364 Commits

Include renames

Author	SHA1	Message	Date
dtourolleandClaude Opus 5	6acc73a9fa	Drag a collection onto another to nest it ... The tree could be built nested but never rearranged: `set_parent` existed, with its cycle check and its tests, and nothing in the UI called it. A collection created in the wrong place stayed there. Each row is already a drop target, so it becomes a `DragArea` too — wrapped at the instantiation site the way the grid's cells are, which keeps the row's own TouchArea nested underneath and a click still selecting. `allow-move`, not copy: a collection has one parent, unlike a photograph, which is filed in as many collections as you like. The drop is handed only the target's id, so the source is remembered from the press that precedes the drag — Slint builds the payload through a `pure` binding, which must not have side effects. An image drag always fills `dragging`, so an empty payload with a remembered row is unambiguously a rearrangement; the row is taken rather than read, or a later empty drop would move a collection nobody touched. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 22:54:50 +02:00
dtourolleandClaude Opus 5	edcaf42ded	Show only the photographs taken in the period you are looking at ... The library could be narrowed by rating, flag and availability, but not by when a photograph was taken — so finding a fortnight meant scrolling to it and holding position. The range rides on `RatingFilter` for the reason `local_only` already does: every query path threads that one struct, so the count in the header cannot claim a total the grid does not draw. Undated images are excluded whenever either end is set — they cannot be inside or outside a span, and drawing them made the range look as though it had not applied. Taken from the timeline rather than typed into two date fields. Finding the period is what the histogram is for, and having found it the user should not have to read the dates off the axis and key them back in. The histogram keeps drawing the full extent while the range is on, or there would be nowhere to widen back out from. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 22:38:26 +02:00
dtourolleandClaude Opus 5	d2d5d6f22b	Keep the selection visible when the grid scrolls under it ... Scrolling rebuilds every cell, and a fresh cell carries `selected: false`. `sync_badges` and `sync_ratings` refilled what the rebuild cleared; nothing refilled the selection, so the ticks vanished on every scroll. The selection itself was never lost — it is a set of image ids and survives untouched — which made this worse than losing it: the header buttons still acted on forty photographs the user could no longer see were held. `load_window` reaches the selection through a `Weak` handle set at wiring. Weak because the two controllers are joined only through the window, and an `Rc` each way would leak both; absent, the grid draws nothing selected, which is what it did before. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 22:30:37 +02:00
dtourolleandClaude Opus 5	02d629922f	Draw the date histogram over the collection you are looking at ... The timeline counted the whole library whatever the grid was showing, so opening a collection left a fortnight in Arosa as one column of a fifteen-year axis — an axis describing photographs that were not on screen. Scope the buckets and the span to the same collection and rating filter the grid uses. `timeline_range` counts `images` alone and cannot express the membership join, so the scoped query lives beside the other scoped readers in the UI and shares their descendants-of-scope rule. `catalog_span` now delegates to the same scoped reader. Zoom and scrub measured the full library while the bars were scoped, so a scrub could land on an instant the collection did not contain and send the view somewhere the user had not asked to go. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 22:15:59 +02:00
dtourolleandClaude Opus 5	09f6cf8c0f	Name the sidecar the drain could not deliver ... The write path learned to say which file the server refused; the drain path — the one that runs for work queued while offline — still reported only a count. That is the path that matters most, because everything it carries was made with no connection and exists on one device. With it named, the failure on the tablet is: draining PhotosRaw/2026/2026-08-03/_MG_9221.drsc: permission denied on a credential that pushes the catalog to the same library root in the same pass. So this is not the account and not the app password: one path is writable and another is not, which points at the server — a read-only share over that folder, or a file access control rule on the extension — rather than at anything this side can retry its way out of. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 14:21:00 +02:00
dtourolleandClaude Opus 5	6a7ed37aed	Say which sidecar the server refused, and where ... A queued sidecar failing to upload was reported as a count and a reason — "1 queued sidecar(s) still undelivered: permission denied" — with the path only at debug level, which the app filters out by default. That is unsynced user work: a rating or an edit that exists on one device and nowhere else. Which photograph it belongs to, and which path the server refused, is the whole of what makes the failure actionable, and without it a 403 on a single file reads the same as a whole library failing to sync. Observed on the tablet, where writes to the derived folder succeed on the same credential — the catalog pushes fine — while one sidecar beside its image is refused. That combination says the path matters and the account does not, so the path is the thing worth printing. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 14:11:30 +02:00
dtourolleandClaude Opus 5	dea826811e	Render a blown highlight white instead of magenta ... Every clipped sky came out bright pink. Measured, not guessed: developing _MG_8596.CR2 and looking at the export, the subject renders correctly and only the saturated region is wrong. A fully clipped pixel reaches the shader as (1, 1, 1) — three photosites that stopped counting, carrying no colour at all. The as-shot multipliers are not neutral, so balancing sends it to (1.93, 1.00, 1.68) on this body, and the camera matrix turns that into R 2.88, G 0.51, B 2.03. Red and blue clip at one; green, whose matrix row is far less positive-heavy, does not. Red and blue high with green low is magenta. Nothing upstream was at fault, which is why the two previous attempts missed it: the white balance is correct, the matrix is correct, and the sensor normalisation is correct. The input simply was not a colour, and correct arithmetic on a non-colour produces a confident wrong answer. So saturation is detected before the balance is applied — the last 1.5% of range, smoothstepped rather than switched, because a hard threshold draws a visible rim around every highlight and a backlit edge on skin is where that shows. Above it the pixel is pulled to the neutral of its own brightness, so it keeps its luminance and loses only the cast. Verified end to end on the file itself: the sky is white, the skin, the black dresses and the stone are unchanged. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 13:17:15 +02:00
dtourolleandClaude Opus 5	d2024da368	Scroll the develop column as one, histogram and all ... Only the sliders scrolled. The capture metadata, the histogram, the geometry controls and copy-and-paste all sat above them in a fixed layout, so on a 280px column in portrait they took the height the sliders needed — and the histogram, which is the instrument the sliders are judged against, could neither be scrolled to nor scrolled past. The Flickable moves out of `AdjustPanel` and around the whole column. Nesting one inside the other was not an option: a slider drag already has to be won against one scroller, and a second would give it a third thing to be lost to. `slider-dragging` becomes an `out` property for the same reason. The arbitration is unchanged and still necessary — a track stands the scroller down as soon as a finger touches it, or every attempt to drag a slider would scroll the column instead — but the scroller obeying it now lives a level up. The scroll gutter stays where it is, as padding inside the content: it is what guarantees somewhere to put a thumb that means "scroll" and nothing else, and it matters more now that it serves the entire column. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 12:53:04 +02:00
dtourolleandClaude Opus 5	c9c5c43aa1	Carry the library across when its home moves ... The previous commit moved durable data out of Android's cache directory, and on its own that would have been an upgrade that quietly discarded work. The app looks in the new location, finds nothing, and rescans a library of tens of thousands of images over the network — while the old copy, including every offline rating and edit that had not yet synced, sits in a directory the system is free to delete. So the account's directory is moved once at startup, before anything opens a store. A rename rather than a copy: both are inside the app's own data on one filesystem, so it is atomic and cannot half-finish. An existing destination wins and the move is skipped — that covers a second run and a fresh install, and in neither case may this overwrite live data. A failed move is logged, not fatal. The cost is a rescan, which is recoverable; refusing to start is not. Two tests, against ordinary directories rather than the platform's idea of a cache: one puts an unsynced sidecar in the old location and asserts it is readable in the new one afterwards, the other pins that live data is never overwritten. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 12:32:20 +02:00
dtourolle	4e62b89d17	Merge branch 'android-collections' ... Touch multi-selection in the grid, filing a selection into a collection without a drag, and taking a collection offline from a held row.	2026-08-17 12:27:49 +02:00
dtourolleandClaude Opus 5	d913e50948	Select photographs with a finger, and take a collection with you ... Two things a tablet could not do. Both existed for a pointer and had no touch form at all, which on Android meant the collection sidebar was somewhere to look at rather than somewhere to file into. **Selecting more than one.** Ctrl-click and shift-click are the only ways into a multi-selection, and touch has neither. Holding a cell now enters selection mode, where a tap toggles — reported to Rust as a ctrl-press, so it goes through the same `apply_press` as everything else rather than growing a second copy of the selection rules. A double tap takes the run between where selecting began and there: the touch form of shift-click, and the reason the anchor from *before* the double tap has to be remembered, since both of its taps move the anchor onto the cell being tapped. A "Select" button does the same thing where a gesture would go undiscovered (FR-UI-4). **Filing without a drag.** A one-finger drag beginning in the grid belongs to the Flickable that scrolls it — that is the arbitration working, not a bug to route around — so the selection can now be filed from a sheet listing the sidebar's own rows. Copy by default, as the drag has always been; moving out of the collection being shown is a switch, because it is the one that takes something away. **Taking a collection offline.** The machinery was there and reachable only by scoping the grid to a collection and finding a button behind a disclosure. Holding a collection's name now asks the question directly, and the tray on a row and the header button ask the same one — three affordances doing two different things is how a user comes to avoid all three. The question is asked rather than a toggle flipped because both answers are expensive: one downloads gigabytes, the other deletes them, and the counts and sizes go in the buttons where they are read before the tap. `Cache::release` is new and is the destructive half `unpin` deliberately is not. "Remove the local copies" is asked by someone whose device is full, and withdrawing a promise while leaving the bytes for a future eviction to notice is not an answer to it. It unpins before forgetting, or the next pin fetch would dutifully download everything it just deleted. The sidebar's trays read `tier_actual`, never `tier_desired`: the question is whether these will open on the aeroplane, and a pin whose download has not run yet answers no. TRACES: FR-CAT-7 | FR-NC-6a | FR-NC-6b | FR-NC-6c | FR-UI-2 | FR-UI-3 | FR-UI-4 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 12:27:29 +02:00
dtourolleandClaude Opus 5	b8e9793908	Keep offline work out of a directory Android empties ... The catalog, the sidecar cache and the export outbox were all landing in the app's cache directory on Android, where the system deletes them without asking under storage pressure. `catalog_path` derived its base from `XDG_DATA_HOME` or `HOME`, and neither is set on Android, so it fell through to `temp_dir()` — which resolves there to `/data/user/0/<pkg>/cache`. Confirmed on the tablet: the app logs its catalog under `cache/darkroom/...`. What sits beside that catalog is not disposable. `sidecars/` is the commit point for every rating and edit made with no connection — the whole mechanism that makes offline culling safe — and `outbox/` holds exports the user has already been told succeeded. A day of sorting on a train, evicted by an OS housekeeping pass before it ever reached the server, is the worst failure this application can have, and it would leave no error and no trace. The fallback is now `SessionStore::data_dir()`, the persistent per-app directory the Android entry point establishes before any store opens — the same one credentials and sessions already use. Desktop is untouched: the XDG data location is still preferred, so nobody's catalog moves. Two tests. One asserts no durable path contains `/cache/` or `/tmp/`; the other pins the outbox to the catalog's parent, because three call sites derive their location that way and a change here moves all of them at once. Found while verifying that offline browsing works on the tablet with wifi disabled — which it does: the app cold-starts with no network, reports the scan failure without crashing, and serves its grid from the local store. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 12:23:18 +02:00
dtourolleandClaude Opus 5	44f0a4971b	Show the folder picker on the platform that needs it most, and upload at once ... Two faults, both of my own making, reported from the tablet as "I cannot select a location" and "it does not upload". **The picker button was gated on `target-selected == 1`.** That index was Remote's position while both targets were offered. Making the target list platform-aware narrowed Android's to Remote alone, so Remote became index 0 and the button disappeared — on the one platform where the picker is the *only* way to set a destination, since a device folder is not reachable there at all. It is gated on a boolean derived from the target now. An index into a list whose length varies is not a fact about the target, and writing it as one is what made a correct change break the thing it was meant to fix. **A queued export waited for a sync pass.** Staging first is deliberate — an export is finished on disk the moment it is written, and offline is then just a longer queue — but nothing drained the outbox until the next sync, so "Queued for Exports" sat unchanged and read, fairly, as an upload that never happened. A finished batch that wrote anything now drains immediately. The sync-pass drain stays: the first makes an upload feel immediate, the second is what eventually delivers the exports made in a tunnel. Committed without the parallel session's in-flight collection work, which is mid-save and does not compile; verified by stashing it and building this tree alone. 281 dr-ui tests pass, clippy clean. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 12:01:12 +02:00
dtourolleandClaude Opus 5	a8b28136a6	Merge the batch export, and settle the seven-branch merge ... Resolves the last of the parallel work. Two conflicts worth recording, because both were semantic rather than textual: `render_for_export` gained a colour space on master while the batch branch was rewriting the single-image export path around it. Kept both: the batch request supersedes the synchronous path, and the space still has to be chosen at render time because the conversion happens in the shader before the clip to 0..1. `render_open_frame` takes it as an argument rather than reaching for a controller it does not hold. The map-wait moved into `readback::await_mapping` on one branch while another was editing the constant it used, so `READBACK_POLL_LIMIT` survived the merge with no callers. Removed rather than left for clippy to find later. 1164 tests pass, clippy clean, fmt clean. Traceability 53.0% -> 54.3%. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 10:11:44 +02:00
dtourolle	cfff6a3302	Merge branch 'zero-copy-display' ... # Conflicts: # core/dr-gpu/src/adjust.rs # ui/dr-ui/src/develop.rs	2026-08-17 10:04:14 +02:00
dtourolle	9fc8721fa8	Merge branch 'histogram' ... # Conflicts: # ui/dr-ui/src/lib.rs	2026-08-17 10:00:27 +02:00
dtourolleandClaude Opus 5	7d3c8c521f	Export a whole selection, on a thread that is not the interface's ... The export button rendered, resampled and encoded a 24 MP frame on the UI thread and the window was dead for all of it. That was written down as a known compromise, on the grounds that a batch is what makes the wait intolerable rather than merely noticeable. This is the batch, so the compromise comes due. The grid's selection now exports (FR-EXP-7). A worker thread takes a clone of the `GpuContext` — an `Arc` pair over a device and a queue — and opens each photograph for itself: fetch, sidecar, decode, demosaic, render at full size, resample, sharpen, encode, write. Nothing of that touches the interface, which keeps drawing throughout, and the progress goes where every other background job's does: one row in the activity register, with a count and a bar. Why the worker does not borrow the session it could have had. A `DevelopSession` owns the `AdjustPass` the canvas renders from, so handing it to a worker would stop the develop view drawing for the length of the batch — the same freeze, moved. Opening a session per image instead costs a `Demosaicer` and an `AdjustPass` each time round, and the pipeline cache is per-pass so the composed shader is recompiled per image rather than once for the run. Against a full-resolution decode, render and encode that is a few percent, and it keeps this file out of the pipeline `develop` owns. A reusable export pass is the obvious next economy if a profile ever says so. The open image is the exception, and it is why the develop button is not simply a one-image batch. Its edit lives in the interface's session and may not have reached a sidecar yet, so a worker that re-opened the file would export the saved version rather than the one on screen. That frame is therefore rendered by the caller and handed over as `Source::Rendered`; everything after the render — the Lanczos reduction, the encode, the write, which is the larger half of the wait and all of its variance — still leaves the UI thread. So the develop export is no longer synchronous, but it is not fully off-thread either, and the doc comment says so rather than claiming otherwise. Cancellation (NFR-ARCH-3) is an `AtomicBool` read between stages, and the export button becomes the cancel button while a run is live — a batch that could only be stopped by not touching the selection would be a trap. Waits on another worker use `recv_timeout` rather than `recv`, so a cancelled batch sitting on a forty-megabyte download gives up within 100 ms instead of when the transfer finishes. The honest bound is worse than that: a frame already in render has no interior stopping point, so the worst case is one image. Closing that needs the render itself to become interruptible, which is NFR-ARCH-2's scheduler and not a finer poll here. Failures are per image and typed (NFR-ARCH-4). One unreadable body, one folder that cannot be written, one server that went away — each is a message on the channel, a line in the log, and a count in the summary, and the batch carries on. A run with any failure keeps its row until it is cleared, because that is the row somebody came to the list to find; a cancelled run does not, because they asked for it. Two collisions that look alike and are not. `CollisionPolicy` is the user's answer to "a file of this name was already there", and Overwrite is a fine answer to that. It is not an answer to "the frame I exported four seconds ago was also called this" — two folders in a library each holding an IMG_0001 is ordinary — so a name the run has already issued is always stepped past whatever the policy says about the folder. Both halves are held by tests. Supporting changes, each smaller than it sounds. `open_session` comes out of `load_bytes` so the worker shares the JPEG-versus-RAW routing rather than carrying a copy that would drift; the half that builds a `slint::Image` stays behind, where it belongs. `LibraryController::selected_image_paths` answers from the catalog rather than from the loaded window, because selection is by id and survives a scrub — a selection made before scrolling routinely names photographs no row holds. `cache_context_for` takes an id for the same reason, so a batch reads the originals cache instead of re-downloading three hundred files. `format_date` is shared so `{date}` and the timeline agree about what day a photograph was taken. Left undone, deliberately: the batch is sequential, where FR-EXP-7 asks for all available cores. Four full-resolution frames in flight is tens of megabytes each and a straightforward way to exhaust a tablet, and the GPU is shared with the interface in any case. Also undone: exporting with a chosen preset rather than the current export settings — that is FR-EXP-5's machinery, which does not exist yet. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 09:59:57 +02:00
