Repository navigation
Conversation
…se port and PCM transaction
…ACK before arming
…, never browser PCM
…integration owner
…without legacy owner deadlocks
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: d62c6d4436
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| staged.getDynamicObject()->setProperty("mainOutputs",juce::var(outputPins)); | ||
| return publishRevisedTopology(staged,expectedEpoch,expectedRevision, |
There was a problem hiding this comment.
Refresh the cloned route's revision before publishing MAIN outputs
After PATCH CH1 succeeds, choosing any MAIN output pair fails because sourceTemplate_ retains the command's pre-publication revision, while the accepted publisher revision has already advanced by one. This clone changes only mainOutputs; publishRevisedTopology() checks the current revision argument but then calls apply(), which reads the stale revision from staged, causing RealtimeRoutePublisher::publish() to reject it. Consequently, the new MAIN assignment workflow cannot succeed even with valid hardware pins and muted outputs. The failed attempt also clears channelIdPcmEpoch_ before publication, disabling the previously admitted ChannelId renderer. Update the staged revision to expectedRevision before invoking the compiler, as NativeVerifiedSourcePatch::prepare() does for input patches.
Useful? React with 👍 / 👎.
…instead of stale ENGINE symbols
…ect stale fabricated adapters
… output, not retired variant URLs
…rrent approved GUI regressions stay PR-required
…le-preview PR coverage
P0 ArSonMix — single-owner working audio product recovery
Base main:
6084d1e232f829fced708b7c4d8ca3d9522d4b57(already includes ENGINE #266 + First Real Sound #408 selective integration, and Windows/Chromium PASS on that exact base).Real implementation, not a GUI simulation
/preview/More/Audio Setup workflow. Enumerates enabled JUCE output endpoint masks and lets the user patch the chosen stereo pair while muted.routingMainOutputcommand in the existing WebView2→JUCEarMixActionbridge; nativerepatchDryRoutingLabMainOutputs()uses the sameLiveDryRoutingLab::publishRevisedTopologycompiler, graph publisher, VST3 lease and callback. Source persists as the native accepted topology and validates sparse hardware output mask.Remove product-blocking agent bureaucracy without weakening safety
--scopeclassification of actual PR changed files, still selecting required native MSVC/CTest, Chromium and exact-SHA Product Qualification.Engineering reference behavior
Grounded in Adobe Audition's Audio Hardware and multitrack device I/O, FL Studio's Audio Settings and mixer hardware I/O routing, and JUCE AudioDeviceManager/AudioIODeviceCallback. Driver and plugin latency are nonzero; goal is practical lowest stable monitoring latency, no UI-callback DSP or unbounded RT work.
Acceptance, strict evidence split
No stable release declaration before exact Windows artifact + physical audio testing. This is the one substantive product-recovery PR; do not split into workstream PRs.
Next: on exact-head green Windows qualification, land one integrated branch and produce user-testable EXE/ZIP using the main Windows packaging workflow; resolve first-sound physical issues from real diagnostics, not guessed audio patches.
2026-11-08 reconciliation: ARCHIVED, NOT MERGED
Historical P0 recovery proposal has been incorporated or displaced by the merged #418 native integration and single-owner rules in current AGENTS.md. Its old M4/CI coordination edits conflict with today's approved source. Residual hardware issues are still open in #412/#15.
Current authority: main, #418. Historical branch retained; hardware FAT not inferred.