Stalled "Merge Reviews" (Re: What Fedora makes sucking for me - or why I am NOT Fedora

[Orcan Ogetbil]

--- On Tue, 12/9/08, Behdad Esfahbod wrote:
> 
> First, it's not like you wake up in the morning and
> boot your Fedora and it
> pops up a message saying "Merge reviews are not
> fixied, shutting system
> down."??  What's the problem really, other than
> some cruft left in Bugzilla?
> 
> If you were implying that the unreviewed packages should
> have been removed
> from Fedora, go ahead, remove cairo, fontconfig, freetype,
> pango, and vte.
> 

I have been thinking about this issue lately. Right now, if someone makes a new package (s)he has to throw it into a pool of ~800 other packages waiting for an assignee and hope that someone will notice the package.
Approximately half of these packages are "Merge Reviews".

- I joined the community this September. Nowhere in the guidelines I saw the definition of a "Merge Review". I had to ask people at #fedora-devel to figure that out.

- Out of the 8 "Merge Review"s I have done so far, only 1 maintainer cared to reply and fix the issues. Some "Merge Reviews", that I have done, only ask for changing 1-2 lines in the SPEC file. 

>From my experience, I can tell that there is an apparent dismissal on the "Merge Review" packages by the maintainers. How can we accelerate this process? A review consists of two main contributors. A reviewer and a maintainer. Here are some suggestions:

1- Finding the Reviewer: Whenever someone decides to contribute to Fedora, (s)he should be encouraged to do at least one "Merge Review". It would be fruitful if this is put in the guidelines and the sponsors check this before approving someone's sponsorship.

If some contributor does a "Merge Review", his/her packages should be pushed upward on the Review Requests list[1]. That way they'll have a better chance to get reviewed. I think this will encourage some people to do "Merge Reviews".

2- What can be done to force these maintainers to fix their packages? If a maintainer has an ongoing "Merge Review" with a "package review" flag set to "?", (s)he shall not file a new "Review Request" until that "Merge Review" is resolved. If a reviewer takes time to make a full review, the maintainer should at least respect that, or more harshly, should be made to respect that.

As I said there are ~800 packages awaiting a reviewer, and this number is not decreasing. An action should be taken soon.
-Orcan

[1] http://tinyurl.com/fedora-rev-req