Repository navigation
performance: the song decides whether parts loop; no checkbox - #250
Merged
Merged
Conversation
The performance checkbox bundled four things: waits active or inert, part jumps allowed, the agent's fast path and stage tools, and an automatic agent-session switch. Only the first is a real choice, and the song already states it: a part loops because it ends in waitForSignal. - Live playback always honours the waits; loopParts(false) in the song makes the waits after it play straight through (reset per compile). An export - an OfflineAudioContext render - never waits, as before. - Jumping to a part by name works whenever the song plays. - The agent is on stage whenever the song plays and has parts: part names and "next" are dispatched by the panel, other text is a stage turn (stage prompt, low effort, playhead state). Stopped, it is an ordinary chat. The producer also has list_parts/go_to_part/send_signal. No automatic session switch: /new starts a fresh one (worth it before a show). - The checkbox and window.togglePerformanceMode are gone. Tests: sequencer (loopParts inert waits; compiler marks waits after loopParts(false)), the worklet spec (loops by default, loopParts(false) plays through), the stage-agent e2e (stage while playing, ordinary chat when stopped), roles; song-api.md updated. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Deploying webassemblymusic with
|
| Latest commit: |
44fe368
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://b41e73dc.webassemblymusic.pages.dev |
| Branch Preview URL: | https://song-driven-performance.webassemblymusic.pages.dev |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This removes the performance checkbox. Whether a part loops is now decided by the song, and navigating between parts always works while it plays.
Why. The checkbox bundled four things:
Only the first is a real choice, and the song already states it: a part loops because it ends in
waitForSignal.What changes:
loopParts(false)makes every wait after it play straight through; it resets on each compile. An export (anOfflineAudioContextrender) never waits, as before.list_parts,go_to_partandsend_signal./newstarts a fresh session, which is worth doing before a show.window.togglePerformanceMode.song-api.mdhas a new "Performance: looping parts and signals" section, includingloopParts.Tests:
loopParts(false)play through; the compiler marks the waits after it; each compile starts looping again.test-sequencer16/16.loopParts(false)plays through. Chromium passes locally.test-agent-tools146/146;test-quickjs-sandbox19/19; headless studio 3/3.🤖 Generated with Claude Code