fix(design): renumber colliding ledger ids DL-327..330 to DL-334..337 - #926
fix(design): renumber colliding ledger ids DL-327..330 to DL-334..337#926rigel-mintaka wants to merge 1 commit into
Conversation
`main` is red on `design-ledger-gate:check`: the delivery→EventFabric cutover record's ledger block (merged as #905) claimed DL-327..333, but four of those ids were already taken by records that merged first, so the ledger carries four duplicates and the gate fails on every branch rebased onto current `main`. Because the pre-push gate runs `moon ci`, this blocks every lane's submit — and a stale clone still reads green, so it is easy to miss. The four cutover rows move to the free tail; the earlier claimants keep their ids because already-merged records cite them and other lanes implement from those citations. | id | kept by | cutover row moved to | | --- | --- | --- | | DL-327 | SubjectService token principal (#868) | DL-334 (OQ-1 ack timing) | | DL-328 | gateway-credentials at-rest encryption (#859) | DL-335 (OQ-2 callback) | | DL-329 | compass-stack supervision | DL-336 (OQ-3 p1 DB role) | | DL-330 | apple-container macOS runner | DL-337 (OQ-3 p2 recovery) | DL-331/332/333 were already unique and keep their ids, so the cutover block is now contiguous at DL-331..337. Its record's `Ledger-impact` line and the four in-prose references are updated to match, including the DL-333 row's own citation of the reconnect trigger it rests on (DL-330 → DL-337). No ruling is altered — this is an identifier reconciliation. Same shape as the earlier DL-319/DL-320 → DL-322/DL-323 reconciliation. Verified: `moon run design-ledger-gate:check` passes, and the ledger parses to 304 rows with zero duplicate ids (it previously reported four). Ledger-impact: renumbers DL-327..330 to DL-334..337; no ruling changes. Refs RIG-3107 Co-authored-by: Matt Wilkinson <matt@rigel.build>
|
Merging to
After your PR is submitted to the merge queue, this comment will be automatically updated with its status. If the PR fails, failure details will also be posted here |
|
Compass engineering docs preview: https://compass-server-ledger-id-col.compass-eng-docs.pages.dev Deployed from Changed pages: |
|
Superseded by #927 — closing in favor of it. Both PRs carry a byte-identical ledger renumber (same md5 over the changed #927 additionally handles the fact that #905 merged at 04:10Z, so its record is frozen: it adds a post-freeze annotation carrying the id mapping and leaves the in-body No defect in this PR — it executed the routed instruction, which omitted the freeze consideration. Structural follow-up for the recurring collision is RIG-3364. |
1 participant
This PR is part of a stack containing 2 PRs:
mainmainis red ondesign-ledger-gate:check: the delivery→EventFabric cutoverrecord's ledger block (merged as #905) claimed DL-327..333, but four of those
ids were already taken by records that merged first, so the ledger carries four
duplicates and the gate fails on every branch rebased onto current
main.Because the pre-push gate runs
moon ci, this blocks every lane's submit — anda stale clone still reads green, so it is easy to miss.
The four cutover rows move to the free tail; the earlier claimants keep their
ids because already-merged records cite them and other lanes implement from
those citations.
DL-331/332/333 were already unique and keep their ids, so the cutover block is
now contiguous at DL-331..337. Its record's
Ledger-impactline and the fourin-prose references are updated to match, including the DL-333 row's own
citation of the reconnect trigger it rests on (DL-330 → DL-337).
No ruling is altered — this is an identifier reconciliation. Same shape as the
earlier DL-319/DL-320 → DL-322/DL-323 reconciliation.
Verified:
moon run design-ledger-gate:checkpasses, and the ledger parses to304 rows with zero duplicate ids (it previously reported four).
Ledger-impact: renumbers DL-327..330 to DL-334..337; no ruling changes.
Refs RIG-3107
Co-authored-by: Matt Wilkinson matt@rigel.build