The term cloud computing is becoming problematic, as it becomes evident this is just the next evolution of what we used to call the mainframe.
All we're doing is using virtualization on server farms, meaning your applications and data don't have to live in a particular location, freeing capacity and simplifying things.
But this is a very big deal. Virtualization eliminates a key advantage of Windows, something Microsoft must respond to. It accentuates scaling advantages that Linux, with its Unix heritage, has long had, which is why Red Hat is doing all it can to explore the meaning of that.
One way is through Infinispan, an open source project meant to reduce the bottleneck databases now have on cloud computing. By building a new data structure format, one compatible with a virtualized environment, Linux could do to Oracle what it has already done to Windows, make it less relevant in high-end computing.
The need for such tools is only going to accelerate as companies like Eucalyptus push clouds as something anyone can have. This is computing evolution in action. Something becomes possible, then commonplace, and it eventually gets cheap as chips.
What happened to Silicon Graphics in the 1990s, high-end graphics moving from dedicated workstations to chips and software, is going to happen in this decade to Microsoft's and Oracle's market advantages. Open source is grinding down those advantages through cloud computing.
It's the big story of this decade.