Build around open source, not just on it

"Adoption of open-source application development tools will dampen total market growth, especially for Java tools," they write, even while predicting total development tool sales will hit $5 billion in 2009.

I've got another memo here from your friends at IBM. The solution is to build around open source development, rather than going right at it.

IBM is doing a rewrite of its Rational tools to make them more compatible with Eclipse, the open source development system IBM supports.

This idea has always been central to Eclipse. It's a common store of tools that member companies, and others, can build around, add to, and profit from.

Critics may say that IBM is changing Eclipse with this move, embracing it, making it part of its own keiretsu. In fact it's the other way around. Eclipse, like open source generally, is changing IBM, profoundly.

It may be the biggest untold story of the decade and, if you grew up in technology during the 1960s or 1970s one of the most surprising. IBM fought the U.S. for 30 years over its desire to control computing, yet its greatest success has come from releasing that control.

This is what Jonathan Schwartz' vision for Sun is supposedly all about, embracing the change embodied in open source, supporting it, and seeking to profit from it.

But the IBM example shows Sun may be going about this in the wrong way. IBM doesn't control Eclipse. IBM doesn't own Eclipse, in the way Sun now owns mySQL.

IBM supports a variety of open source projects, in its hardware, in its software, with its money. Then it builds proprietary advantages around that support.

It works. IBM may be a widows-and-orphans stock, with a P/E of just 15, but the widows aren't eating cat food, and the orphans have shoes with souls.

Sun's P/E may be higher, but building around seems to work better than control in the long run.