[QGIS-Developer] stale bot :-(

[Nyall Dawson]

On Wed, 23 Dec 2020 at 00:14, Karsten Tebling <tebling at masuch.de> wrote:
>
> Hello,
>
> could you tell me what is considered weed? In most cases I can't provide
> sample data because I do not own the data or a coworker had the crash
> and "only" saved the report details.
>
> In the "QGIS unexpectedly ended"-dialog it reads "You can also send us a
> helpful bug report using the Copy Report button and opening a ticket at
> QGIS Issue Tracker." - the thing is, the "copy report"-button only
> activates after some text is inserted into "Tell us something about when
> you got the crash" and without knowing that, you just assume the "copy
> report"-button is broken too and close the dialog. In addition some of
> my workmates can't even read english, they just close the dialog
> immediately. I also noticed that one of my reports was closed because I
> only posted the stacktrace and a short sentence about what my coworker
> was roughly doing as far as he could remember and I couldn't provide
> sample data. So I guess posting the stack trace and a short explanation
> is considered weed and not a "helpful bug report"?
>
> Could you add more information to the "QGIS unexpectedly ended"-dialog,
> so it is easier for you to pinpoint the issue without having the user to
> create sample data, a sample project, a new user profile? That would be
> my wish for christmas.

That sounds ideal, but unfortunately things just don't work that way.
If your bug report is not reproducible there's a very low chance it
will get fixed.

So if you want your bug fixed, the motivation sits with you to make it
as easy as possible for the bug triaging team/developers to reproduce.
Honestly, if we (developers) can't reproduce a bug in <5 minutes,
we'll just move to the next ticket in the queue and you've missed your
chance at a fix...

And that's exactly what stale bot and the feedback tag is designed to
assist with -- it helps push the responsibility back to the bug
submitter to make sure there's sufficient detail and a reproducible
test case in the ticket. It's actually in place to avoid frustration
caused by the "I submitted this ticket 12 months ago, why has nothing
been done?!?!?" situation.

Nyall


>
> greetz
>
> Karsten
>
>
> Am 22.12.2020 um 09:37 schrieb Richard Duivenvoorde:
> > I received some "this issue will be closed because of lack of feedback",
> > but I think current 'stale bot implementation' is a little rude on the edges.
> >
> > (did I miss some discussion about this here?)
> >
> > I received some "this issue will be closed because of lack of feedback", only because I commented on them. But some are just plain valid:
> >
> > As an example:
> > https://github.com/qgis/QGIS/issues/39107#issuecomment-749298135
> > that is just a valid use-case (or maybe feature request?)
> >
> > This one:
> > https://github.com/qgis/QGIS/issues/37984
> > is just valid (I even installed Oracle to check this...)
> >
> > This one: also valid and still needed:
> > https://github.com/qgis/QGIS/issues/37956
> >
> > Weeding the issues is not nice, so removing non-valid, old or non responding is fine, but now I think the bot is pulling the veggies instead of the weed :-)
> >
> > Regards,
> >
> > Richard Duivenvoorde
> >
> >
> >
> >
> >
> > _______________________________________________
> > QGIS-Developer mailing list
> > QGIS-Developer at lists.osgeo.org
> > List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> > Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> _______________________________________________
> QGIS-Developer mailing list
> QGIS-Developer at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer