v1.2 research briefs + proposal

[LeoHChen]

Summary

Overnight research output for v1.2 planning. Read the PROPOSAL.md first — it synthesizes the three briefs into a concrete v1.2 commitment recommendation with 5 open decisions for you.

Briefs

• WINAMP_INSPIRATION.md (Research: Winamp 2.x/5 features + design language to inspire v1.2 #117) — 10 ranked follow-up proposals. Top 5 fold into already-filed v1.2 issues; bottom 5 are v1.3 candidates.
• PERF_COLD_LAUNCH.md (Profile cold launch + first-screen render with Instruments (research, feeds A) #106) — top 3 wins target ~640ms → ~400ms time-to-populated-paint on a 5k-track drive. Sim-side numbers; user should re-validate on real hardware before locking in.
• TAG_WRITING_LIBRARY.md (Pick + document a tag-writing library for ID3 edits (research, feeds G) #113) — recommends `ID3TagEditor` (MIT, Swift 6, pure Swift, mp3-only). m4a + flac deferred to v1.3.

v1.2 commitment recommendation (from PROPOSAL.md §2)

• In: Performance: cold-launch + first-screen render speedup #107, Performance: scrolling smoothness in Albums and Artists grids #108, Design consistency sweep: migrate stragglers to WinampTheme primitives #110, Charcoal Phosphor polish: tighten edge cases the v1.1 audit missed #111, SwiftUI player screen ("None") parity with skinned player #112, ID3 edit sheet (Tier 1) + AI-suggested fixes (Tier 2) #114-tier1, v1.2 public App Store launch: screenshots, promo copy, listing review #118 (7 issues)
• Defer to v1.3: Performance: parallelize indexer metadata extraction #109 (indexer parallelization), Bulk ID3 cleanup tooling (stretch — may slip to v1.3) #115 (bulk ID3 cleanup)
• Tier 2 AI ID3 (within ID3 edit sheet (Tier 1) + AI-suggested fixes (Tier 2) #114): ship in v1.2 if on track; slip to v1.3 if not

Open decisions for you

1. Approve commit-to-v1.2 list?
2. File 4 new v1.3 candidates from Winamp follow-ups?
3. Tier 2 AI ID3 — v1.2 stretch, v1.3 punt acceptable?
4. Real-device perf re-measurement — you do it, or release agent captures Instruments trace?
5. Llama easter egg — yes / no / designer's call?

Test plan

• You read PROPOSAL.md
• You answer the 5 open decisions
• On approval, this PR merges as-is (research artifact); implementation work happens in separate per-issue PRs

🤖 Generated with Claude Code