Package Review Stats for the week ending January 18th, 2009

Mamoru Tasaka mtasaka at ioa.s.u-tokyo.ac.jp
Thu Jan 29 04:07:10 UTC 2009


Robert Scheck wrote, at 01/29/2009 07:19 AM +9:00:
> On Mon, 19 Jan 2009, Brian Pepple wrote:
>> Review Requests: 66
>> Merge Reviews: 10
>> Total reviews modified: 76
> 
> Wow, so much reviews happened in one week. Unfortunately so much quantity,
> but much lesser quality. Do we really need quantity high scores rather the
> quality ones?
> 

Well, although it may be off-topic from what is currently being discussed about
package review stats on this mailing list, however I want to write my opinion
on this.

You may think that the completed number "76" on one week seems large,
however actually from my quick check
- ~265 new review requests are submitted on 2009 January
- Usually I don't care Merge reviews, however even if I don't count
  the number of Merge reviews there are still ~442 reviews which
  are assigned to anybody
  * among them ~72 requests blocks NEEDSPONSOR
And I guess the number of unfinished review requests (or even non-touched requests)
is growing continuously instead of much effort of reviewers.

I won't say that the quantity of review requests one reviewer did can sacrifice
quality of reviews, however it is also true that the current state is almost
unacceptable that there are "so many" unfinished review requests ("so many"
is from my feeling) and some of them are left untouched for months.
(As a sponsor member, I also note that I feel that it is especially
 undesirable that "so many" NEEDSPONSOR-blocking review requests are left
 untouched, because this can take new contributors' will away)

We need much more reviewers and much more reviews to be completed within one week.
I hope that we can find some way to satisfy both quantity and quality.

Mamoru




More information about the fedora-devel-list mailing list