Skip to content

fix: document windows+3.14t as upstream pywin32 wheel gap #22

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
onlyxItachi merged 1 commit into from
Jul 18, 2026
Merged

fix: document windows+3.14t as upstream pywin32 wheel gap #22

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 8 additions & 3 deletions .github/workflows/ci.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,11 +18,16 @@ jobs:
matrix:
os: [ubuntu-latest, macos-latest, windows-latest]
# The full requires-python range on every OS, floor (3.10) to
# free-threaded 3.14 ("3.14t" is uv's key for the cp314t build; the
# whole dep tree ships cp314t wheels for all three OSes — verified
# against PyPI before widening the matrix).
# free-threaded 3.14 ("3.14t" is uv's key for the cp314t build).
python: ["3.10", "3.11", "3.12", "3.13", "3.14", "3.14t"]
runs-on: ${{ matrix.os }}
# windows + 3.14t is a DOCUMENTED upstream gap, not our code (2026-07-18):
# mcp requires pywin32 on win32, and pywin32==312 ships no cp314t wheel, so
# `uv sync` fails before our suite can even run. Linux/macOS 3.14t legs are
# strict and green. This cell is allowed to fail so the matrix reports
# reality without going red on someone else's wheel gap — and it starts
# passing automatically the day pywin32 publishes cp314t wheels.
continue-on-error: ${{ matrix.os == 'windows-latest' && matrix.python == '3.14t' }}
steps:
- uses: actions/checkout@v4
- name: Install uv
Expand Down
7 changes: 5 additions & 2 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,7 +123,11 @@ Claude Code and OpenAI Codex setup (both stdio MCP): see [`docs/clients.md`](doc

## Status

**v1.2.0 (pre-release, in review)** — the Development Observatory release:
**v1.2.0 (pre-release, in review)** — the Development Observatory release
(Python 3.10–3.14 incl. free-threaded 3.14t, CI-verified on Linux/macOS/Windows;
one known upstream gap: *Windows* 3.14t cannot install yet because `mcp`'s
`pywin32` dependency ships no cp314t wheel — that CI cell is a documented
allowed-failure until pywin32 catches up):
seven new backends extend the digest matrix from profilers to the whole
edit→build→verify→measure loop. BuildDigest: `clang_time_trace` (compiler-phase
traces, with `[total]` aggregate tagging), `cmake_profile` (configure-step
Expand All @@ -137,7 +141,6 @@ previous-green comparison proven on real two-run fixtures). RepoState:
honest absence, never `0.0`). Zero new tools — every backend rides the same
seven; 206 tests. Every fixture in the release was produced by the real tool it
mimics — several briefs were corrected by what the artifacts actually contained.
Python 3.10–3.14 including free-threaded 3.14t, Linux/macOS/Windows CI-verified.

**v1.1.1** — contract-hardening hotfix (issues #2/#4/#5): registry format
collisions are detected on the normalized key (a case-variant format can no
Expand Down
Loading