Re: Stop issue verification?

[Jean Abou Samra]

Le 02/11/2020 à 21:03, Michael Käppler a écrit :
> Am 02.11.2020 um 19:47 schrieb Jean Abou Samra:
> > Le 02/11/2020 à 18:18, Jean Abou Samra a écrit :
> >
> > > Thanks. I'm going ahead to change the labels right now.
> > Merge request put up at
> > https://gitlab.com/lilypond/lilypond/-/merge_requests/493
> > and labels renamed and deleted as appropriate, except for one,
> > namely Status::started. While verifying what I was doing I
> > noticed that 51 open issues have this label (as well as,
> > somewhat surprisingly, 18 closed ones). Most of them
> > haven't been touched in years; the most recent is from
> > two months ago though: 
> > https://gitlab.com/lilypond/lilypond/-/issues/4599
> > Of course, the wide majority doesn't have an assignee.
> >
> > List at
> > https://gitlab.com/lilypond/lilypond/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=Status%3A%3AStarted 
> >
> >
> >
> > Is it fine to remove this information? I think it is, but
> > it seems best to have confirmation as deleting a label
> > can't be undone (although most of them have were marked
> > with this label on SourceForge, which makes it appear in
> > comments).
> Hi Jean,
> thanks for working on this!
> I've had a quick glance on the list of open issues with 
> 'Status::Started'.
>
> It seems the majority are unfinished patches, where it is often not
> clear if the original assignees
> did not have time to incorporate review comments or did not want to work
> further on the patch at all.
> But if anybody does want to work on one of them, he can assign himself
> (again).
> So I think it is fine to remove 'Status::started', as well.
>
> Cheers,
> Michael
Hi all,

So, I did remove Status::Started.

At https://gitlab.com/lilypond/lilypond/-/merge_requests/493there's
some discussion about wether to reopen issues in the event of a
regression. Current CG contains a policy that this should never
be done, always opening a new issue. This dates back to the
Google tracker era, where the work was somehow more organized
than today if I read the archives correctly. I think we don't
need a policy anymore for such a thing: people are perfectly
capable of using their best judgement -- corrected by others
if necessary -- and reopening in some cases definitely makes
sense, such as if the fix was invalidated. What do you think
about removing the following sentence from
http://lilypond.org/doc/latest/Documentation/contributor/issue-classification?

This  means that nobody should ever need look at the report again -- if  there 
is any information in the issue that should be kept, open a  new issue for that 
info.


Thanks,
Jean