diff options
| author | Pjotr <wrk@bosa.lan> | 2026-11-07 17:17:19 +0000 |
|---|---|---|
| committer | Pjotr <wrk@bosa.lan> | 2026-11-07 17:17:19 +0000 |
| commit | 2307d5e2db5ce8efae19a647bd48b4311be34607 (patch) | |
| tree | aefc0a48a38cd69fd0f38f8c163ad77ad8d8317a /test/python/test_groups.py | |
| parent | 9fa03035b9f21352c1c8b93681cc1bc1ce4cf6ff (diff) | |
| download | genenetwork4-2307d5e2db5ce8efae19a647bd48b4311be34607.tar.gz | |
version: document the provenance fields, and assert they agree
* README.md (/version): the example carries dirty/source_digest/branch, and the prose says why a stamp can lag a commit -- a binary built from a dirty tree must not claim that commit -- and that build.rs re-stamps on a commit because it watches the ref, not just .git/HEAD. * docs/www/ai-instruction.md: comparing against a clone means comparing the part before "-dirty", with source_digest for the exact build. * tests/version.rs: `dirty` mirrors the suffix, branch is non-empty, and source_digest is 16 hex digits or unknown.
Diffstat (limited to 'test/python/test_groups.py')
0 files changed, 0 insertions, 0 deletions
