mirror of
https://github.com/storytold/lightcraft.git
synced 2026-11-03 06:59:53 -06:00
06c1b9ae07
* CRW-LIB: decode Canon CRW through the shared craft-crw crate LightCraft decodes Canon CRW (the CIFF raws of the EOS D30, D60, 10D, 300D and the PowerShot G1/G6) through craft-libs' `craft-crw`, the clean-room decoder derived and verified on the `clean-room-crw-format-decode` branch and moved there with its frozen tables and derivation record. `crates/raw/src/vendor/crw.rs` is now the adapter: it maps the crate's header and samples onto `RawImage` (CFA, levels, active area, orientation, as-shot white balance, the make/model/time records a CRW keeps instead of Exif) and converts the capture time through chrono. The dependency is a local path (`../craft-libs/crates/crw`) until the craft-libs branch merges; then it becomes a git revision. Everything LightCraft-side from that branch comes along unchanged: `probe_bytes` fills metadata gaps from the decoder's records so CRW imports name their camera; the EOS D60 bundled camera profile and its chart-measured dual-illuminant matrices (`chart_table.rs`, `spectral::Camera` carries its illuminant pair) give the D60 a Kelvin white-balance scale; CRW joins the JPEG-fitted starting look, the catalog's raw formats and `lightcraft-cli calibrate`; `RENDER_CACHE_VERSION` 71; the CC0 CRW corpus entries for `cargo xtask corpus`; coverage, parity, gaps, raw-parity, target-app-parity and ROADMAP updates. The G5 stays preview-only (the crate refuses it with the reason) and the robustness test keeps its CRW sample. What does not come along: the decoder, its tables and `crw_oracle.rs` (now craft-crw's digest-pinned sample test) and the derivation record, which lives in craft-libs `crates/crw/clean-room/`. `clean-room/crw/` here keeps only the EOS D60 chart-calibration document and its capture list, with a README pointing at the record. Gates: fmt, clippy -D warnings (raw all targets; engine, catalog, cli), lightcraft-raw tests (475 + 134, the corpus files decoding through the adapter), catalog tests, engine lib tests (1,300 pass; the two failures, `tests_import::recursive_discovery_…` and `originals::tests::write_targets_pin_absent_paths_…`, are in files unchanged from main), `cargo xtask parity`, `assets`, `layers`, `coverage.py --check`. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * no message * no message * Update documentation after moving code to craft-libs * Review coment fixes * CRW: pin craft-crw f3b62ec, a truncated entropy stream is Corrupt craft-libs #472 makes craft_crw::decode refuse a stream that ends more than Limits::max_stream_end_fill (64 samples, one block) before the frame does, instead of filling the rest with predictor values and returning success. The adapter already maps craft_crw::Error::Corrupt to RawError::Corrupt; a regression test proves a truncated CRW now fails at this boundary (it decoded as a flat frame on the old pin) while its header still reads. Only craft-crw moves; craft-ui/docpath/update/window stay on c5a1cbf. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * Code review comment fix * Add some tests --------- Co-authored-by: Robert Woldberg <314159+rwoldberg@users.noreply.github.com> Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
19 KiB
19 KiB
LightCraft roadmap
Stage: alpha · next: beta, ~14 points and ~340–600 h away
Last reviewed: 2026-10-10 · Last updated: 2026-10-24 · Change: minor (Canon CRW decoded clean-room, 6 bodies verified, 557 models verified; gap 9: CI runs the tests on Linux x64 since #2667; macOS, Windows and the GPU equivalence tests still don't; the lens profile database priority links its tracking issue #329; Print: Japanese photo and postcard paper sizes — L判, 2L判, KG判, はがき, 往復はがき, 六切, 四切, 名刺; Minolta MRW decoded, 6 bodies verified; RAW: 85% of Adobe's models decode, 39% verified after the camera-coverage sweep #1327; German catalog 91% → 97%: interface strings and What's New lines translated) · Target: Adobe Lightroom Classic 15.5 / Lightroom 9.6
One-page summary of how close LightCraft is to Adobe Lightroom. Agents: read this, then pick work from docs/gaps.md. The full assessment, method and evidence are in docs/target-app-parity.md.
