Open Source, Good. Patents, Bad.
Last night I attended the VLAB presentation, “Bringing Great Open Source Ideas to Market.” The format was a presentation by Marc Fleury, founder and CEO of JBoss, followed by a panel discussion moderated by an industry analyst with Fleury, three other open source advocates and Sequoia Partner Doug Leone, appropriately dressed in black.
Fleury’s presentation extolled the virtues of his “third generation” model, “professional open source.” Basically they hire the developers so they can tightly control the code base. This allows better support as they can put fixes into the project (vs. “second generation” companies like Red Hat who have to lobby to get fixes into the next distribution). He tirelessly pitched other attractive attributes of open source businesses that have “critical mass,” such as the predictability of maintenance revenue and the ease of customer acquisition with a large user base.
As the presentation turned to business models, Fleury noted that there are perhaps six (out of 94,000!) open source projects that have the critical mass to support a business. How do you get critical mass? Fleury says “get lucky” – an answer that comes off as modest and dodges the question. Later in the discussion, panelists cited solving a real problem, having a strong leader and getting community involvement.
The moderator asked each panelist if Open Source is good for the software industry. Not surprisingly, they all said it is good. Leone, in carefully chosen words, said “What is good for customers is ultimately good for the industry.” Later he noted that “Open Source is of interest at the lower levels of the IT organization. The closer you get to someone with a budget, the more they care about solving their problem, and the less they care about the source code.”
When the moderator asked about the role of patents in open source, the temperature of the room dropped 10 degrees. All panelists agreed – software patents are bad. IBM’s recent contribution is nothing more than a ploy to slow down Microsoft’s forthcoming litigation. Leone quipped that the way to make money off open source is to sue open source companies. (Note to event organizers, you can never go wrong having a Sequoia Partner on the panel.) Fleury, who is surely the world’s most charismatic French software developer, quoted a line he attributed to Scott McNeally on open source and patents, “Open Source is like using a dirty needle – you never know what you might inject.”
As the event turned to audience questions, another VC in the audience asked (1) whether customers wanted integrated LAMP stacks and thus (2) whether the value-add of open source would move away from code and towards assembly, certification etc. – basically the SourceLabs/Spikesource business plan. Fleury answered “Yes and No.” The rest of the panel remained oddly silent.
As the evening wore on, a fatigued and hungry Leone grew tired of being agreeable and could no longer forebear. Paraphrasing he said that open source will become pervasive over the next several years, but open source companies will be small and marginal businesses that will ultimately consolidate.
Next month the topic is “The Future of MicroContent and Mobile Device Applications.” The topic will include RSS, and yes, PodCasting.