!080794 CAFC rules that software is patentable In what I hope will be a big boost to my software prior art database and search services (shameless plug - this is my newsletter :-), the United States Court of Appeals for the Federal Circuit (CAFC) has issued a decision that once and for all (if such a phrase is meaningful in the legal world) declares that SOFTWARE IS AN INVENTION THAT CAN BE PATENTED. From today's New York Times column by Teresa Riordan, page D2. ---------- In 1992, three inventors - Kuriappan Alappat, Edward Averill and James Larsen - went to court to challenge the Patent Office's rejection of their application for a patent on a software-driven device for generating a computer display; the case being known as the Alappat case. Under the Patent Office's reasoning, "a programmed general-purpose computer could never be viewed as patentable subject matter", the court [CAFC] wrote in the Alappat decision, adding: "This reasoning is without basis in law. The Supreme Court has never held that a programmed computer may never be entitled to patent protection". Similar CAFC reasoning has been that a software program can be patented, because, by setting many little switches in a computer's microprocessor, it literally creates a new circuit and therefore a new machine. According to Donald Chisum, a patent law expert and author of Chisum, "The court has broadly affirmed the patentability of software - as long as it is operating a machine or causing a physical result". Peter K. Trzyna, a patent lawyer with the Chicago law firm Keck, Mahin & Cate who specialize in financial computer products, said the Alappat case was being closely watched by the banking industry. "People who ignored patents on Wall Street, people who ignored patents for financial services, will have no choice but to deal with them now". ============================================================================== It is interesting that in such news stories, the Patent and Trademark Office is portrayed as having one fixed, strict, semi-tough policy against software patents, with repeated mentions to PTO action in the late 1980's to tighten its definitions of what it considered to be patentable software. For in reality, what matters is how the examiners interpret all of the PTO, CAFC and other precedents on this issue. For the most part, the examiners interpret them in many different ways. Unfortunately for the Alappat group, they apparently were assigned an examiner who strictly enforces the guidelines, while I have seen patents issued just in the last year that indicate that many examiners loosely enforce these same guidelines. The Air Force, for example, last year received a software patent for nothing more than the difference in source code between two versions of a public domain computational physics computer program, which to me is even more "software-y" than the Alappat technology. One got through, one didn't. But at least the CAFC has clearly ruled on the issue, which will probably turn the flood of software patent applications now entering the Patent Office into a deluge of biblical proportions, leading to even more Compton's-like shenanigans and lawsuits in the years to come. If the majority (>75%) of software patent applications continue to cite no non-patent prior art (especially journal articles and theses), then examiners will continue to be making obviousness and novelty determinations nearly in the blind. If the software community really is concerned about bad software patents, it better start educating itself about the need to submit better prior art with applications. Or hire me :-) Stay tuned and get yourself a law degree. Greg Aharonian Internet Patent News Service