Message347153
| Author |
steve.dower |
| Recipients |
dstufft, eric.araujo, p-ganssle, paul.moore, steve.dower, tim.golden, vstinner, zach.ware |
| Date |
2019年07月02日.20:06:22 |
| SpamBayes Score |
-1.0 |
| Marked as misclassified |
Yes |
| Message-id |
<1562097983.47.0.889642211937.issue37481@roundup.psfhosted.org> |
| In-reply-to |
| Content |
For the record, I am 100% in favor of deprecating and removing bdist_wininst and bdist_msi from distutils as soon as possible, as they are maintenance burdens and bdist_wininst in particular often attracts AV false positives.
setuptools is welcome to either do the same or restore the functionality, I don't have any preference if someone else is maintaining it (but I'd suggest removal there too until someone wants to maintain it).
Some of our distributions of Python already exclude bdist_wininst, so it can't be blindly relied on. And I don't know that it was ever upgraded for PEP 514. So far I have heard zero complaints about this, though as Paul Ganssle pointed out on the PR they're not likely to be using the "newer" packages.
So consider me +1 for deprecate in 3.8 (we can still do that, right? it's just docs and a warning on use) and remove in 3.9 or 3.10 as appropriate. |
|