dtourolleandClaude Opus 5	0233df4bf2	See what the highlights are doing: a live histogram (FR-DSP-7) ... Exposure, blacks and whites were set by eye. Nothing said a highlight had blown — the canvas shows white where a channel is at 250 and white where it is at 255, and the difference is the whole question. **Counted on the GPU, not on the readback.** There is a full frame sitting in CPU memory on every canvas update right now — `AdjustPass::read_output`, the bridge spike S1 removes — and walking it would have been thirty lines and no shader. FR-DSP-7 states the mechanism and not just the feature: "these derive from a GPU-side reduction into a small buffer. Per-frame CPU readback of image data is prohibited." A histogram founded on the bridge would be correct today and deleted by S1, and would meanwhile be the reason the bridge could not go. What crosses the bus here is 4104 bytes whatever the image size. The reduction tallies into workgroup memory first and merges once per workgroup. A photograph is not noise: a clear sky puts tens of thousands of adjacent pixels in one bin, and contending for that single global atomic serialises the dispatch. **On the settled frame only.** `render_now` already knows whether a gesture is still moving — `draft` is the flag `redraw` derives from `was_coalesced` — so the dispatch and its transfer happen once when the slider stops rather than on each of the forty frames a drag emits. Nothing is lost: a histogram flickering past under a finger is not a reading anyone takes. FR-DSP-7 requires exactly this, that it not extend the FR-DSP-3 frame budget. Luma is weighted in 8.8 fixed point — 54, 183, 19, summing to 256 exactly — rather than in floats. Not thrift: it makes the shader's arithmetic reproducible bit for bit, which is what lets the test below be an `assert_eq` against a CPU count rather than a tolerance. ARCH §6.13's line about integer state, applied where it happens to also be free. **What the numbers were checked against.** A flat frame must put all 4096 pixels in one bin and one only. A 256-wide ramp must occupy every level with exactly the same count, which is what catches an off-by-one in the quantisation — a `floor` where a rounding was needed shifts the whole photograph one bin left and looks like nothing at all. And a 101x37 frame of seeded pseudo-random pixels — deliberately not a multiple of the 16x16 workgroup, so the edge tiles run off the image — is compared slot for slot against a second, obvious CPU implementation. Exact equality, no tolerance. The CPU version is a deliberate reimplementation rather than shared code: the bugs worth catching here are ones shared code would commit identically on both sides. Above that, the presentation arithmetic is unit-tested headless, because it is where a wrong answer is invisible. A histogram of the wrong shape looks exactly as plausible as one of the right shape. So: 64 columns because it divides 256 and an uneven fold draws an even ramp as a comb; the peak excludes the end columns, or a night scene scaled against its own black spike is a flat line with no information in it; heights are clamped into the plot; and "0%" is kept distinct from "<0.1%" and from "—", since an indicator reading "clipped" over a figure reading "none" is a panel contradicting itself. Clipping counts a *pixel* with any channel at an extreme, not a channel. Any, because a blown red has no gradation left in it however much green and blue still hold — and it is the saturated highlight, the sunset and the red jersey, that clips first and recovers worst. Per pixel, because counting channels can report 200% of a frame clipped, and a percentage above 100 is a readout nobody trusts again. Two affordances for it, which NFR-A11Y-3 asks for: a bar standing at the end of the plot the tones are piling against, and a figure saying how much. Either alone reads. The panel sits directly under the capture metadata and above every control, because it is what the controls are judged against. It is hand-built rather than generated, and ARCH §4.3a is untroubled: a histogram is not an operation — no parameters, changes nothing, answers a question rather than asking one — and nothing in it reads a parameter out of a descriptor. Three plot colours and a neutral luma trace join the palette. That is the swatch's exception rather than a second one: a per-channel histogram has to say which channel, and no achromatic treatment distinguishes red from blue, so the hue is data exactly as the image beside it is. Held well back from full strength for the reason the theme preamble gives. The bounded, non-parking map wait moves out of `AdjustPass` into `readback::await_mapping`, shared with the histogram's transfer. Thirty lines of load-bearing reasoning about frozen interfaces and lost devices, and two copies of it would have drifted. The histogram describes the frame on the canvas, so it is in the output colour space FR-DSP-7 asks for, and when zoomed it describes the visible region — a photographer inspecting a highlight at 4x is asking about that highlight. A device that cannot build the reduction loses the histogram and keeps the photograph. Still to do for FR-DSP-7: the pixel colour readout under the cursor. 324 tests pass, clippy and fmt clean. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 09:55:40 +02:00
dtourolleandClaude Opus 5	cf8f5b632f	Show the develop frame itself, instead of a photocopy of it ... The oldest open item in the project (ARCH §6.1, spike S1, AC-8). Every frame in develop was read off the GPU into a `SharedPixelBuffer` and handed back to Slint to upload again: ~7 ms at 4K against a 0.28 ms compute pass, 96% of the frame spent carrying pixels to the CPU and back so they could be drawn where they already were. Slint 1.17 will adopt a `wgpu::Texture` directly, and the whole of what that needs is arrangement rather than code. **One device, made before the window.** A texture belongs to the device that allocated it, so the compute passes and the compositor cannot each open their own. `GpuContext::new_shared` opens one and hands back the instance and adapter alongside it; `dr_ui::shared_gpu` gives all four to `BackendSelector::require_wgpu_29(WGPUConfiguration::Manual { .. })`. That call has to come before the first window, because creating one selects a backend for you — which is why the GPU is now opened at the top of `run` rather than two hundred lines down beside the other controllers. dr-gpu still names no UI type. It hands out raw wgpu and does not ask who is compositing (ARCH §6.5a). **Vulkan only on the shared path**, where headless keeps its GL fallback. wgpu's GL backend reaches its display through EGL at instance creation, and before a window exists there is no display handle to give it — so a GL instance cannot later produce the window surface Slint needs from it. A machine with no Vulkan gets no shared device and browses without develop, which is the same degradation as no adapter at all. **`renderer-femtovg` becomes `renderer-femtovg-wgpu`.** The old one is FemtoVG over OpenGL and cannot be handed a wgpu texture at all. It is not kept alongside as a fallback: FemtoVG-over-GL has no branch for an imported texture, falls through to "render this image into a buffer", gets nothing, and draws nothing — a blank canvas with no error, which is worse than the failure it would be papering over. The consequence is stated plainly in the manifest: the desktop app now needs a working wgpu adapter to open a window. **Two output textures, not one, and this is the part that is not obvious.** Slint repaints when the image property *changes*, and it decides that with `PartialEq` — which for two images over the same `wgpu::Texture` says "unchanged". A pass that reused a single target would have rendered every slider move correctly on the GPU and shown none of them: right, and invisible. `AdjustPass` alternates between two targets, so consecutive frames are genuinely different values. It also settles the read-while-write question that one queue was already answering. `RENDER_ATTACHMENT` is added to both render targets. Neither pass uses it; Slint rejects an imported texture without it, on the reasoning that a compositor handed a texture may need to draw into it. **`AdjustPass::read_output` is deleted rather than gated.** It and `export_pixels` were the same transfer under two names, and the comments explaining why they were separate are the point of the whole criterion: reading pixels back to *display* them is the defect, reading them back to *encode a file* is the only way a file is made. The display twin is now gone outright, which is stronger than a feature flag — it cannot be turned back on. `export_pixels` is untouched and still ungated. The `readback` feature comes off dr-ui, darkroom-desktop and darkroom-android; it stays in dr-gpu, where it still gates `RenderTarget::read_pixels` and the segmentation field readback. `examples/develop` moves to `export_pixels`, which is honest — it writes a PPM — and so no longer needs the feature. Four tests, each named for what it protects and each of which fails without a screen if the property it guards breaks: - the adjust target satisfies every condition Slint's import checks, asserted in the crate that owns the descriptor, because a descriptor that drifts fails at runtime on a real display and nothing else would notice; - consecutive renders are different textures, and the third is the first again, so the alternation is a rotation and not an allocation per frame; - the develop canvas has no CPU pixel buffer and does have a wgpu texture — AC-8 itself, in the terms Slint uses; - consecutive frames compare unequal as `slint::Image`, which is the property the repaint actually depends on. The zoom test's readback moves into the test module. It has to: there is no library function that copies a displayed frame to the CPU any more, and that is the point — the round-trip now exists in the test binary and nowhere a shipping build can reach. **What is not proven.** No GUI was run. What is verified is that the texture satisfies the import contract, that the import succeeds, that the canvas is a texture rather than a buffer, and that consecutive frames are distinguishable. What is unverified is everything that needs a display: that Slint's FemtoVG wgpu renderer adopts the Manual configuration on a real surface, that the picture appears the right way up and the right colour, and the frame timing that motivated the whole exercise. Android is untouched by testing — the android backend routes a WGPU29 request to Skia, whose wgpu surface does handle imported textures, but that is read from the source, not observed. 56 dr-gpu tests and 255 dr-ui tests pass, clippy clean under `-D warnings`, fmt clean. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 09:54:00 +02:00
dtourolleandClaude Opus 5	4b36ca66aa	Render an export into the colour space its file will claim ... The colour-managed export branch left one call site deliberately unfixed, and this is it. `render_for_export` composed with the default sRGB shader, so a Display P3 export failed with an accurate error rather than producing a mislabelled file — the right way to leave a half-finished path, and no way to leave it. The space is chosen at render time because that is the only time it can be: the conversion happens in the shader, before the clip to 0..1, so by the time pixels reach an encoder they are in exactly one space and the only honest thing left is to label them. `Frame::in_space` carries which, and a mismatch between what was rendered and what was asked for stays a typed error. Also regenerates the traceability matrix over the four merged branches. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 09:34:57 +02:00
dtourolle	8e330a24e9	Merge branch 'undo-redo'	2026-08-17 09:26:23 +02:00
