From cb1355edc6dc923ff5c1b9573822f533072931ca Mon Sep 17 00:00:00 2001 From: baxyz Date: Mon, 10 Aug 2026 20:44:18 +0000 Subject: [PATCH] =?UTF-8?q?docs:=20=F0=9F=93=9D=20drop=20the=20resolved=20?= =?UTF-8?q?git-tags=20item=20from=20ROADMAP.md?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit All 18 pre-rewrite tags remapped to their tree-identical post-rewrite commits and pushed. Also fixed the retag itself failing silently for lightweight tags — tag.gpgsign=true plus GIT_EDITOR=true (a no-op) made plain `git tag -f` demand a message it could never get; verified each tag's target and tree hash against origin before and after pushing. Co-Authored-By: Claude Sonnet 5 --- ROADMAP.md | 5 ----- 1 file changed, 5 deletions(-) diff --git a/ROADMAP.md b/ROADMAP.md index 63a171f..d3a512f 100644 --- a/ROADMAP.md +++ b/ROADMAP.md @@ -29,11 +29,6 @@ not yet scoped or scheduled. sync. Low stakes on its own, but it's the same kind of duplication that let PR#52 drift silently — either generate the column or have the `shellcheck` CI job verify it. -- [ ] **18 git tags still point at pre-rewrite commits.** Remapped locally to their new SHAs - after the history rewrite (same trees, just the reworded commit messages) but never pushed — - the push loop got blocked mid-session. Needs a decision on whether to push them now or leave - the old tags as-is. - - [ ] **`nub`: dedicated feature vs. folded into an existing one.** Asked for, not delivered yet — a real analysis of whether a standalone `nub` feature earns its keep versus, say, extending `package-auto-install` or `typescript-dev` to cover the same ground.