gitea: the package seats take a publish with a digest and no parts as tag-or-refuse (A-490, A-490a) #707

Merged
mesh-tools merged 5 commits from fix/gitea-publish-by-digest into main 2026-11-19 21:40:03 +00:00
5 Commits
Author SHA1 Message Date
jschoubben ac2cd3749c Registry holders: the distribution holder keeps trunk's check; the package holders run the kit's held-under-another-digest case on both seats (A-490j, A-490a)
Builds: design 47 G A-490j, A-490a

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 21:39:35 +01:00
jschoubben 4bdcedd0f0 Registry holders take a one-part publish made while the build runs: a reserved artifact with no digest yet accepts it (A-490j)
Builds: design 47 G A-490j, §381, A-415

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 21:34:20 +01:00
jschoubben f85199f624 gitea: the not-stored refusal for another digest applies to the no-parts publish only; the one-part publish keeps trunk's refusal
Builds: design 47 G A-490, A-490a

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 17:09:16 +01:00
jschoubben 13d43a7de0 gitea: a no-parts publish of a version held under another digest is refused as not stored, named by its digest; conformance runs on both package seats
The builder's republish recognises a refusal by the words not stored.
Builds: design 47 G A-490, A-490a

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 17:04:55 +01:00
jschoubben 5dfb357333 gitea: the package seats take a publish with a digest and no parts (held says published, else refused not stored)
The builder's publish again sends version and digest and no parts; the package seats refused it as not one part.
Both registry holders now run the publish conformance cases.

Builds: design 47 G A-73, A-417, A-490

Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
2026-11-19 16:51:10 +01:00