Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!mailrus!iuvax!bsu-cs!bsu-ucs!sysc From: sysc@bsu-ucs.uucp Newsgroups: comp.arch Subject: Re: Patents and Architecture Message-ID: <32590@bsu-ucs.uucp> Date: 29 Jun 90 10:59:20 GMT References: <4742@sunquest.UUCP> <63036@sgi.sgi.com> Organization: Ball State University, Muncie, In - Univ. Computing Svc's Lines: 84 In article <63036@sgi.sgi.com>, karsh@trifolium.esd.sgi.com (Bruce Karsh) writes: > In article <4742@sunquest.UUCP> terry@sunquest.UUCP > (Terry Friedrichsen) writes: >>2) You can patent almost anything, old or new. Just try it; write up >>a description of some clever piece of software which embodies a process >>which is well-known to the industry. Then give it to a patent lawyer to >>cast into the appropriate "legal description". After that, even YOU >>won't recognize what in blazes is being described. The patent examiner >>doesn't have a chance. > > This just isn't true. The patent examiners are not dummys. The language > of a patent's claims are sometimes hard to read because they are more precise > than the writing we are normally used to. The language of a patent's claims is not normally hard to read just because it is more _precise_ than we are used to. I imagine that lawyers are sometimes capable of writing precise language, but that is not usually what they do. Patent claims are often written not primarily for precision so much as for _generality_. A patent is often written in such a way that the language of its principal claims can cover as much as possible. It is true that the subject matter of the invention must be described with a certain amount of precision, but one of the principal jobs of patent attorneys is to make sure that the claims are not too narrow, but are broad enough to maximize profits, which has the concommitant effect of reducing precision. >>If several other folks were using the techniques YEARS ago, it sounds >>like it was probably an obvious idea. That, unfortunately, is up to >>(expensive) litigation to decide. In the US, that is unfortunately the case. Some countries have systems that can simplify this somewhat and (possibly) reduce the costs. For example, (West) Germany publishes patent materials at prior stages of the process to allow for objections before a patent is finally granted. (My recollection of the details is slightly hazy and there was a change in the system a few years ago, so I may be mixing up what happened before the change and after.) They used to publish unexamined applications (or at least applications in early stages of examination), provisional patents and final patents. If I recall correctly, they now forgo the provisional patent publication, but there is still an objection period during which anyone can object to a patent and have it called back for administrative review (as opposed to litigation). If this works out (and I admit I don't have any numbers on this at all), then their system should have the effect of essentially recruiting the people who would have the greatest stake in pointing out any prior art as research assistants to the (overworked) patent examiners. My memory is hazy on this, but if I do recall correctly, the application can be narrowed under this system without losing priority, but broadening requires a new application with consequent loss of priority. Thus, the German system seems to me to have significant advantages for everyone (except possibly the lawyers :-): the examiners get (free) help from highly motivated researchers, holders of prior art get an opportunity to point it out at a stage much earlier and MUCH less expensive than defending themselves against a patent-infringement suit, and the applicant gets a chance to narrow his/her/its claims rather than having the entire patent nullified. > You mean that they claim to have been using it years ago. Were they keeping > it a secret? Patents are supposed to discourage secrecy. If its useful > and they didn't publish it and they didn't make a product out of it, then > how sympathetic should we be if someone gets a patent. Or else the examiners overlooked some prior art (or misjudged obviousness :-). > It's a young field and there will be some mistakes made. Since software So you do agree that is possible. Now all someone has to do is convince you that it has ever actually occurred. :-) > As we get more precedent and experience with patents, the abuses will decrease. > Other fields of intellectual endeavor have had patents for a long time and > it has been good for them. If lots of good patents are filed, the bad ones > will have a much harder time getting past the examiners. As software patent examiners see more software patent applications, they should get better, but basically, a patent examiner must be an expert in the field and as long as software is a relatively young field, experts will be both a LOT better paid and a LOT more likely to be happy in actual development than as patent examiners, so we are not likely to see top people examining patents for some time. ---- Paul Neubauer sysc@bsu-ucs.uucp or SYSC@BSUVAX1.BITNET