- cross-posted to:
- selfhosted@lemmy.world
- cross-posted to:
- selfhosted@lemmy.world
You must log in or # to comment.
I think it was time on October 26, 2018.
long past time. what’s taking people so long
what’s taking people so long
besides the fact that there’s no decent alternative for workloads of a certain size/scale?
Which almost nobody needs?
What is missing from GitLab?
A platform that is down for almost a whole day every month does not seem like a decent alternative for those workloads
Honestly. GitHub has been playing catch-up to Gitlab’s features for like decade to me. I still prefer their ci runner model to this day.
If you have to ask the question, you haven’t made a serious comparison of the two.
No, I have just migrated my company to self-hosted GL and found a better experience with literally no downsides.
I think it is a fair question, because I really don’t see what could be the deal-breaker. Is it copilot, or the terrible GitHub actions?
You don’t see the deal breaker because your requirements are clearly low enough that open-core software works for your use case. Your experience is not universal. Your inexperience is leading you to believe you’re more knowledgeable than you really are. I don’t need to explain myself to someone that engages in bad faith arguments.
Let me guess, you chose Gitlab because you’re a DHH fanboy?
You’re spending more effort to be a smug prick than it would take to just rattle off an example or three. Sounds more to me like you don’t know what the fuck you’re talking about and just want to publicly stroke your… Ego.
“Sounds to me like…”
Nobody asked you? Why do we care what you think?
Sounds to me like you’re insecure and looking for a free handout instead of being able to come up with a coherent reason for why you’re owed something you’ve not earned.
Not the person, but get off your high horse, and either explain what you mean, or shut up.
Thats a lot of words to say nothing of value.
Thats a lot of words to say nothing of value.
Projection ain’t just for the cinemas.
That’s a lot of words over two comments to avoid answering a simple question
Thanks for continuing to put your inexperience on display. If you’re so certain you’re right, you don’t need my answer.
There’s always Azure DevOps. Wait…
Yep. Pretty sure it’s just inertia/network effect at this point.
Decentralisation. I don’t want to fall in the same trap picking a huge platform which can only be ran by a large entity. Federation in Forgejo is taking quite long.
We made a PoC on top of Solid which does work so it is clearly possible, but the code quality is too low to trust the solution.
Perhaps we should use git in the distributed manner it was designed
Yeah, too many people don’t know about having multiple remotes
One thing, any chance you’ve heard of a way to have two remotes, one basically ignored gitignore (personal self-hosted storage) vs front-facing with github or similar?
Hmm not natively but I think you could hack something together using git plumbing, perhaps just some hooks. I would still advise against adding secrets to version control even if it stays completely local though, it’s still “bad hygiene” and something you should try to avoid. You don’t need to go with a full-fledged secret management solution, just keep your secrets in
.envs or similar external files and use a password manager to store your live keys. Of course it all depends on the specific context, if it’s something bound to loopback I’m not going to stop you from runningadmin:adminand committing that to your personal docker-compose stash, but if it’s protecting something you care about don’t put it in git.
gitlab?
They’ve been pushing a lot of AI slop lately, and in general aim to turn gitlab into an AI-coding-first platform, where humans only coordinate AIs. Search for gitlab act 2 if you wanna know more
When Nightmare Eclipse got banned from GitHub they were also banned from Gitlab because Microsoft ordered them to. I would not use Gitlab.
Why even rely on a web interface for git, just use git
Pull requests and issue trackers
I can do a pull request via email
Why do issue trackers need to be connected?
Genuinely, pull requests being off platform without good integrated diff and merge conflict resolution tooling just doesn’t work at a certain scale.
Like linux kernel scale?
I would argue that Linux kernel scale basically doesn’t work. It is a huge barrier to would-be developers to get into. You might argue that’s a feature, but it’s pretty hard to argue that it doesn’t add a lot of friction to the process
Reason #74 to do so
