Hacker News Reader

Comment by the__alchemist | Hacker News Reader

We are in the disapora phase; there is a steady stream of these announcements, each with a different GitHub alternative. I speculate that within a few months, the communities will have settled on a single dominant one. I'm curious if it will be one of the existing ones, or something new. Perhaps a well-known company or individual will announce one; it will have good marketing, and dominate.

View on HN

This has been going on for a decade, at the beginning it was projects moving to Gitlab now there's a lot of alternative projects but GitHub is still the only one that counts for discoverability. This is a very small minority of projects that move away from Github and it's way too early to declare GitHub doomed.

No different than everyone talking about the next “iPhone Killer” when someone other than Apple releases a phone. Although, I think that rhetoric has largely died down.

Gitlab did seem like a hope. But they very quickly became an even more massive and slow SPA javascript app than even github was.

Github is fine for discoverability but as a development platform I think it's going to die. Public issues/PRs are a cesspool now and going to get worse, and agentic workflows are going to drive companies to want to hide how the sausage is made. People will gradually migrate to alternatives and mirror to Github while it remains relevant.

I always stayed with GitHub because it just worked the best. GitLab was slow and janky. gitea and its various forks lacked features and felt a step backwards. Sourcehut workflow is far too opinionated for my liking. Don't even get me started on GNU Savannah.

Some parts of the Free Software/Open Source crowd has always bemoaned the rise of GitHub, "because obviously you should use Free Software, its your ethical duty!" Most people just use what works best, including many Free Software devs. There is a loud minority (even louder in bubbles like HN) but for most people it's just one factor out of many, at best.

The reason GitHub became dominant is fairly simply: it just worked the best. Doesn't mean it was perfect (remember how long it took for line numbers to not be copied from code examples?) but the alternatives were even worse.

It's interesting to see how badly they're messing it up. You'd think that making a new react-based frontend for a fairly uncomplicated issue tracker wouldn't be too hard, but seems like it is. Some initial bugs after a rewrite are normal, but ... it's been a year? I still regularly just see closed issues in my issue overview. The back button is basically just broken. These are not obscure heisenbugs: these are bugs you find after using it for five minutes. The entire experience is just so janky.

I don't think Github is dying at this moment. I do think that the regression of UX quality is a necessary pre-condition for its death. Like many things its death will happen "very gradually, and then suddenly all at once". Sourceforge once seemed omnipresent and that changed very quickly.[1] But who knows where things will end up in five or ten years?

[1]: I'd like to pre-empt the inevitable "that's because of the adware" comment that someone always seems to post: that's a false history. The adware happened well after it already lost its position and was the desperate attempt of a declining struggling platform for income.

I’d guess most revenue comes from enterprise accounts which are not public.

Different devs have different preferred ways to work and collaborate. I doubt the FOSS community will converge on a single solution. I think we’re at a point of re-decentralization, where devs will move their projects to the forge that satisfies their personal/group requirements for control, hosting jurisdiction, corporate vs community ownership, workflow, and uptime.

This is due to increasing competition in the source forge space. It’s good that different niches can be served by their preferred choice, even if it will be less convenient for devs who want to contribute a patch on a more obscure platform.

The bigger question is whether we want a single dominant replacement, or whether it just means we'll be back in the same place in 5 years.

> I speculate that within a few months, the communities will have settled on a single dominant one.

The solutions on the roadmap are not centralized as GitHub. There is a real initiative to promote federation so we would not need to rely on one entity.

I love this, and hope it works out this way. Maybe another way to frame it: In 2 years, what will the "Learn Python for Beginners" tutorials direct the user towards? Maybe there will not be a consensus, but my pattern-matching brain finds one!

> I speculate that within a few months, the communities will have settled on a single dominant one.

I really hope not. Heterogeneity is really valuable in this space, and there's really no "one size fits all" model.

I really liked GitHub and I would also pay more for it, but that does not seem to be a priority. On safari the whole PR review is barely useable any longer because of bad performance without gaining any discoverable to me new features. Obviously a lot of man hours went in to ruining the product but I can’t understand why

On the upside I guess they use git internally as well so maybe they could just find a usable commit and revert all the crap they changed

Isn't that pretty much GitLab? But then most people still prefer GitHub anyway.

GitLab is too heavyweight for many projects. It’s great for corporations or big organizations like GNOME, but it’s slow and difficult to administer. It has an important place in the ecosystem, but I doubt many small projects will choose it over simpler alternatives like Codeberg.

Gitlab is worse than GitHub in every way.

At least GitHub adds new features over time.

Gitlab has been removing features in favor of more expensive plans even after explicitly saying they wouldn’t do so.

Gitlab works fine for me. Been using it at work for a few years and recently moved all my personal repos there

> At least GitHub adds new features over time.

Not as quickly as they add anti-features, imho.

Personally, I prefer the CI/CD setup on GitLab over GitHub Actions.

Horses for courses I guess ¯\_(ツ)_/¯

Gitlab is part of the reason I'm thinking along these lines: It has been around for a while, as a known, reasonably popular alternative to GitHub. So, I expected the announcement to be "We moved to GitLab", Yet, what I observe is "We moved to CodeHouse" or "We moved to Source-Base" The self-hosting here with mirrors to two one I'm not familiar with is another direction.

I think people are wary of moving to gitlab because its a similarly large platform and dont want to repeat their mistakes

gitlab has also gone full slop

It looks like all that they’re doing is griping over frontends and interfaces to do all the custodial work other than version control (ie., all baked-in git provisions).

How do you speculate the candidacy for email.

The settling on a dominant one does not happen - self-hosting becomes more popular.

> Perhaps a well-known company or individual will announce one; it will have good marketing, and dominate.

Hah, exactly what we’re attempting with Tangled! Some big announcements to come fairly soon. We’re positioning ourselves to be the next social collab platform—focused solely on indies & communities.

... and it will be SourceForge. finally.

Reply on news.ycombinator.com