4 ms·

Suggestion for creating a startup: For private repositories gitolite is really awesome, but I think there should be a company about it. What I mean is that git

by antirez 16y ago

Suggestion for creating a startup: For private repositories gitolite is really awesome, but I think there should be a company about it. What I mean is that github or some other company should offer hosted gitolite instances, in the following form: You pay something like price_of_small_linode_instace+price_of_service, to get: A virtual machine with gitolite up and running, with a good web interface and online documentation. I think that the web interface part is the piece this company should not opensource. With this web interface you can create repositories and users, and modify the gitolite settings in general. A backup service, in some other different VM and network, so that if something goes wrong with your gitolite VM you can ask a free (and automated) restore, possibly with hour granularity for the latest 24 hours, and less granularity for bigger periods and so forth. Note that this gives the user virtually infinite repositories... if they don't upload movies to their instance, as there is a lot of space even in a small linode instance for git repositories. This way this is a "pure product" as there is no need to repair or install manually virtual machines, and the company can focus into evolving the offering, developing a better version of gitolite, of the web interface, and so forth. p.s. I mentioned linode but can well be ec2 or whatever is better suited for this stuff.

icefox 16y ago
What do people really think of this idea? Would you buy it? I have written a web interface for managing git repos called GitHaven (my personal home install is here: http://git.meyerhome.net:8080/ http://git.meyerhome.net:8080/). It has the usual goodies, private/public repos etc. I have some friend's companies that run it in the cloud already. Do you think it would actually make money i.e. Would you be willing to join me in a startup on it (in boston)? Would it pass the y-combinator test?
- tptacek 16y ago
  Probably not. Github is on a path to become the market leader, has a critical mass of customers, and has a business model that harnesses those customers to benefit from network effects. This idea is akin to building a business offering private white-label Stack Overflow clones. Key point: the actual software is not what makes Stack Overflow and Github valuable. Simpler reaction: Github:FI is already priced at a rounding error for the kinds of companies from which most of the revenue in this space will originate.
  - icefox 16y ago
    From the research I have done there are three major types of git server buyers: 1) We are so small that we don't even think twice and buy some private repos on GitHub. (i.e. no market there) 2) We are so big that we can't possibly even fathom the idea of having our code on someone else's server so GitHub.com is out of the question. Sadly GitHub:Fi is per/user and because we are really big that means we have lots of devs so it is too expensive in most cases. (i.e. enterprise sale yuck) 3) For medium sized companies that either have some devs (more then 5) or a ton of repos or large repos they want an alternative to github. Many are slapping up gitolite which while good is a far cry from github. Like I mentioned I already have found companies that want this solution and they are using GitHaven. The real question is if this market is too small or should I just move on to another project, open source GitHaven and let everyone (who wants to) slap it up as a frontend to their existing gitolite installs?
    - tptacek 16y ago
      Most products targeted at developers are per-seat priced (or have pricing derived from per-seat pricing); it's how you do value pricing when value scales with the number of developers. Look at how Atlassian does it; that's still effectively per-seat licensing (it requires prospects to think about how many devs they'll have). You can try to avoid these schemes, but you have to do it and not go out of business.
      - sdesol 16y ago
        Yeah the per-seat point is really a moot point when it comes to enterprise. ClearCase is $4000 a license and a lot and I mean a lot of large companies use it. I know of a company that is currently spending $2.5 million a year on ClearCase licenses and they really want to get off it but ClearCase is so entrenched/integrated into their work flow that this is not a trivial task. Github is not going to solve this problem for them. The problem with Github:Fi is the value proposition is extremely weak when it comes to meeting enterprise needs. - They have an issue tracker that doesn't support custom fields. This is critical for companies with 500+ developers and/or very diverse product lines. - They have a wiki but you can get confluence for $12,000 for the entire company which is more feature rich. It has decent searching. - They have blogging but this is of no value in the enterprise world. The majority of employees really don't have the time or care about blogging internally. - They have Gist which offers no real value for the same reason why blogging isn't that big of a deal. When it's all over, their only strong value proposition for the enterprise world is their intuitive push and pull interface. But the problem with this is they are competing directly with free solutions like: - gitorious - gitolite + internal web interface - gerrit I'm obviously speculating here, but based on how their pricing model changed: old: 1000 per user (initial purchase) + 200 something a year per user subscription cost new: 250 per user if less than 25 users and 200 per user if more than 25 users I think they realized they over estimated their value proposition initially. If Github wants to seriously break into the enterprise world, they are either going to have to lower their per seat cost and/or strengthen their value proposition.

