_upgraded() {
+    let mut c = Catalog::new();
+    let op = photo(&mut c);
+    c.apply(op.clone()).unwrap();
+    let m = MemStore::new();
+    m.set(LOG, format!("{}\n", encode_record(1, &op)).into_bytes());
+    let (_, loaded, r) = Journal::open(Box::new(m.clone())).unwrap();
+    assert_eq!(loaded.to_snapshot(), c.to_snapshot());
+    assert_eq!(r.upgraded_from, Some(0));
+    assert_eq!(snapshot_version(&m), u64::from(VERSION));
+}
+
+/// A new library gets a snapshot in this format at once, so an older build never sees a bare log
+/// it would read in part.
+#[test]
+fn new_library_is_written_in_this_format() {
+    let m = MemStore::new();
+    let (_, _, r) = Journal::open(Box::new(m.clone())).unwrap();
+    assert!(r.created);
+    assert_eq!(r.upgraded_from, None);
+    assert_eq!(snapshot_version(&m), u64::from(VERSION));
+}
+
+fn assert_refused_untouched(m: &MemStore) {
+    let before = m.files.lock().unwrap().clone();
+    let e = Journal::open(Box::new(m.clone())).err().expect("refused");
+    assert!(matches!(e, CatalogError::Newer(_)), "{e:?}");
+    let msg = e.to_string();
+    assert!(msg.contains("newer version of LightCraft") && msg.contains("left unchanged"), "{msg}");
+    assert_eq!(*m.files.lock().unwrap(), before, "nothing modified, nothing renamed");
+}
+
+#[test]
+fn newer_snapshot_is_refused_untouched() {
+    let (base, log, _) = legacy_parts();
+    let m = MemStore::new();
+    m.set(
+        SNAPSHOT,
+        format!("{{\"format\":\"lightcraft-catalog\",\"version\":{},\"seq\":2,\"catalog\":{}}}\n", VERSION + 1, base.to_snapshot()).into_bytes(),
+    );
+    m.set(LOG, log.into_bytes());
+    assert_refused_untouched(&m);
+    // even when its catalog doesn't parse as this version's
+    let m = MemStore::new();
+    m.set(
+        SNAPSHOT,
+        format!("{{\"format\":\"lightcraft-catalog\",\"version\":{},\"seq\":2,\"catalog\":{{\"photos\":7}}}}\n", VERSION + 1).into_bytes(),
+    );
+    assert_refused_untouched(&m);
+    let e = Journal::open(Box::new(m)).err().unwrap().to_string();
+    assert!(e.contains(&format!("v{}", VERSION + 1)), "{e}");
+}
+
+/// A log record with a valid CRC whose op this version doesn't know: refused, not taken for a
+/// torn tail (truncated) or damage (renamed) — wherever it is in the log.
+#[test]
+fn unknown_op_in_the_log_is_refused_untouched() {
+    let (base, log, _) = legacy_parts();
+    let future = |seq: u64| {
+        let body = r#"{"op":"fromTheFuture","id":1,"what":"something new"}"#;
+        format!("{{\"seq\":{seq},\"crc\":{},\"op\":{body}}}\n", crc32fast::hash(body.as_bytes()))
+    };
+    let snap = format!("{{\"format\":\"lightcraft-catalog\",\"version\":{VERSION},\"seq\":2,\"catalog\":{}}}\n", base.to_snapshot());
+    let id = base.photos().next().unwrap().id;
+    let after = format!("{}\n", encode_record(5, &Op::SetRating { id, rating: 1 }));
+    for log in [format!("{log}{}", future(4)), format!("{log}{}{after}", future(4)), format!("{}{log}", future(3))] {
+        let m = MemStore::new();
+        m.set(SNAPSHOT, snap.clone().into_bytes());
+        m.set(LOG, log.into_bytes());
+        assert_refused_untouched(&m);
+    }
+    // a known op with a field of an unknown shape is "newer" too
+    let body = r#"{"op":"setRating","id":1,"rating":"five stars"}"#;
+    let m = MemStore::new();
+    m.set(SNAPSHOT, snap.into_bytes());
+    m.set(LOG, format!("{log}{{\"seq\":4,\"crc\":{},\"op\":{body}}}\n", crc32fast::hash(body.as_bytes())).into_bytes());
+    assert_refused_untouched(&m);
+}
+
+/// A CRC mismatch is still damage / a torn tail (not "newer").
+#[test]
+fn bad_crc_is_still_a_torn_tail() {
+    let (base, log, _) = legacy_parts();
+    let m = MemStore::new();
+    m.set(
+        SNAPSHOT,
+        format!("{{\"format\":\"lightcraft-catalog\",\"version\":{VERSION},\"seq\":2,\"catalog\":{}}}\n", base.to_snapshot()).into_bytes(),
+    );
+    let body = r#"{"op":"fromTheFuture"}"#;
+    m.set(LOG, format!("{log}{{\"seq\":4,\"crc\":1,\"op\":{body}}}\n").into_bytes());
+    let (_, _, r) = Journal::open(Box::new(m)).unwrap();
+    assert!(r.torn_bytes > 0 && r.damaged.is_none(), "{r:?}");
+}
+
+/// Guard for the version rule (see `journal` → *Format versions*): adding an [`Op`] variant breaks
+/// this match. When it does, bump `journal::VERSION`, add a row to the format table, and add the
+/// variant here under the new version.
+#[test]
+fn op_variants_are_versioned() {
+    fn since(op: &Op) -> u32 {
+        match op {
+            Op::AddPhoto { .. }
+            | Op::RemovePhoto { .. }
+            | Op::SetRating { .. }
+            | Op::SetFlag { .. }
+            | Op::SetLabel { .. }
+            | Op::SetDevelop { .. }
+            | Op::SetMeta { .. }
+            | Op::SetDeleted { .. }
+            | Op::SetLocal { .. }
+            | Op::SetVersions { .. }
+            | Op::SetHistory { .. }
+            | Op::PushHistory { .. }
+            | Op::AddAlbum { .. }
+            | Op::RemoveAlbum { .. }
+            | Op::RenameAlbum { .. }
+            | Op::MoveAlbum { .. }
+            | Op::SetAlbumPhotos { .. }
+            | Op::SetAlbumCover { .. }
+            | Op::SetAlbumRules { .. }
+            | Op::AddStack { .. }
+            | Op::RemoveStack { .. }
+            | Op::SetStack { .. }
+            | Op::SetCaptured { .. }
+            | Op::SetAnalysis { .. }
+            | Op::SetFile { .. }
+            | Op::Relink { .. }
+            | Op::SetContent { .. }
+            | Op::SetLabelName { .. }
+            | Op::Batch { .. } => 1,
+            Op::SetBrowsed { .. } => 2,
+            Op::SetAlbumOrder { .. } => 3,
+        }
+    }
+    let newest = since(&Op::SetAlbumOrder { id: AlbumId(0), order: None });
+    assert_eq!(newest, VERSION, "the newest op's version must be the current format version");
+}
@@ -59,6 +59,7 @@ fn op_for(c: &mut Catalog, kind: u8, a: u8, b: u8) -> Op {
                    cover: None,
                    smart: None,
                    quick: false,
+                    order: None,
                },
            }
        }
@@ -77,6 +78,7 @@ fn op_for(c: &mut Catalog, kind: u8, a: u8, b: u8) -> Op {
        13 => Op::Batch { ops: c.remove_from_stacks_ops(&[pid]) },
        14 => c.set_top_ops(pid).unwrap_or(Op::Batch { ops: vec![] }),
        9 if b.is_multiple_of(5) => Op::SetLabelName { label: ColorLabel::ALL[a as usize % 5], name: (a % 3 != 1).then(|| format!("Label {b}")) },
+        9 if b % 7 == 3 => Op::SetAlbumOrder { id: aid, order: (!b.is_multiple_of(3)).then_some(u32::from(b % 5)) },
        9 if b.is_multiple_of(2) => {
            Op::SetFile { id: pid, file_name: format!("Renamed_{b}.jpg"), source: Source::File { path: format!("/p/Renamed_{b}.jpg") } }
        }
@@ -274,3 +276,53 @@ fn fs_store_roundtrip() {
    assert_eq!(c3.to_snapshot(), c.to_snapshot());
    let _ = std::fs::remove_dir_all(&dir);
 }
+
+/// The streamed snapshot is byte-for-byte the format written before streaming (issue #37), so
+/// existing libraries load and older builds can read new snapshots; FsStore and MemStore agree.
+#[test]
+fn streamed_snapshot_keeps_the_on_disk_format() {
+    let (m, _, full) = sample_catalog_with_log();
+    let (mut j, c, _) = open(&m);
+    assert_eq!(c.to_snapshot(), full.to_snapshot());
+    j.snapshot(&c).unwrap();
+    let legacy = format!(
+        "{{\"format\":\"lightcraft-catalog\",\"version\":{},\"seq\":{},\"catalog\":{}}}\n",
+        crate::journal::VERSION,
+        j.seq(),
+        c.to_snapshot()
+    );
+    assert_eq!(String::from_utf8(m.get(SNAPSHOT).unwrap()).unwrap(), legacy);
+    assert_eq!(j.stats().last_snapshot.bytes, legacy.len() as u64);
+
+    let dir = std::env::temp_dir().join(format!("lc-journal-stream-{}", std::process::id()));
+    let _ = std::fs::remove_dir_all(&dir);
+    std::fs::create_dir_all(&dir).unwrap();
+    // an old-style snapshot (written by a build before streaming) loads, and is rewritten the same
+    std::fs::write(dir.join(SNAPSHOT), legacy.as_bytes()).unwrap();
+    let (mut fj, fc, r) = Journal::open(Box::new(FsStore::open(&dir).unwrap())).unwrap();
+    assert_eq!((fc.to_snapshot(), r.snapshot_seq), (c.to_snapshot(), j.seq()));
+    fj.snapshot(&fc).unwrap();
+    assert_eq!(std::fs::read(dir.join(SNAPSHOT)).unwrap(), legacy.as_bytes());
+    drop(fj);
+    let _ = std::fs::remove_dir_all(&dir);
+}
+
+/// A snapshot that fails while streaming (serialisation or I/O error) leaves the old file in
+/// place and removes its temp file.
+#[test]
+fn failed_streamed_write_keeps_the_old_file() {
+    let dir = std::env::temp_dir().join(format!("lc-journal-streamfail-{}", std::process::id()));
+    let _ = std::fs::remove_dir_all(&dir);
+    let mut s = FsStore::open(&dir).unwrap();
+    s.write_atomic(SNAPSHOT, b"old").unwrap();
+    let r = s.write_atomic_with(SNAPSHOT, &mut |w| {
+        w.write_all(b"partial new content")?;
+        Err(std::io::Error::other("serialisation failed"))
+    });
+    assert!(r.is_err());
+    assert_eq!(std::fs::read(dir.join(SNAPSHOT)).unwrap(), b"old");
+    assert!(!dir.join(format!("{SNAPSHOT}.tmp")).exists());
+    assert_eq!(s.write_atomic_with(SNAPSHOT, &mut |w| w.write_all(b"new")).unwrap(), 3);
+    assert_eq!(std::fs::read(dir.join(SNAPSHOT)).unwrap(), b"new");
+    let _ = std::fs::remove_dir_all(&dir);
+}
@@ -0,0 +1,267 @@
+//! Forgetting untouched Local browse records (see [`crate::local`]).
+
+use std::sync::Arc;
+
+use lightcraft_develop::DevelopSettings;
+
+use crate::*;
+
+const OLD: &str = "2026-08-01T10:00:00";
+const NOW: &str = "2026-10-05T10:00:00";
+
+/// A Local record of `/pics/<folder>/IMG_<i>.JPG` as a browse catalogues it (with its baseline).
+fn local(c: &mut Catalog, folder: &str, i: u32) -> PhotoId {
+    let id = c.alloc_photo_id();
+    let mut p = Photo::new(id, Source::File { path: format!("/pics/{folder}/IMG_{i}.JPG") }, &format!("IMG_{i}.JPG"), "JPEG", 60, 40, OLD);
+    p.meta.camera = "Cam".into();
+    p.meta.creator = "From EXIF".into(); // file metadata is part of the baseline, not a user change
+    p.focal_and_wb_for_test();
+    p.local = true;
+    p.set_local_baseline();
+    c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+    id
+}
+
+impl Photo {
+    fn focal_and_wb_for_test(&mut self) {
+        self.meta.focal_mm = Some(35.3);
+        self.as_shot_wb = Some((5123.4, 3.1));
+    }
+}
+
+fn browsed(c: &mut Catalog, folder: &str, at: &str) {
+    c.apply(Op::SetBrowsed { folder: format!("/pics/{folder}"), at: Some(at.into()) }).unwrap();
+}
+
+fn plan(c: &Catalog, days: u32) -> ForgetPlan {
+    c.forget_local_plan(NOW, days, &|_| false)
+}
+
+fn run(c: &mut Catalog, days: u32) -> ForgetPlan {
+    let p = plan(c, days);
+    for op in p.ops() {
+        c.apply(op).unwrap();
+    }
+    p
+}
+
+#[test]
+fn untouched_records_of_old_folders_are_forgotten() {
+    let mut c = Catalog::new();
+    let old: Vec<_> = (0..5).map(|i| local(&mut c, "old", i)).collect();
+    let recent: Vec<_> = (0..3).map(|i| local(&mut c, "recent", i)).collect();
+    browsed(&mut c, "old", OLD);
+    browsed(&mut c, "recent", "2026-10-01T09:00:00");
+    // a library photo in the old folder is never a candidate
+    let lib = c.alloc_photo_id();
+    let mut p = Photo::new(lib, Source::File { path: "/pics/old/LIB.JPG".into() }, "LIB.JPG", "JPEG", 6, 4, OLD);
+    p.set_local_baseline();
+    c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+
+    let dry = plan(&c, 30);
+    assert_eq!((dry.local, dry.evict.len(), dry.kept_recent), (8, 5, 3), "{dry:?}");
+    assert_eq!(dry.evict, old);
+    assert_eq!(c.len(), 9, "a plan changes nothing");
+    let done = run(&mut c, 30);
+    assert_eq!(done.evict.len(), 5);
+    assert!(old.iter().all(|id| c.photo(*id).is_none()));
+    assert!(recent.iter().all(|id| c.photo(*id).is_some()));
+    assert!(c.photo(lib).is_some());
+    // the folder has no Local records left: its time goes
+    assert_eq!(done.unstamp, vec!["/pics/old".to_string()]);
+    assert_eq!(c.last_browsed("/pics/old"), None);
+    assert_eq!(c.last_browsed("/pics/recent/"), Some("2026-10-01T09:00:00"));
+    // 0 days = never
+    browsed(&mut c, "recent", OLD);
+    assert!(plan(&c, 0).evict.is_empty());
+    assert_eq!(plan(&c, 30).evict.len(), 3);
+}
+
+#[test]
+fn any_user_change_keeps_a_record() {
+    type Change = Box<dyn Fn(&mut Catalog, PhotoId)>;
+    let set = |f: fn(&mut Photo)| -> Change {
+        Box::new(move |c: &mut Catalog, id: PhotoId| {
+            let mut p = (**c.photo(id).unwrap()).clone();
+            f(&mut p);
+            c.apply(Op::RemovePhoto { id }).unwrap();
+            c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+        })
+    };
+    let op = |f: fn(PhotoId) -> Op| -> Change { Box::new(move |c: &mut Catalog, id: PhotoId| drop(c.apply(f(id)).unwrap())) };
+    let edited = || {
+        let mut d = DevelopSettings::default();
+        d.light.exposure = 0.3;
+        Arc::new(d)
+    };
+    let changes: Vec<(&str, Change)> = vec![
+        ("rating", op(|id| Op::SetRating { id, rating: 3 })),
+        ("flag", op(|id| Op::SetFlag { id, flag: Flag::Reject })),
+        ("label", op(|id| Op::SetLabel { id, label: Some(ColorLabel::Red) })),
+        ("develop", Box::new(move |c, id| drop(c.apply(Op::SetDevelop { id, settings: edited(), label: "Exposure".into(), edited: None }).unwrap()))),
+        ("crop", set(|p| Arc::make_mut(&mut p.develop).crop.flip_h = true)),
+        (
+            "history only",
+            Box::new(|c, id| {
+                let step = HistoryStep { label: "Reset".into(), settings: Arc::new(DevelopSettings::default()) };
+                drop(c.apply(Op::PushHistory { id, step }).unwrap())
+            }),
+        ),
+        ("versions", set(|p| p.versions.push(Version { name: "v".into(), created: NOW.into(), settings: p.develop.clone(), auto: false }))),
+        ("title", set(|p| p.meta.title = "T".into())),
+        ("caption", set(|p| p.meta.caption = "C".into())),
+        ("keywords", set(|p| p.meta.keywords.push("k".into()))),
+        ("location", set(|p| p.meta.city = "Oslo".into())),
+        ("gps", set(|p| p.meta.gps = Some((1.0, 2.0)))),
+        ("copyright", set(|p| p.meta.copyright = "© me".into())),
+        ("creator edit", set(|p| p.meta.creator = "Me".into())),
+        ("capture time", op(|id| Op::SetCaptured { id, captured: Some("2020-01-01T00:00:00".into()) })),
+        ("rename", set(|p| p.file_name = "renamed.JPG".into())),
+        ("culling", op(|id| Op::SetAnalysis { id, analysis: Some(Analysis::default()) })),
+        ("deleted", op(|id| Op::SetDeleted { id, deleted: true })),
+        ("pick", op(|id| Op::SetFlag { id, flag: Flag::Pick })),
+        (
+            "album",
+            Box::new(|c, id| {
+                let a = c.alloc_album_id();
+                let mut al = Album::new(a, "A");
+                al.photos.push(id);
+                drop(c.apply(Op::AddAlbum { album: al }).unwrap())
+            }),
+        ),
+        (
+            "stack",
+            Box::new(|c, id| {
+                let other = local(c, "elsewhere", 99);
+                let s = c.alloc_stack_id();
+                drop(c.apply(Op::AddStack { stack: Stack { id: s, photos: vec![id, other], collapsed: false } }).unwrap())
+            }),
+        ),
+        (
+            "virtual copy",
+            Box::new(|c, id| {
+                let mut v = (**c.photo(id).unwrap()).clone();
+                v.id = c.alloc_photo_id();
+                v.copy_of = Some(id);
+                v.copy_name = Some("Copy 1".into());
+                v.local = false;
+                drop(c.apply(Op::AddPhoto { photo: Box::new(v) }).unwrap())
+            }),
+        ),
+    ];
+    for (what, change) in changes {
+        let mut c = Catalog::new();
+        let id = local(&mut c, "old", 1);
+        let control = local(&mut c, "old", 2);
+        browsed(&mut c, "old", OLD);
+        change(&mut c, id);
+        let p = plan(&c, 30);
+        assert_eq!(p.evict, vec![control], "{what}: {p:?}");
+        assert_eq!(p.kept_touched + p.kept_in_use, 1, "{what}: {p:?}");
+        // the stamp stays: a record in the folder remains
+        assert!(p.unstamp.is_empty(), "{what}");
+    }
+    // a change undone (back to the browse state) is untouched again
+    let mut c = Catalog::new();
+    let id = local(&mut c, "old", 1);
+    browsed(&mut c, "old", OLD);
+    let inv = c.apply(Op::SetRating { id, rating: 2 }).unwrap();
+    assert!(plan(&c, 30).evict.is_empty());
+    c.apply(inv).unwrap();
+    assert_eq!(plan(&c, 30).evict, vec![id]);
+}
+
+/// Catalogs from before last-browsed times (and baselines): nothing is forgotten on the first
+/// run — the folders are stamped now — and baseline-less records are judged by their data.
+#[test]
+fn upgrade_without_times_forgets_nothing_then_counts_from_then() {
+    let mut c = Catalog::new();
+    let a = local(&mut c, "a", 1);
+    // legacy records: no baseline
+    let legacy = |c: &mut Catalog, i: u32, f: fn(&mut Photo)| {
+        let id = c.alloc_photo_id();
+        let mut p = Photo::new(id, Source::File { path: format!("/pics/a/L{i}.JPG") }, "L.JPG", "JPEG", 6, 4, OLD);
+        p.meta.camera = "Cam".into();
+        p.local = true;
+        f(&mut p);
+        c.apply(Op::AddPhoto { photo: Box::new(p) }).unwrap();
+        id
+    };
+    let clean = legacy(&mut c, 1, |_| {});
+    let rated = legacy(&mut c, 2, |p| p.rating = 1);
+    let with_creator = legacy(&mut c, 3, |p| p.meta.creator = "maybe the user".into());
+    let p = run(&mut c, 30);
+    assert!(p.evict.is_empty(), "{p:?}");
+    assert_eq!(p.stamp, vec!["/pics/a".to_string()]);
+    assert_eq!(c.last_browsed("/pics/a"), Some(NOW));
+    // 30+ days later
+    let later = c.forget_local_plan("2026-11-06T10:00:00", 30, &|_| false);
+    let mut evict = later.evict.clone();
+    evict.sort();
+    assert_eq!(evict, vec![a, clean]);
+    assert_eq!(later.kept_touched, 2, "{rated:?} {with_creator:?}");
+}
+
+#[test]
+fn in_use_records_are_kept() {
+    let mut c = Catalog::new();
+    let a = local(&mut c, "old", 1);
+    let b = local(&mut c, "old", 2);
+    browsed(&mut c, "old", OLD);
+    let p = c.forget_local_plan(NOW, 30, &|p| p.id == a);
+    assert_eq!((p.evict.clone(), p.kept_in_use), (vec![b], 1));
+}
+
+/// The ops are journaled: a crash right after forgetting reloads the same catalog, and a
+/// fingerprint survives the snapshot round trip (a reloaded record is still untouched).
+#[test]
+fn forgetting_is_journaled_and_replays() {
+    let m = MemStore::new();
+    let (mut j, mut c, _) = Journal::open(Box::new(m.clone())).unwrap();
+    let mut ops = Vec::new();
+    for i in 0..10 {
+        local(&mut c, if i % 2 == 0 { "old" } else { "new" }, i);
+    }
+    // log what was built so far: the ops are the records as added
+    for p in c.photos() {
+        ops.push(Op::AddPhoto { photo: Box::new((**p).clone()) });
+    }
+    ops.push(Op::SetBrowsed { folder: "/pics/old".into(), at: Some(OLD.into()) });
+    ops.push(Op::SetBrowsed { folder: "/pics/new".into(), at: Some(NOW.into()) });
+    browsed(&mut c, "old", OLD);
+    browsed(&mut c, "new", NOW);
+    j.append(&ops).unwrap();
+    j.snapshot(&c).unwrap();
+    drop(j);
+    // reload from the snapshot: baselines still match
+    let (mut j, mut c, _) = Journal::open(Box::new(m.clone())).unwrap();
+    let p = c.forget_local_plan(NOW, 30, &|_| false);
+    assert_eq!(p.evict.len(), 5, "{p:?}");
+    let ops = p.ops();
+    for op in &ops {
+        c.apply(op.clone()).unwrap();
+    }
+    j.append(&ops).unwrap();
+    let expect = c.to_snapshot();
+    drop(j); // crash: replay the log
+    let (_, c2, r) = Journal::open(Box::new(m)).unwrap();
+    assert_eq!(r.failed, 0);
+    assert_eq!(c2.to_snapshot(), expect);
+    assert_eq!(c2.len(), 5);
+    // nothing left to forget
+    assert!(c2.forget_local_plan(NOW, 30, &|_| false).evict.is_empty());
+}
+
+#[test]
+fn op_photo_ids_cover_nested_ops() {
+    let op = Op::Batch {
+        ops: vec![
+            Op::SetRating { id: PhotoId(1), rating: 1 },
+            Op::SetAlbumPhotos { id: AlbumId(1), photos: vec![PhotoId(2), PhotoId(3)] },
+            Op::Batch { ops: vec![Op::RemovePhoto { id: PhotoId(4) }] },
+        ],
+    };
+    let mut ids = Vec::new();
+    op.photo_ids(&mut |id| ids.push(id.0));
+    assert_eq!(ids, vec![1, 2, 3, 4]);
+}
@@ -0,0 +1,93 @@
+//! Issue #99: one process per library.
+
+use std::path::PathBuf;
+
+use crate::lock::{LOCK, OWNER};
+use crate::*;
+
+fn temp_dir(tag: &str) -> PathBuf {
+    let d = std::env::temp_dir().join(format!("lc-lock-{tag}-{}", std::process::id()));
+    let _ = std::fs::remove_dir_all(&d);
+    std::fs::create_dir_all(&d).unwrap();
+    d
+}
+
+#[test]
+fn second_opener_is_refused_until_the_first_lets_go() {
+    let dir = temp_dir("twice");
+    let first = LibraryLock::acquire(&dir, "LightCraft").unwrap();
+    assert!(first.held());
+    let owner: LockOwner = serde_json::from_slice(&std::fs::read(dir.join(OWNER)).unwrap()).unwrap();
+    assert_eq!((owner.pid, owner.program.as_str()), (std::process::id(), "LightCraft"));
+
+    let e = LibraryLock::acquire(&dir, "lightcraft-cli").unwrap_err();
+    let LockError::InUse(Some(who)) = &e else { panic!("{e:?}") };
+    assert_eq!(who.pid, std::process::id());
+    let msg = e.to_string();
+    assert!(msg.contains("already open in LightCraft") && msg.contains(&format!("process {}", std::process::id())), "{msg}");
+
+    drop(first);
+    assert!(!dir.join(OWNER).exists(), "the owner note goes with the lock");
+    assert!(dir.join(LOCK).exists(), "the lock file itself may stay: only a held lock refuses");
+    let again = LibraryLock::acquire(&dir, "lightcraft-cli").unwrap();
+    assert!(again.held());
+    drop(again);
+    let _ = std::fs::remove_dir_all(&dir);
+}
+
+/// Files left behind by a crash (lock file + owner note) don't lock anyone out.
+#[test]
+fn leftover_lock_files_are_not_a_lock() {
+    let dir = temp_dir("stale");
+    std::fs::write(dir.join(LOCK), b"").unwrap();
+    std::fs::write(dir.join(OWNER), br#"{"pid":999999,"host":"elsewhere","program":"LightCraft","version":"0.2.0","since":1}"#).unwrap();
+    let l = LibraryLock::acquire(&dir, "LightCraft").unwrap();
+    assert!(l.held());
+    let owner: LockOwner = serde_json::from_slice(&std::fs::read(dir.join(OWNER)).unwrap()).unwrap();
+    assert_eq!(owner.pid, std::process::id(), "the note now names this process");
+    drop(l);
+    let _ = std::fs::remove_dir_all(&dir);
+}
+
+/// Another process holds the lock; when it is killed (a crash), the OS releases it.
+#[test]
+fn lock_held_by_another_process_is_released_when_it_dies() {
+    let dir = temp_dir("child");
+    let exe = std::env::current_exe().unwrap();
+    let mut child = std::process::Command::new(exe)
+        .args(["--exact", "tests_lock::lock_holder_child", "--ignored", "--test-threads=1"])
+        .env("LC_LOCK_CHILD_DIR", &dir)
+        .stdout(std::process::Stdio::null())
+        .stderr(std::process::Stdio::null())
+        .spawn()
+        .unwrap();
+    let deadline = std::time::Instant::now() + std::time::Duration::from_secs(60);
+    while !dir.join("child-locked").exists() {
+        assert!(child.try_wait().unwrap().is_none(), "child exited before locking");
+        assert!(std::time::Instant::now() < deadline, "child never locked");
+        std::thread::sleep(std::time::Duration::from_millis(20));
+    }
+    let e = LibraryLock::acquire(&dir, "LightCraft").unwrap_err();
+    let LockError::InUse(Some(who)) = &e else { panic!("{e:?}") };
+    assert_eq!(who.pid, child.id());
+    assert!(e.to_string().contains("on this computer") || who.host.is_empty(), "{e}");
+
+    child.kill().unwrap(); // SIGKILL: no clean-up runs, like a crash
+    child.wait().unwrap();
+    assert!(dir.join(OWNER).exists(), "the crashed holder's note is left behind");
+    let l = LibraryLock::acquire(&dir, "LightCraft").expect("the OS released the dead process's lock");
+    assert!(l.held());
+    drop(l);
+    let _ = std::fs::remove_dir_all(&dir);
+}
+
+/// Run by [`lock_held_by_another_process_is_released_when_it_dies`] in a child process: take the
+/// lock, say so, and wait to be killed.
+#[test]
+#[ignore = "helper process for lock_held_by_another_process_is_released_when_it_dies"]
+fn lock_holder_child() {
+    let Some(dir) = std::env::var_os("LC_LOCK_CHILD_DIR") else { return };
+    let _l = LibraryLock::acquire(std::path::Path::new(&dir), "lightcraft-cli").unwrap();
+    std::fs::write(std::path::Path::new(&dir).join("child-locked"), b"").unwrap();
+    std::thread::sleep(std::time::Duration::from_secs(120));
+}
@@ -0,0 +1,223 @@
+//! Issue #101: an append that fails part-way (disk full, a dropped network share) and is then
+//! retried must not read back as log damage, and must lose nothing.
+
+use std::sync::atomic::{AtomicBool, Ordering};
+use std::sync::{Arc, Mutex};
+
+use crate::journal::LOG;
+use crate::*;
+
+/// A store whose next append writes only a prefix of its data, then fails.
+#[derive(Clone, Default)]
+struct Torn {
+    files: MemStore,
+    /// The next append writes this many bytes, then returns an error.
+    torn_at: Arc<Mutex<Option<usize>>>,
+    /// `truncate` fails (the cut-back after a failed append can't be done right away).
+    truncate_fails: Arc<AtomicBool>,
+}
+
+impl Torn {
+    fn tear_next_append(&self, bytes: usize) {
+        *self.torn_at.lock().unwrap() = Some(bytes);
+    }
+}
+
+impl Store for Torn {
+    fn read(&mut self, name: &str) -> std::io::Result<Option<Vec<u8>>> {
+        self.files.read(name)
+    }
+    fn write_atomic(&mut self, name: &str, data: &[u8]) -> std::io::Result<()> {
+        self.files.write_atomic(name, data)
+    }
+    fn append(&mut self, name: &str, data: &[u8]) -> std::io::Result<()> {
+        if let Some(n) = self.torn_at.lock().unwrap().take() {
+            self.files.append(name, &data[..n.min(data.len())])?;
+            return Err(std::io::Error::other("disk full"));
+        }
+        self.files.append(name, data)
+    }
+    fn truncate(&mut self, name: &str, len: u64) -> std::io::Result<()> {
+        if self.truncate_fails.load(Ordering::SeqCst) {
+            return Err(std::io::Error::other("share unavailable"));
+        }
+        self.files.truncate(name, len)
+    }
+    fn describe(&self) -> String {
+        "torn".into()
+    }
+}
+
+fn add_photo(c: &mut Catalog) -> Op {
+    let id = c.alloc_photo_id();
+    Op::AddPhoto { photo: Box::new(Photo::new(id, Source::File { path: format!("/p/{}.jpg", id.0) }, "p.jpg", "JPEG", 4, 3, "2026-01-01")) }
+}
+
+/// Apply `ops` to the live catalog (like a command) and return them for the journal.
+fn apply(c: &mut Catalog, ops: Vec<Op>) -> Vec<Op> {
+    for op in &ops {
+        c.apply(op.clone()).unwrap();
+    }
+    ops
+}
+
+/// Every tear position of a batch, then the retry (same ops, same seqs, like the engine's queue)
+/// and more appends: reloading replays everything, reports no damage and no torn bytes.
+#[test]
+fn partial_append_then_retry_loads_cleanly() {
+    let probe = MemStore::new();
+    let (mut j, mut c, _) = Journal::open(Box::new(probe.clone())).unwrap();
+    let add = add_photo(&mut c);
+    let first = apply(&mut c, vec![add]);
+    j.append(&first).unwrap();
+    let id = c.photos().next().unwrap().id;
+    let batch = apply(&mut c, vec![Op::SetRating { id, rating: 3 }, Op::SetFlag { id, flag: Flag::Pick }, Op::SetRating { id, rating: 4 }]);
+    let before = probe.get(LOG).unwrap().len();
+    j.append(&batch).unwrap();
+    let batch_len = probe.get(LOG).unwrap().len() - before;
+
+    for tear in 1..batch_len {
+        let t = Torn::default();
+        let (mut j, mut c, _) = Journal::open(Box::new(t.clone())).unwrap();
+        let add = add_photo(&mut c);
+        let first = apply(&mut c, vec![add]);
+        j.append(&first).unwrap();
+        let good_len = t.files.get(LOG).unwrap().len();
+        let batch = apply(&mut c, vec![Op::SetRating { id, rating: 3 }, Op::SetFlag { id, flag: Flag::Pick }, Op::SetRating { id, rating: 4 }]);
+        t.tear_next_append(tear);
+        assert!(j.append(&batch).is_err(), "tear at {tear}");
+        // cut back to the last whole record
+        assert_eq!(t.files.get(LOG).unwrap().len(), good_len, "tear at {tear}");
+        j.append(&batch).unwrap();
+        let more = apply(&mut c, vec![Op::SetLabel { id, label: Some(ColorLabel::ALL[0]) }]);
+        j.append(&more).unwrap();
+        drop(j);
+
+        let (_, loaded, r) = Journal::open(Box::new(t.clone())).unwrap();
+        assert!(r.damaged.is_none(), "tear at {tear}: {r:?}");
+        assert_eq!((r.torn_bytes, r.torn_fragments, r.replayed), (0, 0, 5), "tear at {tear}");
+        assert_eq!(loaded.to_snapshot(), c.to_snapshot(), "tear at {tear}");
+    }
+}
+
+/// When cutting back fails too (the share is gone), the next append cuts back first; until that
+/// works nothing more is appended after the fragment.
