809 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
dtourolleandClaude Opus 5	d7a81375ee	List the steps, and let a photographer step straight to one ... Undo answers "take back the last thing", which is the question asked about a mistake just noticed. It is the wrong instrument for one noticed six adjustments later: eight presses, each changing the picture, with no way to see how far back the mistake is without passing through it. A step is a whole state, so arriving from six away costs what arriving from one does — which is what makes a row worth making clickable rather than decorative. `Edit::Discrete` had to go for the list to be worth drawing. Seventeen call sites recorded the same anonymous step, which is fine for deciding whether two changes are one gesture and useless for a panel: seventeen rows reading "Discrete" is not a history. Every variant now carries enough to name itself, and the compiler enumerated the sites that had to start saying so. A step that moved a parameter is still named out of the descriptor, so an operation added as a YAML declaration appears in the history correctly named with nothing written for it (FR-DEV-3c). Choosing a film stock was not undoable at all. The pick went straight to `choose_film`, which nothing on the history's path ever sees. `pick_film` records it, and is separate because the same call is also how a *restored* edit gets its tables back — recording that would push a step for the undo the photographer had just asked for. The list is rebuilt off a revision rather than off every redraw. A drag ends in a redraw per frame while folding into one step, so the unconditional version would tear down and recreate every row sixty times a second to arrive back at the list already on screen. The counter is process-wide: a per-instance one starts every photograph at the same number, so a frontend holding "the revision I last drew" would keep the previous image's steps on screen — invisible while every image opens with one identical row, and a wrong-photograph bug the moment persisted history means it does not. The step names that no descriptor can supply are constants with a roll, and a test walks the roll rather than a second copy of it. `resolve` splits so that "is this catalogued?" can be asked: `derive` turns `history.mask_toggled` into "Mask Toggled", which names a field rather than an act and, being perfectly readable, is a mistake nobody would look at twice. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 23:04:35 +02:00
dtourolleandClaude Opus 5	b89f1cfece	Snapshot the whole edit in the history, so a drawn mask can be taken back ... The undo stack snapshotted a `Preset` — the parameter map — and a mask layer is deliberately not a parameter. So drawing one changed nothing the history could see: `record` returned `false`, no step opened, and the layer the photographer had just painted had no way back. The interface went on calling `record` in good faith, including from the mask controls, and nothing failed. A film stock went missing the same way. The snapshot is an `EditState` now, so the history is complete by construction rather than by anyone keeping a list in their head. `undo` and `redo` return a `Step` rather than a `bool`. Stepping is not the only outcome a caller has to act on — a step across a change of film leaves the graph without its tables, and only the caller can bake them — and a `bool` would let that be dropped by writing nothing at all, which is the shape of mistake this module had already made once. `DevelopSession` settles the debt either way; a step that found nowhere to go is left alone, since clearing the film because undo hit the floor would take the stock off the picture. Five tests, all of which fail against the old snapshot: a drawn layer is undoable and redoable, a layer's own settings are a step of their own, a change of stock is a step and names what it needs baked back, clearing the film is undoable, and an exposure move does not deep-copy the mask stack. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:54:55 +02:00
dtourolleandClaude Opus 5	93efdf27a6	Keep the film stock when an edit is saved ... `Version::update` is the write path an automatic save goes through. It copied the parameters and the masks and said nothing about the film, so a photograph developed on a stock was written back without it and opened the next time without its emulsion. Nothing reported a failure — the line was simply not there. It is the third of the three routines that captured "the edit" and the only one that got it wrong, which is the argument for not having three. All of them now destructure one `EditState`, so `from_graph`, `update` and `apply` cannot disagree about what an edit consists of, and the next part of one cannot be lost by anybody writing a line too few. Two tests, both of which fail without the fix: the field survives `update`, and the stock survives the round trip through the file. `apply` returns the `FilmRebake` it always implicitly owed, so `apply_version` now reads the debt off the call rather than off `version.film` — and pays it in both directions, since a version with no film has to clear the adjust pass too or it keeps textures bound that nothing will sample. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:53:00 +02:00
dtourolleandClaude Opus 5	d562ceaaf4	Show a face beside each name in the people rail ... The rail was drawing an empty square for every person: cover was hardcoded to a default image. That is the one place a portrait matters most, because the rail is how the user decides which unnamed group to open first, and a list of "Unnamed (24 faces)" rows tells them nothing. The portrait is the person's confirmed face with the largest crop_px -- the most source pixels the face actually occupied, so the one they have the best chance of recognising -- falling back to a suggestion so a freshly clustered group still has a face beside it. Cached in the controller, because every mutating action reloads the whole screen and cutting a portrait costs a JPEG decode per person. Without the cache, confirming one face would re-decode a proxy for every person in the library, and the rail does not change when a suggestion is accepted. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:52:38 +02:00
dtourolleandClaude Opus 5	5622a58ce3	Give an edit one complete state, and make omitting part of it a compile error ... An edit used to be a bag of scalars. That stopped being true when the mask stack and the film stock arrived, both deliberately held apart from `ops` because a layer is not a scalar and a stock is not a scalar — and nothing announced the change. What happened instead is that three routines each captured "the edit" and each captured a different subset of it. `EditState` is all of it: the parameter map, the masks, the stock. What keeps it complete is not a comment. `EditGraph::state` destructures the graph exhaustively, `EditGraph::set_state` destructures the state exhaustively, and the fields are public so every construction site is a struct literal naming all of them. Adding a fifth kind of graph state — FR-DEV-8's spot removal is the one already asked for — fails to compile until somebody has decided whether an undo has to put it back. Verified both ways round by adding a field to each type and watching five call sites refuse to build. A compiler error rather than a runtime check, because the failure being prevented is silence: the missing halves produced no panic, no warning and no failing test. The stack is now shared rather than owned, and that is about the drag path rather than memory: `state` runs on every parameter change, which during a drag is once a frame, and deep-copying a painted brush sixty times a second to record an exposure move would be a cost paid for nothing. `masks_mut` is the one door a stack is modified through, so it clones on write. `FilmRebake` is the one thing a caller is still owed. Restoring a stock always needed the profile database this crate does not link (ARCH §6.5a); it was a comment before, and it is a `#[must_use]` return value now. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:50:11 +02:00
dtourolleandClaude Opus 5	2944c1b704	Document the run marker and cross-device face sync ... Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:42:28 +02:00
dtourolleandClaude Opus 5	96d07da15f	Sync face data as sealed shards, so a second device does not re-index ... Indexing 23,500 images is about two hours of CPU, and the result is byte-identical on every device: the same model over the same proxy produces the same embedding. Paying for it once per account rather than once per device is the point. Shards rather than the catalog snapshot, because the snapshot goes up whole on every sync and a fully indexed library carries roughly 30 MB of embeddings. That is exactly the cost the thumbnail store's 25 MB cap exists to bound, so face shards use the same cap -- imported from dr_thumbs rather than restated, since the number is a statement about sync cost and the two must not drift apart. The split follows the one already there: bulk immutable data in sealed shards, small mutable data in the catalog snapshot. Faces, landmarks, embeddings and run markers shard; people, names and assignments ride the catalog and merge by uuid. Keyed on oc:fileid throughout, never on image_id, because a row id means nothing on another device. The run marker travels with the faces it describes. Without it a receiving device cannot tell an image with no faces from one never examined, and would re-detect every landscape it had just adopted. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:41:50 +02:00
dtourolleandClaude Opus 5	4a82753d22	Show the detector the photograph, not the sensor's scanlines ... "Find subjects" was handed the proxy in the sensor's own orientation, so every frame shot on a body held sideways reached the model lying on its side — and a model trained on upright photographs is very bad at those. Measured end to end on a 22 MP frame of two people and a dog: `person 0.36` and nothing else, against `dog 0.82, person 0.61, person 0.49` for the same pixels stood up. Nothing failed; the panel simply offered one poor subject where there were three good ones. The orientation was never dropped on purpose. The proxy is deliberately rendered through a *neutral* graph — the detection has to survive an exposure change, or every slider would invalidate the masks built on it — and neutral took the file's orientation with it along with everything else. Landscape frames were unaffected, which is why it stood for as long as it did. The turn is `Orientation::source_pixel`, the same function the grid's thumbnails already go through, so the detector and the thumbnailer now agree about which way is up rather than holding two opinions. What it is turned by is `Framing::effective_orientation` — the file's EXIF tag and the photographer's own rotations composed into one permutation, by the group law rather than by adding the turns, which is a distinction `Framing` already had to make and had already tested. Rotating the picture and pressing the button again therefore does what it looks like it does. The proxy stays in sensor space and the masks come back into it. That is not a detail to be tidied later: the generated shader samples the mask array at `uv_src`, *after* the framing map, so a mask stored upright would sit a quarter turn off the subject it was drawn around. That is a wrong mask rather than a weak one, and nothing announces it. So the picture is stood up for the model and laid back down for everything else, and `upright`/`lay_down` are returned as a pair because calling one and forgetting the other is silent. Both directions are the one function: `upright` gathers through `source_pixel` and `lay_down` scatters through it. A quarter turn is a bijection of the pixel grid, so the round trip is exact — no filter, no resampling, and no hole to fill — and an inverse written out by hand would be a second thing to keep in step, whose way of being wrong is a mask mirrored about the wrong axis, which still looks like a mask. The orientation joins the confidence and the tiling flag in the segmentation signature, and for the same reason: turning the photograph changes what the model recognises, so two runs either side of a rotation are different instance lists. Two that happened to come out the same length would otherwise share a signature and a stored layer would be silently re-indexed from one into the other. The refine pass had it too — it re-runs the model over a crop rendered in the same sensor space — so it makes the same turn, and would otherwise have handed back a worse mask than the one it was asked to improve, on the subject the photographer had just pointed at. `dr-gpu`'s `local` example is fixed with it. It exists to be the shipping path with pictures attached, and a diagnostic that reproduces the bug it is meant to catch is a trap for whoever reads it next. Seven tests. The round trip is the identity over all eight EXIF tags on a non-square asymmetric grid; a turn carries whole pixels rather than shearing the channels apart; a sideways frame reaches the model upright; a box comes back in sensor pixels, worked out by hand for the one turn a portrait frame actually writes; a restored box still reads low-to-high for every tag, since the rest of the pipeline takes `x1 - x0` without checking the sign; and the eight tags cannot collapse into one signature key. The existing composition test now runs against `effective_orientation` itself, over all 8 x 16 baseline-and-user pairs. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:40:58 +02:00
dtourolleandClaude Opus 5	336a1fd296	Write down what the Android renderer change cost, and what else is owed ... Three compromises were taken deliberately over the last few days and none of them was written anywhere a future reader would look. So `docs/technical-debt.md`, and two corrections to the architecture document that the Android change made untrue the moment it landed. **TD-1, the Android readback.** §12/6.1 says GPU results never round-trip through the CPU, and the develop view on Android now does exactly that. That is worth recording as a breach with reasons rather than quietly leaving a constraint the code no longer honours — the next person to read §6.1 and then `DevelopSession::render` would otherwise conclude one of them is a mistake. The entry carries the device measurements that forced it, why setting `preTransform` is not a fix available to us, and the three separate things any one of which would remove it. **TD-2 and TD-3**, the serial thumbnail fetch and the unbounded drain, were found while chasing the tearing and are still outstanding. Both have a known shape for the fix; neither is a bug, and neither should be discovered again from scratch. The architecture document said the app renders "through wgpu to Vulkan on both Linux and Android", which stopped being true at 6267802, and §12/6.1 claimed a constraint with no exceptions. Both now say what the code does and point at the debt entry for why. The numbers are labelled with what they are. TD-1's readback cost has *not* been measured on the device and says so, and TD-3's figures are from a debug build and say so — a documented measurement that quietly turns out to be the wrong build is worse than no measurement. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:36:37 +02:00
dtourolleandClaude Opus 5	26a1eb7e28	Record that face detection has run, not just what it found ... An image with no faces in it was indistinguishable from one that had never been looked at, so every landscape, still life and document scan in the library was re-detected on every pass, for ever. In a real library that is most of it: on the 23,527-image test library, 64 of the first 110 images indexed contain no face at all. Schema v9 adds face_index, a run marker per (image, model) carrying the face count and the proxy edge it read. Keyed on the model, so a model change puts every image back in the queue by itself. That makes a coverage figure possible, which is the thing a user actually wants to see. The audit also splits the outstanding set by whether a proxy exists, because 23,417 awaiting a proxy and 110 ready to index are different problems, and telling the user to run indexing again would not fix the first. The Identity screen gains Index faces, Stop, and the coverage line. examples/face_index.rs is the same check and sweep without a window, which is the right shape for an overnight pass. Measured on the real library in release: 3.5 images/second, 110 images and 125 faces in 30 seconds, and a second run correctly finds nothing left to do. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 22:13:41 +02:00
dtourolle	e0cb968e04	Merge remote-tracking branch 'origin/master' into worktree-spot-removal ... # Conflicts: # docs/traceability.md	2026-08-26 21:56:32 +02:00
dtourolle	bc07c32611	Merge branch 'master' into worktree-spot-removal ... # Conflicts: # docs/traceability.md	2026-08-26 21:51:47 +02:00
