cli: duplicate #187 commits stacked on test/argv-matrix-185 #219

Closed
opened 2026-09-08 02:28:11 +09:00 by natsukium · 1 comment
Owner

The #187 work (cli: publish the finite option vocabularies through clap and cli: snapshot every visible help page) exists a second time as unlabeled commits stacked on test/argv-matrix-185 (worktree issue-185).

Evidence: git branch --contains 645708f5 reports test/argv-matrix-185 and fix/schema-finite-strings-178; 645708f5 and its parent 915e5284 are a second copy of the same change based on 105b0899 (git merge-base 645708f5 origin/main = 105b0899), i.e. the #187 work is stacked onto the #185 branch rather than onto a branch of its own. git -C .claude/worktrees/issue-185 status --short additionally shows M crates/felis-cli/src/tests.rs with 77 uncommitted insertions — the same file both #185 and #187 touch.

Failure mechanism: a PR opened from test/argv-matrix-185 carries the #187 change as unrelated commits. Those commits are now on main (d7b61a1b, 1f8fa55c), so a rebase onto origin/main will either drop them or conflict on cli_output.rs / cli_sessions.rs / tests.rs, and the uncommitted 77 lines will be silently omitted from that PR.

Correction: rebase test/argv-matrix-185 onto the current origin/main so the two duplicate #187 commits fall away as already-applied, resolve or commit the pending tests.rs hunk deliberately, and verify the remaining diff contains only the #185 argv-matrix work before opening its PR.

Deferred from #187 because it concerns another branch's hygiene, not #187's acceptance criteria.

Split from #187.

The #187 work (`cli: publish the finite option vocabularies through clap` and `cli: snapshot every visible help page`) exists a second time as unlabeled commits stacked on `test/argv-matrix-185` (worktree `issue-185`). Evidence: `git branch --contains 645708f5` reports `test/argv-matrix-185` and `fix/schema-finite-strings-178`; 645708f5 and its parent 915e5284 are a second copy of the same change based on 105b0899 (`git merge-base 645708f5 origin/main` = 105b0899), i.e. the #187 work is stacked onto the #185 branch rather than onto a branch of its own. `git -C .claude/worktrees/issue-185 status --short` additionally shows ` M crates/felis-cli/src/tests.rs` with 77 uncommitted insertions — the same file both #185 and #187 touch. Failure mechanism: a PR opened from `test/argv-matrix-185` carries the #187 change as unrelated commits. Those commits are now on `main` (d7b61a1b, 1f8fa55c), so a rebase onto `origin/main` will either drop them or conflict on `cli_output.rs` / `cli_sessions.rs` / `tests.rs`, and the uncommitted 77 lines will be silently omitted from that PR. Correction: rebase `test/argv-matrix-185` onto the current `origin/main` so the two duplicate #187 commits fall away as already-applied, resolve or commit the pending `tests.rs` hunk deliberately, and verify the remaining diff contains only the #185 argv-matrix work before opening its PR. Deferred from #187 because it concerns another branch's hygiene, not #187's acceptance criteria. Split from #187.
Author
Owner

Closing: the duplicate commits were an artifact of a pipeline resume that replayed the already-merged #187 chain in a fresh worktree stacked on a stale branch, not a state of main or of any surviving branch. Those worktrees and local branches are deleted; main carries #187 exactly once (915e5284, 645708f5 via PR 198).

Closing: the duplicate commits were an artifact of a pipeline resume that replayed the already-merged #187 chain in a fresh worktree stacked on a stale branch, not a state of `main` or of any surviving branch. Those worktrees and local branches are deleted; `main` carries #187 exactly once (915e5284, 645708f5 via PR 198).
Sign in to join this conversation.
No description provided.