+#[test]
+fn failed_cut_back_is_done_before_the_next_append() {
+    let t = Torn::default();
+    let (mut j, mut c, _) = Journal::open(Box::new(t.clone())).unwrap();
+    let add = add_photo(&mut c);
+    let first = apply(&mut c, vec![add]);
+    j.append(&first).unwrap();
+    let good_len = t.files.get(LOG).unwrap().len();
+    let id = c.photos().next().unwrap().id;
+    let batch = apply(&mut c, vec![Op::SetRating { id, rating: 2 }, Op::SetRating { id, rating: 5 }]);
+    t.truncate_fails.store(true, Ordering::SeqCst);
+    t.tear_next_append(30);
+    assert!(j.append(&batch).is_err());
+    assert_eq!(t.files.get(LOG).unwrap().len(), good_len + 30, "the fragment is still there");
+    // still unreachable: the retry fails without writing anything after the fragment
+    assert!(j.append(&batch).is_err());
+    assert_eq!(t.files.get(LOG).unwrap().len(), good_len + 30);
+    t.truncate_fails.store(false, Ordering::SeqCst);
+    j.append(&batch).unwrap();
+    drop(j);
+    let (_, loaded, r) = Journal::open(Box::new(t.clone())).unwrap();
+    assert!(r.damaged.is_none(), "{r:?}");
+    assert_eq!(r.replayed, 3);
+    assert_eq!(loaded.to_snapshot(), c.to_snapshot());
+}
+
+/// Logs written before the fix: the fragment of the failed append is glued to the first retried
+/// record on one line, followed by the rest of the retry. They load in full (the fragment is
+/// skipped), instead of stopping at the fragment and setting the later records aside as damage.
+#[test]
+fn legacy_fragment_glued_to_the_retry_is_skipped() {
+    let m = MemStore::new();
+    let (mut j, mut c, _) = Journal::open(Box::new(m.clone())).unwrap();
+    let add = add_photo(&mut c);
+    let first = apply(&mut c, vec![add]);
+    j.append(&first).unwrap();
+    let good = m.get(LOG).unwrap();
+    let id = c.photos().next().unwrap().id;
+    let batch = apply(
+        &mut c,
+        vec![
+            Op::SetRating { id, rating: 3 },
+            Op::SetFlag { id, flag: Flag::Pick },
+            Op::SetLabelName { label: ColorLabel::ALL[1], name: Some("Größe ✓".into()) },
+        ],
+    );
+    j.append(&batch).unwrap();
+    let full = m.get(LOG).unwrap();
+    let batch_bytes = &full[good.len()..];
+    let later = apply(&mut c, vec![Op::SetRating { id, rating: 1 }]);
+    j.append(&later).unwrap();
+    let later_bytes = m.get(LOG).unwrap()[full.len()..].to_vec();
+    drop(j);
+
+    // every tear (inside the first record, after it, inside a multi-byte character…)
+    for tear in 1..batch_bytes.len() {
+        if batch_bytes[tear - 1] == b'\n' {
+            continue; // a tear on a record boundary leaves no fragment
+        }
+        let mut log = good.clone();
+        log.extend_from_slice(&batch_bytes[..tear]);
+        log.extend_from_slice(batch_bytes);
+        log.extend_from_slice(&later_bytes);
+        let t = MemStore::new();
+        t.set(LOG, log);
+        let (_, loaded, r) = Journal::open(Box::new(t)).unwrap();
+        assert!(r.damaged.is_none(), "tear at {tear}: {r:?}");
+        assert_eq!(r.torn_fragments, 1, "tear at {tear}");
+        assert_eq!(loaded.to_snapshot(), c.to_snapshot(), "tear at {tear}");
+    }
+
+    // a damaged record whose tail happens to be a whole record from the future is still damage
+    let mut log = good.clone();
+    log.extend_from_slice(b"garbage");
+    log.extend_from_slice(&later_bytes);
+    log.extend_from_slice(&later_bytes);
+    let t = MemStore::new();
+    t.set(LOG, log);
+    let (_, _, r) = Journal::open(Box::new(t)).unwrap();
+    assert!(r.damaged.is_some(), "{r:?}");
+}
+
+/// A snapshot can absorb ops whose append failed: it records them as saved (`seq` moves past
+/// them), and appending after it continues from there.
+#[test]
+fn snapshot_with_unlogged_ops_saves_them() {
+    let t = Torn::default();
+    let (mut j, mut c, _) = Journal::open(Box::new(t.clone())).unwrap();
+    let add = add_photo(&mut c);
+    let first = apply(&mut c, vec![add]);
+    j.append(&first).unwrap();
+    let id = c.photos().next().unwrap().id;
+    let queued = apply(&mut c, vec![Op::SetRating { id, rating: 2 }, Op::SetFlag { id, flag: Flag::Reject }]);
+    t.tear_next_append(10);
+    assert!(j.append(&queued).is_err());
+    j.snapshot_with_unlogged(&c, queued.len() as u64).unwrap();
+    assert_eq!(j.seq(), 3);
+    let after = apply(&mut c, vec![Op::SetRating { id, rating: 5 }]);
+    j.append(&after).unwrap();
+    drop(j);
+    let (j, loaded, r) = Journal::open(Box::new(t.clone())).unwrap();
+    assert_eq!((r.snapshot_seq, r.replayed), (3, 1));
+    assert_eq!(j.seq(), 4);
+    assert_eq!(loaded.to_snapshot(), c.to_snapshot());
+}
+
+/// `FsStore::truncate` on a file that was never created (an append that failed first) is a no-op.
+#[test]
+fn fs_store_truncates_a_missing_file_quietly() {
+    let dir = std::env::temp_dir().join(format!("lc-torn-{}", std::process::id()));
+    let _ = std::fs::remove_dir_all(&dir);
+    let mut s = FsStore::open(&dir).unwrap();
+    s.truncate(LOG, 0).unwrap();
+    assert!(!dir.join(LOG).exists());
+    s.append(LOG, b"abc\n").unwrap();
+    s.truncate(LOG, 0).unwrap();
+    assert_eq!(std::fs::read(dir.join(LOG)).unwrap(), b"");
+    let _ = std::fs::remove_dir_all(&dir);
+}
@@ -12,6 +12,7 @@ workspace = true
 lightcraft-geom = { workspace = true }
 lightcraft-color = { workspace = true }
 lightcraft-raster = { workspace = true }
+lightcraft-heif = { workspace = true, optional = true }
 serde = { workspace = true }
 bytemuck = { workspace = true }
 thiserror = { workspace = true }
@@ -34,6 +35,10 @@ ravif = { workspace = true, optional = true }
 default = ["parallel", "jxl", "avif"]
 parallel = ["dep:rayon", "lightcraft-raster/parallel", "jpeg-decoder/rayon"]
 jxl = ["dep:jxl-oxide"]
+# HEIC/HEIF decode via the optional lightcraft-heif crate (heic-rs, pure Rust). Off by default so
+# distributors choose it at build time (HEVC patents are a distributor's call, same policy as
+# PhotoCraft); `cargo xtask ci` clippies and tests it (`--features heif`), releases don't ship it yet.
+heif = ["dep:lightcraft-heif"]
 ## AVIF encoding via ravif/rav1e (pure Rust, BSD). Native targets only; a no-op on wasm32.
 avif = ["dep:ravif"]

@@ -14,6 +14,7 @@ let d: Decoded = decode(&bytes, DecodeOptions::default())?;
 let working: Rgb32f = to_working(&d);                    // linear Rec.2020 D65 (Bradford)
 let fast = decode(&bytes, DecodeOptions::fit(2048, 2048))?; // JPEG: DCT-domain 1/2..1/8 scaling
 let t: Thumbnail = decode_thumbnail(&bytes, 256)?;       // embedded EXIF/MPF preview or scaled decode
+let h: Header = read_header(&bytes)?;                    // stored width/height + orientation, no pixels decoded

 let icc = icc::write_named(NamedSpace::DisplayP3);       // or icc::write_matrix_trc(&space, &Trc::…)
 let meta = EncodeMeta { icc: Some(&icc), exif: Some(&exif), xmp: Some(&xmp) };
@@ -24,6 +25,12 @@ encode_webp_lossless(&img, &meta)?;
 encode_avif(&img, 70, 6, &meta)?;                        // native + feature `avif`
 ```

+Thumbnail fallback decoding limits the source to 64 million pixels by default.
+Non-JPEG decoders allocate at source resolution before resizing, even for a tiny
+thumbnail. Use `decode_thumbnail_with` and `ThumbnailOptions::max_pixels` to set
+a different source budget. Oversized sources return `Error::TooLarge`; embedded
+JPEG previews can still be used without decoding the full source image.
+
 `Format::RawTiffLike` (DNG, CR2, NEF, ARW, PEF, ORF, RW2, SRW, …) and `Format::RawOther` (CR3, RAF,
 CRW, MRW, X3F) are detected so the engine can route them to `lightcraft-raw`; `decode` returns
 `Error::Unsupported` for them.
@@ -50,6 +57,10 @@ or CMYK profile converted by the CMS straight to linear Rec.2020), `Container` (
 EXIF `R03` Adobe RGB hint), `Untagged` (assumed sRGB; untagged float data assumed linear),
 `IccUnsupported` (profile present but unusable → **sRGB fallback**), `Naive` (CMYK without profile).

+`decode_jpeg_with_fallback(bytes, opts, space)` lets an enclosing RAW container supply the colour
+space of a JPEG with no ICC or EXIF metadata. It applies the transfer curve before linear-light
+resizing; a JPEG carrying its own metadata retains its existing interpretation.
+
 ICC profiles are parsed with `moxcms` (v2/v4, `curv`/`para` TRCs, `chad`); sRGB, Display P3,
 Adobe RGB (1998), ProPhoto (ROMM) and Rec.2020 are recognised by colorants. `icc::write_matrix_trc`
 emits v4 display profiles (D50 PCS, Bradford `chad`) for export embedding.
@@ -323,7 +323,9 @@ fn cmyk(raw: &Raw, ch: usize, icc_bytes: Option<&[u8]>, info: Option<&icc::IccIn
    }
    // Naive: RGB = (1 - C)(1 - K) … treated as sRGB-encoded.
    let px = samples
-        .chunks_exact(4)
+        .as_chunks::<4>()
+        .0
+        .iter()
        .map(|c| {
            let k = 1.0 - c[3];
            [(1.0 - c[0]) * k, (1.0 - c[1]) * k, (1.0 - c[2]) * k].map(lightcraft_color::transfer::srgb_to_linear)
@@ -91,7 +91,7 @@ pub fn encode_jpeg(img: &EncodeImage, quality: u8, subsampling: ChromaSubsamplin
    let n = img.width as usize * img.height as usize;
    let (data, ct): (std::borrow::Cow<[u8]>, jpeg_encoder::ColorType) = match img.channels {
        1 => (data[..n].into(), jpeg_encoder::ColorType::Luma),
-        2 => (data.chunks_exact(2).take(n).map(|c| c[0]).collect::<Vec<_>>().into(), jpeg_encoder::ColorType::Luma),
+        2 => (data.as_chunks::<2>().0.iter().take(n).map(|c| c[0]).collect::<Vec<_>>().into(), jpeg_encoder::ColorType::Luma),
        3 => (data[..n * 3].into(), jpeg_encoder::ColorType::Rgb),
        _ => (data[..n * 4].into(), jpeg_encoder::ColorType::Rgba),
    };
@@ -254,7 +254,7 @@ pub fn encode_tiff(img: &EncodeImage, compression: TiffCompression, meta: &Encod
        }};
    }
    fn expand_ga<T: Copy>(s: &[T], n: usize) -> Vec<T> {
-        s.chunks_exact(2).take(n).flat_map(|c| [c[0], c[0], c[0], c[1]]).collect()
+        s.as_chunks::<2>().0.iter().take(n).flat_map(|c| [c[0], c[0], c[0], c[1]]).collect()
    }
    match img.samples {
        Samples::U8(s) => match ch {
@@ -0,0 +1,125 @@
+//! HEIF / HEIC (the iPhone and Mac photo format), read-only.
+//!
+//! Decoding lives in the optional `lightcraft-heif` crate (heic-rs, a pure-Rust HEVC
+//! still-picture decoder), enabled by this crate's `heif` feature — off by default because HEVC
+//! is patent-encumbered and whether a build carries an HEVC decoder is the distributor's call
+//! (the same policy as the sibling PhotoCraft). Without the feature, HEIF files are still
+//! detected and opening one is an [`Error::Unsupported`] error, never a panic.
+//!
+//! HEIF records orientation in the container (`irot`/`imir`, plus a `clap` crop), not in EXIF,
+//! so the decoder applies them and `orientation` is reported as 1: the EXIF Orientation tag only
+//! mirrors what the container says, and applying it as well would turn the photo twice.
+//! The samples come back in the container's own primaries with its ICC profile (if any); like
+//! every decoder here, the colour interpretation happens in [`crate::convert::finish`].
+
+#[cfg(feature = "heif")]
+use crate::convert::{Buf, Meta, Model, Raw, finish};
+use crate::{DecodeOptions, Decoded, Error, Format, Result};
+
+const F: Format = Format::Heif;
+
+/// The reason given when this build has no HEIF decoder.
+#[cfg(not(feature = "heif"))]
+pub(crate) const NOT_IN_BUILD: &str = "HEIC/HEIF support isn't included in this build of LightCraft";
+
+/// The reason a HEIF file that no build can open is refused (this crate decodes HEVC stills
+/// only; the wrapper names the exact construct in its own errors).
+#[cfg(feature = "heif")]
+const NOT_DECODED: &str = "the image is not a decodable HEVC still (image sequence, overlay, alpha-less grid or AVIF-coded item)";
+
+#[cfg(not(feature = "heif"))]
+pub(crate) fn decode(_bytes: &[u8], _opts: &DecodeOptions) -> Result<Decoded> {
+    Err(Error::Unsupported(F, NOT_IN_BUILD))
+}
+
+#[cfg(feature = "heif")]
+fn err(e: lightcraft_heif::Error) -> Error {
+    match e {
+        // This crate's Unsupported takes a fixed reason; the dynamic wording lives in the
+        // lightcraft-heif error itself.
+        lightcraft_heif::Error::Unsupported(_) => Error::Unsupported(F, NOT_DECODED),
+        lightcraft_heif::Error::Limit(m) | lightcraft_heif::Error::Malformed(m) => Error::Malformed(F, m),
+    }
+}
+
+#[cfg(feature = "heif")]
+pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
+    // The container alone: the declared size is checked before any pixel is decoded.
+    let info = lightcraft_heif::probe(bytes).map_err(err)?;
+    if (info.width as u64).saturating_mul(info.height as u64) > opts.max_pixels {
+        return Err(Error::TooLarge(info.width as u64, info.height as u64));
+    }
+    let options = lightcraft_heif::Options { max_pixels: opts.max_pixels, apply_transforms: true };
+    let decoded = lightcraft_heif::decode(bytes, &options).map_err(err)?;
+    // The decoder re-reads the container; trust its output's shape, not the probe's.
+    let buf = if decoded.sixteen_bit {
+        let mut v = Vec::with_capacity(decoded.data.len() / 2);
+        for pair in decoded.data.as_chunks::<2>().0 {
+            v.push(u16::from_ne_bytes(*pair));
+        }
+        Buf::U16(v)
+    } else {
+        Buf::U8(decoded.data)
+    };
+    let raw = Raw {
+        width: decoded.width as usize,
+        height: decoded.height as usize,
+        model: Model::Rgb,
+        alpha: decoded.has_alpha,
+        premultiplied: false,
+        buf,
+        bit_depth: if decoded.sixteen_bit { 16 } else { 8 },
+    };
+    let meta = Meta { icc: decoded.icc, exif: decoded.exif, xmp: decoded.xmp, orientation: Some(1), ..Default::default() };
+    finish(F, raw, meta, (decoded.width, decoded.height), opts)
+}
+
+#[cfg(test)]
+mod tests {
+    use crate::{Format, sniff};
+
+    /// The container header both builds recognise, whatever the feature says.
+    const HEIC_HEADER: &[u8] = b"\0\0\0\x18ftypheic\0\0\0\0mif1heic";
+
+    #[test]
+    fn heif_is_always_recognised() {
+        assert_eq!(sniff(HEIC_HEADER), Some(Format::Heif));
+    }
+
+    #[test]
+    fn can_decode_follows_the_feature() {
+        assert_eq!(Format::Heif.can_decode(), cfg!(feature = "heif"));
+    }
+
+    #[cfg(not(feature = "heif"))]
+    #[test]
+    fn without_the_feature_heif_is_a_clear_unsupported_error() {
+        let r = crate::decode(HEIC_HEADER, crate::DecodeOptions::default());
+        assert!(matches!(&r, Err(crate::Error::Unsupported(Format::Heif, why)) if why.contains("isn't included in this build")), "{r:?}");
+    }
+
+    /// Found by PhotoCraft's `decode_heif` fuzz target: a malformed box makes heic-rs 0.1.1 slice
+    /// out of range (`boxes.rs:130`). It must be an error, not a crash.
+    #[cfg(feature = "heif")]
+    const HEIC_RS_BOX_PANIC: [u8; 72] = [
+        0x00, 0x00, 0x00, 0x24, 0x66, 0x74, 0x79, 0x70, 0x68, 0x65, 0x69, 0x63, 0x00, 0x00, 0x00, 0x00, 0x6d, 0x69, 0x66, 0x31, 0x4d, 0x69, 0x50,
+        0x72, 0x6d, 0x69, 0x61, 0x66, 0x4d, 0x69, 0x48, 0x42, 0x72, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x6d, 0x65, 0x74, 0x61, 0x00, 0x00,
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x10, 0x75, 0x75, 0x69, 0x64, 0x00, 0x00, 0x00, 0x00, 0x03, 0x08, 0x08, 0x08, 0x00, 0x03, 0x01, 0x03, 0x70,
+        0x00, 0xa8, 0x00,
+    ];
+
+    #[cfg(feature = "heif")]
+    #[test]
+    fn a_heic_rs_panic_is_a_malformed_error() {
+        let r = crate::decode(&HEIC_RS_BOX_PANIC, crate::DecodeOptions::default());
+        assert!(matches!(&r, Err(crate::Error::Malformed(Format::Heif, _))), "{r:?}");
+    }
+
+    #[cfg(feature = "heif")]
+    #[test]
+    fn an_image_sequence_is_unsupported() {
+        // An ftyp with no meta box: a video track, not a still.
+        let r = crate::decode(b"\0\0\0\x18ftypmsf1\0\0\0\0msf1hevc", crate::DecodeOptions::default());
+        assert!(matches!(&r, Err(crate::Error::Unsupported(Format::Heif, _))), "{r:?}");
+    }
+}
@@ -146,7 +146,19 @@ pub fn write_matrix_trc(space: &RgbSpace, trc: &Trc) -> Vec<u8> {
    };
    p.description = Some(ProfileText::Localizable(vec![LocalizableString::new("en".into(), "US".into(), desc)]));
    p.copyright = Some(ProfileText::Localizable(vec![LocalizableString::new("en".into(), "US".into(), "No copyright, use freely".into())]));
-    align_tags(&p.encode().unwrap_or_default())
+    align_tags(&fixed_date(p.encode().unwrap_or_default()))
+}
+
+/// Zero the header's creation date/time (ICC.1 §7.2.8, bytes 24..36) so a profile's bytes depend
+/// only on its colour space and curve. `moxcms` stamps the wall clock there on every `encode`
+/// (ignoring `ColorProfile::creation_date_time`), which made two exports of one photo differ in
+/// the embedded profile and nowhere else, so output hashes could not be compared (#180). Readers
+/// do not interpret the field; all zeros is the conventional "unspecified" value.
+fn fixed_date(mut b: Vec<u8>) -> Vec<u8> {
+    if let Some(date) = b.get_mut(24..36) {
+        date.fill(0);
+    }
+    b
 }

 /// Re-lay a profile's tag data on 4-byte boundaries (ICC.1 §7.3.1). The encoder packs tags back to
@@ -236,7 +248,7 @@ pub(crate) fn cms_to_linear_rec2020(icc: &[u8], samples: &[f32], channels: usize
        for (s, d) in src_chunks.zip(dst_chunks) {
            t.transform(s, d).ok()?;
        }
-        Some(out.chunks_exact(3).map(|c| [c[0], c[1], c[2]]).collect())
+        Some(out.as_chunks::<3>().0.iter().map(|c| [c[0], c[1], c[2]]).collect())
    };
    std::panic::catch_unwind(run).ok().flatten()
 }
@@ -300,6 +312,23 @@ mod tests {
        let _ = SRGB;
    }

+    /// Two profiles for one space are byte-identical, and the header date is fixed: exports of the
+    /// same photo with the same settings hash the same (#180).
+    #[test]
+    fn written_profiles_are_deterministic() {
+        for n in NamedSpace::ALL {
+            let a = write_named(n);
+            let b = write_named(n);
+            assert_eq!(a, b, "{n:?}");
+            assert_eq!(&a[24..36], &[0u8; 12], "{n:?} header date/time");
+            // the profile ID is "not computed", not a hash of a timestamp
+            assert_eq!(&a[84..100], &[0u8; 16], "{n:?} profile ID");
+        }
+        let lin = write_matrix_trc(&PROPHOTO, &Trc::Linear);
+        assert_eq!(lin, write_matrix_trc(&PROPHOTO, &Trc::Linear));
+        assert_eq!(&lin[24..36], &[0u8; 12]);
+    }
+
    #[test]
    fn garbage_is_none() {
        assert!(parse(&[]).is_none());
@@ -105,12 +105,67 @@ pub(crate) fn parse_markers(b: &[u8]) -> Option<Markers> {
 }

 pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
+    decode_with_fallback(bytes, opts, None)
+}
+
+/// Stored dimensions and EXIF orientation from the markers, without decoding the scan data. The
+/// file is refused, as a decode would refuse it, when it has no frame header, an unsupported frame
+/// type or component count, or when its markers don't run to an end-of-image marker after the
+/// image data (a truncated file). Damage *inside* the entropy-coded data can't be seen this way.
+pub(crate) fn header(bytes: &[u8]) -> Result<(u32, u32, u16)> {
+    let m = parse_markers(bytes).ok_or_else(|| Error::Malformed(F, "missing SOI".into()))?;
+    if m.components == 0 {
+        return Err(Error::Malformed(F, "no frame header".into()));
+    }
+    check_size(F, m.width as u64, m.height as u64, &DecodeOptions::default())?;
+    if !matches!(m.sof, 0xC0..=0xC3) {
+        return Err(Error::Unsupported(F, "arithmetic-coded or hierarchical JPEG"));
+    }
+    if !matches!(m.components, 1 | 3 | 4) {
+        return Err(Error::Malformed(F, format!("{} colour components", m.components)));
+    }
+    if m.first_scan_components == 0 || !reaches_eoi(bytes) {
+        return Err(Error::Malformed(F, "truncated: the image data doesn't end with an end-of-image marker".into()));
+    }
+    let orientation = m.exif.as_deref().map(exif::summarize).unwrap_or_default().orientation.unwrap_or(1);
+    Ok((m.width, m.height, orientation))
+}
+
+/// Whether the markers after SOI, followed through every segment and scan, reach an EOI marker
+/// (the first one: the primary image's, ahead of any MPF images appended to the file). Segments
+/// are skipped by their length; in scan data `FF 00` (stuffing), `FF FF` (fill) and RSTn are not
+/// markers that end it.
+fn reaches_eoi(b: &[u8]) -> bool {
+    let mut p = 2usize;
+    loop {
+        // entropy-coded data (or garbage between segments, as `parse_markers` tolerates it)
+        let Some(skip) = b.get(p..).and_then(find_ff) else { return false };
+        p = p.saturating_add(skip);
+        let Some(&marker) = b.get(p.saturating_add(1)) else { return false };
+        p = match marker {
+            0xD9 => return true,
+            0xFF => p.saturating_add(1),
+            0x00 | 0x01 | 0xD0..=0xD8 => p.saturating_add(2),
+            _ => {
+                let Some(len) = b.get(p.saturating_add(2)..p.saturating_add(4)).map(|l| u16::from_be_bytes([l[0], l[1]]) as usize) else {
+                    return false;
+                };
+                if len < 2 {
+                    return false;
+                }
+                p.saturating_add(2).saturating_add(len)
+            }
+        };
+    }
+}
+
+pub(crate) fn decode_with_fallback(bytes: &[u8], opts: &DecodeOptions, fallback: Option<crate::NamedSpace>) -> Result<Decoded> {
    let m = parse_markers(bytes).ok_or_else(|| Error::Malformed(F, "missing SOI".into()))?;
    if m.components == 0 {
        return Err(Error::Malformed(F, "no frame header".into()));
    }
    check_size(F, m.width as u64, m.height as u64, opts)?;
-    let scale_to = opts.max_size.filter(|&(mw, mh)| mw > 0 && mh > 0 && (m.width >= 2 * mw || m.height >= 2 * mh));
+    let scale_to = opts.max_size.filter(|&(mw, mh)| mw > 0 && mh > 0 && (m.width >= mw.saturating_mul(2) || m.height >= mh.saturating_mul(2)));
    // zune-jpeg handles the common cases fastest; jpeg-decoder covers DCT scaling, CMYK/YCCK,
    // 12-bit and non-interleaved sequential scans (which zune-jpeg 0.5 mis-decodes with subsampling).
    let non_interleaved = matches!(m.sof, 0xC0 | 0xC1) && m.first_scan_components < m.components;
@@ -119,7 +174,10 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    } else {
        decode_zune(bytes, &m)?
    };
-    let meta = Meta { icc: m.icc, exif: m.exif, xmp: m.xmp, ..Default::default() };
+    // A raw container may carry the colour declaration of an otherwise untagged preview.
+    // Apply it before linearization/resizing; existing JPEG metadata always wins.
+    let hint = fallback.filter(|_| m.icc.is_none() && m.exif.is_none()).map(|space| crate::SourceSpace::named(space, crate::SpaceOrigin::Container));
+    let meta = Meta { icc: m.icc, exif: m.exif, xmp: m.xmp, hint, ..Default::default() };
    finish(F, raw, meta, (m.width, m.height), opts)
 }

@@ -159,7 +217,7 @@ fn decode_jpeg_decoder(bytes: &[u8], m: &Markers, scale_to: Option<(u32, u32)>)
    let (model, buf, depth) = match info.pixel_format {
        PixelFormat::L8 => (Model::Gray, Buf::U8(px), 8),
        PixelFormat::L16 => {
-            let v = px.chunks_exact(2).map(|c| u16::from_be_bytes([c[0], c[1]])).collect();
+            let v = px.as_chunks::<2>().0.iter().map(|c| u16::from_be_bytes([c[0], c[1]])).collect();
            (Model::Gray, Buf::U16(v), 16)
        }
        PixelFormat::RGB24 => (Model::Rgb, Buf::U8(px), 8),
@@ -168,6 +226,21 @@ fn decode_jpeg_decoder(bytes: &[u8], m: &Markers, scale_to: Option<(u32, u32)>)
    Ok(Raw { width: w, height: h, model, alpha: false, premultiplied: false, buf, bit_depth: depth })
 }

+/// The first `0xFF` byte, eight bytes at a time (scan data has one every few hundred bytes: this
+/// is most of [`header`]'s time).
+fn find_ff(b: &[u8]) -> Option<usize> {
+    const ONES: u64 = 0x0101_0101_0101_0101;
+    let (words, tail) = b.as_chunks::<8>();
+    for (i, w) in words.iter().enumerate() {
+        // a zero byte in !w is an 0xFF byte in w
+        let x = !u64::from_ne_bytes(*w);
+        if x.wrapping_sub(ONES) & !x & (ONES << 7) != 0 {
+            return w.iter().position(|&v| v == 0xFF).map(|j| i * 8 + j);
+        }
+    }
+    tail.iter().position(|&v| v == 0xFF).map(|j| words.len() * 8 + j)
+}
+
 /// The box to request from `jpeg-decoder::scale` so the decoded image still covers `mw × mh` once
 /// fitted with the source aspect ratio.
 fn scaled_request(w: u32, h: u32, mw: u32, mh: u32) -> (u16, u16) {
@@ -226,7 +299,7 @@ fn mpf_images(m: &Markers) -> Vec<(usize, usize)> {
    let Some(e) = entries.iter().find(|e| e.tag == 0xB002) else { return vec![] };
    let Some(list) = t.bytes(e) else { return vec![] };
    let mut out = Vec::new();
-    for (i, rec) in list.chunks_exact(16).enumerate().take(16) {
+    for (i, rec) in list.as_chunks::<16>().0.iter().enumerate().take(16) {
        if i == 0 {
            continue;
        }
@@ -246,12 +319,75 @@ fn mpf_images(m: &Markers) -> Vec<(usize, usize)> {
 mod tests {
    use super::*;

+    #[test]
+    fn container_fallback_matches_tagged_jpeg_before_resizing_and_keeps_own_metadata() {
+        use crate::{ChromaSubsampling, EncodeImage, EncodeMeta, NamedSpace, Samples, SpaceOrigin, encode_jpeg};
+        let pixels: Vec<u8> = (0..16 * 16).flat_map(|i| if i % 2 == 0 { [120, 180, 90] } else { [30, 100, 200] }).collect();
+        let encode = |icc, exif| {
+            encode_jpeg(
+                &EncodeImage::new(16, 16, 3, Samples::U8(&pixels)),
+                100,
+                ChromaSubsampling::S444,
+                &EncodeMeta { icc, exif, ..Default::default() },
+            )
+            .unwrap()
+        };
+        let untagged = encode(None, None);
+        let adobe = crate::icc::write_named(NamedSpace::AdobeRgb);
+        let tagged = encode(Some(&adobe), None);
+        for opts in [DecodeOptions::default(), DecodeOptions::fit(3, 3)] {
+            let with = crate::decode_jpeg_with_fallback(&untagged, opts, NamedSpace::AdobeRgb).unwrap();
+            let expected = crate::decode(&tagged, opts).unwrap();
+            assert_eq!(with.space.origin, SpaceOrigin::Container);
+            let (a, b) = (with.to_working(), expected.to_working());
+            for (p, q) in a.data.iter().zip(&b.data) {
+                assert!(p.iter().zip(q).all(|(a, b)| (a - b).abs() < 0.0001), "{p:?} != {q:?}");
+            }
+        }
+        let srgb = crate::icc::write_named(NamedSpace::Srgb);
+        let exif = crate::exif::minimal_exif(6);
+        for bytes in [encode(Some(&srgb), None), encode(None, Some(&exif)), encode(Some(b"bad ICC"), None)] {
+            let opts = DecodeOptions::fit(5, 5);
+            let with = crate::decode_jpeg_with_fallback(&bytes, opts, NamedSpace::AdobeRgb).unwrap();
+            let expected = crate::decode(&bytes, opts).unwrap();
+            assert_eq!(with.space, expected.space);
+            assert_eq!(with.image, expected.image);
+            assert_eq!(with.orientation, expected.orientation);
+        }
+        assert!(crate::decode_jpeg_with_fallback(b"bad JPEG", DecodeOptions::default(), NamedSpace::AdobeRgb).is_err());
+    }
+
    #[test]
    fn scaled_request_covers() {
        assert_eq!(scaled_request(6000, 4000, 256, 256), (256, 171));
        assert_eq!(scaled_request(100, 100, 256, 256), (100, 100));
    }

+    #[test]
+    fn find_ff_matches_a_plain_search() {
+        let mut b: Vec<u8> = (0..300u32).map(|i| (i * 7 % 255) as u8).collect();
+        for at in [None, Some(0), Some(7), Some(8), Some(15), Some(203), Some(296), Some(299)] {
+            if let Some(i) = at {
+                b[i] = 0xFF;
+            }
+            for start in [0, 1, 5, 9, 290, 299, 300] {
+                let s = &b[start..];
+                assert_eq!(find_ff(s), s.iter().position(|&v| v == 0xFF), "{at:?} from {start}");
+            }
+        }
+    }
+
+    #[test]
+    fn unbounded_fit_box_decodes_at_full_size() {
+        // a full-size load asks for the largest box (`usize::MAX` saturated to `u32::MAX`): no DCT scaling,
+        // and no overflow working out whether to scale
+        use crate::{ChromaSubsampling, EncodeImage, EncodeMeta, Samples, encode_jpeg};
+        let pixels = vec![128u8; 32 * 24 * 3];
+        let jpeg = encode_jpeg(&EncodeImage::new(32, 24, 3, Samples::U8(&pixels)), 90, ChromaSubsampling::S444, &EncodeMeta::default()).unwrap();
+        let d = crate::decode(&jpeg, DecodeOptions::fit(u32::MAX, u32::MAX)).unwrap();
+        assert_eq!((d.width, d.height), (32, 24));
+    }
+
    #[test]
    fn markers_on_garbage() {
        assert!(parse_markers(b"").is_none());
@@ -274,7 +274,7 @@ pub fn encode(
    {
        let rows: Vec<(usize, Vec<&mut [f32]>)> = {
            let mut its: Vec<_> = full.iter_mut().map(|p| p.chunks_mut(width)).collect();
-            (0..height).map(|y| (y, its.iter_mut().map(|it| it.next().unwrap()).collect())).collect()
+            (0..height).map_while(|y| Some((y, its.iter_mut().map(|it| it.next()).collect::<Option<_>>()?))).collect()
        };
        let conv = |(y, mut out): (usize, Vec<&mut [f32]>)| {
            let src = &data[y * width * channels..(y + 1) * width * channels];
@@ -6,16 +6,20 @@
 //!   profiles (matrix/TRC exactly, LUT/CMYK via the `moxcms` CMS), 16-bit and float precision are
 //!   preserved. EXIF/XMP/ICC blobs are passed through untouched; orientation is reported, never applied.
 //! - [`to_working`] converts a decode result to linear Rec.2020 D65 (the pipeline working space).
+//! - [`read_header`] reads a file's stored dimensions and orientation from its headers, without
+//!   decoding pixels (import probes).
 //! - [`decode_thumbnail`] is the fast path for grid thumbnails (EXIF thumbnail or DCT-scaled decode).
 //! - [`encode`] writes JPEG, PNG, TIFF, lossless WebP and (native, feature `avif`) AVIF, embedding
 //!   ICC/EXIF/XMP. [`icc::write_matrix_trc`] builds profiles for export.
 //!
 //! No decoder panics on malformed input (property-tested).
 #![forbid(unsafe_code)]
+#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)]

 mod convert;
 pub mod encode;
 pub mod exif;
+mod heif;
 pub mod icc;
 mod jpeg;
 pub mod jpeg_par;
@@ -155,6 +159,60 @@ pub fn decode(bytes: &[u8], opts: DecodeOptions) -> Result<Decoded> {
    std::panic::catch_unwind(|| decode_unguarded(bytes, format, &opts)).unwrap_or_else(|_| Err(Error::Malformed(format, "decoder panicked".into())))
 }

+/// What a file's headers say about its image: enough to catalogue it without decoding pixels.
