> Please report this upstream to https://github.com/pypa/pip > This does not sound Debian-specific at all. > > I can't reproduce the bug, without writing a proxy that causes a failure > like you had, which is far beyond the effort I'm willing to put in here. > You're in a much better position to advocate for this bug upstream than > I am.
I agree with all of that. I would report upstream if the bug tracker were hosted in a more open and neutral place. Microsoft venues are places I will not go to interact, apart from searching to see if a report already exists. Github in particular has access problems. Debian shelters users with this bug reporting policy¹: “Don't file bugs upstream If you file a bug in Debian, don't send a copy to the upstream software maintainers yourself, as it is possible that the bug exists only in Debian. If necessary, the maintainer of the package will forward the bug upstream.” I don’t expect anyone to mirror it upstream because I think everyone should equally have the option to avoid Microsoft’s platform, and also I agree with the Debian project principle to not impose work on others. This report has the upstream tag so upstream developers can filter on that tag and so downstream devs can filter it out. I just noticed the --log and --log-file options function correctly in one case. Part of the problem might be interplay with a bug² where pip continues even if there is a problem with the log file location for the user running it. Reproducing it myself might be non-trivial since I would need an unlimited internet connection to test more installations (which do not indicate their size prior to action). So perhaps the report could be closed on that basis anyway. There are several more upstream bugs to report against pip*. I intend to report them in Debian’s BTS using the upstream tag, unless no one wants them. I realise that they may never be seen by someone who would work them. ¹ https://www.debian.org/Bugs/Reporting ² https://github.com/pypa/pip/issues/114