1. As others have said: BitBucket.com. Every time these guys are on here promoting this service they act like they've never heard of it. But they're in this space? Doesn't make sense.
2. I don't know how good I feel trusting my source code to a sorta-startup run by a small group of fulltime students. I don't think you should quit school to run this startup, but I do think this is an awfully serious service to try to offer in your spare time.
Yes, it's a developers job to ensure that he has backups -- and Git repos lend themselves to this very well. But that doesn't absolve you of the responsibility to treat your user's code as if they have no other copies and no other backups.
3. Your post about why the world needs another Git hosting service just explains why you think Github isn't the right fit. It doesn't really make your case.
4. When you're talking to engineers and you write this big setup paragraph about how much work it was just to share a git repo, and all of us are familiar with the myriad of hosting options, it just makes you look like you either don't know about your competition or you're hoping we don't. Not an effective strategy.
5. Github will give free repo's to students. And if you're not a student and you want a private repo paying a small amount for it is not a bad thing. In fact, it's a good thing. My code is valuable to me. I want to pay you a fair sum for a fair service.
I don't think you should quit what you're doing if you're having fun, learning, and believe in your idea. But GitHub is a huge brand with a ton of goodwill from the community. BitBucket is own by Atlassian which generally makes great things. Both have fantastic GUI's for Windows & Mac and great websites. So far I haven't heard anything that differentiates you. Don't treat me like I'm dumb, just show me why you're better.
1) We are aware of BitBucket; it's a great service, but it limits us on the number of collaborators we can have, thereby stifling productive group work. As college students, we're always collaborating, be it for assignments or our own side-projects, and BitBucket doesn't make sense for this.
2) Yes, we are college students, and I'm sorry you feel that college students can't be trusted, but I urge you to reconsider this opinion. Its not fair to apply such a generalization to all students.
3) We chose to address the primary service that we ourselves used for private repositories before Legit Teams. We understand that there are a myriad of services available for git hosting, but we are unable to address them all in one post; we limited the scope of the article so we could get our point across succinctly.
4) I think there is a slight misunderstanding here. When we described setting up git hosting, we meant on our own servers and not using a third party hosting platform. These paragraphs were simply laying the groundwork for why we created Legit Teams—they were the context.
5) We are aware of this; they offer a free micro-plan to students and educational institutions. This, however, still limits us to five private repositories. The whole point of Legit Teams is to get rid of limits on repositories and collaboration, which we feel are large pain points when developing with others.
Not necessarily a reason to avoid the niche, but the reasons you listed to enter aren't very good (namely, too expensive for lots of small projects) as BitBucket solves that problem.
Your price points are all wrong. No one cares about the difference between $3 and $6. In 2012 dollars, we're talking penny candy.
A single price point would be less confusing, but I don't really understand the appeal. Anyone who doesn't trust github and/or needs that level of secrecy will be willing to shell out the $10 a month for a dedicated instance that they can install git on themselves.
In short, I don't think you can be competitive with github because I think the frictional market you're trying to target doesn't exist in the capacity that you would like to think it does.
That's exactly what I thought. 2 paragraphs in I was expecting a Bitbucket mention and a reason why it didn't work for them and found none. Search the page and there's no mention of it at all. Then I wrote the article off as an uninformed rant.
For the segment of people you're targeting, there is very little difference between $3 and $6 a month, let alone $3, $4, $5, and $6 a month. We're talking portions of a Starbucks coffee here.
I'd focus on why you're different than the (already loved and highly used) existing options out there. Then I'd think about revamping the pricing to have each plan be more differentiated.
So why should we be paying approximately $1/repository when we don’t need all the open-source-focused features? And why does it make sense to pay per repository when the real cost is in space and IO? Our whole set of repositories probably aren’t more than 500 MB. And 500 MB, at current prices, costs merely a few cents. IO, in like manner, is negligible; we only push our repositories to the server once in a while. So why $10 or $20 a month?
If you EVER take one thing away from HN, 'tis this: NEVER compete on price. It is a race to the bottom. There will always be someone willing to provide the service for less than you. Just don't do it.
