Skip to content

Session Handoff [auto-2731]: Deduplicate pr_review.py wait's Liveness Triple and Held Q_FULL Read (#2731) #2733

Closed

Description

@ptr727

Next steps, in priority order

  1. Fix Deduplicate pr_review.py wait's Liveness Triple and Its Back-to-Back Held Q_FULL Reads #2731 in scripts/pr_review.py wait. Done looks like both of the issue's items landed on develop through a feature pull request whose body carries Closes on promotion: #2731:
    • One pure helper, taking the payload and min_rounds, computes the liveness triple (head_review_done, answered_outside_review, reviewer_login_drift), and both main's first Q_LIVE evaluation and live_state call it, so the first stop decision and every later one read the same conditions.
    • A held wait opens its poll on the snapshot it already read rather than a second back-to-back Q_FULL read, with the held-wait tests that count reads updated to the new count.
  2. Run the script's test suite and the hub's full local gate set, then local-strict-review, before the push.

External blockers

None known at pick time.

Internal dependencies

  • Step 1's helper must keep every stop reading main and live_state take today. The issue's risk is a reading present at one site and missing from the other, so a test asserting both sites reach the same stop decision on the same payload guards the refactor.
  • The read-count change and the held-wait test updates land together, since the tests count reads.

State

The parked decision queue

None. Count 0.

What the last round did

Nothing on this lane. This is its first link.

What not to repeat

Nothing recorded yet.

New learnings

None yet.

Activity

  1. added
    handoffA link in the session handoff chain, one open issue per track
    on Oct 10, 2026
  2. ptr727 commented on Oct 10, 2026

    @ptr727
    OwnerAuthor

    Close-out

    The lane is complete, so no successor link follows.

    • Merged: Share pr_review.py wait's Liveness Readings and Open the Held Poll on Its Snapshot #2734 was squash-merged into develop as f905191. Its body carries Closes on promotion: #2731, so Deduplicate pr_review.py wait's Liveness Triple and Its Back-to-Back Held Q_FULL Reads #2731 stays open until develop is promoted to main. The run's scope was develop, so no promotion was opened or merged.
    • What it fixed (Deduplicate pr_review.py wait's Liveness Triple and Its Back-to-Back Held Q_FULL Reads #2731):
      • A new pure helper, liveness(pr, min_rounds), computes the three liveness readings. main's first Q_LIVE evaluation and live_state both call it.
      • A held wait now opens its poll on the snapshot it already read rather than making a second back-to-back Q_FULL read.
      • The held-wait tests that count reads were updated. Two tests were added: one confirms the first evaluation and every poll go through the same helper, and one confirms live_state returns liveness of the same payload.
    • Review:
      • local-strict-review ran two recorded passes. The first found one introduced finding, a stale test name and docstring, which was fixed in a170f96. The second found nothing.
      • Copilot raised one finding, which misread live_state's third positional argument as min_rounds when it is num. CodeRabbit raised one finding asking to restore the removed Q_FULL read. Both were declined with code-path evidence in their threads and resolved.
    • Filed along the way: nothing. No lesson issue and no decision issue.
    • Worktree and branch: both were removed after the merge. The remote branch's tip was checked against the merged head before it was deleted.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    handoffA link in the session handoff chain, one open issue per track

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions