Skip to content

chore(console): bump objectui pin 0cf8f0f7 → b1204af0 (22 commits, 8 releasing changesets, 0 breaking) - #6694

Merged
os-project-manager merged 1 commit into
mainfrom
claude/issue-6663-objectui-pin-bump
Aug 8, 2026
Merged

chore(console): bump objectui pin 0cf8f0f7 → b1204af0 (22 commits, 8 releasing changesets, 0 breaking)#6694
os-project-manager merged 1 commit into
mainfrom
claude/issue-6663-objectui-pin-bump

Conversation

@os-project-manager

Copy link
Copy Markdown
Collaborator

Fixes #6663

Moves .objectui-sha from 0cf8f0f70d10 to the objectui origin/main head measured at execution time, b1204af0a1f7, together with the single @objectstack/console changeset that scripts/bump-objectui.sh generates. The file surface is exactly those two files — 36 insertions, 1 deletion. Nothing else rides this bump.

SHA and interval

item value
old pin 0cf8f0f70d1040a63316e0c55cb2fd2e87989d73 (set by #6476)
new pin b1204af0a1f7256514bab4479f5e9c915fb91449
head commit docs(create-plugin): 按 route B 收缩到不漂移的部分,产物一律指向 buildPluginFiles() (#3715) (#3760), 2026-08-08T20:17:37Z
ahead_by 22
behind_by 0 — the pin is a true ancestor of objectui main
measured at 2026-08-08, re-fetched immediately before running the script; drift on re-measure = 0

Every count quoted on the ticket is superseded. The steward's three readings (4 → 8 → 12 commits) had all expired by claim time; the measured interval is 22 commits / 8 releasing changesets. 918888a3, the SHA in the ticket title, is not the end of the interval — it is the fourth commit inside it.

Ancestry was confirmed locally with git merge-base --is-ancestor (exit 0) plus git rev-list --count in both directions (22 forward, 0 back), in a dedicated objectui worktree checked out at origin/main — not the shared checkout, whose HEAD moves under a running task.

Digest

scripts/bump-objectui.sh reported:

→ objectui pin: 0cf8f0f70d10 → b1204af0a1f7
→ 8 releasing changeset(s), 0 breaking (0 declared major, 0 annotated), 0 release-nothing, 14 commit(s) without a changeset
→ wrote changeset console-b1204af0a1f7.md (@objectstack/console: patch)

The 8 releasing changesets objectui declared over the interval, all patch:

changeset package
system-hub-permissions-leg-3655.md @object-ui/console
plugin-report-react-19-peer-3690.md @object-ui/plugin-report
create-plugin-runnable-test-stack.md @object-ui/create-plugin
create-plugin-build-deps-anchored-3742.md @object-ui/create-plugin
data-table-row-menu-empty-guard.md @object-ui/components
dataset-selection-spec-type-3613.md @object-ui/data-objectstack
nav-sys-objects-canonical-route-3739.md @object-ui/app-shell
pivot-compare-stacking-3614.md @object-ui/plugin-dashboard

The remaining 14 commits carry no .changeset/*.md; the generated changeset itemizes them by subject rather than counting them, per #6174.

#6099 re-verification — clean, and verified twice independently

#6099's failure mode is a v17-window breaking change declared minor with the breakage stated only in the changeset body, which the digest cannot flag as major. Re-verified against the interval actually bumped, not the one on the ticket:

  1. By hand — all 8 changeset bodies read in full with git show, scanning for BREAKING, breaking change, 不兼容, 破坏性, migration required. Zero hits. All 8 frontmatter blocks declare patch; there is no minor and no major in the range.
  2. By the script — the digest independently reports 0 breaking (0 declared major, 0 annotated).

Both agree, so there is nothing to mis-classify and the resolved bump is patch. Consequently — and unlike #6476, whose generated changeset carried two **BREAKING** markers and needed an ADR-0087 disposition — check:adr-0087-registration passes untouched here: this PR adds no declared-breaking changeset (1 non-breaking changeset(s) seen). No ledger entry was added or needed.

This does NOT unblock #6208, and this PR does not claim it does

The ticket title reads required before the open rc release PR #6208 can go green. Per the steward's 11:2xZ correction, that is true but not sufficient, so the justification here rests on the bump's own merits only:

scripts/check-objectui-pin-fresh.mjs rules stale whenever the compare status is ahead, independently of whether the range declares changesets#3340 red condition 1, "the judging ref is AHEAD of the pin — frontend commits are unbundled". A 22-commit lag is enough on its own.

Separately, no gate on #6208 has ever produced a verdict: every workflow run on changeset-release/main concludes action_required with zero jobs executed, which is a maintainer workflow-approval matter and outside this PR's reach. Landing this bump does not change that, and the absence of a red pin gate there must not be read as the pin having passed.

Ordering: this must merge before #6208. That PR regenerates on every push to main and must pick up this bump's @objectstack/console changeset.

Gates run locally

All run after git commitcheck-adr-0087-registration reads committed state (git show HEAD:), so staging alone leaves it red.

gate result
pnpm install --frozen-lockfile pass — confirms no lockfile regeneration is part of this bump
pnpm lint (ESLint) pass, exit 0
pnpm check:objectui-pin-fresh pass — verdict flipped to FRESH: ".objectui-sha is objectui main (objectui@b1204af0a1f7)"
pnpm check:objectui-changeset pass — digest + range self-tests green
pnpm check:console-sha pass (skip branch) — "No console dist … skipping"; the artifact is gitignored and built by build-console.sh
pnpm check:empty-changeset pass — 48 assertions, "1 declaring changeset(s) added"
pnpm check:adr-0087-registration pass — 100 assertions, no declared-breaking changeset
pnpm check:nul-bytes pass — 6237 tracked text files scanned, no raw ASCII control bytes
remaining 27 gates of the ESLint job all pass — slot-lookup, query-options-erasure, verify-stand-in, doc-authoring, docs-audit-scope, role-word, quick-reference-counts, adr-anchors, org-identifier, authz-resolver, service-providers, route-envelope, error-code-casing, wildcard-fallthrough, meta-type-normalized, init-service-contract, durability-log-level, startup-registry-verdict, release-notes, release-body, node-version, workflow-status-functions, shard-attestation, published-files, engine-double-contract, kernel-hook-pairs, resume-authority-declared, merge-driver, spec-parsed-alias

The gate list was enumerated from .github/workflows/lint.yml rather than guessed. The TypeScript Type Check job is left to CI: this diff changes zero TypeScript files, and CI's conclusion on that job is the binding one.

Deliberately not done

scripts/bump-objectui.sh closes by reminding that a pin move should be followed by pnpm sdui:manifest (the ADR-0082 D4 declaration-parity ratchet). It was not run here. It needs Playwright chromium and a full objectui build, and its outputs land outside the two files this ticket allows. Per #5960 it is an on-demand maintenance step deliberately kept out of CI (docs/releases-maintenance.md → "After the pin moves"); recorded here so this bump does not swallow it silently.

No changeset beyond the generated one, and no skip-changeset label: this PR releases @objectstack/console.


Generated by Claude Code

docs(create-plugin): 按 route B 收缩到不漂移的部分,产物一律指向 buildPluginFiles() (#3715) (#3760)

objectui@b1204af0a1f7256514bab4479f5e9c915fb91449
@vercel

vercel Bot commented Aug 8, 2026

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment
Project Deployment Actions Updated (UTC)
objectstack Ignored Ignored Aug 8, 2026 12:55pm

Request Review

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

documentation Improvements or additions to documentation size/s tooling

Projects

None yet

2 participants