[Development] Long-lived P1 issues

[Jason McDonald]

On Tue, 3 Nov 2020 at 18:39, Allan Sandfeld Jensen <kde at carewolf.com> wrote:

> On Dienstag, 3. November 2020 05:34:02 CET Jason McDonald wrote:
> > If an issue is not important enough to get attention within a year, is it
> > really P1?
>
> But how many of those are accepted? P1 is just set on the assumption the
> bug
> is geniune, for instance reported crashes usually get P1. But if it can't
> be
> reproduced it might never get accepted.
>

That's a good point.  708 of the 1175 Open P1s are in the Reported state,
i.e. if we're sticking to the defined lifecycle, those bugs are
unconfirmed/untraiged.

I find this rather worrying, since it seems to suggest that the majority of
bugs that our users/customers believe to be P1 aren't getting triaged in a
timely manner (or at all in a lot of cases).

>
> Also many modules are not maintained, is it in a module with an active
> maintainer of group?
>

Clicking around randomly for a few minutes suggests that these are
overwhelmingly in modules that aren't marked as inactive. I don't have any
other way to tell which modules have active maintainers.


> I was more confused when I found P0 that had been open for 10 months. That
> wasn't really a P0.
>

That's also a little worrying. When I was Qt's release manager way back in
2008-10, I always had the list of open P0s open and regularly checked
progress because those were the issues blocking the releases I was
responsible for.  At that time there were periodic discussions of having
some automated process to ping the assignees of idle P0s after a short
timeout.

Cheers,
--
Dr. Jason McDonald
(macadder on FreeNode)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.qt-project.org/pipermail/development/attachments/20201103/7eb4a8e3/attachment.html>