!082994 PTO not to appeal Alappat decision favoring software patents The US Patent and Trademark Office will not appeal a decision by a federal circuit court that opens wider the door for software patents. In a case filed by Tektronix (in re Alappat), the courts clearly indicated its approval of patents for software. The PTO could have appealed the decision to the US Supreme Court, but "I don't think we are going to appeal it", a patent official said, adding that guidelines will soon be issued to patent examiners on how to implement the Alappat decision. A formal decision (about appealing) is expected in the next several weeks. Alexander Johnson, a lawyer with the firm that represented Tektronix, said the patent office also has indicated to him that it will not appeal the decision. ==================== The gist of the above appeared in an article in the August 29, 1994 issue of Electronic Engineering Times. The article exhibits the same misperceptions about software patents that few in the patent community feel like dispelling. In the original article, we see "The USPTO will not appeal a decision by a federal court last month that opens the door for the first time to software patents". Given that 14,000 software patents have been issued by the USPTO since 1970, obviously Alappat will only widen a door that is already wide open. The only reason that Alappat et.al. ended up in court was that they were assigned a tough examiner. There are hundreds of issued software patents far more purely and trivially software than the technology involved in Alappat, and yet they were issued. Unfortunately, most of the press accounts of Alappat, written by reporters infrequently (if at all) supplied information by the patent professional organizations, exaggerated the importance of Alappat in the context of the reality of the software patents being issued. The main importance of Alappat is for those whose thoughts on software patents are based for the most part on court precedents, which now clearly are in favor of software patents. With regards to the guidelines the PTO plans to prepare for examiners, I hope that the guidelines are drafted in light of in re Alappat and other court cases, AS WELL AS in light of the 14,000 software patents issued to date. For example, last year the Air Force was awarded a software patent for nothing more than the difference in source code between two versions of a public domain computational physics program it had developed. Clearly that is rather a low standard of inventiveness for a software patent, but as it was issued, the new guidelines should either reflect that issuance as acceptable, or create criteria and a threshold to prevent such software patents from being issued. One guideline I would like to suggest for software patents is that no software patent should be issued that references no prior US patents, no prior foreign patents, and no prior non-patent literature, or conversely all issued software patents must reference at least one prior US patent and one prior literature item. Such a strange rule (for those familiar with with the probability distributions of literature references to software concepts) is needed because a fair number of software patents are issued that violate this proposed rule, patents that in the context of sufficient existing prior art would not have been issued. Greg Aharonian Internet Patent News Service (for subscription info, send 'help' to patents@world.std.com) (for prior art search services info, send 'prior' to patents@world.std.com)