On Thu, 17 Mar 2005 09:18:14 +0800, Ricky <[EMAIL PROTECTED]> wrote: > The algorithm does not become exclusively locked. It may be used by someone > else without having to pay a license if the new product beig patented shows > an innovative use of the algorithm. Software does not merely express an > algorithm. It must express it in a specific application which is innovative. > Take the case of mousetraps. The inventor of the original mousetrap use oter > prior inventions and concepts to come up with an innovation. So he made use > of the knowledge of the physical properties of metals. The knowledge about > how springs operate. Some knowledge about the behavior of mice. All of this > knowledge are equivalent to algorithms.
Fallacy. As defined by any mathematics book - an algorithm is a finite set of well-defined instructions for accomplishing some task which, given an initial state, will result in a corresponding recognizable end-state. All algorithms are mathematical in nature - which in itself is a separate class from all the other bodies of knowledge (although science uses the same mathematical laws). Given this is the case - software is just the same - a well-defined, finite set of mathematical instructions expressed in another language appropriate for the machine that would run these instructions. > Coming back to software, Boole could not patent symbolic logic. But Univac > could patent an application of boolean logic in its application to computer > software since it places the knowlodge within the context of a particular > application that could be used to create a product. > > Should someone would stumble a better algorithm on his own (and this > does happen, given that it's mere aggregation of algorithms that you > get to do even bigger tasks, express them into programming language, > and now you have software) that would infringe on the patented > combination of algorithms then that user cannot even use the better > algorithm he created. The developer is then forced to use other > algorithms that may not be even the best rather than the most obvious > or optimal given that someone has patented a particular order of > algorithms. > Your missing the point. I can always apply the algorithm contained within > someone else's aggregate if I am creating a new application that does > something unique and original with it. Algorithms cannot be patented in > themselves. The developer would not be forced to pay a for a licence. But this is the effect of software patents - you are locking the specific use of a particular combination and expression of such algorithms for a given application. Do note that patents lock the ideas to be the exclusive domain of a patent holder - which he may elect to continue being locked to himself or allow other people to use it given his or her conditions. Given the nature that these algorithms can be picked up or stumbled upon from any mathematics book or journal - patents on software would effectively disallow the usage of these algorithms when aggregated in a certain form which could be used in whatever task (e.g. using Djikstra's algorithm in conjunction with Bayesian statistics to find the shortest path within a hospital as well as alternate routes. This might be obvious to some, or can be stumbled upon, or can be even improved). > > Software are real products. It is something we interact with everyday. > > Therefore, it is something patents can protect. > > Given this reasoning - even news articles and books would qualify for patents. > > Paolo, I don't think you understand enough about patents. Please check out > the WIPO website here: > http://www.wipo.int/about-ip/en/patents.html > > Rage is merely countering your argument that A prouct needs to be physical in > order to be patented. The fact that something is not made of wood, metal or > plastic does not render it "not-physical." A process may be patented as well > and this is the sense under which software falls. A program represents a > process design - a way of doing things - which required labour and thought on > the part of the creator. So what you are saying is that a process may not be > patented. Unfortunately I do understand enough about patents. The usual problem arises on how one would define the provisions for something to be qualified as patentable. The US Supreme Court Diehr case in 1981 (computerization in heating of rubber) reversed the traditional notion that expressions of ideas or algorithms to be unpatentable to become patentable - which in turn paved the way to what software patents in the US is today. Incredibly stupid but that mistake costed a lot. Look at the mess they are in now. As to whether a process can be patented - yes it can be patented. But the provisions of patent qualification would require that something novel should come out of the process - and that something should be within the realm of the tangible, and not a mere translation from one intellectual form to another. Given that algorithms and software are just within the same class of each other (remaining within the intellectual realm being finite sets of instructions - and software are just translations of algorithms or an aggregation thereof - which itself already reeks prior art) - no form of matter came out, nothing novel there. > You also cannot equate a book or newspaper article to a process. A book, > article, or photograph does not need an application in order to exist. They > exist upon the point of creation when the act of creation came about. A > process though must shown to be workable in order to qualify. So a software > designer must spend time and effort testing and demonstrating his process to > show how it works and convince the patent office that it is worthy of a > patent. If I were to write complete nonsense code, I could easily get this > copyrighted since there is no need to show that it works. Take the first line > of the poem "Jabberwocky" by Lewis Carrol: "'Twas brillig in the slithy > toves." Carrol is expressing a thought here using gibberish. Under strict > English language rules, this statement would have produced a list of error > messages if one were to apply the rigour of programming to it. On the other > hand, Carrol created a unique statement that is worthy of copyright simply > because he created it. Whether or not the audience understands it is > irrelevant. > I'm arguing that the process of generating books or newspaper articles do take effort as well. Software by itself does not do anything either - you would need a tangible interface (called hardware) for which you could derive work from it. > Again software must be more than a mere aggregation concepts. The fact that > these concepts are applied and put together in a novel manner with benefit to > someone else is what makes the software patentable. Another real world > application. I can use VBScript to create a short statement to sort fields in > a database. I cannot use this in a patent application simply because I am not > creating something new and novel. By the same token, I cannot patent the > concept 1+1=2 in any programming language. If I were to create a program > which uses the expression: "A=1+1" as part of a novel application, then I > could patent my program. So the existence of the concept "1+1=2" does not > invalidate my patent in any way nor does it prevent others from using "1+1=2" > as a concept in other processes. Well, mixing together a couple of algorithms and directing them for a certain task is not within the sole realm of software. Expressing them in another intellectual manner does not constitute it to become any novel than the original form, or the unaggregated forms where it came from. > The software itself may not be physical but its application will have real > physical effects - numbers displayed on a monitor, an invoice being printed, > a rocket being launched. So even if software may not be "physical" it exists > within a physical framework with real-world effects. Then by that alone it can't qualify for patents. You can, however, patent the hardware or the process that produced the hardware that software uses (which is done as standard practice by many manufacturers). Then again you'd be subject to the standard rules of patent application - that such new matter (hardware) must be innovative, must not constitute prior art - highly doubtful unless you can show evidence of such.. > Trying to construe one method that works in one class of work as > appropriate for another class of work is just folly. > > What this boils down to is that people against Software Patents fail to grasp > the real intent of patents and how the patent ensures the rights of the > creator to benefit from his invention. Not all software is patentable. But > all software is copyrightable. It is necessary to understand this better > before condemning Software Patents. > Many still fail to grasp that the intentions and qualifications of patenting certain devices do not qualify when applied to software. -- Paolo Alexis Falcone [EMAIL PROTECTED] -- Philippine Linux Users' Group (PLUG) Mailing List plug@lists.q-linux.com (#PLUG @ irc.free.net.ph) Official Website: http://plug.linux.org.ph Searchable Archives: http://marc.free.net.ph . To leave, go to http://lists.q-linux.com/mailman/listinfo/plug . Are you a Linux newbie? To join the newbie list, go to http://lists.q-linux.com/mailman/listinfo/ph-linux-newbie
