With Snap To › Layers on, other layers, selections, shapes and boards now snap
to an artboard's edges and centre, as in Photoshop. An artboard is a group, and
groups weren't snap targets, so a board offered nothing; its children still
offer their own edges. A board being moved (in the exclude list) doesn't snap
to itself.
Test: an_artboard_offers_its_own_edges.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The shell checks an opened document's missing fonts on the UI thread.
missing_fonts locked the shared text engine even for a document with no
type layers, and the engine's first use waits for the installed-font scan
started at launch, which takes minutes on some machines: the window froze
at 2% or 95% on the first open after every launch. A document without type
now returns at once, and a background open with type layers waits for the
scan on its worker ("Loading fonts") so the later UI-thread check is quick.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
DDS textures can now be saved, not just opened: uncompressed (BGRA, BGRX, L8, A8L8; lossless) or BC1/BC2/BC3 (DXT1/3/5) from a clean-room block encoder, with an optional box-filtered mip chain, in the legacy header. Save As and Save a Copy ask with DDS Options; scripts pass `ddsCompression`/`ddsMipmaps` (file commands, control `app.save`, RPC `doc.save`, MCP `doc_save`/`doc_export`), and the CLI takes `--dds-compression`/`--dds-mipmaps`. Targa and DDS answers travel together as `FormatChoices` through the Save As flow.
Review follow-ups:
- File › Save behind DDS or Targa Options records a plain Save with its answer, and replaying `file.save` applies it (#3531).
- The block encoder finds the principal axis from every channel, so a channel that varies on its own can't hide the others' variation.
- DDS joins the shared round-trip suite; the header is built before the output buffer is allocated.
- Enter on a focused Cancel in DDS, Targa or TIFF Options cancels instead of saving.
On wasm32 a near-1 GiB .abr arrives as base64 and is copied several times
before decoding; with a 1 GiB sample budget the import could pass 4 GiB and
abort the tab instead of failing cleanly. Pick the decoder budget by pointer
width (like MAX_DECODED_BYTES) and derive the engine's file cap from it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Found while testing crates/engine/src/file_cmds.rs:690 (xml_unescape), used by li_values, find_attr and read_file_info
Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
An average of the four tones left Black alone light gray (0.67 where
Calculations reads 0.13); the channel now gets the same luminosity
Calculations reads. Other modes are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Found while testing crates/engine/src/file_cmds.rs:768
Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
The round trip compared `as_f64` on both sides, which is `None` for a
string, so the assertion held whatever came back for `size`, `color` and
`location`. Numbers are still compared by value -- one can go out as a
float and come back as an integer -- and anything else is compared as it
is. The test passes, so the strings were being preserved; they just
were not being checked.
Found in review of #7742.
With one layer selected and no pixel selection, Align used the canvas
bounds, so a layer inside an artboard jumped out of its board (to the
middle of the canvas, between other boards). As in Photoshop, it now
aligns to the artboard holding it; a layer outside any artboard, and an
artboard itself, still align to the canvas. An explicit `to` is unchanged.
The first pass sent every Kind::Long field through long_exact -- all 26
of them -- on the strength of three reports. One of those three turned
out to need the opposite: edit.stroke ceils its width, so 2.4 renders at
3 and the integer field can hold the result faithfully. Refusing it
would drop a step that is expressible, which is why #7696 normalises it
instead.
So the refusal is now a field's own kind, LongExact, carried by the two
fields whose command uses the setting as written: Crystallize's ClSz
(clamp 3..300, no rounding) and Ripple's Amnt (clamp -999..999, no
rounding). The other 24 keep long() until someone has read each against
its command.
`Kind::Long` wrote its param through `long()`, which rounds. Photoshop's field
really is an integer, so a fractional setting cannot be written -- but rounding
it put a different edit in the file than the one recorded, with `skipped` empty
and nothing to say so. Measured through `actions.save` and back:
filter.pixelate.crystallize cellSize 5.4 -> ClSz 5 -> cellSize 5
filter.distort.ripple amount 81.4 -> Amnt 81 -> amount 81
edit.stroke width 2.4 -> Wdth 2 -> width 2
The reports measure what that does to the pixels: 385 to 583 differing colour
components at every bit depth, with import reporting no warnings at all.
`long_exact` returns `None` for a value the field cannot hold -- a fraction, or
a magnitude past `i32` -- and `fields_to` passes that up, so the step takes the
path a step with no Photoshop equivalent already takes: strict export names it
and fails, `skipUnsupported` leaves it out and lists it. Refusing rather than
exporting a near miss is what the rest of the file does; `pixel_target` already
refuses a filter aimed at a mask because the plain event has no field for one.
The strict error now covers both cases: a step with no equivalent event, and a
step with a setting its event cannot hold.
`long()` stays for the hand-written mappings that use it (Levels input/output,
`Thsh`, `Tlrn`, `Qlty`, `NCol`, `Dpth`, the Color Balance triples): each is its
own command scope, and most take whole numbers from a range already clamped.
Fixes#7669, #7670, #7671.
- Only a font picked from the open menu joins the recent fonts; the wheel and the
arrow keys on the closed field step past many families and no longer fill the list.
- The style picked under a family's ">" is applied in the Glyphs panel and in
Character/Paragraph Style Options, as the options bar and Character panel already did.
- Monospace by name matches whole words and only where the font's tables say nothing,
so "Encode Sans" is a sans serif, not a "code" font.
- The favourite and recent-font commands are no longer recorded into Actions, so
playing an action back doesn't rewrite the user's font lists.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Retain upstream contributor entries and Ratul’s original entry while
refreshing the UI scale presets branch against current main. Preserve
all scale presets, catalog labels, tests, documentation and coauthor history.
Validation: 5,290 engine/UI tests passed, 32 ignored; fmt, Clippy, layers,
wasm, scorecard generation/check and panic_hunt passed.