+#[derive(Clone, Copy, Debug, PartialEq, Eq)]
+pub struct Header {
+    pub format: Format,
+    /// Full-resolution stored dimensions, orientation **not** applied ([`Decoded::source_width`]).
+    pub width: u32,
+    pub height: u32,
+    /// EXIF/TIFF orientation 1..=8, 1 when absent ([`Decoded::orientation`]).
+    pub orientation: u16,
+}
+
+/// Read the stored dimensions and orientation of any format [`decode`] supports, agreeing with
+/// what a decode reports, without decoding pixels: JPEG, PNG, TIFF, WebP and PSD are read from
+/// their headers (markers, chunks, IFD, resources). GIF, BMP and JPEG XL fall back to a small
+/// decode.
+///
+/// The header readers also refuse what a decode would refuse before reaching the pixels (an
+/// unsupported layout, compression or frame type) and files whose image data is cut short
+/// (truncated: the data runs past the end of the file, or a JPEG's markers don't reach its
+/// end-of-image marker). Damage *inside* compressed data that is all present is only found by a
+/// decode. Malformed input yields an error, never a panic (as for [`decode`], a panic inside a
+/// third-party parser is caught on unwinding targets).
+pub fn read_header(bytes: &[u8]) -> Result<Header> {
+    let format = sniff(bytes).ok_or(Error::UnknownFormat)?;
+    std::panic::catch_unwind(|| read_header_unguarded(bytes, format))
+        .unwrap_or_else(|_| Err(Error::Malformed(format, "header reader panicked".into())))
+}
+
+/// [`read_header`] without the panic guard (for fuzzing our own code paths).
+#[doc(hidden)]
+pub fn read_header_unguarded(bytes: &[u8], format: Format) -> Result<Header> {
+    let (width, height, orientation) = match format {
+        Format::Jpeg => jpeg::header(bytes)?,
+        Format::Png => png_codec::header(bytes)?,
+        Format::Tiff => tiff_codec::header(bytes)?,
+        Format::WebP => webp::header(bytes)?,
+        Format::Psd => psd::header(bytes)?,
+        // rare here (or, for JPEG XL, without a header reader yet): a small decode, as before
+        _ => {
+            let d = decode_unguarded(bytes, format, &DecodeOptions::fit(64, 64))?;
+            (d.source_width, d.source_height, d.orientation)
+        }
+    };
+    Ok(Header { format, width, height, orientation })
+}
+
+/// Decode a JPEG preview with its enclosing container's colour space as a fallback.
+/// A JPEG carrying ICC or EXIF metadata keeps its own interpretation. The fallback is applied
+/// before transfer decoding and linear-light resizing, never to already-resampled pixels.
+pub fn decode_jpeg_with_fallback(bytes: &[u8], opts: DecodeOptions, fallback: NamedSpace) -> Result<Decoded> {
+    std::panic::catch_unwind(|| jpeg::decode_with_fallback(bytes, &opts, Some(fallback)))
+        .unwrap_or_else(|_| Err(Error::Malformed(Format::Jpeg, "decoder panicked".into())))
+}
+
 /// [`decode`] without the panic guard (for fuzzing our own code paths).
 #[doc(hidden)]
 pub fn decode_unguarded(bytes: &[u8], format: Format, opts: &DecodeOptions) -> Result<Decoded> {
@@ -170,8 +228,8 @@ pub fn decode_unguarded(bytes: &[u8], format: Format, opts: &DecodeOptions) -> R
        Format::Jxl => jxl::decode(bytes, &opts),
        #[cfg(not(feature = "jxl"))]
        Format::Jxl => Err(Error::Unsupported(format, "built without the `jxl` feature")),
+        Format::Heif => heif::decode(bytes, &opts),
        Format::Avif => Err(Error::Unsupported(format, "no pure-Rust, permissively licensed AV1 decoder yet")),
-        Format::Heif => Err(Error::Unsupported(format, "no pure-Rust, permissively licensed HEVC decoder yet")),
        Format::RawTiffLike | Format::RawOther => Err(Error::Unsupported(format, "camera raw: decode with lightcraft-raw")),
    }
 }
@@ -216,11 +274,19 @@ pub struct ThumbnailOptions {
    /// smaller size). Defaults to `max_edge`, i.e. only previews that need no upscaling. Lower it
    /// (e.g. 160 for typical EXIF thumbnails) for an instant placeholder.
    pub min_embedded_edge: u32,
+    /// Maximum source pixels for fallback decoding (default 64 MP). Non-JPEG formats
+    /// decode at source resolution before resizing, so output size does not bound memory.
+    #[serde(default = "thumbnail_max_pixels")]
+    pub max_pixels: u64,
+}
+
+fn thumbnail_max_pixels() -> u64 {
+    64_000_000
 }

 impl ThumbnailOptions {
    pub fn new(max_edge: u32) -> Self {
-        ThumbnailOptions { max_edge, min_embedded_edge: max_edge }
+        ThumbnailOptions { max_edge, min_embedded_edge: max_edge, max_pixels: thumbnail_max_pixels() }
    }
 }

@@ -239,7 +305,7 @@ pub fn decode_thumbnail_with(bytes: &[u8], opts: &ThumbnailOptions) -> Result<Th
    {
        return Ok(t);
    }
-    let d = decode(bytes, DecodeOptions::fit(max_edge, max_edge))?;
+    let d = decode(bytes, DecodeOptions { max_size: Some((max_edge, max_edge)), max_pixels: opts.max_pixels })?;
    Ok(Thumbnail {
        image: d.to_srgb8(),
        orientation: d.orientation,
@@ -37,7 +37,7 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    };
    buf.truncate(out.buffer_size());
    let (samples, depth) = match out.bit_depth {
-        png::BitDepth::Sixteen => (Buf::U16(buf.chunks_exact(2).map(|c| u16::from_be_bytes([c[0], c[1]])).collect()), 16),
+        png::BitDepth::Sixteen => (Buf::U16(buf.as_chunks::<2>().0.iter().map(|c| u16::from_be_bytes([c[0], c[1]])).collect()), 16),
        _ => (Buf::U8(buf), info.bit_depth as u8),
    };
    let raw = Raw { width: out.width as usize, height: out.height as usize, model, alpha, premultiplied: false, buf: samples, bit_depth: depth };
@@ -57,6 +57,75 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    finish(F, raw, meta, (w, h), opts)
 }

+/// Stored dimensions and EXIF orientation (`eXIf`, wherever it is) from the chunk headers, without
+/// inflating the image data. Refused, as a decode would refuse it: an invalid IHDR, an indexed
+/// image without a palette, no IDAT, or a file that ends before the chunk after the last IDAT
+/// begins (truncated image data). Chunk CRCs and the deflate stream are not checked.
+pub(crate) fn header(b: &[u8]) -> Result<(u32, u32, u16)> {
+    if !b.starts_with(b"\x89PNG\r\n\x1a\n") {
+        return Err(err("bad signature"));
+    }
+    let be32 = |p: usize| b.get(p..p.checked_add(4)?).map(|v| u32::from_be_bytes([v[0], v[1], v[2], v[3]]));
+    if be32(8) != Some(13) || b.get(12..16) != Some(b"IHDR") {
+        return Err(err("missing IHDR"));
+    }
+    let ihdr = b.get(16..29).ok_or_else(|| err("truncated IHDR"))?;
+    let (w, h) = (u32::from_be_bytes([ihdr[0], ihdr[1], ihdr[2], ihdr[3]]), u32::from_be_bytes([ihdr[4], ihdr[5], ihdr[6], ihdr[7]]));
+    let (depth, color, compression, filter, interlace) = (ihdr[8], ihdr[9], ihdr[10], ihdr[11], ihdr[12]);
+    if w > i32::MAX as u32 || h > i32::MAX as u32 {
+        return Err(err("dimensions out of range"));
+    }
+    check_size(F, w as u64, h as u64, &DecodeOptions::default())?;
+    let depth_ok = match color {
+        0 => matches!(depth, 1 | 2 | 4 | 8 | 16),
+        3 => matches!(depth, 1 | 2 | 4 | 8),
+        2 | 4 | 6 => matches!(depth, 8 | 16),
+        _ => false,
+    };
+    if !depth_ok || compression != 0 || filter != 0 || interlace > 1 {
+        return Err(err("invalid IHDR"));
+    }
+    let (mut p, mut palette, mut idat, mut after_idat, mut exif) = (8usize, false, false, false, None);
+    while let Some(len) = be32(p) {
+        let Some(kind) = p.checked_add(4).and_then(|k| b.get(k..k.checked_add(4)?)) else { break };
+        if idat && kind != b"IDAT" {
+            // a chunk follows the image data: the IDAT run is complete
+            after_idat = true;
+        }
+        if kind == b"IEND" {
+            break;
+        }
+        let start = p.saturating_add(8);
+        let data = (len <= i32::MAX as u32).then(|| start.checked_add(len as usize).and_then(|end| b.get(start..end))).flatten();
+        let Some(data) = data else {
+            if after_idat {
+                break; // a damaged chunk after the image data: a decode ignores it too
+            }
+            return Err(err("truncated"));
+        };
+        match kind {
+            b"PLTE" => palette = true,
+            b"IDAT" if after_idat => {} // a stray IDAT after other chunks: the first run is the image
+            b"IDAT" => {
+                if color == 3 && !palette {
+                    return Err(err("indexed image without a palette"));
+                }
+                idat = true;
+            }
+            b"eXIf" if exif.is_none() => exif = Some(data.strip_prefix(b"Exif\0\0").unwrap_or(data)),
+            _ => {}
+        }
+        p = start.saturating_add(data.len()).saturating_add(4);
+    }
+    if !idat {
+        return Err(err("no image data"));
+    }
+    if !after_idat {
+        return Err(err("truncated image data"));
+    }
+    Ok((w, h, exif.map(crate::exif::summarize).unwrap_or_default().orientation.unwrap_or(1)))
+}
+
 fn container_hint(info: &png::Info) -> Option<SourceSpace> {
    if let Some(c) = &info.coding_independent_code_points {
        let named = match c.color_primaries {
@@ -69,8 +69,8 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    let (icc, exif, xmp) = resources(res);

    // Layer and mask information: only the layer count sign matters (merged transparency).
-    let lm_len = if psb { r.u64()? as usize } else { r.u32()? as usize };
-    let lm = r.take(lm_len)?;
+    let lm_len = if psb { r.u64()? } else { u64::from(r.u32()?) };
+    let lm = r.take(usize::try_from(lm_len).map_err(|_| err("truncated"))?)?;
    let merged_transparency = if lm.len() >= if psb { 10 } else { 6 } {
        let off = if psb { 8 } else { 4 };
        let count = i16::from_be_bytes([lm[off], lm[off + 1]]);
@@ -88,33 +88,6 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
        1 => return Err(Error::Unsupported(F, "1-bit bitmap PSD")),
        _ => return Err(err("bad depth")),
    };
-    let plane = width * height * bpc;
-    let row_bytes = width * bpc;
-    let mut planes: Vec<Vec<u8>> = Vec::with_capacity(channels);
-    match compression {
-        0 => {
-            for _ in 0..channels {
-                planes.push(r.take(plane)?.to_vec());
-            }
-        }
-        1 => {
-            let counts_len = channels * height;
-            let mut counts = Vec::with_capacity(counts_len);
-            for _ in 0..counts_len {
-                counts.push(if psb { r.u32()? as usize } else { r.u16()? as usize });
-            }
-            for c in 0..channels {
-                let mut out = Vec::with_capacity(plane);
-                for y in 0..height {
-                    let src = r.take(counts[c * height + y])?;
-                    unpackbits(src, row_bytes, &mut out);
-                }
-                planes.push(out);
-            }
-        }
-        _ => return Err(Error::Unsupported(F, "ZIP-compressed merged image data")),
-    }
-
    let (model, n_color) = match mode {
        1 | 8 => (Model::Gray, 1),
        2 => (Model::Rgb, 1),
@@ -128,50 +101,106 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
        return Err(err("too few channels"));
    }
    let alpha = merged_transparency && channels > n_color && mode != 2;
+    // Only the colour (+ alpha) planes are decoded; extra channels (spot colours, masks) are
+    // stored after them and never touched.
    let used = n_color + alpha as usize;
-    let n = width * height;
+    let too_large = || Error::TooLarge(width as u64, height as u64);
+    let n = width.checked_mul(height).ok_or_else(too_large)?;
+    let row_bytes = width.checked_mul(bpc).ok_or_else(too_large)?;
+    let plane = n.checked_mul(bpc).ok_or_else(too_large)?;
+    let mut planes: Vec<Vec<u8>> = Vec::with_capacity(used);
+    match compression {
+        0 => {
+            // Raw planes are borrowed from the input, so their size is bounded by the file.
+            for _ in 0..used {
+                planes.push(r.take(plane)?.to_vec());
+            }
+        }
+        1 => {
+            // Row byte-count table: one entry per row per channel (all channels, since the data
+            // follows it). Taken from the input before anything is allocated from it.
+            let entry = if psb { 4 } else { 2 };
+            let table_len = channels.checked_mul(height).and_then(|c| c.checked_mul(entry)).ok_or_else(|| err("truncated"))?;
+            let table = r.take(table_len)?;
+            let count = |i: usize| -> usize {
+                let at = i.saturating_mul(entry);
+                match table.get(at..at.saturating_add(entry)) {
+                    Some(&[a, b]) => usize::from(u16::from_be_bytes([a, b])),
+                    Some(&[a, b, c, d]) => u32::from_be_bytes([a, b, c, d]) as usize,
+                    _ => 0,
+                }
+            };
+            // PackBits expands at most 64× (a 2-byte repeat run yields 128 bytes), so a row needs
+            // at least this many encoded bytes. Requiring it (and the data being present) before
+            // allocating bounds the decoded planes by the file size: a tiny file can no longer
+            // declare gigabytes of zero-padded or repeat-run rows.
+            let min_row = row_bytes.div_ceil(128).saturating_mul(2);
+            let mut data_len = 0usize;
+            for i in 0..used.saturating_mul(height) {
+                let c = count(i);
+                if c < min_row {
+                    return Err(err("RLE row too short for its width"));
+                }
+                data_len = data_len.saturating_add(c);
+            }
+            if data_len > bytes.len().saturating_sub(r.p) {
+                return Err(err("truncated"));
+            }
+            for c in 0..used {
+                let mut out = zeroed::<u8>(plane, too_large)?;
+                for (y, row) in out.chunks_exact_mut(row_bytes).enumerate() {
+                    let src = r.take(count(c * height + y))?;
+                    unpackbits(src, row)?;
+                }
+                planes.push(out);
+            }
+        }
+        _ => return Err(Error::Unsupported(F, "ZIP-compressed merged image data")),
+    }
+    let samples = n.checked_mul(used).ok_or_else(too_large)?;

    // Interleave to device samples.
    let buf = if mode == 2 {
        // Indexed → 8-bit RGB via the 768-byte palette.
-        if color_mode_data.len() < 768 || depth != 8 {
+        let (Some(pal), true) = (color_mode_data.get(..768), depth == 8) else {
            return Err(err("bad indexed palette"));
-        }
-        let mut v = Vec::with_capacity(n * 3);
-        for &i in &planes[0] {
+        };
+        let idx = planes.first().ok_or_else(|| err("missing plane"))?;
+        let mut v = zeroed::<u8>(n.checked_mul(3).ok_or_else(too_large)?, too_large)?;
+        for (o, &i) in v.as_chunks_mut::<3>().0.iter_mut().zip(idx) {
            let i = i as usize;
-            v.extend_from_slice(&[color_mode_data[i], color_mode_data[256 + i], color_mode_data[512 + i]]);
+            *o = [pal[i], pal[256 + i], pal[512 + i]];
        }
        Buf::U8(v)
    } else {
+        let cmyk = model == Model::Cmyk;
        match bpc {
            1 => {
-                let mut v = vec![0u8; n * used];
-                for (c, p) in planes.iter().take(used).enumerate() {
-                    for i in 0..n {
-                        let x = p[i];
+                let mut v = zeroed::<u8>(samples, too_large)?;
+                for (c, p) in planes.iter().enumerate() {
+                    for (o, &x) in v.iter_mut().skip(c).step_by(used).zip(p) {
                        // PSD CMYK stores 255 = no ink; our CMYK convention is 0 = no ink.
-                        v[i * used + c] = if model == Model::Cmyk && c < 4 { 255 - x } else { x };
+                        *o = if cmyk && c < 4 { 255 - x } else { x };
                    }
                }
                Buf::U8(v)
            }
            2 => {
-                let mut v = vec![0u16; n * used];
-                for (c, p) in planes.iter().take(used).enumerate() {
-                    for i in 0..n {
-                        let x = u16::from_be_bytes([p[2 * i], p[2 * i + 1]]);
-                        v[i * used + c] = if model == Model::Cmyk && c < 4 { 65535 - x } else { x };
+                let mut v = zeroed::<u16>(samples, too_large)?;
+                for (c, p) in planes.iter().enumerate() {
+                    for (o, s) in v.iter_mut().skip(c).step_by(used).zip(p.as_chunks::<2>().0) {
+                        let x = u16::from_be_bytes(*s);
+                        *o = if cmyk && c < 4 { 65535 - x } else { x };
                    }
                }
                Buf::U16(v)
            }
            _ => {
-                let mut v = vec![0f32; n * used];
-                for (c, p) in planes.iter().take(used).enumerate() {
-                    for i in 0..n {
-                        let x = f32::from_be_bytes([p[4 * i], p[4 * i + 1], p[4 * i + 2], p[4 * i + 3]]);
-                        v[i * used + c] = if model == Model::Cmyk && c < 4 { 1.0 - x } else { x };
+                let mut v = zeroed::<f32>(samples, too_large)?;
+                for (c, p) in planes.iter().enumerate() {
+                    for (o, s) in v.iter_mut().skip(c).step_by(used).zip(p.as_chunks::<4>().0) {
+                        let x = f32::from_be_bytes(*s);
+                        *o = if cmyk && c < 4 { 1.0 - x } else { x };
                    }
                }
                Buf::F32(v)
@@ -182,25 +211,111 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    finish(F, raw, Meta { icc, exif, xmp, ..Default::default() }, (width as u32, height as u32), opts)
 }

-/// PackBits: append exactly `row_bytes` bytes (zero-padded if the run data is short).
-fn unpackbits(mut src: &[u8], row_bytes: usize, out: &mut Vec<u8>) {
-    let target = out.len() + row_bytes;
-    while out.len() < target && !src.is_empty() {
-        let h = src[0] as i8;
-        src = &src[1..];
-        if h >= 0 {
-            let n = (h as usize + 1).min(src.len()).min(target - out.len());
-            out.extend_from_slice(&src[..n]);
-            src = &src[(h as usize + 1).min(src.len())..];
-        } else if h != -128 {
-            let n = ((1 - h as isize) as usize).min(target - out.len());
-            if let Some(&b) = src.first() {
-                out.extend(std::iter::repeat_n(b, n));
-                src = &src[1..];
+/// Stored dimensions and EXIF orientation (resource 1058) from the header and image resources,
+/// without reading the merged image data. Refused, as a decode would refuse it: an unsupported
+/// depth, colour mode or compression, or merged image data that runs past the end of the file
+/// (truncated). The run-length data itself is not unpacked.
+pub(crate) fn header(bytes: &[u8]) -> Result<(u32, u32, u16)> {
+    let mut r = Rd { b: bytes, p: 0 };
+    if r.take(4)? != b"8BPS" {
+        return Err(err("bad signature"));
+    }
+    let psb = match r.u16()? {
+        1 => false,
+        2 => true,
+        _ => return Err(err("bad version")),
+    };
+    r.take(6)?;
+    let channels = r.u16()? as usize;
+    let height = r.u32()? as usize;
+    let width = r.u32()? as usize;
+    let depth = r.u16()?;
+    let mode = r.u16()?;
+    check_size(F, width as u64, height as u64, &DecodeOptions::default())?;
+    if channels == 0 || channels > 56 {
+        return Err(err("bad channel count"));
+    }
+    let cm_len = r.u32()? as usize;
+    let color_mode_data = r.take(cm_len)?;
+    let res_len = r.u32()? as usize;
+    let (_, exif, _) = resources(r.take(res_len)?);
+    let lm_len = if psb { r.u64()? as usize } else { r.u32()? as usize };
+    r.take(lm_len)?;
+    let compression = r.u16()?;
+    let bpc = match depth {
+        8 => 1usize,
+        16 => 2,
+        32 => 4,
+        1 => return Err(Error::Unsupported(F, "1-bit bitmap PSD")),
+        _ => return Err(err("bad depth")),
+    };
+    let n_color = match mode {
+        1 | 8 | 7 | 2 => 1,
+        3 => 3,
+        4 => 4,
+        9 => return Err(Error::Unsupported(F, "Lab PSD")),
+        _ => return Err(err("unknown colour mode")),
+    };
+    if channels < n_color {
+        return Err(err("too few channels"));
+    }
+    if mode == 2 && (color_mode_data.len() < 768 || depth != 8) {
+        return Err(err("bad indexed palette"));
+    }
+    let rest = bytes.len().saturating_sub(r.p);
+    let need = match compression {
+        0 => width.checked_mul(height).and_then(|n| n.checked_mul(bpc)).and_then(|n| n.checked_mul(channels)),
+        1 => {
+            // the row byte counts, then the rows they count
+            let rows = channels.checked_mul(height).ok_or_else(|| err("truncated"))?;
+            let table = rows.checked_mul(if psb { 4 } else { 2 }).filter(|&t| t <= rest).ok_or_else(|| err("truncated"))?;
+            let mut total = table;
+            for _ in 0..rows {
+                total = total.saturating_add(if psb { r.u32()? as usize } else { r.u16()? as usize });
            }
+            Some(total)
+        }
+        _ => return Err(Error::Unsupported(F, "ZIP-compressed merged image data")),
+    };
+    if need.is_none_or(|n| n > rest) {
+        return Err(err("truncated"));
+    }
+    let orientation = exif.as_deref().map(crate::exif::summarize).unwrap_or_default().orientation.unwrap_or(1);
+    Ok((width as u32, height as u32, orientation))
+}
+
+/// A zero-filled buffer whose allocation failure is an error, not an abort.
+fn zeroed<T: Clone + Default>(len: usize, too_large: impl Fn() -> Error) -> Result<Vec<T>> {
+    let mut v = Vec::new();
+    v.try_reserve_exact(len).map_err(|_| too_large())?;
+    v.resize(len, T::default());
+    Ok(v)
+}
+
+/// PackBits: decode one row into `row` (pre-zeroed, so short run data leaves zero padding).
+/// A run that would write past the end of the row is an error; trailing bytes are ignored.
+fn unpackbits(mut src: &[u8], row: &mut [u8]) -> Result<()> {
+    let mut at = 0;
+    while at < row.len() {
+        let Some((&h, rest)) = src.split_first() else { break };
+        src = rest;
+        let h = h as i8;
+        if h >= 0 {
+            let n = h as usize + 1;
+            let lit = src.get(..n).unwrap_or(src);
+            let dst = row.get_mut(at..at + lit.len()).ok_or_else(|| err("RLE row overruns its width"))?;
+            dst.copy_from_slice(lit);
+            at += lit.len();
+            src = src.get(n..).unwrap_or(&[]);
+        } else if h != -128 {
+            let n = (1 - h as isize) as usize;
+            let Some((&b, rest)) = src.split_first() else { break };
+            src = rest;
+            row.get_mut(at..at + n).ok_or_else(|| err("RLE row overruns its width"))?.fill(b);
+            at += n;
        }
    }
-    out.resize(target, 0);
+    Ok(())
 }

 fn resources(mut b: &[u8]) -> (Option<Vec<u8>>, Option<Vec<u8>>, Option<String>) {
@@ -234,11 +349,89 @@ mod tests {

    #[test]
    fn packbits() {
-        let mut out = Vec::new();
-        unpackbits(&[0xFE, 0xAA, 0x02, 0x80, 0x00, 0x2A], 6, &mut out);
-        assert_eq!(out, vec![0xAA, 0xAA, 0xAA, 0x80, 0x00, 0x2A]);
-        let mut out = Vec::new();
-        unpackbits(&[0x05, 1], 4, &mut out);
-        assert_eq!(out, vec![1, 0, 0, 0]);
+        let mut out = [0u8; 6];
+        unpackbits(&[0xFE, 0xAA, 0x02, 0x80, 0x00, 0x2A], &mut out).unwrap();
+        assert_eq!(out, [0xAA, 0xAA, 0xAA, 0x80, 0x00, 0x2A]);
+        // Short run data is zero-padded.
+        let mut out = [0u8; 4];
+        unpackbits(&[0x01, 1], &mut out).unwrap();
+        assert_eq!(out, [1, 0, 0, 0]);
+        // Truncated literal run: copies what is there.
+        let mut out = [0u8; 4];
+        unpackbits(&[0x03, 7, 8], &mut out).unwrap();
+        assert_eq!(out, [7, 8, 0, 0]);
+    }
+
+    #[test]
+    fn packbits_overrun_is_an_error() {
+        // Repeat run of 128 into a 4-byte row.
+        let mut out = [0u8; 4];
+        assert!(unpackbits(&[0x81, 0xAA], &mut out).is_err());
+        // Literal run of 6 into a 4-byte row.
+        assert!(unpackbits(&[0x05, 1, 2, 3, 4, 5, 6], &mut out).is_err());
+    }
+
+    /// PSD header + empty colour-mode/resources/layer sections + compression id.
+    fn header(channels: u16, w: u32, h: u32, depth: u16, mode: u16, compression: u16) -> Vec<u8> {
+        let mut b = b"8BPS".to_vec();
+        b.extend_from_slice(&1u16.to_be_bytes());
+        b.extend_from_slice(&[0; 6]);
+        b.extend_from_slice(&channels.to_be_bytes());
+        b.extend_from_slice(&h.to_be_bytes());
+        b.extend_from_slice(&w.to_be_bytes());
+        b.extend_from_slice(&depth.to_be_bytes());
+        b.extend_from_slice(&mode.to_be_bytes());
+        b.extend_from_slice(&[0; 12]); // colour mode data, resources, layer & mask: all empty
+        b.extend_from_slice(&compression.to_be_bytes());
+        b
+    }
+
+    #[test]
+    fn huge_rle_with_truncated_count_table_is_rejected_before_allocating() {
+        // 1 × 2^30 px, 56 channels: the row table alone would be 56 Gi entries; the file has 4 bytes.
+        let mut b = header(56, 1, 1 << 30, 32, 3, 1);
+        b.extend_from_slice(&[0, 0, 0, 0]);
+        let e = decode(&b, &DecodeOptions::default()).unwrap_err();
+        assert!(matches!(e, Error::Malformed(_, ref m) if m == "truncated"), "{e:?}");
+    }
+
+    #[test]
+    fn huge_rle_of_empty_rows_is_rejected_before_allocating() {
+        // 2^30 × 1 px RGB at 32 bits: every row count is 0, which used to zero-pad 3 × 4 GiB.
+        let mut b = header(3, 1 << 30, 1, 32, 3, 1);
+        b.extend_from_slice(&[0; 6]);
+        let e = decode(&b, &DecodeOptions::default()).unwrap_err();
+        assert!(matches!(e, Error::Malformed(_, ref m) if m.contains("too short")), "{e:?}");
+        // 1 Gpx RGB 8-bit whose (plausible) row counts claim more data than the file holds.
+        let mut b = header(3, 1 << 20, 1 << 10, 8, 3, 1);
+        b.extend_from_slice(&[0xFF; 3 * 1024 * 2]);
+        let e = decode(&b, &DecodeOptions::default()).unwrap_err();
+        assert!(matches!(e, Error::Malformed(_, ref m) if m == "truncated"), "{e:?}");
+    }
+
+    #[test]
+    fn rle_repeat_run_past_row_width_is_an_error() {
+        // 4 × 1 gray 8-bit; the single row is a 128-byte repeat run.
+        let mut b = header(1, 4, 1, 8, 1, 1);
+        b.extend_from_slice(&2u16.to_be_bytes());
+        b.extend_from_slice(&[0x81, 0x55]);
+        let e = decode(&b, &DecodeOptions::default()).unwrap_err();
+        assert!(matches!(e, Error::Malformed(_, ref m) if m.contains("overruns")), "{e:?}");
+    }
+
+    #[test]
+    fn rle_gray16_and_extra_channels_decode() {
+        // 2 × 2 gray 16-bit with an extra (unused) channel whose data is absent: only the used
+        // plane is decoded. Row: repeat run of 4 bytes 0x80 -> samples 0x8080.
+        let mut b = header(2, 2, 2, 16, 1, 1);
+        for _ in 0..4 {
+            b.extend_from_slice(&2u16.to_be_bytes());
+        }
+        b.extend_from_slice(&[0xFD, 0x80, 0xFD, 0x80]);
+        let d = decode(&b, &DecodeOptions::default()).unwrap();
+        assert_eq!((d.width, d.height), (2, 2));
+        let p = d.image.get(1, 1);
+        let q = d.image.get(0, 0);
+        assert!(p[0] > 0.0 && (p[0] - q[0]).abs() < 1e-6, "{p:?}");
    }
 }
@@ -66,7 +66,8 @@ impl Format {
    pub fn can_decode(self) -> bool {
        match self {
            Format::Jxl => cfg!(feature = "jxl"),
-            Format::Avif | Format::Heif | Format::RawTiffLike | Format::RawOther => false,
+            Format::Heif => cfg!(feature = "heif"),
+            Format::Avif | Format::RawTiffLike | Format::RawOther => false,
            _ => true,
        }
    }
@@ -163,6 +164,9 @@ fn tiff_is_raw(b: &[u8]) -> bool {
    }
    let Some(t) = Tiff::new(b) else { return false };
    let Some(ifd0) = t.first_ifd() else { return false };
+    let sony = t.ifd(ifd0).is_some_and(|(entries, _)| {
+        entries.iter().any(|e| e.tag == 0x010f && t.bytes(e).and_then(|v| v.get(..4)).is_some_and(|v| v.eq_ignore_ascii_case(b"SONY")))
+    });
    let mut stack = vec![ifd0];
    let mut seen = 0;
    while let Some(pos) = stack.pop() {
@@ -171,6 +175,12 @@ fn tiff_is_raw(b: &[u8]) -> bool {
            break;
        }
        let Some((entries, _)) = t.ifd(pos) else { continue };
+        // Sony M/S lossless ARWs retain a dummy CFA pattern, but store linear YCbCr.
+        // Requiring the raw pattern plus this layout keeps ordinary Sony TIFFs as TIFFs.
+        let uint = |tag| entries.iter().find(|e| e.tag == tag).and_then(|e| t.uint(e));
+        if sony && entries.iter().any(|e| e.tag == 0x828e) && uint(0x0103) == Some(7) && uint(0x0106) == Some(6) && uint(0x0115) == Some(3) {
+            return true;
+        }
        for e in &entries {
            match e.tag {
                0xC612 => return true, // DNGVersion
@@ -233,4 +243,38 @@ mod tests {
        b.extend_from_slice(&0u32.to_le_bytes());
        assert_eq!(sniff(&b), Some(Format::RawTiffLike));
    }
+
+    #[test]
+    fn sony_linear_ycbcr_subifd_is_raw_but_ordinary_tiff_is_not() {
+        fn fixture(compression: u32, pattern: bool) -> Vec<u8> {
+            let mut b = b"II*\0\x08\0\0\0".to_vec();
+            let entry = |b: &mut Vec<u8>, tag: u16, typ: u16, count: u32, value: u32| {
+                b.extend_from_slice(&tag.to_le_bytes());
+                b.extend_from_slice(&typ.to_le_bytes());
+                b.extend_from_slice(&count.to_le_bytes());
+                b.extend_from_slice(&value.to_le_bytes());
+            };
+            b.extend_from_slice(&2u16.to_le_bytes());
+            entry(&mut b, 0x010f, 2, 5, 38);
+            entry(&mut b, 0x014a, 4, 1, 44);
+            b.extend_from_slice(&0u32.to_le_bytes());
+            b.extend_from_slice(b"SONY\0\0");
+            b.extend_from_slice(&(if pattern { 4u16 } else { 3 }).to_le_bytes());
+            entry(&mut b, 0x0103, 3, 1, compression);
+            entry(&mut b, 0x0106, 3, 1, 6);
+            entry(&mut b, 0x0115, 3, 1, 3);
+            if pattern {
+                entry(&mut b, 0x828e, 1, 4, u32::MAX);
+            }
+            b.extend_from_slice(&0u32.to_le_bytes());
+            b
+        }
+        let mut raw = fixture(7, true);
+        assert_eq!(sniff(&raw), Some(Format::RawTiffLike));
+        assert_eq!(sniff(&fixture(7, false)), Some(Format::Tiff));
+        assert_eq!(sniff(&fixture(1, true)), Some(Format::Tiff));
+        assert_eq!(sniff(&raw[..38]), Some(Format::Tiff));
+        raw[38..42].copy_from_slice(b"TEST");
+        assert_eq!(sniff(&raw), Some(Format::Tiff));
+    }
 }
@@ -1,5 +1,6 @@
-//! TIFF decode via the `tiff` crate: 8/16/32-bit integer, 16/32/64-bit float, gray/RGB/CMYK/palette,
-//! alpha (associated or not), ICC (34675), XMP (700), orientation (274). First image only.
+//! TIFF decode via the `tiff` crate: 1/2/4/8/16/32/64-bit integer, 16/32/64-bit float, gray/RGB/CMYK,
+//! alpha (associated or not), ICC (34675), XMP (700), orientation (274). First image only. Palette
+//! (indexed-colour) images are refused by the `tiff` crate itself, on every read path.

 use crate::convert::{Buf, Meta, Model, Raw, check_size, finish};
 use crate::{DecodeOptions, Decoded, Error, Format, Result};
@@ -14,10 +15,7 @@ fn err(e: impl std::fmt::Display) -> Error {
 }

 pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
-    let mut limits = Limits::default();
-    limits.decoding_buffer_size = opts.max_pixels.saturating_mul(4 * 8).min(usize::MAX as u64) as usize;
-    limits.ifd_value_size = 32 << 20;
-    let mut d = Decoder::new(std::io::Cursor::new(bytes)).map_err(err)?.with_limits(limits);
+    let mut d = Decoder::new(std::io::Cursor::new(bytes)).map_err(err)?.with_limits(limits(opts));
    let (w, h) = d.dimensions().map_err(err)?;
    check_size(F, w as u64, h as u64, opts)?;
    let ct = d.colortype().map_err(err)?;
@@ -25,26 +23,12 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    let icc = tag_bytes(&mut d, Tag::IccProfile);
    let xmp = tag_bytes(&mut d, Tag::from_u16_exhaustive(700)).map(|v| String::from_utf8_lossy(&v).trim_end_matches('\0').to_string());
    let orientation = d.get_tag_u32(Tag::Orientation).ok().map(|v| v as u16);
-    let photometric = d.get_tag_u32(Tag::PhotometricInterpretation).ok();
    let planar = d.get_tag_u32(Tag::PlanarConfiguration).ok() == Some(2);
    let extra = d.get_tag_u16_vec(Tag::ExtraSamples).ok().unwrap_or_default();
    let colormap = d.get_tag_u16_vec(Tag::ColorMap).ok();
    let compression = d.get_tag_u32(Tag::Compression).ok();

-    let (model, n_color, n_samples, bits) = match ct {
-        ColorType::Gray(b) => (Model::Gray, 1, 1, b),
-        ColorType::GrayA(b) => (Model::Gray, 1, 2, b),
-        ColorType::RGB(b) => (Model::Rgb, 3, 3, b),
-        ColorType::RGBA(b) => (Model::Rgb, 3, 4, b),
-        ColorType::CMYK(b) => (Model::Cmyk, 4, 4, b),
-        ColorType::CMYKA(b) => (Model::Cmyk, 4, 5, b),
-        ColorType::Palette(b) => (Model::Rgb, 1, 1, b),
-        // JPEG-compressed YCbCr is converted to RGB by the JPEG decoder.
-        ColorType::YCbCr(b) if compression == Some(7) => (Model::Rgb, 3, 3, b),
-        ColorType::Multiband { bit_depth, num_samples } if num_samples >= 3 => (Model::Rgb, 3, num_samples as usize, bit_depth),
-        ColorType::Multiband { bit_depth, num_samples } if num_samples >= 1 => (Model::Gray, 1, num_samples as usize, bit_depth),
-        other => return Err(Error::Unsupported(F, unsupported_name(other))),
-    };
+    let (model, n_color, n_samples, bits) = layout(ct, compression, planar)?;
    let has_alpha = n_samples > n_color && !matches!(ct, ColorType::Palette(_));
    let premultiplied = has_alpha && extra.first() == Some(&1);

@@ -52,6 +36,7 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    let (wu, hu) = (w as usize, h as usize);
    let n = wu * hu;
    let mut buf = match res {
+        DecodingResult::U8(v) if bits < 8 => Buf::U8(unpack(&v, wu, hu, n_samples, bits)?),
        DecodingResult::U8(v) => Buf::U8(v),
        DecodingResult::U16(v) => Buf::U16(v),
        // Rare integer widths: keep 16 significant bits (plenty for display-referred data).
@@ -94,14 +79,8 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
        return finish(F, raw, meta, (w, h), opts);
    }

-    // WhiteIsZero → invert.
-    if photometric == Some(0) {
-        match &mut buf {
-            Buf::U8(v) => v.iter_mut().for_each(|x| *x = 255 - *x),
-            Buf::U16(v) => v.iter_mut().for_each(|x| *x = 65535 - *x),
-            Buf::F32(v) => v.iter_mut().for_each(|x| *x = 1.0 - *x),
-        }
-    }
+    // WhiteIsZero needs nothing here: the `tiff` crate inverts those samples as it reads them (and
+    // refuses WhiteIsZero layouts it can't invert), so inverting again would show a negative.

    // Drop extra samples beyond colour + one alpha.
    let keep = n_color + has_alpha as usize;
@@ -113,6 +92,67 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    finish(F, raw, meta, (w, h), opts)
 }

+fn limits(opts: &DecodeOptions) -> Limits {
+    let mut limits = Limits::default();
+    limits.decoding_buffer_size = opts.max_pixels.saturating_mul(4 * 8).min(usize::MAX as u64) as usize;
+    limits.ifd_value_size = 32 << 20;
+    limits
+}
+
+/// The sample layout of a colour type we can decode: (model, colour samples, samples per pixel,
+/// bits per sample).
+fn layout(ct: ColorType, compression: Option<u32>, planar: bool) -> Result<(Model, usize, usize, u8)> {
+    let l = match ct {
+        ColorType::Gray(b) => (Model::Gray, 1, 1, b),
+        ColorType::GrayA(b) => (Model::Gray, 1, 2, b),
+        ColorType::RGB(b) => (Model::Rgb, 3, 3, b),
+        ColorType::RGBA(b) => (Model::Rgb, 3, 4, b),
+        ColorType::CMYK(b) => (Model::Cmyk, 4, 4, b),
+        ColorType::CMYKA(b) => (Model::Cmyk, 4, 5, b),
+        ColorType::Palette(b) => (Model::Rgb, 1, 1, b),
+        // JPEG-compressed YCbCr is converted to RGB by the JPEG decoder.
+        ColorType::YCbCr(b) if compression == Some(7) => (Model::Rgb, 3, 3, b),
+        ColorType::Multiband { bit_depth, num_samples } if num_samples >= 3 => (Model::Rgb, 3, num_samples as usize, bit_depth),
+        ColorType::Multiband { bit_depth, num_samples } if num_samples >= 1 => (Model::Gray, 1, num_samples as usize, bit_depth),
+        other => return Err(Error::Unsupported(F, unsupported_name(other))),
+    };
+    if l.3 < 8 && planar && l.2 > 1 {
+        return Err(Error::Unsupported(F, "planar TIFF with samples narrower than a byte"));
+    }
+    Ok(l)
+}
+
+/// Stored dimensions and orientation from the first IFD, without reading the image data. Refused,
+/// as a decode would refuse it: a layout or compression we can't decode, or strips/tiles that
+/// start past the end of the file (truncated). The compressed data itself is not checked.
+pub(crate) fn header(bytes: &[u8]) -> Result<(u32, u32, u16)> {
+    let mut d = Decoder::new(std::io::Cursor::new(bytes)).map_err(err)?.with_limits(limits(&DecodeOptions::default()));
+    let (w, h) = d.dimensions().map_err(err)?;
+    check_size(F, w as u64, h as u64, &DecodeOptions::default())?;
+    let ct = d.colortype().map_err(err)?;
+    let compression = d.get_tag_u32(Tag::Compression).ok();
+    let planar = d.get_tag_u32(Tag::PlanarConfiguration).ok() == Some(2);
+    layout(ct, compression, planar)?;
+    // what this build of the `tiff` crate decompresses: none, CCITT G4, LZW, JPEG, Deflate, PackBits
+    if !matches!(compression.unwrap_or(1), 1 | 4 | 5 | 7 | 8 | 32773 | 32946) {
+        return Err(Error::Unsupported(F, "compression method"));
+    }
+    let chunks = |offsets: Tag, counts: Tag, d: &mut Decoder<std::io::Cursor<&[u8]>>| {
+        Some((d.get_tag_u64_vec(offsets).ok()?, d.get_tag_u64_vec(counts).ok()?))
+    };
+    let (offsets, counts) = chunks(Tag::StripOffsets, Tag::StripByteCounts, &mut d)
+        .or_else(|| chunks(Tag::TileOffsets, Tag::TileByteCounts, &mut d))
+        .ok_or_else(|| err("no image data"))?;
+    // Only where each chunk starts: some writers overstate StripByteCounts, and the `tiff` crate
+    // decodes such files (it stops once it has the pixels), so a chunk's end proves nothing.
+    let len = bytes.len() as u64;
+    if offsets.is_empty() || offsets.iter().zip(&counts).any(|(&o, &c)| c > 0 && o >= len) {
+        return Err(err("truncated: image data past the end of the file"));
+    }
+    let orientation = d.get_tag_u32(Tag::Orientation).ok().map(|v| v as u16).filter(|o| (1..=8).contains(o)).unwrap_or(1);
+    Ok((w, h, orientation))
+}
+
 /// A BYTE/UNDEFINED tag as bytes (the decoder may surface unknown byte tags as unsigned values).
 fn tag_bytes<R: std::io::Read + std::io::Seek>(d: &mut Decoder<R>, tag: Tag) -> Option<Vec<u8>> {
    use tiff::decoder::ifd::Value;
@@ -140,6 +180,29 @@ fn unsupported_name(ct: ColorType) -> &'static str {
    }
 }

+/// Samples narrower than a byte (1-bit bilevel and fax scans, 2- and 4-bit gray) arrive packed, most
+/// significant bits first, each row padded to a whole byte. Spread them out one per byte, scaled to
+/// 0–255 like any 8-bit sample.
+fn unpack(packed: &[u8], width: usize, height: usize, samples: usize, bits: u8) -> Result<Vec<u8>> {
+    if !matches!(bits, 1 | 2 | 4) {
+        return Err(Error::Unsupported(F, "TIFF samples of 3, 5, 6 or 7 bits"));
+    }
+    let bits = bits as usize;
+    let per_row = width.checked_mul(samples).ok_or_else(|| err("row too wide"))?;
+    let row_bytes = per_row.checked_mul(bits).ok_or_else(|| err("row too wide"))?.div_ceil(8);
+    if row_bytes == 0 || packed.len() / row_bytes < height {
+        return Err(err("sample buffer too short"));
+    }
+    let max = (1u8 << bits) - 1;
+    let scale = 255 / max; // 255, 85, 17: exact
+    let mut out = Vec::with_capacity(per_row.saturating_mul(height));
+    for row in packed.chunks_exact(row_bytes).take(height) {
+        let values = row.iter().flat_map(|&b| (1..=8 / bits).map(move |i| (b >> (8 - bits * i)) & max));
+        out.extend(values.take(per_row).map(|v| v * scale));
+    }
+    Ok(out)
+}
+
 fn interleave(buf: Buf, n: usize, s: usize) -> Buf {
    fn go<T: Copy + Default>(v: Vec<T>, n: usize, s: usize) -> Vec<T> {
        if v.len() < n * s {
@@ -12,6 +12,9 @@ fn err(e: impl std::fmt::Display) -> Error {

 pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    let mut d = image_webp::WebPDecoder::new(std::io::Cursor::new(bytes)).map_err(err)?;
+    // RIFF chunk lengths are untrusted. Metadata cannot exceed the input;
+    // otherwise a tiny truncated file can request gigabytes before read_exact fails.
+    d.set_memory_limit(bytes.len());
    let (w, h) = d.dimensions();
    check_size(F, w as u64, h as u64, opts)?;
    let alpha = d.has_alpha();
@@ -24,3 +27,30 @@ pub(crate) fn decode(bytes: &[u8], opts: &DecodeOptions) -> Result<Decoded> {
    let raw = Raw { width: w as usize, height: h as usize, model: Model::Rgb, alpha, premultiplied: false, buf: Buf::U8(buf), bit_depth: 8 };
    finish(F, raw, Meta { icc, exif, xmp, ..Default::default() }, (w, h), opts)
 }
+
+/// Stored dimensions and EXIF orientation from the RIFF chunks and the bitstream header, without
+/// decoding the image. Refused, as a decode would refuse it: no image chunk (`VP8 `, `VP8L` or an
+/// animation frame), or one that runs past the end of the file (truncated). The compressed data
+/// itself is not checked.
+pub(crate) fn header(bytes: &[u8]) -> Result<(u32, u32, u16)> {
+    let mut p = 12usize;
+    loop {
+        let Some(h) = bytes.get(p..p.saturating_add(8)) else { return Err(err("no image data")) };
+        let size = u32::from_le_bytes([h[4], h[5], h[6], h[7]]) as usize;
+        let end = p.saturating_add(8).saturating_add(size);
+        if matches!(&h[0..4], b"VP8 " | b"VP8L" | b"ANMF") {
+            if end > bytes.len() {
+                return Err(err("truncated image data"));
+            }
+            break;
+        }
+        p = end.saturating_add(size & 1);
+    }
+    let mut d = image_webp::WebPDecoder::new(std::io::Cursor::new(bytes)).map_err(err)?;
+    d.set_memory_limit(bytes.len());
+    let (w, h) = d.dimensions();
+    check_size(F, w as u64, h as u64, &DecodeOptions::default())?;
+    let exif = d.exif_metadata().ok().flatten();
+    let exif = exif.as_deref().map(|e| e.strip_prefix(b"Exif\0\0").unwrap_or(e));
+    Ok((w, h, exif.map(crate::exif::summarize).unwrap_or_default().orientation.unwrap_or(1)))
+}
@@ -30,12 +30,50 @@ fn try_all(bytes: &[u8]) {
    if let Some(f) = sniff(bytes) {
        let _ = decode_unguarded(bytes, f, &small_opts());
        let _ = decode_unguarded(bytes, f, &DecodeOptions { max_size: Some((8, 8)), max_pixels: 1 << 22 });
+        let _ = read_header_unguarded(bytes, f);
    }
    let _ = decode(bytes, small_opts());
-    let _ = decode_thumbnail(bytes, 16);
+    let _ = read_header(bytes);
+    let _ = decode_thumbnail_with(bytes, &ThumbnailOptions { max_pixels: small_opts().max_pixels, ..ThumbnailOptions::new(16) });
    let _ = lightcraft_codecs::icc::parse(bytes);
 }

+#[test]
+fn webp_truncated_metadata_cannot_allocate_declared_chunk_size() {
+    let mut bytes = seeds()[5].clone();
+    let tag = bytes.windows(4).rposition(|w| w == b"XMP ").unwrap();
+    bytes[tag + 4..tag + 8].copy_from_slice(&1_946_157_060u32.to_le_bytes());
+    let decoded = decode(&bytes, small_opts()).unwrap();
+    assert_eq!((decoded.image.width, decoded.image.height), (24, 16));
+    assert!(decoded.xmp.is_none(), "invalid optional metadata is ignored");
+}
+
+#[test]
+fn thumbnail_rejects_huge_gif_before_allocating_source_pixels() {
+    // A logical screen header without a global palette; previously this reached a
+    // multi-gigabyte output allocation before discovering that no frame follows.
+    let bytes = [
+        0x47, 0x49, 0x46, 0x38, 0x39, 0x61, 0x8f, 0xa4, 0xb0, 0x3f, 0x57, 0x14, 0xe8, 0x8a, 0xf8, 0xe0, 0xc7, 0x41, 0x9d, 0x0c, 0x10, 0xde, 0x54,
+        0x24, 0xcb, 0xed, 0xff, 0x4a, 0xe7, 0xa0, 0x82, 0x10, 0xa0, 0xbc, 0xac, 0xe5, 0x33, 0x3f, 0xc9, 0x8d, 0x69, 0xf4, 0x24, 0xce, 0x99, 0x4f,
+        0x0d, 0x47, 0xa0, 0xb0, 0xf4, 0xa8, 0x59, 0xad, 0x29, 0xbd, 0x09, 0x48, 0x63, 0xea, 0x59, 0xb7, 0xb7, 0xf2, 0xa1, 0x90, 0xeb, 0x92, 0xc5,
+        0x9c, 0xde, 0xcc, 0x96, 0x67, 0xc4, 0xe5, 0x8c, 0x03, 0x50, 0xb2, 0x30, 0xf8, 0xf4, 0x69, 0x1d, 0xeb, 0xb9, 0xbd, 0x4e, 0xfb, 0xbd, 0xb6,
+        0x02,
+    ];
+    let result = decode_thumbnail(&bytes, 16);
+    assert!(matches!(result, Err(Error::TooLarge(42_127, 16_304))), "{result:?}");
+}
+
+#[test]
+fn thumbnail_respects_configured_source_pixel_limit() {
+    let image = Rgba8::from_fn(3, 2, |_, _| [40, 80, 120, 255]);
+    let bytes = encode_png(&EncodeImage::rgba8(&image), &EncodeMeta::default()).unwrap();
+    let mut opts = ThumbnailOptions::new(1);
+    opts.max_pixels = 5;
+    assert!(matches!(decode_thumbnail_with(&bytes, &opts), Err(Error::TooLarge(3, 2))));
+    opts.max_pixels = 6;
+    assert_eq!(decode_thumbnail_with(&bytes, &opts).unwrap().image.width, 1);
+}
+
 fn seeds() -> &'static Vec<Vec<u8>> {
    static S: OnceLock<Vec<Vec<u8>>> = OnceLock::new();
    S.get_or_init(|| {
@@ -0,0 +1,366 @@
+//! `read_header` (import probes) agrees with a decode on dimensions and orientation, refuses what a
+//! decode refuses (truncated files included), and returns errors — never panics — on hostile input.
+
+use lightcraft_codecs::exif::minimal_exif;
+use lightcraft_codecs::*;
+use lightcraft_raster::Rgba8;
+
+fn pixels(w: usize, h: usize) -> Rgba8 {
+    Rgba8::from_fn(w, h, |x, y| [(x * 37 % 256) as u8, (y * 53 % 256) as u8, ((x ^ y) * 9 % 256) as u8, (128 + x % 100) as u8])
+}
+
+fn meta(exif: &Option<Vec<u8>>) -> EncodeMeta<'_> {
+    EncodeMeta { exif: exif.as_deref(), xmp: Some("<x/>"), ..Default::default() }
+}
+
+/// A TIFF with an Orientation tag (`encode_tiff` writes none), via the `tiff` crate's encoder.
+fn tiff_oriented(img: &Rgba8, orientation: u16) -> Vec<u8> {
+    use tiff::encoder::{TiffEncoder, colortype};
+    let rgb: Vec<u8> = img.data.iter().flat_map(|p| [p[0], p[1], p[2]]).collect();
+    let mut cur = std::io::Cursor::new(Vec::new());
+    let mut enc = TiffEncoder::new(&mut cur).unwrap();
+    let mut im = enc.new_image::<colortype::RGB8>(img.width as u32, img.height as u32).unwrap();
+    im.encoder().write_tag(tiff::tags::Tag::Orientation, orientation).unwrap();
+    im.write_data(&rgb).unwrap();
+    cur.into_inner()
+}
+
+/// A minimal 8-bit RGB PSD (raw or PackBits-literal merged data) with an EXIF resource.
+fn psd(w: usize, h: usize, orientation: Option<u16>, rle: bool) -> Vec<u8> {
+    let mut b = Vec::new();
+    b.extend_from_slice(b"8BPS");
+    b.extend_from_slice(&1u16.to_be_bytes());
+    b.extend_from_slice(&[0; 6]);
+    b.extend_from_slice(&3u16.to_be_bytes());
+    b.extend_from_slice(&(h as u32).to_be_bytes());
+    b.extend_from_slice(&(w as u32).to_be_bytes());
+    b.extend_from_slice(&8u16.to_be_bytes());
+    b.extend_from_slice(&3u16.to_be_bytes());
+    b.extend_from_slice(&0u32.to_be_bytes()); // colour mode data
+    let mut res = Vec::new();
+    if let Some(o) = orientation {
+        let exif = minimal_exif(o);
+        res.extend_from_slice(b"8BIM");
+        res.extend_from_slice(&1058u16.to_be_bytes());
+        res.extend_from_slice(&[0, 0]); // empty pascal name, padded
+        res.extend_from_slice(&(exif.len() as u32).to_be_bytes());
+        res.extend_from_slice(&exif);
+        if exif.len() % 2 == 1 {
+            res.push(0);
+        }
+    }
+    b.extend_from_slice(&(res.len() as u32).to_be_bytes());
+    b.extend_from_slice(&res);
+    b.extend_from_slice(&0u32.to_be_bytes()); // layer and mask info
+    b.extend_from_slice(&(rle as u16).to_be_bytes());
+    let plane = |c: usize| (0..w * h).map(move |i| ((i * (c + 3) * 11) % 256) as u8);
+    if rle {
+        // every row as literal runs of at most 128 bytes (PackBits' longest)
+        let row_len = w + w.div_ceil(128);
+        for _ in 0..3 * h {
+            b.extend_from_slice(&(row_len as u16).to_be_bytes());
+        }
+        for c in 0..3 {
+            let p: Vec<u8> = plane(c).collect();
+            for row in p.chunks(w) {
+                for run in row.chunks(128) {
+                    b.push(run.len() as u8 - 1);
+                    b.extend_from_slice(run);
+                }
+            }
+        }
+    } else {
+        for c in 0..3 {
+            b.extend(plane(c));
+        }
+    }
+    b
+}
+
+/// Generated files of every format with a header reader, plus a fallback one (BMP via `image`).
+fn fixtures(w: usize, h: usize) -> Vec<(String, Vec<u8>)> {
+    let img = pixels(w, h);
+    let e = EncodeImage::rgba8(&img);
+    let rgb: Vec<u8> = img.data.iter().flat_map(|p| [p[0], p[1], p[2]]).collect();
+    let rgb_e = EncodeImage::new(w as u32, h as u32, 3, Samples::U8(&rgb));
+    let u16s: Vec<u16> = (0..w * h * 3).map(|i| (i * 331) as u16).collect();
+    let mut out = Vec::new();
+    for o in [None, Some(1), Some(6), Some(8)] {
+        let exif = o.map(minimal_exif);
+        let tag = format!("o{}", o.unwrap_or(0));
+        out.push((format!("jpeg 420 {tag}"), encode_jpeg(&rgb_e, 85, ChromaSubsampling::S420, &meta(&exif)).unwrap()));
+        out.push((format!("jpeg 444 {tag}"), encode_jpeg(&rgb_e, 85, ChromaSubsampling::S444, &meta(&exif)).unwrap()));
+        out.push((format!("png rgba {tag}"), encode_png(&e, &meta(&exif)).unwrap()));
+        out.push((format!("png 16-bit {tag}"), encode_png(&EncodeImage::new(w as u32, h as u32, 3, Samples::U16(&u16s)), &meta(&exif)).unwrap()));
+        out.push((format!("webp {tag}"), encode_webp_lossless(&e, &meta(&exif)).unwrap()));
+        out.push((format!("psd {tag}"), psd(w, h, o, false)));
+        out.push((format!("psd rle {tag}"), psd(w, h, o, true)));
+        out.push((format!("tiff {tag}"), tiff_oriented(&img, o.unwrap_or(1))));
+    }
+    out.push(("tiff lzw".into(), encode_tiff(&e, TiffCompression::Lzw, &EncodeMeta::default()).unwrap()));
+    out.push(("tiff packbits".into(), encode_tiff(&e, TiffCompression::PackBits, &EncodeMeta::default()).unwrap()));
+    let mut bmp = Vec::new();
+    image::codecs::bmp::BmpEncoder::new(&mut bmp).encode(&rgb, w as u32, h as u32, image::ExtendedColorType::Rgb8).unwrap();
+    out.push(("bmp".into(), bmp));
+    out
+}
+
+/// What the import probe learnt before it read headers: a small decode.
+fn decoded(bytes: &[u8]) -> Result<Header> {
+    let d = decode(bytes, DecodeOptions::fit(64, 64))?;
+    Ok(Header { format: d.format, width: d.source_width, height: d.source_height, orientation: d.orientation })
+}
+
+#[test]
+fn header_matches_decode() {
+    for (w, h) in [(160, 96), (96, 160), (7, 5)] {
+        for (name, bytes) in fixtures(w, h) {
+            let header = read_header(&bytes).unwrap_or_else(|e| panic!("{name} {w}×{h}: {e}"));
+            assert_eq!(header, decoded(&bytes).unwrap(), "{name} {w}×{h}");
+            assert_eq!((header.width, header.height), (w as u32, h as u32), "{name}");
+        }
+    }
+    let o = |name: &str| fixtures(16, 8).into_iter().find(|f| f.0 == name).map(|f| read_header(&f.1).unwrap().orientation).unwrap();
+    assert_eq!((o("jpeg 420 o6"), o("png rgba o8"), o("webp o6"), o("psd o8"), o("tiff o6"), o("jpeg 444 o0")), (6, 8, 6, 8, 6, 1));
+}
+
+/// Whenever the decode the probe used to run accepts a file (here: every cut of every fixture),
+/// the header agrees with it: reading headers never refuses a file that decodes. The reverse
+/// is checked for truncation, which the header readers detect.
+#[test]
+fn truncated_files_agree_with_decode() {
+    // 160 px wide: large JPEGs take the DCT-scaled (strict) decode path the probe took for photos
+    for (name, bytes) in fixtures(160, 96) {
+        let n = bytes.len();
+        let mut cuts: Vec<usize> = (0..100).map(|i| n * i / 100).collect();
+        cuts.extend(n.saturating_sub(64)..n);
+        for cut in cuts {
+            let part = &bytes[..cut];
+            if let Ok(d) = decoded(part) {
+                assert_eq!(read_header(part).ok(), Some(d), "{name} cut at {cut}/{n}");
+            }
+        }
+        for cut in [n / 4, n / 2, n * 9 / 10] {
+            assert!(read_header(&bytes[..cut]).is_err(), "{name}: cut at {cut}/{n} accepted");
+        }
+    }
+}
+
+#[test]
+fn truncation_details() {
+    let f = fixtures(160, 96);
+    let get = |name: &str| f.iter().find(|x| x.0 == name).map(|x| x.1.clone()).unwrap();
+    let jpeg = get("jpeg 420 o6");
+    assert!(read_header(&jpeg[..jpeg.len() - 2]).is_err(), "a JPEG without its EOI is truncated");
+    let mut trailing = jpeg.clone();
+    trailing.extend_from_slice(&[0; 100]);
+    assert_eq!(read_header(&trailing).unwrap().orientation, 6, "bytes after the EOI are fine");
+    let png = get("png rgba o6");
+    assert!(read_header(&png[..png.len() - 2]).is_ok(), "a damaged IEND is ignored, as by a decode");
+    assert!(read_header(&png[..png.len() - 12]).is_err(), "nothing after the image data: truncated");
+}
+
+/// eXIf after the image data counts, as it does for a decode (which reads the whole file).
+#[test]
+fn png_exif_after_idat() {
+    let img = pixels(9, 4);
+    let plain = encode_png(&EncodeImage::rgba8(&img), &EncodeMeta::default()).unwrap();
+    let exif = minimal_exif(6);
+    let mut chunk = (exif.len() as u32).to_be_bytes().to_vec();
+    chunk.extend_from_slice(b"eXIf");
+    chunk.extend_from_slice(&exif);
+    chunk.extend_from_slice(&[0; 4]); // CRC: not checked by either reader
+    let iend = plain.len() - 12;
+    let mut b = plain[..iend].to_vec();
+    b.extend_from_slice(&chunk);
+    b.extend_from_slice(&plain[iend..]);
+    assert_eq!(read_header(&b).unwrap().orientation, 6);
+}
+
+fn jpeg_segments(sof: u8, w: u16, h: u16, comps: u8) -> Vec<u8> {
+    let mut b = vec![0xFF, 0xD8, 0xFF, sof, 0, 8 + 3 * comps as u16 as u8, 8];
+    b.extend_from_slice(&h.to_be_bytes());
+    b.extend_from_slice(&w.to_be_bytes());
+    b.push(comps);
+    for c in 0..comps {
+        b.extend_from_slice(&[c + 1, 0x11, 0]);
+    }
+    b.extend_from_slice(&[0xFF, 0xDA, 0, 8, 1, 1, 0, 0, 63, 0]);
+    b.extend_from_slice(&[0x12, 0xFF, 0x00, 0x34, 0xFF, 0xD0, 0x56]);
+    b.extend_from_slice(&[0xFF, 0xD9]);
+    b
+}
+
+fn png_ihdr(w: u32, h: u32, depth: u8, color: u8) -> Vec<u8> {
+    let mut b = b"\x89PNG\r\n\x1a\n\0\0\0\x0dIHDR".to_vec();
+    b.extend_from_slice(&w.to_be_bytes());
+    b.extend_from_slice(&h.to_be_bytes());
+    b.extend_from_slice(&[depth, color, 0, 0, 0, 0, 0, 0, 0]);
+    b
+}
+
+fn png_chunk(b: &mut Vec<u8>, kind: &[u8; 4], len: u32, data: &[u8]) {
+    b.extend_from_slice(&len.to_be_bytes());
+    b.extend_from_slice(kind);
+    b.extend_from_slice(data);
+    b.extend_from_slice(&[0; 4]);
+}
+
+#[test]
+fn hostile_headers_are_errors() {
+    // a well-formed synthetic JPEG header is read
+    assert_eq!(read_header(&jpeg_segments(0xC0, 33, 21, 3)).unwrap(), Header { format: Format::Jpeg, width: 33, height: 21, orientation: 1 });
+    let jpeg_bad = [
+        jpeg_segments(0xC0, 0, 21, 3),        // zero width
+        jpeg_segments(0xC0, 33, 0, 3),        // zero height (DNL)
+        jpeg_segments(0xC0, 65535, 65535, 3), // 4.3 Gpx
+        jpeg_segments(0xC9, 33, 21, 3),       // arithmetic coding
+        jpeg_segments(0xC5, 33, 21, 3),       // hierarchical
+        jpeg_segments(0xC0, 33, 21, 2),       // two components
+        jpeg_segments(0xC0, 33, 21, 3)[..30].to_vec(),
+        vec![0xFF, 0xD8, 0xFF, 0xE1, 0, 0],       // segment length 0
+        vec![0xFF, 0xD8, 0xFF, 0xE1, 0xFF, 0xFF], // segment past the end
+        vec![0xFF, 0xD8, 0xFF, 0xD9],
+    ];
+    for (i, b) in jpeg_bad.iter().enumerate() {
+        assert!(read_header(b).is_err(), "JPEG case {i}");
+    }
+    let mut no_sos = jpeg_segments(0xC0, 33, 21, 3);
+    no_sos.truncate(2 + 2 + 17);
+    no_sos.extend_from_slice(&[0xFF, 0xD9]);
+    assert!(read_header(&no_sos).is_err(), "JPEG without a scan");
+
+    let mut ok = png_ihdr(5, 3, 8, 2);
+    png_chunk(&mut ok, b"IDAT", 1, &[0]);
+    png_chunk(&mut ok, b"IEND", 0, &[]);
+    assert_eq!(read_header(&ok).unwrap(), Header { format: Format::Png, width: 5, height: 3, orientation: 1 });
+    let png_case = |ihdr: Vec<u8>, chunks: &[(&[u8; 4], u32, &[u8])]| {
+        let mut b = ihdr;
+        for (k, l, d) in chunks {
+            png_chunk(&mut b, k, *l, d);
+        }
+        b
+    };
+    let idat: (&[u8; 4], u32, &[u8]) = (b"IDAT", 1, &[0]);
+    let iend: (&[u8; 4], u32, &[u8]) = (b"IEND", 0, &[]);
+    let png_bad = [
+        png_case(png_ihdr(0, 3, 8, 2), &[idat, iend]),
+        png_case(png_ihdr(5, 0, 8, 2), &[idat, iend]),
+        png_case(png_ihdr(0x8000_0000, 1, 8, 2), &[idat, iend]),
+        png_case(png_ihdr(65536, 65536, 8, 2), &[idat, iend]),
+        png_case(png_ihdr(5, 3, 4, 2), &[idat, iend]), // RGB at 4 bits
+        png_case(png_ihdr(5, 3, 8, 5), &[idat, iend]), // no colour type 5
+        png_case(png_ihdr(5, 3, 8, 3), &[idat, iend]), // indexed without PLTE
+        png_case(png_ihdr(5, 3, 8, 2), &[iend]),       // no IDAT
+        png_case(png_ihdr(5, 3, 8, 2), &[idat]),       // nothing after the image data
+        png_case(png_ihdr(5, 3, 8, 2), &[(b"IDAT", 0xFFFF_FFFF, &[0]), iend]),
+        png_case(png_ihdr(5, 3, 8, 2), &[(b"tEXt", 0x7FFF_FFFF, &[0]), idat, iend]),
+        png_ihdr(5, 3, 8, 2)[..20].to_vec(),
+        b"\x89PNG\r\n\x1a\n\0\0\0\x0dIHDX\0\0\0\x05\0\0\0\x03\x08\x02\0\0\0".to_vec(),
+    ];
+    for (i, b) in png_bad.iter().enumerate() {
+        assert!(read_header(b).is_err(), "PNG case {i}");
+    }
+
+    let tiff = encode_tiff(&EncodeImage::rgba8(&pixels(6, 4)), TiffCompression::None, &EncodeMeta::default()).unwrap();
+    let ifd = u32::from_le_bytes([tiff[4], tiff[5], tiff[6], tiff[7]]) as usize;
+    let mut tiff_bad = vec![b"II*\0\xff\xff\xff\x7f".to_vec(), b"II*\0\x08\0\0\0".to_vec(), b"II*\0\x04\0\0\0".to_vec()];
+    // cut short after the IFD
+    assert!(ifd < tiff.len() / 2, "the IFD comes before the image data");
+    tiff_bad.push(tiff[..tiff.len() - 10].to_vec());
+    for (i, b) in tiff_bad.iter().enumerate() {
+        assert!(read_header(b).is_err(), "TIFF case {i}");
+    }
+    // a strip offset pointing past the end of the file
+    let t = |b: &[u8], p: usize| u16::from_le_bytes([b[p], b[p + 1]]);
+    let n = t(&tiff, ifd) as usize;
+    let mut far = tiff.clone();
+    for e in 0..n {
+        let p = ifd + 2 + e * 12;
+        if t(&far, p) == 273 && t(&far, p + 2) == 4 && u32::from_le_bytes([far[p + 4], far[p + 5], far[p + 6], far[p + 7]]) == 1 {
+            far[p + 8..p + 12].copy_from_slice(&0x7FFF_0000u32.to_le_bytes());
+        }
+    }
+    assert_ne!(far, tiff, "the strip offset was found");
+    assert!(read_header(&far).is_err(), "strip past the end");
+
+    let webp_bad = [
+        b"RIFF\0\0\0\0WEBP".to_vec(),
+        b"RIFF\0\0\0\0WEBPVP8L\xff\xff\xff\xff\x2f".to_vec(),
+        b"RIFF\0\0\0\0WEBPXXXX\xff\xff\xff\xff".to_vec(),
+        b"RIFF\0\0\0\0WEBPVP8 \x04\0\0\0\0\0\0\0".to_vec(),
+    ];
+    for (i, b) in webp_bad.iter().enumerate() {
+        assert!(read_header(b).is_err(), "WebP case {i}");
+    }
+
+    let good = psd(4, 3, Some(6), true);
+    assert_eq!(read_header(&good).unwrap().orientation, 6);
+    let set = |pos: usize, v: &[u8]| {
+        let mut b = good.clone();
+        b[pos..pos + v.len()].copy_from_slice(v);
+        b
+    };
+    let res_len = u32::from_be_bytes([good[30], good[31], good[32], good[33]]) as usize;
+    let compression = 34 + res_len + 4;
+    let mut psd_bad = vec![
+        set(12, &0u16.to_be_bytes()),           // no channels
+        set(12, &57u16.to_be_bytes()),          // too many
+        set(12, &2u16.to_be_bytes()),           // too few for RGB
+        set(14, &0x7FFF_FFFFu32.to_be_bytes()), // absurd height
+        set(18, &0u32.to_be_bytes()),           // zero width
+        set(22, &1u16.to_be_bytes()),           // 1-bit
+        set(22, &7u16.to_be_bytes()),           // bad depth
+        set(24, &9u16.to_be_bytes()),           // Lab
+        set(24, &77u16.to_be_bytes()),          // unknown mode
+        set(24, &2u16.to_be_bytes()),           // indexed without a palette
+        set(compression, &2u16.to_be_bytes()),  // ZIP
+        set(26, &0xFFFF_FFFFu32.to_be_bytes()), // colour mode data past the end
+        set(30, &0xFFFF_FFF0u32.to_be_bytes()), // resources past the end
+    ];
+    // a 1 × 2^30 RLE image declared in a tiny file: refused without reading 2^30 row counts
+    let mut tall = good.clone();
+    tall[14..18].copy_from_slice(&(1u32 << 30).to_be_bytes());
+    tall[18..22].copy_from_slice(&1u32.to_be_bytes());
+    psd_bad.push(tall);
+    psd_bad.push(good[..good.len() - 1].to_vec());
+    for (i, b) in psd_bad.iter().enumerate() {
+        assert!(read_header(b).is_err(), "PSD case {i}");
+    }
+
+    for b in [&b""[..], b"\xFF\xD8\xFF", b"\x89PNG\r\n\x1a\n", b"II*\0", b"MM\0*", b"8BPS\0\x01", b"GIF89a", b"not an image at all"] {
+        assert!(read_header(b).is_err(), "{b:?}");
+    }
+}
+
+/// The same with a few random bytes changed: whenever the decode accepts the damaged file, the
+/// header agrees with it (deterministic seed).
