bootstrap.py (the assembly root's): a leg moved off a branch has an ignored file the pin tracks overwritten and a merge in progress ended, with no word, and a tracking branch ahead of the pin with no commit of its own is left where it is #709

make bootstrap should not overwrite an ignored file at a path the pin tracks, or end a merge in progress, without saying so, and it should place a tracking branch that is ahead of the pin with no commit of its own where the pin is (or AGENTS-shape.md should stop saying it does). The riders in the same placement code ride with it.

From lane 4's checklist triage of 2026-10-16 (T1), approved by Brett Heap at 2026-10-16T13:36:07Z ("approve all triage recommendations"). One pull request covers the lines below, together with #710.

Parent: #458. Milestone: v1.1.1 (a follow-up filed after Brett Heap's v1.1.1 word of 2026-10-16T13:35:28Z).

Authored after the v1.1.0 work; lands after the v1.1.0 tag unless the release owner (lane 3) says otherwise.

Lands after: lane 2's #634 (ready; it edits templates/assembly-root/AGENTS-shape.md) and lane 5's #694 (filed, no pull request yet; it edits the same templates/assembly-root/scripts/bootstrap.py, at recorded_gitlink, _try_git and the git run). Whichever of these lands second re-gates against the first.

Effort: M: --no-overwrite-ignore on three checkouts and a merge, one operation-in-progress refusal (which may need a ruling, see below), one re-pin. Platform: all (L4-79 is about case-insensitive disks only).

Files: templates/assembly-root/scripts/bootstrap.py and templates/assembly-root/AGENTS-shape.md, both pinned shape copies. The pull request re-pins them once, and re-pins bootstrap.py once for #709 and #710 together, so each project's v1.1.1 update takes one upstream-changed row per file.

Lane openRepoShape-4 claims it at filing; see its CLAIMED comment.

Why this is filed. Two of the lines lose a person's work, and a real user can meet them:

• L4-74. An ignored file at a path the pin tracks is overwritten, silently, by every placement that changes files.
• L4-75. A merge in progress is ended by the move off its branch.

The rest is exotic input: a tracking branch whose tip is its upstream's own tip, a symbolic HEAD outside refs/heads/, and a case-only branch name on a case-insensitive disk.

Line numbers are at c6122e8 (#639); main has moved since.

Lines

• L4-74 (P2: work is lost): every bootstrap placement that changes files OVERWRITES an ignored file at a path the pin tracks, silently: the on-branch merge --ff-only (bump-leg.py, then make bootstrap, leaves the leg detached: bootstrap never fast-forwards a local tracking branch that is behind the new pin, though AGENTS.md, README and bump-leg.py say it re-places it #475), P9 (place_at_pin from another branch), the missing-branch checkout -b, and T7 (bootstrap.py and AGENTS-shape.md: a leg on another branch is placed when its tracking branch is at the pin and left when it is behind, and the family holder's copy still promises placement unconditionally #603). R603 measured it on main and on bootstrap.py: a leg on another branch whose tracking branch is behind the pin is fast-forwarded and checked out as one at the pin already is, and the family holder's AGENTS-shape.md and bump-leg.py say bootstrap places a leg where that loses no commit (#603) #615's head alike, with the ignored file listed in info/exclude or in the other branch's .gitignore. Where: templates/assembly-root/scripts/bootstrap.py:454-455 names the overwrite and the file has no --no-overwrite-ignore. Fix: --no-overwrite-ignore on those checkouts and on the merge; git 2.43 has the flag for both. (source)
• L4-75 (P2: work is lost): moving a leg off a branch with a merge in progress (merge --no-commit) ENDS the merge: MERGE_HEAD is removed and the staged result is carried onto the tracking branch. This happens in P9, on the missing-branch path and in T7. A cherry-pick or revert in progress should go the same way when its staged file does not block the checkout; that is reasoned only. Where: bootstrap.py:286 and :454-456. Fix: refuse to move off a branch with an operation in progress, and say so. (source)

Rides along (backlog; tick it here if the same change covers it)

• L4-76: P9's line (on <branch> at <pin> (branch tip == pin)) does not name the branch HEAD left. T7's does, after bootstrap.py: a leg on another branch whose tracking branch is behind the pin is fast-forwarded and checked out as one at the pin already is, and the family holder's AGENTS-shape.md and bump-leg.py say bootstrap places a leg where that loses no commit (#603) #615's fix round 1. Where: bootstrap.py:414 prints that line and no branch left. (source)
• L4-77: a tracking branch AHEAD of the pin with no commit of its own (its tip is its upstream's remote-tracking ref) is left with not moving an existing branch whenever HEAD is on the branch, on another branch, or on a branch with no commit yet. move_back_to_pin starts only from HEAD detached at the pin. Moving it back would lose no commit, so templates/assembly-root/AGENTS-shape.md:95-96 ("places each leg ... where that loses nothing") is false of these states. Where: bootstrap.py:502; AGENTS-shape.md:95-96 is unchanged. Either the placement moves it back or the sentence says what it leaves. (source)
• L4-78: HEAD as a symbolic ref outside refs/heads/: P9 places the leg, while T7 prints which is neither. This is exotic, and byte-identical to main. Where: bootstrap.py:468-469 (P9) and :380 (which is neither). (source)
• L4-79 (DOUBTFUL): reasoned only, for case-insensitive disks: the current != refs/heads/<branch> compare is case-sensitive. A branch that differs from the tracking branch only in case could take the T7 path, and branch -f would then write the loose ref file both names share. Where: bootstrap.py:469 compares refs/heads/<branch> exactly (read). What branch -f then does on a case-insensitive disk needs a macOS or Windows run, and CI is Linux only since ci: run the test suite on Linux only #597, so the pull request may close it with a note and no change. (source)

What is expected

• No placement overwrites an ignored file at a path the pin tracks. It refuses by name, or leaves the file and says so.
• A merge in progress is not ended by a move off its branch. The placement refuses and says what to finish first. This may need a ruling from Brett Heap: whether it is a refusal (exit 2) or a NOTE that goes on. It is a new refusal in a path that moves files today.
• P9's line names the branch HEAD left, as T7's does.
• A tracking branch ahead of the pin with no commit of its own is moved back, or AGENTS-shape.md:95-96 is reworded to say which states it leaves. The pull request picks one and says which.
• Tests for each: an ignored file in info/exclude, an ignored file in the other branch's .gitignore, a merge --no-commit, a branch whose tip is its upstream's.

Not a duplicate: lane 5's #648 and #682 are the FAMILY root's bootstrap.py (L4-80 went to them), and #694 is a missing git. No issue numbered #650 to #700 cites L4-74 to L4-79 (read at 13:17Z).

Each line above is a line of #458's checklist comment (issuecomment-6201724792). Tick it there too when this closes.

Lane: openreposhape-4 (openRepoShape-4)