dtourolleandClaude Opus 5	8ea92545df	Stop the export folder forgetting itself, and let Android reach one ... Three faults, compounding into an export that could not be made to work on a tablet at all and a destination that appeared to reset on its own. **Switching target destroyed the destination.** One field held both a filesystem path and a remote folder, so changing the target had to clear it — `/home/x/Exports` carried to the server would have offered to create a folder called `home` at the library root. The consequence was that merely looking at the other option threw away the destination already chosen, which reads, correctly, as a setting that will not stick. There are two fields now. Each target remembers where it was pointed and switching is free; `active_destination` picks between them so no caller can reach for the wrong one. **The library root read as "unset".** The picker opens at the root, so confirming it where it opens stored an empty string — indistinguishable from "ask each time" and looking exactly like the picker had done nothing. Empty now means the library root for a server destination, which is a real folder and the one the photographs are already in; it means "ask" only for a device folder, where no path is worth assuming. The page labels it so. **Android defaulted to a target it cannot use.** A device folder there means the Storage Access Framework, which provides no filesystem path (ARCH §6.9) and is not implemented — so the default target could never succeed however the destination was filled in. The export button said "no export folder is set", the settings page offered no way to choose one, and the only way out was to guess that the other target was the working one. The device target is now absent from `ExportTarget::available()` on Android and the default there is the server, which needs no platform work at all. A settings file carrying an unreachable target — copied from a desktop, say — is corrected on read rather than left to fail at the last step. Seven tests, each named for the fault it prevents returning. The compatibility one matters most: a file written before `remote_destination` existed keeps its device path and gains an empty remote one. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 09:25:15 +02:00
dtourolleandClaude Opus 5	7f524d2fd0	Give a mis-drag a way back ... Develop edits now save themselves to a sidecar the moment you leave the image, so until this there was no way to undo one — the mistake was persisted and the only recourse was to remember the old number. The history is a stack of snapshots, because the edit graph is already plain data: `Preset::capture` reduces it to what differs from default and `Preset::apply` puts it back, so undo is those two calls and nothing else. A command object per action, with an inverse beside it, would have been a second thing every operation had to register — and operations are declared in YAML precisely so that a new one needs no code written for it. A snapshot cannot fall behind them. The interesting part is coalescing. A slider drag emits an event per frame and must be one step, not forty. Nothing in the interface reports a gesture boundary — the same wall the render coalescing hit, and it is answered the same way rather than by threading a "finger is down" out of every slider, curve point and crop handle. What stands in for the boundary is the control plus recency: changes to the same control within 700 ms amend one step. Which control is "the same" is asked of the graph, not listed: an operation whose declared presentation claims a parameter is one where a single gesture moves several — a curve point carries an x and a y — so those coalesce as one widget. Nothing in the history names the tone curve. The compromise, and it is a real one: a control let go of and picked up again within the window is one step rather than two. Buying the other answer costs a gesture-boundary signal on every control, which is more surface than the difference is worth. The stack is bounded at 64 states for NFR-RES-1 — a develop session stays open for hours. Sixty-four rather than a byte cap: what is being bounded is steps a photographer would want back, and a byte cap would give the elaborate edit the shallowest history, which is exactly backwards. The session owns its history and every mutator records into it, so the callbacks in `lib.rs` cannot change the edit and forget to — with a dozen generic callbacks that would have been one press of undo away from wrong every time a control was added. Opening a photograph makes its stored edit the floor rather than a step: it is not work done in this sitting, and an undo reaching behind it would discard a previous session's edit and then save that on the way out. Not yet done, from FR-DEV-5: history is per-session and in memory, and there are no named snapshots. What mattered was that a saved mis-drag had no way back at all. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 09:03:34 +02:00
dtourolleandClaude Opus 5	2330ed25e9	Let a grouped slider be dragged, by flattening the panel that drew it ... White balance, highlights and shadows, and the mixer took a press, jumped once, and went dead under the finger. Exposure and contrast dragged perfectly — which is what made it read as a slider bug rather than a layout one. The panel nested. A group's head row drew the *whole* group, repeating over `row.group-len` and indexing back into `root.rows`; every other row drew nothing. So the inner repeater's model was read off the head row, and depended on that row's identity. Moving any parameter in the group rewrites that row — its own value changed, or `group-modified` flipped for its neighbours — which re-evaluated the repeater, rebuilt its items, and destroyed the `TouchArea` holding the live gesture. A lone parameter had no inner repeater, so the five single-parameter operations were never affected. Now each row draws only itself, so an update touches one control and nothing structural. The facet heading comes off `starts-facet`, which Rust already marks on the first row of a run, and the group heading is drawn by the row that heads it. It also retires the old hazard of a head row building every control in its group — thirty-six live TouchAreas behind the mixer's twelve visible ones. The same identity hazard reached the rows themselves through `ModelRc`, which compares by identity rather than contents: a fresh empty model per row per call made every row differ from itself, so `sync_rows` rewrote all of them on every event. `points` now shares one empty model as `choices` already did. Both are held by tests, because the failure is invisible in a still — every value is right and the panel looks perfect. Curve rows are excluded: their points model carries live coordinates, is rebuilt by design, and `sync_rows` writes values through the existing model rather than swapping it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 07:23:01 +02:00
dtourolleandClaude Opus 5	40d4e439a9	Drop the parameter the offline sidecar writer never used ... `write_one_sidecar` took an `Option<&NextcloudBackend>` it ignored. It was a leftover from the shape the write path had before online and offline separated into two functions: the offline one records to the cache and queues, and has no server to talk to by definition. A parameter that is always `None` and always unused says the opposite — that there is a case where it is `Some` — and the next reader has to check. The closure it was threaded through is renamed to say what it does rather than how it is called. `run(None)` needed the reader to know what `None` meant; `queue_all()` is the sentence. Also regenerates the traceability matrix, which now records FR-CAT-9's queue. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 07:17:31 +02:00
dtourolleandClaude Opus 5	cb1d2be240	Choose the export folder by walking the server, not by typing it ... The destination for a Nextcloud export was a text field. Nobody recalls the exact spelling of a path three levels down, and getting it wrong does not fail — `create_dir` makes whatever was typed, so a misremembered folder becomes a new one at the root and the exports are somewhere nobody looks. So it is picked the way the library root is picked, using the same `FolderBrowser` model the launch screen drives: up, into, and "use this folder", confirming the folder currently *shown* rather than one selected in the list. Same rule in both places, so the phrase means one thing. The model is shared; the worker is not. `settings_ui::spawn_folder_list` is a near-twin of the launch screen's, because that one reaches into the `LaunchController` for its session and reports onto the launch screen's error line, while this one is handed credentials and writes to the settings page. Factoring them together needs a function taking both controllers or a trait implemented twice to abstract two call sites — more machinery than the twenty lines it saves. What matters is shared already: navigation behaves identically because both drive the same model. The callbacks are wired in `lib.rs` rather than in `settings_ui::wire`, because listing a remote folder needs credentials and the settings page holds no session on purpose — it is reachable before a library is opened and must not depend on one existing. With no account the picker says to sign in first, rather than showing an empty list that reads as a server with no folders. Details that are decisions rather than accidents: the picker opens at the library root rather than at whatever half-typed path is in the field, which would list nothing and look broken. The listing area is a fixed 180px, since a folder with sixty children would otherwise push the rest of the settings page off the bottom. "Up" is disabled at the root rather than hidden, so the row does not jump as the user navigates. A failed listing leaves the picker open on the folder it was showing — where the user had got to is not something to discard over a dropped request. And the chosen folder saves immediately like every other setting on a page that has no Save button. The poll timer lives on the controller for the reason `LaunchController` keeps its own there: a `slint::Timer` stops when dropped, so one local to the function that starts it would be collected before the listing arrived. Carries in-flight work from a parallel session — a segmentation pass in dr-gpu, a sidecar cache, and the develop panel's continuing changes. 1020 tests pass, fmt clean. One clippy warning remains and is not mine: `sidecar_cache::dir` is unused while that work is in progress. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-17 07:12:01 +02:00
