Open Source grids could be the answer; what's the question?

Even with the open source virtualization/grid potential there is still a lot of talk about how exactly we can leverage all that functionality and potential power. It's all very well having the ability to use a massive computational grid, but what the heck would we do with it?

You know, I've heard this sort of question dozens, if not hundreds of times over the last 10 years or so, and it's a reflection of one of Open Source's main problems - a problem that's also one of its main strengths.

Open Source developers are more inclined towards solving a cool problem, or making something that people say is impossible - while devs working for companies have a more market focus, they're required by the very nature of their jobs to consider commercial sensibilities. This is a short term problem, but a long term strength.

A lot of innovation is driven by the open source side of the fence, with new and interesting solutions coming out of the woodwork. Part of the problem with this is that they're not solving a problem that presently exists.

That's where human nature kicks in, and finds a use; re-phrases a problem, comes at it from a different angle, or views it in a new light - and suddenly, the formally problem-less solution has a problem it can solve. From this side of that point in time of course, we can't see what that problem is, but just wait - it'll show up.