Enterprises -- big customers -- remain the best market for open source today. The advantages of open source, of shared software infrastructure, are so obvious no company can stand against it.
The best analogy I can come up with is a mountain, like the Matterhorn (right, from Wikipedia).
The surface, the crust, is innovation. Below it is stable. Open source shares the cost of maintaining that stability, allowing more money to be spent on the crust, on what matters, on what drives business performance, on the new.
This is why I've never understood the critics who say open source doesn't innovate and thus has no value. It actually enables innovation by not innovating.
This is the whole point of the Eclipse project, of which Oracle is a strategic partner. Companies share the cost of maintaining a firm base of enterprise software, then they each innovate on top of it.
In this model both the cost and profitability of the base are minimized. Before announcing second quarter figures, which were in line with estimates, Red Hat CEO Jim Whitehurst told a journalist that to reach $5 billion in revenue he had to replace $50 billion in value.
This disappoints and saddens open source fans like Florian Mueller, who despite fighting software patents like growth as well as the next guy. He sees what Red Hat is doing as value destruction, noting that Ubuntu's contributions to the core platform are also relatively minor.
I disagree. I see it as value maintenance. That's probably how enterprises see it, too. When they use Red Hat Enterprise Linux they don't have to think much of the center of their software mountain. It's stable, and its cost is minimal. This lets them concentrate on the crust, on services customers will pay for.
If you're not paying monopoly rents, in other words, you have more money for other things. Value moves from the maintenance of the mountain to the products and services being built on top of it.
Oracle, like Microsoft, is dancing as fast as it can to justify continuing monopoly rents. The corporate database turns out to be an even better base to do this than the operating system was for Microsoft. An operating system can be virtualized.
When I have written disparagingly of Oracle this week, it has been based on an understanding that enterprises have seen this movie before and know how it turns out. Thanks to open source they have come to see monopoly rents for base functionality as waste which distracts them from their real focus, their customers.
The only way Oracle can justify its rents is to innovate, and that's what enterprises were looking for from Larry Ellison this week. The real story of OpenWorld was the failure to deliver Fusion. If you're not innovating, enterprises are saying, why am I paying you, when I can use shared infrastructure for less?