Face indexing was compiled into the APK all along — dr-ui takes dr-face with
`inference` on every target, so SCRFD, alignment, MBF, calibration and
clustering were all in there. What was missing was the weights, and on Android
there was no way to supply them.
Route C (docs/faces.md §2.2) says the user obtains the model and the app loads
it. On a desktop that is a real gesture: drop two files in
~/.local/share/darkroom/models/ and indexing starts working. On Android it is
not a gesture at all. `internal_data_path` is app-private, `run-as` needs a
debuggable build, and the in-app fetch route C specifies was never built — so
the settings page reported "no face model is installed" on every launch with
nothing behind the message. Not "off until you supply weights"; off.
So the shape-fixed pair goes into LFS under the APK's assets, assemble-apk.sh
copies it into the package, and `android_main` unpacks it to the shared models
directory before anything asks whether a model is present.
Three things that are not incidental:
The models directory is now shared across accounts rather than per-account.
Weights are identified by `faces.model_id`, not by who is signed in, so two
accounts had no reason to hold two copies — and the unpack runs before any
session exists to key a per-account path off. `face_models` still prefers a
per-account directory when one is populated, so anyone mid-migration keeps the
ability to pin one library to its own pair.
The unpack writes under a temporary name and renames. `face_models` decides
availability on `is_file()` alone, so a copy truncated by the process being
killed would leave a file that passes that test and fails inside tract —
reported to the user as a broken model rather than a missing one.
assemble-apk.sh refuses an LFS pointer. At ~130 bytes it looks exactly like a
model to `cp`, and unchecked it reaches the device and fails in the graph
loader instead of telling someone to run `git lfs pull` — the same guard
dr-segment's build script applies to yolo26n-seg.onnx.
The licensing half is unchanged and recorded in §2.2a: the InsightFace grant is
research-only, this is a private repository and a self-installed build, and
these files come back out before anything is published. The weights are still
not a cargo build input — dr-face has no `models/` directory and no
`embedded-model` feature, and nothing in the build reads them. The APK assembly
step copies two files and is the only thing in the tree that knows they exist.
Verified on device: both models unpack on first launch (2524817 and 13616095
bytes) and the APK carries them at assets/models/.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The Android job ran in gitea.tourolle.paris/dtourolle/darkroom-android:latest,
a tag that had never been pushed. The image existed only as a local
darkroom-android:latest on one machine, so every Android job died at docker
pull with "manifest unknown" before reaching a step. The registry API confirms
it: that manifest is a 404 while the other builder images answer 200.
android-image.yml now builds and pushes it, following KPN's docker.yaml — host
runner rather than a container, so it has the Docker daemon and the host's
cached registry credentials, and a plain-git checkout because that host has no
Node for actions/checkout.
Where it diverges from KPN: that workflow gates on dorny/paths-filter running
inside the builder image, which here would need the very image that is missing.
The tag is the git tree hash of docker/android instead, which changes when and
only when a file there changes. An unrelated push reuses the image, a Dockerfile
edit cannot keep serving a stale latest, and a missing tag rebuilds itself
without a manual step.
The presence probe is curl against the registry API, not `docker manifest
inspect`. The latter exits 1 on this registry even for tags that are plainly
there — jellytau-builder:latest answers HTTP 200 while docker reports "manifest
unknown" for it — and trusting it would have rebuilt 7 GB on every push. The
HEAD request also yields Docker-Content-Digest, so latest is repointed only when
the digests actually disagree, without pulling any layers.
A probe that cannot authenticate falls through to building. Rebuilding when it
was unnecessary costs minutes; skipping a build that was needed is the failure
this commit exists to remove.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Establishes the v0.1 foundations on both platforms:
- dr-types: SourceRef (never a filesystem path — Android SAF has none),
Format, Availability, Validator with ETag quote normalisation
- dr-gpu: wgpu device, compute pass writing a storage texture, resize
- dr-ui: Slint shell with FR-UI-1 adaptive layout, computed in Rust to
avoid a binding loop
- docker/android: pinned toolchain, verified producing API 28 ARM binaries
Measured the cost of the temporary CPU readback path (dr-gpu bench):
compute is 0.06-0.28ms across sizes while readback is 0.63-7.43ms, so
readback is 90-96% of frame time and scales with area. Recorded in
ARCH §6.1 — this is why spike S1 is the priority.
Mitigations pending S1: reuse the staging buffer, apply at most one
resize per frame, and cap render resolution at 2048 on the long edge.
10 tests passing; core crates cross-compile for aarch64-linux-android.
dtourolleandClaude Opus 5