fuck off klausner you fucking nigger loving scumbag bottom feeder.

Correction: you mean I am the only person on _Techdirt_ to mention this kind of stuff. Every time one of these flakey patents gets featured on Slashdot, lots of people come forward with specific memories. Maybe one Slashdotter in ten thousand knows where the bodies are buried in the case of a particular patent, but, given Slashdot’s size, that is enough. However, Slashdot can only subject one patent a day to that kind of concentrated attention, and the patent office issues something like five hundred patents a day. There is not enough collective attention span to subject them all to public examination, criticism, and ridicule.

I believe Mike Masnick is a rather younger man than I am– I expect his memory does not go back so far. In 1981, as an engineering student, I was programming an IBM 1130 computer belonging to the University of Cincinnati’s engineering school, and learning from people who had been programming computers in the 1950’s.

I have done a certain amount of historical research about computers, mostly focusing on the period from 1940-1980. In round numbers, I have identified something like half a million pages of computer magazines between 1960 and 1980, most of them so badly cataloged that the only way to find out what is in them is to read through the lot.

Under the circumstances, it is hardly surprising that patent examiners rarely know what has already been done. If patent examiners were adequately trained, they would spend perhaps ten years getting up to speed with the literature before they presumed to examine patents, but of course no such thing happens. Until a couple of years ago, I would have said that the Patent Office was a species of organized crime, knowingly and intentionally issuing false patents for profit. I was somewhat surprised when the Patent Office made a good-faith effort to comply with KSR v. Teleflex, instead of engaging in a prolonged campaign of covert sabotage. The Patent Office went through an organized crime phase in the early nineteenth century, and the Franklin Institute of Philadelphia (the first national engineering society) had to whip it back to legitimacy. It wasn’t particularly surprising that the whole business was happening again. I was seriously concerned about the number of engineers who, in the utmost extremity of exasperation, proposed to settle the Patent Office problem with rifles. Fortunately, in the aftermath of KSR v. Teleflex, things seem to have changed, and only certain outliers such as the Federal District Court for East Texas still maintain the “Don Corleone” mentality. If the Patent Office can become efficient at revoking existing software patents under Bilski, that will deprive the East Texas court of standing to hear cases involving these patents.

From a legal standpoint, most prior art issues about software are moot. Bilski is not strictly a derivative holding of KSR v. Teleflex, but it is a reasonable intellectual development of the Supreme Court’s reasoning in respect of the computerization of pre-existing paper processes, applying the same reasoning to mental processes. Given that the Supreme Court ruled unanimously in KSR v. Teleflex, no one who is not whistling past a graveyard really thinks they will overturn Bilski. Of course most patent lawyers are whistling past a graveyard, and so is the Federal District Court for East Texas.

I would say that someone who is accused of infringing a patent is in fact the subject of a police prosecution. Patent holders use the methods of a police detective to seek out strangers who might be found to be infringing. They seek injunctions and import bans to damage their targets’ business without benefit of trial. It is not like a case of business partners falling out, or anything like that. The alleged infringer should be entitled to the presumption of innocence. By that standard, the patent process is ludicrous. It should be for the patentee to prove that the purported invention was original and nonobvious. As I have said, the perfunctory character of a patent examination does not begin to meet that requirement.

My working definition of obviousness is whether something is within the paradigm or not. For example, Albert Einstein’s famous response to Heisenberg and Bohr that “God does not play dice with the universe.” is an argument across a paradigm. More than twenty years ago, it was widely accepted that things like sound and video recordings were data, and could in principle be handed in the same way as other data.

One useful measure of obviousness is the kinds of homework/examination problems one finds in textbooks. One can show that students were expected to attack wide ranges of problems within general areas, and to do it as an ordinary matter of routine, within the space of an hourly exam, with a high probability of success, 70% being a C, or a passing grade. To complete the problems, you are expected to break big composite problems down into small simple problems, and then to identify these sub-problems as special cases of more general problems, which have well-known solutions. To put this in concrete terms, consider a problem from elementary calculus, the integral:

F(x) = Int[ 9 * x ^ 2 + 12 * x ^ 3]dx

You are expected to use standard theorems to reduce it to:

F(x) = 9 * Int[ x ^ 2]dx + 12 * Int[ x ^ 3]dx

and then to observe that the integrals Int[x^2]dx and Int[x^3]dx are special cases of Int[x^n]dx, for which there is a general solution.

Int[x ^ n]dx = (x ^ (n+1))/(n+1)

and therefore:

F(x)= 3 * x ^ 3 + 3 * x ^ 4

The same kind of thing applies in a computer systems analysis textbook, or an electronics textbook, or almost any other engineering textbook. You are supposed to take a system and reduce it into conventional components, not to treat it as sui generis. A college course is supposed to teach, not particular solutions, but ways of arriving at solutions. If you had signed up for a freshman course in a technical subject, and attempted to take a pre-KSR v. Teleflex patent examiner’s approach, complaining that the teacher had not showed you exactly how to solve that particular problem, you would surely have flunked the course. The teacher would probably have written on your examination paper: “Behold, the New Bartleby!!!!,” and would have circulated copies to his colleagues as a comic entertainment.

I think it is this level of obviousness that Mike Masnick and company are responding to, though I would not undertake to speak for them. Anyone who has had a technical education has been expected to rapidly resolve textbook problems which look very much like combination patents into their elements in double-time, without even getting much of a pat on the head for it. As one commentator remarked, circa 1970, “…grades, or dog yummies, or small fish, or whatever you throw to students these days.” These kinds of patents are worth a Milk-Bone, not a twenty-year monopoly.

In 1981, a man named Joseph Deken published a book entitled _The Electronic Cottage: Everyday Living with your Personal Computer in the 1980’s_. It was sufficiently popular that it became a Book-of-the-Month-Club Alternate Selection, before going into paperback. This book is too extensive to be readily summarized, but, inter alia, it sets forth the idea that a computer can have cameras, microphones, motion detectors, chemical analyzers, and other sensors, and treat their outputs as data, that the computer can control various machines, etc., etc. The whole robot, so to speak. I have a whole collection of similar books going back to about 1970, written by various popular-science and science-fiction authors. In the film, _2001: A Space Odyssey (1968)_ the HAL 9000 computer delivers a video mail, which it has been concealing in its databanks, and which is the ultimate cause of the computer’s aberrant behavior up to that time.