From 01842bbeee48e136b0581f46548ac575634494f6 Mon Sep 17 00:00:00 2001 From: jochen Date: Fri, 20 Nov 2026 16:55:48 +0100 Subject: [PATCH] Design 47 H: H-A-154, the catalog revision names its base --- 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 16797bc3..3c8e603d 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 @@ -761,3 +761,10 @@ Where two preparers disagreed, the chosen one is named. names of the assignments left out, written when the run starts and read by every step's judgement and by the resume, in the seat mesh-wants' definition and its generated contracts; a change of the seat's definition is a group with the assigner, one check. State the judgement needs lives on the bus, never only in the step's memory. P1, P7. +- **H-A-154 (the catalog revision names its base, 16:55)** Met in the build of H-A-151: the assigner's validate + reads the base through the seat mesh-catalog's verb `revision`, and no field of the revision or of the verb's input + names "the newest kept revision below this one that the take-in pinned"; the indexer's pins are its own state, + which no other module reads (P12). The take-in writes the base on the catalog revision it writes: a field `base`, + the number of the newest kept revision below it that it pinned, absent on the branch's first take-in. Validate's + before case reads that field from the revision it already asks for, so the base the pipeline keeps and the base + validate reads are one and the same, written in one write (P2, P7, P1). No new verb and no second ask.