+1 on the things proposed
Donat On Wed, Mar 9, 2022 at 2:18 AM Ralph Goers <ralph.go...@dslextreme.com> wrote: > > > > > On Mar 8, 2022, at 12:39 PM, Sean Busbey <sbus...@apple.com.INVALID> wrote: > > > > > >> On Mar 8, 2022, at 11:44 AM, Ralph Goers <ralph.go...@dslextreme.com> > >> wrote: > >> > >> > >> > >>> On Mar 8, 2022, at 8:55 AM, Sean Busbey <sbus...@apple.com.INVALID> wrote: > >>> > >>> - suggest all changes as PRs > >> > >> Generally, I am OK with that. As there have been no commits and no other > >> participants I didn’t see the point of doing PRs for all the work I have > >> done so far for 1.10.0 as they would still be awaiting approval. In fact, > >> I sent an email to the Flume private list in November informing them I was > >> moving to CTR for this work due to the lack of activity in the project. I > >> got approval for that from a couple of the active PMC members. However, I > >> am happy to create PRs so you can review what I do before it gets > >> committed from here. > >> > > > > I see. I only meant as a way to get contributions as compared to Jira > > attachments. A move to CTR is significant though, I’ll make a note of it > > while consolidating the contribution guide. Should I describe the plan as > > we would stay CTR indefinitely? Or is this e.g. a short term way to get the > > 1.10 release out the door? > > > > I like self-merged PRs as a foothold for new contributors to get involved > > in reviews (as you noted). Over in Apache Yetus we use “Lazy Consensus” on > > PRs. Essentially, a PR is merged at the sooner of a +1 (from any > > contributor) or 3 days elapsing since posting. I don’t mean to presume it > > would help Flume, but it has worked really well for Yetus at balancing out > > when the small pool of regulars has overlapping contribution time. > > I would hope we would not stay with CTR indefinitely. But to have RTC you > have to have people willing and able to perform reviews. I get very busy > between my $dayjob and Log4j so I can’t always be as timely as one would > like. I have no idea what Tristan’s availability is. Given your other ASF > activities I don’t think getting your commit bit turned on here should be > that hard. But once we have a number of people who can and do perform reviews > it would be best to go back to RTC IMO. And yes, it would be impossible to > get a 1.10.0 release using RTC. > > I would wait to discuss how to implement RTC until we get back to the point > where it makes sense. > > Ralph > >