Headline numbers
| Value | Kind | |
|---|---|---|
| Feature breadth (checklist) | 81.6% of 532 rows (P0 98.0%, P1 95.2%, P2 50.0%) | measured: parity-checklist.md recount |
| Feature breadth (weighted by use) | ~77% | estimated |
| Ready for real work (full target) | ~61% (56–66%) | estimated: weighted sum over the dimensions |
| Mainstream practitioner | ~51% (46–56%) | estimated |
| Essentials user | ~65% (60–70%) | estimated |
| Remaining to beta | ~340–600 h (Opus 5.5 agent hours, ~60% parallel) | estimated |
| Remaining to full parity | ~1,000–1,800 h (~70% parallel) | estimated |
Each audience with its own hours (Opus 5.5 agent wall-clock hours to ~95%):
| Audience | Ready | Opus 5.5 agent wall-clock hours to ~95% | Work that dominates |
|---|---|---|---|
| Full target (ready for real work) | ~61% | 1,000–1,800 h (~70% parallelizes) | raw colour, lens and camera data; AI models; Classic output modules; localization; plug-in ecosystem |
| Mainstream practitioner | ~51% | 450–840 h (~60% parallelizes) | camera colour and render fidelity; CR3 / ORF / HE NEF; lens database; subject / sky masks; Lightroom XMP exchange; stability |
| Essentials user | ~65% | 160–300 h (~50% parallelizes) | default raw look for popular cameras; launch stability; basic-slider feel; opening CR3 / HEIC files people send |
The gap between breadth and readiness is quality and coverage, not missing buttons: camera colour is
estimated rather than measured, CR3 decodes on 26 verified Canon bodies (rarer CRX variants still open as previews), compressed ORF and HE NEF
open as previews, rendering is tuned by eye, AI masks are heuristics, there is no lens database, and
tests run in CI on Linux x64 (every pull request and push to main) and FreeBSD (nightly), not on macOS or Windows. Why alpha: core workflows run end to end, but readiness is below
the ~75% beta bar (it is above the ~40% alpha bar) and Lightroom's main inputs, raw files, still have blocking gaps (compressed ORF and HE NEF open as previews; Canon,
~30% of raw users, decodes CR3 on 26 verified bodies). It passes the alpha gate: all six core workflows
(import, cull, develop, save and reopen, batch looks, export) work end to end on macOS
(docs/roadmap.md). The three readiness numbers and their weights and discounts are in
target-app-parity.md.
Hours are calibrated on this repo's PR history; see
target-app-parity.md.
By dimension
| Dimension | Ready | Breadth | Remaining | Doc |
|---|---|---|---|---|
| Features | ~61% | ~77% (checklist 81.1%) | 600–1,100 h | target-app-parity |
| ↳ RAW cameras, colour, lens profiles | ~60% | 85% of 1,446 Adobe models decode, 39% verified | 255–480 h | raw-parity |
| UI/UX fidelity | ~70% | ~85% | 32–54 h | ui-parity |
| File formats (non-raw) | ~70% | ~85% | 30–55 h | file-format-parity |
| Hardware | ~40% | ~50% | 40–80 h | hardware-parity |
| Localization | ~64% of the 12 key languages | 8 of 12 present | 130–210 h (+135–239 h new scripts) | localization-parity |
| Performance | ~70% | ~80% | 30–50 h | target-app-parity |
| Stability | ~60% | — | 30–60 h | target-app-parity |
| Platforms | ~60% | macOS, Windows, Linux, FreeBSD, web | 25–45 h | target-app-parity |
| Ecosystem / plug-ins | ~15% | ~25% | 40–80 h | target-app-parity |
| AI features | ~20% | ~30% | 120–220 h (overlaps features) | target-app-parity |
Features
| Area | Ready | Breadth | Remaining |
|---|---|---|---|
| Import and library | 80% | 90% | 25–45 h |
| RAW decoding and camera coverage | 77% | 82% | 120–220 h |
| Colour, camera profiles, render fidelity | 35% | 65% | 80–150 h |
| Global develop adjustments | 70% | 93% | 30–60 h |
| Optics, lens profiles, crop and geometry | 50% | 80% | 40–80 h |
| Masking | 45% | 67% | 60–110 h |
| Remove, heal, Enhance | 30% | 45% | 70–130 h |
| Presets, profiles, versions, sync | 85% | 95% | 8–15 h |
| Export | 75% | 89% | 15–30 h |
| Merge and HDR | 60% | 85% | 20–35 h |
| Views, compare, zoom, culling | 70% | 90% | 15–30 h |
| Classic output modules (Map, Book, Slideshow, Print, publish, tethering) | 10% | 15% | 90–150 h |
| Video | 0% | 0% | 25–45 h |
Weights and evidence: target-app-parity.md.
