Headline FeedsLinuxInsider Talkback
|
|
Re: Software Patent Protest Moves from Street to Internet
Posted by: John P. Mello Jr. 2003-08-27 15:23:16
See Full StoryA protest tomorrow in the streets of Brussels, Belgium, over a proposed European law on software patents will spill into the Internet as more than 600 Web sites are expected to modify their operations in support of the dissenters. The online protest is being organized by the Foundation for a Free Information Infrastructure. The group is asking Web sites of every stripe to shut down for the day in support of the protesters and to replace their home page with a protest page.
Re: Software Patent Protest Moves from Street to Internet
Posted by: FrankCire 2003-08-29 13:15:16 In reply to: John P. Mello Jr.
I have a degree in computer science and a degree in law. What I know as a certainty based on this background is that most of the legal conclusions presented by computer scientists when discussing U.S. patent law and practice are simply not true.
Algorithms are not patentable in the U.S. Software is not patentable in the U.S. Business methods are not patentable in the U.S.
A general-purpose computing machine having programming instructions that when executed by the general-purpose computing machine transform the general-purpose computing machine into a special-purpose computing machine is patentable in the U.S.
So, if the special-purpose computer has some program instructions in it that implement an algorithm, then the special-purpose computer is patentable, not the program instructions or the algorithms.
You can also patent computer-implemented processes; however, you are required to show the structure that implements the method. What's the structure? The special-purpose computing machine, of course.
Consider an amplifier implemented by an analog device or by a digital signal processor (DSP). The analog amplifier accepts an input signal with an input amplitude and provides an output signal with an output amplitude greater than the input amplitude. If I come up with a novel and nonobvious way to configure some semiconductor material to create the analog amplifier, then everyone agrees that I should be allowed to seek a patent for the new amplifier. Now I have a DSP that accepts an input signal at an input amplitude and generates an output signal at an output amplitude greater than the input amplitude. If the software I use within the DSP is both novel and nonobvious, why should I not be allowed to seek a patent?
Put another way, an amplifier is a computational device that accepts a string S that is a member of language L and produces a string S' that is also a member of L, not a very complicated computation at all. If I build devices that perform the computation with one device using electrons and one device using programming instructions, why can't I patent each version of the device? Currently, in the U.S. you can get a patent on both. In the European Patent Office, you only get a patent on the device using electrons.
The software community is really grappling with the overall question of whether or not we should be able to claim property rights in intangibles. I for one see no difference in claiming a right in something I can touch (like my car) and something that I can't touch (like my ideas).
Algorithms are not patentable in the U.S. Software is not patentable in the U.S. Business methods are not patentable in the U.S.
A general-purpose computing machine having programming instructions that when executed by the general-purpose computing machine transform the general-purpose computing machine into a special-purpose computing machine is patentable in the U.S.
So, if the special-purpose computer has some program instructions in it that implement an algorithm, then the special-purpose computer is patentable, not the program instructions or the algorithms.
You can also patent computer-implemented processes; however, you are required to show the structure that implements the method. What's the structure? The special-purpose computing machine, of course.
Consider an amplifier implemented by an analog device or by a digital signal processor (DSP). The analog amplifier accepts an input signal with an input amplitude and provides an output signal with an output amplitude greater than the input amplitude. If I come up with a novel and nonobvious way to configure some semiconductor material to create the analog amplifier, then everyone agrees that I should be allowed to seek a patent for the new amplifier. Now I have a DSP that accepts an input signal at an input amplitude and generates an output signal at an output amplitude greater than the input amplitude. If the software I use within the DSP is both novel and nonobvious, why should I not be allowed to seek a patent?
Put another way, an amplifier is a computational device that accepts a string S that is a member of language L and produces a string S' that is also a member of L, not a very complicated computation at all. If I build devices that perform the computation with one device using electrons and one device using programming instructions, why can't I patent each version of the device? Currently, in the U.S. you can get a patent on both. In the European Patent Office, you only get a patent on the device using electrons.
The software community is really grappling with the overall question of whether or not we should be able to claim property rights in intangibles. I for one see no difference in claiming a right in something I can touch (like my car) and something that I can't touch (like my ideas).
