Julie Samuels has an aggravated piece up on the EFF Deeplinks site about the Federal Circuit screwing up Ultramercial v. Hulu, again
. While I agree with her frustration and with the idea that the CAFC is Doing It Wrongtm
, it's not nearly as easy or black-and-white as Samuels makes it out to be.
(Readers should keep in mind that Samuels is an actual lawyer, whereas I am not. I happen to think I'm right but arguing against authority is always opening oneself to the "you're an idiot" response.)
To begin with, the areas of agreement: we agree that the Supreme Court wanted CAFC to reconsider Ultramercial's patent in light of the ruling in Mayo. We agree that patents ought not to apply to mere abstractions. We agree that patents should cover specific implementations, not general classes of procedures. And we agree that in light of all the above, the '545 patent was improperly issued.
But there the agreement ends. I, for example, would have rejected the '545 patent on the grounds that it cites zero non-patent prior art. This shows that the authors and the examiner did not review the relevant and extensive technical literature. Failing to do adequate prior art review should be a priori sufficient for total rejection. Samuels would have the patent invalidated on 101 abstractness grounds.
The problem with using 101 invalidation on any computer-related patent is that we don't have a good definition of "abstract." In computers, I can simulate any machine in software and likewise I can reduce any software to hardware. Take for example the latest cool machine announced by adafruit: a Bitcoin Miner device. Bitcoins have always been described as (and by) mathematical abstractions. They're numbers, one of the most abstract things we know. Creating (mining) them is done by running complex equations, which we also think of as abstract. But wait, what's that picture (copied above)? It's a piece of hardware - a machine - the least abstract thing in the patent world.
For Samuels to say that the Ultramercial patent is too abstract requires her to explain how it would still be abstract if I could buy a machine from adafruit that did what the patent describes. Sorry, it's not that easy.
Second, Samuels seems to think that the Supreme Court's rulings in Mayo and Myriad should make it obvious that CAFC got it wrong this time. Well, no. It's not that easy. As I wrote over a year ago, it appears that SCOTUS mixed up 101 and 103 reasoning in its Mayo decision. Perhaps Samuels can address the question of how Breyer's opinion successfully separates the two sections of the law, or how the opinion provides clear guidance on application of the "natural law" unpatentability principles. But she hasn't done so as far as i can tell. It's not that easy.
And I'm really sorry she mentions Myriad at all. As I blogged at the time, the commentary consensus is that SCOTUS got some basic science in that decision wrong. If they got the science wrong then what do we make of the opinions on which that flawed reasoning rests? Furthermore, the opinion is nominally unanimous, but in fact the raft of concurrences makes it clear that no more than two Justices could agree on any of the major principles.
Therefore, for Samuels to say that it was in any way obvious how CAFC should rule in Ultramercial is just grossly oversimplifying matters. I'm quite sure that the Federal Circuit judges could clearly articulate how they thought they were following the SCOTUS rulings which, so far as this non-lawyer/engineer/scientist/blogger can tell are not providing anything at all like clear and obvious guidance.
Oh, and one more thing Samuels and I agree on: SCOTUS is going to have to rule on Ultramercial directly. Whether they can manage to make things better by doing so is another area where we probably disagree.