Repository navigation
Conversation
…th deterministic tests
…th deterministic tests
… callback detection
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. |
…very across extended virtual callbacks
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1ad5e9f275
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".
…ed bounded ASRC pipeline
…independent callback sizes
… next hardware integration
… stable timer at t=0
…t storm and t=0 stability
…click field report
…endent source/render callbacks
…device JUCE transport
|
RECONCILIATION 2026-10-26 — Converted to DRAFT, not merged. Recovery/clock-bridge prototype branch diverges from current main and its old JuceIndependentDeviceAdapter/IndependentClockPipeline design is no longer the live integration: merged #35 uses JuceIndependentClockDevice, with recovery/watchdog/fail-closed route hardening from #39/#40/#45/#50/#55. Archive this for traceability; re-implement any demonstrated missing tests against current main (do not copy whole old CMake/LiveAudioEngine). P0 hardware acceptance remains OPEN at #15. |
|
Backlog reconciliation — archived as SUPERSEDED (NOT merged). Stacked device recovery helper overlaps #25/#29 and current lifecycle/epoch code. Preserve useful token/clock tests for hardware #15; do not merge obsolete CMake/LiveAudioEngine. Current PRD v5 / ARD v5. Active owners: native #266 / epic #192, UI #265, physical #15, Show #180 and owner GUI #176. This PR's historical commits remain accessible, and no head is blindly merged into modern engine. |
Consolidated P0 engine work
Implement deterministic, generation-token-based device lifecycle policy with bounded retries, fail-closed endpoint loss, stale async event rejection, and stable-state persistence gating. Add provenance-correct capture/render counters with explicit dropped/missing frames for future split-stream adapter; these are NOT hardware driver xruns. All callback-facing operations use bounded atomics, no allocations/locks. Add native Release-active CTest coverage for recovery and split stream telemetry.
Scope and next gate
This is a P0 foundational PR, not completion of integrated native WASAPI independent capture/render streams, device auto-fallback or hardware listening FAT. Existing JUCE joined audio path untouched until the real adapter can be validated. GUI v0.8 and existing COPY FULL REPORT preserved. Require Windows MSVC + 8 native CTests and Chromium CI GREEN before merge.
P0 tracker: #15
Additional native transport added after initial review
IndependentClockPipelineexercises two independent callback timebases with bounded FIR bridge and frame-level telemetry in native CTest.JuceIndependentDeviceAdapterimplements source-only and render-only JUCEAudioDeviceManagerinstances. Capture callback feeds ASRC producer, render callback pulls and dispatches prepared PCM to host processor in bounded 8192-frame chunks; no heap or COM operations occur in callbacks. Adapter header is MSVC-compiled fromLiveAudioEngine.hbut intentionally not activated in the production callback path pending transactional device migration, automatic rollback and hardware verification.Remaining: device migration from joined JUCE into split sessions, AUTO/STABLE runtime selection, real render underrun provenance and complete hardware FAT. Issue #15 remains open; this PR is not a claim to finish P0 or eliminate physical crackling.