Is open source the bubble 2.0 waiting to happen?
Rod Johnson, author of the Spring framework, thinks open source is hot right now, but its a “bubble” ready to burst, according to an article titled What Makes An Open Source Project Successful? by Charles Babcock.
Most open source projects are supported by an army of volunteers who buy into the hype, but “capitalism will inevitably reassert itself” and developers will find they need to put more effort into steady jobs and private lives, leaving “open source zombies”–unsupported, unmaintained projects–he predicts.
This is true, with many a project, that hasn’t built a successful ecosystem. Keep in mind that with the gazillion text editors out there, not all stand the test of time, like Emacs and vim do. Capitalism is always going to win hands down, because money in its essence is important to survival. Go back to Maslow’s Hierarchy of Needs, and realize that open source developers too, need to drive a car, have a roof over their head, start a family, and put food on the table.
So how does open source fund itself? Venture capitalists are active in the market (apparently $160 million has been pumped into OSS companies in the last 12 months), but they always want a return on investment, usually within 2-5 years. If customers alone can’t offshoot the ROI, there’s always going public.
Its great to note that MySQL and JBoss are mentioned, all of which are company backed, thanks to the selling of services, or added value. They help hire people working on the project, and of course, fund them for doing such work.
What caught me was the fact that JasperForge had 30,000 registered developers. As someone actively involved in MySQL Forge, I was a little shocked, so I had to mosey over, and double check. As of this writing, 30,735 registered users exist on JasperForge, but these include forum posters! I breathed a sigh of relief (go journalistic inaccuracies!).
Also of interest, Chris Messina has blogged about The relative value of open source to open services, in where he analysis activeCollab. This software is the open source Basecamp equivalent, with a promise that it will always be free, but apparently is changing the tune of the song. What especially moved me was this comment (that I can’t find on the blog post): “Build a community for the free software during early development and testing, then close it up just as the project matures.” Nasty, I wonder how many open source projects have done this? Its always good to note that forking the project, is also just as easy, but doesn’t promote ubiquity. It also paints the picture to those selling proprietary solutions that the open source ones really cannot be trusted, as they die or have a frequent rate of change.
Jason, the professional brand strategist who commented also has an interesting nugget: “While you’re the driving source behind the project – NOBODY fully owns their own brand. A brand is owned by the community that are a part of it. Without customers, a brand is nothing.”
He’s completely right. Red Hat Linux was a brand, but during the split, they created RHEL and Fedora. Even though there were no community contributions per se, to Red Hat Linux, it was probably the most widely supported commercial distribution that existed in its time; and there was a bloody large community around it. Of course, we won’t get into trademarks now…
What do we take away from all this?
- Don’t make promises that you’re going to break (Ubuntu has a promise to always be free, just like Red Hat Linux was, and I hope they keep it)
- Build the ecosystem up (Community, is key)
- Customers are important (Both to keep your brand alive, and to help you run a successful company)
- VC money won’t last forever
- Developers are human, and have needs like everyone else (hire them, pay them, feed them, and keep them happy)
- Forking is bad, and should be avoided at all costs
- Don’t abuse (or misuse) your community (eventhough they don’t pay you, they’re not second class citizens)
 – yes, IPO money hasn’t been around for a long time now, I’m sure… The tech industry is still getting out of the previous dot-com burst.