Hacker News Reader

Comment by Analemma_ | Hacker News Reader

If they can't stand on their own feet, I imagine somebody would/will buy GitHub at a firesale rather than see it disappear completely. For all their missteps, they have developer mindshare that is the envy of everyone. If we're lucky it would be someone like Google or Microsoft, if we're unlucky it might be Oracle or SalesForce. Whoever it would end up being though, GitHub won't just vanish.

View on HN

Ignoring the fact that the VCs want to get paid and GitHub's employees want to keep working, how big of a skeleton crew would it take to keep a site like GitHub running in "maintenance" mode?

Ask Craigslist, or soon Twitter?

Can you really put something in "maintenance" mode without killing it?

Maintaining the app and not moving forward with new features and adding value (because of lack of resources), I would assume it would just die eventually.

I don't understand this obsession with the first derivative of value. GitHub already provides loads of value. I understand they need resources to keep the lights on, but they could do that by recouping a tiny fee from their users and most everyone would be happy.

don't they already attempt to do that though? i don't consider $7/month a large fee. (granted, they could probably get more subscribers at $3-5, but i'm assuming they landed on $7 for a reason)

They make vast more quantities off of actual businesses. The $7 repo fee is literally a soft wind on the face of Everest.

They're not without alternatives already though. In a changing environment to stay still is to fall behind.

That's not as big an issue when there are network effects -- look at Craigslist. Plus, what killer feature is bitbucket going to add at this point to make everyone want to switch? They're practically the same product. All the major innovation happened years ago.

Clickable code view would be nice, e.g:

http://lxr.linux.no/linux+v4.9/kernel/async.c

This would be wonderful.

Thanks for the idea. This would be awesome indeed. I've created an issue on GitLab for this: https://gitlab.com/gitlab-org/gitlab-ce/issues/25762

And to move is to risk a quick death. I wished existing brands stopped innovating so much. That's e.g. what got Pebble killed.

Doubly so when a brand reaches infrastructure level - like Github, or arguably Dropbox. Just don't screw with the product that's working well. Not breaking things isn't expensive.

...it's a source control company. What additional features could I possibly want? They serve their purpose just fine.

I really hate this, because this is what bloats and kills products. It's never complete, it always needs more features, until it implodes.

It's the corpse model of product development.

Post VC investment, GitHub would never be allowed to exist in maintenance mode. It would be liquidated in one form or another if things were that dire. Once they took the venture capital, the possible outcomes narrowed considerably.

They could go non-profit like the Mozilla foundation.

But not if the owners (investors) want their money back instead.

I appreciate how GitHub's software platform makes migration elsewhere much easier, making me less leveraged by GitHub the company.

That, ironically, makes me REALLY not want them to go under: they don't deserve to.

Isn't the migration easier because it's just a git repo? I'm curious how GitHub's software platform specifically facilitates easy migration.

The comprehensive API helps, though that can theoretically be limited if a ton of people try to migrate.

The git repo itself is trivial to migrate (you already have a full copy, just push it somewhere else).

The problems are issues, code comments, pull request discussions, etc, which may be harder, and their API helps here.

Truth be told, an API is an easy way to get app developers to include support for your platform; you aren't going to save yourself from the reality of web scrapers in any case.

The API certainly helps. Our GitLab can import all the things you mention. https://docs.gitlab.com/ce/workflow/importing/import_project...

Or we could actually luck out, and settle on GitLab or similar.

Not that I hold out too much hope here, but given GitLab's emphasis on the enterprise and self-hosting, it would be amazing if they could put some effort into making disparate GitLab installations behave like one big network (think Six Apart with trackbacks/typekey/etc way back in the day).

Github's huge win is the network effect of having one account and being able to interact on thousands of issue trackers, create PRs, etc. It would super cool if GitLab could achieve some of that same experience without requiring everyone to all be on gitlab.com. It would go a long way toward addressing the criticisms that Github is too centralized.

> Not that I hold out too much hope here, but given GitLab's emphasis on the enterprise and self-hosting, it would be amazing if they could put some effort into making disparate GitLab installations behave like one big network (think Six Apart with trackbacks/typekey/etc way back in the day).

Yeah, I think it'd be awesome if there was some form of optional federation between GitLab instances (sort of like Matrix or NextCloud).

IIRC Gogs/Gitea and GitLab have plans to enable pull request federation, so you could use your own instance to make pull requests on a completely different one.

I also recommend to check out Gitea (community fork of Gogs) if you haven't yet, it's a nice alternative to Gitlab for selfhosting.

Federated merge requests (pull requests) is something we've been wanting to do for a long time [0]. Maybe 2017 is the year to do it.

[0]: https://gitlab.com/gitlab-org/gitlab-ce/issues/4013

See my comment below, we'd love this and look forward to everyone's ideas on this [0]

[0]: https://gitlab.com/gitlab-org/gitlab-ce/issues/4013

I'm not sure Google or Microsoft offer hope. The former seem like it would soon shutter it as not returning added value to shareholders and the latter like they'd subtly degrade access from non MS platforms until people wished Google had bought it and shut it down.

It seems like other smaller companies have exploited the profitable parts of Github's niche out from under them.

I think I here a bell tolling.

> the latter like they'd subtly degrade access from non MS platforms until people wished Google had bought it and shut it down.

Maybe I'm too nice to them, but that doesn't seem like Microsoft's modus operandi anymore. These days they're all about getting people to buy more CPU cycles from Azure, rather than trying to seriously pursue Windows platform lock-in (which is a lost cause and they know it).

I imagine if Microsoft did buy GitHub, they'd have all kinds of offers like "get automated tests and CI to your Azure machines every time you push", or allowing you to host the repo on your own Azure VMs but still use the GitHub interface, stuff like that. They'd probably keep the free GitHub mostly as it is, with some nagging to upsell premium Azure stuff but nothing worse than that.

As for Google, I agree that they tend to get easily distracted and drop things, but I would hope that GitHub is popular enough (both inside and outside Google) that there would be serious pressure not to let it die on the vine. Contrast with abandonware like Wave and Google+ which never got enough mindshare that anyone felt like really fighting for it.

People _relied_ on Google Code

I figured MSFT was going to buy them and have it integrate nicely with Azure. They could do hosted enterprise github's natively on azure and would outclass code commit on AWS. Instead they bought linkedin which doesn't make sense to me.

Microsoft already has Visual Studio Team Services, which is its Git-compatible source code and work-item-tracking system. Like GitLab, VSTS is focusing on enterprise sales first (though there is a free-tier that you can play with). Why would Microsoft want to buy Github and try to refocus it when it already has a product that caters to its enterprise clientele?

Because developers like GitHub?

What missteps?

Supposedly competitors may be cheaper or offer more in the free tier, and I could imagine you'd call Git itself a 'misstep', but otherwise I can only think of incredible cultural and technical achievements by GitHub?

I love GitHub (product + company), however over the past few years there's been a lot of very public internal issues.

That being said, the product itself is stellar.

The political issues inside the company are a lot more troubling than anything involving the product, I agree.

I agree somebody will buy them before they disappear, but I strongly disagree on the purchasers.

I'll take Oracle, SalesForce, or Microsoft over Google any day of the week. The worst Oracle will do is start charging me more, but god only knows what kind of sleazy ad tracking Google will add. No thank you, I'll move everything over to BitBucket.

Fortunately, I don't think Google will bother. They already shutdown Google Code when they couldn't use it to increase ad revenue, so hopefully they'll leave GitHub alone.

The options you named all use google analytics

They even let you add your own analytics key

SalesForce bought heroku and heroku seems to be doing well.

Reply on news.ycombinator.com