Skip to content

release 0.6.0: version bump and changelog cut - #66

Merged
danielplohmann merged 1 commit into
mainfrom
release/0.6.0
Sep 15, 2026
Merged

danielplohmann merged 1 commit into
mainfrom
release/0.6.0

Conversation

@danielplohmann

Copy link
Copy Markdown
Owner

Cuts 0.6.0. Both version strings, ## [Unreleased] renamed to ## [0.6.0] - 2026-09-15 with a fresh empty one above it, and the compare links moved on. No source changes.

This is the first release through publish-release.yml — the workflow has existed since #59 but no tag has ever exercised it.

What is in it

23 entries from five merged PRs plus the release machinery and three dependency bumps:

#60 S_INLINESITE2, separated code, stripped PDBs, MSF block sizes to 32768
#61 single-pass record walk — 3-8x on every listing, output byte-identical
#62 PDB.open maps the file; inline sites placed without holding a module's worth
#65 C7/C11 module-stream signatures stripped; an unknown one no longer parsed as a record
#53 parse-boundary, C13 diagnostics and DBI/GSI error handling

Why minor rather than patch

The Fixed entry for the fused ChangeCodeLengthAndCodeOffset annotation says This moves addresses: the second and later ranges of an inline site built from that opcode now start earlier than 0.5.0 reported, on every rust-lld and clang output. The first range, and every single-range site, are unchanged.

CHANGELOG.md's preamble reserves exactly this case — "A release that made a previously-resolved address resolve differently would be breaking, and would say so here" — so it is called out in the entry rather than left for a reader to infer.

Verified before tagging

  • release_guard.py --tag v0.6.0 extracts the section and reports prerelease=false; --tag v0.7.0 is refused naming both version strings
  • 645 passed, ruff check and ty check clean on the new pins (ruff 0.16.6, ty 0.0.80)
  • python -m build and twine check --strict pass on sdist and wheel
  • the wheel installed into a clean venv and imported with python -P reports 0.6.0 from site-packages with py.typed present, and the purepdb entry point runs; metadata reads Requires-Python: >=3.11, License-Expression: BSD-3-Clause

After merge

Tag v0.6.0 on the merge commit. The push starts the real run, which parks at the pypi environment's required reviewer; a manual run from the same tag rehearses to TestPyPI and creates the project there, which is what converts the pending publisher. Approve after the rehearsal looks right.

🤖 Generated with Claude Code

Both version strings to 0.6.0, `## [Unreleased]` renamed to
`## [0.6.0] - 2026-09-15` with a fresh empty one above it, and the compare
links moved on.

23 entries: the corpus-audit fixes (#60), the single-pass record walk (#61),
the mapped `PDB.open` and streamed inline-site placement (#62), the C7/C11
signature handling (#65) and the parse-boundary hardening (#53), plus the
release workflow itself and three dependency bumps.

Minor rather than patch, and deliberately so: the `Fixed` entry for the fused
`ChangeCodeLengthAndCodeOffset` annotation says **This moves addresses** -- the
second and later ranges of an inline site built from that opcode now start
earlier than 0.5.0 reported, on every rust-lld and clang output. `CHANGELOG.md`'s
preamble reserves that for a breaking change and requires it to be stated, which
that entry does.

Verified before tagging: `release_guard.py --tag v0.6.0` extracts the section
and reports prerelease=false, and refuses v0.7.0 naming both version strings;
645 passed, ruff and ty clean; `python -m build` plus `twine check --strict`
pass on both artefacts; and the wheel, installed into a clean venv and imported
with `python -P`, reports 0.6.0 from site-packages with `py.typed` present and
runs the `purepdb` entry point.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
@danielplohmann
danielplohmann merged commit cf1efdb into main Sep 15, 2026
11 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant