Release 57: bring main into the integration branch (conflicts resolved) #324

2026-10-14T06:22:53Z

mesh-admin commented

2026-10-14 06:22:53 +00:00

What

Brings main (at the time of the merge) into release 57's integration branch feat/57-config-and-provisioning. Main gained commits since the branch was cut, among them #319 (two provider-hold store tests), the rest-at-once walk (ADR 0353), the try verb (issue 686), base lending before a push (issue 694) and the no-move tier fix (issue 648). The plain pull request from main (#322) did not merge cleanly and was closed.

Why

merge-check.sh's store-tests drop check compares a change's testdata/store-tests against origin/main's. The release branch lacked main's new entries, so the check failed every pull request into it.

How it was made

Locally git merge origin/main on top of origin/feat/57-config-and-provisioning (221fe5a6). The merge's whole effect (git diff origin/feat/57-config-and-provisioning HEAD) is pushed here as three commits only because the patch (135 KB) was too large for one forge call; the last commit's tree equals the local merge's tree (3ed03383), checked with git fetch and git diff (empty).

Conflicts and how each was resolved

testdata/store-tests: the only conflict. Both sides appended entries after TestAReportOfABusyContainerIsKeptAndReadBack. Resolved as the union of both lists in the file's order: the branch's three entries (TestPlacementMovesOutOfTheSettingsLayers, TestASettingRoutesPlacementBesideTheLayer, TestADeclarationIsTheSameAfterPlacementMoves) first, then main's eleven (the rest-at-once tests, TestANoMoveTierWaitsForNoNewReport, and #319's TestOneProviderDownAndThreeConsumersFailingAreOneCondition and TestAConsumerFailingBesideAHealthyProviderIsItsOwn). No duplicates; 74 entries.

Every other file merged automatically (including internal/builder/builder.go). No code was changed beyond the merge.

Checks run (on the merge, locally)

gofmt -l over cmd internal examples: clean.
go vet ./...: clean.
MESH_TEST_POSTGRES=postgres://postgres@127.0.0.1:<port>/postgres?sslmode=disable sh merge-check.sh against a throwaway PostgreSQL over TCP (127.0.0.1, dynamic_shared_memory_type=mmap, stopped afterwards): exit 0. Whole suite under the race detector passed; the drop check against origin/main passed; all 74 listed store tests ran against the store and passed, none NOT RUN or FAIL.

Reviewer's checklist

Scope: only main's commits plus the store-tests union; nothing else.
Tests: main's own tests come with it; all pass on the merge.
Composition: no change beyond what main already carries.
No secret, no hand-written database change; no migration added.
Parsers: unchanged beyond main.
gofmt, go vet, merge-check.sh pass (above).
Words follow the glossary.

Co-Authored-By: Claude Opus 5.5 noreply@anthropic.com

## What Brings `main` (at the time of the merge) into release 57's integration branch `feat/57-config-and-provisioning`. Main gained commits since the branch was cut, among them #319 (two provider-hold store tests), the rest-at-once walk (ADR 0353), the `try` verb (issue 686), base lending before a push (issue 694) and the no-move tier fix (issue 648). The plain pull request from main (#322) did not merge cleanly and was closed. ## Why `merge-check.sh`'s store-tests drop check compares a change's `testdata/store-tests` against `origin/main`'s. The release branch lacked main's new entries, so the check failed every pull request into it. ## How it was made Locally `git merge origin/main` on top of `origin/feat/57-config-and-provisioning` (221fe5a6). The merge's whole effect (`git diff origin/feat/57-config-and-provisioning HEAD`) is pushed here as three commits only because the patch (135 KB) was too large for one forge call; the last commit's tree equals the local merge's tree (3ed03383), checked with `git fetch` and `git diff` (empty). ## Conflicts and how each was resolved - `testdata/store-tests`: the only conflict. Both sides appended entries after `TestAReportOfABusyContainerIsKeptAndReadBack`. Resolved as the union of both lists in the file's order: the branch's three entries (`TestPlacementMovesOutOfTheSettingsLayers`, `TestASettingRoutesPlacementBesideTheLayer`, `TestADeclarationIsTheSameAfterPlacementMoves`) first, then main's eleven (the rest-at-once tests, `TestANoMoveTierWaitsForNoNewReport`, and #319's `TestOneProviderDownAndThreeConsumersFailingAreOneCondition` and `TestAConsumerFailingBesideAHealthyProviderIsItsOwn`). No duplicates; 74 entries. Every other file merged automatically (including `internal/builder/builder.go`). No code was changed beyond the merge. ## Checks run (on the merge, locally) - `gofmt -l` over `cmd internal examples`: clean. - `go vet ./...`: clean. - `MESH_TEST_POSTGRES=postgres://postgres@127.0.0.1:<port>/postgres?sslmode=disable sh merge-check.sh` against a throwaway PostgreSQL over TCP (127.0.0.1, `dynamic_shared_memory_type=mmap`, stopped afterwards): exit 0. Whole suite under the race detector passed; the drop check against `origin/main` passed; all 74 listed store tests ran against the store and passed, none NOT RUN or FAIL. ## Reviewer's checklist 1. Scope: only main's commits plus the store-tests union; nothing else. 2. Tests: main's own tests come with it; all pass on the merge. 3. Composition: no change beyond what main already carries. 4. No secret, no hand-written database change; no migration added. 5. Parsers: unchanged beyond main. 6. gofmt, go vet, merge-check.sh pass (above). 7. Words follow the glossary. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