dtourolleandClaude Opus 5	e00c99b864	Let a photograph leave: an export button, and a cache to leave from ... dr-export could turn a frame into bytes and nothing could ask it to. This is the button, and the place the bytes go. **Everything is staged first.** An export bound for the server is written to a local outbox and uploaded afterwards; offline is not a special case, it is the same path with a drain that finds the server absent. Doing it the other way — upload directly, stage only on failure — makes the failure path the one that is rarely exercised and always broken, and a network drop mid-batch leaves some exports existing and some not with nothing recording which. Staged first, an export is finished the moment it is written and the upload is a promise kept later. The outbox sits beside the catalog rather than under the cache. dr_catalog's cache already draws that line: passive entries are a convenience and go under LRU, pinned ones are a promise and never do. An export awaiting upload is a promise — the user was told it succeeded — and sweeping it for disk would destroy the only copy. Bytes are written before the destination record, so a kill between the two leaves an orphan the drain ignores rather than a record pointing at nothing. The status line says "Queued for Exports/2026", never "Exported to Nextcloud", until it has actually landed. There is a test asserting that wording, because the tempting shorter sentence is a claim the app cannot keep. The drain runs on the sync pass, before the shards: a thumbnail shard can be rebuilt from the originals and the catalog is an index, but a queued export exists nowhere else. `DevelopSession::render_for_export` renders the framed size rather than reusing the frame on screen, which is deliberately viewport-sized (FR-DSP-1) — encoding that would hand the user a soft, screen-sized file with nothing to say anything had been lost (FR-EXP-9). One compromise, recorded rather than hidden: the export runs synchronously on the UI thread, so the window is unresponsive for the few hundred milliseconds a full-resolution render and encode takes. Moving a DevelopSession and its GPU pass to a worker is a larger change than one button earns, and it is batch export that makes the wait intolerable rather than merely noticeable. Still missing: the Nextcloud folder *picker*. The destination is typed into Settings for now. `FolderBrowser` in launch.rs is already the reusable model for it — it browses a remote tree and nothing about it is specific to choosing a library root — but wiring it into the settings page needs a listing worker and browser UI there, which is its own piece of work. Carries in-flight work from a parallel session — presets, the develop copy and paste, and the node schema's `presentation` and `enum` support. One misplaced callback in settings_ui.rs is moved from `render` to `wire`: registered in `render` it borrowed a `&SettingsController` into a 'static closure and would not compile, and that file's own docs say render pushes properties while wire connects callbacks. 992 tests pass, clippy and fmt clean. Traceability 48.3% -> 51.0%. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 23:58:17 +02:00
dtourolleandClaude Opus 5	69b12e327f	Let framing say it wants the canvas, instead of the panel knowing ... The generated panel opened with a special case: if op.id == dr_pipeline::framing::ID { continue; } and a paragraph explaining that framing's eight parameters are eight bad controls — four crop edges you would have to type coordinates into, a "rotate" slider running 0..3, two switches — so `GeometryPanel` presents them as the gestures they are instead. Every word of that is true, and none of it was the frontend's to know. It is a fact about the operation, and ARCH §4.3a is explicit that a frontend deciding things by *naming a stage* is the boundary being crossed: a second frontend would have had to learn the same special case, and nothing in the capability output said why it existed. Framing now declares a `Presentation` preferring `WidgetKind::CropOverlay`, with a demand of two-dimensional dragging and — deliberately — no precise pointing, since FR-UI-7 grows the handles to the modality and a crop is forgiving. The widget owns all eight parameters rather than only the rect: a frontend taking this on takes the whole framing control surface, and leaving rotation and the flips behind would scatter them into the generated panel underneath a crop control that already exists. The panel's rule is now general. `supported` answers whether a kind is implemented *anywhere* — drawn in the panel like the tone curve, or hosted on the canvas like the crop — and `is_on_canvas` settles which afterwards, so the two cannot disagree about the same kind. Any stage preferring an on-canvas widget is skipped, with nothing named. A frontend that implements neither still gets the eight sliders: tedious, complete, and the guarantee the whole hint mechanism rests on. **The tests were asserting against the wrong thing.** `rows_of` was a hand-written simulation of the row generator, complete with its own copy of the framing skip, so the suite was checking a second implementation kept in step by hand. It was not in step: giving framing a presentation changed the real panel and the simulation disagreed, which is exactly how a green suite hides a regression. It now calls `rows_from`, and `rows_of_unfiltered` is gone. `framing_is_not_generated_as_sliders` survives but asserts by routing rather than by counting — no row may carry framing's capability index — so it cannot be satisfied by two miscounts cancelling out. Alongside it, an invented stage preferring a widget this frontend lacks, falling back to one the canvas hosts, must also be skipped: if that ever needs a name added to pass, the special case has grown back. Not included: moving the crop overlay's markup out of app.slint into controls.slint. It is cosmetic next to the above and app.slint is in another session's working set. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 23:29:23 +02:00
dtourolleandClaude Opus 5	0a331c717e	Give the controls a vocabulary, and let a node ask for one ... widgets.slint set the rule — screens consume components, and a bare `Theme.*` at a call site means a component is missing — and it set it for chrome only. The controls never got the same treatment, so they were written wherever they were first needed and copied from there. **The slider was private to the develop panel.** `SliderTrack`, with the fifty-line preamble explaining how it wrests a drag away from a Flickable, lived inside adjust.slint and no other screen could reach it. It shows: export quality is a 1-to-100 value, and the settings page offered a free-text box for it, with the range written in a hint and enforced nowhere. `to-float()` answers 0 for anything it cannot parse, so a typo saved a quality of 0 and the page displayed the 0 back as though it had been asked for. The tick-box was written twice, in launch.slint and settings.slint, from the same 18px box and the same handler; the second carried a comment deferring the lift until a third caller appeared. The label-and-hint header was written three times inside settings.slint alone. controls.slint is the input layer beside widgets.slint's chrome layer, and the constraint that makes it reusable is that **nothing in it knows about `ParamRow`** — that struct is the develop panel's flattening of the capability model, and a control that imported it could only ever be used by the develop panel. The primitives take plain numbers; the ParamRow-shaped wrappers stay in the panel that owns the model. 658 lines came out of the three screens. `SliderRow` is the slider-plus-number-box ARCH §4.3 names as the pointer presentation of a bounded scalar, and quality is its first adopter. It commits on gesture end rather than on every movement, because the settings page saves to disk on change and a two-second drag is a couple of hundred writes where a text field committed once. The develop panel keeps the live stream — that is what its pipeline is for — so `SliderTrack` now reports both. **The other half is the descriptor.** FR-DEV-3a and ARCH §4.3a already specify more than was built: an ordered preference list of widgets rather than one, the demands a widget makes, and kinds beyond scalar and bool. - `Presentation.widgets` is now a list, walked by `choose`, falling back to plain sliders. Falling off the end is not an error, and there is a test asserting an operation asking only for an unimplemented widget still yields one control per parameter. - `WidgetDemand` carries what a widget inherently needs — two-dimensional dragging, precise pointing — and no pixels, breakpoints or platform names. - `WidgetKind` grows to the specified set. There is deliberately no `Colour` *kind*: a colour is three numbers, and a value type that is not an `f32` would reach through the graph, the uniform block and the sidecar format to buy what `ColourWheel` over three scalars already describes. Every widget here is a hint over ordinary scalars, which is what keeps the fallback honest. - `ParamKind::Enum` is the one new shape, and it fits because a variant index is exact in binary32. `kind: enum` with a `variants:` list works in `ops/*.yaml`, so a node declaring one gets a segmented control with no UI file edited — which is the promise ops/mod.rs already makes. The panel's dispatch was duplicated: a lone parameter and a grouped one each wrote out their own list of kinds, so `enum` would have had to be added twice and a kind added to one would appear or vanish depending on how many parameters its operation happened to declare. `ParamControl` is now the only such chain. `rows_from` is free-standing rather than a method, which is what lets the FR-DEV-3c acceptance test requirements.md asks for actually be written: an operation the frontend has never heard of, appearing in a generated panel, with no GPU in sight. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 23:21:35 +02:00
dtourolleandClaude Opus 5	e7130ff891	Give the library header somewhere to put six buttons ... A tablet in portrait is 1920 physical pixels at density 400 — 768 logical, which is below the 820 breakpoint, so the library was already taking the compact layout. The compact header simply was not compact: one flat HorizontalLayout holding a menu button, a title, three status readouts and six buttons. Slint's HorizontalLayout has no wrap and no overflow. Given less width than its children want it shrinks each to its minimum and lets the rest run past the edge, so "Change library" arrived as "Change li…", the readouts elided to nothing useful, and the row overflowed anyway. The six actions move into a `HeaderActions` component drawn in one of two places: inline in the header when expanded, and in a disclosure row under it behind a "More" button when compact. One component rather than two copies, because the alternative is six buttons with their visibility rules and callbacks written twice, and the copy that rots is the one behind a disclosure nobody opens while testing. The row closes itself when the window widens, so rotating to landscape does not leave a stray toolbar behind. "More" is a word rather than an ellipsis glyph: "⋯" sitting in a header of elided labels reads as one more truncation, which is the failure this change exists to remove. `LibraryGrid` takes `expanded` to decide, from the window width rather than the device (FR-UI-1) — so a narrow desktop window gets the same treatment. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 23:04:06 +02:00
dtourolleandClaude Opus 5	9e47133304	Run the formatter over the shard-naming change ... `cargo fmt --all -- --check` is a CI gate and 9a51cc8 landed two files past it, so the build has been red on master since regardless of what came after. Whitespace only — a wrapped `Ok(...)` and two `assert_eq!`s split across lines. No logic is touched and the tests are unchanged. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 22:29:28 +02:00
