Re: Handling of sponsorship requests
On Fri, Aug 24, 2012 at 12:06:27AM +0200, Patrick Matthäi wrote:
> Hello list,
>
> as sponsor I realy welcome the decission and changes to handle
> sponsoring requests about our bugtracker, but there are still two
> questions/disadvantages from my side:
>
> a) How should we handle requests, where the maintainer *may* be MIA?
> Just for an example #658114 - a realy simple question is unanswered for
> about 6 months!
> IMO they should be pinged one more time, maybe they have overseen the
> response, but if I want to see a package in Debian, I would track it, so
> I also would see such "maintainers" as possible MIA candidates just
> after their first uploads. So on they may be not qualified to maintain
> packages/bugs within a distribution.
I read on bug 658114 not really a question to be answered but rather a
suggestion on where to find a sponsor, and since Ansgar Burchardt already sent
a cc to the Debian MySQL Maintainers I don't see anything left to do by Mateusz
Kijowski for this. Also, Mateusz Kijowski has been updating the package at
mentors, so I think that Mateusz Kijowski is doing OK by patiently waiting for
a sponsor without sending pointless messages like "I'm still looking for a
sponsor and I got no reply from here and there". So bug 658114 is probably not
the perfect example of what you meant to discuss.
But in general I agree that some requests for sponsorship deserve more
attention from the requestors.
>
> b) Should we assume that the uploader is aware about freeze
No, the sponsor should verify whether the changes are suitable for upload.
> and realy
> delay typical "new upstream release" uploads to unstable?
No, the freeze policy doesn't always forbid that.
> IMO I do not
> think so.
I agree that there is no need to delay every "new upstream release" upload to
unstable, if that is what you meant.
> The process is a bit more complicated but it is still possible
> to update testing packages without unstable upoads.
Not for fixing bugs in wheezy with severity "important". But then one can
still add epoch to revert the upload of a "new upstream release" to unstable.
The really unwanted uploads to unstable during the freeze are the "disruptive"
ones as explained in the freeze policy.
> Surely it is not the cleanest way
I guess that everyone agrees that the preferred way is via unstable.
> but sponsorship requests should be
> processed :)
I agree that Debian could use some more sponsors to get more sponsorship
requests processed. (We both wrote "processed", not "uploaded".)
>
>
> Personaly I would not sponsor packages in the a) case,
I agree that it is OK to ignore requests for sponsorship where the requestor
doesn't respond to review comments.
> so on I think
> they should be closed
Please don't close bug 658114 just yet. In general I agree that it is
reasonable to want to close some really abandoned RFS bugs.
> and some more active and interested mentor could
> do this job.
Sounds like you're interested in doing that. :-)
>
> b) is controverse when testing is freezed,
For many requests for sponsorship it is still OK to upload to unstable.
> but the number of requests
> will just grow and grow!
Sponsors can still continue to sponsor many requests during the freeze.
> Maybe also some warning to the uploader like
> "we are frozen, please only upload important bugfixes to sid if
> required, if not please use experimental"
That would be more restrictive than the freeze policy. I suggest to not issue
such warning.
Regards,
Bart Martens
Reply to: