Introduce Discord presence runtime support and continue composition-root decomposition by moving Jellyfin wiring into dedicated composer modules. This keeps main runtime orchestration thinner while preserving behavior and test coverage across config, runtime, and docs updates.
Split config defaults, option metadata, runtime-option registry, and template sections into domain modules while preserving the composed public API in definitions.ts. This keeps ConfigService behavior stable and makes future config extensions easier to add with focused regression coverage.
- Add M PLUS 1 variable weight font (100-900) to src/renderer/fonts/
- Add @font-face declaration in style.css
- Update default fontFamily for primary and secondary subtitles
- Add macOS (Hiragino Sans) and Windows (Yu Gothic) fallbacks
- Update build script to copy fonts dir to dist/renderer/
Remove configurable isLapis sentence/audio field overrides so sentence cards always map to Sentence and SentenceAudio. Update types and docs to reflect the simplified config surface.
Detach --background launches from terminals with quieter runtime output, make wrapper/plugin overlay start explicit, and allow trailing commas in JSONC configs for safer hot-reload edits. Includes pending Anki/docs/backlog updates in this unreleased batch.
- Merge feature branch changes for n+1 target-only highlight flow
- Extend merged token model and token-merger to mark exactly-one unknown targets
- Thread n+1 candidate metadata through tokenizer and config systems
- Update subtitle renderer/state to route configured colors and new token class
- Resolve merge conflicts in core service tests, including subtitle and subsync behavior