dtourolleandClaude Opus 5	151dcc3c02	Make a file out of a photograph ... Export existed as a settings page and nothing else: format, quality, colour space, five sizing modes, a filename template and a metadata switch, all configurable in detail, and no way to produce a single file. dr-export is the other half. **It returns bytes and a name, and writes nothing.** An export has three destinations with nothing in common — a path on Linux, a SAF document on Android where there is no path at all (ARCH §6.9), and a PUT to a Nextcloud folder — so a crate that opened the file itself would serve one of them and be rewritten for the other two. The caller places the bytes. Resize, then sharpen, then encode, in that order and for a reason: output sharpening compensates for the softening the resample introduced, so its strength scales with how much scaling actually happened, and sharpening before shrinking would throw the result away. Lanczos-3, separable, with weights computed once per output row — FR-EXP-4 asks for Lanczos or better because a box filter turns a distant fence into moiré. Collision handling takes the "is this name taken" test as a closure rather than looking at a directory, because there is no directory it could look at that works everywhere. That shape is not politeness toward Linux: Android's createDocument renames on collision by itself and cannot overwrite at all, so all three CollisionPolicy settings need the answer *before* anything is created. Overwrite, Skip and Increment are each tested, and Increment gives up after ten thousand rather than spinning against a destination that reports everything as taken. Three things are honest rather than done: - **Colour space.** sRGB only. The shader encodes and clips to sRGB before this crate sees a pixel, so tagging a file Display P3 would claim a gamut it does not contain. Refused with a typed error instead of mislabelled; honouring it is a pipeline change (FR-EXP-2). - **AVIF and JPEG XL.** No encoder. libaom and libjxl are C, ravif is slow enough to change what a batch feels like, and the settings page offers both because FR-EXP-1 lists them — so asking for one says so rather than writing a JPEG under a .avif name. - **16-bit TIFF** is a real 16-bit file carrying eight bits of information, because AdjustPass renders to Rgba8Unorm. Widened by *257, not <<8, so white lands on 65535 rather than a quarter-percent grey. Making it mean what it says needs the composer told what format to write. Metadata is not written at all, which satisfies the half of FR-EXP-8 that matters most: strip_location defaults to on, and a file with no EXIF block has no GPS tag. Retaining camera and copyright when asked is not implemented and cannot be faked by omission. Also here: - `AdjustPass::export_pixels`, ungated where `read_output` is behind a feature. The two are the same transfer and opposites in intent: reading pixels back to *display* them is what ARCH §6.1 forbids and AC-8 asserts against, while reading them back to encode a JPEG is the only way a file has ever been made. Separate methods so the instrumentation can count one without counting the other. - `ExportTarget`, so a destination can be a folder on the server. On Android that is the only destination needing no platform work whatsoever — a PUT against create_dir, already on the RemoteBackend trait, behaving identically on both platforms. Switching target clears the destination, since a path is not a remote folder and carrying one across would offer to create a folder called `home` at the library root. Verified end to end rather than by unit test alone: `cargo run -p dr-export --example export` decodes a frame, runs the develop chain on the GPU at full resolution, reads it back, and writes all five formats — 27 ms for a full-size JPEG, 165 ms with a Lanczos reduction to 1200px. ImageMagick agrees the 16-bit TIFF is 16-bit. dr-export cross-compiles clean for aarch64-linux-android; all three encoders are pure Rust, which is why they were chosen. 944 tests pass, clippy and fmt clean. Not yet wired to a button. The develop view has no export action, so nothing in the running app can reach any of this yet. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 22:26:37 +02:00
dtourolleandClaude Opus 5	b08e94405c	Give the activity row a name Slint will accept ... `ActivityItem inherits VerticalLayout` declared `in property <ActivityRow> row`, and every element that can sit in a GridLayout already carries a `row` for its grid placement — so the declaration was an override of a built-in rather than a new property, and the compiler refused it. dr-ui had not built since. Renamed to `job`, which is also the better word: the property is one background job, and `row` described where it was drawn rather than what it holds. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 21:33:45 +02:00
dtourolleandClaude Opus 5	9a51cc88d6	Give a shard's remote name the client that wrote it ... Shard ids are per store: every client fills its own numbering from 0, so "shard 3" names different thumbnails on every device. The derived sync published them into a flat shard-NNNN.sqlite namespace anyway, which left two clients writing one name. Both failures that follow were live. On upload, a client's open shard overwrote a peer's file of the same id — content the peer still believed was published and would never restore, because its own copy was sealed and the name existed. On download, the loop skipped any remote id it already held locally, which is the only safe reading of a name that says nothing about who wrote it, so a client holding shards 0..5 never fetched the peer's 0..5 at all. Between them, two populated clients exchanged almost nothing: only shards numbered above the other's highest. A fresh device worked, having no local shards to collide with, which is why this went unnoticed — it is exactly the case the feature was written for. The name is now shard-<client>-NNNN.sqlite. The client id is minted per store in index.sqlite, beside the numbering it qualifies rather than in settings: a store deleted and rebuilt restarts at shard 0 and must not claim the remote names its predecessor wrote. Since our own ids now say nothing about what we have taken from others, index.sqlite also keeps a ledger of adopted remote names and the size each had when merged. A size rather than a flag, because a peer's sealed shard never returns but its open one grows, and re-merging the grown copy is how the thumbnails it gained since arrive. Flat names already on servers still parse, reporting no owner, so each client adopts them once, and nothing is written under that form again. One whose id and byte size match a local shard is that client's own earlier upload by the same identity argument the upload path already makes for sealed shards, so the rename does not cost every client a re-download of its whole store. Older builds ignore the new names and stop receiving shards until updated; their own uploads are still adopted, so nothing is lost. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 21:15:12 +02:00
dtourolleandClaude Opus 5	f7e8cc99b1	Call them marks, not glyphs, now that they are drawn ... The last comment left over from the icon change in 65e6a96, which moved the star strip off Text and onto drawn paths. "Glyph" now names something this file no longer has. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 21:14:58 +02:00
dtourolleandClaude Opus 5	65e6a96a65	Say what the application is doing, in one bar and one list ... Every background job reported into a window property of its own — library-thumbs-done, library-pin-total, library-syncing — which only the grid ever read. A pin download that outlived the view it was started from drew nothing at all once the user opened an image, and there was no answer anywhere to "what is this busy with", because the answer was spread across eight properties nothing collected. They report to one register now (ui/dr-ui/src/activity.rs). It publishes an aggregate, which draws a three-pixel bar across the top of the shell in every view, and a row per job, which the settings page lists: scans, thumbnail batches, pin and open downloads, sidecar uploads, the sync and the trash. Failures stay on the list until they are cleared; routine successes do not, or a scroll would bury them. The handle removes a still-running job when it drops, so a worker that dies mid-transfer takes its row with it rather than leaving the bar sweeping for the rest of the session. Also carries in-flight work from a parallel session — the drawn icon set and the dr-pipeline ops split. dr-pipeline's build script does not compile at this commit; ui/dr-ui does, with clippy clean and its tests passing. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 18:32:09 +02:00
dtourolleandClaude Opus 5	70435b712e	Walk the grid with the arrow keys, and open with Return ... A cull is thousands of decisions. The grid already bound 0-5, P, X and U so the judgement itself needs no mouse, and then made the photographer reach for one to move to the next frame — which is most of the gesture, most of the time. Arrows now move a cursor, shift+arrow extends the selection, Home/End and PageUp/PageDown cover ground, and Return opens what the cursor is on. The cursor is a library ordinal, not a row of the loaded window. The window is a few screenfuls around wherever the user is looking, so a cursor held as a row would stop at its edge or keep counting into cells belonging to other photographs; walking out of the window reloads it around the new position, exactly as scrolling does. This is the argument the selection already made by keying on ids rather than rows. The anchor had that fault for real: it was a window row, so a shift-click after a scroll extended from whatever image had since drifted into it. It is an ordinal now, and apply_press takes the window's offset to map between the two. A range longer than the loaded window truncates to what is loaded, since selection is by id and an image the catalog has not been asked for has none. The pointer and the keyboard share select_row so the two cannot drift apart. The one deliberate difference: a plain arrow collapses the selection onto the cursor, where a plain click on an already-selected cell leaves it alone — that exception exists so a multi-image drag can start from one of its members, and there is no drag behind a keystroke. The grid scrolls to the cursor only when the cursor leaves the viewport, and then by as little as will do it. Reusing the scrub's seek() would put the cursor's row at the top on every press, which makes a row unreadable as you walk along it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 16:25:13 +02:00
dtourolleandClaude Opus 5	9b2ee0d0eb	Show the colour mixer as three runs of twelve, each row a colour ... The mixer was thirty-six sliders reading "Hue / Sat / Lum" twelve times over with nothing saying which band any row belonged to. The identity was there all along — the descriptor declares param.mixer.orange.sat and BANDS carries orange at 30° — and was discarded on the way out: labels.rs had no mixer entries, so every key fell through to a derived label that yields the bare channel name. A parameter can now say which aspect it adjusts and which subject it adjusts it on, with the subject's hue where the subject is a colour (descriptor::Facet). That is data about what the operation does, not a layout: the mixer genuinely weights pixels around 30°. What to draw from 30°, and in what order to stack the runs, stay in dr-ui (ARCH §4.3a) — develop.rs brings rows sharing an aspect together and marks the first of each, and adjust.slint names the run once and draws a swatch, a track and a readout on one line. Grouped by channel rather than by band because an edit is almost never "everything about orange"; it is the saturation of the greens, made by comparing one channel across neighbouring bands. Twelve band sections put those twelve rows in twelve different places. The swatch is the label, which is what makes twelve rows fit where four did. The band name is not lost: it is the row's accessible label, so the control is not colour-only, and labels.rs is where the mapping is written down — including chartreuse as "Yellow-Green" and spring as "Blue-Green", since nobody hunting foliage scans a list for "Spring". Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 16:14:13 +02:00
