I prefer Forgejo, but both it and Gitea support actions like GitHub's. You can have a nice CI/CD pipeline that runs 100% in-house, for free. I adore it for personal projects.
If you want to run a process after each push to a branch or merge into main or whatever, you describe it in a YAML file in that repo. Configure some workers to run those actions and off you go! I use it for things like running tests and applying Terraform changes.
I understand that part. Mostly interested where the runners are coming from? macOS especially is pretty costly to provide runners for, so who is doing that for free?
We've run Gitea actions (and contributed here and there) for a couple of years, since-by-side with Github. We host in containers on the Gitea side so there are some marginal differences as to what can be run in a job, but our experience has been very positive.
I want to signal boost the following quote from the URL above:
> Forgejo was created in October 2022 after a for profit company took over the Gitea project. It exists under the umbrella of a non-profit organization, Codeberg e.V. and is developed in the interest of the general public. In the year that followed, this difference in governance led to choices that made Forgejo significantly and durably different from Gitea.
If you take it at face value (at your peril), Gitea is about to start enshittification, while Forgejo will not at any point. My personal opinion, is that this is credible.
They are, and always were. I think we’re more accustomized to it though, and know they won’t try to pull some shenanigans with the CE at least. I guess Codeberg didn’t trust Gitea in the same way when they decided to fork, but I think as a result Forgejo would be more sustainable, them being a nonprofit and all.
I bounced away from Gitea because they don't (last time I checked) have OIDC. I started[0] trying to revive-and-drive a previous PR[1] to add it, but the test failures are beyond my motivation to investigate and resolve.
While GitHub and GitLab have dedicated design and front-end teams to improve their UI/UX, Gitea and Forgejo aren't large enough to reach that scale, even after Gitea became a company.
For example, look at the number of issues triaged with "UX" [0] or "UX Paper Cut" [1] on GitLab. It is an order of magnitude larger than you would find in any other FOSS option.
Sorry but the GitLab UI was bad, is bad, the whole software feels clunky and slow to use and everything is nested where in comparison Gitea is simple, intuitive and straightforward, just like the old Github days. I also don't know if it's a good sign that there are a lot of UX issues?
In my experience, the "really good" is that it comes batteries included:
- completely docker based CI/CD which makes reasoning about what it's going to do easier than "read through some minified .js from some rando"
- they do have composable CI/CD akin to the GitHub Actions marketplace, but I haven't used it as much in anger to speak to how valuable it is versus "competitive checkbox feature"
- built-in Terraform State, so no more S3 + Dynamo
- highly configurable JWT claim curation for ease of OIDC based access from the pipelines
- good integration between the platform and multiple Kubernetes clusters
- related to that, a strong "review environment" setup
- they were also hinting at being a Sentry replacement, but regrettably I had to switch back to GitHub before that came out of preview so I don't this second know where it stands
GitLab doesn't have an equivalent of GitHub actions (except an alpha-quality prototype).
GitHub Actions can share runtime environment, which makes them cheap to compose. GitLab components are separately launched Docker containers, which makes them heavyweight and unsuitable for small things (e.g. a CI component can't install a dependency or set configuration for your build, because your build won't be running there).
The components aren't even actual components. They're just YAML templates concatenated with other YAML that appends lines to a bash script. This means you can't write smart integrations that refer to things like "the output path of the Build component", because there's no such entity. It's just some bash with some env var.
Pedantically I think GLCI treats every environment the same, but by review environments I meant "disposable copies of the app such that one could interact with it during merge request review" e.g. https://mr-8675.example.com corresponding to /example/-/merge_request/8675 that would be provisioned when the MR was opened and torn down when the MR was merged or closed
I believe it aligns with this behavior in GitHub: <https://docs.github.com/en/actions/how-tos/deploy/configure-...> with the distinction that it appears from the GH docs that they think of that as "needs administrative approval" whereas GLCI thinks of it as "if the pipeline has permissions to run provisioning, off to the races, because names are free"
GitLab introduced the "deployment tier" I think as a means of communication to other users about the importance of the environment, but control over what credentials were made available to CI/CD was always controlled via <https://docs.gitlab.com/ci/environments/#limit-the-environme...> which partially explains why the only reason to involve a repository administrator would be to install or update a secret needed to deploy successfully
“Really good” under which metric? Because it is slow, even more confusing after the terrible sidebar redesign and, to quote a famous author, its usage does not spark any joy.
Codeberg and gitea, on the other hand, feel great, like early Github. Fast and simple, instead of a product that’s adding feature on top of half-baked feature to capture the sweet corporate $$$.
I have to agree. I recently joined a company using Gitlab, coming from years of GitHub only. I have a soft spot for underdogs but I already found many features with bugs (especially related to hierarchy and inheritance) that makes you feel "meh".
I feel like all new AI tools only integrate with GitHub though, like Claude Code. We're actually thinking of moving from GitLab to GitHub, just for this reason.
Is that a problem with GitLab or a problem that should make you wary of Claude Code though? It's one thing to lock yourself into one LLM provider, but when they start chaining you to other SaSS organizations aren't they just locking you down even more?
