Sponsor shortage

[Ben Rosser]

On Fri, Jul 10, 2015 at 6:44 PM, Michael Schwendt <mschwendt at gmail.com>
wrote:

> On Fri, 10 Jul 2015 15:25:05 -0400, Ben Rosser wrote:
>
> > Speaking as someone who relatively recently went through the process
> though
> > (and whose package(s) sat in the review tracker for two years):
> motivation
> > is hard to come by when it looks like you're not going to get sponsored
> > because (you think) nobody cares about your package.
>
> ??? When exactly in the process does it look like you're not going to
> get sponsored? After a week? After a month? After having done 2-3 reviews
> and linked those reviews in the needsponsor ticket(s)? After having pointed
> at that work in a post to devel@ list?
>
> I would need to see your initial review requests to comment on this
> issue further.
>
>
I don't really want to turn this thread into a "why didn't I get sponsored
sooner?" thread (and in fact, looking back, I'd guess not linking the two
informal reviews I did wasn't a good thing). But since I brought it up...

https://bugzilla.redhat.com/show_bug.cgi?id=823679

In fact, you yourself were the first reviewer. I guess I could have emailed
you privately, but...


> Nowadays, I consider the review queue as very tiresome. I've commented
> on many tickets, I even have made clear behind my name in bugzilla that
> I'm in the packager sponsors group, but nobody has taken that as
> opportunity to email me privately. The time when new contributors did
> that is gone.
>
>
As a potential new contributor, I never felt encouraged to email potential
package sponsors privately, as I didn't know any of them. I never felt
encouraged to bug a SIG (because while I was packaging a piece of Python
software, it was hardly a major piece of software or anything). I wanted to
package a piece of software I was considering using for a project, because
it wasn't already in the distribution and I liked Fedora enough that I
wanted to contribute something upstream. So I followed the procedure(s) on
the wiki, or at least I think I did. I reviewed one or two other packages
informally.

I assumed that eventually someone who was a sponsor would review my package
and tell me what they thought of it. Then maybe ask me to do a few more
things, maybe a few more reviews. Tell me what they thought I needed to
work on, etc.

And, I mean, that eventually did happen. So I guess the system worked. It
just took a long time to work.... longer than I would have expected from
only reading the documentation on the wiki.


> When somebody has not submitted a single ticket in bugzilla, not even
> via ABRT, one can not even be sure the person is using Fedora.
>

So you're saying that reporting bugs against other components in the
distribution that aren't necessarily packages is a thing that we are
looking for in potential new packagers?

That seems reasonable. But this is not indicated on the How To Get
Sponsored page.

> Then, once someone *does* notice your package, maybe they work with you on
> > improving it, ask you to do a few other things, etc. And it becomes a
> more
> > active process on both sides. At least, that's been my understanding of
> the
> > process based on what I've read in those guidelines.
>
> That's one way how to do it. Unfortunately, some of the needsponsor people
> don't respond. Some respond after months, saying they have been busy. No
> notification of that in the ticket. No response! What would happen if the
> package were in the distribution? Somebody would start the non-responsive
> maintainer procedure.
>
>
That makes sense, so perhaps we should apply apply some analogy of the
non-responsive maintainer procedure to new package requests (by unsponsored
contributors, anyway), rather than let their tickets sit in bugzilla for
forever without comment?


> > It wouldn't be a bad thing, IMO, to (automatically?) ping sufficiently
> old
> > tickets with a sort of "what's the status on this" and maybe a link to
> the
> > sponsorship guidelines reminding them that there are other things they
> can
> > do if they still want to become a packager.
>
> It's too tiresome IMO. Pinging is frowned upon in other tickets, too. Just
> be responsive, keep track of a single ticket that is important to you
> because you need it for the sponsorship process. It is severely
> demotivating for reviewers as well as sponsors to get no response to
> reviews they post in the tickets.
>

It's demotivating to packagers, especially new ones, when they get no
reviews over reasonably lengthy periods of time too.

Anyway, I wasn't suggesting we ping more regularly than a year.


> Really, I think you're asking for too much, if you expect all
> single-package-review-quest-with-no-activity guys to be sponsored
> or to be sponsored quickly.
> --
> devel mailing list
> devel at lists.fedoraproject.org
> https://admin.fedoraproject.org/mailman/listinfo/devel
> Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
>

I thought the point of the sponsorship system was to help mentor potential
new contributors until they reached the point where someone thought they
were competent enough and sufficiently well-versed in our guidelines. While
simultaneously providing a hurdle from preventing just anyone from
contributing low-quality packages into the distribution.

So no, I don't think we should just *sponsor* new contributors instantly or
quickly-- I don't want to lower hurdles either. But I also think we
shouldn't just let their tickets bit-rot in bugzilla for more than a year
without any comment. I don't know what the solution is, necessarily, but
it's enough of a problem that someone (else) thought it worth contacting
devel@ about.

Ben Rosser
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/devel/attachments/20150711/180874ee/attachment-0001.html>