Review round (CodeRabbit + an adversarial review, handled together):
- The launch list starts with every row "Checking…" and a worker per path checks it (a
share whose server is gone, or a sleeping disk, can take seconds or never answer; one
that hangs doesn't hold the others). After 5 s a row says the drive isn't answering;
it can still be opened. Once every answer is in, the selection moves to the first
library that can be opened, unless the person moved it.
- File ▸ Open Recent Library and `app.recentLibraries` no longer look at folders either
(the menu is rebuilt every frame): they use the same answers, refreshed after 30 s, when
a library is opened, removed from the list, or fails to open. `app.recentLibraries`
gives `exists: null` and `checking` until a check answered.
- Reasons are found on the worker and put into words on the UI thread, in the interface
language. A check whose worker dies ends as "This library couldn't be checked".
- Each check holds an activity row ("Checking recent libraries"; quick ones don't show).
- Docs: control protocol (`checking`, `available`, `exists`), What's New; fr strings.
Tests: slow check never holds the window; late answers keep the person's selection; the
menu takes its answers from a worker (fails with the old is_dir()); reasons in French; one
hung drive doesn't hold the others; a dying check ends; reopening clears a missing answer.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Bash scripts use [[ ]] for tests, functions name their positional
parameters and end with an explicit return, and the app-specific packaging
check writes its ::error:: annotations to stderr. No behaviour change.
Review follow-up on #2685. CodeRabbit: the library-switch check ran for Merge only, so a window left over from before a switch could still make a library from defaults that described the old one; it now covers both choices (new test). The Mac session: on macOS the switch's own notice can replace the error toast before the test reads it, so the stale-window tests check confirm's error directly. docs/gaps.md gap 13 updated with the parity row (CodeRabbit).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AbegtYaPuwqZ81pGfELsFK
Dropping a wgpu buffer after the device's thread-locals are gone can
abort; the device owns the memory either way.
🤖 Generated with Codebuff
Co-Authored-By: Codebuff <noreply@codebuff.com>
The command sweep ran metadata.downloadPlaceNames with empty params, which
starts a real download from download.geonames.org into the settings folder.
The sweep now skips it (the command has its own loopback-server tests) and
asserts that no place names download was started.
Co-Authored-By: Claude <noreply@anthropic.com>
`median_row` (the running median under process V2's highlights/shadows base) kept a sorted
window of f32 and, per pixel, found the leaving and entering values with two binary searches
on `total_cmp`, then `remove` + `insert`. The searches mispredicted on most steps.
Now the window holds `i32` keys that order exactly like `f32::total_cmp` (the same transform
it applies), both ranks come from one branch-free counting pass the compiler vectorizes, and
one `copy_within` moves the keys between them. Equal keys are equal bits, so the output is
unchanged bit for bit.
24 MP Sony A7 III log luminance, one P-core, interleaved: r = 9 (export) 1977-2027 ms ->
912-958 ms, r = 3 1249-1292 -> 658-724, r = 24 2180-2261 -> 972-1030. All cores, r = 9:
170-188 -> 86-102 ms. The GPU path has its own median and is untouched.
New test: special values (NaN of both signs, signed zeros, infinities), long ties and planes
narrower than the window match a per-pixel sorted median bit for bit.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0121DZuByhKyGLHw5C9JkUue
- fr-formats: "{n} photo{0} déplacée{0} vers « {1} »" (the adjective took a literal "(s)").
- docs/localization-parity.md: 7 languages in "Other languages LightCraft ships" (was 4);
Swedish and Polish ship, so they leave the "lacks" list and the target totals.
- xtask ci_lock (#2302): the lock-setup errors name the way out
(`LIGHTCRAFT_CI_LOCK=off` or `=<dir>`).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The launch window no longer probes each recent library with the lock check
of File > Open Library (which created catalog.lock and wrote
catalog.lock.owner into every free library at every launch, including
other drives, network shares and sleeping disks). A row is now checked by
looking only: the path exists, is a folder, can be listed, holds a
LightCraft catalog, and that catalog opens for reading. A library another
program has open is found when Open is pressed: the open fails with the
lock owner's message, the window stays and greys the row out.
New test: listing (and drawing the window) leaves every library folder
byte-identical (contents and modification times); it fails with the old
probe. Docs (What's New, parity row, control protocol) say so.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AbegtYaPuwqZ81pGfELsFK
Settings > General > At launch: Open the last library (default, as before)
or Ask me. With Ask me the desktop app opens nothing and the library-problem
window lists the recent libraries instead (name, folder, when last opened;
missing, locked or no-longer-library ones greyed out with the reason; the
last one selected): Open, Open Another..., New Library..., Remove from List,
Quit, with arrows / Return / Esc. Alt/Option held in the first 2 s after the
window appears shows the same list once over the library that opened (winit
reports no modifiers before the first frame; macOS only on a change).
File > Open Recent Library > Remove from List and
app.recentLibraries {remove: path} forget one entry; nothing on disk is
touched. The list reports lastOpened, kept under a new settings key
(recentLibrariesOpened) so older files and versions keep their list.
--library and LIGHTCRAFT_LIBRARY keep priority (launch_plan), --memory never
asks, lightcraft-cli never reads ui.json. Settings are saved while the list
shows; command-line files go into whichever library opens.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AbegtYaPuwqZ81pGfELsFK
LRC-LIB-CATALOG-IMPORT. Before File ▸ Import Lightroom Catalog… starts, a
window asks where the catalog goes: merge into the open library (named,
with its photo count; preselected when it is empty) or create a new
library named after the catalog file, next to the open library (both
editable, Choose… picks the folder). The new library is made and opened
through app.newLibrary (its rules: an empty or new folder, refused while
another import runs), then the catalog is imported into it. Cancel does
nothing. Open Library… on a Lightroom catalog goes through the same choice.