Languages
Measured from crates/ui-egui/locales (share of 2,304 catalog messages translated; German counts 3,644 of 3,754, because its What's New lines add keys); none is full
while lower-layer errors stay English. Detail: localization-parity.md.
| Language | Code | Translated | Status |
|---|---|---|---|
| English | en | 100% | full (source) |
| Simplified Chinese | zh-hans | 87% | partial |
| Spanish | es | 86% | partial |
| Hindi | hi | 0% | none (no Devanagari shaping) |
| Arabic | ar | 0% | none (no RTL / shaping) |
| French | fr | 88% | partial |
| Portuguese (Brazil) | pt-br | 85% | partial |
| Indonesian | id | 100% (What's New in English) | partial |
| Japanese | ja | 87% | partial |
| German | de | 97% (3,644/3,754) | partial |
| Korean | ko | 0% | none (no Hangul face) |
| Vietnamese | vi | 0% | none |
Also shipped: Traditional Chinese (87%), Russian (74%), Ukrainian (99%), Italian (92%), Georgian (100%), Finnish (79%).
Upcoming
Ranked toward beta; detail in docs/roadmap.md, gaps in docs/gaps.md.
- Camera colour of our own: measured calibration for the popular models (50–90 h)
- CR3 for every Canon body (25–45 h)
- Render-fidelity suite against Lightroom, then tuning (30–60 h)
Full-resolution zoom(done 2026-10-12)- AI masks: Subject / Sky / People (40–70 h + model decision)
- Compressed ORF and HE NEF (28–55 h)
- Lens profile database of our own (40–80 h, #329)
- Tests in CI on macOS and Windows, and the GPU tests in CI (6–12 h)
- Per-model raw verification (20–35 h)
Progress log
- 2026-10-22: Print page setup and PDF output gain the Japanese photo and postcard papers — L判
(89 × 127 mm), 2L判 (127 × 178), KG判 (102 × 152), はがき (100 × 148), 往復はがき (148 × 200),
六切 (203 × 254), 四切 (254 × 305) and 名刺 (55 × 91) — at their exact metric sizes. The Print dialog
and
print.settake their customary names, and names are matched case-, space- and punctuation-insensitively (L判,l,L-sizeare one paper). - 2026-10-13: Canon CRW (the 2000–2004 CIFF bodies) decodes from sensor data: compression
derived clean-room against a black-box reference and frozen (
craft-libs crates/crw/clean-room/02-crw-format.md); EOS D30/D60/10D/300D and PowerShot G1/G6 verified pixel-exact on 28 of 29 corpus files (18 of them the maintainer's D60 shots); as-shot white balance, black level, rotation and capture time from the CIFF records; the PowerShot G5's anomalous coding stays preview-only. With the camera-coverage sweep and MRW merged, 557 models verified; Canon bucket 95% → 98%. - 2026-10-14: Camera-coverage sweep verifies 30 more models on 40 CC0 files, with file and sensor
hashes pinned (
docs/raw/coverage.py: 515 → 545 verified). Eleven candidates excluded after rechecking identity, provenance and camera-image agreement. Phone and drone DNG tone differences remain tracked in #1277. - 2026-10-18: Minolta / Konica Minolta MRW decodes from sensor data (12-bit packed and 16-bit words,
RGGB and GBRG, as-shot WB from
WBG); DiMAGE 7, 7Hi, A1, A2, A200 and Dynax / Maxxum 7D verified on CC0 files. 551 models verified, 201 preview-only; headline numbers unchanged. - 2026-10-14: 97 Canon CR2, Pentax and Ricoh models verified, one CC0 raw.pixls.us file each pinned and
checked against the camera JPEG (
crates/raw/tests/it/canon_pentax_corpus.rs). 334 models verified, ~50% of photographers on a verified body; headline numbers unchanged. - 2026-10-12: Mamiya MEF (ZD) decodes from sensor data and is imported; verified on a CC0 file. 241 models verified, 35 unsupported; headline numbers unchanged.