dtourolleandClaude Opus 5	ab4a7e00e7	Leave the adjust panel somewhere to scroll from ... A strip down the right-hand edge of the panel that no control reaches, so there is always somewhere to put a thumb that means "scroll" and nothing else. This is the other half of the slider arbitration. A `SliderTrack` stands the Flickable down the moment a finger touches it, which is what makes dragging an adjustment reliable — and the price is that the track can no longer be dragged past. What was left to scroll from was the ~20px band of label between one control and the next, which on a panel that is mostly tracks means aiming rather than reaching. Reserving the space outright is the honest version of what had been left to chance. Padding rather than a spacer element, and that is what makes it work: the strip is inside the Flickable but no child is laid out into it, so nothing puts a TouchArea over it. A press there reaches the Flickable directly, with no arbitration to lose. A full touch target wide (FR-UI-3). A gutter too narrow to hit confidently would be the problem it was added to fix, in a smaller space. It costs the tracks about 44px of a 280px column, which leaves travel enough that the readout still moves a step per pixel at the precisions the descriptors ask for. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 15:54:57 +02:00
dtourolleandClaude Opus 5	67c0237ddd	Make one slider, and take the lids off the develop column ... Two complaints from a tablet, with one cause between them. **Some sliders dragged and others only answered a tap.** They were not the same control. `ParamSlider` read its geometry from a `ParamRow` for the generated panel and `PlainSlider` took plain numbers for the straighten angle, each with its own track, handle, hit area and gesture rules written out separately — and a comment arguing the duplication was safe, because "a slider that dragged differently depending on which panel it sat in would be a worse inconsistency than the duplication". That is exactly what happened. The touch arbitration fixed in the previous commit went into `ParamSlider` and `CurveEditor`; `PlainSlider` kept the old code, so two sliders in the same sidebar behaved differently and which one you got depended on where you were dragging. The duplication failed to survive its first change. There is now one `SliderTrack`, owning the track, the hit area, the claim test, the hover arbitration, click-to-jump and double-click reset. The two wrappers differ only in where their numbers and labels come from. **Nothing in the develop column collapses any more.** Every group was a `Section` with a disclosure triangle, including five operations that carry a single parameter — so the lid was most of the row, wrapping one slider whose own label repeated the heading word for word. A control behind a lid is one the user does not know the pipeline has. `GroupHeading` keeps what the section was actually for: the name, the dot that says something inside differs from its default, and the reset. The reset is now permanently visible rather than appearing on hover, because a hover-only control is one no finger can reach. IMAGE goes back to flat as well. The column as a whole still closes, from the status strip — which is the control that was wanted, at the level it makes sense at. This costs no vertical space: `Section` defaulted to expanded and nothing ever set it otherwise, so the panel was already unfolded and the lids were overhead with no saving behind them. Verified with cargo test -p dr-ui (192), clippy at -D warnings, and an arm64-v8a release build installed on a tablet. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 15:38:23 +02:00
dtourolleandClaude Opus 5	489465faf0	Show the photograph the way it was taken ... Nothing read EXIF orientation, so every frame from a body held sideways lay on its side — in the grid, in develop, and in the read-only preview. The tag is honoured as part of *reading the file*, at the same standing as a RAW's masked-photosite crop, never as an edit. It lives as a baseline on Framing rather than as a starting value for quarter_turns, which is what keeps four things true: a sideways file opens unmodified, reset returns it to upright rather than to the sensor's scan order, its sidecar stays empty, and the rotate button still moves the image 90° whatever the file underneath it says. Framing::effective composes the baseline with the user's own turns through the group law rather than by adding turns and OR-ing flags. The naive version gets one case wrong — an odd baseline turn plus a user mirror — and gets it wrong quietly, because the result is still a plausible orientation. The composition collapses to a single permutation, so obeying the tag costs nothing per pixel. dr_decode::orientation is a header-only IFD walk, separate from metadata() for the reason the entry points are separate at all: the grid asks once per cell and must not build a rawler decoder to get one tag. CR3 and RAF fall back to the full read, being neither TIFF nor JPEG. Written down as FR-DEV-3h. Known gap: thumbnails cached before this stay sideways. The store is keyed by file and size, and its shards sync — invalidating them would have every client re-download 25 MB a shard, which is not this commit's call to make. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 15:37:05 +02:00
dtourolleandClaude Opus 5	03326242a1	Make the CI checks say what they mean, and format the workspace ... The Android job's "Verify minimum API level" step has never verified the minimum API level. It took the first `*.so` anywhere under the target directory, which is a host proc-macro from debug/deps — an x86-64 object built by the runner's gcc, whose .comment section cannot mention Android and so can never contradict the expected value. It now reads the artifact under the target triple, compares against MIN_API parsed from the Dockerfile rather than a second copy of the number, and fails on a mismatch. Both sides are checked non-empty first: two failed parses would otherwise compare equal and pass, which is the same silent success in a new costume. The Android image installs one SDK package per layer and keeps the output. sdkmanager is a JVM program that aborts when it cannot get memory, and the single `> /dev/null` step reported that as a bare "exit code 134" while a retry re-downloaded everything that had already succeeded. tools/ci-local.sh runs all four jobs — desktop, android, layering, traceability — against the host toolchain, which is pinned to the same 1.92.0 CI installs. Its matrix check compares regeneration against the working tree rather than against HEAD: CI starts from a clean checkout, so git's answer is the right one there and reports every local run stale here. The rest is rustfmt across the workspace, and the clippy findings that surfaced once it did: manual_contains in dr-thumbs and collections_ui, a map iterated as pairs for its keys, an index loop over a slice, and two runtime assertions on a constant now made at compile time. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 12:02:51 +02:00
dtourolleandClaude Opus 5	94a2686dcb	Fit the interface to the system bars, the finger and the back key ... Four faults that only show on a device, and one that was hiding on the desktop too. The system bars. Target SDK 36 forces edge-to-edge, so the window spans the display and the develop status strip was drawn underneath the clock and the wifi icons. Slint already computes the inset from Android's OnApplyWindowInsetsListener and exposes it as Window.safe-area-insets; nothing read it. The four views now sit inside a shell placed within the safe area. Every inset is zero on the desktop, so that layout does not move. Sliders under a finger. A Flickable steals any gesture that drifts more than 8 logical pixels along its scrolling axis within half a second of the press, and it steals it by cancelling the child. ParamSlider's axis test correctly declined to claim vertical drags, but nothing told the Flickable to stand down once a drag was claimed — so an adjustment would start moving and then be taken away mid-motion. A mouse holds a horizontal line closely enough to stay under 8px; a finger does not, which is why these worked on the desktop and not on the tablet. The claim now sets `interactive: false` for the rest of the gesture. The tone curve had the same fault and worse: its points are dragged vertically, which is the Flickable's own axis, so every drag was stolen — on the desktop as well. The back gesture. Nothing handled it, so back closed the application from anywhere in it. Android delivers it as Key.Back to the focused item and bubbles it up the ancestors, which is the second reason the shell wraps the views rather than sitting beside them. The order is innermost first: settings, then crop, then zoom, then develop to the grid, then a collection scope. Answering false at the top of the stack leaves Android to close the activity, as it does for every other application there. Escape does the same on a keyboard. back_step is a pure function over a flat NavState so the ordering can be tested without a backend: which of two states is left first is the whole of the feature, and it is the part that is easy to get subtly wrong when spelled out in nested ifs over live properties. Develop's canvas now takes focus on show. Without it the arrow keys did nothing until the canvas was clicked, and Key.Back had no focus item to bubble from. Panels that close. The collections sidebar and the develop column are collapsible from the grid header and the status strip. The layout class now supplies only the default: a panel closed to see more of a photograph stays closed while the window keeps its shape, and the choice is dropped when the class changes, because rotating a tablet asks a different question from the one answered in landscape. IMAGE became a Section, being the only group in the column that could not be put away and the one whose content is read first and needed least. Pinch to zoom on the develop canvas, anchored on the midpoint between the fingers (FR-UI-4). The wheel is the desktop's answer and there is no wheel on a tablet. The develop status strip was 28px against the 44px headers on the library and settings pages either side of it — the one screen where a way out has to be found was the one drawn smallest. All three now agree. Verified with cargo test -p dr-ui (192 passing, 6 new), clippy at -D warnings, and an arm64-v8a release build packaged to an APK. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-16 11:59:31 +02:00
dtourolle	4d78041d1d	Many imorovments	2026-08-12 22:16:15 +02:00
