Forced and SDH get no sidecar name of their own: one language, one
sidecar, and a forced track is ignored where a plain one exists. That
makes the database constraint fall out rather than needing a scheme.
The OpenAI-compatible backend's base URL and model become database rows.
It is not "OpenAI" — it is any endpoint speaking that shape, llama.cpp
included — so which one is in use is something to try and change, not a
property of the deployment. Only the API key stays in the environment.
Records the subtitle decisions settled during Subtitles milestone
planning: global wanted set (pt + en), embedded-track extraction,
providers, ranking, translation backends, sidecar naming, alass sync,
budgets, and stated non-goals. Points §13 item 9 at §15 and adds
arr-subs to the crate lists.
Closes-Issue: #183
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A release covering several episodes is measured by size divided by the
episode count, for the target penalty and the floor alike. An unknown
count divides by one, which fails toward rejecting rather than grabbing.
Refs #179
Untracking a season now clears its episodes' wanted flag, unified search
returns titles only, the library is the homepage and the signal chain moves
into settings, and the wordmark links home.
Refs #169
`resolution_pref` only gated eligibility, so every release was scored
against its own resolution's size band and a 23 GB 4K lost to an
at-target 1080p. Each step up the list is now worth `resolution_step`
points, seeded at 300: five gibibytes of 4K overshoot, so a 4K up to
27 GB wins and a bloated one still does not.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Design for a single-service replacement of Radarr/Sonarr: domain model,
policy engine, sourcing via Prowlarr Torznab, import pipeline, reconcile
loop, crate layout, CI gate and build order.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Miguel Palhas