dtourolleandClaude Opus 5	e14bc34a9e	Ask which frame this was taken on, because grain is enlargement ... A crystal is a fixed size in micrometres. How grainy a photograph looks is therefore not a property of the emulsion alone -- it is film size against output size, and the frame is the half a digital file cannot supply. This assumed 35 mm for everything. The same emulsion on 4x5 averages about 3,800 crystals into the pixel that holds 300 on 35 mm, so it renders roughly 3.5 times smoother at the same print; every large-format photograph was being rendered as grainy as a half-frame. `Format` now carries the real image widths -- the gate, not the nominal inches, since a "4x5" exposes about 121 mm -- and the film node asks for it. It is a genuinely fixed list, unlike the stocks, so it is a declared `enum` parameter and gets its control, its sidecar entry and its undo step for nothing. It is also the first enum in the develop chain, and it broke two tests by being one. A row has to compare equal to itself across two builds or `sync_rows` replaces it on every parameter event -- destroying the elements built from it, including whichever TouchArea holds the current gesture, so the format picker would have fought every slider drag in the panel. `ModelRc` compares by identity and the row built a fresh choices model each call. `no_choices` already shares one empty model for exactly this reason, and the build site already said "see no_choices for why the identity matters". The fix follows it: memoise the model per variant list. Curve rows solve the same problem the other way, writing values through the existing model, which is not needed here -- a variant list is fixed at compile time, so one model can serve forever. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 21:47:19 +02:00
dtourolleandClaude Opus 5	626780276d	Draw with OpenGL on Android, where the driver owns the display rotation ... The grid tore while scrolling on the tablet, in portrait only, and was flawless in landscape. It was not vsync and it was not the grid. Measured on the device, same build, only the tablet rotated: landscape bufferTransform=ROT_180 composition=DEVICE (2) clean portrait bufferTransform=ROT_270 composition=CLIENT (1) torn The panel is mounted landscape — 1920x3000 at installOrientation 3 — so a portrait window needs a 90 degree rotation before scanout. wgpu-hal hardcodes the swapchain's `preTransform` to `IDENTITY` and says so in a comment beside the line: // On Android 10+, libvulkan's `vkQueuePresentKHR` returns // `VK_SUBOPTIMAL_KHR` if not doing pre-rotation ... This is always the // case when the device orientation is anything other than the identity // one, as we unconditionally use `VK_SURFACE_TRANSFORM_IDENTITY_BIT_KHR`. That is gfx-rs/wgpu#3345, and it cannot be fixed by setting the field: `preTransform` is a *promise* that the content is already rotated, so keeping it needs the renderer to rotate what it draws, which wgpu cannot do on Skia's behalf. We do not have to be on that swapchain. `AndroidWindowAdapter` chooses `SkiaRenderer::default_wgpu_29` only because this crate enables `unstable-wgpu-29`; without it `SkiaRenderer::default` resolves — through i-slint-renderer-skia's build script, which selects OpenGL on anything that is not Apple, Windows or wasm — to Skia over OpenGL, where the driver owns the rotation and there is no transform to get wrong. So both renderer features move to the desktop-only dependency, and desktop is untouched. # The cost, stated rather than hidden Skia over OpenGL cannot sample a `wgpu::Texture`, so the develop view's frame comes back through memory: `AdjustPass::export_pixels`, already ungated and already used by the export path, into a `SharedPixelBuffer`. That is the round-trip ARCH §6.1 and AC-8 exist to forbid, and it is the right trade only because of what the alternative actually is — not a faster develop view, but a grid that tears in the orientation a tablet is mostly held in. Two things keep it small. The device is still opened on Android, so demosaic and the adjust pass are untouched on the GPU; only the last hop changes. And `render` fits the pass to the canvas before it runs, so the readback is at viewport resolution, a fraction of the ~7 ms at 4K the original measurement was taken against. Four other explanations died on the way here, each by measurement rather than argument: the present mode (a patch confirmed in the installed binary reached `AutoVsync`, and the rows still duplicated), our shared wgpu device (Slint opened its own, unchanged), Skia's partial rendering (off for GPU surfaces), and client composition itself (unavoidable in portrait on this panel, so it cannot be what distinguishes a torn frame from a clean one). Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 21:47:02 +02:00
dtourolleandClaude Opus 5	3e607222c6	Record the Identity screen in the face spec ... Also notes what building it taught the design: a split has to reject before it confirms, or the next clustering pass undoes it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 21:16:04 +02:00
dtourolleandClaude Opus 5	10b10569e2	Add the Identity screen ... A third top-level screen beside the library and develop, because naming a cluster and pulling a stranger out of it are tasks with their own rhythm and need the whole window. The screen is designed around the clustering being wrong, which is FR-CULL-10 rather than pessimism: grouping over-merges on siblings, on parents and children, and on the same person a decade apart. So Split off sits next to Confirm all rather than behind a menu, the confirm/reject pair is on the face itself, and a group the system found is drawn differently from a person the user has vouched for. Splitting rejects before it confirms. Without that the next clustering pass suggests the face straight back and the user's correction becomes an argument they keep having. Face crops come from the proxies the grid already built, one decode per image rather than per face -- a group photograph holding six faces of one family is one JPEG. Where the calibration is not fitted the screen says confidence is unavailable instead of printing a percentage that looks measured, which is FR-CULL-9's rule at the point it becomes visible. The verdict controls use drawn icons, not tick and cross characters: ui/icons.slint exists because those render as tofu on Android. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 21:05:41 +02:00
dtourolleandClaude Opus 5	8ab9440190	Put the repair tool on the photograph ... A third chip beside Crop and Local, and the mode strip's own comment predicted the shape: a mode that arms a gesture on the canvas and scopes the column. Click a mark to cover it, drag the disc to move the repair, drag the source circle to say where the patch comes from, Delete to remove it. The source starts two and a half radii towards the middle of the frame, which is FR-DEV-8's automatic placement in its cheap form — dust sits on skies and skies are smooth, so it is usually right and always one drag from fixed. Two things are drawn deliberately. The circles are the size the repairs actually are, because whether a disc covers a speck is the whole judgement being made and a fixed-size dot would say nothing about it; the reach around them is padded to a touch target so a spot on a dust mark can still be picked up on a phone. And only the selected repair shows its source: a dusty sky carries a dozen, and two dozen circles with nothing saying which belongs to which is less information rather than more. The panel edits what is stored while the canvas draws what is mapped, and the two are pushed separately for that reason — a slider deriving its value from the drawn radius would move differently at different zoom levels. It is also the one panel built from SliderRow rather than a live track: a repair has no OpId to coalesce a drag under, so a row that fires once per gesture is what keeps undo one step per decision. Verified as far as this environment allows: the strip renders and the column re-scopes, photographed under XWayland. Synthetic clicks do not reach this application, so the gestures are as-written rather than as-felt, and docs/spot-removal.md says so. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:58:13 +02:00
