SM: CollabNet is a commercial company which secures traditional business contracts, right? Companies pay CollabNet versus CollabNet being an OpenSource provider? How does that all work?
BB: At one level we are software as a service. We charge for access on a per-user, per-month basis. Over the past few years we have developed our processes to the point where we can also run this on a customer’s site or network for them and if needed we can also give a client the rights and permissions to run it on their site or network independently. Regardless of the model we will still charge the same price based on that per-user, per-month model. We gain operationally efficiency being a software as a service provider. Our code stack is a combination of open source code and proprietary code which we do license commercially.
SM: Yet not everything in your portfolio is proprietary?
BB: There is an interesting story around the OpenSource stack. It is not just pre-existing projects like Apache and Linux. There is a tool we developed ourselves which we have leveraged a huge community around called Subversion. Originally our goal for Subversion was for it to be a category killer in the version control space. We did not want something that was just a successor to CVS. It was important that it be a tool people would migrate to from available commercial tools, and they have.
SM: What was the rationale for the Subversion project?
BB: Developing Subversion as an OpenSource project had strategic and tactical implications. Strategically we did it because we needed something disruptive. The space definitely needed the tool, yet we did not have the resources to provide it alone. The Subversion tool is a tool which keeps track of the history of your intellectual property. It is a time machine for your source code. To build up the consumer confidence to the level it would have required to be a successful venture would have cost tens of millions of dollars in marketing. Putting it out there as an OpenSource project and getting Apache to adopt it, which they have, shows that it can scale and keep a rich and high fidelity history.
SM: You also get adoption and testing?
BB: You get all that stuff.
SM: How widely adopted is it?
BB: We estimate there are 3-5M Subversion users out there. I hear about companies which have migrated to it all the time.
SM: You explained the strategic reasons. What was the tactical rationale?
BB: Tactically we did it because there was no way we could have built a team large enough to do it ourselves. Tactically we have had 3-4 full time developers on it rather consistently, and they have been able to leverage efforts of a community which has been much larger. Our role has been that of the air traffic controller, working on a core and laying out a roadmap indicating how it all should work, yet ensuring there are plenty of places for others to plug. This includes other companies as well, who also sell support services based on Subversion.
We wanted this to become the default standard for the entire industry. For us it has become the thin edge of the wedge inside of the company. If a company is already using Subversion we can come in and help them support its usage, but by the way we have all these other tools which plug in very cleanly.