It's... ok. But many of the really useful features are paid. E.g. merge trains or mandatory reviews.
I also don't think "it's open source!" is a huge differentiator because it's enormous, difficult to deploy from source and written in Ruby so the chance of being able to actually modify it for some feature you want is near zero.
I think Forgejo is probably a way better option at this point even if it is less mature. It's written in Go so way easier to deploy and edit. And none of the features are paid.
I do like Gitlab but... it's not amazing. I liked Phabricator more (except for its lack of integrated CI).
It isn't. Ruby lacks static typing, and Rails heavily uses generated identifiers, which means navigating a huge codebase like Gitlab is basically impossible unless it's your full time job (or you get lucky). I've tried. I kept finding methods that - based on a grep - were never called from anywhere, and there's no IDE support for something like Find All References.
I'm sure if it was your full time job you'd eventually learn the codebase, but there's no way you can just dip in and add a feature unless you really persevere.
But I did manage to add a few features to the gitlab-runner (used for CI) - because it's written in Go, and Go has static types and pretty great IDE support these days. Night and day.
I've also added a few features to VSCode which is a similarly huge codebase. Again it's written in Typescript which has static types and good IDE support. It would have been effectively impossible if that wasn't the case.
This does not match my experience at all, and I think your "near zero" claim is silly.
> difficult to deploy from source
I won't argue with you here. There are a lot of moving pieces in a Rails deployment. This isn't different from most web app frameworks, but it is difficult.
That said, I've never worked on a Rails app where deployment was any more difficult than a variation on `bin/deploy v123 production`, because I wrote that script and it works 100% of the time.
> and written in Ruby so the chance of being able to actually modify it for some feature you want is near zero
But this is still silly. You just don't know Rails or Ruby well, and don't want to learn them. Fine, but if you hadn't already made that decision, you would find the solution simple enough. No judgement intended -- different framework/language paradigms fit different people differently.
Rails has great IDE support also. Static typing can be a useful language feature, but a lack of same has not ever, in my experience, made it more difficult to understand real-world code.
There is a lot to love about Go too, don't get me wrong. But I would guess that the number of random developers who could drop in and be immediately productive in a Ruby/Rails app, vs a Go webapp, is basically equivalent. The overlap of projects where both would be highly appropriate choices is a bit thin.
[I hire into Ruby/Rails jobs regularly. I often hire senior developers with no Ruby/Rails background, but I do not hire people into these positions who are not open to learning. It takes a senior dev (from the C/Algol family) one day to learn Ruby, and (from a web dev background) a week or less to learn Rails. I have never seen a failure.
I also hire into Go jobs almost as frequently. The hiring criteria is a bit different (less emphasis on web awareness), but I do find it easier to teach Go to a Ruby dev, than Ruby to a Go dev. Make of that what you will.]
I am not trying to start trouble, or a heated debate, but I did want to say that my experience was the same as OPs and I am also coming from a static typing background so that likely explains my having a similar experience and expectations. I did for sure use RubyMine for attempting a change, so not "vim and yolo" but rather world class tooling and trying to discern where any random symbol came from was oppressively hard
But I was responding specifically to "in Ruby, so the chance of being able to actually modify it ... is near zero", which does not address the real issue.
It's perfectly possible to write simple, clear code in Ruby (and Rails!), but I'll concede that GitLab is not the best example of that.
If OP had said ~"... and the GitLab codebase is large and can be difficult to navigate and make drop-in contributions to ... also I have an aversion to dynamically-typed languages" :) ... then I wouldn't have bothered commenting.
I have an aversion to dynamically-typed languages because of these problems. It's not some random preference.
> You don't want to learn Ruby or Rails
Learning Ruby or Rails wasn't the problem. The Ruby language itself is fairly trivial. The issue is the lack of static types, and the fact that you can't even fall back to grep.
I know Python very well but it is almost as difficult to edit large Python codebases with no type hints. (It's not quite as bad because most Python code is greppable.)
Ruby and Rails work perfectly well for many many people, and you have chosen not to be one of them. That's a valid choice, but it's your choice and nothing more. It really isn't much about Ruby or Rails.
I grep through Rails code bases all the time. It is my first-choice method of discovery. In the very rare cases where it does not work immediately, I set a breakpoint and run from the REPL. This never does not work, even in the GitLab code base.
I have my criticisms of Ruby, and Rails, too. But your "near zero" comment is a shallow dismissal that captures your biases and presents them as some kind of informed truth. It is not.
I'll plug another option Gitpatch, however it's pretty early beta and not open-source yet, but most likely will be under AGPL at some point. It has built-in patch stacks (aka stacked PRs) and probably faster than any other Git host out there.
disclosure: I'm the author.
GitLab has a ton of options, And I find myself a bit overwhelmed by the user interface. It really needs a UX lead to simplify and create a better information architecture.
And, if you don't like something there's a very good chance you could be the change you want to see - they have a pretty welcoming contribution culture. Even if you don't want to change something, being able to read the source for it goes a long way toward aligning your understanding of the behavior, and that's not a diss on their usually pretty good documentation