dtourolleandClaude Opus 5	2ac069a6b3	Index the library's faces, and group them into people ... Wires dr-face to dr-catalog: a background sweep that reads the proxy the grid already built, detects, aligns, embeds and stores, then a clustering pass that turns those embeddings into suggested people. Detection runs on the Large thumbnail tier and nowhere else. That is what makes the feature affordable -- a browsed library has already paid for its proxies, so face indexing adds no RAW decode that was not already happening -- and it is why an image whose proxy is missing is skipped rather than fetched: requesting one here would put face indexing on the network path FR-CULL-8 keeps it off. The sweep keeps no cursor. It asks the catalog what is missing, so it resumes after process death with no repeated work beyond the in-flight image, and cancelling is dropping the receiver. recluster writes only the suggested half. Confirmed faces go in as anchors and come back untouched, and a cluster of one stays nameless -- naming every stray face would fill the People view with noise the user then has to dismiss. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:49:30 +02:00
dtourolleandClaude Opus 5	871a0eac28	Check the image has the commands before CI finds out it does not ... Two failures in a row were the same shape: a command the workflow calls was not in the image. git-lfs, then file(1). Each cost a full run to learn, and the android job is expensive to be wrong in -- the step that fails is at the end, so every attempt paid twenty-eight minutes of cross-compile first to reach the line that could not work. Both were visible in ten seconds from here. `docker run <image> command -v file` is the whole diagnosis; it just never occurred to anybody to ask before pushing. So the question gets asked automatically. This reads the `run:` blocks out of the workflows, pulls the commands worth doubting -- the ones a minimal Debian plausibly lacks, not `cd` -- and checks each against the image its job declares. It does not run the workflow and is not a replacement for one. It answers exactly the question that was expensive to answer. `git lfs` is handled specially and the comment says why: it is a subcommand, so the first word of the line is `git`, which is always there. Taking first words alone would have missed the original bug -- and did, in the first version of this script. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 20:31:25 +02:00
dtourolleandClaude Opus 5	d06b24c485	Regenerate the traceability matrix after the 0.7.0 work ... Seven more tags -- 621 against 614 -- from the print path, the film table rebuild and the projector lamp. Coverage is unchanged at 51.4% (91/177): the new tags land on requirements that already had one. Nine rows move. As before, the matrix links to line numbers, so a commit that inserts anything above a tag rewrites its row without changing what the row says. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 20:28:27 +02:00
dtourolleandClaude Opus 5	2958444835	Let undo reach the repairs before the tool can make one ... A history state was a Preset — a map of scalars — which was right while every edit in the graph was a parameter. A spot is not one, and undo is the first thing anyone does with a repair: place it, dislike it, take it back. Left as it was, that press would have stepped some unrelated slider and left the spot on the photograph, which reads as undo being broken rather than as undo being absent. So a state is now the pair, params and spot set. The same door is the one the mask stack will come through: mask edits are outside undo today for exactly this reason, and FR-DEV-5 is not finished until they are not. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:27:34 +02:00
dtourolleandClaude Opus 5	f00b3ae924	Take the tone from the hole and the texture from beside it ... A clone gets the texture right and the level wrong. Dust on a gradient sky is copied from a patch a little lighter than the hole it fills, and the repair reads as a disc even though every grain in it is correct — which is why FR-DEV-8 asks for heal and not only for clone. Heal adds the membrane: the difference between the two neighbourhoods, sampled at twenty-four points around the rim and interpolated across the disc by inverse square distance. Solving the Poisson problem properly is tens of Jacobi iterations, and an iteration here is a dispatch — sixty dispatches to remove a dust spot is not a frame budget. The closed form costs one loop over the rim, no state, and no second pass. The spec called for mean-value weights; inverse squares are two transcendentals per sample cheaper and agree wherever the boundary difference varies smoothly, which is every repair anyone makes. What decides whether that trade holds is the measurement, so the measurement is the test: on a ramp steep enough to leave a clone wrong by 38 levels out of 255, the heal is wrong by 0. docs/spot-removal.md §6.1 records what shipped and what it would take to go back. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:22:10 +02:00
dtourolleandClaude Opus 5	5323608051	Draw the repairs, before anything sharpens what they removed ... A spot set now composes detail passes of its own, one per round, and they go ahead of every operation's kernel. That placement is the decision worth recording: a sharpening pass reads a neighbourhood, so sharpening a dust mark before removing it smears its edge into pixels the repair's disc does not cover, and what survives is a faint over-sharpened ring around an otherwise perfect patch. It also disagrees with ARCH §5.2, which draws spot removal after clarity — docs/spot-removal.md §5.1 is where that is argued out. Every length reaching the shader is in render pixels, converted here where the framing is in scope. Both the centre and the source go through `Framing::output_at` — the same map the fused pass applies to every pixel — so a rotated photograph rotates the offset with no trigonometry, and the radius is found by mapping a point one radius above the centre and measuring, rather than by multiplying by a ratio this function has no business knowing about. The tests turn and crop the frame and expect the mark to stay gone, which is the property that arrangement buys. compose_full now takes the spot set, because a photograph with a repair and no sharpening still has a detail stage: a fused pass that encoded its own output there would quantise twice and bind to a texture of the wrong format. compose_detail_for takes the source size for the same kind of reason — a RenderScale describes the region on screen, and a spot is stored against the photograph. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:17:51 +02:00
dtourolleandClaude Opus 5	00e78dc2ac	Cluster faces into people, and calibrate what a similarity means ... FR-CULL-9 forbids thresholding a bare cosine anywhere in the subsystem, so calibrate fits P(same person) per library and reports whether the fit is trustworthy. Two details carry most of the weight. The fit runs against a 200-bin histogram rather than a pair list: a 25,000-face library has ~3e8 pairs and no gradient descent is running over that. And a fresh library has no valid calibration, because the positives have to come from user confirmations or burst siblings -- bootstrapping them from high cosine would fit the calibration to the belief it was supposed to test. Clustering defends against the over-merging FR-CULL-10 warns about with constraints rather than a better threshold: two faces in one photograph never merge, and two groups confirmed as different people never merge. Average link rather than single link, so one strong edge cannot weld two families together. Calibration is defined once, in dr-face, and dr-catalog re-exports it. Two implementations of one probability model is exactly how a number comes to mean the wrong thing. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:10:33 +02:00
dtourolleandClaude Opus 5	aac3136407	Store faces and the people they belong to ... Schema v8: people, faces, face_person, face_person_rejected, and the per-library calibration. Follows catalog.md 10.1 with two additions the spec work turned up. crop_px, because at the 1024px proxy tier a group shot reaches the embedder at ~50 source pixels upsampled to 112 and a portrait at 340. FR-CULL-9 names face size as an axis along which an uncalibrated similarity misbehaves, so it is a stored feature rather than a UI hint. face_person_rejected, because rejection is not the absence of an assignment. Without it the next clustering pass re-suggests exactly the face the user just pushed away, and the tool feels broken. record_detections replaces rather than appends, since DetectFaces is coalesced per image -- and carries confirmations across the replacement by box overlap, so re-indexing with a better model cannot discard the user's own labelling. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:02:52 +02:00
