Skip to content

closure_drift 1.1.0

Choose a tag to compare

@luizfnsilva luizfnsilva released this 09 Oct 22:50
· 218 commits to main since this release

What 1.1.0 had to meet was written first (tests/PREREGISTRATION.md §16). A reviewer who did not write it found 13 problems in the candidate; they are in docs/REVIEW-1.1.0.md as delivered, and each fixed one has a proof that was red before the fix. DOI: 10.5281/zenodo.23271569

  • --published FILE: a local list of the versions you released, one per line. Only their tags are publication points; a tag left off is left out of the analysis, not shown unpublished. Answers the first failure reported from outside (ruff discussion 29122).
  • Version sources: Chart.yaml, gemspec, go.mod (a Go module's version is its whole tag).
  • The refusal for want of a label comes before the scan.
  • Opt-in: --label-equality version, --modes.
  • Changed only under --at commits: the repository's tags vote on the version file.

199 proofs, 56 mutants, 315 adversarial cases, 17 properties against an independent oracle, on Linux (Python 3.9–3.14), macOS and Windows (tests/RECORD.md). The 100 projects of the regression corpus give the recorded answer, as predicted before the run.

Try it in 30 seconds: discussion #9.