by antirez 16y ago

icefox 16y ago

What do people really think of this idea? Would you buy it? I have written a web interface for managing git repos called GitHaven (my personal home install is here: http://git.meyerhome.net:8080/ http://git.meyerhome.net:8080/). It has the usual goodies, private/public repos etc. I have some friend's companies that run it in the cloud already. Do you think it would actually make money i.e. Would you be willing to join me in a startup on it (in boston)? Would it pass the y-combinator test?

tptacek 16y ago
Probably not. Github is on a path to become the market leader, has a critical mass of customers, and has a business model that harnesses those customers to benefit from network effects. This idea is akin to building a business offering private white-label Stack Overflow clones. Key point: the actual software is not what makes Stack Overflow and Github valuable. Simpler reaction: Github:FI is already priced at a rounding error for the kinds of companies from which most of the revenue in this space will originate.
- icefox 16y ago
  From the research I have done there are three major types of git server buyers: 1) We are so small that we don't even think twice and buy some private repos on GitHub. (i.e. no market there) 2) We are so big that we can't possibly even fathom the idea of having our code on someone else's server so GitHub.com is out of the question. Sadly GitHub:Fi is per/user and because we are really big that means we have lots of devs so it is too expensive in most cases. (i.e. enterprise sale yuck) 3) For medium sized companies that either have some devs (more then 5) or a ton of repos or large repos they want an alternative to github. Many are slapping up gitolite which while good is a far cry from github. Like I mentioned I already have found companies that want this solution and they are using GitHaven. The real question is if this market is too small or should I just move on to another project, open source GitHaven and let everyone (who wants to) slap it up as a frontend to their existing gitolite installs?
  - tptacek 16y ago
    Most products targeted at developers are per-seat priced (or have pricing derived from per-seat pricing); it's how you do value pricing when value scales with the number of developers. Look at how Atlassian does it; that's still effectively per-seat licensing (it requires prospects to think about how many devs they'll have). You can try to avoid these schemes, but you have to do it and not go out of business.
    - sdesol 16y ago
      Yeah the per-seat point is really a moot point when it comes to enterprise. ClearCase is $4000 a license and a lot and I mean a lot of large companies use it. I know of a company that is currently spending $2.5 million a year on ClearCase licenses and they really want to get off it but ClearCase is so entrenched/integrated into their work flow that this is not a trivial task. Github is not going to solve this problem for them. The problem with Github:Fi is the value proposition is extremely weak when it comes to meeting enterprise needs. - They have an issue tracker that doesn't support custom fields. This is critical for companies with 500+ developers and/or very diverse product lines. - They have a wiki but you can get confluence for $12,000 for the entire company which is more feature rich. It has decent searching. - They have blogging but this is of no value in the enterprise world. The majority of employees really don't have the time or care about blogging internally. - They have Gist which offers no real value for the same reason why blogging isn't that big of a deal. When it's all over, their only strong value proposition for the enterprise world is their intuitive push and pull interface. But the problem with this is they are competing directly with free solutions like: - gitorious - gitolite + internal web interface - gerrit I'm obviously speculating here, but based on how their pricing model changed: old: 1000 per user (initial purchase) + 200 something a year per user subscription cost new: 250 per user if less than 25 users and 200 per user if more than 25 users I think they realized they over estimated their value proposition initially. If Github wants to seriously break into the enterprise world, they are either going to have to lower their per seat cost and/or strengthen their value proposition.