Repository navigation
origin/master is a single orphan commit with no ancestor in common with the previous 2545-commit history (old history is safe) #2328
Description
origin/master currently consists of one commit, and it has no ancestor in common with the previous master. Nothing is lost — the old history is safe on another remote branch — but the state is worth a decision rather than a surprise.
What the repository looks like now
| commit | commits in history | timestamp | |
|---|---|---|---|
| master, before | ca4234e20 — fix(ci): repair never-green Coq kernel build gate (#2321) |
2545 | 2026-08-21 18:00:32 +0700 |
| master, now | fead099c2 — fix(coq): PhiFloat.v — bare B2R/B754_finite resolved to BinarySingleN |
1, and it is also the root | 2026-08-21 18:48:37 +0700 |
git merge-base origin/master <any older branch> exits 1: the histories share no commit. This is a fresh orphan root, not a rewrite — no rebase or filter produces a one-commit history whose root is its own tip.
Nothing is lost
- The previous 2545-commit history is reachable from
origin/fix/coq-phifloat-binary64-name-collision(2547 commits, containsca4234e20). - Working branches are unaffected:
claude/igla-fpga-improvements-3f5e1ais intact at 2556 commits, 0/0 against its ownoriginref. - The research repository
gHashTag/trinity-fpgais a different remote and is untouched.
Recovery, if that is what you want
Deliberately not performed automatically — restoring a shared branch is irreversible for anyone who has already fetched, and it discards the new commit unless preserved first:
git fetch origin
git branch orphan-master-fead099c2 origin/master
git push origin ca4234e20:refs/heads/master --force-with-lease--force-with-lease rather than --force, so that if master has moved again since this issue was filed the push fails instead of repeating the event.
How it was noticed, and what changed here
git status reported the divergence as 126 / 117 one wave and 2556 / 1 the next — as a number, merely more divergence. The categorical fact is that git merge-base stopped returning anything, so tri audit now gates on merge-base rather than on ahead/behind counts, and prints a distinct failure for "histories are unrelated".
The evidence is written to docs/reports/MASTER-ORPHANED-W971.md because the only local proof of the previous tip was git reflog show origin/master, which expires.
Filed by the autonomous wave loop — informational; nothing was modified on master.
Activity
specs/automation: inngest-probe-suite.t27 -- safe probe of every serv…
specs/automation: inngest-probe-suite.t27 -- safe probe of every serv…
gHashTag commentedon Sep 24, 2026
A boundary drafted from the paths this issue already names, so a bee can take it. Nothing is reserved until this is in the issue body - the Queen reads the body, not comments. Check the paths first: a wrong boundary reserves files the work does not own.
Boundary
docs/reports/MASTER-ORPHANED-W971.mdtri-claw commentedon Oct 9, 2026
The Queen cannot take this issue yet. A bee starts only on an issue whose body has every section below, and this one lacks:
- specs/x.t27); without it the Queen reserves nothing and no bee can take the issueFR-001: ... MUST ...exit 0, or a backticked commandShe tried to write them 2 times and could not (last: not a task: This issue is an informational report about the repository state (origin/master being an orphan commit) with recovery instructions provided, but it explicitly states "nothing was modified on master" a). Add them and the next round takes it. The rule is
specs/policy/issue_shape.t27.gHashTag commentedon Oct 17, 2026
Closing as resolved. The state this issue describes no longer holds on
origin/master. Checked 2026-10-17:fead099c2(the commit this issue calls a one-commit orphan root) now has parentca4234e20, the previous master tip:git log -1 --format=%p fead099c2printsca4234e20.git merge-base --is-ancestor ca4234e20 origin/masterexits 0, andgit merge-base --is-ancestor fead099c2 origin/masterexits 0.git rev-list --count origin/master= 6288, so the 2545-commit history is part of master again, under the 2026-04-04 root9cb1f9b7e.No recovery push is needed. (
ea15cd54c, a second root dated 2026-07-05, comes from a merged side history and is unrelated to this event.)PR #3556 also cross-references this issue. Reopen if the orphan state comes back.