Spike: station-to-station show syndication — rank the three shapes, recommend the sequence

[genwave-radio]

📻 Origin

THOUGHTS.md, Freaky Friday 08-14 honorable mention + Dean's follow-up questions ('so many questions :)'). Triage 2026-08-22: Dean ruled the spike itself ranks the shapes rather than pre-committing.

🎯 The spike's deliverable

A scored comparison of the three possible shapes of 'a show airs on more than one station', plus a recommended build sequence:

1. Definition-only pack — a show ships as structure: format clock, persona, segment skeleton, script templates. Each station re-renders in its own voices, music from its own library. No audio crosses stations → the copyright question dissolves; reuses the existing catalog-pack muscle (personas/themes/avatar/icon/font kinds all precedent).
2. Definition + pre-rendered patter audio — the host's actual voice travels; music stays local. Show 'sounds like' its creator everywhere. New questions: voice/likeness consent, pack size, staleness (dated references baked into audio).
3. Live relay with a break-cue protocol — origin station streams, receivers carry it live, an in-band/side-channel signal ('break in X seconds') lets locals drop in their own IDs — the real radio tradition, and the full latency/protocol/copyright problem.

📏 Score each shape on

• Effort and what already exists (catalog kinds, format clocks/shows via gh-feat: dayparting — NAMED shows over the existing schedule ('The Breakfast Show with DJ X') #383, input.harbor precedent from gh-Grab the Mic: live human handoff via Liquidsoap harbor input #533)
• Copyright exposure and whose responsibility it is (the gh-spike: public-station registry — is a broadcast-rights waiver enough of a gate? #207 waiver spike is precedent for the click-through model; note whether music-carrying relay is platform risk or streamer risk)
• What the community shelf can actually carry today vs what needs new infra (registry build-out gh-registry: public station registry — the registry itself + Admin UI publish flow (templated PR) #244)
• The local-station-ID gap problem for live relay (cue signal design sketch, not implementation)

🏠 Also answer in the spike

• Where it lives: Dean's instinct is the community catalog ('it's where you already go for everything else') — confirm or argue.
• Priority: filed P3; the spike says whether the walking-skeleton shape (likely definition-first) is cheap enough to justify a bump.

🚫 Non-goals

No code. The spike output is a written recommendation ready for /design.