On Friday, December 11, 2009, two more patents bit the dust thanks to the ruling of the United States Federal District Court for the Northern District of California. This case came to my attention through Docket Report, which provides a remarkably detailed and informative daily summary of patent litigation events. In the matter of Fuzzysharp Technologies, Inc. v. 3D Labs, Inc., LTD (Case No: C 07-5948 SBA) District Judge Saundra Brown Armstrong granted Summary Judgment for the defendant and ruled that the relevant claims of U.S. Patent No. 6,172,679 and U.S. Patent No. 6,618,047 were invalid for failure to satisfy the machine or transformation test set forth by the United States Court of Appeals for the Federal Circuit in Bilski. It seems inevitable that the Supreme Court will alter in at least some way the Bilski test, which Justice Sotomayor continually referred to as “extreme” during oral arguments and which also troubled Justice Breyer and others. In the meantime, because the Patent Office and District Courts continue to churn these cases, a lot of work will be wasted and money needlessly expended only to have to start all over with a new test. Talk about ridiculous!
The claims in question (claims 1 and 12 of the ‘047 patent; and claims 1, 4 and 5 of the ‘679 patent) and the patents in general seem to be fairly typical representations of patents issued in the post State Street era, although the lineage of these patents date back to well before State Street was decided in 1998, although neither were granted until after State Street was decided. Nonetheless, these and many other patents granted in the post State Street era are now in question and falling, not because the underlying technology is not patentable, but because the way it is described satisfies a test different from the one we live with currently. Essentially, the Federal Circuit decided in Bilski that they didn’t like the previous test and didn’t care that many individuals and companies ordered their affairs according to State Street. By changing the rules the Federal Circuit showed enormous indifference to the need for settled law, and now daily the decision punishes those who played by the existing rules because they could not foresee the reconfiguration of settled law that occurred in Bilski. This is fundamentally unfair, make no mistake about it, and I wish more had been made about this at the Supreme Court.
In reading the Background section of the opinion the ruling should not come as a surprise. Judge Armstrong explained the patents, summarizing them by saying:
The claims in both patents are ‘method’ or ‘process’ claims drawn to mathematical algorithms that can be used to reduce the number of calculations required to determine whether a 3D surface is visible or invisible on a display screen. According to Fuzzysharp, reducing the number of calculations decreases the processing time necessary to form a digital image, thereby enhancing the speed with which the image can be displayed.
Oddly, the cite given in the decision is to the Plaintiff’s opposition papers. Perhaps the cite is incorrect, or perhaps this statement was taken out of context, I just don’t know. But characterizing method claims as being drawn to mathematical algorithms is probably not the best thing to do in order to support patentability.
After going through the Federal Circuit’s machine or transformation test and a summary of the arguments made by the parties, the district court observed that the salient question is not whether the claims are tied to a computer, but rather whether the claims in question are “tied to a particular machine,” as was explicitly stated in the Bilski decision. The district court then went through a series of cases where other district court Judges ruled similarly, and also went through a series of decisions from the Board of Patent Appeals and Interferences that also reached similar conclusions, namely that recitation of a computer, standing alone, is not enough to overcome a Bilski rejection.
Judge Armstrong, prior to reaching her ultimate conclusion, explained:
The claim language clearly states that these claims are drawn to mathematical calculations and algorithms for calculating whether certain surfaces are visible or invisible in 3D computer graphics. This is exemplified by the language of the claims, which specify a sequence of calculations that involve “identifying,” “comparing,” “determining,” and “ignoring” data. Though the calculations may be “performed on a computer,” they are not tied to any particular computer. For these reasons, the claims of the ‘047 and ‘679 Patent fail to pass muster under the Bilski machine implementation test for patentability under 35 U.S.C. § 101.
So the fact that a method or process may be performed on a computer is not enough. I dare say that strict adherence to the Federal Circuit test in Bilski would compel a similar ruling that a method or process is not patentable even if it must be performed on a computer. Thus, the take home lesson moving forward must be that it is not enough to recite a computer, or even articulate an invention that necessarily must and only can be performed on a computer. At least for now these types of inventions must be described with a level of particularity that explains the innovation on a system level. The reality is that for the moment we cannot articulate these types of inventions as software. Yes, they are software and that reality does not and will not change. Regardless of the outcome of the Bilski case at the Supreme Court software will continue to be patentable in the United States, we just now cannot call it software and may in the future be unable to do so as well depending on the ruling issued by the Supreme Court. Whether we call it software or not, it directs a computer, which is a machine, to undertake action in a predetermined way. So the focus needs to be on the pieces, parts and overall architecture and not the method steps.
One final thought on this case. The defendant, not surprisingly, argued that the claims in the patents in question were nothing more than “mathematical formulas and algorithms that fail Bilski’s machine-or-transformation test.” Surprisingly, the plaintiff conceded that claims in question were are not transformative, but nevertheless were tied to a particular machine; namely, a computer. The “tied to a computer” argument with patent applications written and issued when State Street was the law of the land are generally speaking going to have a very difficult time satisfying the machine part of the Bilski test. Yes, it is conceivable, but the weight of authority at the district court level and from the Board of Patent Appeals within the US Patent Office is telling a different story. Yes, patent examiners do frequently allow you to get claims with little more than a computer being recited, and you should take them because they may well stand the test of time after the Supreme Court issues its Bilski decision.
Having said all of this, it is becoming increasingly clear that at least some effort, and more than token effort, needs to be placed on demonstrating patent claims are transformative. In this particular case, where the invention relates to a method of reducing the visibility related computations in 3-D computer graphics, I think a transformation argument could have been mounted, or at least could have been attempted. I know there are a lot of decisions that go into which arguments are made and how, and I don’t want to engage in Monday (or Tuesday) morning Quarterbacking, but I do think it is important for us all, perhaps particularly clients, to understand that the machine test when strictly followed is disaster waiting to happen for many, if not most, pre-Bilski software related patents.
Now, one final thought on software patents in general. The Supreme Court is going to at a minimum adjust the Bilski test. We all know that for fact. There is no reason for the Supreme Court to take a case from the Federal Circuit for the purpose of saying “nice job, you got it right.” It has now been many years since the Supreme Court has affirmed the Federal Circuit, and Bilski will not reverse that trend. The only responsible thing for the Patent Office to do is to give applicants the option to suspend prosecution pending a decision. For those who wrote and filed applications before Bilski that would likely be welcomed, otherwise rejections will be finally made and the game of stringing out prosecution will begin, waiting for the Supreme Court to fix this mess. It would also be wise for district courts to stand down for a few months as well, otherwise they will waste time and resources now and eventually get the case kicked back to them for consideration in light of a newly minted Supreme Court decision.