Uh oh!
There was an error while loading. Please reload this page.
docs(runtime): retire the dead objectui-lockstep rename promise in packages.ts - #7904
Conversation
…ckages.ts The `unhiddenApps` / `unhideError` comment at packages.ts promised a rename riding "the objectui follow-up card, together." That card was #6955, whose dev measured the rename out rather than in: the server still emits the old names and objectui reads neither field anywhere (zero grep hits repo-wide), so the PM ratified "not at all" as option A — a zero-reader diagnostic rename buys no capability. Reword the comment to record that settled state, cite #6955, and note the option-B escape hatch (a standalone producer-side rename card) for anyone who later wants the vocabulary tidy on its own merits. Comment-only; zero behaviour change. Fixes#7578. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01B3Kurx8qufrDzNjk4rag7V
The latest updates on your projects. Learn more about Vercel for GitHub. 1 Skipped Deployment
|
📓 Docs Drift CheckThis PR changes 1 package(s): 20 hand-written doc(s) reference the affected code and may need an implementation-accuracy re-verification:
⛔ 2 release-owned page(s) also reference the affected code. These are read-only:
|
Uh oh!
There was an error while loading. Please reload this page.
Fixes#7578
What
Comment-only reword at
packages/runtime/src/domains/packages.ts(the block documenting why thepublish-draftsresponse keeps theunhiddenApps/unhideErrorwire-field spelling). The old text promised: "The rename rides the objectui follow-up card, together."That follow-up card was #6955. Its dev measured the rename out rather than in: the server still emits the old names (unchanged emit sites — not touched by this PR) and objectui reads neither field anywhere (zero grep hits repo-wide). Under the card's own "rename in lockstep, or not at all" rule, the resolution was not at all — PM-ratified as option A: a rename of a zero-reader diagnostic payload buys no capability (startup-scope discipline).
The comment now records that settled state instead of pointing the next reader at work that will never arrive: the spellings are kept deliberately and permanently, cites #6955 for the measurement + ruling, and notes the option-B escape hatch (a standalone producer-side rename card) for anyone who later wants the vocabulary tidy on its own merits.
Scope proof
One file, one hunk, comment-only:
Zero behaviour change. The three emit sites the card called out as off-limits (
unhideErrorassignment,unhiddenAppsassignment, themetadata:reloadedannounce read) are untouched — confirmed by diff review, not just by not editing them.Tests
node scripts/check-nul-bytes.mjs— OK (7230 files scanned).pnpm check:changeset-gate-self-tests— green (empty-changeset, ADR-0087 registration, changeset-no-major self-tests).pnpm check:route-envelope— green (9 route modules audited, no regression).pnpm --filter '@objectstack/runtime^...' buildthenpnpm --filter '@objectstack/runtime' build— green (build closure first, per AGENTS.md [skill] 新 worktree 里第一次验证之前必须先 build 依赖闭包 —— AGENTS.md §9 的陈旧产物陷阱当日连咬三个 dev(假红 + 假绿两个方向) #6371).pnpm --filter '@objectstack/runtime' typecheck— green, no errors.pnpm --filter '@objectstack/runtime' test— 137 test files, 2086 tests, all passed (includespackages-capability-gate.test.ts,packages-uninstall-envelope.test.ts,packages-readonly-gate.test.ts).No new tests: comment-only diff has no observable behaviour to test.
Changeset
skip-changesetlabel applied — comment-only diff releases nothing.Generated by Claude Code