In the past, Ahmet and I spent some time each week reviewing and pinging pull requests. This did not happen the past few weeks due to some vacations and travel. I do think pinging is effective for many of the PRs at least.
On Thu, Oct 12, 2017 at 1:20 PM, Lukasz Cwik <[email protected]> wrote: > My experience is that it takes a good amount of time to review PRs and a > good portion of my time spent contributing to this project is by reviewing > PRs. > I currently have 3 out of 10 PRs that are older then 2 weeks so in my > experience pinging people to about progress has been pretty effective. > Out of those older PRs, 2 of those PRs I have heard back from the authors > and that they would attempt to get back to it soon. > > On Wed, Oct 11, 2017 at 8:09 PM, Kenneth Knowles <[email protected]> wrote: > > > Hi all, > > > > We have hit 100 open pull requests today*. It is an arbitrary number, > but a > > good excuse to note the upward trend. In part, I think it is simply > having > > more changes happening, which is cool. But it is also due to review > > latency. Sorting by "last updated" the first two pages range from 6+ > months > > to 16 days ago. > > > > We may, first of all, need a sweep to close stalled / no-go PRs. > > > > After that, having a triage process where someone drops in on PRs and > asks > > "any update?" has not been terrifically helpful in my experience (and > also > > obscures how stale PRs are) but is perhaps the most active measure we've > > taken in the past. > > > > Gitbox will probably make it easier to see who is requested to review a > PR > > and whether it is waiting on the reviewer or the author. That may help. > > > > Any other thoughts? > > > > Kenn > > > > *I'm part of the problem; 16 of them contain the phrase "R: @kennknowles" > > and I also have ~4 outgoing PRs that have stalled > > >
