Yes, I heard this rumor right after I moved from github to gitlab. Well if I have to go elsewhere at least gitlab will archive my abandoned free account for me :)
FWIW, I found them easier to deal wit than github, so will hang tight to see how this plays out.
Nothing they’ve done since they were created has ever moved them in a more open source friendly direction, and they’ve broken a ton of promises both implicit and explicit along the way.
GitHub OTOH has only become more open source friendly (minus the AI stuff, but I suspect Gitlab is no better on that front).
One feature area where Gitlab is still better for realworld stuff is CI (Gitlab CI vs Github Actions). Yes, you can do most things on both, but Gitlab CI makes a lot more sense.
In general, Github still feels like it's built for hobby coders (focusing on simplicity instead of configurability - which doesn't have to be a bad thing) while Gitlab feels like it's built for professional teams from the ground up.
I have used Gitlab CI basically daily for over 5 years and it makes sense. I would need to think hard to come up with something that seems fundamentally wrong.
I have never used Github Actions. Can you explain or give some examples what doesn't make sense?
IIRC Github Actions started as a 'visual editor' where you would drop and arrange 'Actions' and define the data flow between actions, but what most people want from a CI system is just a script/config file in their git repo which defines what command line tools to run, and to group those commands in jobs dependending on each other so that some jobs can run in parallel (which Github Actions only implemented as an afterthought after users demanded it).
To reuse code, Gitlab CI has simple template files which you can import into your toplevel .gitlab-ci.yml, and you have an inheritance system to derive new jobs from other jobs. That's a very simple and powerful system.
Code reuse in Github works with above mentioned 'actions' where each action seems to be a whole repository of stuff instead of a single file like in Gitlab CI.
Gitlab CI seems to be designed by people who know what they do and what their users need, while Github Actions seems to be designed by architecture astronauts, and has only afterwards and reluctantly been hammered into a shape where it does the things most users expect.
GitHub Actions feels like it was first designed to let people customize the GitHub Pages deploy flow (since GitHub by default only offered Jekyll as a static site generator, and Jekyll is Ruby tooling and not lightweight to run at all) and as a CI tool second, being molded into behaving like one after Travis CI went bad for open source projects.
Gitlab CI actually seems like it was made for CI in the first place.
I'm pretty sure of the same, and that feature actually leaks into the implementation. Right after the initial introduction of the current Github Actions, we translated some Azure Devops scripts to Actions and a lot of the structure and most keywords where nearly identical. As well as the interface when running the CI.
If I remember correctly GitHub CI is pretty much a straight port of Microsoft’s existing Azure DevOps CI, done pretty soon after the acquisition. The rest of Azure DevOps UX is kinda insane so it’s no surprise the CI is a bit of a pain too.
Try and test a Github Action locally - it's an engineering project up there with the Space Shuttle.
Repositories around the world are filled with endless commits of "test1", "test2", "test3" trying to debug their actions in prod.
Right, but Gitlab does have the excellent built-in pipeline editor that will visualize and validate your pipelines for you.
It can also render the complete pipeline config (making it easy to run and debug the problematic parts locally just by copying the relevant parts, even if they're hidden in and include somewhere).
I find them both equally bad just in different ways.
Compare the gitlab UI with phabricator for example. The workflows are mostly a strange mixture of whatever github made up on the back of a napkin and Stakeholder-consultant slop.
GitLab has accepted my patches…do you have a timeline for when Github will do the same? Sure, maybe the directions are different, but the baselines couldn't be more different either.
Neat…though considering how far removed it is from the actual behaviors of the forge rather than things that are essentially "bikeshed topics", I'm still not very convinced that Github is even in the same league as GitLab in "OSS friendliness".
In my experience, Gitlab is a lot more stable than Github. My last job was on Github, and we had an outage a couple times a month at least. We even had a Slack emote for it! My current job is on Gitlab, and we haven’t had a single outage in the year that we’ve been on them.
Second that. My job before last moved to a locally hosted enterprise Github instance, which promptly ate itself. The specs required to run it were also impressive, something like 64gb minimum to boot but more was strongly recommended.
Haha, I keep getting burned by GitHub outages even as a private contributor with my personal account... speaking of which, I expect one outage soon, this week so far it has been available always when I needed it...
GitHub is open source friendly on paper, but almost nothing they do is actually open source, or even source available. Contrast that to GitLab who are actually open core, and the vast majority of their software if publicly available for free with a very permissive license.
