I do think it would have been more productive if this discussion had happened earlier. I generally expect for non-trivial PRs that the first version have issues (certainly most/all of mine do.) As committers/PMC members, we do have a role as gatekeepers and we shouldn't merge PRs with issues. But the default in those cases should hopefully be to work with the original contributor to make any necessary changes or in some cases explain why the PR is likely to never be merged (because of some inherent issue to the approach).
Unfortunately, this often takes more time and effort than just fixing things ourselves, but that's part of what's necessary for building a healthy and sustainable open source project. -- Michael Mior mm...@apache.org Le mar. 28 août 2018 à 08:31, Zoltan Haindrich <k...@rxd.hu> a écrit : > Hello, > > I've found a simplification issue a while ago; written an ignored test for > it - and opened a new jira to have the fix its own ticket: > https://issues.apache.org/jira/browse/CALCITE-2327 - because it was not > connected to the actual ticket. > > * I've submitted a patch to disable this oversimplification because that > seemed to be the right move to me. > * The review process became stuck after a while... > * Today I've got a mail that my PR is got closed by Vladimir Sitnikov...I > went to see what's up... > * Seems like Vladimir have just committed a different patch - and closed > the ticket (which is still assigned to me) and that's it... > * I feel that it would have been better to at least contact me prior to > just throwing a completely different patch at the problem - ...because > unfortunately this new patch > still contains an bug. > > I've looked at the last few commits to see if this a "pattern": for > CALCITE-2271 something similar happened...Alexey Makhmutov have submitted a > PR in April; 2 days ago > Vladimir have continued and submitted his own version of it. > > Sorry, but I don't think the above process is "right"... > > cheers, > Zoltan >
