Larry Ellison is a genius at changing the subject. (To the right, the famous "crasher squirrel," explained here by Wikipedia.)
By this I mean few companies are more consistent in their brand and image management than Oracle (ORCL). They don't start arguments without reason.
Oracle is starting a controversy today because it has a problem it can't explain away.
Clouds are a big problem for Oracle because they only need one database, controlling an entire corporate system, rather than instances of the programs on each server. That's because they “virtualize” applications, using and re-using the same infrastructure no matter what operating system the original program was written in.
Oracle prices based on servers. And this is not just a pricing problem. As one cloud executive told me recently, “why should I buy Oracle licenses when I can run a single community version of mySQL?”
OK, Oracle has changed mySQL to “open core,” and is charging plenty for the “enterprise” version, so that's one finger in the dike. But it's just one finger. There are plenty of other open source databases out there, and the fact remains – each cloud just needs one database.
In fact clouds don't really need expensive servers at all. Google (GOOG) makes its own PCs, from parts. One of the very big cloud ideas, distributed processing, is based on the idea that instead of buying a big, expensive computer to solve big problems, you just cobble together a bunch of cheaper, slower ones.
But look, squirrel. By getting the media to focus on HP's (HPQ) buy of Autonomy, whether the price was right and whether Oracle might have gotten it half-off, Ellison and Hurd take advantage of the media's natural attention deficit.
Don't like the news? Change it to the story you want.
Autonomy and HP are not the issue here. Clouds are the issue. Oracle's entire business model is deeply threatened by clouds which are growing in size, complexity, and attractiveness to enterprises at an accelerating rate.
This does not mean big companies like Proctor & Gamble (PG) are going to buy sticks tomorrow and run their businesses off clouds the next day. It doesn't work that way. But lots of applications that used to require Oracle database servers are moving toward clouds. Big web sites. Animation. Simulation. As with open source itself, it starts with a test project, moves on to new projects, and eventually heads straight for legacy systems – and Oracle's legacy revenues.
And this may be the biggest problem of all. If Oracle is seen as “legacy,” rather than cutting edge, the cliff HP recently fell off gets a lot closer. So Oracle investors are wise not to look at the squirrel, and instead demand a strategy.