+#[test]
+fn mutated_files_agree_with_decode() {
+    let mut s: u64 = 0x1234_5678_9abc_def1;
+    let mut rnd = || {
+        s ^= s << 13;
+        s ^= s >> 7;
+        s ^= s << 17;
+        s
+    };
+    let (mut decodable, mut disagreements) = (0, Vec::new());
+    for (name, bytes) in fixtures(160, 96) {
+        for _ in 0..300 {
+            let mut b = bytes.clone();
+            for _ in 0..1 + rnd() % 3 {
+                let i = rnd() as usize % b.len();
+                b[i] = rnd() as u8;
+            }
+            if let Ok(d) = decoded(&b) {
+                decodable += 1;
+                let h = read_header(&b);
+                if h.as_ref().ok() != Some(&d) {
+                    disagreements.push(format!("{name}: decode {d:?}, header {h:?}"));
+                }
+            }
+        }
+    }
+    assert!(decodable > 1000, "{decodable}");
+    assert!(disagreements.is_empty(), "{disagreements:#?}");
+}
@@ -245,6 +245,80 @@ fn tiff_cmyk_and_orientation() {
    assert!(red[0] > 0.99 && red[1] < 0.01 && red[2] < 0.01, "{red:?}");
 }

+/// A one-strip, uncompressed, single-sample TIFF (little-endian) whose samples are `bits` wide and
+/// already packed into `strip`, one padded row after another.
+fn packed_tiff(width: u16, height: u16, bits: u16, photometric: u16, strip: &[u8]) -> Vec<u8> {
+    const STRIP_AT: u32 = 8 + 2 + 9 * 12 + 4;
+    let short = |tag: u16, v: u16| [&tag.to_le_bytes()[..], &3u16.to_le_bytes(), &1u32.to_le_bytes(), &v.to_le_bytes(), &[0, 0]].concat();
+    let long = |tag: u16, v: u32| [&tag.to_le_bytes()[..], &4u16.to_le_bytes(), &1u32.to_le_bytes(), &v.to_le_bytes()].concat();
+    let mut b = b"II*\0".to_vec();
+    b.extend(8u32.to_le_bytes());
+    b.extend(9u16.to_le_bytes());
+    b.extend(short(256, width));
+    b.extend(short(257, height));
+    b.extend(short(258, bits));
+    b.extend(short(259, 1));
+    b.extend(short(262, photometric));
+    b.extend(long(273, STRIP_AT));
+    b.extend(short(277, 1));
+    b.extend(short(278, height));
+    b.extend(long(279, strip.len() as u32));
+    b.extend(0u32.to_le_bytes());
+    assert_eq!(b.len(), STRIP_AT as usize);
+    b.extend_from_slice(strip);
+    b
+}
+
+/// WhiteIsZero (photometric 0) gray: the lowest sample is white. The `tiff` crate already inverts these
+/// samples, and a second inversion here used to show every WhiteIsZero scan as a negative.
+#[test]
+fn tiff_white_is_zero_is_inverted_once() {
+    // Two pixels: the lowest sample, then the highest, at 8 and 16 bits.
+    for (bits, strip) in [(8, vec![0u8, 255]), (16, [0u16.to_le_bytes(), 65535u16.to_le_bytes()].concat())] {
+        for (photometric, first) in [(0, 1.0), (1, 0.0)] {
+            let d = decode(&packed_tiff(2, 1, bits, photometric, &strip), DecodeOptions::default()).unwrap();
+            let (a, b) = (d.image.get(0, 0)[1], d.image.get(1, 0)[1]);
+            assert!((a - first).abs() < 1e-6 && (b - (1.0 - first)).abs() < 1e-6, "{bits}-bit, photometric {photometric}: {a} {b}");
+        }
+    }
+}
+
+/// 1-, 2- and 4-bit samples (bilevel scans, fax images, low-depth gray) arrive packed eight, four or
+/// two to a byte with each row padded to a byte; they used to be rejected as "sample buffer too short".
+#[test]
+fn tiff_samples_narrower_than_a_byte() {
+    let gray = |d: &Decoded, x: usize, y: usize| d.image.get(x, y)[1];
+
+    // 10 × 2 bilevel: rows are 2 bytes, the last 6 bits of each are padding (set, to prove they are ignored).
+    // Row 0 alternates set/clear bits; row 1 is clear but for its last pixel.
+    let rows = [0b1010_1010, 0b1011_1111, 0b0000_0000, 0b0111_1111];
+    // What a set bit decodes to: white for BlackIsZero (1), black for WhiteIsZero (0).
+    for (photometric, set_bit) in [(1, 1.0), (0, 0.0)] {
+        let d = decode(&packed_tiff(10, 2, 1, photometric, &rows), DecodeOptions::default()).unwrap();
+        assert_eq!((d.image.width, d.image.height, d.bit_depth), (10, 2, 1));
+        assert!(d.grayscale);
+        let clear_bit = 1.0 - set_bit;
+        for x in 0..10 {
+            let want = (if x % 2 == 0 { set_bit } else { clear_bit }, if x == 9 { set_bit } else { clear_bit });
+            assert!((gray(&d, x, 0) - want.0).abs() < 1e-6, "row 0, x {x}, photometric {photometric}");
+            assert!((gray(&d, x, 1) - want.1).abs() < 1e-6, "row 1, x {x}, photometric {photometric}");
+        }
+    }
+
+    // 2-bit: 0, 1, 2, 3 → 0, 85, 170, 255. 4-bit: 0, 15, 8 → 0, 255, 136 (the row's last nibble is padding).
+    let d = decode(&packed_tiff(4, 1, 2, 1, &[0b00_01_10_11]), DecodeOptions::default()).unwrap();
+    for (x, v) in [0, 85, 170, 255].into_iter().enumerate() {
+        assert!((gray(&d, x, 0) - srgb_to_linear(v as f32 / 255.0)).abs() < 1e-6, "2-bit x {x}");
+    }
+    let d = decode(&packed_tiff(3, 1, 4, 1, &[0x0F, 0x8F]), DecodeOptions::default()).unwrap();
+    for (x, v) in [0, 255, 136].into_iter().enumerate() {
+        assert!((gray(&d, x, 0) - srgb_to_linear(v as f32 / 255.0)).abs() < 1e-6, "4-bit x {x}");
+    }
+
+    // A strip shorter than its rows is an error, never a panic.
+    assert!(decode(&packed_tiff(10, 2, 1, 1, &rows[..3]), DecodeOptions::default()).is_err());
+}
+
 #[test]
 fn webp_lossless_exact() {
    let img = gradient(31, 19);
@@ -461,7 +535,7 @@ fn thumbnail_accepts_smaller_preview_when_asked() {
        encode_jpeg(&EncodeImage::rgba8(&main), 80, ChromaSubsampling::S420, &EncodeMeta { exif: Some(&exif), ..Default::default() }).unwrap();
    let t = decode_thumbnail(&bytes, 256).unwrap();
    assert_eq!(t.source, ThumbnailSource::Scaled);
-    let t = decode_thumbnail_with(&bytes, &ThumbnailOptions { max_edge: 256, min_embedded_edge: 160 }).unwrap();
+    let t = decode_thumbnail_with(&bytes, &ThumbnailOptions { min_embedded_edge: 160, ..ThumbnailOptions::new(256) }).unwrap();
    assert_eq!(t.source, ThumbnailSource::ExifThumbnail);
    assert_eq!((t.image.width, t.image.height), (160, 120));
 }
@@ -2,7 +2,9 @@
 //!
 //! The Planckian locus uses the Kim et al. (2002) cubic approximation (1667–25000 K); outside that
 //! range it is extended linearly in mired space. Tint is a signed offset perpendicular to the locus in
-//! CIE 1960 (u, v): **positive tint = magenta** (below the locus), `Duv = −tint / TINT_SCALE`.
+//! CIE 1960 (u, v). Tint describes the **correction**: positive adds magenta, negative adds green.
+//! A positive correction neutralises a green illuminant above the locus, so its source white uses
+//! `Duv = tint / TINT_SCALE`. Bradford adaptation removes that source cast.

 use crate::{Mat3, RgbSpace, Xy, bradford};

@@ -65,11 +67,12 @@ fn locus_normal(t: f64) -> (f64, f64) {
    if ny < 0.0 { (-nx, -ny) } else { (nx, ny) }
 }

-/// Chromaticity of the white point for temperature `t` (K) and `tint`.
+/// Source illuminant chromaticity for temperature `t` (K) and correction `tint`.
+/// Positive tint means a green source white, whose correction adds magenta to the image.
 pub fn temp_tint_to_xy(t: f64, tint: f64) -> Xy {
    let (u, v) = locus_uv(t);
    let (nu, nv) = locus_normal(t);
-    let duv = -tint / TINT_SCALE;
+    let duv = tint / TINT_SCALE;
    uv_to_xy(u + nu * duv, v + nv * duv)
 }

@@ -104,7 +107,7 @@ pub fn xy_to_temp_tint(p: Xy) -> (f64, f64) {
    let (lu, lv) = locus_uv(t);
    let (nu, nv) = locus_normal(t);
    let duv = (u - lu) * nu + (v - lv) * nv;
-    (t, -duv * TINT_SCALE)
+    (t, duv * TINT_SCALE)
 }

 /// Matrix (linear RGB in `space` → same space) that white-balances a scene lit by `src` so that it
@@ -141,10 +144,18 @@ mod tests {
        let warm = temp_tint_to_xy(3000.0, 0.0);
        let cool = temp_tint_to_xy(9000.0, 0.0);
        assert!(warm.x > cool.x);
-        // positive tint is magenta: lower v
+        // A positive magenta correction specifies a greener source illuminant: higher v.
        let (_, v0) = xy_to_uv(temp_tint_to_xy(5000.0, 0.0));
        let (_, v1) = xy_to_uv(temp_tint_to_xy(5000.0, 50.0));
-        assert!(v1 < v0);
+        assert!(v1 > v0);
+    }
+
+    /// Daylight lies above the Planckian locus, so D65 reads as a small positive tint, as
+    /// Lightroom's Daylight preset (5500 K, +10) has it (#188).
+    #[test]
+    fn daylight_has_a_positive_tint() {
+        let (_, tint) = xy_to_temp_tint(crate::D65);
+        assert!((5.0..15.0).contains(&tint), "{tint}");
    }

    #[test]
@@ -9,6 +9,7 @@
 //!
 //! The pipeline's working space is **linear Rec.2020, D65** ([`WORKING`]).
 #![forbid(unsafe_code)]
+#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)]

 pub mod cct;
 pub mod perceptual;
@@ -125,14 +126,17 @@ pub const WORKING: RgbSpace = REC2020;

 impl RgbSpace {
    /// Matrix taking linear RGB in this space to XYZ (white Y = 1) relative to its own white.
+    /// Degenerate (collinear) primaries, which readers reject before building a space from a
+    /// file, leave the primaries unscaled instead of panicking.
    pub fn to_xyz(&self) -> Mat3 {
        let p = [self.r.to_xyz(), self.g.to_xyz(), self.b.to_xyz()];
        let m = Mat3([[p[0][0], p[1][0], p[2][0]], [p[0][1], p[1][1], p[2][1]], [p[0][2], p[1][2], p[2][2]]]);
-        let s = m.inverse().expect("primaries are independent").apply(self.white.to_xyz());
+        let s = m.inverse().map_or([1.0; 3], |inv| inv.apply(self.white.to_xyz()));
        m.mul(&Mat3::diag(s[0], s[1], s[2]))
    }
+    /// Inverse of [`to_xyz`](Self::to_xyz); the identity for degenerate primaries.
    pub fn from_xyz(&self) -> Mat3 {
-        self.to_xyz().inverse().expect("invertible")
+        self.to_xyz().inverse().unwrap_or(Mat3::IDENTITY)
    }
    /// Luminance weights (the Y row of `to_xyz`).
    pub fn luma(&self) -> [f64; 3] {
@@ -154,7 +158,8 @@ pub fn bradford(src: Xy, dst: Xy) -> Mat3 {
    }
    let s = BRADFORD.apply(src.to_xyz());
    let d = BRADFORD.apply(dst.to_xyz());
-    let inv = BRADFORD.inverse().expect("invertible");
+    // a constant, invertible matrix: the fallback (no adaptation) can't happen
+    let Some(inv) = BRADFORD.inverse() else { return Mat3::IDENTITY };
    inv.mul(&Mat3::diag(d[0] / s[0], d[1] / s[1], d[2] / s[2])).mul(&BRADFORD)
 }

@@ -28,9 +28,10 @@ fn mats() -> &'static OkMats {
        let to_lms = XYZ_TO_LMS.mul(&REC2020.to_xyz());
        OkMats {
            to_lms: to_lms.to_f32(),
-            from_lms: to_lms.inverse().expect("invertible").to_f32(),
+            // constant, invertible matrices (round trips tested below): the identity fallback can't happen
+            from_lms: to_lms.inverse().unwrap_or(Mat3::IDENTITY).to_f32(),
            to_lab: LMS_TO_LAB.to_f32(),
-            from_lab: LMS_TO_LAB.inverse().expect("invertible").to_f32(),
+            from_lab: LMS_TO_LAB.inverse().unwrap_or(Mat3::IDENTITY).to_f32(),
        }
    })
 }
@@ -0,0 +1,26 @@
+[package]
+name = "lightcraft-denoise-core"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
+description = "LightCraft denoise model metadata, Bayer tiling and cache formats (L1; no inference backend)"
+
+[lints]
+workspace = true
+
+[features]
+default = []
+rawnind-model = []
+
+[dependencies]
+lightcraft-raster = { workspace = true }
+serde = { workspace = true }
+serde_json = { workspace = true }
+sha2 = { workspace = true }
+thiserror = { workspace = true }
+miniz_oxide = { workspace = true }
+rayon = { workspace = true }
+web-time = { workspace = true }
+
+[dev-dependencies]
+proptest = { workspace = true }
@@ -0,0 +1,7 @@
+# lightcraft-denoise-core
+
+L1 model metadata, bounded archive/cache formats and Bayer packing/tiling helpers.
+No weights or inference backend. See [the user workflow](../../docs/denoise.md).
+`lightcraft-denoise` re-exports these types for CPU/GPU users; the engine uses this
+crate directly when inference is disabled. The default download catalog is empty.
+Run `cargo test -p lightcraft-denoise-core`; optional catalog checks use `--features rawnind-model`.
@@ -0,0 +1,279 @@
+//! Taking one file out of a `.zip` (a `.dtmodel` is one), the little of the format a model download needs: stored and
+//! deflated entries, found by name, checked against their CRC-32 and a size limit, written next to where they are
+//! going and moved into place. It is a stranger's file, so every length and offset in it is checked, a name is only
+//! ever compared (never used as a path), and zip64, encrypted and multi-disk archives are refused.
+
+use std::io::{Read, Seek, SeekFrom, Write};
+use std::path::Path;
+
+/// How far from the end the end-of-central-directory record can start (a 22-byte record and a 64 KiB comment).
+const EOCD_SEARCH: u64 = 22 + 65535;
+/// Largest central directory we read.
+const MAX_DIRECTORY: u64 = 1 << 20;
+
+#[derive(Debug, PartialEq, thiserror::Error)]
+pub enum ArchiveError {
+    #[error("the archive could not be read: {0}")]
+    Io(String),
+    #[error("not a usable archive: {0}")]
+    Format(String),
+    #[error("the archive has no file called `{0}`")]
+    Missing(String),
+}
+
+impl From<std::io::Error> for ArchiveError {
+    fn from(e: std::io::Error) -> Self {
+        ArchiveError::Io(e.to_string())
+    }
+}
+
+fn format<T>(why: &str) -> Result<T, ArchiveError> {
+    Err(ArchiveError::Format(why.into()))
+}
+
+fn u16_at(b: &[u8], at: usize) -> Option<u16> {
+    Some(u16::from_le_bytes(b.get(at..at.checked_add(2)?)?.try_into().ok()?))
+}
+
+fn u32_at(b: &[u8], at: usize) -> Option<u32> {
+    Some(u32::from_le_bytes(b.get(at..at.checked_add(4)?)?.try_into().ok()?))
+}
+
+/// CRC-32 (IEEE) of `bytes`.
+fn crc32(bytes: &[u8]) -> u32 {
+    let mut table = [0u32; 256];
+    for (i, t) in table.iter_mut().enumerate() {
+        let mut c = i as u32;
+        for _ in 0..8 {
+            c = if c & 1 != 0 { 0xedb8_8320 ^ (c >> 1) } else { c >> 1 };
+        }
+        *t = c;
+    }
+    !bytes.iter().fold(!0u32, |c, &b| table[((c ^ u32::from(b)) & 0xff) as usize] ^ (c >> 8))
+}
+
+struct Entry {
+    method: u16,
+    crc: u32,
+    compressed: u64,
+    size: u64,
+    local_header: u64,
+}
+
+/// Find the entry called `name` (exactly, or as the last part of a longer name: `folder/name`) in the central directory.
+fn find(f: &mut (impl Read + Seek), name: &str) -> Result<Entry, ArchiveError> {
+    let len = f.seek(SeekFrom::End(0))?;
+    let tail_len = len.min(EOCD_SEARCH);
+    f.seek(SeekFrom::Start(len - tail_len))?;
+    let mut tail = vec![0u8; tail_len as usize];
+    f.read_exact(&mut tail)?;
+    let eocd = (0..tail.len().saturating_sub(21)).rev().find(|&i| u32_at(&tail, i) == Some(0x0605_4b50));
+    let Some(e) = eocd else { return format("this is not a zip file") };
+    let (entries, size, offset) = (u16_at(&tail, e + 10), u32_at(&tail, e + 12), u32_at(&tail, e + 16));
+    let (Some(entries), Some(size), Some(offset)) = (entries, size, offset) else { return format("the end of the archive is cut short") };
+    if entries == 0xffff || size == u32::MAX || offset == u32::MAX {
+        return format("zip64 archives are not supported");
+    }
+    if u64::from(size) > MAX_DIRECTORY || u64::from(offset) + u64::from(size) > len {
+        return format("the list of files is not where it says");
+    }
+    f.seek(SeekFrom::Start(u64::from(offset)))?;
+    let mut dir = vec![0u8; size as usize];
+    f.read_exact(&mut dir)?;
+    let mut at = 0usize;
+    for _ in 0..entries {
+        if u32_at(&dir, at) != Some(0x0201_4b50) {
+            return format("the list of files is damaged");
+        }
+        let get = |o: usize| at.checked_add(o);
+        let (flags, method, crc) = (u16_at(&dir, at + 8), u16_at(&dir, at + 10), u32_at(&dir, at + 16));
+        let (compressed, uncompressed) = (u32_at(&dir, at + 20), u32_at(&dir, at + 24));
+        let (n, x, c) = (u16_at(&dir, at + 28), u16_at(&dir, at + 30), u16_at(&dir, at + 32));
+        let local = u32_at(&dir, at + 42);
+        let (Some(flags), Some(method), Some(crc), Some(compressed), Some(uncompressed), Some(n), Some(x), Some(c), Some(local), Some(start)) =
+            (flags, method, crc, compressed, uncompressed, n, x, c, local, get(46))
+        else {
+            return format("the list of files is cut short");
+        };
+        let Some(entry_name) = dir.get(start..start + usize::from(n)) else { return format("a file name is cut short") };
+        let matches = entry_name == name.as_bytes() || entry_name.strip_suffix(name.as_bytes()).is_some_and(|p| p.ends_with(b"/"));
+        if matches {
+            if flags & 1 != 0 {
+                return format("encrypted files are not supported");
+            }
+            if compressed == u32::MAX || uncompressed == u32::MAX {
+                return format("zip64 files are not supported");
+            }
+            return Ok(Entry { method, crc, compressed: u64::from(compressed), size: u64::from(uncompressed), local_header: u64::from(local) });
+        }
+        at = start + usize::from(n) + usize::from(x) + usize::from(c);
+    }
+    Err(ArchiveError::Missing(name.into()))
+}
+
+/// Take the file called `name` out of the archive at `archive` and write it to `dest` (through a temporary file next
+/// to it); `max_bytes` bounds its size. Returns the number of bytes written.
+pub fn extract(archive: &Path, name: &str, dest: &Path, max_bytes: u64) -> Result<u64, ArchiveError> {
+    let mut f = std::io::BufReader::new(std::fs::File::open(archive)?);
+    let e = find(&mut f, name)?;
+    if e.size > max_bytes || e.compressed > max_bytes {
+        return format("the file in the archive is larger than allowed");
+    }
+    f.seek(SeekFrom::Start(e.local_header))?;
+    let mut head = [0u8; 30];
+    f.read_exact(&mut head)?;
+    if u32_at(&head, 0) != Some(0x0403_4b50) {
+        return format("the file's header is not where it says");
+    }
+    let skip = u64::from(u16_at(&head, 26).unwrap_or(0)) + u64::from(u16_at(&head, 28).unwrap_or(0));
+    f.seek_relative(i64::try_from(skip).map_err(|_| ArchiveError::Format("a header is too long".into()))?)?;
+    let len = usize::try_from(e.compressed)
+        .ok()
+        .filter(|&n| n <= 256 * 1024 * 1024)
+        .ok_or_else(|| ArchiveError::Format("the compressed model is too large".into()))?;
+    if e.size > 256 * 1024 * 1024 {
+        return format("the extracted model is too large");
+    }
+    let mut packed = Vec::new();
+    packed.try_reserve_exact(len).map_err(|_| ArchiveError::Format("not enough memory for the model archive".into()))?;
+    packed.resize(len, 0u8);
+    f.read_exact(&mut packed)?;
+    let data = match e.method {
+        0 => packed,
+        8 => miniz_oxide::inflate::decompress_to_vec_with_limit(&packed, max_bytes.min(256 * 1024 * 1024) as usize)
+            .map_err(|err| ArchiveError::Format(format!("{err:?}")))?,
+        m => return format(&format!("compression method {m} is not supported")),
+    };
+    if data.len() as u64 != e.size || crc32(&data) != e.crc {
+        return format("the file in the archive is damaged");
+    }
+    let tmp = dest.with_extension("part");
+    let written = (|| -> std::io::Result<()> {
+        let mut out = std::fs::File::create(&tmp)?;
+        out.write_all(&data)?;
+        out.flush()?;
+        drop(out);
+        std::fs::rename(&tmp, dest)
+    })();
+    if let Err(err) = written {
+        let _ = std::fs::remove_file(&tmp);
+        return Err(err.into());
+    }
+    Ok(data.len() as u64)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    /// A zip with the given (name, bytes, deflate) entries, written by hand.
+    pub(crate) fn make_zip(files: &[(&str, &[u8], bool)]) -> Vec<u8> {
+        let mut out = Vec::new();
+        let mut dir = Vec::new();
+        for (name, data, deflate) in files {
+            let packed = if *deflate { miniz_oxide::deflate::compress_to_vec(data, 6) } else { data.to_vec() };
+            let offset = out.len() as u32;
+            let method: u16 = if *deflate { 8 } else { 0 };
+            out.extend_from_slice(&0x0403_4b50u32.to_le_bytes());
+            out.extend_from_slice(&[20, 0, 0, 0]);
+            out.extend_from_slice(&method.to_le_bytes());
+            out.extend_from_slice(&[0, 0, 0, 0]);
+            out.extend_from_slice(&crc32(data).to_le_bytes());
+            out.extend_from_slice(&(packed.len() as u32).to_le_bytes());
+            out.extend_from_slice(&(data.len() as u32).to_le_bytes());
+            out.extend_from_slice(&(name.len() as u16).to_le_bytes());
+            out.extend_from_slice(&[0, 0]);
+            out.extend_from_slice(name.as_bytes());
+            out.extend_from_slice(&packed);
+            dir.extend_from_slice(&0x0201_4b50u32.to_le_bytes());
+            dir.extend_from_slice(&[20, 0, 20, 0, 0, 0]);
+            dir.extend_from_slice(&method.to_le_bytes());
+            dir.extend_from_slice(&[0, 0, 0, 0]);
+            dir.extend_from_slice(&crc32(data).to_le_bytes());
+            dir.extend_from_slice(&(packed.len() as u32).to_le_bytes());
+            dir.extend_from_slice(&(data.len() as u32).to_le_bytes());
+            dir.extend_from_slice(&(name.len() as u16).to_le_bytes());
+            dir.extend_from_slice(&[0; 12]);
+            dir.extend_from_slice(&offset.to_le_bytes());
+            dir.extend_from_slice(name.as_bytes());
+        }
+        let cd_offset = out.len() as u32;
+        out.extend_from_slice(&dir);
+        out.extend_from_slice(&0x0605_4b50u32.to_le_bytes());
+        out.extend_from_slice(&[0, 0, 0, 0]);
+        out.extend_from_slice(&(files.len() as u16).to_le_bytes());
+        out.extend_from_slice(&(files.len() as u16).to_le_bytes());
+        out.extend_from_slice(&(dir.len() as u32).to_le_bytes());
+        out.extend_from_slice(&cd_offset.to_le_bytes());
+        out.extend_from_slice(&[0, 0]);
+        out
+    }
+
+    fn temp(name: &str) -> std::path::PathBuf {
+        std::env::temp_dir().join(format!("lc-denoise-zip-{name}-{}", std::process::id()))
+    }
+
+    #[test]
+    fn a_stored_and_a_deflated_file_come_out_whole() {
+        let big: Vec<u8> = (0..200_000u32).map(|i| (i * 7 % 251) as u8).collect();
+        let zip = make_zip(&[("m/config.json", b"{}", false), ("m/model.onnx", &big, true), ("m/other.onnx", b"zz", true)]);
+        let (a, out) = (temp("a.zip"), temp("a.onnx"));
+        std::fs::write(&a, &zip).unwrap();
+        assert_eq!(extract(&a, "model.onnx", &out, 1 << 20).unwrap(), big.len() as u64);
+        assert_eq!(std::fs::read(&out).unwrap(), big);
+        assert_eq!(extract(&a, "m/config.json", &out, 1 << 20).unwrap(), 2);
+        assert_eq!(std::fs::read(&out).unwrap(), b"{}");
+        // a name must match whole path parts: `del.onnx` is not `model.onnx`
+        assert!(matches!(extract(&a, "del.onnx", &out, 1 << 20), Err(ArchiveError::Missing(_))));
+        assert!(matches!(extract(&a, "nothing", &out, 1 << 20), Err(ArchiveError::Missing(_))));
+        // the size limit applies
+        assert!(matches!(extract(&a, "model.onnx", &out, 1000), Err(ArchiveError::Format(_))));
+        for p in [a, out] {
+            let _ = std::fs::remove_file(p);
+        }
+    }
+
+    /// Opt-in: `LC_DENOISE_DTMODEL=<rawdenoise-nind.dtmodel> cargo test -p lightcraft-denoise -- --ignored archive`
+    #[test]
+    #[ignore = "needs the real download: set LC_DENOISE_DTMODEL"]
+    fn the_real_download_gives_the_model_it_should() {
+        let Some(path) = std::env::var_os("LC_DENOISE_DTMODEL") else { return };
+        let known = crate::known::find("rawnind-bayer").unwrap();
+        let (d, entry) = (known.download.unwrap(), known.entry.unwrap());
+        assert_eq!(crate::hash::sha256_file(Path::new(&path)).unwrap(), d.sha256);
+        let out = temp("real.onnx");
+        let n = extract(Path::new(&path), entry, &out, crate::manifest::MAX_MODEL_BYTES).unwrap();
+        assert_eq!(Some(n), known.manifest.size_bytes);
+        assert_eq!(Some(crate::hash::sha256_file(&out).unwrap()), known.manifest.sha256);
+        let _ = std::fs::remove_file(out);
+    }
+
+    #[test]
+    fn damaged_and_hostile_archives_are_errors_not_panics() {
+        let zip = make_zip(&[("m/model.onnx", &vec![9u8; 5000], true)]);
+        let (a, out) = (temp("b.zip"), temp("b.onnx"));
+        // cut anywhere
+        for n in (0..zip.len()).step_by(3) {
+            std::fs::write(&a, &zip[..n]).unwrap();
+            assert!(extract(&a, "model.onnx", &out, 1 << 20).is_err(), "cut at {n}");
+        }
+        // one byte changed anywhere: an error, or (when it falls in an unused field) the right file, never a panic
+        for at in 0..zip.len() {
+            let mut bad = zip.clone();
+            bad[at] ^= 0x5a;
+            std::fs::write(&a, &bad).unwrap();
+            if extract(&a, "model.onnx", &out, 1 << 20).is_ok() {
+                assert_eq!(std::fs::read(&out).unwrap(), vec![9u8; 5000], "a changed byte at {at} gave another file");
+            }
+        }
+        // not a zip at all, and an empty file
+        for junk in [Vec::new(), b"PK".to_vec(), vec![0u8; 100_000]] {
+            std::fs::write(&a, junk).unwrap();
+            assert!(extract(&a, "model.onnx", &out, 1 << 20).is_err());
+        }
+        assert!(extract(&temp("missing.zip"), "model.onnx", &out, 1 << 20).is_err());
+        for p in [a, out] {
+            let _ = std::fs::remove_file(p);
+        }
+    }
+}
@@ -0,0 +1,191 @@
+//! Bayer geometry: which way a mosaic is laid out, and how a patch of it is packed into the four planes
+//! (red, green on the red row, green on the blue row, blue) a denoise model takes.
+//!
+//! Every Bayer layout is the same 2 × 2 cell shifted by one pixel at most, so a layout is just where its
+//! first red sample is. The cell grid starts one pixel before the picture when that red sample is in
+//! column or row 1, and the picture is read with its edge mirrored (without repeating the edge sample, so
+//! the colour of a mirrored sample is the colour that belongs at that place).
+
+/// Where the first red sample of a Bayer mosaic sits, `(dx, dy)` in `{0, 1}²`: RGGB is `(0, 0)`, GRBG `(1, 0)`,
+/// GBRG `(0, 1)`, BGGR `(1, 1)`.
+#[derive(Clone, Copy, Debug, PartialEq, Eq)]
+pub struct Layout {
+    pub dx: usize,
+    pub dy: usize,
+}
+
+impl Layout {
+    /// The layout of a 2 × 2 colour pattern, row-major, colours 0 = red, 1 = green, 2 = blue; `None` when
+    /// it is not a Bayer cell (one red, one blue, the greens on the other diagonal).
+    pub fn from_cell(cell: [u8; 4]) -> Option<Layout> {
+        let at = |c: u8| cell.iter().position(|&v| v == c);
+        let red = at(0)?;
+        let blue = at(2)?;
+        let greens = cell.iter().filter(|&&v| v == 1).count();
+        // red and blue sit on one diagonal: indices 0+3 or 1+2
+        if greens != 2 || red + blue != 3 {
+            return None;
+        }
+        Some(Layout { dx: red % 2, dy: red / 2 })
+    }
+
+    /// The cell grid's origin: `-1` along an axis whose first red sample is at 1, so that cell 0 starts on a red column or row.
+    fn origin(self) -> (isize, isize) {
+        (isize::try_from(self.dx).unwrap_or(isize::MAX).saturating_neg(), isize::try_from(self.dy).unwrap_or(isize::MAX).saturating_neg())
+    }
+
+    /// Cells across and down covering a `width × height` mosaic.
+    pub fn cells(self, width: usize, height: usize) -> (usize, usize) {
+        let cells = |n: usize, phase: usize| n / 2 + (n % 2 + phase.min(1)).div_ceil(2);
+        (cells(width, self.dx), cells(height, self.dy))
+    }
+
+    /// The mosaic position of cell `(cx, cy)`'s top-left sample (may lie outside the picture).
