Monday, March 12, 2007

What a difference a year makes

Last week, I spent a couple of days attending the Open Source Think Tank. Great setting (Silverado in Napa), fantastic organization by Andrew Aitken and Mark Radcliffe (great wine and weather, for starters), spectacular list of attendees (from the open source perspective).

I attended the same event last year. Pretty much the same people. Back then, we talked about what open source will be in 2010. The answer: everywhere. Open source will be the dominant way of developing and distributing software. Open source will conquer the world. I left that meeting complaining to Andrew I did not learn anything... I felt there were issues we should have discussed, instead of just talking about how good we were doing.

This year, everything changed. The useless hype is gone. The entire two days were about the issues facing open source. About licensing and GPL3 and what the OSI can do to make things simple for OSS customers. About Oracle and RedHat. About the Novell and Microsoft pact. About monetizing and sustaining the business model.

It was absolutely great and extremely useful.

At the end, the outcome was exactly the same of last year: open source will be the
dominant way of developing and distributing software in the next years. But we did not talk about it. Open source is so mainstream that CIOs are considering it by default. It is treated exactly as proprietary software. They adopt it because it is better and cheaper, only if well supported. In that case, they want to pay. OSS is treated as commercial software. They do not even care about the source code. The field is totally leveled. We made it there. Since the game is on quality, price and support, open source companies will win hands down.

As an example, SourceFire went public last Friday. Great IPO, a gem in this confused stock market. Today up 13% while I write. 400M+ of market cap. An open source software company. A dual licensing company. I remember I heard Martin Roesch talking at JavaOne few years back and explaining the model that works in open source. The one that allows you to build public companies... It was fascinating. There were three companies in the first paper I read about dual licensing: SourceFire (public), MySQL (about to go public) and SleepyCat (acquired by Oracle). It should tell you something. Dual licensing is the model of Funambol, SugarCRM, Zimbra, among others. We just started a few years later but we are catching up fast...

The Think Tank reinforced all I knew about open source: the anxiety around open source licensing and indemnification is going to stick around. Dual licensing takes it all away. It is the model forward, the one CIOs go for. And the market is just following, as usual.

The big change in the last year is that the hype is gone. Open source is mature and mainstream. Proprietary vendors are adding open source wherever they can. Dual licensing is the way. It is clearly winning. Results are solid.

Let me add
to this self-celebration a funny episode during the Think Tank, which happened when Microsoft was accused to "split Linux". The Microsoft representative stood up and shouted "that is not true!!". He later went on stage with the Novell representative (nice move by the organizers, since they were the key sponsors ;-) and explained what the deal meant. Sorry, I forgot what he said, since I had my Marketing BS filter set to on. But I recall someone from the audience asking why Steve Ballmer explained the deal saying "Linux uses our intellectual property". And the answer about Steve Ballmer being hard to control and that "he recognized his statement was counterproductive". Yeah, right... Talking about a guy who does not know what he says, while he runs the largest software company on the planet. Pleeease. Just try to be nice open source citizens: it will be the dominant way of developing and distributing software in a few years and you can clearly use dual licensing, you know?