Return to BSD News archive
Xref: sserve comp.org.eff.talk:9206 misc.int-property:550 comp.unix.bsd:5839 Newsgroups: comp.org.eff.talk,misc.int-property,alt.suit.att-bsdi,comp.unix.bsd Path: sserve!manuel!munnari.oz.au!spool.mu.edu!decwrl!csus.edu!netcom.com!mcgregor From: mcgregor@netcom.com (Scott Mcgregor) Subject: Re: Patents: What they are. What they aren't. Other factors. Message-ID: <1992Sep30.035812.124@netcom.com> Organization: Netcom - Online Communication Services (408 241-9760 guest) References: <1992Sep26.161204.24573@rwwa.COM> <a=+p43j.mcgregor@netcom.com> <5204.Sep2920.32.3192@virtualnews.nyu.edu> Date: Wed, 30 Sep 1992 03:58:12 GMT Lines: 174 In article <5204.Sep2920.32.3192@virtualnews.nyu.edu> brnstnd@nyu.edu (D. J. Bernstein) writes: >Scott says that opposition to patents on mental processes---including >mental processes applied in an inessential way to physical objects--- >implies opposition to physical process patents. This is a rehash of his >old argument from comp.patents. As before Scott ignores the crucial word >``inessential.'' Actually, I said nothing of the kind, either in my most recent posting or past posting. Indeed, I would not say it because I don't agree with you that they are "innessential" I did quote this word however, from Mr. Bernstein's own posting. To be clear let me say what I do believe in the terms that Mr. Bernstein uses above. Opposition to patents on mental processes does in on way imply opposition to physical process patents. However, opposition to processes that can be partly performed by human brain, but which determine results that are then used to alter objects in the PHYSICAL world, does indeed muddy the question of whether the process is mental or physical. And so while a ban against pure mental patents would not have any effect on physical process patents, a ban that extended to this mixed arena could indeed undermine a large amount of physical process patents. For Mr. Bernstein, this seems to not be a concern. He feels that he can describe some applications as "inessential" and others as "essential". But I don't think that this can be codified, and indeed, I think that decisions such as Diamond vs. Diehr (where apparently Bernstein would say the physical world changes are essential) and the pythagorean example he cites (where the courts seemed to agree the application was inessential) show that this is still case by case law and not codifiable in any unambiguous way. Mr. Bernstein claims that changing the signals on a modem control line is inessential, as is changing polarities on magnetic media or changing phosphorescent displays on a cathode ray tube. But not vulcanization of rubber in Diamond vs. Diehr. But should we accept his characterization of what is essential or inessential? Let's look at this a bit further. When a telegraph changes the signals on a control line that is apparently essential and deserving of a patent. But clearly the determination of whether to send a long or short is a mental task, interrupting the current to change the signal on the line is (was) a trivial application of laws of nature. Similarly, the changing of polarities on a cassette tape or video tape would seem to be essential, but not on a disk drive using a compression chip. And the changing of phosphors in a TV is essential, but in XOR is not. At the same time, the complex set of instructions as to heating and cooling times and temperatures--which could be kept on a clipboard or memorized by an operator just as easily as stored in a computer program--but these are not essentially mental in Bernstein's view, and the simple physical processes (heating and cooling with traditional mechanisms) are elevated to the essential classification. These distinctions may be entirely clear to Mr. Bernstein and others, but I don't think that they would be obvious to the average judge, attourney or jury member. Indeed, it is my suspicion that whether you find such a thing to be essential or inessential depends upon your bias coming into the question. If what you personally care about is the physical world manifestations (cured rubber, a shorter data transmission, more storage to a fixed area of media) then you are likely to regard the physical application as "essential" to the delivering of the useful value the patent claims. But if you personally value the steps themselves, you would find the physical world application "inessential" even in the rubber curing process. This is essentially a bias of theory vs. practice. Bernstein illustrates this vividly in his next point. >It is essential that, to cure rubber, you actually take a physical piece >of rubber and do something to it. You can't complete the rubber-curing >process without physical action. > >It is not essential that, to apply the LZW algorithm, you actually take >a physical disk and compress the bits on it. The fact that LZW applies >to data on disks, on paper, on tapes, in memory, and in your