Oracle has just sued Google over implementation methods used in the Dalvik virtual machine at the heart of the Android operating system. The complaint alleges knowing and willful infringement of seven patents originally issued to Sun Microsystems.
Oracle has retained Boies, Schiller & Flexner. One wonders if they’ll be any more competent than they were on the SCO lawsuit…
Results of a quick skim of the patents follow.
Here’s the list:
The six patents we can see are all mobile-Java implementation methods. Most seem to relate to optimization techniques, compilation, and JNI, though the claim language is general and vague enough in some cases that it is difficult to be certain.
I’m not seeing any algorithmic depth here – Google’s defense will almost certainly in part be that all this stuff fails the obviousness bar. There is also an interesting question as to whether Oracle has met its obligation under law to notify Google and allow it reasonable time to cure the infringement (removing these techniques from the Dalvik machine) before suing.
The commenter who tossed the complaint link at me notes that Steve Jobs and Larry Ellison are best buddies. So this could in part be Apple, rapidly losing the market-share battle to Android, striking back by proxy. I’m guessing Ellison told his lawyers to throw anything at Google that they thought might stick, and do it yesterday.
On a slightly different level, I think this is Apple implicitly conceding that it can’t beat Android with product design and needs to stop the competition before iOS gets kerb-stomped.
This post may be updated as I learn more.
UPDATE: On rereading, I see that there are copyright as well as patent claims; they’ll have to pierce Google’s clean-room defense to make those stick. Also, while they do seek an injunction, I see no request for a temporary restraining order; this suggests that Oracle’s lawyers know they don’t have an open-and-shut case and are wary of overplaying their hand. But if that’s so, why the actual-damages claim? Puzzling.
UPDATE2: A commenter found a link to the ‘720 patent. These weakens the argument that the complaint was a sloppy rush job.
UPDATE3: There’s prior art for the ‘720 patent in Emacs. Happens I know that part of the Emacs codebase of old and can confirm the author is correct. One down, six to go.
UPDATE4: I was incorrect in thinking the direct monetary damages claim is unsustainable; Oracle inherited J2ME, which it licenses to some handset makers.
UPDATE5: J2ME gives Oracle more skin in the mobile-Java game than I knew. This decreases the likelihood that Larry Ellison is doing a favor for his best buddy Steve.’
UPDATE6: Predating the ‘476 patent, RFC 86.0 is probably grounds for an obviousness challenge.
UPDATE7: It has been alleged that the technique described in the ‘205 patent was described in Efficient implementation of the smalltalk-80 system (1984)