Allow hatch-vcs 0.5 (fix #3612) #3613
Conversation
Keep SemVer-bounding `hatch-vcs` as agreed in urllib3#3412 (comment), but stop pinning it all the way to the patch release.
There was a problem hiding this comment.
Thank you! LGTM. The CI failure is unrelated.
Stories like https://lwn.net/Articles/1020576/ make me reticent to unpin completely, especially since hatch-vcs did not reach 1.0 yet. I've added a commit to pin setuptools-scm, for consistency, as it's way more active than hatch-vcs itself.
Fair enough. Downstream in Fedora, we will (necessarily) keep removing or editing the pins as needed, and – as in this PR – letting you know whether or not anything catches fire when we do so. 😄 |
Keep SemVer-bounding
hatch-vcsas agreed in #3412 (comment), but stop pinning it all the way to the patch release. Fixes #3612.I’m happy to simply expand this version bound for now, although I would like to suggest that it might not really be necessary to upper-bound the version of
hatch-vcsafter all. There are 79 packages in Fedora that havehatch-vcsas a build dependency, andpython-urllib3is the only one that upper-bounds the version ofhatch-vcs. Furthermore, we found no regressions in any of those packages when we updated from 0.4.0 to 0.5.0.