Opt-out: Settings ▸ General ▸ When importing a Lightroom catalog (Ask
where to import it / Always merge into the open library), saved as
ui.json settings.lightroomImport (lenient, default ask); the window's
"Don't ask again" picks Always merge. A session in memory merges without
asking, as before.
file.importLightroom takes {into?: open|new, name?, location?} to choose
without the window. UI layer only: library.importLightroom, the engine
and lightroom_import.rs are unchanged.
Tests: decision table (setting x library on disk/in memory x photos x
into), params, default name; headless window / merge / new library /
cancel / empty library / setting / Don't ask again persisted / folder
not empty / into / in memory / stale library; tests_open_lightroom
follows the new window. Format messages in all 17 catalogs, plain
strings in French. Docs: lightroom-catalog-import, control-protocol,
parity-checklist, whats-new.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AbegtYaPuwqZ81pGfELsFK
A colour-range mask compares every pixel against every sample, so
render and export cost is pixels x samples. The interactive sampler
stops at 5, but mask.add / mask.update (colorRange) passed their
`samples` array straight through, so a few hundred samples on a large
render turned into billions of operations and hung rendering and
export. The project already caps similar lists (MAX_STROKE_DABS,
MAX_DETAIL).
The samples list from automation is now capped at 64.
Test: a request with 1000 samples is cut to 64; a two-sample request is
unchanged.
Reported-by: BS.ehlas.cz
`lightcraft-cli commands` (text and --json) now sorts the registry by id
before printing, so each family's commands sit together. The engine's
registry order, used by engine.commands over MCP and the control channel,
is unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
.avif has no AV1 decoder in any build and .heic/.heif only decode when
the codecs' heif feature is enabled, so listing them let web users pick
files import would immediately report as failed. accept_list() now mirrors
lightcraft_codecs::Format::Heif::can_decode() and never offers .avif.
EXTENSIONS itself is unchanged, so files on disk are still recognised and
imported-or-reported as before.
The File/Edit/View headings sat only egui's default 8 px apart because frameless
buttons get no padding of their own. The test lays out the in-window menu bar and
checks every pair of neighbouring headings is at least 24 px apart; without the
item-spacing fix the gap is 4 px and the assertion fails.
parse_rgb checked only the length, and u8::from_str_radix accepts a
leading '+', so "#+f+f+f" set the overlay to #0f0f0f. Require six hex
digits, as 80e439f6 does for the other #rrggbb parsers.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
fmt_gap wrote whole minutes past an hour, which only drag values could hit
before; with the culling value fields taking any gap up to a day, a typed
3601 s showed as "1 h 0 min" and editing that text gave 3600 s. Past an hour it
now writes "1 h 0 min 1 s" when there are seconds left (the Auto-Stack slider
shares the formatter).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The engine already applies photo.setMeta to all selected photos via
targets(), but the Info panel and All Metadata dialog always show a
single photo's values with no mixed-value indication.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Measured on 68 photos straightened by hand in Lightroom Classic and 300 left as shot, Auto
straighten turned 79 of the 300 by more than 1° (65 by more than 2°): the edges of planks, a
branch or a bird's wing were taken for the horizon. It now answers only for at most 10° and on at
least 3 agreeing lines (Upright Level is unchanged): 19 of the 300, with the same 14 of the 68
levelled within 0.5° and 1 instead of 3 more than 1° off. When it finds nothing, the app says so
instead of leaving the photo without a word. LR-CROP-AUTO goes to 🟡: it still leaves most tilted
photos alone.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Two corpus tests fail on main with the CC0 corpus present (they skip without it, so no CI
has run them):
- camera_preview::tests::corpus_z8_uses_the_bundled_profile / corpus_z6ii_…: the test fitted
the camera look with `lens: None`. Since 1ee751ebd (Nikon's in-file distortion for the
Z 6II and Z 8, the same day as the tests) the camera's JPEG is matched through that warp,
which files.rs reads with `lens_geometry` and removes from the opcodes before binning the
proxy. The test now does the same; the app path was right all along.
- corpus_raw_decodes: coverage-1639.fff has no JPEG and hands on its 8-bit RGB image as a
baseline TIFF (`rgb_preview`, b5dc500c5); the test only accepted JPEG / JPEG XL previews.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- docs/parity-checklist.md, docs/gaps.md, docs/target-app-parity.md: rows two PRs both
changed had been kept twice by the doc resolution. Each row is now upstream main's row
plus every merged PR's own edit (3-way against the PR's merge base), so main's newer
text is kept (LR-LIB-FLAG #1013, Lanczos-3 in LR-MERGE-HDR / -PANO, LR-EXP-DIM #1620,
gaps 8 and 10) and the PRs' additions too; LR-BEHAV-LOCALE / -LOCALIZE take main's
Georgian and #1846's Polish. `cargo xtask parity --write` refreshed the summary.
- #1313 added "Show Frame Rate", a switch; #2195 names switches for what they show
(menubar::tests::toggles_are_named_for_the_thing_not_show): now "Frame Rate".
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Find Missing Photos took a missing photo's file name with Path::file_name,
which doesn't split on `\` on macOS or Linux, so a library made on Windows
had photos it could only find by content: those with no hash on record
(Lightroom-catalog imports, files edited since) were never found. Both the
missing photos and the files found on disk are now named by one rule: the
part after the last `/` or `\`, in lower case.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
