Two changes that drop the friction in self-hosting RefBoard so the install
story becomes "docker compose up, open the URL".
1. JWT_SECRET is now optional. On first boot the backend generates a
64-byte random secret and persists it in the existing settings table.
process.env.JWT_SECRET still wins when set, so ops setups that manage
secrets out-of-band are unaffected. The prod-throws-without-env guard
is gone (auto-generation is a strictly safer default than the previous
hardcoded dev fallback).
2. STORAGE_BACKEND=fs|minio picks between MinIO (default, unchanged) and
a new local-filesystem adapter. The FS adapter exposes a fake minioClient
that mirrors the methods RefBoard calls (statObject, getObject,
getPartialObject, listObjectsV2, putObject, removeObject(s), bucketExists,
makeBucket), so consumers swap require('./minio') for require('./storage')
and nothing else changes. Sidecar .mime files hold Content-Type so the
range-aware media proxy still serves the right response headers.
examples/compose/minimal-fs.yml is the single-container variant that uses
the FS adapter. The default docker-compose.yml still spins up MinIO.
README quick-start collapses to one block (cp .env, docker compose pull,
docker compose up -d). The first user you register on the Login screen is
auto-promoted to admin, same as before.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Move PDF page count check before MinIO upload and DB record creation
to prevent orphaned objects when >500 page PDFs are rejected
- Use job type label (PDF page / Video) in media-worker error messages
instead of hardcoded "Video processing failed"
- Replace LIKE-based idempotency check with exact JSON match to prevent
page 1 matching page 11/12/etc substring collisions
- Background media worker: polls media_jobs table, runs ffprobe+ffmpeg
with concurrency limit, generates video posters, emits socket events
- Non-blocking video upload: stores file + enqueues job, returns immediately
- Poster hydration on board load: GET /boards/:id injects poster/dimensions
from DB into canvas_state video objects
- SpatialGrid: fixed-cell (512px) spatial hash for O(nearby) culling instead
of O(all) item scanning, eliminates setTimeout violations
- Canvas-based video rendering: draws video frames to offscreen canvas then
uploads to GPU, completely eliminates GL_INVALID_OPERATION errors from
PixiJS VideoSource auto-update mechanism
- Server poster upgrade path: culling ticker and applyProcessedMedia() both
upgrade client-captured posters to server posters when available
- Pause restores server poster (paused video behaves like an image)
- Selection drag-end persistence: onObjectDragEnd broadcasts + saves + undo
- Live media:job:update socket handler patches scene data + VideoSprite
dimensions without broadcast fanout