Surface the S5a media_kind discriminator in the library UI. A generic
uploaded file (media_kind='file') is download+share only — no player and
no timeline to trim, so Play/Edit/Edit-details are media-only; it shows a
neutral file glyph instead of an (always-empty) thumbnail and no duration
chip (container/size still read). The "Shared with me" tab's file row is
gated the same way. Re-download is now hidden for edit clips and uploads
too (neither has a re-fetchable source — the backend already rejects both
with 409), removing a latent click-to-409 on edit rows.
FE-only: the serializer already emits media_kind (S5a). Adds media_kind to
the DownloadAsset type and "upload" to the job_kind union; DOM tests cover
the file-row presentation (glyph, no img, no duration).
Enlarge every non-title text on the card (channel, status/format/size/date meta,
source-ref line, progress readout) one step (xs→sm) plus their inline icons, and
grow the row action buttons (padding + icon) via the actions wrapper so the shared
IconBtn stays smaller in the modals. Buttons keep the title's line — the wrapper is
shrink-0/no-wrap, so the text column shrinks instead.
The completed download rows already surfaced size + duration; add the concrete
container (MP4/M4A/WebM…) as a compact chip so the library reads as a media list
with a visible format — the audio-vs-video distinction the running-only phase pill
can't show at rest. Pure display of the already-serialized asset.container; locked
by a small DOM test (present when ready, absent before the worker fills it).
The download/edit child (yt-dlp + ffmpeg) shares one LAN box with the API,
Postgres, and Plex playback. A CPU re-encode pegged every core for ~30s and
starved web requests + playback until it finished. Full-scope S4 mitigations:
- Child self-de-prioritizes at startup (inherited by every ffmpeg/yt-dlp
descendant): CPU nice increment + best-effort idle ionice class + an
OMP_NUM_THREADS budget, all config-gated and independently disable-able.
- Cap the browser-safe re-encode's encoder threads (`-threads`) so a transcode
reserves a core for foreground work instead of claiming all of them.
- Throttle the retention GC's whole-tree orphan sweep (yield IO after each
batch of files) so its rglob can't monopolize disk IO in the API process.
- Idle-back-off the Download Center (2s→15s when nothing queued/running) and
UsageBar (2s→15s when running_jobs is 0) polls; both snap back to 2s on the
next fetch after an enqueue (invalidateDownloads refetches immediately).
Tests: encode_thread_cap (pin/reserve-core/edge), -threads flag presence,
_deprioritize (nice/ionice/OMP branches), GC sweep throttle batching.
Pure code-motion, DownloadCenter 1088->359 L. Group by kind, mirroring the S4
plexCards/plexViews split:
- components/downloadRow.tsx (309 L): the user-list presentation + shared row
primitives — StatusPill, Thumb, IconBtn, SourceRef, ChannelLine, DownloadRow,
UsageBar (+ jobTitle/jobChannel/displayUrl helpers, status/phase constants).
- components/downloadModals.tsx (178 L): the per-item user modals — MetaEditModal
(rename/metadata) and ErrorModal (failure detail + retry).
- components/downloadAdmin.tsx (267 L): the admin "System" tab — AdminSystem plus
QuotaModal and QuotaUserPicker.
Every extracted piece is prop-driven with no shell-state reach-back, so the shell
just imports them back; the tab logic, mutations and modal state stay in
DownloadCenter. Behaviour-preserving.
Gate green (tsc first try, knip, lint 0-err, 247 vitest). Self-E2E (localdev):
Queue/Library/Shared/System tabs render, UsageBar, admin storage/footprint/job
list, MetaEditModal open/close, 0 console errors.
peter