I wonder if Github's ultimate plan is to make money from something other than charging for storage.
After all in the long run, while charging for storage is immediately profitable (and so a good way to sustain the business), there may be a bigger pie hidden a little further once you reach a critical mass of users...
Then it would make sense to charge as little as possible, to increase usage until you reach that critical mass.
I can't quite figure out what the bigger pie is for Github - but then I haven't spent that much time thinking about it. Anyone got some ideas?
Next up for GitHub is seeing how big the pie will be for our upcoming Firewall Install product, allowing companies to install their own copy of the site for internal use.
Sounds like a great idea (and definitely a huge pie), but enterprise sales can be very tricky. Good luck breaking in! May you forever doom the Visual Source Safes and ClearCases of that world to oblivion :-) They truly deserve it.
On the good side, you should be able to price very competitively, and these days, everyone will be looking to cut costs.
I didn't say that it doesn't work, I'm just saying that maybe there's a better business model hidden behind it. What if 50% of all the open source projects in the world and their forks are all on github. Does that open up some other ways to make money?
The disk allocation was very tight (especially for companies moving legacy repos to it) - it was 3gb space for a 'large' account intended for 25 users and 50 repos.
swombat is saying there may be a hidden business model that even the GitHub crew haven't discovered yet, because the conditions that enable it haven't yet been reached.
A great improvement, and I love my GitHub free account!
But I still can't justify using GitHub for my private source code. I've been coding for quite some time; my personal repository (along with related assets like images and sounds) consumes 1.7GB.
To hold that much data at GitHub, I'd have to purchase a Medium account at $22/month. But it only costs 36 cents per month to host my source code with jgit and Amazon S3!
Granted, GitHub has a lot of great features, but that's a pretty big gap for me.
Imho, you shouldn't be storing images and sounds on your SCM... that's not really what it's for. We store most of those externally, in a non-version-controlled environment. That really greatly reduces the space requirements (like, by a factor of 20). It also makes it much quicker to clone the repo on a new machine.
That is hard to stomach. If you're writing, for example, a game and said game has custom interface graphics, you certainly want to version those graphics along with you code. If you don't, it will be difficult to roll back to a working (and visually correct) past build.
If git is slow when I use it manage lots of binary content, that sounds to me like a problem with git rather than a problem with my workflow.
And, FWIW, git isn't too bad for managing all my data since I've broken my work into many repositories.
Oh, git isn't slow when storing binary data on it.. it's slower to clone (because you have to download all that data again), but not to use.
However, unless you are building a game or something like that, storing binaries on git will really bloat your repo so it's worth avoiding it if reasonable.
If it's assets that you do need to revert to, fair enough. But there are many "compiled" binaries that don't really need to be in the VCS, imho... and if you can err on the side of including too few binaries, you'll save a lot of space.
I would not agree on that, it seems to me that github is used differently than sourceforge or the like. The vast majority of github repos is ruby and javascript of which the majority is most likely rails related. Lots of people also use it to host their private scripts and little tools, you don't have that on sourceforge where most of the big opensource apps are hosted. Maybe Rubyforge will suffer. Anyway, I like github a lot and appreciate the upgrade, but I don't think it will overshadow sourceforge anytime soon.
After all in the long run, while charging for storage is immediately profitable (and so a good way to sustain the business), there may be a bigger pie hidden a little further once you reach a critical mass of users...
Then it would make sense to charge as little as possible, to increase usage until you reach that critical mass.
I can't quite figure out what the bigger pie is for Github - but then I haven't spent that much time thinking about it. Anyone got some ideas?