Re: [PATCH v1 (RFC)] docs: discourage users from using bugzilla.kernel.org

On Wed, Jan 13, 2021 at 01:17:10PM +0200, Jani Nikula wrote:

quoted

quoted

Well, that said, a lot of stuff sent to the _proper_ mailing lists also never
receives a response

Good point.

There's a school of thought that this is actually a feature. If there's
no attention, the reports on the list will just fade away and be
forgotten. Whereas in bugzilla, someone needs to actively resolve even
the ignored and forgotten bugs. (Or it needs to be automated.)

FWIW, it's easy for me to script this, if there is consensus that a bug that
hasn't seen any activity for longer than N months should be auto-closed with
some apologetic comment like:

    This bug has aged out and will be auto-closed. Sorry that it didn't work out.

    If this issue is still present in recent kernel releases, you may need to
    reach out directly to subsystem maintainers in order to get their attention
    regarding this problem.

    Please do not re-open this bug without the above step, as it will simply
    get auto-closed again in the future.

Attending to a bug database of thousands of open bugs takes a huge
amount of effort, and if the bugs aren't being fixed, a lot of that
effort is just wasted. If a bug doesn't get fixed now (or soon-ish),
what are the chances it'll get fixed months or years down the line?

Well, it's *possible* that someone comes across that bug during their research
and adds enough additional information to get it fixed. However, this is
extremely unlikely and it's better to just open a new bug anyway.

Just musing, has anyone else seen a shift in bug reports from "I'm part
of the community, and I want to help improve this stuff" towards "I'm a
customer and I demand support"? I don't think the kernel community can
really cater to the latter very well, and would be better directed at
distro bug trackers.

I haven't seen any specific change like that. One good thing about bugzilla is
that people who do file bugs have already overcome significant barriers to
sign up and navigate the quaint, decades-old bugzilla interface -- so their
bug reports tend to be generally well-written.

-K