Uh oh!
There was an error while loading. Please reload this page.
docs(metadata-protocol,objectql): stop hardcoding the shared authoring-rule count in comments (#7491) - #8046
Conversation
…g-rule count in comments
Three comments asserted that `AUTHORING_RULES` holds 26 shared author-time
rules. The registry declares 38, so all three were wrong.
Rather than re-hardcode 38 — the same statement that just went stale — the
count is dropped and the registry is left as the authority ("all the shared
`AUTHORING_RULES`"). The sentences each turn on ALL of the rules being
disengaged, never on how many there are, so nothing is lost and the wording
cannot expire when rule 39 lands. This matches how the prose docs already
refer to the table (content/docs/deployment/validating-metadata.mdx), which
names the registry and its file and states no count.
Comment-only; no behaviour change.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019Uu3qrB98cMbJ48ae3z9S7The latest updates on your projects. Learn more about Vercel for GitHub. 1 Skipped Deployment
|
📓 Docs Drift CheckThis PR changes 2 package(s): 16 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#7491
domain:metadataPM seat on the dev's behalf — branch pushed, no PR appeared, so the GitHub API is still unreachable from that container.The defect
Comments hardcoded a count of the shared
AUTHORING_RULES. The count was stale.⭐ The card said two sites. The dev found three —
packages/metadata-protocol/src/protocol.ts(×2) andpackages/objectql/src/plugin.ts(×1). The brief asked it to verify that count rather than trust it, and verifying it was worth doing.⭐ The fix removes the number rather than correcting it
The commit subject is the point: "stop hardcoding the shared authoring-rule count." A comment carrying a hand-maintained count of a registry will go stale again — that is what this card is. Correcting
26to the current number would have re-armed the same trap for whoever reads it next.Diff is 3 lines across 2 files. That is the whole change.
Why it was dispatched as the third concurrent editor of
protocol.tsUnder the maintainer's mutually-exclusive-region exemption, this lane runs a per-hot-file editor cap. I raised it 2 → 3 this round (stated on #7491,
5266987227) because the queue's remaining cards all land inmetadata-protocol— the lane was idle for lack of disjoint work, not work.⇒ This card was chosen for the third slot precisely because it is a comment-only change: very nearly the smallest conflict surface available in an 11k-line file, and textually distant from the two sibling regions (#8027's
sys_metadataoverlay path, #7893'sfieldwrite door). The brief told the dev to keep it that way and it did — no adjacent tidying, no reflow beyond the correction.Anchoring note
The card cited
protocol.ts:2632. That line number was stale three times over before this dispatch: it was already at:2698onorigin/main, then #7931 (7372d46) moved it, then #8015 (e3c8ed0) moved it again ~30 minutes before dispatch. The dev was told to find it by comment text, never by line number.Changeset
None — a comment-only change releases nothing.⚠️ If
Check Changesetobjects, the sanctioned route is theskip-changesetlabel (which the PM applies), ⛔ not a newly added empty-frontmatter changeset — those are rejected (#5471) because an all-empty set stalls the release silently and greenly (#4898).Generated by Claude Code