Hi David, thanks a lot for your response and the updates to the wiki. One more question:
Do we need to introduce the "add-community-reviewed-LGTM" and "remove-community-reviewed-LGTM" flinkbot commands? Can't community members use the regular GH review feature, and approve the PR, and Flinkbot detects the approvals? On Fri, Apr 11, 2025 at 5:16 PM David Radley <david_rad...@uk.ibm.com> wrote: > Hi Robert, > Thanks for the feedback. I have updated the flip with your suggestions. > > Comments on your suggestions : > > * “Another benefit related to this is that it is easier for the PMC to > screen > for committer candidates based on their review involvement and quality.” > Good idea. > * “Update the wiki process” – I have updated with a link > * Comments around “community-reviewed-requires-deep-review”. I was > looking for a way to indicate that the community review would not add much > benefit and requires subject matter expect due to the complexity / far > reaching implication of a change. These sort of changes I feedback on but > would not approve. The community review is still valuable in these cases as > well. Let’s leave this one out for now. > * *community-reviewed-suggest-close* How can community reviewers > tell Flinkbot that they want to close a PR? will we introduce a new > command? Yes I will add details to the Flip > * *community-health-initiative-reviewed* I will change to have the > same prefix. > * On the performance metrics thought. Not evil - I like it that we can > evidence what is working and what is not working about all aspects of the > process. So we can make targeted changes. I think this applies to all parts > of the process. We are gathering a snapshot of how the process is working > currently, I hope that we can get a lot of insight from trawling Git. I > hope that we can get more insight when we have these labels as it will call > out the impact of the community reviewers. We could split on a user basis, > it is probably worth doing on a project basis as well to get project wide > insights as to what might be working well. > > Kind regards, David > > From: Robert Metzger <rmetz...@apache.org> > Date: Thursday, 10 April 2025 at 08:45 > To: dev@flink.apache.org <dev@flink.apache.org> > Subject: [EXTERNAL] Re: [DISCUSS] FLIP-518: Introduce a community review > process > Thanks for the FLIP. > > In the benefits section, you are mentioning: > > > - community reviewing then becomes a respected way to contribute to > Flink on the road to becoming a committer. So, it is in the > contributor’s > interest to review. > > > Another benefit related to this is that it is easier for the PMC to screen > for committer candidates based on their review involvement and quality. > > "*Update the wiki process"* > > Which process do you mean? Can you share a link? > > > *community-reviewed-requires-deep-review* > > I am not convinced that this is a good label to introduce, because it is an > "easy way out" for community reviewers. If the community review process is > for contributors to behave like committers, then the community should also > be involved in deep reviews. Otherwise, the PMC can not assess if a > community reviewer is ready for committership. > > > *community-reviewed-suggest-close* > > How can community reviewers tell Flinkbot that they want to close a PR? > will we introduce a new command? > > > *c**ommunity-health-initiative-reviewed* > > I like that the other labels are all prefixed by `community-reviewed`, this > one isn't. > What's the benefit of adding a CHI-specific label, instead of just having > one process for community reviews? CHI members are part of this? > > I know this is a bit of an evil "let's create performance metrics" thought, > but what do you think about measuring the accept rate of community > reviewers? E.g. for a user a, what % of their approved PRs have been merged > w/o further feedback by a committer? > I guess one problem of this idea is that folks can focus on just approving > typo fixes. E..g there will be an incentive for people to open typo fix > PRs, and there will be an incentive for folks to approve those. Just an > idea. Maybe for v2 of this. > > Best, > Robert > > > On Mon, Mar 31, 2025 at 5:33 PM Doğuşcan Namal <namal.dogus...@gmail.com> > wrote: > > > Yep, ship it. > > > > I don’t see any harm on start following the suggested process since it > only > > adds a few labels to the existing PRs. > > > > We could learn more by time and improve the process on the way. > > > > QQ. Do the proposed flinkbot commands work already? > > > > On Fri, 21 Mar 2025 at 15:32, David Radley <david_rad...@uk.ibm.com> > > wrote: > > > > > I would like to start a discussion around FLIP-518< > > > https://cwiki.apache.org/confluence/x/_wuWF>< > https://cwiki.apache.org/confluence/x/_wuWF%3e> which proposes a new > > > community review process. Please let me know if you support this idea, > or > > > would like changes. > > > > > > We hope this process improvement will > > > > > > * Encourage more people in the community review PRs, by formally > > > recognising reviews as contributions. > > > * Reducing the workload on committers. > > > * Reducing our technical backlog. > > > > > > > > > Kind regards, David. > > > > > > Unless otherwise stated above: > > > > > > IBM United Kingdom Limited > > > Registered in England and Wales with number 741598 > > > Registered office: Building C, IBM Hursley Office, Hursley Park Road, > > > Winchester, Hampshire SO21 2JN > > > > > > > Unless otherwise stated above: > > IBM United Kingdom Limited > Registered in England and Wales with number 741598 > Registered office: Building C, IBM Hursley Office, Hursley Park Road, > Winchester, Hampshire SO21 2JN >
