Review queue/FESCo after the merge

[Christopher Aillon]

Thorsten Leemhuis wrote:
> And 1108 open reviews in total (including lots of merge reviews (¹))
> according to
> http://fedoraproject.org/wiki/PackageMaintainers/InProgressReviewRequests
> 
> which redirects to
> https://bugzilla.redhat.com/buglist.cgi?query_format=advanced&short_desc_type=allwordssubstr&short_desc=&product=Fedora&component=Package+Review&query_format=advanced&bug_status=ASSIGNED&bug_status=FAILS_QA&bug_status=MODIFIED&bug_status=NEEDINFO&bug_status=NEW&bug_status=ON_DEV&bug_status=ON_QA&bug_status=PASSES_QA&bug_status=POST&bug_status=RELEASE_PENDING&bug_status=VERIFIED&long_desc_type=substring&long_desc=&bug_file_loc_type=allwordssubstr&bug_file_loc=&status_whiteboard_type=allwordssubstr&status_whiteboard=&fixed_in_type=allwordssubstr&fixed_in=&qa_whiteboard_type=allwordssubstr&qa_whiteboard=&keywords_type=allwords&keywords=&emailassigned_to1=1&emailtype1=exact&email1=&emailassigned_to2=1&emailreporter2=1&emailqa_contact2=1&emailcc2=1&emailtype2=exact&email2=&bugidtype=include&bug_id=&votes=&changedin=&chfieldfrom=&chfieldto=Now&chfieldvalue=&cmdtype=doit&order=Reuse+same+sort+as+last+time&field0-0-0=flagtypes.name&type0-0-0=notsubstring&value0-0-0=fedora-review%2B
&f
> ield0-1-0=bug_id&type0-1-0=notregexp&value0-1-0=%5E163776%24&field0-2-0=bug_id&type0-2-0=notregexp&value0-2-0=%5E163778%24&field0-3-0=bug_id&type0-3-0=notregexp&value0-3-0=%5E163779%24&field0-4-0=bug_id&type0-4-0=notregexp&value0-4-0=%5E177841%24
> 
> Some of them are quite old. And the list of course doesn't include those
> bugs that got closed as the packager lost interest over time.
> 
> 
> I think we have a problem here. I'm actually wondering what FESCo (and
> the Board) thinks about that.

Thanks for the mail, Thorsten.  I've actually been thinking about this 
pretty recently.

We have a problem, I agree.  It's a problem I'm happy to have in a way 
because it means we're growing fast.

Part of the problem is the review process itself.  It encompasses 
several pages, many of the items are duplicated, etc.  It's just unruly. 
  And the more packaging guidelines we have, the worse it will get.  I 
tried doing a few package reviews again recently and it annoyed me to no 
end to have to do this.  And I understand the benefits here of doing the 
reviews.

I think the ideal way to fix this is to have a web app that people 
submit packages to for review.  This web app will build the SRPM in 
koji, can check the md5sum of the tarball vs upstream, can run rpmlint, 
make sure the various specfile tags are in the right format, etc etc etc 
-- as many things that we can automate in the review process we should 
automate.  Once it passes the robot review, it auto files the bugzilla 
ticket with a link to the review, and then a human can worry about 
looking at the few things (s)he needs to.  This will greatly improve the 
process for everyone IMO.

Now I just need to convince someone to do this work :)