Well, I don’t get to make that decision. If the maintainers choose to do that, I’ll follow, but there’s a good chance that a lot of the other contributors won’t. For something in rapid development with a lot of community contributions, you want that barrier of entry to be as low as possible.
So if it was up to me (and it’s not), I would say no. I would be open to an official mirror somewhere else, and perhaps moving to a separate feature/bug tracking system (esp. if it’s easier for the community to report bugs), which imo is the biggest barrier to moving the repo.
I guess I’m not particularly worried about it since the project is FOSS and the difficulty in switching is pretty low.
@sugar_in_your_tea As a non-lemmy-dev, I don’t get to participate in that decision either, no matter how strong I think the arguments are.
I’m not convinced that the difficulty in switching is low; as you say, bug/issue tracking is a big barrier, but other features are part of the #EEE strategy [4], and switching later when MS upsets the community like Musk or Huffman will be difficult.
An official mirror would be a good start to make a future move easier.
There are mirrors of the Lemmy code on Gitea and Gitlab. They are linked in the readme. We also hope to migrate development to Gitea once federation is implemented.
I get that but it still sounds awesome 😉 Of course Gitea federating with other Gitea (or Forgejo) instances would be awesome and make it a right candidate to host another fediverse project.
It’s really not that difficult. The things that need to be migrated are:
links to repository - tedious, but the GH repo could be left up with links to the new repo and code mirroring could be configured; I’ve seen that done pretty often, and some projects go the opposite route (e.g. the Linux kernel has a GitHub mirror for… reasons)
issues - the GitHub CLI and API both offer a programmatic way to pull down issues, so those can be migrated with a script to whatever the new solution is; I could whip something up in an afternoon if needed
actual code hosting - just change remotes and push; this is like a 5-min thing
CI - the current CI situation is really simple and easy to port
And, that’s it. If needed, the whole process could be done in a couple days.
Well, I don’t get to make that decision. If the maintainers choose to do that, I’ll follow, but there’s a good chance that a lot of the other contributors won’t. For something in rapid development with a lot of community contributions, you want that barrier of entry to be as low as possible.
So if it was up to me (and it’s not), I would say no. I would be open to an official mirror somewhere else, and perhaps moving to a separate feature/bug tracking system (esp. if it’s easier for the community to report bugs), which imo is the biggest barrier to moving the repo.
I guess I’m not particularly worried about it since the project is FOSS and the difficulty in switching is pretty low.
@sugar_in_your_tea As a non-lemmy-dev, I don’t get to participate in that decision either, no matter how strong I think the arguments are.
I’m not convinced that the difficulty in switching is low; as you say, bug/issue tracking is a big barrier, but other features are part of the #EEE strategy [4], and switching later when MS upsets the community like Musk or Huffman will be difficult.
An official mirror would be a good start to make a future move easier.
@lemmy
[4] https://en.wikipedia.org/wiki/Embrace%2c_extend%2c_and_extinguish
There are mirrors of the Lemmy code on Gitea and Gitlab. They are linked in the readme. We also hope to migrate development to Gitea once federation is implemented.
That is awesome to hear! Lemmy federating with the code forge it’s hosted on sounds awesome!
Im not talking about federation between Lemmy and Gitea. Not sure how that would work or if it would make sense.
I get that but it still sounds awesome 😉 Of course Gitea federating with other Gitea (or Forgejo) instances would be awesome and make it a right candidate to host another fediverse project.
It’s really not that difficult. The things that need to be migrated are:
And, that’s it. If needed, the whole process could be done in a couple days.
So I don’t see any kind of urgency here.