From 325bf58483aff32976180a9d13c1f298c8460d3e Mon Sep 17 00:00:00 2001 From: jochen Date: Thu, 19 Nov 2026 16:44:24 +0100 Subject: [PATCH] Answer A-490: a package holder takes the no-parts publish as tag-or-refuse and never reads the artifact store Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy --- 03-DESIGN/01-to-be/47-milestones/G-22-lead-answers-2.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/03-DESIGN/01-to-be/47-milestones/G-22-lead-answers-2.md b/03-DESIGN/01-to-be/47-milestones/G-22-lead-answers-2.md index 362a9fb5..6cec0c63 100644 --- a/03-DESIGN/01-to-be/47-milestones/G-22-lead-answers-2.md +++ b/03-DESIGN/01-to-be/47-milestones/G-22-lead-answers-2.md @@ -2048,3 +2048,10 @@ the principle that settles it, cited as "design 47 G A-n". conformance kit has a case that refuses a puller which does not read the runs that started from its module's own `steps` state, every build module's test runs it, and the catalog's failure-kit check refuses a build module whose tests do not. (From review round 3 of the build-node step-ends fix.) +- **A-490** (A-73, A-417; principle 7) A package holder (`go-package-registry`, `npm-package-registry`) takes the + no-parts publish as tag-or-refuse: when the forge holds the package's name at the version with that digest it answers + published and moves nothing, else it refuses `not stored`, naming the digest. It never reads the artifact store: a + package's bytes reach the registry only through the build module's one-part publish while it builds, and no copy of + a package is kept anywhere else. A package whose build's publish was refused is built again + (`builder_build_again`), never published from a copy. (Live 2026-11-19: the sdk's build publish refused at 12:25 by + the retired `callers` read; the builder's no-parts republish refused for the parts form.)