Accelerating issue fixes and PR's


#1

Several of the staff are currently involved in looking at ways to become more transparent about issue and PR status, as well as better organizing our Github repositories.

If you have suggestions on what we could do better in this regard, or comments on what we are doing well, I’d appreciate hearing them.


#2

Semi-related, as plugin developers please review pull requests, especially to the API, but also to SC/SF/SV if you are willing and have the knowledge. It is hard to know what to change without feedback and it is better to change things before they are merged into the API. The more people who review and check out pull requests, there is a lower chance of bugs and other issues being introduced into production branches. In your review for pull requests, consider throwing your use case as well, if you have one, for the changes as it can provide examples for others of why this is useful/important, act as a possible avenue of testing for the PR creator, and maybe even provide some motivation for the creator as well.

If you want to help out even more, testing the changes locally is a great way to help a contributor finish their PR and get it ready to merged. If you need the new API for something in your plugin, you could even make a special version and test it against the changes to see if all works properly in-game and that you like the API.

As a whole, this can help pull requests be completed much quicker and as a result, possibly merged into bleeding faster as well.


#3

#5

Remind bot may be helpful, but as for stale I prefer a human touch, even if it is canned responses, for fear of the project seeming too impersonal, as real life can get in the way for both our users and developers.

A few lines about issues: issues are important (even if a few issues are reference to PR) but looking at the issues page (especially SpongeCommon/API) I think it’s them we should apply Meronat’s comment to. They are just too many to be handled by the Sponge team alone.

There is actually a ton less then there was recently, Phit’s been doing an amazing job, and I’ve been tasked to assist him, I just figured I’d reach out to the community first, to see if there are any untapped ideas.

Edit: Just to show one of the preliminary ideas already in action, Phit has been testing all 1.10 issues to see if they are reproducible on 1.12. He would appreciate any help received, and there is a project tracking the state of this task: https://github.com/SpongePowered/SpongeForge/projects