- 2026-10-12: Epson ERF (R-D1, R-D1s, R-D1x) decodes from sensor data, white balance measured from each file's preview; 3 bodies verified on CC0 files. 240 models verified; headline numbers unchanged.
- 2026-10-11: Leica Q3 verified by a recorded check of 38 private DNGs (all decode and develop at full size; mean ΔE to the camera JPEG 6.3–13.9). 236 models verified; headline numbers unchanged.
- 2026-10-12: 22 more Canon CR3 bodies verified (EOS R, RP, R3, R5, R6, R6 Mark II / III, R7, R10, R50,
R50 V, 90D, 250D, 850D, M200, M50 Mark II, M6 Mark II, 1D X Mark III, PowerShot G5 X II, G7 X III, SX70,
V1): one CC0 file each pinned and checked against the camera JPEG; all 110 CC0 CR3s on raw.pixls.us
decode. The EOS R6 Mark III gets its as-shot white balance (
ColorData66). 259 models verified; RAW decoding 77% → 83%; mainstream ~49% → ~51%, essentials ~64% → ~65%, full ~61%. - 2026-10-12: 117 Nikon and Fujifilm bodies verified, one CC0 raw.pixls.us file each pinned and checked
against the camera JPEG (
crates/raw/tests/it/body_corpus.rs); the sweep found COOLPIX 5700 / 8400 / 8800 decoding wrongly. 354 models verified, ~58% of photographers on a verified body; headline numbers unchanged. - 2026-10-11: Canon EOS R5 Mark II C-RAW decodes from sensor data (57% of its files opened as a preview): the CRX QP map's first-symbol literal and limited-length escape; four CC0 R5 Mark II files pinned, the QP maps of 33,665 private C-RAW files decode exactly (cr3.md). 235 models verified; RAW decoding 76% → 77%; headline numbers unchanged.
- 2026-10-10 (UI docking): Library, Tools, and Presets panels can be regrouped, split, floated within the app window, closed, and reopened. Tools keeps the current Edit/Crop/Masking mode; the photo view and filmstrip stay together. Panel moves and reset are available as UI commands, and saved UI preferences retain arrangements. Widget state styling uses the shared
craft-uitheme helper, retaining the dark palette, typography, spacing and per-state borders. No readiness percentage change. - 2026-10-10 (RAW): RAW measured model by model against Adobe's 1,446-model list (234 verified, 943 unverified, 233 preview-only, 36 unsupported; ~84% of photographers' cameras decode, ~40% on a verified body). Full ~61%, mainstream ~49%, essentials ~64%; beta ~14 points away.
- 2026-10-10 (final): full ready is a weighted sum over the dimensions (~59%; no new evidence); mainstream practitioner ~51% (450–840 h), essentials user ~63% (160–300 h); alpha gate passes.
