Try before the review, fail first on behaviour, survey every reader of a changed value (playbooks 07 and 08) #433

Merged
mesh-tools merged 3 commits from docs/review-speed into main 2026-10-19 11:08:51 +00:00
Member

The operator asked (2026-10-19, after milestone R3b of design 47) why a review almost never takes one round, and to change what speeds the process up. R3b took two review rounds and one more cycle after a failed try, for three causes this changes:

  • A test that fails first only by not compiling has not failed first: the agent takes out the change's effect and sees an assertion fail (playbook 07 step 1, playbook 08 step 3). R3b: a new field on a check request was proven by a test of its helper alone.
  • The prepare survey lists every other reader and writer of each value a change adds or changes, with a criterion for each that could treat it wrongly (playbook 07 step 0 item 5). R3b: the router's sync ended every condition the controller did not list.
  • The branches are brought up to the trunk, pushed and tried at the end of step 1, before the review; the reviewer reads the tries' results; step 3 pushes the reviewed heads and tries again only a head the review changed (playbooks 07 and 08, and the overview's line for 08). R3b: the reviewer reasoned about a bus test's timing, wrongly, and the try after the review found it.

The mesh instruction section how-work-is-done states the old order; its change is proposed to the operator beside this pull request (claude_code_instructions_propose).

Review: 3 rounds (round 1: three sentences left in 07 that still said the old order and the trunk merge at the review's start; round 2: a promise step 3 did not keep). hq's merge-check.sh passed.

Reviewed-by: reviewer-r3b@shanks at 6c607504de

The operator asked (2026-10-19, after milestone R3b of design 47) why a review almost never takes one round, and to change what speeds the process up. R3b took two review rounds and one more cycle after a failed try, for three causes this changes: - A test that fails first only by not compiling has not failed first: the agent takes out the change's effect and sees an assertion fail (playbook 07 step 1, playbook 08 step 3). R3b: a new field on a check request was proven by a test of its helper alone. - The prepare survey lists every other reader and writer of each value a change adds or changes, with a criterion for each that could treat it wrongly (playbook 07 step 0 item 5). R3b: the router's sync ended every condition the controller did not list. - The branches are brought up to the trunk, pushed and tried at the end of step 1, before the review; the reviewer reads the tries' results; step 3 pushes the reviewed heads and tries again only a head the review changed (playbooks 07 and 08, and the overview's line for 08). R3b: the reviewer reasoned about a bus test's timing, wrongly, and the try after the review found it. The mesh instruction section how-work-is-done states the old order; its change is proposed to the operator beside this pull request (claude_code_instructions_propose). Review: 3 rounds (round 1: three sentences left in 07 that still said the old order and the trunk merge at the review's start; round 2: a promise step 3 did not keep). hq's merge-check.sh passed. Reviewed-by: reviewer-r3b@shanks at 6c607504de1bf902efbdd7320f06df444e9bc420
mesh-tools added 3 commits 2026-10-19 11:05:40 +00:00
Say a moved trunk is met by the merge check, not promised to step 3 (review round 2)
mesh/merge-gate head 6c607504de1b merged into trunk 7e1c0f4c1564 — pass: the change touches no module of the mesh's graph
mesh/repo-check head 6c607504de1b merged into trunk 7e1c0f4c1564 — pass: its merge-check.sh passed
mesh/delivery delivered
6c607504de
Author
Member

Delivery novox/hq@6c607504de1b — delivered since 2026-10-19T11:08:52Z

Delivery plan — builds nothing: the change touches no module of the mesh's graph

Transitions

  • 2026-10-19 11:05 (new) → proposed (announced): novox/hq#433's head announced
  • 2026-10-19 11:06 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit
  • 2026-10-19 11:06 checked → rejected (refused): the gate failed or could not run, or the repository's own check failed
  • 2026-10-19 11:07 rejected → proposed (recheck): judged again against trunk 7e1c0f4c15 (it was rejected against trunk )
  • 2026-10-19 11:07 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit
  • 2026-10-19 11:07 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
  • 2026-10-19 11:08 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
  • 2026-10-19 11:08 published → delivered (done): nothing for a walk to move

The commit's note under refs/notes/mesh-plan keeps every transition: git log --notes=mesh-plan.

<!-- mesh-delivery:view --> **Delivery** `novox/hq@6c607504de1b` — **delivered** since 2026-10-19T11:08:52Z **Delivery plan** — builds nothing: the change touches no module of the mesh's graph **Transitions** - 2026-10-19 11:05 (new) → proposed (announced): novox/hq#433's head announced - 2026-10-19 11:06 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit - 2026-10-19 11:06 checked → rejected (refused): the gate failed or could not run, or the repository's own check failed - 2026-10-19 11:07 rejected → proposed (recheck): judged again against trunk 7e1c0f4c1564 (it was rejected against trunk ) - 2026-10-19 11:07 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit - 2026-10-19 11:07 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail - 2026-10-19 11:08 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move - 2026-10-19 11:08 published → delivered (done): nothing for a walk to move The commit's note under `refs/notes/mesh-plan` keeps every transition: `git log --notes=mesh-plan`.
Author
Member

Merge check at 6c607504

  • mesh/merge-gate: ERROR — the check could not run: the facts snapshot cannot be read, and a check without it judges nothing: cannot reach the artifact store at novox.internal:5100: Get "http://novox.internal:5100/v2/facts/manifests/latest": dial tcp: lookup novox.internal on 10.10.0.3:53: read udp 10.10.0.3:43562->10.10.0.3:53: i/o timeout
  • mesh/repo-check: ERROR — the check could not run: the facts snapshot cannot be read, and a check without it judges nothing: cannot reach the artifact store at novox.internal:5100: Get "http://novox.internal:5100/v2/facts/manifests/latest": dial tcp: lookup novox.internal on 10.10.0.3:53: read udp 10.10.0.3:43562->10.10.0.3:53: i/o timeout

Run by the build seat on shanks as build-1792407943147608571 (builds --log build-1792407943147608571).

**Merge check** at `6c607504` - `mesh/merge-gate`: **ERROR** — the check could not run: the facts snapshot cannot be read, and a check without it judges nothing: cannot reach the artifact store at novox.internal:5100: Get "http://novox.internal:5100/v2/facts/manifests/latest": dial tcp: lookup novox.internal on 10.10.0.3:53: read udp 10.10.0.3:43562->10.10.0.3:53: i/o timeout - `mesh/repo-check`: **ERROR** — the check could not run: the facts snapshot cannot be read, and a check without it judges nothing: cannot reach the artifact store at novox.internal:5100: Get "http://novox.internal:5100/v2/facts/manifests/latest": dial tcp: lookup novox.internal on 10.10.0.3:53: read udp 10.10.0.3:43562->10.10.0.3:53: i/o timeout Run by the build seat on shanks as `build-1792407943147608571` (`builds --log build-1792407943147608571`).
mesh-tools merged commit f269e83f5d into main 2026-10-19 11:08:51 +00:00
mesh-tools deleted branch docs/review-speed 2026-10-19 11:08:51 +00:00
Sign in to join this conversation.
No Reviewers
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: novox/hq#433