dtourolleandClaude Opus 5	6997c0f7ac	Let a detail pass carry a list, not only a kernel ... Every neighbourhood pass so far has been a convolution, whose whole description fits in the uniform block because its structure fixes how many numbers it needs. Spot removal is not that shape: sixty-four repairs and one repair are the same shader with a different buffer behind it. So a pass may declare `storage`, which arrives at binding 3 as `array<vec4<f32>>` with `arrayLength` in scope. The alternative — packing the list into uniforms — needs a fixed maximum paid for on every frame, a composer that can emit vec4 fields because a uniform array's stride is 16 whatever it holds, and it gives the next operation that wants a table nothing to build on. The property worth having is what stays out of the generated source: the count is in the buffer, so placing the tenth spot uploads 512 bytes and reuses the compiled pipeline, exactly as moving a slider does for the fused pass. `changing_the_list_does_not_recompile` is that, asserted. One bind group entry rather than two more layouts, and one placeholder buffer allocated in `new` rather than sixteen bytes per pass per frame — a zero-length storage buffer cannot be bound, and per-frame allocation is what this module's documentation exists to refuse. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 20:00:24 +02:00
dtourolleandClaude Opus 5	19981c1033	Detect, align and embed faces with SCRFD and MobileFaceNet ... Ports the pipeline from the C++ reference in ../scene-actor-extraction (MIT, same author). End to end on real portraits it separates identities the way the reference's fitted calibration says it should: 0.596 between distinct photographs of one person, 0.05 between different people, either side of MBF's 0.267 boundary. Three things are structural rather than incidental: Aligned112 can only be built by align::warp, so Embedder::embed cannot be handed an unaligned bounding-box crop. That mistake yields 512 plausible unit-norm numbers and no error, so the type system refuses it instead. Embedding carries its ModelId and cosine() returns None across models, because a cross-model similarity is the one mistake that produces plausible garbage rather than a failure. The model-free half -- alignment, embedding arithmetic, f16 storage -- sits outside the inference feature and is covered by 11 tests that need no weights on the machine. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 19:57:56 +02:00
dtourolleandClaude Opus 5	44444f7768	Write the repairs down, one line each, and merge them by id ... A spot is eight numbers, so it goes in the version block as a line rather than in a block of its own the way a mask does — sixty-four blocks would bury the rest of the file. A line per spot rather than one line for the set, because the line is what a diff shows and what a merge resolves. The parse arm sits ahead of the op.param arm deliberately: without it, `spot.abc123 = 0.4 0.6 …` reads as an operation called `spot` whose value will not parse, and the repair is dropped with a warning about a corrupt number. The prefix is safe precisely because a spot is not an operation, so no ops/ declaration can claim the name. Merging is merge_masks by id, one level down, and it is where the derived ids earn their keep: two devices that removed different marks hold different ids and both survive, while two that removed the same piece of dust hold the same id and the merge sees the one repair it is. A spot both sides dragged resolves whole to the higher revision — eight numbers describe one disc, and half of each is a repair neither photographer made. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 19:52:54 +02:00
dtourolleandClaude Opus 5	72410f39c6	Answer M1: tract loads both face graphs once their dims are pinned ... Neither InsightFace export parses as shipped -- SCRFD fails at its input node, ArcFace at the first Conv -- which is the same wall dr-segment hit on YOLO's dynamic export. Both load cleanly with the input dims frozen, so the pure-Rust runtime holds for the face pipeline too. tools/fix-face-model-shapes.sh does the freezing, and exists so the artefact is reproducible rather than a binary someone once produced. It takes two forms because the two graphs need different ones: ArcFace's batch is a named dim_param, SCRFD's H and W are dynamic but unnamed. Also notes YuNet loading with no intervention, which matters for the licence question in faces.md 2.3. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 19:51:02 +02:00
dtourolleandClaude Opus 5	97479a0512	Hold the repairs a photographer makes, and say which may share a pass ... A spot is a disc, a source offset and four numbers, and it lives beside `ops` for the reason `masks` and `film` do: the operation trait is ParamId -> f32, and a list of repairs is neither scalar nor fixed. Two decisions here are not obvious. The id is derived from the position rather than counted, because two devices editing offline would each mint `spot3` for different marks and the sidecar merge would then treat two repairs as one — from the position, two devices that removed the same piece of dust agree, and two that removed different ones do not. And every length is in the frame's isotropic units, not a mixture of those and shorter-edge fractions: one unit for the radius, the feather and the offset agrees on a landscape frame and on a portrait one, where a mixture only agrees on the first. `rounds` is the arithmetic that keeps a source from reading a destination. Every spot in one pass reads the photograph as it stood before that pass, so a spot sourcing from an earlier spot's destination would copy the mark that spot was removing. Grouping is not a pass per spot — that is sixty-four dispatches for a case that almost never arises — it is a new round only when the sources actually collide. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 19:50:00 +02:00
dtourolleandClaude Opus 5	ec740115b6	Spec the face pipeline on SCRFD and MobileFaceNet ... FR-CULL-8..12 specify the subsystem in terms of "a 512-dimension embedding from a stated model" and stop there, because D13 was open. This names the models, and grounds them in the measurements and the working C++ pipeline in ../scene-actor-extraction rather than in a literature reading. The licensing half of D13 stays open, but with a route through it: the InsightFace weights are non-commercial and cannot be committed, so the app ships the code and the user fetches the model. faces.model_id already makes that a survivable choice. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 19:44:39 +02:00
dtourolleandClaude Opus 5	8d8d6491ad	Say how spot removal is going to work before writing it ... FR-DEV-8 is the last develop requirement with nothing behind it. The pipeline it lands on already has most of the parts — the neighbourhood stage, the mask crate's normalised coordinates, the gradient handles' canvas drags, the merge-by-id rule — so the spec is mostly about the four things that are genuinely new, and about the two places where the obvious implementation is the wrong one: a Poisson solve is sixty dispatches per spot, and a per-frame readback to find out what colour a sky is would undo ARCH §6.1. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-08-26 19:43:30 +02:00
dtourolleandClaude Opus 5	f82c69bc6b	Say which version this is: 0.7.0 ... Film simulation is a feature, not a fix: five Ilford stocks, a grain model that counts silver rather than adding noise, and the pipeline and UI to drive them. 0.6.0 was tagged thirty-one commits ago and does not describe any of that. The Android versionCode follows from this without being restated -- package.sh packs MAJOR*10000 + MINOR*100 + PATCH, so 0.7.0 is 700, above the 600 already installed on devices and therefore an upgrade rather than a refusal. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> v0.7.0	2026-08-26 18:14:23 +02:00
