Design 47 H: H-A-333 and H-A-334, journal lines at every said end; H2's part answers

This commit is contained in:
2026-11-21 21:55:48 +01:00
parent 04fb76dd53
commit de9f20c0e1
@@ -2007,3 +2007,17 @@ Where two preparers disagreed, the chosen one is named.
mesh/check, and merge-pull merges its pull requests like any other; until built, hq's pull requests are merged by
the operator at the forge. Owed, the next milestone, in the indexer (the registration's source-only form), the gate
and the checker. merges-need-no-approval, P7.
- **H-A-333 (an ask placed is a journal line too; every said end writes its line, 21:55)** Amends H-A-320 (1): the
runner writes a line when it places an ask ("<pipeline> <mover> <step> asked on <seat>") as well as when a step
ends, and the ended line is written wherever a step-ended event is said (the unanswered-bound end, the end at
placing, a supersede, a re-pin, the approval's end, the start pass, the verbs that end a step), in the one place
that says the event; the builder writes its answered line in the one place an ask is answered (BuildStep's finish),
naming the pipeline, the movers, the step and the outcome, the per-entry "existing build / built now" line beside
it. P-failure-loud.
- **H-A-334 (H2's part: two answers, 21:55)** For design 47 H2's part: (1) hq as a watched source is a design line, not an
operator question: the operator wants no merge put to them (21:48) and hq's checks were meant to be the mesh's (ADR
0527), as H-A-332 says; (2) the destroy-level answers' channel stays the approve-level bend of H-A-237 until a
research effort designs the destroy proof's channel (an operator question only when that research is proposed);
(3) the error kind "away" widening ADR 0664 §1 and ADR 0668 §1 is a progressive insight on ADR 0668 proposed by the
record writer; (4) the items of section 10 rows 41 to 102 marked H2 go into a second part, H2b, written next; design
47's main file's size is a separate reshaping run. act-on-the-intent, P7.