Xqt added a comment.
I introduced Ruff two years ago as a replacement for pydocstyle. Pydocstyle was no longer maintained, and its own GitHub repository explicitly recommends Ruff as a replacement. For me, the interesting question is therefore which tools actually give Pywikibot the best maintenance and code-review workflow, rather than who develops or owns them. I don't think that being maintained by a for-profit company is, by itself, a good reason to reject Ruff. Pywikibot itself benefits from commercial companies providing infrastructure and services to open-source projects. We use GitHub and GitHub Actions, previously we used Travis and AppVeyor for CI, and use commercial services such as Codecov for coverage reporting. More importantly, Pywikibot itself can be and is used commercially, just like Wikipedia content can be reused commercially. So I don't think the commercial status of the company behind a tool is such a meaningful criterion here. On the other hand, I don't see a strong reason to prefer Ruff's isort implementation over isort itself. We already use Ruff for its other checks, so having import sorting integrated into Ruff is convenient, but I don't consider that a decisive argument. The isort package is actively maintained, and I don't see any significant advantage in using Ruff's implementation here. Even saving two seconds of test time is not particularly important. TASK DETAIL https://phabricator.wikimedia.org/T434949 EMAIL PREFERENCES https://phabricator.wikimedia.org/settings/panel/emailpreferences/
_______________________________________________ pywikibot-bugs mailing list -- [email protected] To unsubscribe send an email to [email protected]
