v0.3.63
Fixed
- Add-to-queue toast no longer says "all already queued" when new items were actually added (
cursor.lastrowidisNoneafterexecutemanyper Python sqlite3 — readMAX(id)after the insert instead).
Performance
- First-time poster resolution for a batch of new items now runs 8 paths in parallel instead of one-at-a-time, cutting initial render of ~30 new items from 30+ s to a few seconds.
Docker images
Pinned to this release:
docker pull ghcr.io/i-ial9000/shrinkerr:0.3.63 # portable (amd64 + arm64, CPU)
docker pull ghcr.io/i-ial9000/shrinkerr:0.3.63-nvenc # x86_64 + NVIDIA GPU (NVENC)
docker pull ghcr.io/i-ial9000/shrinkerr:0.3.63-edge # as above but ffmpeg master
docker pull ghcr.io/i-ial9000/shrinkerr:0.3.63-edge-nvencFloating tags (:latest, :edge, :nvenc, :edge-nvenc) track the most recent release of each lineage.