I would imagine that these guys are very sad and depressed right now, given all the harsh -- and correct -- feedback that they've gotten here. I'll try to add something different:
The various competitors in this space compete apparently on public/private visibility, price, storage quotas, and user+repo counts. If you compete on these dimensions you will have a very difficult time differentiating yourself -- there simply is not enough difference between one repo provider and another on these bases to matter. Cost of switching, I would imagine, is relatively easy -- `git remote add`, `git push`, and done?
There is also, perhaps, competition on UI/UX quality. This can be a differentiating factor for some, but I think not as strong as it is for other markets, given the strong technical background of the target market. If you had some original and compelling offer which had a poorer but workable UI (which improved rapidly over time), that could negate the advantage that e.g. Github has over you on interface.
I don't have any better ideas for this space -- the generic advice for you I would give is to innovate in some different, not-already-explored way. Can you provide something completely different from more users, more repos, more storage, lower price?
We are not sad; we realize this service is not meant for everyone and we understand that we'll receive criticism. In doing so, we hope to better Legit Teams.
We don't view this as a business that's going to make us millions of dollars so much as a product that we'd love for both us and others to use. It solves an issue we were experiencing and, at the least, we'll use our own service proudly.
Your points about competing on other dimensions do make sense, but as you seem to imply yourself, it is hard to find exactly what those other dimensions are. We're searching for this ourselves, but based on feedback we've received from our friends, this is at least a preliminary service that they'd like to use. We plan to iterate from here.
If your team is small enough then BitBucket is a great option for closed-source projects. The only difference is the lack of GitHub's interface and neat features like graphs. I wasn't too impressed with the code-review features for pull requests, but I don't remember GitHub being significantly better either.
Come on, there are many options in the market that suits well in every mentioned need, just to point, http://www.codeplane.com just right the same thing.
Ok, you can create a new service, but you shouldn't act like you're the only one around that do what you do.
When I was in College (just graduated a few months ago), GitHub applied a $7 coupon to my account, which covers a micro account. All I did was ask about student pricing. They should probably offer an educational discount with .edu email verification.
If you want private repo hosting on the cheap, I would suggest using Gitolite [1] on a VPS [2].
The current version includes self-service key management, regular users are able to create repos in their own namespace, and easy forking of repositories. It's extensible, you can write your own commands in any language.
Granted, it has minuses. It only has a command line interface (no GUI), requires a little setup and editing of configuration files and reading of docs.
I want my code stored in a safe place that sees regular backups in case of disk failure, or I'd throw it on my own VPS. Your service's website does not provide any information addressing security.
[+] [-] encoderer|13 years ago|reply
2. I don't know how good I feel trusting my source code to a sorta-startup run by a small group of fulltime students. I don't think you should quit school to run this startup, but I do think this is an awfully serious service to try to offer in your spare time.
Yes, it's a developers job to ensure that he has backups -- and Git repos lend themselves to this very well. But that doesn't absolve you of the responsibility to treat your user's code as if they have no other copies and no other backups.
3. Your post about why the world needs another Git hosting service just explains why you think Github isn't the right fit. It doesn't really make your case.
4. When you're talking to engineers and you write this big setup paragraph about how much work it was just to share a git repo, and all of us are familiar with the myriad of hosting options, it just makes you look like you either don't know about your competition or you're hoping we don't. Not an effective strategy.
5. Github will give free repo's to students. And if you're not a student and you want a private repo paying a small amount for it is not a bad thing. In fact, it's a good thing. My code is valuable to me. I want to pay you a fair sum for a fair service.
I don't think you should quit what you're doing if you're having fun, learning, and believe in your idea. But GitHub is a huge brand with a ton of goodwill from the community. BitBucket is own by Atlassian which generally makes great things. Both have fantastic GUI's for Windows & Mac and great websites. So far I haven't heard anything that differentiates you. Don't treat me like I'm dumb, just show me why you're better.
[+] [-] mey|13 years ago|reply
[+] [-] ayushsood|13 years ago|reply
2) Yes, we are college students, and I'm sorry you feel that college students can't be trusted, but I urge you to reconsider this opinion. Its not fair to apply such a generalization to all students.
3) We chose to address the primary service that we ourselves used for private repositories before Legit Teams. We understand that there are a myriad of services available for git hosting, but we are unable to address them all in one post; we limited the scope of the article so we could get our point across succinctly.
4) I think there is a slight misunderstanding here. When we described setting up git hosting, we meant on our own servers and not using a third party hosting platform. These paragraphs were simply laying the groundwork for why we created Legit Teams—they were the context.
