Merge main into fix/issue-102-format-version

Issue #101: catalog.log: cut a partial append back before retrying

crates/catalog/src/journal.rs

@@ -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.
@@ -10,6 +11,12 @@
 //! 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
@@ -27,6 +34,32 @@
 //!
 //! 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` |
+//!
+//! 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};
@@ -34,7 +67,9 @@ 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 = 2;
 
 /// When [`Journal::wants_snapshot`] says it's time to compact the log.
 #[derive(Clone, Copy, Debug)]
@@ -62,10 +97,16 @@ 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
@@ -140,6 +181,9 @@ pub struct Journal {
     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.
@@ -168,18 +212,58 @@ 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
@@ -205,10 +289,16 @@ fn write_snapshot(store: &mut dyn Store, seq: u64, catalog: &Catalog) -> std::io
     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,
 }
@@ -221,15 +311,20 @@ 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 {
@@ -242,6 +337,7 @@ impl Journal {
             stats: PersistStats::default(),
             pending: None,
             retry_at_records: 0,
+            tail_dirty: false,
         };
         let log = log.unwrap_or_default();
 
@@ -261,7 +357,22 @@ 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;
@@ -277,7 +388,15 @@ 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);
                     }
@@ -301,10 +420,34 @@ 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))
@@ -323,7 +466,20 @@ 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;
@@ -343,12 +499,22 @@ impl Journal {
 
     /// Write a snapshot of `catalog` (which must reflect every appended op) and reset the log.
     pub fn snapshot(&mut self, catalog: &Catalog) -> Result<()> {
+        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 mut timing = match write_snapshot(self.store.as_mut(), self.seq, catalog) {
+        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
@@ -357,6 +523,9 @@ impl Journal {
                 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)?;
@@ -365,10 +534,10 @@ impl Journal {
         timing.blocking_ms = timing.total_ms;
         timing.records = self.log_records;
         self.record_snapshot(timing);
-        self.snapshot_seq = self.seq;
         self.log_records = 0;
         self.log_bytes = 0;
         self.retry_at_records = 0;
+        self.tail_dirty = false;
         Ok(())
     }
 
@@ -459,6 +628,7 @@ impl Journal {
         self.log_records = p.tail_records;
         self.log_bytes = p.tail.len() as u64;
         self.retry_at_records = 0;
+        self.tail_dirty = false;
         Ok(())
     }
 

crates/catalog/src/lib.rs

@@ -5,6 +5,11 @@
 //!   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)]
 
@@ -47,6 +52,10 @@ 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>;
@@ -644,6 +653,10 @@ mod tests;
 #[cfg(test)]
 mod tests_background;
 #[cfg(test)]
+mod tests_format_version;
+#[cfg(test)]
 mod tests_journal;
 #[cfg(test)]
 mod tests_local;
+#[cfg(test)]
+mod tests_torn_append;

crates/catalog/src/store.rs

@@ -28,9 +28,12 @@ pub trait Store: Send {
         self.write_atomic(name, &buf)?;
         Ok(buf.len() as u64)
     }
-    /// Append and make durable (fsync) before returning.
+    /// 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;
@@ -132,15 +135,26 @@ impl Store for FsStore {
             self.appender = Some((name.to_string(), f));
         }
         let Some((_, f)) = self.appender.as_mut() else { return Err(io::Error::other("appender not open")) };
-        f.write_all(data)?;
-        f.sync_data()
+        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()
     }

crates/catalog/src/tests_background.rs

@@ -180,7 +180,8 @@ fn failed_background_snapshot_loses_nothing_and_retries_later() {
     wait_written(&j);
     assert!(j.poll().is_err());
     assert_eq!((j.snapshot_seq(), j.stats().failed_snapshots), (0, 1));
-    assert!(g.files.get(SNAPSHOT).is_none());
+    // 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());

crates/catalog/src/tests_format_version.rs

@@ -0,0 +1,186 @@
+//! 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));
+}
+
+#[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,
+        }
+    }
+    let newest = since(&Op::SetBrowsed { folder: String::new(), at: None });
+    assert_eq!(newest, VERSION, "the newest op's version must be the current format version");
+}

