2001-12-28 12:00:00

Blog entry on 28 Dec 2001

John Robb is right: Open source is a flop, at least so far. I think what John means is that open source as a business model is a flop. Clearly, there is plenty of outstanding open source software, so the concept itself cannot be reasonably identified as a flop. But as others have said, this is all about expectation setting.

Remember Star Wars episode I? Before its release, we were all told that this movie would be really, really, really good. By the time it came out, we all expected this film to be life-changing and far better than sex. The movie debuted to mixed reviews. I don't think anything could live up to that level of hype, even though IMHO The Phantom Menace was a good film.

The expectation-setting for open source was even higher. I was at the O'Reilly Open Source convention in August of 1998 where ESR promised that Linux would displace Windows on the server within six months. That's just one example of the big, big expectations created for the open source business world. For another example, consider the VA Linux IPO. The incredible rise of that stock is evidence of hype that was utterly unprecedented.

Unfortunately, the gap between those expectations and reality is huge. So far, open source has not come close to delivering on the hype. Windows 2000 is doing just fine. The LNUX stock price is an immaterial fraction of its high, and the company behind it has left the open source world to focus on a proprietary version of SourceForge.

The underlying wave of change is probably still coming. It is only natural that the margins for certain kinds of mainstream software will decrease. In basically every mass-market industry, some things become commodities. That has happened to web browsers, email clients, TCP stacks, and a few other items. Office suites and operating systems are still very, very high margin items. This will change someday, but it will take longer than we thought, and open source may or may not be the real driver of that change.

Remember folks, I bought into this hype in a big way, so I can speak from experience. We tried to make AbiWord function as a business, and we lost several hundred thousand dollars doing so. The results were painful and clear: The cost of software construction is still very high. Open source is the lowest-margin way to run a software business, and low-margin businesses are very hard. A few businesses have done it, but for most products, the math just doesn't work. We convinced ourselves that it would, mostly by speaking in terms of truly huge market segments, but the truth is that those segments are not won on the basis of price.

I'm not bitter, nor am I "anti-open-source". I like Linux, and I still love watching AbiWord move forward with such energy and momentum. I learned a lot, and somehow my company survived anyway.

SourceGear has always made its money in the world of proprietary software. Rumors of the death of this business model have been greatly exaggerated.