+    pub fn position(self, cx: isize, cy: isize) -> (isize, isize) {
+        let (ox, oy) = self.origin();
+        (ox.saturating_add(cx.saturating_mul(2)), oy.saturating_add(cy.saturating_mul(2)))
+    }
+}
+
+/// `i` mirrored into `0..n` (ping-pong, edge not repeated), so `-1` is `1` and `n` is `n - 2`. `n == 0` gives 0.
+pub fn reflect(i: isize, n: usize) -> usize {
+    if n <= 1 {
+        return 0;
+    }
+    if let Some(period) = isize::try_from(n).ok().and_then(|n| (n - 1).checked_mul(2)) {
+        let m = i.rem_euclid(period);
+        return (if m < n as isize { m } else { period - m }) as usize;
+    }
+    let period = 2 * (n as i128 - 1);
+    let m = (i as i128).rem_euclid(period);
+    // The reflected value is in 0..n, so this conversion always fits usize.
+    (if m < n as i128 { m } else { period - m }) as usize
+}
+
+/// Pack the `tile × tile` cells whose top-left cell is `(cx0, cy0)` into `out` as four planes `[R, G1, G2, B]`
+/// (`4 · tile²` values, NCHW). G1 is the green beside the red sample, G2 the green under it. Samples outside
+/// the picture are mirrored. Returns `false` (and writes nothing) when the sizes do not fit.
+pub fn pack_tile(mosaic: &[f32], width: usize, height: usize, layout: Layout, cx0: isize, cy0: isize, tile: usize, out: &mut [f32]) -> bool {
+    let Some(plane) = tile.checked_mul(tile) else { return false };
+    let Some(values) = plane.checked_mul(4) else { return false };
+    if width == 0
+        || height == 0
+        || tile == 0
+        || layout.dx > 1
+        || layout.dy > 1
+        || width.checked_mul(height) != Some(mosaic.len())
+        || out.len() != values
+    {
+        return false;
+    }
+    let Ok(last) = isize::try_from(tile - 1) else { return false };
+    let (ox, oy) = layout.origin();
+    for (start, origin) in [(cx0, ox), (cy0, oy)] {
+        if start.checked_mul(2).and_then(|v| v.checked_add(origin)).is_none()
+            || start.checked_add(last).and_then(|v| v.checked_mul(2)).and_then(|v| v.checked_add(origin)).and_then(|v| v.checked_add(1)).is_none()
+        {
+            return false;
+        }
+    }
+    // the four samples of a cell, as (dx, dy) from its top-left sample
+    const SUB: [(isize, isize); 4] = [(0, 0), (1, 0), (0, 1), (1, 1)];
+    for (p, &(sx, sy)) in SUB.iter().enumerate() {
+        let Some(dst) = out.get_mut(p * plane..(p + 1) * plane) else { return false };
+        for (j, row) in dst.chunks_exact_mut(tile).enumerate() {
+            let (_, y) = layout.position(cx0, cy0 + j as isize);
+            let ry = reflect(y + sy, height);
+            let Some(src) = mosaic.get(ry * width..(ry + 1) * width) else { return false };
+            for (i, v) in row.iter_mut().enumerate() {
+                let (x, _) = layout.position(cx0 + i as isize, cy0);
+                *v = src.get(reflect(x + sx, width)).copied().unwrap_or(0.0);
+            }
+        }
+    }
+    true
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn hostile_sizes_and_positions_never_panic_or_write_output() {
+        let layout = Layout::from_cell([0, 1, 1, 2]).unwrap();
+        let mut out = [7.0; 4];
+        for (w, h, tile, x) in [(usize::MAX, 2, 1, 0), (1, 1, usize::MAX, 0), (1, 1, 0, 0), (1, 1, 1, isize::MAX), (1, 1, 1, isize::MIN)] {
+            assert!(!pack_tile(&[0.5], w, h, layout, x, 0, tile, &mut out));
+            assert_eq!(out, [7.0; 4]);
+        }
+        for n in [1, 2, isize::MAX as usize, usize::MAX] {
+            for i in [isize::MIN, -1, 0, isize::MAX] {
+                assert!(reflect(i, n) < n);
+            }
+        }
+        assert_eq!(layout.cells(usize::MAX, usize::MAX), (usize::MAX / 2 + 1, usize::MAX / 2 + 1));
+    }
+
+    #[test]
+    fn layouts_from_cells() {
+        assert_eq!(Layout::from_cell([0, 1, 1, 2]), Some(Layout { dx: 0, dy: 0 }));
+        assert_eq!(Layout::from_cell([1, 0, 2, 1]), Some(Layout { dx: 1, dy: 0 }));
+        assert_eq!(Layout::from_cell([1, 2, 0, 1]), Some(Layout { dx: 0, dy: 1 }));
+        assert_eq!(Layout::from_cell([2, 1, 1, 0]), Some(Layout { dx: 1, dy: 1 }));
+        // not Bayer: two reds, a missing blue, greens on the wrong diagonal
+        assert_eq!(Layout::from_cell([0, 0, 1, 2]), None);
+        assert_eq!(Layout::from_cell([0, 1, 2, 1]), None);
+        assert_eq!(Layout::from_cell([1, 1, 1, 1]), None);
+    }
+
+    #[test]
+    fn reflection_keeps_the_colour_of_a_site() {
+        assert_eq!(reflect(-1, 10), 1);
+        assert_eq!(reflect(-2, 10), 2);
+        assert_eq!(reflect(10, 10), 8);
+        assert_eq!(reflect(11, 10), 7);
+        assert_eq!(reflect(25, 10), 7);
+        assert_eq!(reflect(5, 0), 0);
+        assert_eq!(reflect(5, 1), 0);
+        for i in -40..60 {
+            assert_eq!(reflect(i, 11) % 2, (i.rem_euclid(2)) as usize, "parity at {i}");
+        }
+    }
+
+    #[test]
+    fn packing_puts_each_colour_in_its_plane_for_every_layout() {
+        // a mosaic whose sample value says which colour it is: R = 10, G = 20, B = 30
+        for cell in [[0u8, 1, 1, 2], [1, 0, 2, 1], [1, 2, 0, 1], [2, 1, 1, 0]] {
+            let layout = Layout::from_cell(cell).unwrap();
+            let (w, h) = (10usize, 8usize);
+            let mosaic: Vec<f32> = (0..w * h).map(|i| [10.0, 20.0, 30.0][cell[(i / w % 2) * 2 + i % w % 2] as usize]).collect();
+            let (nx, ny) = layout.cells(w, h);
+            let tile = nx.max(ny);
+            let mut out = vec![0.0; 4 * tile * tile];
+            assert!(pack_tile(&mosaic, w, h, layout, 0, 0, tile, &mut out));
+            let plane = tile * tile;
+            for (p, want) in [10.0, 20.0, 20.0, 30.0].into_iter().enumerate() {
+                assert!(out[p * plane..(p + 1) * plane].iter().all(|&v| v == want), "layout {cell:?} plane {p}");
+            }
+        }
+    }
+
+    #[test]
+    fn g1_is_beside_the_red_sample_and_g2_under_it() {
+        // RGGB with a distinct value at every site
+        let (w, h) = (4, 4);
+        let mosaic: Vec<f32> = (0..16).map(|i| i as f32).collect();
+        let mut out = vec![0.0; 4 * 4];
+        assert!(pack_tile(&mosaic, w, h, Layout { dx: 0, dy: 0 }, 0, 0, 2, &mut out));
+        // cell (0,0): R = site (0,0)=0, G1 = site (1,0)=1, G2 = site (0,1)=4, B = site (1,1)=5
+        assert_eq!([out[0], out[4], out[8], out[12]], [0.0, 1.0, 4.0, 5.0]);
+        // cell (1,0): sites (2,0)=2, (3,0)=3, (2,1)=6, (3,1)=7
+        assert_eq!([out[1], out[5], out[9], out[13]], [2.0, 3.0, 6.0, 7.0]);
+    }
+
+    #[test]
+    fn packing_rejects_wrong_sizes() {
+        let mut out = vec![0.0; 16];
+        assert!(!pack_tile(&[0.0; 15], 4, 4, Layout { dx: 0, dy: 0 }, 0, 0, 2, &mut out));
+        assert!(!pack_tile(&[0.0; 16], 4, 4, Layout { dx: 0, dy: 0 }, 0, 0, 3, &mut out));
+        assert!(!pack_tile(&[], 0, 0, Layout { dx: 0, dy: 0 }, 0, 0, 2, &mut out));
+    }
+}
@@ -0,0 +1,46 @@
+//! SHA-256 of a model file, streamed (a model can be hundreds of megabytes).
+
+use std::io::{self, Read};
+use std::path::Path;
+
+use sha2::{Digest, Sha256};
+
+/// Lowercase hex SHA-256 of everything `r` yields.
+pub fn sha256_reader<R: Read>(mut r: R) -> io::Result<String> {
+    let mut h = Sha256::new();
+    let mut buf = vec![0u8; 1 << 20];
+    loop {
+        let n = r.read(&mut buf)?;
+        if n == 0 {
+            break;
+        }
+        h.update(buf.get(..n).unwrap_or(&[]));
+    }
+    Ok(hex(&h.finalize()))
+}
+
+/// Lowercase hex SHA-256 of a file.
+pub fn sha256_file(path: &Path) -> io::Result<String> {
+    sha256_reader(io::BufReader::new(std::fs::File::open(path)?))
+}
+
+/// Lowercase hex SHA-256 of `bytes`.
+pub fn sha256_hex(bytes: &[u8]) -> String {
+    hex(&Sha256::digest(bytes))
+}
+
+fn hex(bytes: &[u8]) -> String {
+    bytes.iter().map(|b| format!("{b:02x}")).collect()
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn known_digests() {
+        assert_eq!(sha256_hex(b""), "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855");
+        assert_eq!(sha256_hex(b"abc"), "ba7816bf8f01cfea414140de5dae2223b00361a396177a9cb410ff61f20015ad");
+        assert_eq!(sha256_reader(&b"abc"[..]).unwrap(), sha256_hex(b"abc"));
+    }
+}
@@ -0,0 +1,128 @@
+//! The denoise models LightCraft knows about: what each one is, who may use it, and where to fetch it.
+//!
+//! None is part of LightCraft. A model with an address here can be downloaded when the user asks and has accepted
+//! its terms; anything else is brought by the user as an `.onnx` file with a `denoise-model.json` beside it.
+
+#[cfg(any(feature = "rawnind-model", test))]
+use crate::licence::{Commercial, Licence};
+
+use crate::manifest::DenoiserManifest;
+#[cfg(any(feature = "rawnind-model", test))]
+use crate::manifest::{Domain, Gain};
+
+/// A known model, and how to get it.
+pub struct Known {
+    pub manifest: DenoiserManifest,
+    /// What to fetch (an archive or the `.onnx` itself); `None` when the user must bring the file.
+    pub download: Option<Download>,
+    /// The `.onnx` inside the downloaded archive (`None`: the download is the `.onnx`).
+    pub entry: Option<&'static str>,
+}
+
+/// RawNIND's UtNet2 for Bayer sensors, as published for darktable (`rawdenoise-nind.dtmodel`, a zip with a Bayer and a
+/// linear model). Its weights are GPL-3.0; the training pictures are CC BY 4.0 and CC0.
+#[cfg(any(feature = "rawnind-model", test))]
+fn rawnind_bayer() -> Known {
+    let manifest = DenoiserManifest {
+        id: "rawnind-bayer".into(),
+        name: "RawNIND UtNet2 (Bayer)".into(),
+        version: "1.0".into(),
+        licence: Licence {
+            name: "GPL-3.0".into(),
+            commercial: Commercial::Yes,
+            url: Some("https://www.gnu.org/licenses/gpl-3.0.html".into()),
+            notice: "The weights are published under the GNU General Public License v3. LightCraft does not include them: they \
+                     are downloaded to your computer at your request and used by it only. If you pass the model file on, the \
+                     GPL applies to that copy. Commercial use is subject to the GPL terms, including applicable source, licence and notice obligations on redistribution."
+                .into(),
+        },
+        source: Some("https://github.com/darktable-org/darktable-ai/tree/master/models/rawdenoise-nind".into()),
+        sha256: Some("da27509dab6a2915da67e988acd86cf71f9d5bbc8d1aa0ed32933578a887b901".into()),
+        size_bytes: Some(31_056_425),
+        provenance: "Trained on RawNIND, real photographs of the same scenes taken noisy and clean (images CC BY 4.0 and CC0, \
+                     published on Wikimedia Commons and by the University of Louvain). Paper: arXiv 2501.08924."
+            .into(),
+        domain: Domain::BayerToRgb,
+        tile: 512,
+        overlap: 64,
+        gain: Gain::MatchMean { nominal: 1.0e6, max_deviation: 0.05 },
+    };
+    let download = Download {
+        id: manifest.id.clone(),
+        url: "https://github.com/darktable-org/darktable-ai/releases/download/release-5.6.0/rawdenoise-nind.dtmodel".into(),
+        file_name: "rawdenoise-nind.dtmodel".into(),
+        size_bytes: 57_700_134,
+        sha256: "d71b5f1e727c85a359e6f74dca9e2016c9d8fc3e2f7ac3e9b347d80ceca969af".into(),
+    };
+    Known { manifest, download: Some(download), entry: Some("model_bayer.onnx") }
+}
+
+/// Every model LightCraft knows.
+pub fn all() -> Vec<Known> {
+    #[cfg(feature = "rawnind-model")]
+    {
+        vec![rawnind_bayer()]
+    }
+    #[cfg(not(feature = "rawnind-model"))]
+    {
+        Vec::new()
+    }
+}
+
+/// The known model with this id.
+pub fn find(id: &str) -> Option<Known> {
+    all().into_iter().find(|k| k.manifest.id == id)
+}
+
+/// The known model whose `.onnx` has this SHA-256.
+pub fn by_sha256(sha256: &str) -> Option<DenoiserManifest> {
+    all().into_iter().map(|k| k.manifest).find(|m| m.sha256.as_deref() == Some(sha256))
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::manifest::validate;
+
+    #[test]
+    fn known_models_are_valid_unique_and_downloaded_over_https() {
+        let all = all();
+        assert_eq!(all.is_empty(), !cfg!(feature = "rawnind-model"));
+        validate(&rawnind_bayer().manifest).unwrap();
+        for (i, k) in all.iter().enumerate() {
+            validate(&k.manifest).unwrap();
+            assert!(all.iter().skip(i + 1).all(|o| o.manifest.id != k.manifest.id), "unique ids");
+            assert!(!k.manifest.licence.name.is_empty() && !k.manifest.licence.notice.is_empty(), "terms are shown before use");
+            if let Some(d) = &k.download {
+                assert_eq!(d.id, k.manifest.id);
+                assert!(d.url.starts_with("https://") && d.sha256.len() == 64 && d.size_bytes > 0);
+                // a name that cannot leave the staging folder
+                assert!(!d.file_name.contains(['/', '\\']) && !d.file_name.starts_with('.'));
+            }
+            // a model that may not be used commercially is never offered for download
+            if k.manifest.licence.commercial == Commercial::No {
+                assert!(k.download.is_none(), "{}", k.manifest.id);
+            }
+        }
+        assert_eq!(find("rawnind-bayer").is_some(), cfg!(feature = "rawnind-model"));
+        assert!(find("nothing").is_none());
+        assert_eq!(
+            by_sha256("da27509dab6a2915da67e988acd86cf71f9d5bbc8d1aa0ed32933578a887b901").map(|m| m.id),
+            if cfg!(feature = "rawnind-model") { Some("rawnind-bayer".into()) } else { None }
+        );
+    }
+}
+
+#[derive(Clone, Debug, PartialEq)]
+pub struct Download {
+    pub id: String,
+    pub url: String,
+    /// The name the file is kept under while it waits for the user to accept its terms.
+    pub file_name: String,
+    pub size_bytes: u64,
+    pub sha256: String,
+}
+
+pub fn host(url: &str) -> &str {
+    url.strip_prefix("https://").and_then(|r| r.split('/').next()).unwrap_or("")
+}
@@ -0,0 +1,16 @@
+//! Model metadata, bounded archive/cache formats and Bayer tiling, independent of inference.
+//! Used by featureless engine and web builds; no neural-network or matrix backend is linked.
+#![forbid(unsafe_code)]
+#![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)]
+pub mod archive;
+pub mod bayer;
+pub mod hash;
+pub mod known;
+pub mod licence;
+pub mod manifest;
+pub mod product;
+pub mod run;
+pub mod tiles;
+
+pub use manifest::{DenoiserManifest, Domain, Gain, ManifestError};
+pub use run::{Control, Error, Params, TileRunner, denoise_bayer};
@@ -0,0 +1,34 @@
+use serde::{Deserialize, Serialize};
+
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
+pub enum Commercial {
+    Yes,
+    /// Research or personal use only. Never bundled or redistributed by LightCraft.
+    No,
+    /// The licence or the training data is unclear: the user is told so before installing.
+    Unknown,
+}
+
+#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
+#[serde(default, rename_all = "camelCase")]
+pub struct Licence {
+    /// SPDX id or a short name ("MIT", "Apache-2.0", "InsightFace non-commercial research").
+    pub name: String,
+    pub commercial: Commercial,
+    pub url: Option<String>,
+    /// Plain-language terms shown before the model is enabled.
+    pub notice: String,
+}
+
+impl Default for Licence {
+    fn default() -> Self {
+        Self { name: String::new(), commercial: Commercial::Unknown, url: None, notice: String::new() }
+    }
+}
+
+pub fn valid_id(id: &str) -> bool {
+    let mut bytes = id.bytes();
+    let first_ok = bytes.next().is_some_and(|b| b.is_ascii_lowercase() || b.is_ascii_digit());
+    first_ok && id.len() <= 64 && bytes.all(|b| b.is_ascii_lowercase() || b.is_ascii_digit() || matches!(b, b'.' | b'_' | b'-'))
+}
@@ -0,0 +1,217 @@
+//! A denoise model's manifest: what the model takes and gives, how big its tiles are and who may use it, as
+//! plain data.
+//!
+//! It comes from a `denoise-model.json` next to the `.onnx` file or from the user's `catalog.json`. Both are
+//! untrusted: [`validate`] must pass before a manifest is used, and [`parse`] applies it after reading JSON of
+//! bounded size. The licence, id and address rules are the face models' (`crate::licence`).
+
+use crate::licence::{Licence, valid_id};
+use serde::{Deserialize, Serialize};
+
+/// Largest `denoise-model.json` we read.
+pub const MAX_MANIFEST_BYTES: usize = 64 * 1024;
+/// Largest model file we accept (bytes).
+pub const MAX_MODEL_BYTES: u64 = 2 * 1024 * 1024 * 1024;
+
+/// What the model takes and what it gives.
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
+pub enum Domain {
+    /// The sensor's Bayer mosaic (black subtracted, white at 1, not white balanced) packed as four planes
+    /// `[R, G1, G2, B]` of its 2 × 2 cells, forced to RGGB; gives camera RGB at the mosaic's resolution, so
+    /// the picture is denoised and demosaiced in one step and must not be demosaiced again.
+    BayerToRgb,
+}
+
+/// How the model's output scale is brought back to the input's.
+#[derive(Clone, Copy, Debug, PartialEq, Serialize, Deserialize)]
+#[serde(tag = "kind", rename_all = "camelCase")]
+pub enum Gain {
+    /// The output already has the input's scale.
+    None,
+    /// The output has a scale of its own, `nominal` times the input's (a million, for one model) and a little
+    /// different from tile to tile: each tile is scaled so that its channel means equal the input's. A tile too
+    /// dark to measure, or whose measured scale is further than `maxDeviation` (a fraction, 0.05 = 5 %) from
+    /// `nominal`, is scaled as if it were within that limit.
+    MatchMean { nominal: f32, max_deviation: f32 },
+}
+
+#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
+pub struct DenoiserManifest {
+    /// Lowercase letters, digits, `.` `_` `-`.
+    pub id: String,
+    pub name: String,
+    pub version: String,
+    #[serde(default)]
+    pub licence: Licence,
+    /// Where the weights come from (an `https://` page or file), for the user's information.
+    #[serde(default)]
+    pub source: Option<String>,
+    /// Lowercase hex SHA-256 of the `.onnx` file, when known.
+    #[serde(default)]
+    pub sha256: Option<String>,
+    #[serde(default)]
+    pub size_bytes: Option<u64>,
+    /// What the model was trained on, in plain words; says "undisclosed" when it is.
+    #[serde(default)]
+    pub provenance: String,
+    pub domain: Domain,
+    /// Side of the square the model runs on, in packed cells (the file's fixed input size).
+    pub tile: u32,
+    /// Cells neighbouring tiles share; the picture is blended across them.
+    pub overlap: u32,
+    pub gain: Gain,
+}
+
+#[derive(Debug, PartialEq, thiserror::Error)]
+pub enum ManifestError {
+    #[error("`{0}` is not valid: {1}")]
+    Field(&'static str, String),
+    #[error("not a denoise-model manifest: {0}")]
+    Parse(String),
+    #[error("a manifest may be at most {MAX_MANIFEST_BYTES} bytes")]
+    TooLarge,
+}
+
+fn bad<T>(field: &'static str, why: impl Into<String>) -> Result<T, ManifestError> {
+    Err(ManifestError::Field(field, why.into()))
+}
+
+fn text(field: &'static str, s: &str, min: usize, max: usize) -> Result<(), ManifestError> {
+    let n = s.chars().count();
+    if n < min || n > max {
+        return bad(field, format!("must be {min} to {max} characters"));
+    }
+    if s.chars().any(|c| c.is_control() && c != '\n') {
+        return bad(field, "must not contain control characters");
+    }
+    Ok(())
+}
+
+fn url(field: &'static str, u: &Option<String>) -> Result<(), ManifestError> {
+    match u {
+        None => Ok(()),
+        Some(u) if u.len() <= 500 && u.starts_with("https://") && !u.chars().any(|c| c.is_control() || c.is_whitespace()) => Ok(()),
+        Some(_) => bad(field, "must be an https:// address of at most 500 characters"),
+    }
+}
+
+/// Check a manifest from outside before trusting it.
+pub fn validate(m: &DenoiserManifest) -> Result<(), ManifestError> {
+    if !valid_id(&m.id) {
+        return bad("id", "1 to 64 lowercase letters, digits, `.`, `_` or `-`, starting with a letter or digit");
+    }
+    text("name", &m.name, 1, 80)?;
+    text("version", &m.version, 1, 40)?;
+    text("provenance", &m.provenance, 0, 1000)?;
+    text("licence.name", &m.licence.name, 0, 80)?;
+    text("licence.notice", &m.licence.notice, 0, 2000)?;
+    url("licence.url", &m.licence.url)?;
+    url("source", &m.source)?;
+    if let Some(h) = &m.sha256
+        && !(h.len() == 64 && h.bytes().all(|b| b.is_ascii_digit() || (b'a'..=b'f').contains(&b)))
+    {
+        return bad("sha256", "must be 64 lowercase hex digits");
+    }
+    if let Some(n) = m.size_bytes
+        && (n == 0 || n > MAX_MODEL_BYTES)
+    {
+        return bad("sizeBytes", format!("must be between 1 and {MAX_MODEL_BYTES}"));
+    }
+    // a U-Net with four poolings wants a multiple of 16; nothing smaller than 64 or larger than 2048 is a tile
+    if !(64..=2048).contains(&m.tile) || !m.tile.is_multiple_of(16) {
+        return bad("tile", "must be a multiple of 16 from 64 to 2048");
+    }
+    // tiles must overlap by less than half, or one tile's blend ramps would run into its other neighbour's
+    if !m.overlap.is_multiple_of(2) || m.overlap < 16 || m.overlap.saturating_mul(2) >= m.tile {
+        return bad("overlap", "must be even, at least 16 and less than half the tile");
+    }
+    if let Gain::MatchMean { nominal, max_deviation } = m.gain {
+        if !(nominal.is_finite() && (1e-6..=1e12).contains(&nominal)) {
+            return bad("gain.nominal", "must be between 0.000001 and 1000000000000");
+        }
+        if !(max_deviation.is_finite() && (0.0..=0.5).contains(&max_deviation)) {
+            return bad("gain.maxDeviation", "must be between 0 and 0.5");
+        }
+    }
+    Ok(())
+}
+
+/// Read and validate a `denoise-model.json`.
+pub fn parse(bytes: &[u8]) -> Result<DenoiserManifest, ManifestError> {
+    if bytes.len() > MAX_MANIFEST_BYTES {
+        return Err(ManifestError::TooLarge);
+    }
+    let bytes = bytes.strip_prefix(&[0xEF, 0xBB, 0xBF]).unwrap_or(bytes);
+    let m: DenoiserManifest = serde_json::from_slice(bytes).map_err(|e| ManifestError::Parse(e.to_string()))?;
+    validate(&m)?;
+    Ok(m)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    pub(crate) fn sample() -> DenoiserManifest {
+        DenoiserManifest {
+            id: "nind-bayer".into(),
+            name: "Bayer denoiser".into(),
+            version: "1".into(),
+            licence: Licence::default(),
+            source: Some("https://example.org/model.onnx".into()),
+            sha256: None,
+            size_bytes: Some(31_000_000),
+            provenance: "trained on real raw noise pairs".into(),
+            domain: Domain::BayerToRgb,
+            tile: 512,
+            overlap: 64,
+            gain: Gain::MatchMean { nominal: 1.0e6, max_deviation: 0.05 },
+        }
+    }
+
+    #[test]
+    fn a_good_manifest_round_trips() {
+        let m = sample();
+        let json = serde_json::to_vec(&m).unwrap();
+        assert_eq!(parse(&json).unwrap(), m);
+    }
+
+    #[test]
+    fn bad_fields_are_named() {
+        let cases: Vec<(&str, Box<dyn Fn(&mut DenoiserManifest)>)> = vec![
+            ("id", Box::new(|m| m.id = "../evil".into())),
+            ("id", Box::new(|m| m.id = String::new())),
+            ("name", Box::new(|m| m.name = "x\u{7}".into())),
+            ("source", Box::new(|m| m.source = Some("http://insecure".into()))),
+            ("sha256", Box::new(|m| m.sha256 = Some("XYZ".into()))),
+            ("sizeBytes", Box::new(|m| m.size_bytes = Some(0))),
+            ("tile", Box::new(|m| m.tile = 500)),
+            ("tile", Box::new(|m| m.tile = 4096)),
+            ("overlap", Box::new(|m| m.overlap = 300)),
+            ("overlap", Box::new(|m| m.overlap = 63)),
+            ("gain.maxDeviation", Box::new(|m| m.gain = Gain::MatchMean { nominal: 1.0, max_deviation: f32::NAN })),
+            ("gain.maxDeviation", Box::new(|m| m.gain = Gain::MatchMean { nominal: 1.0, max_deviation: 2.0 })),
+            ("gain.nominal", Box::new(|m| m.gain = Gain::MatchMean { nominal: 0.0, max_deviation: 0.1 })),
+            ("gain.nominal", Box::new(|m| m.gain = Gain::MatchMean { nominal: f32::INFINITY, max_deviation: 0.1 })),
+        ];
+        for (field, edit) in cases {
+            let mut m = sample();
+            edit(&mut m);
+            match validate(&m) {
+                Err(ManifestError::Field(f, _)) => assert_eq!(f, field),
+                other => panic!("{field}: {other:?}"),
+            }
+        }
+    }
+
+    #[test]
+    fn parse_is_bounded_and_tolerant() {
+        assert_eq!(parse(&vec![b' '; MAX_MANIFEST_BYTES + 1]), Err(ManifestError::TooLarge));
+        assert!(matches!(parse(b"{ nope"), Err(ManifestError::Parse(_))));
+        assert!(matches!(parse(b""), Err(ManifestError::Parse(_))));
+        let mut with_bom = vec![0xEF, 0xBB, 0xBF];
+        with_bom.extend(serde_json::to_vec(&sample()).unwrap());
+        assert!(parse(&with_bom).is_ok());
+    }
+}
@@ -0,0 +1,616 @@
+//! The denoised picture as cached data, and the blend the Amount slider controls.
+//!
+//! A product is camera RGB at the sensor's resolution (what the model gave, before the raw's own opcodes and
+//! crop), kept as half floats in strips of rows, each strip byte-shuffled and deflated with its own checksum.
+//! It is derived data: any problem reading it (missing, truncated, corrupt, made for another photo or model) is
+//! just a cache miss, never a panic, and the picture can be made again.
+//!
+//! File: `LCDN`, version, header length, header JSON (`width`, `height`, `strip`, `key`), then for each strip its
+//! compressed length and CRC-32, then the strips' bytes. `key` says what the product was made from (the photo's
+//! content, the model, this code's version, the settings it depends on); a file with another key is stale.
+
+use std::io::{Read, Write};
+use std::path::Path;
+
+use lightcraft_raster::Rgb32f;
+use rayon::prelude::*;
+use serde::{Deserialize, Serialize};
+
+const MAGIC: &[u8; 4] = b"LCDN";
+const VERSION: u8 = 1;
+const MAX_HEADER: usize = 4096;
+const MAX_SIDE: usize = 1 << 17;
+/// Most pixels a product may have (the denoiser's own limit, [`crate::run::MAX_PIXELS`]).
+const MAX_PIXELS: usize = crate::run::MAX_PIXELS;
+/// Rows per strip: about a megapixel of three planes at 24 MP widths.
+const STRIP_ROWS: usize = 128;
+/// Bumped when the meaning of a product changes (included in keys by the caller).
+pub const ALGORITHM: u32 = 1;
+
+#[derive(Debug, thiserror::Error)]
+pub enum ProductError {
+    #[error("the denoised picture could not be read or written: {0}")]
+    Io(String),
+    #[error("not a usable denoised picture: {0}")]
+    Format(String),
+    /// A product made from something else (another photo state, model or version).
+    #[error("the denoised picture is out of date")]
+    Stale,
+}
+
+impl From<std::io::Error> for ProductError {
+    fn from(e: std::io::Error) -> Self {
+        ProductError::Io(e.to_string())
+    }
+}
+
+#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
+pub struct Header {
+    pub width: usize,
+    pub height: usize,
+    pub strip: usize,
+    pub key: String,
+}
+
+fn allocate<T: Clone>(n: usize, value: T) -> Result<Vec<T>, ProductError> {
+    let mut v = Vec::new();
+    v.try_reserve_exact(n).map_err(|_| ProductError::Format("not enough memory for the denoised picture".into()))?;
+    v.resize(n, value);
+    Ok(v)
+}
+
+/// `f32` as IEEE half-float bits, rounded to nearest even; beyond the range it saturates, NaN becomes 0.
+pub fn to_f16(v: f32) -> u16 {
+    if !v.is_finite() {
+        return if v.is_nan() {
+            0
+        } else if v > 0.0 {
+            0x7bff
+        } else {
+            0xfbff
+        };
+    }
+    let b = v.to_bits();
+    let sign = ((b >> 16) & 0x8000) as u16;
+    let exp = ((b >> 23) & 0xff) as i32 - 127 + 15;
+    let man = b & 0x7f_ffff;
+    if exp >= 31 {
+        return sign | 0x7bff;
+    }
+    if exp <= 0 {
+        if exp < -10 {
+            return sign;
+        }
+        let man = man | 0x80_0000;
+        let shift = (14 - exp) as u32;
+        let (half, rem, halfway) = (man >> shift, man & ((1u32 << shift) - 1), 1u32 << (shift - 1));
+        let mut h = half as u16;
+        if rem > halfway || (rem == halfway && h & 1 == 1) {
+            h += 1;
+        }
+        return sign | h;
+    }
+    let mut h = (((exp as u32) << 10) | (man >> 13)) as u16;
+    let rem = man & 0x1fff;
+    if rem > 0x1000 || (rem == 0x1000 && h & 1 == 1) {
+        h += 1;
+    }
+    if h >= 0x7c00 {
+        h = 0x7bff;
+    }
+    sign | h
+}
+
+/// Half-float bits as `f32`.
+pub fn from_f16(h: u16) -> f32 {
+    let sign = if h & 0x8000 != 0 { -1.0f32 } else { 1.0 };
+    let exp = i32::from((h >> 10) & 0x1f);
+    let man = f32::from(h & 0x3ff);
+    match exp {
+        0 => sign * man * 2f32.powi(-24),
+        31 => sign * 65504.0,
+        _ => sign * (1.0 + man / 1024.0) * 2f32.powi(exp - 15),
+    }
+}
+
+fn crc(bytes: &[u8]) -> u32 {
+    // CRC-32 (IEEE), bitwise: strips are small and this runs once per strip
+    let mut table = [0u32; 256];
+    for (i, t) in table.iter_mut().enumerate() {
+        let mut c = i as u32;
+        for _ in 0..8 {
+            c = if c & 1 != 0 { 0xedb8_8320 ^ (c >> 1) } else { c >> 1 };
+        }
+        *t = c;
+    }
+    !bytes.iter().fold(!0u32, |c, &b| table[((c ^ u32::from(b)) & 0xff) as usize] ^ (c >> 8))
+}
+
+/// One strip as bytes: for each plane (R, G, B) the low bytes of its half floats, then the high bytes.
+fn encode_strip(img: &Rgb32f, y0: usize, y1: usize) -> Result<Vec<u8>, ProductError> {
+    let n = (y1 - y0) * img.width;
+    let mut raw = allocate(6 * n, 0u8)?;
+    let rows = img.data.get(y0 * img.width..y1 * img.width).unwrap_or(&[]);
+    for (i, px) in rows.iter().enumerate() {
+        for (c, &v) in px.iter().enumerate() {
+            let h = to_f16(v);
+            if let Some(lo) = raw.get_mut(2 * c * n + i) {
+                *lo = h as u8;
+            }
+            if let Some(hi) = raw.get_mut((2 * c + 1) * n + i) {
+                *hi = (h >> 8) as u8;
+            }
+        }
+    }
+    Ok(miniz_oxide::deflate::compress_to_vec(&raw, 3))
+}
+
+/// Write `img` to `path` (through a temporary file next to it, so a reader never sees half a product).
