osdir.com


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[DISCUSS] [Contributing] (1) - Pull Request Template


Hi all!

This thread is dedicated to discuss improvements to the way we encourage
contributions. It is spun out of the proposal *"A more structured approach
to reviews and contributions".*

Please keep this thread focused on the contribution side, there are
separate discussions about the reviews!


*Points that were brought up about contributions so far*


*Adjust the contribution guide to not encourage "single commit" pull
requests*

Instead, factor out unrelated changes, cleanup, preparatory refactoring
into individual commits for easier reviews.

  - I personally fully subscribe to this, it is really helpful to the
review process and not really much extra work.


*Simplify the Pull Request template. The template is overwhelming and hard
to work with.*

  - Previous suggestion: remove the introductory text and reference the
contribution guide website
  - We had that before, there was a significant number of new contributors
that did not check out the contribution guide, hence we added a condensed
form of the contribution guide in the template so that everyone is sure to
see it. It creates mild overhead for frequent contributors, but PRs are
better described since.

  - We could think to remove the introduction text, and could push back on
badly described pull requests as a first thing during reviews

  - I personally found the "sensitivity checklist" (does the PR touch X or
Y) quite helpful, it implicitly brought it to the attention of contributors.
  - With a more principled review process, we could let this checklist be
part of the "needs extra attention" checks and remove it from the
contribution PR template.


*Putting work on reviewers versus contributors*

We have many more contributions than active and experiences reviewers.
Adding more reviews is needed, but there will probably always be more
contributors than reviewers.

As a consequence, it can make sense to ask the contributor for a little it
of extra work (like better description, better separation of changes into
commits) if that helps the reviewers to be more efficient and better handle
contributions. In my experience, most contributors understand that and are
willing to help do these things if it helps getting the contribution merged.

How do other people see this?

Best,
Stephan