dtourolleandClaude Opus 5	fa12afed18	Keep originals on this device, by pin and by use ... Fills in `image_cache`, which the previous commit's "On this device" filter read but nothing wrote. Also carries in-flight work that shared these files: the Android TLS root store, the settings page, and a regenerated traceability report. # Two populations, deliberately separate An original is kept here for one of two reasons, and conflating them produces the exact failure the feature exists to prevent. **Pinned** originals were asked for. Pinning a collection before a trip is a promise, so pinned rows are never evicted and never counted against the budget — a cap that could silently delete a pinned trip would make pinning worthless, because it could not be relied on without checking. **Passively cached** originals are a side effect of working: develop already downloads the whole file, so keeping it costs no bandwidth and saves the entire transfer next time. This population is what the budget bounds, evicted least-recently-used, because it otherwise grows until a day of culling fills a disk. Sharing one budget would let a large pin starve the passive cache, or let browsing evict a pin. They are separate. # What was built `dr_catalog::cache` owns the bookkeeping — held tier, size, last use, pinned — and writes the bytes; deciding to download stays with the caller, which is what keeps a crate with no network out of the network's business. Files are written to a temporary and renamed, so a dropped connection cannot leave a truncated file recorded as a complete original. They are named by image id, not filename: `Photos/IMG_0001.CR2` and `Trips/IMG_0001.CR2` are different photographs, and a flat cache keyed on the name would serve one for the other. `spawn_full_fetch` became read-through. A hit is a disk read; a miss stores what it downloads and enforces the budget. A cache that cannot be opened is a miss, not a failure to open the photograph. Pinning writes intent — `tier_desired` — without downloading, so the button responds immediately, and `spawn_pin_fetch` fills it in sequentially afterwards. Sequential because these are tens of megabytes each: the lanes that make the thumbnail sweep fast buy little against one connection's bandwidth and cost a great deal of memory. A pin interrupted by a lost connection resumes from where it stopped. Schema v5 adds `pinned` and `path`. `pinned` is a column rather than something inferred from `pinned_by_rule`, which is ON DELETE SET NULL and so cannot answer for an image whose rule was deleted. A v4 catalog migrates in place; existing rows default to unpinned, the safe direction. The budget and "keep opened originals" come from the settings page rather than a constant, and are applied at startup rather than only on change — a cache capped at 2 GB last session would otherwise spend this one filling to the default. Turning off keeping leaves what is already cached readable: those bytes are paid for, and refusing them would re-download images sitting right there, including pinned ones. Also removes a doubled `#[test]` introduced in the previous commit. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-11 21:12:01 +02:00
dtourolleandClaude Opus 5	cd75e5a4c6	Run the library from local data when the server is unreachable ... Also carries in-flight work that shared these files: the zoom structure-key fix in the adjust pipeline, nearest-neighbour filtering past 1:1, the timeline scrub marker correction, the 423-Locked retry in the metadata sweep, and the thumbnail size-class migration. # Offline mode (FR-CAT-9) The app previously assumed the server was reachable and treated its absence as a series of unrelated per-operation failures. A launch without a connection produced an empty grid, even with a complete catalog on disk and every thumbnail already in the shards. Reachability is now inferred from traffic the app was already making, rather than probed for. `RemoteError::indicates_offline` draws the line that makes this possible: a dead connection is offline, a 403 or a 500 is not — the server answered, so blanking the library over one forbidden file would be a worse error than the one being reported. `Reachability` turns those outcomes into a state, so a library browsing happily never issues a probe at all. Going offline takes one failure, because the user is already experiencing it. Coming back requires evidence — a completed scan or a fetched thumbnail — with a capped exponential backoff behind the manual retry, so twelve sweep lanes failing together do not schedule twelve immediate probes. What keeps working: the catalog opens even when the scan that normally provides it failed, so the grid fills from the last successful scan. Thumbnails come from the shards. Rating, flagging and collecting are catalog writes that never touched the network. What stops is opening an original that was never stored locally, and it now says so in those words instead of reporting "network error: connection refused" over a photograph. Work that is pure network is refused rather than left to fail slowly: the metadata sweep, derived sync, and sidecar writes. The sweep would otherwise spend a timeout per image across the whole library while the progress bar implied something was happening. Deferring sidecars is a real gap rather than a hidden one — a rating made offline reaches its sidecar only when that image is judged again while connected — and it is recorded as such at the call site. # The "On this device" filter A chip beside the rating filters, narrowing the grid to images whose original is held locally. It composes with the rating terms rather than replacing them, so "five-star frames I can actually edit on this train" is one filter. The predicate is SQL, like the rating terms and for the same reason: the count in the header has to agree with the cells drawn. It reads `image_cache.tier_actual`, which nothing writes yet — the next commit fills it. Until then the chip honestly reports zero. `Tier` gains an explicit on-disk encoding. The variants are ordered by generosity and the derived `Ord` invites reordering them, which would silently reinterpret every cached row; the round-trip test is what holds the two in agreement. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-11 20:36:50 +02:00
dtourolleandClaude Opus 5	f6a100863e	Place the timeline marker where the pointer actually is ... The marker was drawn from a bucket index while a click reported a fraction of the track. Those are different quantities: bars each occupy one equal slot whatever span of time they cover, and the index snapped to the containing bucket's start edge, so the marker landed at the top of whichever slot held the instant — close enough to pass on a dense uniform axis, plainly wrong on a sparse one, and never under the click. Send the fraction instead, computed as the exact inverse of the interpolation the scrub handler applies, and position the marker from it. Marker and click are now the same quantity by construction. Scrolling the grid also left the marker behind: only an explicit scrub ever wrote the position, so the axis claimed to say "when you are" and stopped being true the moment the wheel moved. Map the first visible row back to a capture time and move the marker with it. That runs on every scroll event rather than behind the window-reload guard, which fires a few times per screenful and would make the marker advance in jerks. Only the marker moves, not the bars: rebuilding those means a GROUP BY aggregate over the library, far too much for a flick, and they do not change as the grid scrolls anyway. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-11 20:08:32 +02:00
dtourolleandClaude Opus 5	8b7c1e7f10	Open the sign-in URL through an ACTION_VIEW Intent on Android ... The previous commit made the missing launcher honest; this gives Android a real one, so Login Flow v2 can complete on device. Builds `new Intent(ACTION_VIEW, Uri.parse(url))` and hands it to `startActivity` over JNI. The JavaVM and Activity come from ndk_context, which android-activity's glue populates at startup — the same handle Slint's backend uses, so there is no second VM to reconcile. The login worker is a plain std::thread and therefore unknown to the JVM, where any JNI call would abort the process. jni 0.22 scopes attachment to a closure rather than returning a guard, so the whole Intent is built and dispatched inside `attach_current_thread` and the thread detaches on the way out. Names use `jni_str!` and signatures `jni_sig!`, both compile-time: a typo is a build error rather than a NoSuchMethodError on the device. A pending Java exception is checked and cleared before returning, since leaving one pending makes the next JNI call fail somewhere unrelated; in practice it means ActivityNotFoundException, i.e. no browser installed. jni is pinned to 0.22 to match Slint's Android backend. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-09 22:35:38 +02:00
dtourolleandClaude Opus 5	0876977133	Report a missing browser launcher instead of faking success ... `open_in_browser` gated its xdg-open path on `target_os = "linux"`, which is false on Android — that is its own target_os. Android therefore took the fallback arm, which discarded the URL and returned `Ok(())`. Login Flow v2 cannot complete without a browser: the user approves the sign-in there and `auth::poll` waits for that approval. Claiming success meant the UI showed "Approve the sign-in in your browser" with no browser open, the poll waited for an approval that could never arrive, and the worker eventually dropped its channel — surfacing as "sign-in failed unexpectedly", which pointed at the network rather than at the real cause. The server had in fact been contacted successfully. Gate on `unix && !android && !macos` so the arm matches what xdg-open actually implies, and return Unsupported from the fallback. The caller treats it as terminal rather than swallowing it with `let _ =`. Android gets a real Intent-based launcher in the next commit; until then the failure is at least honest about what happened. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-09 22:22:00 +02:00
dtourolleandClaude Opus 5	d49b4b41de	Add thumbnail size classes and grid zoom; fix the scrub ordinal ... The grid now zooms, which needs thumbnails at two resolutions rather than one, and exposed a scrub that landed in the wrong place. **Two thumbnail size classes.** `ThumbSize::Grid` (256px, ~10 KB) and `Large` (1024px, ~45 KB), with the class part of the store key so both coexist. Storing everything large would take the reference library from ~200 MB to ~860 MB, and shards sync, so that is transfer cost on every device rather than only disk. A store written before the class existed migrates in place: its entries are all grid-sized, which is what the column defaults to, so nothing already fetched is discarded. `forget` now drops every size for an image. Reading a single row left the other class's bytes on the shard's tally for good, sealing it early on space nothing occupied. **Grid zoom.** Ctrl+wheel and pinch resize cells between 90px and 420px in geometric steps, so the gesture feels the same at either end where a fixed pixel step would be imperceptible at 400px and violent at 90px. Crossing 256px switches to the large class, so a zoomed cell is sharp rather than upscaled. Columns and window capacity already derived from cell size, so the grid reflows for free. **The scrub landed about half a library too high.** It counted only dated images while the grid shows all of them — 10,733 dated against 19,841 rows — and ignored `shadowed_by`. Verified against the live catalog: the old formula gave 10,887, the new one 10,732, the true grid position 10,732. The scrub's count and the grid's window must use identical predicates and ordering; a test now fails if they diverge. **Timeline gestures are continuous.** Scrub and pan were quantised to whole buckets, so a slow drag did nothing until it crossed a boundary and then jumped a month. Both work in fractions of the visible span now, and pinch-to-zoom arrives for tablet, where there is no wheel to reach the axis with. The pinch accumulator was wrong on first writing: it took at most one step per update, so an 8x spread — three doublings — yielded one zoom level. `log2().trunc()` now extracts every whole doubling and carries the remainder. The original test asserted the wrong number and defended it in a comment, which is worth remembering: a test can entrench a bug as readily as catch one. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>	2026-08-09 21:58:32 +02:00