First and foremost, let’s remember that for Oracle as much or more than any other company, this will be all about business, and not ideology. That means dollars and cents, which means that it’s either direct revenues lost or available that are at issue, or an indirect economic motivation that has to do with a threat to, or an opportunity for, profits in some other fashion.
Any time you buy a company as large as Sun and try to integrate it into as different a business as Oracle, you are likely to do an asset by asset evaluation. Some elements may be incorporated closely into your own products, services or business units, while others may be sold or simply allowed to languish.
So let’s start with this aspect of integrating the two companies. Oracle paid over $6.5 billion for Sun before you take into account the debt that Oracle assumed and will have to repay. That means that Ellison and his Board have to justify this expenditure to Oracle’s stockholders. It does that by causing the value of its stock to go up, rather than down, as a result of the purchase. And it does that by (among other things) causing its profit margin to increase rather than decrease as a result of the purchase. One way to do that is to be sure that the ongoing revenue from maintaining Java rises, rather than falls. If it can’t do that, it should sell Java outright, unless it builds value in other Oracle products and services in some other fashion.
Clearly, Java was one of the Sun crown jewels, so the question is, what to do with it? Since Sun would presumably be quite happy to license the patents to Google, that suggests that the motivation is monetary rather than strategic. It’s less clear to me what the strategic value would be to Oracle to prevent Google for incorporating Java into Android, or to impede the marketplace generally from relying on Java.
The problem for Oracle is that the Java elements that Google used were “clean room” versions of Java developed by a third party. But while clean room versions may not violate Oracles’ copyrights, creating a clean room version may have little or not impact on avoiding patent infringement. Interestingly enough, Oracle is alleging copyright infringement as well as patent infringement. But while Google may be able to rewrite the elements to beat the copyright allegations going forward, it will still be liable for damages, if Sun wins, for the copies of Android that are already in circulation.
Why did Oracle bring its suit now and not the day it acquired Sun? Well, Android has now taken hold well enough that the Oracle suit won't hold it back - which also means that if Oracle wins, it's rewards could be substantial - and continue to become more so. And perhaps it thought it couldn’t wait much longer without being accused of setting a patent trap (lawyers call it “laches,” meaning that if you allow the other person to become more and more embroiled that you shouldn’t be able to profit from your keeping your mouth shut). As it is, since Oracle’s rights will be derivative of those it received from Sun, the clock has been ticking in this respect already for some time.
Another clue to Oracle’s motivations may come by example. Some months ago, Oracle announced that it would begin charging license fees for its OpenOffice compatibility plugin. In the case of OO itself, they couldn't sue anyone, because Sun had earlier made a non-assertion covenant to OASIS, the developer of the ODF document format that OO implements. In Java's case, there is no such pledge - although interestingly enough, the range of applicable, non-pledged patents might have been smaller, or perhaps non-existent, had Java ended up becoming a standard, as might have happened a decade ago. But that's another story.
In comparison to Java, the OO plugin is very small economic potatoes indeed. So while there could have been some strategic reasons for Oracle’s decision to charge for OO (I offered some thoughts on that topic here), Oracle’s Java decision seems to confirm that Oracle may simply be taking a bean counter’s approach to it’s acquisition, making sure that it maximizes the return for it’s stockholders on every asset on a case by case basis.
Here's a final thought: there's more than one way to recover on patent litigation. Could there be Google patents that Oracle would like to license at little or not cost, in exchange for a settlement?
Check out the Alexandria Project Now