- 2026-10-10: full re-measure, progress docs restructured: stage alpha, ready ~55%, checklist
81.1%; new docs (target-app-parity, gaps, roadmap, architecture, raw / file-format / hardware / UI
/ localization parity);
docs/parity.mdrenameddocs/parity-checklist.md. Landed today: Nikon lossy-after-split NEF (#633), ORF 12.8 bpp (#651), HDR gain map JPEG / PQ AVIF / float TIFF export (#646), Lightroom-like Auto (#628), activity stack. - 2026-10-11: Sony ILCA-99M2 (α99 II) camera colour: bundled profile from 144 photos; on 30 withheld photos mean ΔE vs the camera JPEG 5.11 → 4.87, 25 of them closer.
- 2026-10-09: Sony ILCE-7CR camera colour: bundled profile from 125 photos; on 26 withheld photos mean ΔE vs the camera JPEG 4.07 → 3.29. HEIC in release builds (#609), Samsung uncompressed SRW, contact-sheet PDF export, Ukrainian and French UI, release v0.5.0.
- 2026-10-08: independent Rust CR3 lossless and version 0x100/0x200 C-RAW decoding, six exact full-sensor regressions on M50 / R100 / R8; guarded CR3 camera-JPEG colour fitting (cr3.md).
- 2026-10-08: Fujifilm lossless / lossy compressed RAF: pure-Rust striped decoder, full sensor arrays verified on 17 CC0 and 7 supplied files (13 bodies). Fujifilm camera colour: bundled X-H2S / X-T4 profiles from 201 / 545 photos; on 150 withheld files mean ΔE X-H2S 1.70 → 1.62, X-T4 2.43 → 2.34. AI RAW denoise (#384), Lightroom catalog import, German, Spanish and Russian UI.
- 2026-10-07: Panasonic RW2 / Leica RWL / Panasonic RAW decode in every raw format, established on 178 CC0 files from 118 bodies; RW2 file-local camera look (140 of 174 files). Faces: recognition in pure Rust and People view.
- 2026-10-03 – 10-05: community PRs (#42–#51, #60) and 25+ issues: compressed NEF, Move import, folder templates, Local roots and cleanup, 85k-photo grid and catalog performance, background compaction, failed-save errors, GPU export hardening. Added the honest Where we stand assessment and tracker rows for camera colour, camera coverage and render fidelity.
- 2026-10-02: parity estimate added (
cargo xtask parityprints weighted completion); milestones refreshed. - 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.
- 2026-09-30 (later): app running with the full Lightroom-style UI; pipeline v0; DNG / CR2 / ARW; README showcase. ≈ 8 h elapsed.
- 2026-09-30: roadmap created; M0 in progress; research docs (Lightroom reference, Rust imaging ecosystem) complete.
Revision history
| Date | Change | Summary |
|---|---|---|
| 2026-10-24 | minor | Priority 7 (lens profile database) links its tracking issue #329; headline numbers unchanged |
| 2026-10-24 | patch | Alpha rationale: tests run in CI on Linux x64 (#2667) and FreeBSD nightly, not on macOS or Windows; ranked item 8 now reads "Tests in CI on macOS and Windows, and the GPU tests in CI"; percentages and hours unchanged |
| 2026-10-22 | minor | Print: Japanese photo and postcard paper sizes (L判, 2L判, KG判, はがき, 往復はがき, 六切, 四切, 名刺) in the page setup, Print dialog and PDF output; headline numbers unchanged |
| 2026-10-19 | minor | Indonesian interface (id): every plain and formats message, plus 69 command labels, locale, menu command and tests; What's New left in English; Localization ~55% → ~64% (mean of the 12 key languages), 8 of 12 present |
| 2026-10-13 | minor | Canon CRW decoded (clean-room): 6 bodies verified pixel-exact, G5 preview; 551 models verified after merging the coverage sweep |
| 2026-10-12 | minor | Italian interface (it, #1192): 2,201 of 2,398 messages, locale, menu command and tests; headline numbers unchanged |
| 2026-10-13 | minor | Sony ILCA-99M2 camera colour profile in the log; numbers unchanged |
| 2026-10-14 | minor | Localization ~52% → ~55% (mean of the 12 key languages): de, es, fr, ja and uk catalogs filled |
| 2026-10-16 | minor | German 91% → 97% (217 interface strings and 862 What's New lines added); headline numbers unchanged |
| 2026-10-14 | minor | 21 Sony and Leica models verified (258 models, 18%); headline numbers unchanged |
| 2026-10-12 | minor | UI/UX remaining 34–58 h → 32–54 h (canvas tools by keyboard); headline numbers unchanged |
| 2026-10-14 | minor | 17 uncompressed Olympus and Samsung models verified on one CC0 file each (254 models verified, 18%); headline numbers unchanged |
| 2026-10-12 | minor | Mamiya MEF decoded and imported: ZD verified (241); headline numbers unchanged |
| 2026-10-12 | minor | Epson ERF decoded: 3 more verified models (240); headline numbers unchanged |
| 2026-10-12 | minor | Georgian (ka) interface language added (every catalog message) |
| 2026-10-17 | minor | Finnish (fi) interface language added: 2,610 of 3,308 messages (79%), locale, menu command and tests; headline numbers unchanged |
| 2026-10-14 | minor | Full-resolution zoom (gap 4) closed: soft proofing gets zoom windows; headline numbers unchanged |
| 2026-10-12 | minor | UI/UX remaining 40–70 h → 34–58 h (accessibility audit landed); headline numbers unchanged |
| 2026-10-18 | minor | Minolta MRW decodes from sensor data; 6 Konica Minolta bodies verified; headline numbers unchanged |
| 2026-10-14 | minor | Camera-coverage sweep (#1327): 30 models verified on 40 CC0 files (coverage.py 515 → 545 verified); sensor verification does not establish colour parity (#1277); headline numbers unchanged |
| 2026-10-12 | minor | 22 more CR3 bodies verified, EOS R6 Mark III white balance: RAW decoding 77% → 83%, RAW ~56% → ~60%; mainstream 49% → 51%, essentials 64% → 65%; full unchanged at ~61% |
| 2026-10-12 | minor | 117 Nikon and Fujifilm bodies verified (354 models, 24%); headline numbers unchanged |
| 2026-10-11 | minor | EOS R5 Mark II CR3 verified: CR3 on four verified bodies, RAW decoding 76% → 77%; headline numbers unchanged |
| 2026-10-10 | minor | RAW measured model by model; RAW decode and colour as explicit rows: full 59% → 61%, mainstream 51% → 49%, essentials 63% → 64%; beta ~14 points |
| 2026-10-10 | minor | Full ready restored to the additive weighted sum (~59%; no new evidence); hours per audience; beta ~16 points away |
| 2026-10-10 | minor | Mainstream practitioner and essentials user numbers; full ready 55% → 46% (method written down); beta 340–600 h |
| 2026-10-10 | minor | Alpha gate added (core-workflow check): passes, stays alpha |
| 2026-10-10 | major | Progress docs restructured; re-measured; Where we stand, Parity estimate and Totals moved to docs/target-app-parity.md, milestones and risks to docs/roadmap.md, raw coverage to docs/raw-parity.md, catalog migration to docs/file-format-parity.md, CJK notes to docs/localization-parity.md |
| 2026-10-08 | minor | Where we stand by dimension and by kind of user |
| 2026-10-02 | major | Parity estimate and effort table |
| 2026-09-30 | major | Created |
| 2026-10-14 | minor | 97 Canon CR2, Pentax and Ricoh models verified (334 models, 23%); headline numbers unchanged |
| 2026-10-14 | minor | RAW: 82% of Adobe's models decode, 36% verified (integration batch 36's CC0 per-model checks) |