All of these repositories are exemplar clones with no modifications to exemplar’s C++ code on the main branch.
At least for beman.indirect, there was a plausible reason for doing this; River was implementing indirect on a feature branch, which would have allowed for making use of the GitHub Actions CI system. However, users can now use beman-local-ci for this purpose.
I feel that these repositories clutter up the bemanproject organization, and that we should reserve the github.com/bemanproject namespace for libraries that have content. Developers who are not yet at the stage where they have feedback-ready code should develop locally or put the repository under their own GitHub username.
Please do not – these are valid projects in work. They are not 100% unmodified clones with maybe the exception of transform_join. We should try to communicate with the author before we delete. I have uncommitted changes on to/from chars personally.
Not disagreeing that these are on a time clock, but we can wait a bit more.
Rather than deleting, I think the path forward for these is to mark them as private and then contact the author to ask them to migrate them into their own user’s GitHub account until such time as they have code that’s ready for public perusal. I’ve already done this with the author of indirect. It’s fine to have work-in-progress projects but they shouldn’t squat in the github.com/bemanproject namespace.
Well I’m not a super fan of that because now without an invite no one can step up and work on it. As I recall Linus from NYC was also originally working on indirect. The other 2 are myself and Jan S. we’ll get there soon enough. transform we need to contact the author – they are a newish volunteer.
Anyway, people should be mostly looking at the web matrix of libraries – which these aren’t in as of yet because they’re too preliminary. I’m not against culling, I just want to be extremely lenient for stuff that’s in the early stages.
I agree with @ednolan here. We should actively trim the dead branches. If folks want to step up and work on these things, they can say so and we can support that. Otherwise, they’re just creating a worse experience for people browsing the repositories.
We discussed this at today’s meeting and resolved to add a “sandbox-” prefix to repositories that are barely more than exemplar clones. This differentiator will help Beman library users differentiate between libraries ready for active feedback and those that are stale/stunted. Leaving them there allows potential contributors to pick up where the last person left off.
Thanks. I actually have a volunteer to work on the library. I’ll admit that looking at the details now I think we might as well delete and start over on the repo to get the latest. I guess Linus never actually committed anything?