Skip to content

refactor(mv8): pure native FX transaction plans and exact instance readback - #786

Open
masarray wants to merge 3 commits into
mainfrom
refactor/mv8-fx-transaction-view-model-20261118
Open

masarray wants to merge 3 commits into
mainfrom
refactor/mv8-fx-transaction-view-model-20261118

Conversation

@masarray

@masarray masarray commented Nov 18, 2026 •

Copy link
Copy Markdown
Owner

MV-8 — FX Rack Transaction Lifecycle

User outcome: Plugin insertion, movement/removal and rack/slot bypass gain deterministic, auditable fail-closed preflight without adding another controller/PCM owner. Editor navigation after LOAD_INSERT is allowed only after the original protocol.createController reports its correlated native ACK plus an independently accepted, exact lease-bearing FX graph readback.

Actual implementation

  • Extend existing web/approved-fx-rack-view-model.js with pure nativeInsertPlan, nativeEditPlan, nativeInsertSettlement, nativePowerPlan, and nativePowerAckResult; no DOM, event emission, timers, plugin ownership or accepted graph mutation in the VM.
  • Remove the duplicate bridge-side preflight and settlement policy. One original controller.intent, arMixAction transport, ACK/readback controller, 8/15s timeout and installed PCM publisher remain.
  • Fail closed on invalid/rounded >2^53 ChannelId, wrong slot, stale instance lease, ambiguous duplicate processor identities, cross-channel destination error and invalid FX request. A missing native ChannelId target cannot silently fall through to a legacy M4 routing command.
  • Independent global/rack/slot bypass state remains intact; compatibility M4 routing/ACK and logical/offline .smix remain unchanged.
  • Five targeted regression tests exercise positive/negative native FX load/edit/lease/power/bridge ownership paths. Existing native FX / Guitar FX / operator editor contract assertions were reconciled to verify the new pure-VM boundary rather than removed bridge code. All legacy assertions still require the real controller, readback and editor identity.

Scope: 4 JS/test files (two test files + bridge + FX ViewModel) only; NO JUCE/C++/ASIO/DSP/VST3 code, no .smix schema, HTML/CSS/PNG or other concurrent IO/ImmerSIB/SonKu changes. ONE approved /preview/ UI and ONE ChannelId/EdgeId native renderer.

Status: KODE DIBUAT / DRAFT PR. Source syntax and pure preflight PASS, exact-HEAD fast native/UI and full Chromium REQUIRED before merge. Windows hardware FAT/installer NOT RUN. Aman dipakai? BELUM sebagai installer baru. Tindakan Anda: TIDAK PERLU MELAKUKAN APA-APA.

Next: CI diagnosis and exact-SHA browser qualification, merge if protected check passes; continue original native modular-monolith #527 M2–M6 separately.

@cloudflare-workers-and-pages

cloudflare-workers-and-pages Bot commented Nov 18, 2026 •

Copy link
Copy Markdown

Deploying arsonmix with  Cloudflare Pages  Cloudflare Pages

Latest commit: fdd2a4b
Status: ✅  Deploy successful!
Preview URL: https://bdd7d676.arsonmix.pages.dev
Branch Preview URL: https://refactor-mv8-fx-transaction.arsonmix.pages.dev

View logs

Copy link
Copy Markdown
Owner Author

/arsonmix-ci ui 9529a74

@masarray
masarray marked this pull request as ready for review November 18, 2026 07:16
@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Nov 18, 2026 •

Copy link
Copy Markdown
Codex Review Summary: 🔄 Running

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

Review Status Commit Review trigger
📝 Code Review 🔄 Running since 2026-11-18T07:21:31.613416Z fdd2a4b Manual request
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.

Copy link
Copy Markdown
Owner Author

MV-8 safety correction — MOVE is never REMOVE

Source-review found that an earlier pure nativeEditPlan inferred MOVE from destinationId !== null. If a user intended to MOVE an FX but the view-to-ChannelId destination mapping vanished, that could incorrectly choose REMOVE_INSERT — unsafe for a live FX chain. This has been repaired in exact HEAD fdd2a4b0c1ae7325b7aad9a23cb4c807f97fae66:

  • Explicit move: !!target intent arrives from the original bridge.
  • move:true with a missing/noncanonical destination rejects without issuing ANY native transaction.
  • move:false with an unexpected destination also rejects. Only a deliberate genuine REMOVE may dispatch REMOVE_INSERT.
  • Added deterministic negative cases for lost ChannelId and malformed move intent; preserved the one real native controller and prior FX instance leases.

Source syntax and pure local check PASS. Earlier exact-SHA full Chromium and fast checks do not qualify this new commit; use this SHA's automatic Fast UI/Native, one full Chromium and required merge gate. Hardware FAT/installer NOT RUN.

Copy link
Copy Markdown
Owner Author

/arsonmix-ci ui fdd2a4b

Copy link
Copy Markdown
Owner Author

@codex review

Please review exact MV-8 FX Rack HEAD fdd2a4b0c1ae7325b7aad9a23cb4c807f97fae66, especially: a lost destination ChannelId during FX MOVE must fail closed and never route to REMOVE; native LOAD_INSERT must correlate original controller ACK, exact catalog and live instance readback before editor navigation; bypass cannot change rack/slot/global power outside accepted PCM. Maintain only the existing native dispatcher and approved /preview/.

This branch has not been deployed

No deployments
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