5) We are aware of this; they offer a free micro-plan to students and educational institutions. This, however, still limits us to five private repositories. The whole point of Legit Teams is to get rid of limits on repositories and collaboration, which we feel are large pain points when developing with others.
[+] [-] dchuk|13 years ago|reply
Not necessarily a reason to avoid the niche, but the reasons you listed to enter aren't very good (namely, too expensive for lots of small projects) as BitBucket solves that problem.
Never ever ever ever compete on price.
[+] [-] debacle|13 years ago|reply
A single price point would be less confusing, but I don't really understand the appeal. Anyone who doesn't trust github and/or needs that level of secrecy will be willing to shell out the $10 a month for a dedicated instance that they can install git on themselves.
In short, I don't think you can be competitive with github because I think the frictional market you're trying to target doesn't exist in the capacity that you would like to think it does.
[+] [-] di|13 years ago|reply
https://github.com/edu
[+] [-] bjxrn|13 years ago|reply
[+] [-] w33ble|13 years ago|reply
[+] [-] sonier|13 years ago|reply
I host all my private work with them and have never had a problem.
[+] [-] timjahn|13 years ago|reply
I'd focus on why you're different than the (already loved and highly used) existing options out there. Then I'd think about revamping the pricing to have each plan be more differentiated.
[+] [-] glavata|13 years ago|reply
[+] [-] zxypoo|13 years ago|reply
Gitblit? http://gitblit.com/
[+] [-] oz|13 years ago|reply
If you EVER take one thing away from HN, 'tis this: NEVER compete on price. It is a race to the bottom. There will always be someone willing to provide the service for less than you. Just don't do it.
[+] [-] jessedhillon|13 years ago|reply
The various competitors in this space compete apparently on public/private visibility, price, storage quotas, and user+repo counts. If you compete on these dimensions you will have a very difficult time differentiating yourself -- there simply is not enough difference between one repo provider and another on these bases to matter. Cost of switching, I would imagine, is relatively easy -- `git remote add`, `git push`, and done?
There is also, perhaps, competition on UI/UX quality. This can be a differentiating factor for some, but I think not as strong as it is for other markets, given the strong technical background of the target market. If you had some original and compelling offer which had a poorer but workable UI (which improved rapidly over time), that could negate the advantage that e.g. Github has over you on interface.
I don't have any better ideas for this space -- the generic advice for you I would give is to innovate in some different, not-already-explored way. Can you provide something completely different from more users, more repos, more storage, lower price?
[+] [-] ayushsood|13 years ago|reply
We don't view this as a business that's going to make us millions of dollars so much as a product that we'd love for both us and others to use. It solves an issue we were experiencing and, at the least, we'll use our own service proudly.
Your points about competing on other dimensions do make sense, but as you seem to imply yourself, it is hard to find exactly what those other dimensions are. We're searching for this ourselves, but based on feedback we've received from our friends, this is at least a preliminary service that they'd like to use. We plan to iterate from here.
[+] [-] akurilin|13 years ago|reply
[+] [-] kidh0|13 years ago|reply
Ok, you can create a new service, but you shouldn't act like you're the only one around that do what you do.
[+] [-] tylerhowarth|13 years ago|reply
[+] [-] kcbanner|13 years ago|reply
[+] [-] tlrobinson|13 years ago|reply
I'll gladly pay a reasonable fee for hosted services so I don't have to worry about such things.
[+] [-] Apreche|13 years ago|reply
[+] [-] juddlyon|13 years ago|reply
- http://beanstalkapp.com/
- http://www.deployhq.com/
[+] [-] ericcholis|13 years ago|reply
[+] [-] csense|13 years ago|reply
The current version includes self-service key management, regular users are able to create repos in their own namespace, and easy forking of repositories. It's extensible, you can write your own commands in any language.
Granted, it has minuses. It only has a command line interface (no GUI), requires a little setup and editing of configuration files and reading of docs.
[1] https://github.com/sitaramc/gitolite
[2] http://prgmr.com/
[+] [-] jakebellacera|13 years ago|reply
[+] [-] azernik|13 years ago|reply
[+] [-] runako|13 years ago|reply
[+] [-] zmjones|13 years ago|reply