Hi everyone,
Our Analytics crew have worked out how to generate a graph that gives us a view into our code review backlog: http://gerrit-stats.wmflabs.org/graphs/mediawiki
The red line is roughly the equivalent of this search in the Gerrit search box: is:open -CodeReview=+2 -CodeReview=+1 -CodeReview=-1 -CodeReview=-2 project:^mediawiki.*
...which, in English, means "everything in the mediawiki/* projects that hasn't been marked with a positive or negative review yet, and hasn't been merged or abandoned yet"
These numbers seem to be +/- 10 revisions, and not evenly off over the history, so bear that in mind as you look at the numbers. In particular, it seems to paint a slightly rosier picture for how we're doing on keeping up with the backlog than we are.[2]
That said, we seem to be doing pretty good on keeping up - better than I thought had you asked me before I had the graph staring me in the face. We still have quite a backlog, but it appears to be shrinking by a modest amount. Our peak backlog appears to be mid July. For those of you that have been reviewing, thanks for keeping up!
As of this writing, there's 207 revisions that have neither positive nor negative reviews associated with them. That's still seems like a pretty big number. 30 of those are more than a month old, and some date back to May.
How is the process working for everyone? Is stuff getting reviewed quickly enough? How has it been for the reviewers out there?
Rob
[1] https://github.com/wikimedia/limn [2] For those interested in the gory details. Unfortunately, it's not a perfect history due to the way Gerrit stores the history of approvals (or rather, the fact that Gerrit doesn't store the "history", just the current approval state for any given patchset). In addition to known discrepancies, there may well be other issues. In tracking it over the past couple of days, it looks like the last few days are slightly undercounted (relative to the historical numbers), as they drift upward every day so. Everything prior to August 12 is stable, though, so we seem to be getting *consistent* numbers for everything before August 12 (though quite possibly overcounted by 10-ish revisions). It also more-or-less lines up with the few manual datapoints that I have.