Xref: utzoo sci.crypt:5334 comp.compression:893 misc.legal:27455 Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!think.com!samsung!uunet!brunix!ted From: ted@cs.brown.edu (Tony Davis) Newsgroups: sci.crypt,comp.compression,misc.legal Subject: Re: Software Patents, Another Side Keywords: software patents Message-ID: <79767@brunix.UUCP> Date: 1 Jul 91 01:51:37 GMT References: <7792@cactus.org> <79617@brunix.UUCP> <7819@cactus.org> Sender: news@brunix.UUCP Reply-To: ted@cs.brown.edu (Tony Davis) Organization: Brown University Department of Computer Science Lines: 152 Here I'll try to be less confrontational, but realize that I disagree with you on some basic issues. Also realize that I remain undecided about whether patents on software should be allowed. I'm not trying to argue against you; I'm trying to filter out telling arguments one way or the other. Admittedly, my current bias is against software patents, because I have yet to see the benefits that they would supply. >>Considerable >>anger is raised because the patent office is granting patents on algorithms >>that everyone already knows and uses. > >It is simply not enough for "everyone" to "know and use" techniques. For >ideas to be preserved for society, they must be PUBLISHED. If they ARE >published, then a patent cannot be granted to anyone else. If they are >NOT published, a patent may be available. True, if something is published that can prevent later patents on that thing. By the same token if something is commonly known to practitioners in the field, (published or not) that knowledge should also prevent the patent (according to the PTO's own rules). >Of course. Many people would like to claim that patenting as a trivial >process, ... Certainly I'm not among them. That patenting is non-trivial is part of the problem. >>Software is not protected because it is difficult to disassemble. It is >>easy to disassemble. Any knowledgable programmer can do it, however, it >>is usually not worth their time. If someone wants a program they buy it >>or steal it. > >The point is that, while "anyone" could disassemble A piece of code, only >LARGE companies can afford to disassemble MANY pieces of code. Thus, by >distributing only object code, large companies place the individual >programmer at a disadvantage. I disagree. To disassemble as many pieces of code as you care to, you need to know the object code format and how to write a parser for it. >>If they want the ideas in a program, currently they are >>free. Software patents make the ideas costly. > >Well, this is part of the discussion. I CAN and DO argue that so-called >"hardware" patents DO apply to software realizations NOW. > To some extent, I agree that "ideas are free." But an "idea" in a >program is (probably!) a great deal more than just "an idea." Not only >did someone "have" the idea, they had the background to "have" the idea, >the time to think about it, test it, compare it to other options, select >it and use it. All this is research and it is NOT free. SOMEBODY paid >for this. It is very convenient to be able to use someone else's research, >but that hardly encourages more research. It is precisely this attitude >which patents are intended to protect against, and it is the legitimate >role of society to take an interest in doing so. Perhaps we have a different idea of what research is. Or you are talking about bigger 'ideas' than this paragraph suggests. 'Having an idea, thinking about it, testing it, comparing it to other options, selecting it and using it' sounds to me like data structure selection, i.e., this is programming. What constitutes a different enough data structure that it is worthy of patent protection? Surely you aren't willing to grant a patent for every demonstrably different data structure? This is where the crux of the issue lies. Whether or not someone paid for the work is irrelevant. Someone pays for everything, but that doesn't mean it is or should be patentable. Nor does it mean that someone will unduly benefit from another's work. I am not opposed to property rights for programmers. I am opposed to patents preventing me from benefitting from my own knowledge. >>Writing, music, broadcasting, and publishing are all covered by copyright > >Yes, these things ARE copyrighted. My point is that copyright provides >protection for those things. But, by the very nature of software, >copyright DOES NOT provide protection for programmers. If it did, then >nobody would much object to distributing source code. Thus, we have the >case that many creative individuals in society have access to adequate >protection for their work, while programmers do not. Again, are >programmers to be legally inferior? Programmers should have the same protection that musicians and writers have if the nature of the their work is the same as the work of musicians and writers. If this is not enough protection to make them 'legally equal' then patent law is not the way to get them more protection; it is very likely to backfire. The question is does programming constitute MORE creativity than writing? >>In fact, software can do things that the underlying hardware cannot do. >>Software can draw a rabbit on a screen, hardware cannot. That the software >>draws the rabbit by manipulating the hardware does not give the hardware >>the ability to draw the rabbit. > >No, it is the HARDWARE which does EVERYTHING. Software is only the >sequencing information, stored in some physical part of the machine, in >some physical representation, which tells the OTHER hardware which things >to do when. (In fact, "software" in the machine is not really "software" >at all; is IS a part of the machine.) If software is only sequencing information, then it shouldn't be patentable. After all we don't allow patents on cars that can drive from Boston to Atlanta versus cars that can drive from Boston to Philadelphia. The only difference between the two is the sequence of hardware operations that are applied, right? What difference does it make that the sequence of operations is stored within the machine itself? Yes, I believe a patent should be allowed on the mechanism for doing the storage, but how does putting a sequence of operations in physical storage make the sequence itself patentable? The closer you tie software to the machine, the less argument there is for patenting software, because the closer you get to the machine the less difference there is between two pieces of software. >>It is not the purpose of software to select hardware operations. > >Yes, it is. It is if you accept, a priori, the machine perspective. Suppose we select, a priori, the mathematical perspective. Then the purpose of software is to describe an algorithm, an unpatentable idea. What is it about tying an algorithm to a machine which suddenly makes unpatentable algorithms patentable? >>However, I have my suspicions that software simply cannot be >>characterized in such an 'agreeable' manner. > >It is not necessary that everyone AGREE on the truth. The truth is, or >is not, independent of agreement. True, as long as we can agree on who should be allowed to define the truth, we don't need to agree on what the truth is :-). >Software IS intimately associated >with the actual functioning of real physical machines, it controls >physical operations, and is little more than a vastly expanded washing- >machine controller, or player-piano roll. Of course, printed software >can be literature (for people). But software which is intended to >function (that is, virtually ALL software), when it functions, is >logically a part of the machine. Are you saying that player-piano rolls are patentable? Yes, the player-piano is patentable. Suppose we grant that from one perspective software is logically part of the machine. What does that tell us about whether or not software should be patentable? So far you've been saying that software's machine role is simply sequencing of other portions of the machine. Is sequencing patentable or is only 'complex enough' or 'different enough' sequencing patentable? Tony Davis ted@cs.brown.edu