diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index d55fb10..b9d8317 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -92,9 +92,20 @@ and has not been verified, and against which libei versions, is in ## Releasing -Versions are SemVer and live in two places — `pyproject.toml` and -`src/libei/__init__.py` — which have to agree with each other and with the -tag. Nothing enforces that yet. +Versions are SemVer and live in four places. Each one is checked rather than +trusted, so a bump that misses one fails somewhere instead of shipping: + +- `pyproject.toml` — what pip and the built artifacts use. +- `src/libei/__init__.py` — what a caller introspects, held to + `pyproject.toml` by + `tests/test_documentation_shape.py::test_version_is_declared_identically_in_both_places`. +- `README.md` and `docs/developers/verification.md` — both state it in prose, + and `test_status_pages_name_the_current_version` fails if either stops + naming the current one. +- **The release tag.** CI's `build` job reads the version back out of the + built sdist — not out of the source — and fails the tag push when the two + disagree, so a PyPI number is never spent on a tag describing a different + tree. A release is an annotated, `v`-prefixed tag. Pushing it is the whole of it; PyPI is the only place a release is published, and no GitHub Release is cut: diff --git a/README.md b/README.md index fa2b378..8bfa263 100644 --- a/README.md +++ b/README.md @@ -1,5 +1,6 @@ # python-libei +[![Contributing](https://img.shields.io/badge/contributions-welcome-brightgreen.svg)](https://github.com/ctrondlp/python-libei/blob/main/CONTRIBUTING.md) [![CI](https://github.com/ctrondlp/python-libei/actions/workflows/ci.yml/badge.svg?branch=main)](https://github.com/ctrondlp/python-libei/actions/workflows/ci.yml) [![PyPI](https://img.shields.io/pypi/v/python-libei)](https://pypi.org/project/python-libei/) [![License](https://img.shields.io/pypi/l/python-libei)](https://github.com/ctrondlp/python-libei/blob/main/LICENSE) diff --git a/docs/developers/verification.md b/docs/developers/verification.md index 5bcf5b9..3797fdf 100644 --- a/docs/developers/verification.md +++ b/docs/developers/verification.md @@ -37,13 +37,31 @@ By hand: relative pointer, keyboard, absolute pointer, in that order, which is the device race `ei`-side callers must handle; and `Session.Close()` was exercised. No input was injected — emulation is `libei.ei`'s job. -- **`libei.portal.InputCaptureSession` (added 0.5.0), negotiation half only**, +- **`libei.portal.InputCaptureSession` (added 0.5.0), the v1 half**, 2026-09-08, against a real Fedora 44 / GNOME Shell 50.0 session (xdg-desktop-portal 1.21.1). That portal reports InputCapture **version 0**, so this exercised the **v1 `CreateSession` path**: consent dialog approved, a real EIS fd returned, `zones()` answering `(0, [(1920, 1080, 0, 0)])`, and - `Session.Close()` exercised. The v2 `CreateSession2` -> `Start` path has - still never been run live. + `Session.Close()` exercised. + +- **The v2 half, `CreateSession2` -> `Start`**, 2026-10-04, against a real + Fedora 45 / GNOME Shell 51.0 session (xdg-desktop-portal 1.22.1, + xdg-desktop-portal-gnome 51.0, libei 1.6.0, Python 3.15rc2). That portal + reports InputCapture **version 2**, so the version gate sent this run down + the half the 2026-09-08 one could not reach -- and v1 is not reachable there + at all any more, which makes the two dates a pair rather than a repeat. Every + call was driven live, in order: `CreateSession2` (0.03s -- a plain call, with + the session existing before anything is asked), `Start` (12.94s -- the + consent dialog, approved), `ConnectToEIS` (0.10s -- a real EIS fd; `fstat` + reports a socket), `GetZones` (0.06s, answering `(0, [(1920, 974, 0, 0)])`) + and `Session.Close` (0.03s): 13.07s end to end. Nothing was captured, and + nothing could have been -- the run stopped at negotiation, and by the shipped + spec it is `Enable` that arms capture, not `Start`. + + So both negotiation routes have now been run against a real compositor. Two + things remain untested live on this route: persistence, since `persist_mode` + and the `restore_token` `Start` returns are v2 additions this run did not + use -- and the capture half itself. **The capture half remains unverified**, and unlike the rest of `libei.portal` its consent dialog is not the whole reason why. Approving a