dtourolleandClaude Opus 5	f14176de29	Give the projector lamp a name instead of a warning ... Every bake of a Vision3 stock logged: unknown illuminant "K75P", falling back to D55 K75P is a cinema xenon short-arc lamp, and Kodak 2383 and 2393 -- the projection print films those stocks print onto -- name it as the light their result is looked at under. It was never implemented, so it fell through to the unknown branch. The fallback was the right family: a xenon arc sits near 6000 K, close to daylight and nothing like the tungsten enlarger above it. So the pixels do not move. What changes is that D55 is now a documented choice rather than the consolation prize for an unrecognised string, with the approximation stated -- an arc has line structure a Planckian curve cannot express, and the residue of that is small here because the viewing step adapts the white point out either way. The test is the point of the commit. A profile naming a light nobody implemented should fail the suite, not whisper into a log that only gets read when somebody happens to be looking for something else -- which is how this was found. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 17:41:34 +02:00
dtourolleandClaude Opus 5	4b50648870	Rebuild the film tables when the film's own sliders move ... Push and print exposure did nothing at all. The parameter was set and the tables were never rebuilt, so the shader went on running the stock as it had been baked. The check asked the wrong list: self.rows() // one entry per *parameter*, tab-filtered .get(op_index as usize) // indexed by an *operation* index .map(|_| ()) .and( ...the real check... ) `op_index` counts over the scoped capabilities -- it is what `lookup` resolves a slider through -- so capabilities is the only list to ask. Indexing `rows()` served no purpose, and past its end the `and` short-circuited to None and the rebake silently never happened. Both halves of that are worth saying. It was wrong, and it was convoluted, and the convolution is what hid the wrongness: a one-line check would have been obviously right or obviously broken. This is a class of bug the suite cannot reach. The tables are rebuilt in the interface layer, in response to a control, and every test either side of it passed throughout -- dr-film computed the pushed curves correctly and the shader rendered whatever it was handed. Only moving the slider showed it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 17:35:25 +02:00
dtourolleandClaude Opus 5	ff1a3e0e80	Print the black-and-white negatives instead of showing the scan ... Choosing Ilford HP5 Plus showed an inverted grey frame. So did Double-X. They are negatives, and upstream leaves `target_print` null on every monochrome stock, so nothing was ever printed and the scan was all there was. A colour negative at least announces itself -- the orange mask says plainly that you are looking at a negative. A monochrome one just looks broken. They print on Kodak 2302 now, which is a monochrome print film and is what such a negative is actually printed onto; Double-X onto 2302 is the standard cine chain. For the Ilford stocks it stands in for an Ilford paper, which nobody has measured, and is at least the right kind of material. The scan is still reachable through the Scanned/Printed toggle. It is a thing to choose now rather than the only thing on offer. `every_shipped_stock_bakes` did not catch this, and could not: it derives "should this be inverted?" from the stock's kind *and whether it names a paper*, so it looked at an inverted HP5, concluded that was right for an unprinted negative, and passed. The assertion was self-consistent and the situation was still wrong. The new test asserts the thing that actually matters -- a camera negative must name a paper, that paper must be a printing stock, and it must be the same kind of material, so a monochrome negative cannot end up on colour paper. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 17:35:25 +02:00
dtourolleandClaude Opus 5	ce6458547a	Develop longer, from the measurements rather than from a contrast slider ... Pushing was not a thing to simulate. It was measured data being thrown away: Double-X and 2302 each ship five characteristic curves, one per development time, and this shipped the 6.5-minute column and discarded four. All five now ship and interpolate. The axis is real. Double-X runs 4 to 12 minutes, and across it the average gradient goes 0.472 to 1.034 while Dmax goes 1.19 to 2.56. The control is in stops, because that is what a photographer means, and one stop is a factor of about 1.41 in time. That mapping is checked rather than assumed: against Double-X's own axis it lands within 2% of the 9-minute column for +1, and near 12 minutes for +2, which are the times the datasheet gives for exactly that. There is a test. **Pushing must not recover shadow detail, and this does not.** Across the whole measured range the speed point moves about a third of a stop while the gradient doubles; three stops under mid-grey, density goes from 0.008 to 0.035, which is still nothing. Developing longer multiplies what was already recorded and cannot record what never hit the film. A push built as added exposure or global contrast brightens those shadows instead and looks convincing until someone who shoots film sees it, so that property has a test of its own. Interpolated in *log* time, because development is multiplicative: 4 to 5 minutes is the same amount of push as 9 to 12, and interpolating linearly would bunch the control at one end. Clamped at both ends, because past the published range there is no data and extrapolating a contrast curve invents an emulsion nobody tested. A stock measured at one process ignores the control entirely rather than inventing a curve for it -- Portra 800's pushes are separate *measured* profiles, which is the honest way to offer those. Costs nothing per pixel and changes no shader. The curves are a per-stock table, so the interpolation happens on the CPU at bake time, where choosing a stock and moving its sliders already rebakes. The Vulkan shader is untouched. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 17:35:25 +02:00
dtourolleandClaude Opus 5	8f863ec56f	Put file(1) in the Android image too ... The android job now fetches the model and cross-compiles the whole app -- 28 minutes of it -- and then dies on file: not found (exit 127) `Verify minimum API level` reads the linked API out of the .so's ELF notes with file(1), and the image has never had it. Like git-lfs, the absence could not show until something got that far: every previous run panicked in dr-segment's build script long before this line, so the step that was going to fail never ran. Audited the rest of what the remaining steps invoke against the image rather than find the next one the same expensive way -- zip, keytool, base64, mktemp, shred, find, sed, awk, and aapt2/zipalign/apksigner/d8 from build-tools are all present. file was the only gap left. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 14:21:05 +02:00
dtourolleandClaude Opus 5	b9ee29f3c7	Regenerate the traceability matrix for the Ilford stocks ... Same drift as before and for the same reason: the matrix links to line numbers, and the grain work moved lines in files that carry tags. The stocks bring six new files and ten new tags -- 227 scanned against 221, 614 found against 604 -- all on requirements that were already covered, so coverage is 51.4% (91/177) either side. Twelve rows move and nothing else changes. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 10:14:29 +02:00
dtourolleandClaude Opus 5	40e6334bb1	Sign the APK with a real key when one is configured ... The APK has been debug-signed with a key generated on the spot, which is right for putting a build on a test device and useless for anything else: a different signature every run, so nothing can ever update in place. Four secrets now select a real signature -- ANDROID_KEYSTORE_BASE64 and its password, alias and key password. The names are JellyTau's, because that repo already signs its Android build this way against this same runner and one convention across both is one thing to remember. Absence of the secrets is not an error. A fork or a branch build has no access to them and should still produce an installable APK, so the debug path stays exactly as it was. The reverse is an error: if a keystore is supplied and cannot be read, the build fails rather than quietly falling back to a debug key, because a release that is silently debug-signed is worse than no release. Passwords reach apksigner and keytool as `env:`, never `pass:`. `pass:` puts the password in the process table for anything on the box to read. The keystore is written to a 0700 mktemp directory and never into the workspace, which is both what actions/cache saves and what the upload step globs. Also: upload-artifact drops from v4 to v3. v4 was a guess about what this Gitea supports. v3 is what JellyTau uploads its APK with on this runner today, which makes it the version known to work rather than the one that ought to. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 10:13:50 +02:00