+pub fn write(path: &Path, img: &Rgb32f, key: &str) -> Result<(), ProductError> {
+    let pixels = img.width.checked_mul(img.height).filter(|&n| n <= MAX_PIXELS);
+    if img.width == 0 || img.height == 0 || img.width > MAX_SIDE || img.height > MAX_SIDE || pixels != Some(img.data.len()) {
+        return Err(ProductError::Format("the picture has no usable size".into()));
+    }
+    let strips: Vec<(usize, usize)> = (0..img.height).step_by(STRIP_ROWS).map(|y| (y, (y + STRIP_ROWS).min(img.height))).collect();
+    let packed: Vec<Vec<u8>> = strips.par_iter().map(|&(a, b)| encode_strip(img, a, b)).collect::<Result<_, _>>()?;
+    let header = serde_json::to_vec(&Header { width: img.width, height: img.height, strip: STRIP_ROWS, key: key.to_string() })
+        .map_err(|e| ProductError::Format(e.to_string()))?;
+    let tmp = path.with_extension("tmp");
+    let result = (|| -> Result<(), ProductError> {
+        if let Some(dir) = path.parent() {
+            std::fs::create_dir_all(dir)?;
+        }
+        let mut f = std::io::BufWriter::new(std::fs::File::create(&tmp)?);
+        f.write_all(MAGIC)?;
+        f.write_all(&[VERSION, 0, 0, 0])?;
+        f.write_all(&(header.len() as u32).to_le_bytes())?;
+        f.write_all(&header)?;
+        for p in &packed {
+            f.write_all(&(p.len() as u32).to_le_bytes())?;
+            f.write_all(&crc(p).to_le_bytes())?;
+        }
+        for p in &packed {
+            f.write_all(p)?;
+        }
+        f.flush()?;
+        drop(f);
+        std::fs::rename(&tmp, path)?;
+        Ok(())
+    })();
+    if result.is_err() {
+        let _ = std::fs::remove_file(&tmp);
+    }
+    result
+}
+
+/// The parsed start of a product file: header, each strip's (compressed length, CRC-32), and the file length the
+/// start says the whole file has.
+struct Start {
+    header: Header,
+    table: Vec<(usize, u32)>,
+    expected_len: u64,
+}
+
+fn read_start(r: &mut impl Read) -> Result<Start, ProductError> {
+    let mut fixed = [0u8; 12];
+    r.read_exact(&mut fixed)?;
+    if fixed.get(0..4) != Some(&MAGIC[..]) || fixed.get(4) != Some(&VERSION) {
+        return Err(ProductError::Format("not a denoised picture of this version".into()));
+    }
+    let len = fixed.get(8..12).and_then(|b| <[u8; 4]>::try_from(b).ok()).map(u32::from_le_bytes).unwrap_or(u32::MAX) as usize;
+    if len > MAX_HEADER {
+        return Err(ProductError::Format("the header is too large".into()));
+    }
+    let mut hb = allocate(len, 0u8)?;
+    r.read_exact(&mut hb)?;
+    let h: Header = serde_json::from_slice(&hb).map_err(|e| ProductError::Format(e.to_string()))?;
+    let pixels = h.width.saturating_mul(h.height);
+    if h.width == 0 || h.height == 0 || h.width > MAX_SIDE || h.height > MAX_SIDE || pixels > MAX_PIXELS || h.strip != STRIP_ROWS {
+        return Err(ProductError::Format("the size is not usable".into()));
+    }
+    let n = h.height.div_ceil(h.strip);
+    let mut table = Vec::with_capacity(n);
+    let mut total = 12 + len as u64 + 8 * n as u64;
+    for _ in 0..n {
+        let mut e = [0u8; 8];
+        r.read_exact(&mut e)?;
+        let l = u32::from_le_bytes([e[0], e[1], e[2], e[3]]) as usize;
+        let c = u32::from_le_bytes([e[4], e[5], e[6], e[7]]);
+        // a strip never compresses to much more than its raw size
+        if l > 12 * h.strip * h.width + 4096 {
+            return Err(ProductError::Format("a strip is larger than it can be".into()));
+        }
+        total += l as u64;
+        table.push((l, c));
+    }
+    Ok(Start { header: h, table, expected_len: total })
+}
+
+/// The header of the product at `path`, when it is one.
+pub fn read_header_at(path: &Path) -> Result<Header, ProductError> {
+    let mut f = std::io::BufReader::new(std::fs::File::open(path)?);
+    Ok(read_start(&mut f)?.header)
+}
+
+/// Whether `path` holds a complete-looking product made from `key` (the header and the file's length are checked;
+/// strips are checked when read).
+pub fn is_current(path: &Path, key: &str) -> bool {
+    let check = || -> Result<bool, ProductError> {
+        let mut f = std::io::BufReader::new(std::fs::File::open(path)?);
+        let s = read_start(&mut f)?;
+        Ok(s.header.key == key && std::fs::metadata(path)?.len() == s.expected_len)
+    };
+    check().unwrap_or(false)
+}
+
+/// A rectangle of a product, in the product's pixels.
+#[derive(Clone, Copy, Debug, PartialEq, Eq)]
+pub struct Window {
+    pub x: usize,
+    pub y: usize,
+    pub width: usize,
+    pub height: usize,
+}
+
+/// Half floats as `f32`, for all 65536 bit patterns (a table is much faster than computing each).
+fn f16_table() -> &'static [f32] {
+    static TABLE: std::sync::OnceLock<Vec<f32>> = std::sync::OnceLock::new();
+    TABLE.get_or_init(|| (0..=u16::MAX).map(from_f16).collect())
+}
+
+/// Strips decoded together (they are inflated in parallel, then folded in order).
+const GROUP: usize = 8;
+
+/// One strip, checked and inflated: for each plane the low bytes of its half floats, then the high bytes.
+fn decode_strip(index: usize, packed: &[u8], expect_crc: u32, n: usize) -> Result<Vec<u8>, ProductError> {
+    if crc(packed) != expect_crc {
+        return Err(ProductError::Format(format!("strip {index} is damaged")));
+    }
+    let raw = miniz_oxide::inflate::decompress_to_vec_with_limit(packed, 6 * n).map_err(|e| ProductError::Format(format!("strip {index}: {e:?}")))?;
+    if raw.len() != 6 * n {
+        return Err(ProductError::Format(format!("strip {index} has the wrong size")));
+    }
+    Ok(raw)
+}
+
+/// Read the product at `path`, made from `key`, with every `factor × factor` block of pixels averaged into one
+/// (`factor` 1 = full size). The result is camera RGB.
+///
+/// With a `window`, only that rectangle is read and the blocks start at its corner; only whole blocks are made, so
+/// the result is `window.width / factor × window.height / factor` (the same blocks a binned development of the same
+/// crop makes). Without one the whole product is read and blocks at the right and bottom edges may be smaller.
+///
+/// With `clip` (white is 1), a block in which any sample of a colour is at or above it takes that colour's
+/// maximum instead of the mean, so a clipped highlight stays clipped for highlight reconstruction.
+pub fn read_window(path: &Path, key: &str, window: Option<Window>, factor: usize, clip: Option<f32>) -> Result<Rgb32f, ProductError> {
+    let factor = factor.clamp(1, 1 << 12);
+    let mut f = std::io::BufReader::new(std::fs::File::open(path)?);
+    let Start { header: h, table, expected_len } = read_start(&mut f)?;
+    if h.key != key {
+        return Err(ProductError::Stale);
+    }
+    if std::fs::metadata(path)?.len() != expected_len {
+        return Err(ProductError::Format("the file is not the length its header says".into()));
+    }
+    let w = window.unwrap_or(Window { x: 0, y: 0, width: h.width, height: h.height });
+    let fits = |a: usize, b: usize, limit: usize| a.checked_add(b).is_some_and(|e| e <= limit);
+    if w.width == 0 || w.height == 0 || !fits(w.x, w.width, h.width) || !fits(w.y, w.height, h.height) {
+        return Err(ProductError::Format("the window does not lie inside the picture".into()));
+    }
+    // whole blocks for a window; partial ones at the edges when the whole picture is read
+    let whole = window.is_none();
+    let (ow, oh) = if whole { (w.width.div_ceil(factor), w.height.div_ceil(factor)) } else { (w.width / factor, w.height / factor) };
+    if ow == 0 || oh == 0 {
+        return Err(ProductError::Format("the window is smaller than one block".into()));
+    }
+    // the picture rows and columns the blocks cover
+    let (x_end, y_end) = if whole { (w.x + w.width, w.y + w.height) } else { (w.x + ow * factor, w.y + oh * factor) };
+    let mut out = Rgb32f { width: ow, height: oh, data: allocate(ow * oh, [0.0; 3])? };
+    let lut = f16_table();
+    let value = |raw: &[u8], c: usize, n: usize, i: usize| -> f32 {
+        let (lo, hi) = (raw.get(2 * c * n + i).copied().unwrap_or(0), raw.get((2 * c + 1) * n + i).copied().unwrap_or(0));
+        lut.get(usize::from(lo) | usize::from(hi) << 8).copied().unwrap_or(0.0)
+    };
+    // which strips are needed, and where each starts in the file (after the header and the strip table)
+    let wanted = |s: usize| s * h.strip < y_end && (s + 1) * h.strip > w.y;
+    let (mut sum, mut peak) = (allocate(ow, [0f32; 3])?, allocate(ow, [f32::MIN; 3])?);
+    let (mut rows_in, mut oy) = (0usize, 0usize);
+    let strips: Vec<usize> = (0..table.len()).filter(|&s| wanted(s)).collect();
+    // skip the strips before the first one wanted
+    let skipped: usize = table.iter().take(strips.first().copied().unwrap_or(0)).map(|e| e.0).sum();
+    f.seek_relative(i64::try_from(skipped).map_err(|_| ProductError::Format("the file is too large".into()))?)?;
+    for group in strips.chunks(GROUP) {
+        let mut packed = Vec::with_capacity(group.len());
+        for &s in group {
+            let Some(&(len, _)) = table.get(s) else { return Err(ProductError::Format("a strip is missing".into())) };
+            let mut p = allocate(len, 0u8)?;
+            f.read_exact(&mut p)?;
+            packed.push(p);
+        }
+        let decoded: Vec<Result<Vec<u8>, ProductError>> = group
+            .par_iter()
+            .zip(packed.par_iter())
+            .map(|(&s, p)| {
+                let rows = h.height.saturating_sub(s * h.strip).min(h.strip);
+                decode_strip(s, p, table.get(s).map_or(0, |e| e.1), rows * h.width)
+            })
+            .collect();
+        for (&s, raw) in group.iter().zip(decoded) {
+            let raw = raw?;
+            let y0 = s * h.strip;
+            let rows = h.height.saturating_sub(y0).min(h.strip);
+            let n = rows * h.width;
+            for r in 0..rows {
+                let y = y0 + r;
+                if y < w.y || y >= y_end {
+                    continue;
+                }
+                let base = r * h.width;
+                for (ox, (sm, pk)) in sum.iter_mut().zip(peak.iter_mut()).enumerate() {
+                    let x0 = w.x + ox * factor;
+                    let x1 = (x0 + factor).min(x_end);
+                    for x in x0..x1 {
+                        for c in 0..3 {
+                            let v = value(&raw, c, n, base + x);
+                            if let (Some(s_), Some(p_)) = (sm.get_mut(c), pk.get_mut(c)) {
+                                *s_ += v;
+                                *p_ = p_.max(v);
+                            }
+                        }
+                    }
+                }
+                rows_in += 1;
+                if rows_in == factor || y + 1 == y_end {
+                    for (ox, (sm, pk)) in sum.iter_mut().zip(peak.iter_mut()).enumerate() {
+                        // the block's size at the right and bottom edges may be smaller
+                        let bw = (x_end - (w.x + ox * factor)).min(factor);
+                        let k = 1.0 / (rows_in * bw).max(1) as f32;
+                        if let Some(o) = out.data.get_mut(oy * ow + ox) {
+                            for c in 0..3 {
+                                let (s_, p_) = (sm.get(c).copied().unwrap_or(0.0), pk.get(c).copied().unwrap_or(0.0));
+                                if let Some(slot) = o.get_mut(c) {
+                                    *slot = if factor > 1 && clip.is_some_and(|t| p_ >= t) { p_ } else { s_ * k };
+                                }
+                            }
+                        }
+                        *sm = [0.0; 3];
+                        *pk = [f32::MIN; 3];
+                    }
+                    rows_in = 0;
+                    oy += 1;
+                }
+            }
+        }
+    }
+    Ok(out)
+}
+
+/// [`read_window`] of the whole product: every `factor × factor` block averaged into one.
+pub fn read(path: &Path, key: &str, factor: usize) -> Result<Rgb32f, ProductError> {
+    read_window(path, key, None, factor, None)
+}
+
+/// `base + (denoised − base) · amount` for every pixel (`amount` 0 to 1), in place in `base`. The two pictures must
+/// be the same size; if they are not, `base` is left as it was and `false` is returned.
+pub fn blend(base: &mut Rgb32f, denoised: &Rgb32f, amount: f32) -> bool {
+    if base.width != denoised.width || base.height != denoised.height {
+        return false;
+    }
+    let a = if amount.is_finite() { amount.clamp(0.0, 1.0) } else { 0.0 };
+    base.data.par_iter_mut().zip(denoised.data.par_iter()).for_each(|(b, d)| {
+        for c in 0..3 {
+            b[c] += (d[c] - b[c]) * a;
+        }
+    });
+    true
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn oversized_product_header_is_rejected_before_allocating_pixels() {
+        let p = temp("oversized-header");
+        let header = serde_json::to_vec(&Header { width: 10001, height: 10000, strip: STRIP_ROWS, key: "k".into() }).unwrap();
+        let mut bytes = Vec::from(MAGIC.as_slice());
+        bytes.extend_from_slice(&[VERSION, 0, 0, 0]);
+        bytes.extend_from_slice(&(header.len() as u32).to_le_bytes());
+        bytes.extend_from_slice(&header);
+        std::fs::write(&p, bytes).unwrap();
+        assert!(read_header_at(&p).is_err());
+        let _ = std::fs::remove_file(p);
+    }
+
+    fn temp(name: &str) -> std::path::PathBuf {
+        std::env::temp_dir().join(format!("lc-denoise-{name}-{}.lcdn", std::process::id()))
+    }
+
+    fn picture(w: usize, h: usize) -> Rgb32f {
+        Rgb32f {
+            width: w,
+            height: h,
+            data: (0..w * h).map(|i| [(i % w) as f32 / w as f32, (i / w) as f32 / h as f32, 0.001 + 0.5 * ((i % 7) as f32 / 7.0)]).collect(),
+        }
+    }
+
+    #[test]
+    fn half_floats_round_trip_closely() {
+        for v in [0.0f32, 1.0, -1.0, 0.5, 0.333_333, 1e-4, 6.1e-5, 3e-6, 123.456, 65000.0, -0.002] {
+            let back = from_f16(to_f16(v));
+            assert!((back - v).abs() <= v.abs() * 0.001 + 1e-7, "{v} -> {back}");
+        }
+        assert_eq!(from_f16(to_f16(1e10)), 65504.0);
+        assert_eq!(to_f16(f32::NAN), 0);
+        assert_eq!(from_f16(to_f16(f32::NEG_INFINITY)), -65504.0);
+    }
+
+    #[test]
+    fn a_product_survives_a_round_trip_at_full_and_reduced_size() {
+        let (w, h) = (301, 517);
+        let img = picture(w, h);
+        let p = temp("round");
+        write(&p, &img, "k1").unwrap();
+        assert!(is_current(&p, "k1") && !is_current(&p, "k2"));
+        assert_eq!(read_header_at(&p).unwrap().width, w);
+        let full = read(&p, "k1", 1).unwrap();
+        assert_eq!((full.width, full.height), (w, h));
+        for (a, b) in full.data.iter().zip(&img.data) {
+            for c in 0..3 {
+                assert!((a[c] - b[c]).abs() < 0.002, "{a:?} vs {b:?}");
+            }
+        }
+        // averaged into blocks of 4: the sizes round up, and the mean of the picture is kept
+        let small = read(&p, "k1", 4).unwrap();
+        assert_eq!((small.width, small.height), (w.div_ceil(4), h.div_ceil(4)));
+        let mean = |i: &Rgb32f| i.data.iter().map(|p| f64::from(p[1])).sum::<f64>() / i.data.len() as f64;
+        assert!((mean(&small) - mean(&img)).abs() < 0.01);
+        // a factor that does not divide the strip height is the same average
+        let odd = read(&p, "k1", 5).unwrap();
+        assert_eq!((odd.width, odd.height), (w.div_ceil(5), h.div_ceil(5)));
+        assert!((mean(&odd) - mean(&img)).abs() < 0.01);
+        assert!(matches!(read(&p, "other", 1), Err(ProductError::Stale)));
+        std::fs::remove_file(&p).unwrap();
+    }
+
+    /// What binning a window of `img` by `k` means, written the slow obvious way.
+    fn blocks(img: &Rgb32f, win: Window, k: usize, clip: Option<f32>) -> Rgb32f {
+        let (ow, oh) = (win.width / k, win.height / k);
+        let mut out = Rgb32f { width: ow, height: oh, data: vec![[0.0; 3]; ow * oh] };
+        for by in 0..oh {
+            for bx in 0..ow {
+                for c in 0..3 {
+                    let vals: Vec<f32> = (0..k * k).map(|i| img.data[(win.y + by * k + i / k) * img.width + win.x + bx * k + i % k][c]).collect();
+                    let max = vals.iter().copied().fold(f32::MIN, f32::max);
+                    out.data[by * ow + bx][c] =
+                        if k > 1 && clip.is_some_and(|t| max >= t) { max } else { vals.iter().sum::<f32>() / vals.len() as f32 };
+                }
+            }
+        }
+        out
+    }
+
+    #[test]
+    fn a_window_is_read_as_the_blocks_of_that_crop() {
+        let (w, h) = (301, 517);
+        let mut img = picture(w, h);
+        // a clipped sample in the middle of a block
+        img.data[(130 + 2) * w + 61 + 1] = [1.0, 0.4, 0.4];
+        let p = temp("window");
+        write(&p, &img, "k1").unwrap();
+        let near = |a: &Rgb32f, b: &Rgb32f| {
+            assert_eq!((a.width, a.height), (b.width, b.height));
+            for (x, y) in a.data.iter().zip(&b.data) {
+                for c in 0..3 {
+                    assert!((x[c] - y[c]).abs() < 0.003, "{x:?} vs {y:?}");
+                }
+            }
+        };
+        // windows at odd places, across strip boundaries (128 rows), with factors that do and do not divide them
+        for (win, k) in [
+            (Window { x: 3, y: 5, width: 200, height: 300 }, 1),
+            (Window { x: 3, y: 5, width: 200, height: 300 }, 4),
+            (Window { x: 61, y: 130, width: 121, height: 250 }, 5),
+            (Window { x: 0, y: 0, width: w, height: h }, 2),
+            (Window { x: 1, y: 380, width: 90, height: 137 }, 3),
+        ] {
+            let got = read_window(&p, "k1", Some(win), k, None).unwrap();
+            near(&got, &blocks(&img, win, k, None));
+        }
+        // with a clip level the block with the clipped sample keeps its highest value, the others stay means
+        let win = Window { x: 61, y: 130, width: 40, height: 20 };
+        let clipped = read_window(&p, "k1", Some(win), 4, Some(0.99)).unwrap();
+        near(&clipped, &blocks(&img, win, 4, Some(0.99)));
+        assert!(clipped.data[0][0] > 0.99, "the clipped block stays clipped: {:?}", clipped.data[0]);
+        let plain = read_window(&p, "k1", Some(win), 4, None).unwrap();
+        assert!(plain.data[0][0] < 0.99, "an average falls below the clip level");
+        // windows that do not fit, or are smaller than a block, are errors
+        for bad in [
+            Window { x: 0, y: 0, width: 0, height: 5 },
+            Window { x: 290, y: 0, width: 20, height: 5 },
+            Window { x: 0, y: 510, width: 5, height: 10 },
+            Window { x: usize::MAX, y: 0, width: 2, height: 2 },
+            Window { x: 0, y: 0, width: 3, height: 3 },
+        ] {
+            assert!(read_window(&p, "k1", Some(bad), 4, None).is_err(), "{bad:?}");
+        }
+        std::fs::remove_file(&p).unwrap();
+    }
+
+    #[test]
+    fn a_smooth_picture_compresses_well() {
+        let img = Rgb32f {
+            width: 512,
+            height: 512,
+            data: (0..512 * 512)
+                .map(|i| {
+                    let v = 0.2 + 0.0001 * (i % 512) as f32;
+                    [v, v * 0.8, v * 0.6]
+                })
+                .collect(),
+        };
+        let p = temp("size");
+        write(&p, &img, "k").unwrap();
+        let bytes = std::fs::metadata(&p).unwrap().len() as usize;
+        assert!(bytes < 512 * 512 * 6 / 2, "{bytes} bytes for a smooth picture");
+        std::fs::remove_file(&p).unwrap();
+    }
+
+    #[test]
+    fn damaged_files_are_errors_not_panics() {
+        let img = picture(100, 300);
+        let p = temp("damage");
+        write(&p, &img, "k").unwrap();
+        let good = std::fs::read(&p).unwrap();
+        // truncated at every length
+        for cut in (0..good.len()).step_by(37) {
+            std::fs::write(&p, &good[..cut]).unwrap();
+            assert!(read(&p, "k", 1).is_err(), "cut at {cut}");
+            assert!(!is_current(&p, "k"));
+        }
+        // a flipped byte anywhere is caught by a checksum, the header parse, or the key
+        for at in (0..good.len()).step_by(53) {
+            let mut bad = good.clone();
+            bad[at] ^= 0xff;
+            std::fs::write(&p, &bad).unwrap();
+            let _ = read(&p, "k", 1); // must not panic; most are errors
+        }
+        // a header that claims an absurd size
+        let mut huge = b"LCDN\x01\0\0\0".to_vec();
+        let hdr = br#"{"width":4000000000,"height":4000000000,"strip":128,"key":"k"}"#;
+        huge.extend((hdr.len() as u32).to_le_bytes());
+        huge.extend(hdr);
+        std::fs::write(&p, &huge).unwrap();
+        assert!(matches!(read(&p, "k", 1), Err(ProductError::Format(_))));
+        std::fs::write(&p, b"").unwrap();
+        assert!(read(&p, "k", 1).is_err());
+        assert!(read(&temp("missing"), "k", 1).is_err());
+        std::fs::remove_file(&p).unwrap();
+    }
+
+    #[test]
+    fn writing_rejects_unusable_pictures() {
+        let p = temp("bad");
+        assert!(write(&p, &Rgb32f { width: 0, height: 0, data: vec![] }, "k").is_err());
+        assert!(write(&p, &Rgb32f { width: 3, height: 3, data: vec![[0.0; 3]; 8] }, "k").is_err());
+        assert!(!p.exists());
+    }
+
+    #[test]
+    fn blend_mixes_by_amount_and_rejects_other_sizes() {
+        let mut base = Rgb32f { width: 2, height: 1, data: vec![[0.0; 3], [1.0; 3]] };
+        let dn = Rgb32f { width: 2, height: 1, data: vec![[1.0; 3], [0.0; 3]] };
+        assert!(blend(&mut base, &dn, 0.25));
+        assert_eq!(base.data, vec![[0.25; 3], [0.75; 3]]);
+        assert!(blend(&mut base, &dn, f32::NAN));
+        assert_eq!(base.data, vec![[0.25; 3], [0.75; 3]]);
+        assert!(!blend(&mut base, &Rgb32f { width: 1, height: 1, data: vec![[0.0; 3]] }, 1.0));
+        assert!(blend(&mut base, &dn, 1.0));
+        assert_eq!(base.data, dn.data);
+    }
+
+    #[test]
+    fn crc_matches_the_standard_check_value() {
+        assert_eq!(crc(b"123456789"), 0xcbf4_3926);
+    }
+}
@@ -0,0 +1,589 @@
+//! Denoising a whole Bayer mosaic with a tile model: cut it into overlapping tiles, run the model on each, bring
+//! each tile's scale back to the input's, keep clipped highlights as they were, and blend the tiles together.
+//!
+//! The model itself is a [`TileRunner`], so everything around it (packing, tiling, scale matching, highlight
+//! protection, blending, cancelling, progress) is tested without one.
+
+use std::sync::atomic::{AtomicBool, Ordering};
+
+use lightcraft_raster::Rgb32f;
+use rayon::prelude::*;
+
+use crate::bayer::{Layout, pack_tile};
+use crate::manifest::Gain;
+use crate::tiles::{starts, weight};
+
+/// Most pixels one picture may have (guards the allocations that follow from its size).
+pub const MAX_PIXELS: usize = 100_000_000;
+/// A tile whose mean signal is below this (white = 1) is too dark to measure its scale from.
+const MIN_SIGNAL: f32 = 0.003;
+
+#[derive(Debug, thiserror::Error)]
+pub enum Error {
+    #[error("{0}")]
+    Input(String),
+    #[error("the model gave {0}")]
+    Model(String),
+    #[error("the model could not run: {0}")]
+    Runtime(String),
+    #[error("the model could not be loaded: {0}")]
+    Load(String),
+    #[error("cancelled")]
+    Cancelled,
+}
+
+fn allocate<T: Clone>(n: usize, value: T) -> Result<Vec<T>, Error> {
+    let mut v = Vec::new();
+    v.try_reserve_exact(n).map_err(|_| Error::Input("not enough memory for denoise".into()))?;
+    v.resize(n, value);
+    Ok(v)
+}
+
+/// Something that runs the model on one tile.
+pub trait TileRunner: Sync {
+    /// `input` is the tile as four planes `[R, G1, G2, B]` of `tile × tile` cells (NCHW, batch 1); the answer is
+    /// camera RGB, three planes of `2·tile × 2·tile` pixels (CHW), in the model's own scale.
+    fn run(&self, input: &[f32]) -> Result<Vec<f32>, Error>;
+}
+
+pub struct Params {
+    /// Side of the model's square, in cells.
+    pub tile: usize,
+    /// Cells neighbouring tiles share.
+    pub overlap: usize,
+    pub gain: Gain,
+    /// White is 1; samples at or above this (minus a little) are clipped and keep their value.
+    pub clip: Option<f32>,
+    /// Tiles run at once (each model call may itself be single-threaded).
+    pub parallel: usize,
+}
+
+#[derive(Default)]
+pub struct Control<'a> {
+    pub cancel: Option<&'a AtomicBool>,
+    /// `(tiles done, tiles in all)`.
+    pub progress: Option<&'a (dyn Fn(usize, usize) + Sync)>,
+}
+
+/// The tile grid of a picture: `(starts across, starts down)`.
+pub fn grid(width: usize, height: usize, layout: Layout, tile: usize, overlap: usize) -> (Vec<usize>, Vec<usize>) {
+    let (nx, ny) = layout.cells(width, height);
+    (starts(nx, tile, overlap), starts(ny, tile, overlap))
+}
+
+/// How many model calls [`denoise_bayer`] makes for a picture.
+pub fn tile_count(width: usize, height: usize, layout: Layout, tile: usize, overlap: usize) -> usize {
+    let (sx, sy) = grid(width, height, layout, tile, overlap);
+    sx.len().saturating_mul(sy.len())
+}
+
+/// Denoise (and demosaic) the normalised mosaic `mosaic` of `width × height` samples, laid out as `layout`.
+/// Gives camera RGB at the mosaic's resolution.
+pub fn denoise_bayer(
+    mosaic: &[f32],
+    width: usize,
+    height: usize,
+    layout: Layout,
+    runner: &dyn TileRunner,
+    p: &Params,
+    ctl: &Control<'_>,
+) -> Result<Rgb32f, Error> {
+    let pixels = width
+        .checked_mul(height)
+        .filter(|&n| n > 0 && n <= MAX_PIXELS)
+        .ok_or_else(|| Error::Input("the picture has no size, or is too large".into()))?;
+    if mosaic.len() != pixels {
+        return Err(Error::Input("the mosaic does not have width × height samples".into()));
+    }
+    if !(16..=2048).contains(&p.tile) || p.overlap >= p.tile.div_ceil(2) || layout.dx > 1 || layout.dy > 1 {
+        return Err(Error::Input("the tile size and overlap do not fit".into()));
+    }
+    if let Gain::MatchMean { nominal, max_deviation } = p.gain
+        && !(nominal.is_finite() && (1e-6..=1e12).contains(&nominal) && max_deviation.is_finite() && (0.0..=0.5).contains(&max_deviation))
+    {
+        return Err(Error::Input("the model gain is not finite or within its supported range".into()));
+    }
+    if p.clip.is_some_and(|v| !v.is_finite() || v <= 0.0) {
+        return Err(Error::Input("the highlight clip must be finite and positive".into()));
+    }
+    let (nx, ny) = layout.cells(width, height);
+    let (sx, sy) = (starts(nx, p.tile, p.overlap), starts(ny, p.tile, p.overlap));
+    let (cols, rows) = (sx.len(), sy.len());
+    if cols == 0 || rows == 0 {
+        return Err(Error::Input("the picture needs too many tiles".into()));
+    }
+    let tiles: Vec<(usize, usize, bool, bool, bool, bool)> = sy
+        .iter()
+        .enumerate()
+        .flat_map(|(j, &y0)| sx.iter().enumerate().map(move |(i, &x0)| (x0, y0, i > 0, i + 1 < cols, j > 0, j + 1 < rows)))
+        .collect();
+    let total = tiles.len();
+
+    let mut acc = allocate(pixels, [0f32; 3])?;
+    let mut wsum = allocate(nx * ny, 0f32)?;
+    let group = p.parallel.clamp(1, 4);
+    let mut done = 0;
+    // where the time goes, printed under `LIGHTCRAFT_PROFILE`
+    let (mut t_wait, mut t_total) = (std::time::Duration::ZERO, std::time::Duration::ZERO);
+    let began = web_time::Instant::now();
+
+    // One group of tiles at a time: pack, run the model, check and bring each to the input's scale, in parallel.
+    // Blending (sequential: the tiles add into one picture, in tile order so the result does not depend on timing) of
+    // the previous group runs while the next group is on the model, so neither the card nor the cores wait for it.
+    type Tile = (usize, usize, bool, bool, bool, bool);
+    let compute = |chunk: &[Tile]| -> Vec<Result<Vec<f32>, Error>> {
+        chunk
+            .par_iter()
+            .map(|&(x0, y0, bx, ax, by, ay)| {
+                let mut input = allocate(4 * p.tile * p.tile, 0f32)?;
+                if !pack_tile(mosaic, width, height, layout, x0 as isize, y0 as isize, p.tile, &mut input) {
+                    return Err(Error::Input("a tile could not be packed".into()));
+                }
+                let mut out = runner.run(&input)?;
+                let side = 2 * p.tile;
+                if out.len() != 3 * side * side {
+                    return Err(Error::Model(format!("{} numbers for a tile, not the {} expected", out.len(), 3 * side * side)));
+                }
+                if out.iter().any(|v| !v.is_finite()) {
+                    return Err(Error::Model("numbers that are not finite".into()));
+                }
+                finish_tile(&input, &mut out, p, (bx, ax, by, ay));
+                Ok(out)
+            })
+            .collect()
+    };
+    let mut blend = |chunk: &[Tile], outs: Vec<Result<Vec<f32>, Error>>| -> Result<(), Error> {
+        for (&(x0, y0, bx, ax, by, ay), r) in chunk.iter().zip(outs) {
+            blend_tile(&r?, &mut acc, &mut wsum, (width, height), (nx, ny), layout, (x0, y0), p, (bx, ax, by, ay));
+            done += 1;
+            if let Some(f) = ctl.progress {
+                f(done, total);
+            }
+        }
+        Ok(())
+    };
+    let mut previous: Option<(&[Tile], Vec<Result<Vec<f32>, Error>>)> = None;
+    for chunk in tiles.chunks(group) {
+        if ctl.cancel.is_some_and(|c| c.load(Ordering::Relaxed)) {
+            return Err(Error::Cancelled);
+        }
+        let (outs, blended) = rayon::join(
+            || compute(chunk),
+            || match previous.take() {
+                Some((c, o)) => {
+                    let started = web_time::Instant::now();
+                    let r = blend(c, o);
+                    (r, started.elapsed())
+                }
+                None => (Ok(()), std::time::Duration::ZERO),
+            },
+        );
+        blended.0?;
+        t_wait += blended.1;
+        previous = Some((chunk, outs));
+    }
+    if let Some((c, o)) = previous.take() {
+        let started = web_time::Instant::now();
+        blend(c, o)?;
+        t_wait += started.elapsed();
+    }
+    t_total += began.elapsed();
+    let started = web_time::Instant::now();
+    // normalise by the weights
+    let (ox, oy) = (layout.position(0, 0).0, layout.position(0, 0).1);
+    acc.par_chunks_mut(width).enumerate().for_each(|(y, row)| {
+        let cy = ((y as isize - oy) / 2) as usize;
+        for (x, px) in row.iter_mut().enumerate() {
+            let cx = ((x as isize - ox) / 2) as usize;
+            let w = wsum.get(cy * nx + cx).copied().unwrap_or(0.0);
+            if w > 0.0 {
+                *px = px.map(|v| v / w);
+            }
+        }
+    });
+    if std::env::var_os("LIGHTCRAFT_PROFILE").is_some() {
+        eprintln!(
+            "[denoise] {total} tiles, {} at once: tiles {:.0} ms (blending, {:.0} ms of it, overlaps the model), normalise {:.0} ms",
+            group,
+            t_total.as_secs_f64() * 1e3,
+            t_wait.as_secs_f64() * 1e3,
+            started.elapsed().as_secs_f64() * 1e3
+        );
+    }
+    Ok(Rgb32f { width, height, data: acc })
+}
+
+/// Bring a tile's output to the input's scale and keep clipped highlights as they were (in place).
