[PR] fix: preserve new Side Chat sessions and stabilize Intel release checks [maka]

["Astro-Han (via GitHub)" <gi...@apache.org>]

Astro-Han opened a new pull request, #6111:
URL: https://github.com/apache/maka/pull/6111

   ## Summary
   
   Fix the two failures observed on `main` at `8686793e5`:
   
   - Side Chat creates a fork and admits it into the preload catalog, but a subsequent local catalog read can still return the older snapshot. Treating that snapshot as complete removes the new fork. Reuse the existing invalidation version to withhold absence authority until the catalog catches up, while continuing to publish observed running activity. The regression passes through the real local service and preload reconciliation and fails on the previous implementation.
   - The Intel macOS Nightly exceeded the real Host regression's 10-second startup deadline while release contract files ran concurrently. On the same Intel runner, the unchanged parallel suite reached Ready in 6.1 seconds and serial execution in 3.1 seconds; all 220 tests passed in both runs. Serialize the release contracts with one test-runner flag, retaining every test and deadline. Also include child output in startup timeout errors. The historical timeout did not preserve child output, so its exact internal stall cannot be reconstructed.
   
   No stored-data migration or public API change. Existing Electron acceptance coverage remains unchanged.
   
   Evidence: [failed main CI](https://github.com/apache/maka/actions/runs/38556979642), [failed Intel Nightly](https://github.com/apache/maka/actions/runs/38619621522/job/116045101181), [Intel parallel/serial comparison](https://github.com/apache/maka/actions/runs/38686263917).
   
   ## Verification
   
   - Full build and Desktop typecheck passed.
   - Local service and preload catalog suites: 94 passed. The new fork regression fails without the fix; existing continuous-invalidation tests still verify live activity publication.
   - Final `npm run check:release`: 221 passed. The new timeout-output regression fails without the diagnostic change.
   - `npm run format`, `npm run lint`, and `git diff --check` passed.
   - Intel comparison: both 220-test suites passed; serial execution took 48 seconds versus 29 seconds in parallel.
   - Electron E2E is left to PR CI. The final updated suite has not been run on native Intel hardware; the remote comparison used the unchanged tests and the same serial runner setting.
   
   ## AI use
   
   - [ ] No generative tool made a substantive contribution
   - [x] Generative tooling made a substantive contribution
   
   Tool(s) and scope: Codex investigated the CI logs, implemented both changes, and ran the verification.
   
   ## Checklist
   
   - [x] Tests cover the change and fail without it
   - [x] Lint, format, typecheck and the affected suites pass locally
   
   Does this PR entail a change in behavior?
   
   - [x] Yes — described under Summary above
   - [ ] No
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@maka.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@maka.apache.org
For additional commands, e-mail: commits-help@maka.apache.org

Re: [PR] fix: preserve new Side Chat sessions and stabilize Intel release checks [maka]

["Astro-Han (via GitHub)" <gi...@apache.org>]

Astro-Han merged PR #6111:
URL: https://github.com/apache/maka/pull/6111


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@maka.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@maka.apache.org
For additional commands, e-mail: commits-help@maka.apache.org