On Sat, 7 Jun 2025, Laurent Cheylus wrote:
> ----- Mail original -----
> > Hi ports@
> >
> > [cc daniel@ as maintener]
> >
> > update for textproc/py-black to the latest version 25.1.0
> > Changelog: https://github.com/psf/black/releases/tag/25.1.0
> >
> > - update Makefile and distinfo for this version
> > - add a patch to fix tests with click >= 8.2.0
> >
> > Build and tests OK on current/amd64 with Python 3.12.10
>
> Daniel, in your CVS commit for textproc/py-black version 25.1.0, you forgot to add my 3 patches to fix tests for this version.
>
> Laurent
>
Hello Laurent, I did not forget your patches. I have committed my own
update that's been sitting in my tree for some time. You'll note I got rid
of the S/b/beta/ bits in PKGNAME in my commit which I don't think are
needed anymore.
I'll take a look at your patches as time permits. But if they only fix
black regress tests and not real issues in black then I might just wait
until upstream integrates them in a release.
A few comments on any futures updates to black that would make doing
updates easier:
1) please test reverse deps, and if you did do that please state that you
did. I generally don't commit updates without at least doing light testing
of consumers.
2) a consumer that is important to me is spyder, it would be really great
to test that updates to black continue to work from within spyder.
3) a past update broke pdm-backends when black was installed. I don't know
why but would be interested to understand why that was an issue in the
past so we don't accidentally cause the same breakage. For this update, I
did test that all pdm-backends did at least build. I'm not sure if you did
the same, but if you have the energy to figure this one out it would be
super useful. Maybe an update to pdm-backends would be useful given the
past interactions with black. I just don't know.
Thanks for submitting your updates for black. It's appreciated.
No comments:
Post a Comment