Which is worse: losing a visionary, or losing the people who make visions possible? The ongoing drama of Sun's acquisition by Oracle may tell us.

InternetNews.com weighed in the other day about the way IBM has been attempting to benefit from the way folks have jumped ship from Sun, with some notes about how such shakeups affect the talent pool available:

Talent defections are common in acquisitions. Losing the JRuby crew isn't quite as bad as losing James Gosling, the creator of Java. He remains firmly with Sun but his departure would be devastating if it did happen.

So which would be worse: losing Gosling, or losing the rank-and-file talent that make possible visions like Gosling's in a practical way? I'm tempted to say they're both about as bad, but there's a part of me that argues more strongly for the latter being far worse.

One of the paradoxes of this industry is how, ironically enough, visionaries are in no short supply. Everyone wants to be one; many people already think they are. What's harder to come by are people who are willing to put their heads down and do a great deal of gruntwork to make visions like Gosling's come true. They will do that, but only if the guidance is there -- and that guidance doesn't have to come in the form of charismatic Steve Jobs-ian figures, but rather good planners. (In fact, I'd argue that having someone in the Jobs mold as a project leader is a detriment, not a boon, because it too easily allows charisma and force of will to substitute for planning and strategy.)

Sun's position as users of open source doesn't completely mitigate the effects of losing talent, either. If you need someone to work on your vision, your project, you can only cover so much of the territory yourself -- or with contributions from a community that is mostly thinking of scratching its own hundred different itches. Such an approach is best for diversifying existing work which already has a firm foundation, and not creating something wholly new in the absence of firm guidance for how to do that.

Two things become clear now. First is that a company is not a portfolio of ideas or even products, but people. Proprietary or open source, it doesn't matter what they produce -- the biggest asset is the brains, and when they're gone, everything goes with it. Having those brains produce open source code elsewhere is not a substitute for having them in your fold, taking your guidance.

The second: Sun missed a sure thing with IBM. Badly.

InformationWeek has published an in-depth report on Google's upcoming Chrome OS. Download the report here (registration required).

Follow me and the rest of InformationWeek on Twitter.