crates/catalog/src/tests_journal.rs

@@ -283,7 +283,12 @@ fn streamed_snapshot_keeps_the_on_disk_format() {
     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\":1,\"seq\":{},\"catalog\":{}}}\n", j.seq(), c.to_snapshot());
+    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);
 

crates/catalog/src/tests_torn_append.rs

@@ -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);
+}

crates/engine/src/library.rs

@@ -362,16 +362,35 @@ impl Session {
     }
 
     /// Flush everything and write a snapshot (on quit). Ends an open interaction first.
+    ///
+    /// The snapshot is written even when appending the queued ops fails: it is a fresh file
+    /// holding everything in memory, so it saves those ops too (the log handle may be the only
+    /// thing that's broken). Fails only if nothing could be saved; the ops then stay queued.
     pub fn close_library(&mut self) -> Result<()> {
         if self.library.is_none() {
             return Ok(());
         }
         let _ = self.end_interaction();
-        self.persist()?;
+        let persisted = self.persist();
         self.save_view();
+        let unlogged = if persisted.is_err() { self.pending_log.len() as u64 } else { 0 };
         let Some(lib) = self.library.as_mut() else { return Ok(()) };
-        lib.journal.snapshot(&self.catalog)?;
-        Ok(())
+        let before = lib.journal.seq();
+        let snapshot = lib.journal.snapshot_with_unlogged(&self.catalog, unlogged);
+        if unlogged > 0 && lib.journal.seq() == before + unlogged {
+            // the snapshot holds the queued ops: they are saved, never append them again
+            log::info!("library: {unlogged} queued change(s) saved in the closing snapshot");
+            self.pending_log.clear();
+            lib.unsaved_error = None;
+            lib.retry_at = None;
+            lib.last_error = None;
+        }
+        match (persisted, snapshot) {
+            (_, Ok(())) => Ok(()),
+            // neither the log nor the snapshot took the queued ops
+            (Err(EngineError::NotSaved(e)), Err(s)) if lib.journal.seq() == before => Err(EngineError::NotSaved(format!("{e}; snapshot: {s}"))),
+            (_, Err(s)) => Err(s.into()),
+        }
     }
 
     fn view_json(&self) -> Vec<u8> {

crates/engine/src/tests_persist.rs

@@ -17,6 +17,8 @@ struct Flaky {
     fail: Arc<AtomicBool>,
     /// Only snapshots fail (appends work).
     snapshots_only: bool,
+    /// Only appends fail (a dead log handle; fresh files can still be written).
+    appends_only: bool,
 }
 
 impl Flaky {
@@ -39,7 +41,7 @@ impl Store for Flaky {
         self.files.read(name)
     }
     fn write_atomic(&mut self, name: &str, data: &[u8]) -> std::io::Result<()> {
-        if self.failing() {
+        if self.failing() && !self.appends_only {
             return Err(std::io::Error::other("disk full"));
         }
         self.files.write_atomic(name, data)
@@ -145,3 +147,47 @@ fn failed_snapshot_does_not_fail_the_command() {
     store.set_failing(false);
     assert_eq!(open(&store, false).catalog.albums().count(), 2);
 }
+
+/// Issue #101: quitting while the log can't be appended to still saves the queued changes in the
+/// closing snapshot (a fresh file), and reports success; they are not appended again later.
+#[test]
+fn close_saves_queued_changes_in_the_snapshot_when_appends_fail() {
+    let store = Flaky { appends_only: true, ..Default::default() };
+    let mut s = open(&store, true);
+    let id = s.selection.active.unwrap();
+    store.set_failing(true);
+    assert!(matches!(s.execute("photo.rate", &json!({"rating": 4})), Err(EngineError::NotSaved(_))));
+    assert!(matches!(s.execute("photo.flag", &json!({"flag": "pick"})), Err(EngineError::NotSaved(_))));
+    assert_eq!(s.unsaved().map(|u| u.0), Some(2));
+    s.close_library().unwrap();
+    assert_eq!(s.unsaved(), None);
+    assert!(s.pending_log.is_empty());
+    let expect = s.catalog.to_snapshot();
+    drop(s);
+    store.set_failing(false);
+    let mut s2 = open(&store, true);
+    assert_eq!(s2.catalog.to_snapshot(), expect);
+    let p = s2.catalog.photo(id).unwrap();
+    assert_eq!((p.rating, p.flag), (4, Flag::Pick));
+    // later edits continue after the absorbed ops
+    s2.execute("photo.rate", &json!({"rating": 2})).unwrap();
+    let expect = s2.catalog.to_snapshot();
+    drop(s2);
+    assert_eq!(open(&store, true).catalog.to_snapshot(), expect);
+}
+
+/// When neither the log nor the snapshot can be written, closing fails with `NotSaved` and the
+/// changes stay queued (a later retry can still save them).
+#[test]
+fn close_fails_when_nothing_can_be_saved() {
+    let store = Flaky::default();
+    let mut s = open(&store, true);
+    store.set_failing(true);
+    assert!(s.execute("photo.rate", &json!({"rating": 4})).is_err());
+    let e = s.close_library().unwrap_err();
+    assert!(matches!(e, EngineError::NotSaved(_)), "{e:?}");
+    assert_eq!(s.unsaved().map(|u| u.0), Some(1));
+    store.set_failing(false);
+    s.close_library().unwrap();
+    assert_eq!(s.unsaved(), None);
+}

docs/parity.md

@@ -488,7 +488,7 @@ under the UI command that wraps it (a no-op error); the UI command now wins.
 
 | Id | Feature | Tier | Status | Evidence | Missing / notes |
 |---|---|---|---|---|---|
-| LR-BEHAV-AUTOSAVE | Instant autosave | P0 | ✅ | `crates/catalog/src/journal.rs`, `crates/engine/src/library.rs` | every change fsynced before the command returns; a failed write fails the command ("saved in memory but not written to disk"), keeps the change queued and retries; top-bar warning until saved |
+| LR-BEHAV-AUTOSAVE | Instant autosave | P0 | ✅ | `crates/catalog/src/journal.rs`, `crates/engine/src/library.rs` | every change fsynced before the command returns; a failed write fails the command ("saved in memory but not written to disk"), keeps the change queued and retries; top-bar warning until saved; a write that fails part-way is cut back to the last whole record so the retry never reads back as damage (logs from older builds with such a fragment load in full); quitting saves queued changes in the closing snapshot even when the log can't be appended to; versioned catalog format (`journal::VERSION`): older libraries are upgraded on open, a library from a newer LightCraft is refused untouched ("written by a newer version") |
 | LR-BEHAV-UNDO | Global undo | P0 | ✅ | `cmd:edit.undo`, `crates/engine/src/tests.rs` (`rating_flag_undo_redo`) | covers ratings, albums, deletes, edits |
 | LR-BEHAV-MULTISELECT | Multi-selection | P0 | ✅ | `cmd:library.select` (`replace`/`add`/`toggle`/`range`), `cmd:library.selectAll` | |
 | LR-BEHAV-BATCH | Batch apply to selection | P0 | ✅ | `cmd:photo.rate`, `cmd:develop.paste`, `cmd:preset.apply`, `cmd:app.export` | |

docs/whats-new.md

@@ -31,6 +31,14 @@
   sidecar that isn't valid XMP is copied to `<name>.xmp.bak-<time>` first. With the default stem naming, a raw and a
   JPEG with the same name (`IMG_0001.CR3` + `IMG_0001.JPG`) no longer share one sidecar: the raw keeps `IMG_0001.xmp`,
   the JPEG uses `IMG_0001.JPG.xmp`.
+- A save that fails part-way (a full disk, a network share that drops) no longer looks like a damaged catalog
+  afterwards (issue #101): the partial write is cut off before LightCraft retries, so the next launch replays every
+  change. Catalogs already holding such a fragment load in full. Quitting while the catalog log can't be written
+  still saves your queued changes in the closing snapshot.
+- The catalog has a format version (issue #102). Opening a library from an older LightCraft upgrades it; a library
+  written by a newer LightCraft is refused with "this library was written by a newer version of LightCraft" and left
+  untouched — older versions no longer read part of it as a damaged log. Once this version has opened a library,
+  LightCraft 0.2.0 and older refuse it ("unsupported format … v2").
 - Canon CR2 photos from the EOS 7D, 50D, 60D, 550D, 600D, 1200D, 1300D, 5D Mark II and 1D Mark IV (and other
   models whose sensor starts on a green-blue row) no longer come out magenta (issue #85): the colour-filter
   layout is read from each file instead of assumed.