From 0787704bbaba57173129de3b1ae35c870f04edeb Mon Sep 17 00:00:00 2001 From: jochen Date: Fri, 20 Nov 2026 15:37:55 +0100 Subject: [PATCH] Design 47 H: H-A-145, a kept stream is brought to the definition's source filter in place --- 03-DESIGN/01-to-be/47-milestones/H-lead-answers.md | 8 ++++++++ 1 file changed, 8 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 31f829b4..aa46d39a 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 @@ -697,3 +697,11 @@ Where two preparers disagreed, the chosen one is named. client happens to follow. Only before the first registrations are held do the gate's source's branches stand in, in the order the source gives them (§532). An unlisted module is said on every path of the judgement, whether the machine composes, fails or could not run. P3. +- **H-A-145 (a kept stream whose source filter the definition changed, 15:42)** Met live as the last link's third fault: + the kept stream of the runner's `pipeline-ended` was made at G's definition with the source filter of a three-part + subject; H1 moved the event to `.` (§746), the runner says two parts, the stream took nothing in + since 08:13, and the assigner, which hears a kept event from its kept stream, heard no end all day while the indexer, + on the events stream itself, heard every one. The grants part brings a held kept stream to the definition's source + filter in place, keeping its messages, and says so in its journal (held and composed filter); it never makes the + stream again for that, so nothing kept is lost; the same for `pipeline-started` and `wait-held`. A consumer's filters + it already adds in place. P3, P7.