dtourolleandClaude Opus 5	e9b3598841	Add five Ilford stocks, and say plainly that they are constructed ... They were asked for and they are here, but not on the same footing as the Kodak profiles, and the files say so in their first line. What I had claimed, and had to withdraw: that Delta 100 is "quoted around 9" and HP5 "around 12". Ilford publish no such figures. The word granularity does not occur anywhere in their technical information -- grain is described as "fine" and "finest" and nothing more. That claim was in this crate's documentation as though it came from a datasheet; it is corrected there too. Two further traps found while looking: - Kodak colour negatives publish Print Grain Index, not RMS granularity. PGI is a perceptual scale from viewer surveys -- 25 is roughly the threshold of visibility, four units a just-noticeable difference -- and Kodak state it cannot be compared to RMS. So a Portra number cannot be dropped into the granularity field, and none has been. - RMS proper is published mostly for black-and-white, reversal and motion picture stocks. Every shipped stock therefore still carries the same default, which means grain does not yet tell one film from another. That is per-stock data, not code, and is now written down where somebody will find it. So the Ilford profiles are built rather than extracted, and each part rests on something different: speed published and exact -- ISO 400/27 for HP5 is a fact contrast ISO 6:1993's normal development, average gradient 0.62 spectral borrowed from Kodak Double-X, a *measured* panchromatic negative, shifted by the speed difference. Conventional panchromatic sensitisation is much alike across black-and-white films, and this is far better founded than reading pixels off a printed curve silver neutral, which is not an approximation: developed silver absorbs flat, and Double-X's measurement is flat granularity estimated, ordered by each film's known relative grain They render as a film of that speed and contrast. They are not a measurement of that emulsion, and the two stocks that share a speed differ only in the estimated part. `every_shipped_stock_bakes` is tightened to match, because a constructed profile fails in a way a measured one does not: the curve parses, bakes, and sits entirely off one end of its own exposure range, rendering every frame black or blown while passing a finiteness check. It now asserts mid-grey lands somewhere photographic and that the tone response runs the way the stock's kind says it should. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 10:08:51 +02:00
dtourolleandClaude Opus 5	4b2ee0ac50	Count the silver instead of adding noise ... An emulsion is a suspension of crystals. Light sensitises some; development turns a sensitised one opaque, all or nothing. So a patch of film's density is a *count* of developed grains, and a count of independent yes/no events has a variance whether or not anyone wanted texture: mean = D variance = D * (Dmax - u * D) / N That expression is the whole feature. It peaks in the middle of the density range and vanishes at both ends -- clear film has nothing developed to vary, black film has nothing left to develop -- so grain lives in the midtones as a consequence rather than as a "midtone bias" slider. I was wrong earlier that this needs the detail stage. Nothing in it reads a neighbouring pixel; the only reason to move it was that grain must be fixed in film space rather than screen space, and that solves itself: N is grains *per pixel*, so it scales with the film a pixel covers. Zoom out, each pixel averages more grains, less variance -- correct, with nothing super-sampled and nothing filtered. It stays in the fused pass. Grain goes on the density and *before* the dye, which is the physical order and not cosmetic. Perturbing the finished colour -- what an effect does -- tints highlights wrong, because that noise never passes through the dye. Crystal habit lives in `rms_granularity`, the number every datasheet publishes, now a profile field. It measures exactly what differs between a cubic emulsion and a tabular one: at equal speed, tabular crystals present more area per unit silver, so the film reads finer. Delta 100 is quoted near 9 where HP5 is near 12, and that gap *is* the habit. Adding a stock whose grain is its whole reputation is therefore editing one line, not writing a model. Three things this cost, all of them worth writing down: - The default granularity is a colour negative's, blue coarsest. Applied to Tri-X it put *colour* speckle on a black and white photograph. Monochrome stocks collapse it at parse, where every other per-layer table is already replicated from the one measured channel. - Helpers cannot read uniforms. The composer prefixes a uniform with its operation's id and rewrites references inside a fragment body only; helpers are shared and deduplicated, so a bare `gn0` names nothing. `film_lut` already took its size as an argument for this reason, and now says so. - The end-to-end test compares the shader against the CPU model, and grain is stochastic, so that comparison now runs with grain off. Which means a grain that never left the CPU would look exactly like a passing suite -- hence a second test that grain off is bit-identical, one grain per pixel moves it, and ten thousand move it less. Not here, deliberately: no grain slider. The parameters are physical and `rms_granularity` is the honest place to scale one from, but its range wants choosing rather than guessing. Nor a film format -- 35 mm is assumed, and medium format at the same stock is far less grainy per unit of picture. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 10:08:51 +02:00
dtourolleandClaude Opus 5	1d38015a7b	Put git-lfs in the Android image, which never had it ... The Android job has never fetched the model. Not because of the header collision the desktop job hit -- that one is fixed and the desktop job now pulls all 11 MB -- but because the image has no git-lfs at all: git: 'lfs' is not a git command. See 'git --help'. The fetch step dies on its first line, `git lfs install --local`, before any of the auth handling runs. The build then panics in dr-segment's build script with a message telling you to run `git lfs install && git lfs pull` -- advice that could not have worked, because the client it names was never in the image to run. Both jobs failing their fetch step at the same time made this look like one bug with one cause. It was two, in two different images, and the desktop one was noisier: it had a client, so it got as far as an HTTP error worth reading. The android one had nothing to say beyond the name of a missing command. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 06:52:50 +02:00
dtourolleandClaude Opus 5	56bdd457dd	Stop the test build filling the runner's disk ... The desktop job died mid-link with LLVM reporting "IO failure on output stream", which reads like a compiler crash and is not one: underneath it is `No space left on device`. The runner ran out of disk while linking. Worth knowing what it was spending it on. `target/debug` was 24 GB against `target/release`'s 2.6 GB -- the test build is roughly ninety percent of the footprint -- and of that, 15 GB was debug info in `debug/deps` and 3.6 GB was incremental state. Neither buys anything here. Nothing attaches a debugger to a CI run, and incremental compilation exists to make the second build in a working tree fast, which is not a thing a fresh checkout ever has. With both off the same tree is 3.3 GB, `debug/deps` 2.8 GB, and the test binaries build unchanged. Backtraces keep function names and lose file and line numbers; if a failure ever needs those, DEBUG=1 gives line tables back for a fraction of the 15 GB. A `df -h` either side of the expensive steps, so the next time this happens it says so in one line rather than as an error from LLVM. This is a mitigation and it should not be mistaken for the fix. It bounds what this job asks for; it cannot help if the runner is full of anything else, and 24 GB of build output is not obviously the largest thing on a host that also keeps every cached target directory this workflow has ever saved. If it fails here again, the disk needs looking at on draco-x86. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-26 01:13:32 +02:00
