Versions and Releases¶
The asyncstdlib
versioning closely follows
versioning of the Python standard library.
New versions are published via PyPI and Conda-Forge
for installation via pip
and conda
.
Versioning and feature coverage¶
The asyncstdlib
mimics the versioning of the Python standard library:
Major and Minor version indicate which Python feature set is supported, and
Patch version indicates the iteration of this feature set.
For example, asyncstdlib
version 3.9.2 provides the feature set of Python 3.9,
such as cache()
added in 3.9
and cached_property()
added previously.
The asyncstdlib.asynctools
feature set does not follow a strict version model.
New features may be added at minor or patch releases.
Release workflow¶
Note
This section is only relevant for maintainers of asyncstdlib
.
Releases are performed manually but should happen at least when an important fix or major feature is added. Most releases will bump the patch version number; only bump the minor or major version number to match a new Python release.
- Review all changes added by the new releases:
Naming of functions/classes/parameters
Docs are up to date and consistent
Unittests cover all obvious cases
- Bump the version number:
Adjust and commit
asyncstdlib.__init__.__version__
Create a git tag such as
git tag -a "v3.9.2" -m "description"
Push the commit and tags to github
- Publish the new release:
Create a new GitHub release from the recent version tag
PyPI will be automatically published to via GitHub actions
Handle the automatic PR at the Conda-Forge asyncstdlib recipe