On Tue., May 21, 2019, 11:29 Antoine Pitrou, <antoine@python.org> wrote:

Le 21/05/2019 à 20:16, Brett Cannon a écrit :
>
>
> On Tue., May 21, 2019, 09:10 Christian Heimes, <christian@python.org
> <mailto:christian@python.org>> wrote:
>
>     On 21/05/2019 17.31, Antoine Pitrou wrote:
>     >
>     > As I said, if the main annoyance with nntplib is the sporadic test
>     > failures, then the relevant tests can be disabled on CI.
>     >
>     > NNTP itself is still used, even if less and less.
>
>     I don't like the idea to drop a third of the test cases for nntplib
>     -- and the 30% that actually test that Python's NNTP library can
>     talk to an actual NNTP server. IMHO it's more beneficial for core
>     development and for nntplib to have it maintained by somebody that
>     cares about the library.
>
>
> Ditto for me. If the code isn't going to be properly tested then it's
> already unsupported, at which point we're back to wanting to deprecate it.

So I should never have added those tests and then we wouldn't be talking
about removing nntplib.

Not necessarily. I suspect it still would have been listed, you would have objected, and someone may have looked at the tests to point out it was only unit tests with no integration tests against a live service like e.g. the pythontest.net-related tests and so it was already neglected.

It also doesn't help that no one is listed in the experts index for the module either.

-Brett


Regards

Antoine.
_______________________________________________
Python-Dev mailing list
Python-Dev@python.org
https://mail.python.org/mailman/listinfo/python-dev
Unsubscribe: https://mail.python.org/mailman/options/python-dev/brett%40python.org