Before GitHub

> What GitHub Gave Us

To me one of the clear things that GitHub gave us was a structure around a person rather than a project. To me it felt liberating to quickly create a repository attached to my name than it was to go through the (what felt to me) very serious process of coming up with a project name and reserving it on sourceforge just to get a cvs or svn repository (along with website, mailing lists, issue tracking(?), etc, etc...). It felt like the mental load of "oh this is just a quick thing" was a lot easier with github.

> It gave projects issue trackers, pull requests, release pages, wikis, organization pages, API access, webhooks, and later CI.

Although it didn't give us this all at once. I still remember when we created a new user account in order to simulate an organisation, before they existed. I distinctly recall discussing with friends if we wanted to set up a bug tracker software for our project with the assumption that "GitHub will probably release one in a few months anyway". In the end we just kept a text file committed in the repository. Issues were announced a few months later.

I am still so salty that Git won out for the average project over Fossil. Sure Git has some performance advantages for massive codebases like the Linux Kernel, but the vast majority of projects will never run into performance limits from their VCS. Fossil’s internal tools (wiki, forum, tickets<issues>, etc) are just so useful to have versioned with your code in one file.

I use Fossil for all my freelance work and it so easily allows me to get right back into the context of a project, niche details and agreements had with a client, etc. No need to pollute the codebase or gather together a million emails or notetaking software just to get back up to speed.

It can still change, I hate the notion that because Git is so culturally embedded we couldn’t ever switch. Fossil makes it super easy to switch and the workflow is actually easier coming from Git.

> But maybe the most underappreciated thing GitHub did was archival work: GitHub became a library. It became an index of a huge part of the software commons because even abandoned projects remained findable.

I think this is a bad thing actually. Having something that's centralized but helpful-99%-of-the-time atrophies our collective archival skills. If everything had to be seeded by someone to keep it alive, everyone would be better at holding on to their copies of the things they really cared about instead of being able to assume they can just check it out again when they want to.

There should be no single place that something can be taken down. When a project on GitHub gets DMCAed it takes everyones' forks with it too. Just look at what happened when Nintendo took down the popular Switch emulators in 2024, where archival/continuation efforts consisted of people figuring out who had the latest revision checked out and sharing it. That was only possible because they were very popular project: https://news.ycombinator.com/item?id=40254602

Aside: I really love the Splatoon-ish header/footer animation on this site! Very unintrusive, adds a lot to the vibe, and also quickly gets out of your way as soon as you scroll down. I'm totally going to rip this off lol

I absolutely loved Trac. Getting a Trac setup as step 1 in starting a new open source project was just an unbelievable amount of friction.

Fun fact: Django is still running on Trac today, and has been for more than 20 years now: https://code.djangoproject.com/timeline

(I was not involved in setting that one up, though it's possible I helped get the private Trac that pre-dated it running, I honestly can't remember!)

Reading this and mitchellh's post I was curious about code archival services, and found a few projects.

GitHub has their own: https://archiveprogram.github.com/

Software Heritage is a non-profit funded by UNESCO: https://www.softwareheritage.org/2019/08/05/saving-and-refer...

Although they're mostly the code / commit history, not so much surrounding metadata like issues, PRs, discussions, wiki, etc.

This got me thinking about code.google.com, I can't believe Google dropped the ball that hard.

I think I was one of the first people to try Flask. I learned Python so I could take advantage of AppEngine for free and easy modern hosting, which put me in the right spot when Flask launched. I've long been an admirer of Armin's, and recognized his domain before I clicked the link. As he points out, in those days, you didn't default to GitHub.

His post is a response to Mitchell's, from just a few hours ago. I'm impressed with how quickly he wrote a long-form, high-quality, well-reasoned post.

An important part to me that gets overlooked is shared logins. Rust runs the tests of all known Rust projects in a tool called `crater`. I was analyzing a run identifying projects relying on internals of Cargo and opening issues. When making 200 issues by hand, it is a big help when the process is low friction: I had credentials for the site and allowing blank templates. Any time I came across a self-hosted instance, I usually ended up giving up.

It's fun to read stuff like this and then reflect on the journeys of the projects I've been involved with. Most of my open source work has been done with self-hosted infra. My main example is Xfce: back when I started with them in 2004, we had a SVN server, using (I think) CVSweb's then-new SVN support for the web interface, and... maybe that was it?

My memory is telling me that I set up Bugzilla at some point after I joined the core team, though that may have been someone else. When git started becoming a big deal, I spearheaded converting our big SVN repo into many git repositories, and set up the cgit web interface for it. We were still using Bugzilla at that point.

I left the project in 2009 or 2010 or so (I'd joined a small startup and didn't have much time for OSS, sadly), and rejoined in 2022. In the intervening years they'd stood up a GitLab instance with CI runners, and had migrated everything from Bugzilla to GitLab issues.

It's still a very small team (handful of active people), and the infra is mostly managed by one person. It's all very doable, even for small teams. We're very lucky that our infra is generously donated/sponsored, though we also probably get just enough in regular donations that we could pay for it ourselves if we had to. I really appreciate that we're not dependent on Github/Microsoft for anything. Seriously, if you told me 20 years ago that Microsoft of all companies was going to own the largest OSS code forge in the world, I would have thrown up. It still doesn't sit well with me.

