Skip to content

FX rack master power: bypass all effectively and restore each insert's last state (GUI + native engine) - #421

Closed
masarray wants to merge 20 commits into
mainfrom
fix/fx-rack-master-bypass-restore-20261105
Closed

masarray wants to merge 20 commits into
mainfrom
fix/fx-rack-master-bypass-restore-20261105

Conversation

@masarray

@masarray masarray commented Nov 5, 2026 •

Copy link
Copy Markdown
Owner

FX Rack master header power: parent bypass, preserve individual state

Requested operator semantics: when FX rack HEADER power is OFF, every occupied FX slot is effectively OFF and displays its power button OFF. HEADER ON restores the EXACT stored previous state for each insert (an individually OFF plugin stays OFF). Header is NOT a destructive mass toggle of the individual flags.

Architecture

  • ONE approved Pages + JUCE WebView2 mixer; V28 historical author source remains immutable. The official build assembler applies a fail-closed, source-anchored FX semantic transform to the same approved DOM; no second UI, CSS override or fake PCM.
  • Existing FX strip state remains fxOn (rack) + slotBypass[i] (remembered individual); effective ON = fxOn && !slotBypass[i]. FX slot row and ARIA power now follow this effective state, not the stored individual flag.
  • Actual realtime engine: independent per-channel rack gate in LiveChannelFxGraph and native DynamicChannelFxLeaseBank; the audio callback passes the OR of rack/master bypass and individual insert bypass into existing 8-ms wet/dry bypass envelope. No RT allocations, blocking, VST re-instantiation or direct editing of individual bypass preferences.
  • Graph reuse and native insert append retain the master rack state. M4 CH1 original native VST chain follows the same rack gate.
  • Native control: accepted ChannelId epoch/revision/token validation, dedicated native ACK plus subsequent independent PCM producer readback; no optimistic UI mutation. Both header and individual FX power actions are routed, and rejected/stale commands fail closed.
  • Approved native graph readback now conveys rack bypass separately from each plugin's original individual bypassed flag; the GUI recomputes effective power and restores its display when header is re-enabled.

Qualification

  • New Chromium test tests/approved_fx_rack_master_power_smoke.py on actual built /preview/ at 390/1366/1680: add second FX, slot 1 OFF, slot 2 ON, header OFF → both power buttons OFF and all effective false; header ON → first stays OFF, second ON; plugin IDs and CH2 unchanged.
  • C++ real FxGraph/ChannelId lease tests exercise saved individual state, epoch/revision, reuse and native append.
  • CI must be SUCCESS on exact PR head incl Windows JUCE native, RT CTest, full CSS V28 / approved preview single owner, WIRE/fader/FX smoke.
  • Physical Windows VST3/audio FAT still needs an actual device; green CI is not physical audio proof.

Do not merge on failing CI; do not claim all rack state migrated into stable release without exact Windows qualification.


2026-11-08 reconciliation: ARCHIVED, NOT MERGED

FX header master bypass, independent per-slot remembered state and native rack bypass are present in current main with an approved GUI composer and native lease gate. This branch's old UI/engine change set is superseded by the integrated #418 lineage; do not replay duplicate ownership.

Current approved authority: main, merged native integration #418. Existing branch history remains available; this closure is not evidence of physical audio FAT.

…euse and native append, rejects stale revision
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Nov 5, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-11-05T09:04:29.425472Z e9f2f08 PR opened
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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: e9f2f08e66

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".

Comment on lines +109 to +110
row.rackBypassed=b.rackBypassed?b.rackBypassed:
std::make_shared<std::atomic<bool>>(false); // CONTROL allocation

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Preserve rack bypass across native graph re-admission

When an original channel/AUX rack is bypassed and the operator performs a structural channel repatch with outputs disarmed, LiveAudioEngine::repatchDryRoutingLabChannel() calls admitChannelIdFromAcceptedM4(), which rebuilds the bank through UnifiedLegacyProductionAdmission::compileAndCommit(). That path obtains every assignment from captureLegacy(), which never populates rackBypassed, so this fallback initializes every rack to ON. Native rack commands update only the accepted bank, leaving no corresponding bypass state in the M4 graph to recover. Consequently, an unrelated routing edit silently re-enables previously bypassed effects when output resumes. Preserve the accepted rack gates through this re-admission path as well as through reuseRevision() and appendOwned().

AGENTS.md reference: AGENTS.md:L37-L37

Useful? React with 👍 / 👎.

masarray added a commit that referenced this pull request Nov 5, 2026
… reconcile #419

Publish a single control-thread mixerSession choosing admitted ChannelId or verified M4, route 10CH controls through only its PCM authority, correlate per-parameter M4 ACK to subsequent independent producer readback, preserve physical MAIN gain/mute with native ChannelId, fail closed unsafe JS FX rack fan-out pending #421 atomic native gate. Transfer #419 mobile GUI source exact blobs from main; no DSP callbacks or alternate preview.
masarray added a commit that referenced this pull request Nov 5, 2026
…418

Retain single production JUCE PCM engine and approved GUI. This unattached staging commit adds native rack overrides, real ChannelId lease readback, M4 callback bypass and host CSS semantics. Follow-up integration commit wires the existing #418 operator protocol and single mixerSession before publishing branch.
masarray added a commit that referenced this pull request Nov 5, 2026
…ernel golden coverage

Reconcile #418 native-only bridge with selected #421 per-rack atomic engine gate. Both M4 compatibility and ChannelId use exact accepted producer readback for rack ON/OFF without destroying individual slot states. Project legacy M4 scalars, physical source and wire only from mixerSession; add negative ACK/epoch gating, C++ mixerSession test, multi-channel native sample golden tests. Preserve one JUCE callback, public /preview/ visual design, and unmerged P1 #422. Physical FAT and complete gesture-to-PCM certification remain outstanding.
@masarray masarray closed this Nov 8, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant