On Fri, May 10, 2019 at 10:23 AM Justin Mclean <[email protected]> wrote:
> Hi, > > > I believe it to be just normal that things take a while to establish > themselves. > > I first made that pull request 2 weeks ago, that seems a long time to get > something into a repo. But then again all ASF projects I work on use CTR, > where any committer can commit right away, and others can review when they > need and/or when coming up to a release, rather than the process here. > I understand the frustration (I actually gave a whole talk on it), in other projects I have to wait for months or years to get something committed :( I've seen in so many projects (not going to name names here but I've discussed it on the relevant project lists at the time) where the reviews just never happen. "We'll get to it later", "Yes, we still need to fix it", "We'll just create a follow-up Jira" etc. but it never happens. It also puts the burden of getting the quality right on someone else and allows anyone to commit any quality (obviously not implying anything about the quality of stuff you've contributed so far). This - in my personal opinion - is very bad and comes from my experience. I'm happy to hear that your experience with CTR has been more positive though! > I’m sort of a little confused to why we need to have two people(?) need to > approve a PR before it get submitted, IMO that stops people being able to > work effectively on it and collaborate and seems to just prolong the > process. This process also seem to enforce use of the GitHub UI which if at > all possible I’d rather avoid. I go along with it if I must but I really > not sure what advantage it gives over having a discussion here on the > mailing list. It seems to fragment the discussion and people not following > along on GitHub may be missing context. > Two people? I think we agreed that only a single +1 is needed. Dmitry asked the same recently. So whoever does the review can also commit. I also don't see why you have to use the Github interface. You can just attach a patch to Jira. > I appreciated the feedback given on the PRs but so far it seems that the > PR needs to be of a very high standard before it can be committed. My > concern is that may be an artificial barrier to entry, especially when > starting out. IMO It good to have a few things not 100% polished in your > repo as that give people work to do if they are inclined. I think my earlier comment relates to this. Looking at the current and past contributions I don't think anyone has really been blocked by unreasonable requests/high standard only by a lack of time. > Thanks, > Justin > >
