zookeeper-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Norbert Kalmar <nkal...@cloudera.com.INVALID>
Subject Re: [Suggestion] Use Co-authored-by in commit messages
Date Wed, 08 May 2019 12:58:45 GMT
Thanks Maoling, I also think encouraging code review as well is a good
idea, but, unfortunately I have a "but" :)
I see two issues with including reviewers in the commit message.
First, I don't think there is a method to automate this, although I think
the commit script the committers are using can be modified to include it.
Otherwise doing manually would complicate merging PRs for committers.
My other, bigger issue is that there is nothing to track this information.
At least I am not aware of anything. What I mean is Github tracks authors
of the commits. But what would we use the reviewers information? If you
just want to check reviewers for whatever reason, there is a filter for
that already on github, in the Pull Request view. And this would also make
the commit message more "bloated".

I'm not saying we shouldn't do this (not a -1 from my side), I just have my
concerns mentioned above.

Is there any Apache project doing this? Just out of curiosity.

Regards,
Norbert


On Wed, May 8, 2019 at 2:34 PM Justin Ling Mao <maoling199210191@sina.com>
wrote:

> +1,A very good Suggestion.Thanks Norbert.I also suggest about the sign-off
> of the Reviewers' name.For the incentive, if someone participate in the
> review of PR, no matter whether he/she is a committer, we all need include
> his/her name?
>
> ----- Original Message -----
> From: Norbert Kalmar <nkalmar@cloudera.com.INVALID>
> To: dev@zookeeper.apache.org
> Subject: [Suggestion] Use Co-authored-by in commit messages
> Date: 2019-05-08 17:36
>
> Hi Devs,
> I've got this idea from HBase.
> So: when there is a patch that is abandoned by its original author for any
> reason, and it can no longer be merged, someone comes by, and asks to
> continue to work on it. Usually the reply is to use the change freely or no
> reply at all. Either way, what people end up doing is a new pull request,
> and (correct me if I'm wrong) we do not have a standardized method how to
> indicate, or even to indicate at all the original author.
> My proposal is to use github's feature of Co-author, which is a way of
> attributing multiple authors of a given commit. See more details here:
> https://help.github.com/en/articles/creating-a-commit-with-multiple-authors
> I wouldn't think this needs to be forced or anything on future PRs, but
> it's a nice thing to have. And if someone sees an old patch, this could
> give more incentive to continue to work on it, knowing there's a guideline
> in the HowToContribute guide to credit him/her.
> I can update the guide at
> https://cwiki.apache.org/confluence/display/ZOOKEEPER/HowToContribute if
> the reception is positive.
> Regards,
> Norbert
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message