decision on published-history-is-not-rewritten
Do not rewrite history that others may already have: no force-push to shared branches, no amending or rebasing commits once pushed, no deleting published releases or tags. Correct a mistake with a new commit, a new release or a new entry that says what was wrong. Rewriting is fine for work nobody else has seen yet. Reasoning: other people, machines and archives hold copies of published history; rewriting it breaks their copies, can silently drop their work, and makes the record of what happened unreliable. A correction that sits on top of the mistake keeps both visible, which is also how this ledger handles its own errors. Protect the main branch against force-pushes so the rule does not depend on memory. Considered and rejected: force-pushing after 'checking nobody pulled' (the agent cannot see every copy, including mirrors and archives); rewriting to remove an exposed secret as the only fix (the secret is already out and must be replaced; rewriting alone gives false comfort).
- id
- ebc1007ae2104f59f4099944c9a2c505852b206e5af427be1555ef1e4e16e501
- prev
- 6458d12d0ac65b57f59ed2d4cd1a25c40728c098a6dbaababe7ef2f609d707b5
- signed
- by ed25519:0f2cd708dc0274a8b82fbac5fccd948aa2dd895bfaa44d37539091eb0ad3129c
- time
- 2026-11-18T20:04:46Z