head, >should make it clear that the physical application of LZW to bits on >disk is not essential to the process. Note how Bernstein has changed the focus of value in the rubber curing case (finished result) vs. the LZW case (intermediate result). As stated, Bernstein is correct, writing the data to disk is "inessential" to LZW. But if the patent claims instead a method for increasing the practical storage limits for a fixed area of magnetic media (the finished result) then the actual physical part is "essential" and the particular means (LZW) is just that, a means to achieve the result. The fact that LZW could be performed by a mental task instead of by means of a physical process involving electrons is "inessential" to the final result. >In article <a=+p43j.mcgregor@netcom.com> mcgregor@netcom.com (Scott Mcgregor) writes: >> Now real physical effects are happening: digital signals on a control >> line in a modem are changing, magnetic polarities in disks are >> changing, phosphorescent displays are changing all according to >> certain processes. > >The United States Supreme Court calls such effects ``insignificant'' >and, rightly, ignores them in determining whether a process is >patentable. I believe that Mr. Bernstein is merely opining on whether the court considers these "insignificant". They are not insignificant in the context of the telgraph, cassette tape, and TV. I believe that Mr. Bernstein is extrapolating from the "pythagorean" comment. But without specific cases, it is unclear whether such extrapolation is appropriate, or whether the reasoning in Diamond vs. Diehr is more appropriate. The specific examples I have raised, modem compression vs. telegraph, data compression vs. tape recorder, and XOR vs. TV reflects this ambiguity, as does the Diamond vs. Diehr process itself. >If you can't patent LZW-in-your-head, then you also can't patent >LZW-writing-to-disk. The ``writing to disk'' part is just too trivial. Says Mr. Bernstein. I beg to differ. First of all, getting a physical device to perform an process a process equivalent to a mental process is not trivial, especially subject to sufficient accuracy and performance requirements to make the process "useful". Second, there's a big difference in value between the mentally completed but unrecorded result of an LZW compression and an actual disk compressed with it. It is trivial to put rubber in an oven and set the heater to a particular temperature and time too. But doing it, and not just thinking it makes all the difference in the world to the end user. >Put differently: If you could make a mathematical algorithm patentable >simply by adding ``and then we write the result to a storage device'' to >it, then any novice patent lawyer could turn any mathematical algorithm >into something patentable. This would defeat the entire purpose of >prohibiting such patents. Indeed, this is the Diamond vs. Diehr reasoning in reverse. The Diamond vs. Diehr reasoning says that if you failed to allow a process simply because part of it used an algorithm performed on a computer you could create a huge hole for infringement of physical patents as people simply specified that the process controller was a computer. >I'm treading on solid legal ground here. Read what the courts have said >about patenting the Pythagorean theorem; it's exactly what I say above. > I don't disagree with Mr. Bernstein's quotation of the Pythagorean theorem example. I merely claim that the large majority of likely software patent cases would fall between the clear "pythagorean theorem" case and the Diamond vs. Diehr case. Maybe not to Mr. Bernstein, but to the average legislator, PTO examiner, judge, attourney and prospective jury member. Its very hard to make clear why digital technology should be treated one way and analogous analog technology another, or why an ASIC chip may be different from an equivalent general purpose CPU running a software program--but these are precisely the issues that would be raised and which such people (who are not likely to be as computer knowledgable as Mr. Bernstein) will have to decide. Some may therefore rush to the conclusion that it is all the more necessary to ban such potential patents to avoid this. But because such average people will implement the ban there is indeed risk that they will go too far and invalidate a good deal of other electronics and process patents. I don't think that this means that defining the issue in a way that yields desirable results is hopeless, but I think it is difficult, and that overly simple solutions could have drastic results. If software related patent reform comes about, and is more specifically limited due to discussions such as these, my objectives will be satisfied--I am not an opponent of patent reform, just fearful of inexpert surgery. -- Scott L. McGregor mcgregor@netcom.com President tel: 408-985-1824 Prescient Software, Inc. fax: 408-985-1936 3494 Yuba Avenue San Jose, CA 95117