[vdmo]

Summary

Fixes LIMITATIONS.md item 7, which was previously deliberately left as "not fixed — this needs a real design answer, not a guess."

validate_invariants rejected every message type outright if its self-reported pir_invariants.residual exceeded 1e-12 — including ControlParameter, Heartbeat, MusicalContext and ArtistTracking, none of which carry a genuine multi-value invariant. The Resolume bridge computed that residual as a Prouhet-Thue-Morse-signed sum over its last 1–4 recently-set, often unrelated parameter values. A single layer's opacity change (one value) always failed — a lone value can never cancel by construction. Matched multi-value updates passed or failed depending purely on which arbitrary window a given message happened to land on at send time.

The design answer

Equilibrium is a property of a scene — a set of related values a sender deliberately constructed to balance together (the Kraken example: 8 tentacles' wobbliness, paired so the PTM-signed sum cancels) — not of one independent fact like "opacity is now 0.85." Only SceneUpdate can genuinely carry that. Every other message type is no longer gated on residual/coherence-gap.

This isn't a loosened check, it's a correctly-scoped one. The conductor already has the real mechanism for tracking equilibrium across a sequence of independent updates: it recomputes actual aggregate equilibrium from the real accumulated scene state after every message (SceneState::update_equilibrium_status), and the shipped default rule file already acts on that real value (if equilibrium > 1e-12 then escalate(...) in kraken-tentacle.only-pchi). The message-level gate was duplicating that check, badly, over data that was never a valid invariant claim.

Test plan

• 4 new tests: SceneUpdate still rejects a real equilibrium violation (regression guard — this is exactly the case the check exists for); ControlParameter/Heartbeat/MusicalContext/ArtistTracking are no longer gated on residual
• cargo test --workspace: 29/29 pass
• cargo clippy --workspace --all-targets -- -D warnings: clean
• Live re-run of the exact documented scenario: a real conductor + the real (fixed, Fix Resolume bridge: wire-format bugs meant zero messages ever reached the conductor #3) Resolume bridge. A single opacity change, previously always rejected → now applied. An 11-message burst across mixed parameters, previously 5/11 applied and 6/11 dropped at random → now 25/25 applied, 0 dropped, across both scenarios combined

🤖 Generated with Claude Code