[Python-Dev] Issue20846 [was: Cherry-pick between Python 3.4 RC2 and final?]

[Nick Coghlan]

On 4 Mar 2014 09:49, "Ned Deily" <nad at acm.org> wrote:
>
> In article
> <CADiSq7ftyqu9+RgLCk54gMTWkTCcak97pVAfDJSzQ9NjQfuc8g at mail.gmail.com>,
>  Nick Coghlan <ncoghlan at gmail.com> wrote:
> > On 4 Mar 2014 08:40, "Mark Lawrence" <breamoreboy at yahoo.co.uk> wrote:
> > > Will this impact on the decision http://bugs.python.org/issue20846 ?
> > No. I never claimed pip was bug free (any more than CPython itself is),
> > merely the best available option.
> >
> > File a bug against pip (assuming there isn't one already), get it fixed,
> > and we'll bundle the fixed version with the next CPython maintenance
> > release.
>
> FTR, the behavior documented in Issue20846 has nothing to do with either
> pip, setuptools, or python3.4.  See updates to the issue.
>
> One could argue about whether pip should check and, by default, not
> offer to download and install packages that are marked in PyPI as not
> being compatible with the Python version being used, e.g. trying to
> install a Python 2 only package with a Python 3 or vice versa.  AFAIK,
> neither pip nor easy_install (setuptools) has ever done that.  And, in
> any case, it would not have helped with the package in question as it
> does not specify in its package metadata (as shown on PyPI) with which
> versions of Pythons it is supported.  I don't know whether a request for
> such a feature has ever been made for pip; I didn't see any on a quick
> search of the pip tracker.  But this is off-topic for python-dev.

As one final comment on this: allowing proper environmental constraints
that installers can reliably enforce is one of the goals of the metadata
2.0 effort. The existing metadata is not considered sufficiently accurate
to allow it to be used as the basis for automated enforcement.

Regards,
Nick.

>
> --
>  Ned Deily,
>  nad at acm.org
>
> _______________________________________________
> Python-Dev mailing list
> Python-Dev at python.org
> https://mail.python.org/mailman/listinfo/python-dev
> Unsubscribe:
https://mail.python.org/mailman/options/python-dev/ncoghlan%40gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-dev/attachments/20140304/b707be64/attachment.html>