dtourolleandClaude Opus 5	834b219c3f	Publish the Android APK as a build artefact ... The android job proved the app links for aarch64 and then threw the result away. There was no APK anywhere in CI and no upload step in the repo at all, so a green run left nothing anybody could install -- the artefact list was empty by construction, not by failure. It now assembles the APK with the script package.sh uses and uploads it. The .so comes from the build the API-level check already ran; -o only adds a copy of it where the packaging step looks, so this costs one copy rather than a second twenty-minute cross-compile. The signing key is the part worth being careful about. KEYSTORE points at a mktemp directory rather than its default under target-android, because that directory is precisely what actions/cache saves and restores -- the default would have written a private key into the build cache and kept it there. Nothing but the .apk is uploaded. A fresh debug key each run is the right trade for an artefact meant to reach a test device: the only thing a stable key buys is installing over a previous build without uninstalling first, and a key that survives in cache storage to buy it is a bad exchange. if-no-files-found: error because the failure being guarded against is a green run with an empty artefact list, which reads as success right up until somebody goes looking for the file. Debug-signed, arm64-v8a only -- the ABI the job already builds. Neither is a release story; this is a build you can install. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-25 23:38:20 +02:00
dtourolleandClaude Opus 5	5741ec5e00	Lift the APK assembly out of package.sh so CI can run it too ... package.sh does two things: it decides how the host reaches the image, and it assembles an APK once inside it. Only the first half is host-specific. CI already runs in that image, so the second half was about to be copied into a workflow step -- two copies of aapt2/zipalign/apksigner ordering, drifting apart at whatever rate the toolchain moves. So it moves to docker/android/assemble-apk.sh, which assumes it is inside the image and takes its paths from the environment, because the callers disagree about them: the container mounts the repo at /work, the runner checks it out wherever it likes. Every default reproduces what package.sh did, so the host path is unchanged. Two things stop being hard-coded on the way. The build-tools version and the compile SDK are resolved from what is installed rather than written out as 36.0.0 and android-36 -- the versions are Dockerfile ARGs, and a second copy is a second thing to miss when they move. --min-sdk-version now comes from that same ARG instead of a literal 28, which is the number the API-level check in CI already reads. The intermediates are removed at the end. They were harmless in a cache directory nobody looks at; beside a published artefact they are four more files for a glob to pick up by mistake. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-25 23:38:08 +02:00
dtourolleandClaude Opus 5	8b07a90e71	Regenerate the traceability matrix the clippy pass moved ... `traceability-check` fails on master: the committed matrix does not match what the generator produces, so the gate that exists to keep the two in step is the thing reporting they are not. Nothing was traced or untraced. Coverage is 51.4% (91/177) before and after, the same 604 tags against the same 177 requirements; every one of the 32 changed rows is a line number that moved when the clippy warnings were cleared -- `adjust.rs:2150` is now 2164, 632 is 651, 751 is 770. The matrix links to lines, so touching a file above a tag rewrites its row without changing what it says. Regenerated with `cargo run -p traceability -- report`, which is what the failing step tells you to run. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-25 23:21:13 +02:00
dtourolleandClaude Opus 5	606f85df34	Send the LFS object endpoint one Authorization header, not two ... The model fetch has been failing on every run with LFS: Client error: .../info/lfs/objects/0672d7a7... which reads like a rejected credential and is not one. The object is on the server and downloads fine; what fails is the shape of the request. `git lfs pull` makes two calls. The first, to `/info/lfs/objects/batch`, succeeds -- and Gitea answers it with a short-lived `Bearer` JWT scoped to that one object, for git-lfs to use on the second. git-lfs sends that JWT *and* the `Authorization` header this step had installed in git config, and two `Authorization` headers is a 400 from Gitea. Hence a client error on the object one step after the batch call it just made successfully, which is what made this look like an auth problem rather than a duplication. Confirmed directly against the server: the JWT alone on that URL is a 200, the JWT plus any second `Authorization` is a 400, and a lone token header that is merely wrong is a 401 -- so the scheme was never the issue. `lfs: true` on the checkout fails the same way and for the same reason, because actions/checkout persists a header of its own; the comment here blaming a credential the endpoint would not accept was wrong on both counts. So the headers are stripped -- checkout's included, since nothing later in either job talks to the remote -- and the token is handed to git-lfs as an ordinary credential instead. It authenticates the batch call and leaves the per-object JWT alone. This is what fails the Android job today: the build script sees a 133-byte pointer and panics by design, which is the message it is supposed to give and the one nobody could act on. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-25 22:55:26 +02:00
dtourolleandClaude Opus 5	56978fdf35	Clear the clippy warnings that were failing CI before this branch ... Nothing here is film simulation. These are lints that fail master today, under the -D warnings CI runs with, mostly from a toolchain that learned new ones rather than from anybody's code -- `is_multiple_of` and the derivable `Default` did not exist as lints when this was written. They are fixed rather than allowed, and by hand rather than by trusting `cargo clippy --fix` wholesale: its automatic pass split a derive in two and left a stray blank line, which is the sort of thing that is correct and still wrong to commit. The four that needed a decision rather than a rewrite: - The distance transform's inner loop writes through its iterator now. `q` stays, because it is the position the parabola is evaluated at as well as the index it is written to -- the lint is about the write. - `to_source` and `to_proto` take `self` by value. Their receiver is `Copy`, so this is the same machine code and the honest signature. - The export path's return type is five levels deep and now has a name, plus a line saying why the `Option` wraps the `Result`: `None` is cancellation, which is not a failure and has no error to report. - A test fills a range instead of looping over one. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-25 22:35:02 +02:00
dtourolleandClaude Opus 5	3b5952769b	Emit floats an f32 can hold, and drop the format! that formats nothing ... CI runs cargo fmt --check and clippy -D warnings, and this branch had never been through either. Both would have failed it. The bulk was the generated colour tables: eight significant figures where an f32 carries about 7.2, so the eighth is noise that rounds away at compile time and clippy's excessive_precision says so 109 times over. Fixed in the generator rather than only in the file, so it stays fixed -- and the file is trimmed in place rather than re-derived, because regenerating it needs a colour-science stack that has nothing to do with the defect. The format! in the composer is mine too, from extracting the rendering tail: the braces in it were escaped because the text used to live inside a larger template, and once extracted the escapes are noise and the call formats nothing. Also here, and clearly not mine: an unused import and a shadowed binding in dr-gpu, and an unused import in a test. They are pre-existing -- clippy has been failing on master before this branch existed, on lints like is_multiple_of that arrived with a toolchain rather than with anyone's code. Fixed because CI cannot go green around them, and called out because a merge commit is a bad place to quietly edit someone else's crate. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-25 22:28:14 +02:00