From 6f26f97fdbf5c5a677a815a10b6044be79832520 Mon Sep 17 00:00:00 2001 From: jochen Date: Thu, 1 Oct 2026 22:25:31 +0200 Subject: [PATCH] Issue 189: the plan's half is built; the moved word stays for a decision --- .../00-report.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/04-ISSUES/189-a-rebuild-from-the-same-commit-is-not-a-move/00-report.md b/04-ISSUES/189-a-rebuild-from-the-same-commit-is-not-a-move/00-report.md index f2537ce..868a3d9 100644 --- a/04-ISSUES/189-a-rebuild-from-the-same-commit-is-not-a-move/00-report.md +++ b/04-ISSUES/189-a-rebuild-from-the-same-commit-is-not-a-move/00-report.md @@ -33,3 +33,11 @@ the artifact. *How this would be checked:* a controller test registering a build of an unchanged commit with a new artifact digest for a module whose policy rolls out: the machines are sent; `status` shows no machine behind afterwards. + +## Built in part, 2026-10-01 + +mesh-controller 203 and 204: a plan sends the machines of every module in a built tier whose policy +rolls out, once, moved commit or not, and waits for the ones a later tier is built by. After a merge +nothing is left for a hand to push, except what a *record* policy leaves by design. What remains for +a decision is the catalogue's own word: *moved* should follow the artifact, so a module rebuilt +outside any plan — by `build` by hand — rolls out the same way. -- 2.54.0