Design 47 H2: H-A-445, how a declined step reaches the runner and is placed again

This commit is contained in:
2026-11-22 01:56:29 +01:00
parent e67a8f1bae
commit 1324d2183d
@@ -614,3 +614,12 @@ The lead's one-line answers to the questions H2's develop and review raise (impl
not-needed as ended with that outcome (its record's acts, gate_record, delivery_status), never as a refusal: a pull
request closed unmerged is a fact of the forge, not a fault of the act. The scopes value form {token, scopes} is the
gitea module's own and stands. P-failure-loud (a refusal names a fault), P3.
- **H-A-445 (how a declined step reaches the runner and is placed again, 02:04)** H-A-375. (1) The build-step part's
StepEnded gains the outcome `declined` with `needs` (the seats the package needs) and `lacks` (what the node lacked),
said by the holder's step run in place of could-not-run "this node lacks X" (ADR 0552 §2's form), so the runner hears
it as every end. (2) The step asks ride one shared queue, so the runner puts a declined step again under a new ask id
with `not-on`, the nodes that declined it (a holder on a listed node hands it back at once, unacknowledged, and the
queue delivers it on); each node declines at most once; the nodes that have the need are the holders of the seat the
need names (the seat-holders view) less those that declined; when none is left the step ends could-not-run naming the
need and the nodes asked. §777's same-id rule holds for a retry of one ask, not for a new ask after a decline. P3,
ADR 0552 §2, §4.