One talks about open source because it's the de facto home of open source. The other is actually open source.
I'll wait to jump ship until I see who buys them. It could end up being a huge positive for a gitlab. I have been very disappointed in their strategy the past few years and I think they squandered an enormous opportunity and amount of Goodwill with developers. If they got bought by somebody good, then I think it could end up being a massive positive.
I used Gitlab at a previous job maybe four years ago and really liked the UI. Switching to Github at the new gig felt like a huge step backwards. That said, the product and business news I have seen regarding Gitlab since then has almost all been negative. Hopefully they are able to turn things around because at one point I really hoped they would overtake Github and thought it might happen.
To elaborate a bit more; first things first - Gitea is still MIT and open source. Not open core, full open source.
The main reason for Forgejo is moreso that Gitea as a project was taken over by a company instead of being run as a non-profit. Some of the dev team felt uncomfortable with that and forked it.
Personally I haven't seen much reason to switch from Gitea to Forgejo - this is the sort of ideological issue that I'd rather kick the can down the road on until Gitea Ltd goes bad (and in an assumption of good faith, I'll assume that it won't.)
It's not that difficult to move git repositories around after all.
The ideological difference between the two projects really shows on their landing pages. Forgejo has a cute fox drawn by a real artist whose name is credited in the website's footer; Gitea has AI-generated images of a robot in the clouds or in a skyline (it becomes really obvious when you look close)
Indeed. And the cute fox almost doesn't need the credits. When you recognize David Revoy's style on a project page, you know the project is probably a community-driven effort, and is worth checking out if you value that.
For years, "self-hosting" Gitea wasn't done because it was missing a bunch of useful collaboration features. Now, it looks like that gap has been closed. All of the specific features mentioned in that issue seem to have been fixed, and the big remaining task is figuring out below to actually migrate all the existing data out of GitHub -- which doesn't seem to be super high on the priority list.
Many options , older companies like IBM, Google, SAP, Oracle or even Salesforce (already own heroku in dev tooling space so not far fetched ) with stable or slowing market presence in engineering departments
Mid sized newer companies likes Hashicorp or datadog or vercel who target developers as customers .
Gitlab gives access to a large audience of developers to cross sell most dev tools so all these orgs can get a lot of returns paying more than the standalone value of gitlab itself.
The best fit would be companies like Hashicorp who have strong open source pedigree so users won’t be turned off and leave
Yes they did, i should have clarified, as IBM is becoming like Broadcom as an umbrella organization for all sorts of companies, the ibm core is different beast than some of the acquisitions they have been making
In my mind just like LinkedIn , GitHub and Microsoft are every distinct entities with a lot of differences on how they work , Hashicorp and IBM parent are different and will remain so. Integrating into Hashicorp for Gitlab would be very different than integrating into IBM core with different values for both businesses .
I wasn't thinking about a few focused features or integrations, but more generally. i.e non product things like sales and license packaging and son on.
If an acquisition has to make sense there should be a clear path to monetize it, for IBM core or its HashiCorp unit or any other buyer that will not just be through some light integrations alone, they can achieved with partnerships after all you don't need to buy the organization for it.
I seriously don't understand the deals being made in tech. Most of the makes no sense, not even retrospectively. I get Microsoft buying Github, that was a part of their open source strategy and they've always put a high value on developers.
The way the present their numbers is pretty hard to understand, at least for me, did they lose $28million or make $28million in that same quarter? Either way that seems insanely low, if they're expected to be worth $8billion. The gap between profit and revenue seems to high.
There might be some potential for Gitlab complement your other business, in which case you may not see the lack of profit as that big of an issue. The problem is that if you can't make those $8billions back in future profit, then you're going to start making changes to the Gitlab offerings until they do become profitable.
That might be what the new CEO is suppose to do, pump up those numbers, and make it look like a sane investment.
I was thinking about that as well, given that it seems it would fit in well with the red hat portfolio. They don't as far as I know. Have a good answer for a gitforge, and the phenomenal CI CD offering that gitlab has would be very marketable to Red hat customers.
I would be excited if IBM acquired them and put them under the red hat umbrella, because as history has shown, it may mean that gitlab ends up becoming much more open. They may open up the entire product instead of doing the open core model.
This makes a lot of sense, and is truly frightening.
Google isn't know for its hands-off approach nor long term view for service growths. Gitlab is essential to balance Github's impact, I'd hate it to go in the graveyard.
