From d05384d69689ddf54ef5cd183a35cb6713dc69b9 Mon Sep 17 00:00:00 2001 From: jochen Date: Fri, 20 Nov 2026 19:25:01 +0100 Subject: [PATCH] Design 47 H: H-A-180, a hold for a failed restore point whose cause is gone --- 03-DESIGN/01-to-be/47-milestones/H-lead-answers.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/03-DESIGN/01-to-be/47-milestones/H-lead-answers.md b/03-DESIGN/01-to-be/47-milestones/H-lead-answers.md index e9c1b458..bf8c6ff5 100644 --- a/03-DESIGN/01-to-be/47-milestones/H-lead-answers.md +++ b/03-DESIGN/01-to-be/47-milestones/H-lead-answers.md @@ -944,3 +944,10 @@ Where two preparers disagreed, the chosen one is named. ended could not run ("no node places mesh-delivery-sdk on the branch main") and the pipeline can never pass. A step with nothing to judge ends `nothing to run`, which counts as passed (ADR 0682 §8): the composition step of a mover that no machine holds on the branch ends so, saying why, as the levels of an archive artifact do. P3, P7. +- **H-A-180 (a hold for a failed restore point whose cause is gone, 19:25)** Met after pack76 (19:20): a module + held "for a failed backup" keeps that hold across the engine's start and waits for the operator's answer (back up + again or skip, ADR 0568 §4), and no ask was ever raised for it, so the hold is for good. The engine judges a held + module's cause again at its start and whenever the placements change: where the hold's cause was a recovery seat + with no holder placed, and none is placed now, it moves the module as H-A-177 says (without a point, the finding + raised), no operator's answer needed; where a holder is placed, the point is asked again once; a hold whose point + truly failed (a holder answered and failed) raises the operator's ask it was owed. P3, P7, ADR 0568 §4.