mesh-admin added 3 commits 2026-10-14 06:22:53 +00:00

Bring main into release 57's branch, part 1 of 3: the builder's try outcome and the release walk 3d3ca6e500

So the release branch carries what main gained; parts 2 and 3 follow on this branch.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Bring main into release 57's branch, part 2 of 3: the rest at once and the try verb c88474a127

So the release branch carries what main gained; part 3 follows on this branch.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Bring main into release 57's branch, conflicts resolved

mesh/merge-gate head 4b2eac702ab1 merged into trunk 221fe5a60d1f — pass: the change touches no module of the mesh's graph

Details

mesh/repo-check head 4b2eac702ab1 merged into trunk 221fe5a60d1f — pass: THE CHANGE ALTERS ITS OWN CHECK (merge-check.sh): feat/57-config-and-provisioning'…

Details

mesh/delivery delivered

Details

4b2eac702a

So the release branch carries what main gained, and the store-tests drop check passes again.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

mesh-admin commented

2026-10-14 06:23:21 +00:00

Delivery novox/mesh-controller@4b2eac702ab1 — delivered since 2026-10-14T06:31:59Z

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

Transitions

2026-10-14 06:23 (new) → proposed (announced): novox/mesh-controller#324's head announced
2026-10-14 06:27 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit
2026-10-14 06:27 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
2026-10-14 06:31 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
2026-10-14 06:31 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.

**Delivery** `novox/mesh-controller@4b2eac702ab1` — **delivered** since 2026-10-14T06:31:59Z **Delivery plan** — builds nothing: the change touches no module of the mesh's graph **Transitions** - 2026-10-14 06:23 (new) → proposed (announced): novox/mesh-controller#324's head announced - 2026-10-14 06:27 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit - 2026-10-14 06:27 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail - 2026-10-14 06:31 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move - 2026-10-14 06:31 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-admin commented

2026-10-14 06:27:27 +00:00

Verdict: APPROVE at 4b2eac702a

Reproduced: in a fresh worktree of origin/feat/57-config-and-provisioning (221fe5a6), git merge origin/main (3479a354) conflicted only in testdata/store-tests. I resolved it as the union in file order (the feature branch's three placement tests, then main's eleven). git diff --cached --stat against the head is empty: the head's tree is identical to the true merge, so no code differs from what main carries.
testdata/store-tests has no duplicate lines, and every line of both sides is present.
The head is three linear commits (parts 1, 2, 3) on the feature branch, not a merge commit, so main is not an ancestor of it. The tree is right, but merging this pull request will not make main reachable from the branch history. Nothing must change, but a later merge of main will re-surface the same changes; consider it when merging.
gofmt -l lists only vendor/ files (none in repo code); go vet ./... clean.
merge-check.sh with a throwaway PostgreSQL over 127.0.0.1 (dynamic_shared_memory_type=mmap): exit 0, store tests pass. The server was stopped afterwards.

Verdict: APPROVE at 4b2eac702ab18290f910dfd29963c137280e77c6 1. Reproduced: in a fresh worktree of origin/feat/57-config-and-provisioning (221fe5a6), `git merge origin/main` (3479a354) conflicted only in testdata/store-tests. I resolved it as the union in file order (the feature branch's three placement tests, then main's eleven). `git diff --cached --stat` against the head is empty: the head's tree is identical to the true merge, so no code differs from what main carries. 2. testdata/store-tests has no duplicate lines, and every line of both sides is present. 3. The head is three linear commits (parts 1, 2, 3) on the feature branch, not a merge commit, so main is not an ancestor of it. The tree is right, but merging this pull request will not make main reachable from the branch history. Nothing must change, but a later merge of main will re-surface the same changes; consider it when merging. 4. gofmt -l lists only vendor/ files (none in repo code); go vet ./... clean. 5. merge-check.sh with a throwaway PostgreSQL over 127.0.0.1 (dynamic_shared_memory_type=mmap): exit 0, store tests pass. The server was stopped afterwards.

mesh-admin merged commit 9932adfa00 into feat/57-config-and-provisioning

2026-10-14 06:31:44 +00:00

mesh-admin deleted branch pkg/57-sync-main-1

2026-10-14 06:31:45 +00:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: novox/mesh-controller#324