+fn finish_tile(input: &[f32], out: &mut [f32], p: &Params, edges: (bool, bool, bool, bool)) {
+    let t = p.tile;
+    let side = 2 * t;
+    if let Gain::MatchMean { nominal, max_deviation } = p.gain {
+        // the part of the tile whose scale is measured: away from the edges that blend into a neighbour
+        let m = p.overlap / 2;
+        let (i0, i1) = (if edges.0 { m } else { 0 }, t - if edges.1 { m } else { 0 });
+        let (j0, j1) = (if edges.2 { m } else { 0 }, t - if edges.3 { m } else { 0 });
+        let cells = ((i1 - i0) * (j1 - j0)) as f64;
+        let plane_sum = |pl: usize| -> f64 {
+            (j0..j1)
+                .map(|j| input.get(pl * t * t + j * t + i0..pl * t * t + j * t + i1).map_or(0.0, |r| r.iter().map(|&v| f64::from(v)).sum::<f64>()))
+                .sum()
+        };
+        let in_mean = [plane_sum(0) / cells, (plane_sum(1) + plane_sum(2)) / (2.0 * cells), plane_sum(3) / cells];
+        for (c, &mean_in) in in_mean.iter().enumerate() {
+            let sum_out: f64 = (2 * j0..2 * j1)
+                .map(|y| {
+                    out.get(c * side * side + y * side + 2 * i0..c * side * side + y * side + 2 * i1)
+                        .map_or(0.0, |r| r.iter().map(|&v| f64::from(v)).sum::<f64>())
+                })
+                .sum();
+            let mean_out = sum_out / (4.0 * cells);
+            let measured = if mean_in as f32 >= MIN_SIGNAL && mean_out > 0.0 { (mean_out / mean_in) as f32 } else { nominal };
+            let g = measured.clamp(nominal * (1.0 - max_deviation), nominal * (1.0 + max_deviation));
+            if g.is_finite() && g > 0.0 {
+                let inv = 1.0 / g;
+                if let Some(plane) = out.get_mut(c * side * side..(c + 1) * side * side) {
+                    plane.iter_mut().for_each(|v| *v *= inv);
+                }
+            }
+        }
+    }
+    if let Some(clip) = p.clip {
+        protect_clipped(input, out, t, clip);
+    }
+}
+
+/// Where the mosaic is at (or within a cell of) white, the output becomes the plain cell values again, so a clipped
+/// area stays exactly as clipped as the camera recorded it (highlight recovery later depends on that).
+fn protect_clipped(input: &[f32], out: &mut [f32], t: usize, clip: f32) {
+    let side = 2 * t;
+    let plane = t * t;
+    let (Some(r), Some(g1), Some(g2), Some(b)) =
+        (input.get(0..plane), input.get(plane..2 * plane), input.get(2 * plane..3 * plane), input.get(3 * plane..4 * plane))
+    else {
+        return;
+    };
+    let mut m: Vec<f32> = (0..plane).map(|i| r[i].max(g1[i]).max(g2[i]).max(b[i])).collect();
+    if !m.iter().any(|&v| v >= clip) {
+        return;
+    }
+    // 3×3 maximum over cells (separable)
+    let dilate = |src: &[f32], horizontal: bool| -> Vec<f32> {
+        (0..plane)
+            .map(|i| {
+                let (x, y) = (i % t, i / t);
+                let mut best = src[i];
+                for d in [-1isize, 1] {
+                    let (nx, ny) = if horizontal { (x as isize + d, y as isize) } else { (x as isize, y as isize + d) };
+                    if nx >= 0 && ny >= 0 && (nx as usize) < t && (ny as usize) < t {
+                        best = best.max(src[ny as usize * t + nx as usize]);
+                    }
+                }
+                best
+            })
+            .collect()
+    };
+    m = dilate(&dilate(&m, true), false);
+    let lo = clip - 0.03;
+    for j in 0..t {
+        for i in 0..t {
+            let k = j * t + i;
+            let s = ((m[k] - lo) / (clip - lo)).clamp(0.0, 1.0);
+            if s <= 0.0 {
+                continue;
+            }
+            let s = s * s * (3.0 - 2.0 * s);
+            let plain = [r[k], 0.5 * (g1[k] + g2[k]), b[k]];
+            for (c, &pv) in plain.iter().enumerate() {
+                for (sy, sx) in [(0, 0), (0, 1), (1, 0), (1, 1)] {
+                    if let Some(v) = out.get_mut(c * side * side + (2 * j + sy) * side + 2 * i + sx) {
+                        *v += (pv - *v) * s;
+                    }
+                }
+            }
+        }
+    }
+}
+
+/// Add a finished tile to the picture with its blend weights.
+#[allow(clippy::too_many_arguments)]
+fn blend_tile(
+    out: &[f32],
+    acc: &mut [[f32; 3]],
+    wsum: &mut [f32],
+    size: (usize, usize),
+    cells: (usize, usize),
+    layout: Layout,
+    at: (usize, usize),
+    p: &Params,
+    edges: (bool, bool, bool, bool),
+) {
+    let (width, height) = size;
+    let (nx, ny) = cells;
+    let (t, side) = (p.tile, 2 * p.tile);
+    let fade = p.overlap;
+    let wx: Vec<f32> = (0..t).map(|i| weight(i, t, fade, edges.0, edges.1)).collect();
+    let wy: Vec<f32> = (0..t).map(|j| weight(j, t, fade, edges.2, edges.3)).collect();
+    for j in 0..t {
+        let cy = at.1 + j;
+        for i in 0..t {
+            let cx = at.0 + i;
+            // cells past the picture's edge (a picture smaller than one tile is padded) belong to no pixel
+            if cx >= nx || cy >= ny {
+                continue;
+            }
+            let (Some(&a), Some(&b)) = (wx.get(i), wy.get(j)) else { continue };
+            let w = a * b;
+            if let Some(s) = wsum.get_mut(cy * nx + cx) {
+                *s += w;
+            }
+            let (px0, py0) = layout.position(cx as isize, cy as isize);
+            for (sy, sx) in [(0isize, 0isize), (0, 1), (1, 0), (1, 1)] {
+                let (x, y) = (px0 + sx, py0 + sy);
+                if x < 0 || y < 0 || x as usize >= width || y as usize >= height {
+                    continue;
+                }
+                let o = (2 * j + sy as usize) * side + 2 * i + sx as usize;
+                let (Some(&r), Some(&g), Some(&bl)) = (out.get(o), out.get(side * side + o), out.get(2 * side * side + o)) else { continue };
+                if let Some(px) = acc.get_mut(y as usize * width + x as usize) {
+                    px[0] += w * r;
+                    px[1] += w * g;
+                    px[2] += w * bl;
+                }
+            }
+        }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    /// A stand-in model: the cell's samples as an RGB picture at twice the resolution (a plain 2 × 2 "demosaic"),
+    /// times `scale`, so the output has a scale of its own like the real model's.
+    struct Mock {
+        tile: usize,
+        scale: f32,
+    }
+    impl TileRunner for Mock {
+        fn run(&self, input: &[f32]) -> Result<Vec<f32>, Error> {
+            let (t, side) = (self.tile, 2 * self.tile);
+            let plane = t * t;
+            let mut out = vec![0.0; 3 * side * side];
+            for j in 0..t {
+                for i in 0..t {
+                    let k = j * t + i;
+                    let rgb = [input[k], 0.5 * (input[plane + k] + input[2 * plane + k]), input[3 * plane + k]];
+                    for (c, v) in rgb.iter().enumerate() {
+                        for (sy, sx) in [(0, 0), (0, 1), (1, 0), (1, 1)] {
+                            out[c * side * side + (2 * j + sy) * side + 2 * i + sx] = v * self.scale;
+                        }
+                    }
+                }
+            }
+            Ok(out)
+        }
+    }
+
+    fn params(tile: usize, overlap: usize, gain: Gain) -> Params {
+        Params { tile, overlap, gain, clip: None, parallel: 2 }
+    }
+
+    #[test]
+    fn overflowing_tiles_and_nan_gain_are_errors_before_the_model_runs() {
+        let layout = Layout::from_cell([0, 1, 1, 2]).unwrap();
+        let model = Mock { tile: 16, scale: 1.0 };
+        for (tile, overlap, gain) in [
+            (usize::MAX, 0, Gain::None),
+            (16, usize::MAX, Gain::None),
+            (16, 0, Gain::MatchMean { nominal: 1.0, max_deviation: f32::NAN }),
+            (16, 0, Gain::MatchMean { nominal: f32::INFINITY, max_deviation: 0.1 }),
+        ] {
+            assert!(matches!(denoise_bayer(&[0.5], 1, 1, layout, &model, &params(tile, overlap, gain), &Control::default()), Err(Error::Input(_))));
+        }
+    }
+
+    /// A smooth picture whose colour at a sample is `base[colour] · (1 + shading)`.
+    fn mosaic(w: usize, h: usize, cell: [u8; 4]) -> Vec<f32> {
+        (0..w * h)
+            .map(|i| {
+                let (x, y) = (i % w, i / w);
+                let c = cell[(y % 2) * 2 + x % 2] as usize;
+                [0.5, 0.3, 0.2][c] * (0.6 + 0.4 * ((x as f32 / 23.0).sin() * (y as f32 / 31.0).cos()))
+            })
+            .collect()
+    }
+
+    fn reference(m: &[f32], w: usize, h: usize, layout: Layout) -> Rgb32f {
+        // what the mock gives for the whole picture in one tile
+        let (nx, ny) = layout.cells(w, h);
+        let t = nx.max(ny).div_ceil(16) * 16;
+        denoise_bayer(m, w, h, layout, &Mock { tile: t, scale: 1.0 }, &params(t, 16, Gain::None), &Control::default()).unwrap()
+    }
+
+    #[test]
+    fn tiles_blend_to_the_same_picture_as_one_tile_for_every_layout() {
+        for cell in [[0u8, 1, 1, 2], [1, 0, 2, 1], [1, 2, 0, 1], [2, 1, 1, 0]] {
+            let layout = Layout::from_cell(cell).unwrap();
+            let (w, h) = (150usize, 110usize);
+            let m = mosaic(w, h, cell);
+            let want = reference(&m, w, h, layout);
+            let got = denoise_bayer(&m, w, h, layout, &Mock { tile: 32, scale: 1.0 }, &params(32, 8, Gain::None), &Control::default()).unwrap();
+            assert_eq!((got.width, got.height), (w, h));
+            for (a, b) in got.data.iter().zip(&want.data) {
+                for c in 0..3 {
+                    assert!((a[c] - b[c]).abs() < 1e-5, "layout {cell:?}: {a:?} vs {b:?}");
+                }
+            }
+        }
+    }
+
+    #[test]
+    fn the_picture_does_not_depend_on_how_many_tiles_run_at_once_or_how_long_each_takes() {
+        // tiles that finish in a different order every time: blending still adds them in tile order
+        struct Jittery(Mock);
+        impl TileRunner for Jittery {
+            fn run(&self, input: &[f32]) -> Result<Vec<f32>, Error> {
+                let wait = (input.iter().take(64).sum::<f32>() * 1000.0) as u64 % 7;
+                std::thread::sleep(std::time::Duration::from_millis(wait));
+                self.0.run(input)
+            }
+        }
+        let layout = Layout { dx: 0, dy: 1 };
+        let (w, h) = (150usize, 110usize);
+        let m = mosaic(w, h, [1, 0, 2, 1]);
+        let run = |parallel: usize| {
+            let mut p = params(32, 8, Gain::MatchMean { nominal: 1.0e6, max_deviation: 0.05 });
+            p.parallel = parallel;
+            denoise_bayer(&m, w, h, layout, &Jittery(Mock { tile: 32, scale: 1.01e6 }), &p, &Control::default()).unwrap()
+        };
+        let one = run(1);
+        for parallel in [2, 3, 8] {
+            assert_eq!(run(parallel).data, one.data, "{parallel} at once");
+        }
+    }
+
+    #[test]
+    fn a_failing_tile_or_a_cancel_ends_the_picture_without_a_panic() {
+        struct Fails(Mock, std::sync::atomic::AtomicUsize);
+        impl TileRunner for Fails {
+            fn run(&self, input: &[f32]) -> Result<Vec<f32>, Error> {
+                if self.1.fetch_add(1, Ordering::Relaxed) == 4 {
+                    return Err(Error::Runtime("boom".into()));
+                }
+                self.0.run(input)
+            }
+        }
+        let layout = Layout { dx: 0, dy: 0 };
+        let (w, h) = (150usize, 110usize);
+        let m = mosaic(w, h, [0, 1, 1, 2]);
+        let r = denoise_bayer(
+            &m,
+            w,
+            h,
+            layout,
+            &Fails(Mock { tile: 32, scale: 1.0 }, Default::default()),
+            &params(32, 8, Gain::None),
+            &Control::default(),
+        );
+        assert!(matches!(r, Err(Error::Runtime(_))), "{r:?}");
+        let stop = AtomicBool::new(true);
+        let ctl = Control { cancel: Some(&stop), progress: None };
+        let r = denoise_bayer(&m, w, h, layout, &Mock { tile: 32, scale: 1.0 }, &params(32, 8, Gain::None), &ctl);
+        assert!(matches!(r, Err(Error::Cancelled)));
+    }
+
+    #[test]
+    fn a_scale_of_its_own_is_taken_out_tile_by_tile() {
+        let layout = Layout { dx: 0, dy: 0 };
+        let (w, h) = (150usize, 110usize);
+        let m = mosaic(w, h, [0, 1, 1, 2]);
+        let want = reference(&m, w, h, layout);
+        let gain = Gain::MatchMean { nominal: 1.0e6, max_deviation: 0.05 };
+        // the scale differs a little from nominal; each tile's own scale is measured and removed
+        let got = denoise_bayer(&m, w, h, layout, &Mock { tile: 32, scale: 1.02e6 }, &params(32, 8, gain), &Control::default()).unwrap();
+        let mean = |img: &Rgb32f, c: usize| img.data.iter().map(|p| f64::from(p[c])).sum::<f64>() / img.data.len() as f64;
+        for c in 0..3 {
+            let (a, b) = (mean(&got, c), mean(&want, c));
+            assert!((a - b).abs() / b < 0.01, "channel {c}: {a} vs {b}");
+        }
+        // a scale far from nominal is clamped rather than trusted
+        let wild = denoise_bayer(&m, w, h, layout, &Mock { tile: 32, scale: 3.0e6 }, &params(32, 8, gain), &Control::default()).unwrap();
+        assert!(mean(&wild, 1) > 2.0 * mean(&want, 1));
+    }
+
+    #[test]
+    fn a_picture_smaller_than_a_tile_is_padded_by_mirroring() {
+        let layout = Layout { dx: 1, dy: 1 };
+        let (w, h) = (20usize, 14usize);
+        let m = mosaic(w, h, [2, 1, 1, 0]);
+        let got = denoise_bayer(&m, w, h, layout, &Mock { tile: 64, scale: 1.0 }, &params(64, 16, Gain::None), &Control::default()).unwrap();
+        assert_eq!((got.width, got.height), (w, h));
+        assert!(got.data.iter().all(|p| p.iter().all(|v| v.is_finite() && *v > 0.0)));
+    }
+
+    #[test]
+    fn clipped_highlights_stay_clipped() {
+        // a model that darkens everything by 10 % would pull a clipped plateau below white
+        struct Dim(Mock);
+        impl TileRunner for Dim {
+            fn run(&self, input: &[f32]) -> Result<Vec<f32>, Error> {
+                Ok(self.0.run(input)?.into_iter().map(|v| v * 0.9).collect())
+            }
+        }
+        let layout = Layout { dx: 0, dy: 0 };
+        let (w, h) = (96usize, 96usize);
+        let m: Vec<f32> = (0..w * h).map(|i| if i % w > 40 && i % w < 70 { 1.0 } else { 0.4 }).collect();
+        let mut p = params(32, 8, Gain::None);
+        p.clip = Some(0.99);
+        let got = denoise_bayer(&m, w, h, layout, &Dim(Mock { tile: 32, scale: 1.0 }), &p, &Control::default()).unwrap();
+        let mid = got.data[48 * w + 55];
+        assert!(mid.iter().all(|&v| v > 0.995), "plateau {mid:?}");
+        let dark = got.data[48 * w + 10];
+        assert!(dark.iter().all(|&v| (v - 0.36).abs() < 0.01), "unclipped area still goes through the model: {dark:?}");
+    }
+
+    #[test]
+    fn cancelling_and_progress_work() {
+        let layout = Layout { dx: 0, dy: 0 };
+        let (w, h) = (150usize, 110usize);
+        let m = mosaic(w, h, [0, 1, 1, 2]);
+        let cancel = AtomicBool::new(true);
+        let r = denoise_bayer(
+            &m,
+            w,
+            h,
+            layout,
+            &Mock { tile: 32, scale: 1.0 },
+            &params(32, 8, Gain::None),
+            &Control { cancel: Some(&cancel), progress: None },
+        );
+        assert!(matches!(r, Err(Error::Cancelled)));
+        let seen = std::sync::Mutex::new(Vec::new());
+        let f = |d: usize, n: usize| seen.lock().unwrap().push((d, n));
+        denoise_bayer(&m, w, h, layout, &Mock { tile: 32, scale: 1.0 }, &params(32, 8, Gain::None), &Control { cancel: None, progress: Some(&f) })
+            .unwrap();
+        let seen = seen.into_inner().unwrap();
+        let total = tile_count(w, h, layout, 32, 8);
+        assert!(total > 1);
+        assert_eq!(seen.last(), Some(&(total, total)));
+        assert!(seen.windows(2).all(|p| p[1].0 == p[0].0 + 1));
+    }
+
+    #[test]
+    fn hostile_inputs_are_errors() {
+        let layout = Layout { dx: 0, dy: 0 };
+        let mock = Mock { tile: 32, scale: 1.0 };
+        let p = params(32, 8, Gain::None);
+        assert!(matches!(denoise_bayer(&[], 0, 0, layout, &mock, &p, &Control::default()), Err(Error::Input(_))));
+        assert!(matches!(denoise_bayer(&[0.0; 10], 4, 4, layout, &mock, &p, &Control::default()), Err(Error::Input(_))));
+        assert!(matches!(denoise_bayer(&[0.0; 16], 4, 4, layout, &mock, &params(32, 16, Gain::None), &Control::default()), Err(Error::Input(_))));
+        struct Bad(Vec<f32>);
+        impl TileRunner for Bad {
+            fn run(&self, _: &[f32]) -> Result<Vec<f32>, Error> {
+                Ok(self.0.clone())
+            }
+        }
+        assert!(matches!(denoise_bayer(&[0.5; 16], 4, 4, layout, &Bad(vec![0.0; 5]), &p, &Control::default()), Err(Error::Model(_))));
+        assert!(matches!(denoise_bayer(&[0.5; 16], 4, 4, layout, &Bad(vec![f32::NAN; 3 * 64 * 64]), &p, &Control::default()), Err(Error::Model(_))));
+    }
+}
@@ -0,0 +1,100 @@
+//! Cutting a picture into overlapping tiles and putting the results back together.
+//!
+//! A model runs on a fixed square, so a picture is covered by tiles spaced evenly along each axis, each sharing
+//! at least `overlap` cells with its neighbours. Each tile's influence fades towards its own edges (except at the
+//! picture's edge), and the result is the sum of tile × weight divided by the sum of weights, so any set of
+//! overlapping tiles blends consistently and a seam between two tiles of slightly different scale is a smooth ramp.
+
+/// Where the tiles start along one axis of `n` cells: `[0]` when the axis fits one tile (the tile is then padded by
+/// mirroring), else evenly spaced starts, the first at 0 and the last at `n − tile`, each tile sharing at least
+/// `overlap` cells with the next.
+pub fn starts(n: usize, tile: usize, overlap: usize) -> Vec<usize> {
+    if n <= tile || tile == 0 {
+        return vec![0];
+    }
+    let step = tile.saturating_sub(overlap).max(1);
+    // tiles needed so that start k·spacing with spacing ≤ step reaches n − tile
+    let Some(k) = (n - tile).div_ceil(step).checked_add(1).filter(|&k| k <= 1_000_000) else { return Vec::new() };
+    let last = n - tile;
+    (0..k).map(|i| if k == 1 { 0 } else { ((i as u128 * last as u128 + (k - 1) as u128 / 2) / (k - 1) as u128) as usize }).collect()
+}
+
+/// Weight of the cell at offset `p` (0-based, inside a tile of `tile` cells) along one axis: 1 in the middle, easing
+/// down to a sliver over `fade` cells at an edge that has a neighbour, 1 right up to an edge that is the picture's.
+pub fn weight(p: usize, tile: usize, fade: usize, has_before: bool, has_after: bool) -> f32 {
+    let fade = fade.max(1) as f32;
+    let ramp = |d: usize| {
+        let t = ((d as f32 + 0.5) / fade).min(1.0);
+        // smoothstep, never exactly 0 so that a cell is always covered
+        (t * t * (3.0 - 2.0 * t)).max(1e-3)
+    };
+    let before = if has_before { ramp(p) } else { 1.0 };
+    let after = if has_after { ramp(tile.saturating_sub(1).saturating_sub(p)) } else { 1.0 };
+    before.min(after)
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn hostile_axes_have_bounded_tile_counts_and_checked_spacing() {
+        assert!(starts(usize::MAX, 1, 0).is_empty());
+        assert!(starts(usize::MAX, 2, usize::MAX).is_empty());
+        let s = starts(usize::MAX, usize::MAX / 2, 0);
+        assert_eq!(s.first(), Some(&0));
+        assert_eq!(s.last(), Some(&(usize::MAX - usize::MAX / 2)));
+        assert!(s.windows(2).all(|s| s[0] < s[1]));
+    }
+
+    #[test]
+    fn one_tile_when_it_fits() {
+        assert_eq!(starts(100, 512, 64), vec![0]);
+        assert_eq!(starts(512, 512, 64), vec![0]);
+        assert_eq!(starts(0, 512, 64), vec![0]);
+        assert_eq!(starts(10, 0, 0), vec![0]);
+    }
+
+    #[test]
+    fn tiles_cover_the_axis_and_overlap_enough() {
+        for n in [513usize, 600, 900, 1000, 1024, 1500, 3024, 3600, 5000, 12345] {
+            let s = starts(n, 512, 64);
+            assert_eq!(s[0], 0, "n={n}");
+            assert_eq!(*s.last().unwrap(), n - 512, "n={n}");
+            for w in s.windows(2) {
+                assert!(w[1] > w[0], "starts increase, n={n}");
+                assert!(w[0] + 512 >= w[1] + 64, "at least 64 shared cells between {} and {}, n={n}", w[0], w[1]);
+            }
+        }
+    }
+
+    #[test]
+    fn weights_fade_only_towards_neighbours() {
+        let (t, f) = (512, 64);
+        // a lone tile: full weight everywhere
+        assert!((0..t).all(|p| weight(p, t, f, false, false) == 1.0));
+        // the first tile fades on its far side only
+        assert_eq!(weight(0, t, f, false, true), 1.0);
+        assert!(weight(t - 1, t, f, false, true) < 0.01);
+        // the last tile fades on its near side only
+        assert!(weight(0, t, f, true, false) < 0.01);
+        assert_eq!(weight(t - 1, t, f, true, false), 1.0);
+        // a middle tile is 1 in the middle and everywhere positive
+        assert_eq!(weight(t / 2, t, f, true, true), 1.0);
+        assert!((0..t).all(|p| weight(p, t, f, true, true) > 0.0));
+    }
+
+    #[test]
+    fn two_overlapping_tiles_blend_to_a_partition_of_unity() {
+        // at every cell of the shared region the normalised weights are in [0, 1] and sum to 1
+        let (t, f) = (512usize, 64usize);
+        let s = starts(900, t, 64);
+        assert_eq!(s.len(), 2);
+        for cell in s[1]..s[0] + t {
+            let a = weight(cell - s[0], t, f, false, true);
+            let b = weight(cell - s[1], t, f, true, false);
+            let (na, nb) = (a / (a + b), b / (a + b));
+            assert!((na + nb - 1.0).abs() < 1e-6 && na >= 0.0 && nb >= 0.0);
+        }
+    }
+}
@@ -0,0 +1,32 @@
+[package]
+name = "lightcraft-denoise"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
+description = "LightCraft AI denoise (L1): denoise-model manifests, Bayer packing, tiling and blending, the cache file, and pure-Rust CPU inference"
+
+[lints]
+workspace = true
+
+[features]
+default = []
+runtime = ["dep:faer"]
+# Product-policy switch: absent by default; users may install their own model.
+rawnind-model = ["lightcraft-denoise-core/rawnind-model"]
+
+[dependencies]
+lightcraft-denoise-core = { workspace = true }
+# Safe matrix views over the pure-Rust GEMM backend; no BLAS, cc/build scripts
+# or inner thread pool. Its dependencies include architecture-specific assembly.
+faer = { package = "faer-core", version = "=0.17.1", default-features = false, features = ["std"], optional = true }
+lightcraft-raster = { workspace = true }
+serde = { workspace = true }
+serde_json = { workspace = true }
+sha2 = { workspace = true }
+thiserror = { workspace = true }
+miniz_oxide = { workspace = true }
+rayon = { workspace = true }
+web-time = { workspace = true }
+
+[dev-dependencies]
+proptest = { workspace = true }
@@ -0,0 +1,18 @@
+# lightcraft-denoise
+
+L1 pure-Rust AI RAW denoise primitives and CPU inference. No weights are bundled.
+See [user workflow and limitations](../../docs/denoise.md).
+
+`lightcraft-denoise-core` provides metadata, Bayer helpers and cache formats,
+re-exported here. `manifest` and `known` validate opt-in models; `onnx` reads a bounded operator subset
+into `net`; `cpu` executes it and `reference` supplies scalar checks. `bayer`, `tiles`
+and `run` pack sensor data and blend overlapping tiles while preserving clipped
+highlights. `product` reads/writes the disposable `.lcdn` cache. `runtime` supplies
+the installation self-test. Synthetic networks are generated in Rust for tests.
+
+Run `cargo test -p lightcraft-denoise --features runtime`. The CPU backend is optional;
+featureless and web builds keep the model metadata and cache format without faer/GEMM.
+The `rawnind-model` feature enables the otherwise empty download catalog and requires
+an explicit release-policy decision. Real-model tests are ignored by default and
+require an explicitly supplied local model. The GPU twin is `lightcraft-gpu::nn`;
+model management, background jobs and export integration live in `lightcraft-engine`.
				`@@ -0,0 +1 @@`
				{"version":1,"model":"ILCE-7M4","files":597,"samples":2182876,"matrix":[[1.06975,-0.0709886,-0.000770444],[-0.0169822,1.08012,-0.0651276],[-0.114835,-0.601525,1.74804]],"hue_sat":{"hue_divisions":72,"sat_divisions":5,"val_divisions":5,"data":[[0.0,1.0,1.0],[-0.819008,1.14375,1.0],[-6.07446,1.10974,1.0],[-4.0478,1.04285,1.0],[-0.546764,1.00329,1.0],[0.0,1.0,1.0],[-1.0412,1.10597,1.0],[-7.22645,1.05878,1.0],[-7.30635,0.989387,1.0],[-5.18589,0.974522,1.0],[0.0,1.0,1.0],[-1.02364,1.06303,1.0],[-6.94669,1.01579,1.0],[-7.47614,0.955435,1.0],[-7.02579,0.940854,1.0],[0.0,1.0,1.0],[-0.29852,1.04349,1.0],[-6.37202,1.00021,1.0],[-7.52098,0.935457,1.0],[-7.45702,0.922239,1.0],[0.0,1.0,1.0],[0.367464,1.03266,1.0],[-4.96951,0.996822,1.0],[-7.41706,0.948709,1.0],[-8.04735,0.91672,1.0],[0.0,1.0,1.0],[0.0388419,1.02587,1.0],[-4.47266,1.01282,1.0],[-7.26834,0.987279,1.0],[-9.02713,0.921453,1.0],[0.0,1.0,1.0],[-0.297476,1.02333,1.0],[-4.6323,1.00993,1.0],[-8.57248,0.962973,1.0],[-12.2375,0.872088,1.0],[0.0,1.0,1.0],[-0.785867,1.02786,1.0],[-4.73124,1.00876,1.0],[-9.09856,0.955343,1.0],[-12.4324,0.875472,1.0],[0.0,1.0,1.0],[-0.444072,1.03863,1.0],[-3.94773,1.01784,1.0],[-7.97996,0.975996,1.0],[-10.673,0.917401,1.0],[0.0,1.0,1.0],[1.22651,1.0559,1.0],[-1.89459,1.02934,1.0],[-5.54545,0.993046,1.0],[-8.48877,0.941628,1.0],[0.0,1.0,1.0],[3.24516,1.08292,1.0],[1.20522,1.0577,1.0],[-0.826224,1.01668,1.0],[-3.71884,0.946733,1.0],[0.0,1.0,1.0],[4.68802,1.12039,1.0],[3.23147,1.08619,1.0],[1.84476,1.04,1.0],[0.0970375,0.962298,1.0],[0.0,1.0,1.0],[5.36041,1.16313,1.0],[4.5937,1.10933,1.0],[3.83314,1.05056,1.0],[3.122,0.959876,1.0],[0.0,1.0,1.0],[5.22222,1.17824,1.0],[5.09129,1.1142,1.0],[4.99294,1.04813,1.0],[5.01485,0.957528,1.0],[0.0,1.0,1.0],[4.34382,1.18446,1.0],[4.9457,1.11351,1.0],[5.56715,1.04462,1.0],[6.27579,0.961892,1.0],[0.0,1.0,1.0],[3.5428,1.18985,1.0],[4.9116,1.11394,1.0],[6.11507,1.04098,1.0],[7.03638,0.964331,1.0],[0.0,1.0,1.0],[2.92641,1.20663,1.0],[4.57132,1.1281,1.0],[5.96964,1.04895,1.0],[6.98579,0.96261,1.0],[0.0,1.0,1.0],[2.07123,1.23257,1.0],[3.51736,1.15454,1.0],[4.85051,1.06995,1.0],[5.97805,0.971597,1.0],[0.0,1.0,1.0],[0.869122,1.26762,1.0],[2.22009,1.19021,1.0],[3.54297,1.10536,1.0],[4.31783,1.00254,1.0],[0.0,1.0,1.0],[-0.409355,1.31042,1.0],[0.778039,1.23845,1.0],[1.87845,1.15455,1.0],[1.53759,1.02961,1.0],[0.0,1.0,1.0],[-1.82067,1.35265,1.0],[-0.910751,1.28551,1.0],[-0.100378,1.18585,1.0],[0.0425375,1.01463,1.0],[0.0,1.0,1.0],[-3.26659,1.391,1.0],[-2.50117,1.33263,1.0],[-1.55789,1.18102,1.0],[-0.0853484,1.00594,1.0],[0.0,1.0,1.0],[-4.92446,1.43632,1.0],[-3.92066,1.35798,1.0],[-1.66274,1.11605,1.0],[-0.0358972,1.00178,1.0],[0.0,1.0,1.0],[-5.82474,1.46137,1.0],[-4.22466,1.34716,1.0],[-0.64525,1.04813,1.0],[-0.00357121,1.00023,1.0],[0.0,1.0,1.0],[-7.0672,1.47242,1.0],[-5.38667,1.32663,1.0],[-0.553542,1.02668,1.0],[-0.00188478,1.00008,1.0],[0.0,1.0,1.0],[-8.37522,1.43136,1.0],[-6.13631,1.28174,1.0],[-0.515332,1.01808,1.0],[-0.0015762,1.00005,1.0],[0.0,1.0,1.0],[-10.5196,1.39935,1.0],[-7.49316,1.24442,1.0],[-0.517621,1.01282,1.0],[-0.001445,1.00003,1.0],[0.0,1.0,1.0],[-10.5952,1.3534,1.0],[-7.81548,1.20414,1.0],[-0.63473,1.0111,1.0],[-0.00208267,1.00003,1.0],[0.0,1.0,1.0],[-10.879,1.32051,1.0],[-8.3374,1.18517,1.0],[-0.91817,1.01139,1.0],[-0.00474972,1.00003,1.0],[0.0,1.0,1.0],[-11.8094,1.29525,1.0],[-9.18818,1.17522,1.0],[-0.968552,1.01157,1.0],[-0.00411555,1.00003,1.0],[0.0,1.0,1.0],[-12.2038,1.27777,1.0],[-8.97374,1.15856,1.0],[-0.751018,1.00872,1.0],[-0.00250012,1.00002,1.0],[0.0,1.0,1.0],[-12.0564,1.2709,1.0],[-8.38288,1.15926,1.0],[-0.643806,1.00878,1.0],[-0.00202184,1.00002,1.0],[0.0,1.0,1.0],[-12.8403,1.25618,1.0],[-8.85664,1.14517,1.0],[-0.646111,1.00756,1.0],[-0.00195373,1.00001,1.0],[0.0,1.0,1.0],[-13.2604,1.25094,1.0],[-8.74013,1.11989,1.0],[-0.556009,1.00457,1.0],[-0.00173404,1.00001,1.0],[0.0,1.0,1.0],[-13.3939,1.24388,1.0],[-8.23916,1.11659,1.0],[-0.376536,1.00355,1.0],[-0.000916476,1.00001,1.0],[0.0,1.0,1.0],[-13.6578,1.24212,1.0],[-7.97883,1.11035,1.0],[-0.409859,1.00287,1.0],[-0.0018203,1.0,1.0],[0.0,1.0,1.0],[-15.7225,1.22414,1.0],[-10.2345,1.10217,1.0],[-0.968177,1.00427,1.0],[-0.00650675,1.00001,1.0],[0.0,1.0,1.0],[-16.1557,1.16423,1.0],[-12.721,1.06603,1.0],[-2.02337,1.00334,1.0],[-0.0134173,1.00001,1.0],[0.0,1.0,1.0],[-17.253,1.09714,1.0],[-15.785,1.02184,1.0],[-4.02264,0.996649,1.0],[-0.0279292,0.999956,1.0],[0.0,1.0,1.0],[-17.6206,1.04331,1.0],[-16.2039,0.99401,1.0],[-4.09728,0.990246,1.0],[-0.0269801,0.999896,1.0],[0.0,1.0,1.0],[-17.2302,0.987569,1.0],[-15.1656,0.964114,1.0],[-3.21398,0.983725,1.0],[-0.0202065,0.999807,1.0],[0.0,1.0,1.0],[-16.2184,0.943678,1.0],[-13.4633,0.938141,1.0],[-3.40639,0.971571,1.0],[-0.00923182,0.999563,1.0],[0.0,1.0,1.0],[-14.5653,0.909699,1.0],[-11.6873,0.915733,1.0],[-5.4654,0.942337,1.0],[-0.124331,0.998194,1.0],[0.0,1.0,1.0],[-11.5122,0.906891,1.0],[-9.31547,0.923874,1.0],[-6.12222,0.939832,1.0],[-0.337538,0.995963,1.0],[0.0,1.0,1.0],[-7.77619,0.92294,1.0]Line truncated