nifi-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Joe Witt <joe.w...@gmail.com>
Subject Re: can contributers review PRs?
Date Tue, 24 Mar 2020 16:26:17 GMT
Nissim,

Contributions by anyone (commiter or otherwise) can be both created by non
committers and reviewed by non committers.  When it comes to actually
merging the change you definitely need a committer.

Ultimately whoever merges the changes need to be aware of what they're
doing and ensure the changes represent something workable for the project.

Many PRs that get stalled are adding integrations with new systems or
endpoints.  Often those endpoints cannot be tested without licenses or
complex setups.  This makes the PR pipeline very tough.  Other cases are
changes for which the contributor isn't aware of the side effects or
security implications/etc.. and for which no committer has time to dive
into at that time.

We are likely to enable a Github CI action that will help us maintain
active vs stale PRs such that after a period of time it will automatically
close/mark stale PRs that haven't gotten attention.  Otherwise we have what
we have now which is someone puts up a PR and it might sit indefinitely
(some are very old for instance).

The bigger question at play here might be "What are some things I can do to
ensure my change gets traction?"  This is probably a longer
question/discussion but providing strong evidence of effort on license and
notice, documentation on how to use it for users, documentation that is
detailed for someone else to test it, etc..  All of these things go a long
way.  But before that it is often a good idea to create a JIRA and/or
prompt discussion on mailing lists explaining what you're trying to
accomplish and solicit ideas.  Once a PR shows up without any discussion it
can be really hard to engage because at that point someone has invested a
lot in a particular path and may not appreciate if we share other
concerns/problems/ideas. It just really depends.

Anyway - absolutely anyone can contribute and review.  But a committer is
responsible and necessary to push a commit and they must know what they're
doing and signing the community up to support.

Thanks

On Tue, Mar 24, 2020 at 12:19 PM Nissim Shiman <nshiman@yahoo.com.invalid>
wrote:

> Hello NiFi Devs,
>
> It seem like PRs are reviewed by committers.
>
> Can a contributor review a PR as well?  And, if so, could that count as
> the necessary review for the ticket?
>
> Some tickets have nice functionality, but could become stale if committers
> have their bandwidth pulled in other directions.
> This way the final merge is still done by a committer, but the review step
> has already been handled.
>
> Thanks,
> Nissim SHiman
>

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