What is GitHub’s decline? I’ve used it extensively since 2011 ish (to lazy to look up when) it’s only gotten better. What’s the issue?

What we need is gitlab to finally integrate ActiviyPub so we can fork, comment, open merge request on all gitlab instances from our personal instance. Git is already decentralized, this isn't that hard to do.

I remember this old thing called Bugs Everywhere -- it was a bug/issue tracker which actually lived inside your hg repository. I wonder if we could standardize on something like that? or git notes with an issues ref? or something magical like that?

Then it's BYOR -- bring your own renderer. Trivial CLI bugtrackers, agentic nonsense, pretty web stuff, whatever and the data lives in the repo.

Smaller, decentralised forges actually make lots of sense from a digital sovereignty point of view. Over reliance on a single instance like GitHub is not healthy in the long run. The issue they would have to solve is federation.

I really worry about a bunch of people going over to codeberg. Site's already super slow, but apparently it's quite nice when self-hosted

Anyone who is able to just plop a forgejo instance on their own machines... please do that if possible!

If it wasn’t for SourceForge I’m not sure my life would’ve ended up where it is! They use to promote projects they liked and ended up putting Waterfox on their front page a few times. Really sad when they started blasting people with ads and swapping out installers with adware for popular projects. By that time I moved to Microsoft’s CodePlex, if anyone else remembers that? Felt like I was the only one using it at the time! I remember the connection speeds to it were atrocious, but appreciated they’d share ad revenue from the downloads of a projects page which was nice. I remember it was actually super expensive to offer downloads [for binaries] back then, using these code hosting websites was the only way to do it for “free”

I was posting on Planet Source Code before SourceForge existed.

Great through-history write-up! Thank you.

> That is one of the great ironies of modern Open Source. The distributed version control system won, and then the world standardized on one enormous centralized service for hosting it.

Cycles everywhere indeed. Perhaps we should ->

> GitHub wrote a remarkable chapter of Open Source, and if that chapter is ending, the next one should learn from it and also from what came before.

Indeed! Try to learn from the inevitable iterations to make the next instance at least that slightly better.

... Where the stuff meets the metastuff it seems all works under very similar forces. My thinking is step-by-step - it works on the individual level, and it scales up.

Day to day is step by step and a step today funds the step tomorrow.

This “slowly dying” effect is what happens to every company that gets acquired by big monster slug companies like Microsoft.

What Microsoft acquisitions still have any of their original spark left in them? Or Oracle? Or IBM? Or Google? Etc…

Hell, some Microsoft originals from inside the company like Xbox have even lost their edge.

Money is great and I’m sure I’d take the big check, too, but I’m surprised more tech founders don’t think of their legacy in this way when they decide to sell out.

It’s considered a grand accomplishment to essentially lead the wonderful thing you created to its slow demise and hand it over to apathetic quarterly earnings zombies.

Radicle is a good answer, coupled with a reborn Usenet, maybe Nostr. We have like never before the ability to communicate and cooperate yet most fails to understand and implement that.

Nearly any of us could run an XMPP/Matrix server and federate with friends or Nostr/{0xchat,whitenoise}, all with audio, video, text, file exchange etc, yet less than 1% do that.

Simply people, techies as well, have forgot the meaning of personal ownership and therefore are owned by someone else.

Maybe something like https://fossil-scm.org/home/doc/trunk/www/index.wiki is worth another look. Although Forgejo is probably an easier switch from GitHub. If their federation ideas play out well, that could be a good outcome from all this

If everyone moves off GH, it'll just go back to normal again?

[dead]

> Before GitHub, Open Source was a much smaller world.

Not that much smaller right-before GitHub and right-after it became available.

> but in the number of projects most of us could realistically depend on.

Most FOSS I realistically depend on I don't obtain from GitHub actually.

> There were well-known projects, maintained over long periods of time by a comparatively small number of people.

There were even more not-well-known projects, maintained for less time, by a larger number of people. They just weren't that many of them in one place.

> You knew the names.

You absolutely did not know the names. Post author is just thinking of the names they knew as though those were everybody.

> reputation mattered in a very direct way.

And now it doesn't?

> We took pride (and got frustrated) when the Debian folks came and told us our licensing stuff was murky or the copyright headers were not up to snuff, because they packaged things up.

RedHat was just as popular a distribution; and most users used Windows (like they do today); and the BSD distributions were a thing (although we didn't have Apple's BSD, i.e. MacOS)

Bottom line: Inaccurate description of history.

GitHub is fine. The blog post gestures vaguely at nothing. You put stuff on GitHub. You host repo. You run Actions. It all works.

So what is the issue? It's just nonsense, brain rot.

I'm working pretty hard on building what comes after Github, but I'm going full-tilt boogie and trying to also work out what comes after Git.

I'd love to have a longer conversation with you about how we can seed a better system, because on the off chance I'm successful I have a once-in-a-generation opportunity to fix past mistakes.