On Tue, 15 Mar 2005 19:11:27 +0800, Ricky <[EMAIL PROTECTED]> wrote: > Concepts do not get patented in and of themselves. This is a point which > people are often confused. It is concept plus application that gets patented. > Thus Edison could get a patent for converting electrical energy into visible > light in the specific instance of the incandescent lamp. But he could not > patent the concept in itself since there would be nothing for the patent > clerk to decide on. The LZW algorithm applied specifically to the manner in > which data is stored in a computer memory or hard drive. The algorithm behind > could not be without having the reference to the manner in which data is > organized inside computers. Einstein couldn't patent e=mc2 but the Atomic > bomb could have been if it were developed as a commercial application rather > than as a means to kill innocent civilians effortlessly.
Do note that Edison produced a real, tangible product based from existing concepts available to science in that period. He did not merely translated existing mathematical knowledge into another form of intangible mathematical knowledge. What he has done is truly innovative as he released a new product that can be seen by people. The same cannot be said of software. Despite all hardwork that would be done to translate or aggregate algorithms - software applications still remain to expressions of existing mathematical knowledge, and is still intangible. It's akin to no matter how you aggregate different mathematical formulas (such as aggregating division, addition, subtraction) to produce a new formula (such as Runge Kutta, Simpsons' 2/3, etc) it's still intangible, despite explaining it using english, or via pseudo-code or actual code. > There are many specific instances where you could argue the merits on whether > a patent could exist in particular circumstances. Note the attempts currently > underway to claim patent rights on the JPG compression algorithm. This is one > for the lawyers to decide but it does not invalidate the concept that patents > can and should be applied for in software that deserves to be patented. > > Software is not a mere aggregation of previously existing algorithms. Adobe > Distiller makes use of the "Fast Fourier Transform", and essential part of > the JPG compression algorithm. And yet, it makes a unique application of this > knowledge when embedding image data into a PDF file. Adobe applied for and > owns patents for Distiller and the PDF format itself primarily because they > perceived a need for this feature and created a unique feature compelling > enough to the patent granting authority that the general public would benefit > by having this knowledge made available to the world rather than being locked > away. So in exchange for sharing this knowledge, Adobe agreed to make this > public in exchange for a time limited franchise to exploit this. But it is. Delving further - anyone can use matrix manipulation resolution techniques to produce good hash keys that can be used in compression, as well as cryptography. FFT wasn't locked away in the first place - the mathematics behind matrix manipulation has been available for centuries, and FFT as "discovered" in 1942 is just an improvement over simplistic methods of matrix manipulation (it is fast albeit) at the expense of making the algorithm much more complex. > People often miss this point. Patents make knowledge freely available by > granting the inventor a temporary franchise over the invention. The > Pythagoreans famously kept their knowledge secret in order to increase their > bargaining position as mathematicians and philosophers (who got paid to > advice monarchs.) Pythagoreans alive today would have simply patented their > knoledge and sold the license to any bidder willing to pay the fee. The > benefit, of course is the knowledge became free for exploitation by the > public sooner than would otherwise have happened. > Had the Pythagoreans persisted to this day, all this technology would be non-existent. They hid a basic theorem which has been the cornerstone of geography, mathematics, science. It was only after their demise, coupled with the free exchange of truthful information and criticism from other academic bodies that science was able to flourish. > The other missed point I note is the concept of software code being > copyrighted. This is a completely bogus concept. You can copyright a book or > other publishing medium containing code. You can copyright executable files > in the same sense that you could copyright waveforms on a vinyl disk. But you > cannot copyright the code itself much the same as you cannot copyright the > notes that make up a song. You can however copyright a music sheet wherein > the notes are printed. Same goes with source code. You can't copyright the individual letters or words composing the code - but you can copyright the body of code that contains the letters and/or words. > Source code represents a design and makes no sense in and off itself without > a surrounding reference (a book in which it is contained or a method or > device for compiling and creating executable versions of it.) Thus code in > itself should be patented, not copyrighted. By the same token, you can patent > a design for a machine and submit drawings for this when applying for the > patent. You could then incorporate these drawings into a book and have the > book copyrighted. But the patent for the device remains separate from the > copyright for the book in which the device is published. Source code can also be viewed as a mere expression of algorithms, just like a translation of English text to another language. It can't be considered as a design since it is not ornamental in nature - which is one thing that design patents capture. It cannot be considered as a utility or invention since it is not born of a manufacturing process with real, tangible deliverables (hence the word software). > In a very real sense, patents are less restrictive than copyrights and serve > the public benefit more. As long as you are willing to pay a licence, you can > immediately create a product out of an invention and after a few years, you > no longer have to renew your licence. With a copyright, you have to wait for > 75 years (under current US law, 50 years in most countries) after the creator > passes away to avoid paying a licence. With copyright, all you need to do is simply not to use the offending code and rewrite it using another implementation, akin to translating 1+1=2 from decimal to 1+1=10 in binary, or just paraphrase it if you're using statements from other media such as books, newspapers - then give it proper attribution. With patents, even the mere translation is forbidden. > I truly hope this calarifies the issue for some of you out there. I've been > following this argument for a long time now. It initially came up in the mid > eighties when the issue of software copyrights came up. Up to then, it was > assumed that patents could take care of any IP rights issues on software. > Many arguments went back and forth (including many of the ones I've been > reading in this forum) and in the end, it was decided that published code > could be copyrighted. There was never any real issue of patents until fairly > recently since it was always assumed that this was the only natural means for > dealing with this. As it turns out, some people got hold of the same argument > that others had been using against patents for drugs and thought it would be > pretty neat to apply these to software as well, thus the current drive agains > software patents. > It was due to three US Supreme court decisions that paved the way for software patents: Gottschalk v. Benson, 409 U.S. 63 (1972), Parker v. Flook, 437 U.S. 584 (1978), and Diamond v. Diehr, 450 U.S. 175 (1981). All it took back then was misinterpretation of the tenets of the Patent code which drove what is supposedly within the realm of the mind into something that can be patented. Suddenly all it took to make patents out of algorithms is for someone to make a program - which doesn't bring in any new material to the world but another new utilitarian aggregate expression of the same ideas. > Believe me when I say that it will never happen that software will be free > from patents. It's against the public benefit and it takes too much effort to > protect invention privately when the knowledge will leak out anyway. Patents > allow commercial applications to surface sooner rather than later and we > should be fighting to keep it that way. Unfortunately patents as it exists today has been used unscrupulously to stifle innovation in the software realm. It would've been acceptable if it does really meet the qualification for an invention or design - but software doesn't qualify for these. -- Paolo Alexis Falcone [EMAIL PROTECTED] -- Philippine Linux Users' Group (PLUG) Mailing List [email protected] (#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
