[Python-Dev] PEP 594: Removing dead batteries from the standard library
bcannon at gmail.com
Tue May 21 14:07:08 EDT 2019
On Tue., May 21, 2019, 04:25 Victor Stinner, <vstinner at redhat.com> wrote:
> Le mar. 21 mai 2019 à 13:18, André Malo <nd at perlig.de> a écrit :
> > There's software in production using both. (It doesn't mean it's on pypi
> > even free software).
> > What would be the maintenance burden of those modules anyway? (at least
> > nntp, I guess it's not gonna change).
> The maintenance burden is real even if it's not visible. For example,
> test_nntplib is causing frequently issues on our CI:
> It's failing frequently since 2013, and nobody managed to come with a
> fix.. in 6 years.
> There are 11 open issues with "nntp" in their title (8 with exactly
> "nntplib" in their title).
I would be curious to know how many feature PRs there might exist as that
also is a burden for any of these modules beyond bugs and any general
upkeep we do (e.g. someone on Twitter argued that because someone merged a
new feature that means that a module should stay, but to me that just says
one person cared enough to write a PR and one core dev viewed it as
harmless to merge which isn't necessarily representative of usage).
P.S. I have an idea about under-supported modules in the stdlib and feature
requests, but it's too early to share (nor the appropriate thread of
> test_nntplib uses the public server news.trigofacile.com which is
> operated by Julien ÉLIE. Two years ago, Julien asked me if there is
> any plan to support the NNTP "COMPRESS" command.
> Night gathers, and now my watch begins. It shall not end until my death.
> Python-Dev mailing list
> Python-Dev at python.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Python-Dev