New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
WIP app vs Draft Pull Requests #1506
Comments
👍 |
One other downside of the Draft PRs is that they don't trigger CI runs, I don't really mind |
I'm really grateful you have pointed this out. This is an important downside. If I read the situation correctly, I think there is not sufficient impetus to move forward with this recommendation on this repo. In that case, let us close this for now. If a track independently chooses to uninstall the app and wants to report on the experience, I encourage the report be made here. |
@petertseng Thanks for bringing this up. I forgot about Draft PRs. I'll try it out and see what I think. TL/DR: I think there could be some benefits, but I'm not arguing to keep this issue open. I will say not trigger CI runs, could be beneficial. I know not having the "immediate" feedback is a negative, but on the plus side, it provides a way to not run a potentially unnecessary CI process into a pipeline that could hold up other projects. Cause Exercism's Travis CI shares a build pipeline, so when X number of repos make CI requests, some may have to wait until an opening is made. If we can limit the number of CI builds to just those that are required (up the to repo), then I can see the value in that. |
I remove WIP today without knowing this issue existed. It's been removed from all other projects I work on and I considered Draft PRs to superceded it as a native implementation, so did not consider asking maintainers before doing it. If I'd seen this first I would have checked before doing so - sorry. It seems the Draft PRs do now have CI so I don't feel there is a good reason to have WIP installed and have two systems in place. I also really like to have the minimal org-wide apps installed as possible. However, if people really want it back, I don't mind reinstalling it. |
In this issue, I weigh the pros and cons of the
wip
GitHub app versus Draft PRs.The aim is to consider whether we should leave the
wip
app installed.I invite others to contribute other pros and cons.
I give an opinionated recommendation of what should happen for both this repo and the individual track repos, and also invite others to do the same.
Background
In June of 2018, in exercism/discussions#217, the
wip
GitHub app was installed on all Exercism repos.My interpretation of the purpose is: Allow contributors to prevent a maintainer from merging a PR that is intended not to be merged.
Contributors might use various signaling mechanisms, but the
wip
app is an enforcement of one particular signaling mechanism, since there is no other enforcement in place.In February of 2019, GitHub introduced Draft Pull Requests in https://github.blog/2019-02-14-introducing-draft-pull-requests/.
This has components of both signaling mechanism (there is clear visual distinction of draft PRs) and enforcement (they cannot be merged until marked ready for review).
The two things seem to meet similar needs, so let's compare.
Comparison
What can I do with one that I can't do with the other?
wip
: I can toggle back and forth between states by adding commits, amending commit messages, or editing PR title as appropriate. With a Draft PR, once you mark it ready for review, I found no way to go back.wip
app's status check may not have been marked as required.Opinionated recommendation
The two alternatives seem about equal in my current evaluation.
I give slightly higher weight for increasing the signal value of the check mark vs X indicator, since it is useful at a glance.
I give slightly lower weight to being able to toggle between states since this repo sees WIP work rarely and toggling even more rarely.
So at preference strength 2/10 I would think it's better to remove
wip
app from this repo, and use Draft Pull Requests in the appropriate situations.Regarding individual track repos, I think track maintainers should be empowered to make their own decisions, so it does not make sense for me to make any recommendations applicable to individual track repos, beyond this one:
Track maintainers should consider the pros and cons of the two alternatives and come to a conclusion that makes sense for their individual track.
The text was updated successfully, but these errors were encountered: