Slashdot Mirror
Time To Abolish Software Patents?
gnujoshua writes "Has the time come to abolish software patents? Fortune columnist Roger Parloff reports on a new campaign called End Software Patents, which he views as 'attempting to ride a wave of corporate and judicial disenchantment with aspects of the current patent system.' Ryan Paul of Ars Technica writes that the purpose of the campaign is to 'educate the public and encourage grass-roots patent reform activism in order to promote effective legislative solutions to the software patent problem.' The campaign site is informative and targets many types of readers, and it includes a scholarship contest with a top prize of $10,000.00. We've recently discussed the potential legal re-examination of software patents."
You can bitch and moan all you want about software patents, but the problem is something else. It is the inability of "the little guy" to license patents in a way that doesn't cripple him, or make him subject to the whims of the patent holders.
When patents are easily and fairly licensed, the incentive to use them is increased, and the patent holder reaps the rewards of the increased usage. When they are kept locked down tight and only used as bargaining chips in patent wars, then no one benefits, not even the patent holder.
Patents should be freely licensable if the holder does not currently produce a product based upon the patent. The patent should be negotiable to any other third party who requires it, and it should be available at a reasonable price for reasonable terms. The only time a licensing request should be denied is in the case of gross misconduct of the licensee or if the licensee is a direct competitor to whom providing the patent would materially damage the patent holder. An arbitration agency should be in charge of deciding if a license denial is valid, and to decide if a particular patent holder is denying license requests too often.
Abolition of Software Patents is just plainly nonsense. Not only because companies will madly lobby to destroy any attempt to do so, but because it's not fair for the rightful owners of software patents who developed software (NOT patent trolls).
We all know the answer, there should be some kind of legislation or law (IANAL) that should enforce patent owners to
a) keep developing software based on the patent, not just earn royalties
b) be rightful creators, or earn the patents on a company merge
c) be a real company, not just some patent troll
Where a, b & c should be all true.
My 2c,
W
ATH++
No, we haven't.
We've established that mathematics should not be patentable.
Oh, BTW: you probably meant "an exception to the rule that he who creates something novel should be rewarded".
Any other fields of endevour we should exempt? Not that anyone here doesn't have a personal stake in the outcome.Otherwise it just doesn't make sense, with or without Chewbacca.
Well, let's first see if patents even work as intended.
Ignore this signature. By order.
So then someone can come along, change 1% of the design and sell it as their own? I'm thinking of cars as usual. In software's case, the final product is protected by copyright rather than patents. Individual methods are protected by patents. AFAIK America only introduced software patents in the 90s and things have gone downhill over there since then.
which is totally what she said
The problem isn't software patents. The problem is actually business model patents masquerading as software patents. Another issue is that patent length is standard across industries, when it should vary based on the timescale of innovation. Seven years in software is an epoch; the same for pharmaceuticals would be about a third of the amount of time spent developing a drug.
But the mechanism by which one implements his invention shouldn't matter. The fact that the bar is too low is an entirely separate problem.
1. software patents != copyright
;/) but also means you can do whatever the fuck you want to to ANY piece of program in the world, including reverse engineering the hell out of anything, installing OS X on a non-Apple toaster, freely mixing Linux and leaked windows code, and so on. I would see it as a benefit. The OSS community and the open source / free software model is too powerful for any closed-source corporation (but maybe one) to stop, so simply forking a project and closing the source will mean the fork will die soon.
2. abolishing COPYRIGHT, not PATENTS, would eventually mean that ALL the software will fall under a BSD-style license, which not only means free (but without copyleft
3. Abolishing copyright won't happen any time soon. *Maybe* if Stallman becomes the president.
Slashdot Burying Stories About Slashdot Media Owned
Stallman does not want to abolish copyright, the whole GPL relies on it to keep the source free. If he wanted "the other kind of free", he could already have chosen to use or change to a "BSD-style" license, or release everything to the public domain.
c++;
Well lets say I made a compression algorithm that will lossless compress all data by 1/2 (Yes it is mathmatically impossible, I know) This a new and marvel method. I don't patent it. Microsoft sees it reverse engineeres it and makes their own version and sells it to make billions of dollars as well say Toshiba uses it in their HardDrive technology to double its disk space just with a firmware update, and use this to make an other billions... Now here I am trying to peddal a little WinZip like app where Microsoft and Toshiba has already made my App useless with the technology I created. I would say that I should get some compensation for my creation...
I am not against software patents. I am againt most of the software patents. Software patents in my mind need to be very inovative and considered something where people said you can't do this with that, type of mantality. But most of the patents are not new ideas or something non ovious. Most of them are cases where any good programmer would come up with that method when given the problem.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
There are lots of things which can't be patented - mathematics, scientific discoveries[*], plot devices in novels or films, methods of trolling Slashdot. Why should software algorithms be an exception to the rule?
(* well, except for genes, but that's mad too IMO.)
You have a leap in your logic.
As one who opposes software copyrights, I use the GPL and not the BSD license.
As long as copyright exists, we use it, via the GPL, to prevent others from using it.
When copyright does not exist, the GPL is not necessary, and then the "BSD license"-style freedom takes place.
Choosing the BSD, rather than the GPL is the choice that reflects support of copyright -- it lets others use copyrights on derivatives of your work! If you do not support copyrights, disallow others from using copyright to restrict your software.
Those of us who oppose software copyrights are also pro-GPL, and I do believe Stallman is also in this crowd.
The problem with software patents is that anything that's really novel, like your impossible compression algorithm is basically a mathematical algorithm. Since you can't patent mathematical algorithms, there shouldn't be any need for software patents. I'm not sure if I've ever seen anything really inventive in software that wasn't a mathematical algorithm. There's patents on things like one-click shopping, which aren't mathematical algorithms, but which aren't really all that novel either. And then there's patents that are inventive, like GIF compression, MP3 Compression and others, but which fall under the umbrella of mathematical algorithms.
Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
Well lets say I made a compression algorithm that will lossless compress all data by 1/2 (Yes it is mathmatically impossible, I know) This a new and marvel method. I don't patent it.
Historically, mathematicians (as well as other people like scientists) have never been granted an monopoly on the use of the results of their research, and it's not clear why should that change?
As you acknowledge, it's mathematically impossible, so let's look at a more likely situation: you release your great new application, except big_company comes along and points out a range of other patents of theirs that you are infringing upon.
At best, you might be allowed to cross-licence if you have something they want - in which case, they use your "invention" anyway. Otherwise, you have to stop distributing your product altogether (and hope you don't get sued).
Even if we did accept your hypothetical scenerio - it's not clear that a world where hard drives everywhere have double space is worse than one where the only allowed application of this knowledge is your little app.
I don't understand what you mean by this. If I was a normal troll I'd say you don't know what you're talking about. Personally I don't like copyrights, but I do like the GPL. This is, however, illogical.
Richard Stallman wants everyone to be able to get the source to every computer program they run. He thinks this is very important, because without the source, you cannot modify the program, you can't learn from it, and you can't see what the program actually does. If you don't agree on this, then you won't agree with my next paragraph.
The only thing that stops $BIG_EVIL_COMPANY to take any GPLed open source project, add random proprietary changes to it, and release it without the new source code, is because if they do, they will break the GPL, and the GPL is only enforcible because of the copyright laws. Without the GPL, and in extension: the copyright law, they wouldn't have any obligation to release the source except for goodwill, and of course the other good things that comes with open source and free software. Some companies or individuals doesn't want or need those good things, and thus they oppose open source.
Of course, I'm not Stallman, and I know he doesn't like the current copyright system either, but completely removing copyrights without putting something else in as a replacement would be bad for free-as-in-libre software.
c++;
You're not thinking about the larger ecosystem in your assumption. You created an algorithm, in order to turn the algorithm into a viable product that you can release to market, you need to package it as an app. However, although your algorithm is indeed your own design, the GUI you create potentially impedes on several method patents. You get sued. In order to avoid the lawsuit, you drop the GUI and release a command line tool. However, the file I/O routines you use impede on several software patents. You get sued. That's the problem with software patents, they don't work well in a stack environment. I think copyright is the answer and has been all along. Remember, even if MS clean room reverse engineers your code, unless they can make their reimplementation significantly different from your original source code implementation, which should be impossible since your algo would be incredibly unique, they're still infringing on your copyright. This is the same problem cover bands face when releasing a CD of cover music. The original artist still gets their dues, even if only a passage from the song is used.
Good people do not need laws to tell them to act responsibly, while bad people will find a way around the laws-Plato
In "Math You Can't Use", Ben Klemens makes the point that the software development market is divided almost evenly into three segments: retail, consulting, and in-house. Software patents as currently defined and enforced benefit the first group, retail, but hurt the other two because they do not have the same market dynamics at all.
Patents are an artificial market force created to prevent certain kinds of unfair practices in a centralized, controlled-distribution market. Applied to a decentralized and distributed market such as that for free and open source software, patents create the nightmare scenario of an exponential increase in legal exposure as developers build upon each other's work.
The answer, then, isn't to do away with patents, but to tweak them so they make economic sense again.
Here is Chapter 5 of "Math You Can't Use", and it is well worth reading.
I just purchased the book and am looking forward to reading the rest. A very interesting work.
"We receive as friendly that which agrees with, we resist with dislike that which opposes us" - Faraday
The parent you replied to blindly and stupidly assumes that if copyright did not exist in Law that all software developers would magically decide to release the source. You and I know that if copyright Law were to be abolished, source code would be treated as Trade Secrets by those who don't currently believe in Free Software.
There are three main problem with software patents (and business method patents as well). One is the impossibility of searching the prior art. Another is obviousness to a Person Having Ordinary Skill In The Art (PHOSITA). And a third is that many software patents don't claim innovative solutions to problems, but claim all solutions to a newly discovered problem, even one for which the applicant really hasn't provided a practical, implemented, innovative solution.
The prior art in software is the corpus of all previously written software. To see if something had been done before, it would be necessary for a patent examiner to look through all software that had been written before to determine if a particular technique claimed in a patent application had been used. Programmers don't document their work very thoroughly, much less write journal articles on the techniques they use. And a lot of software technology is just learned by doing and by copying ideas discovered by reading other peoples' code but not formally otherwise documented by them.
Open source software provides one window into the corpus of software, and the IBM Type 4 program libraries (for machines such as the 360 and 1130) have some published source code, but proprietary software is distinctly unavailable for searching in most cases.
The same applies to business method patents. What business writes journal articles about their ways of doing things, especially ways that give them a competitive edge? Again, ideas are learned by doing and copying ideas learned on previous jobs.
One example is the Amazon "One Click" patent. That patent combines the cookie with the open account. You couldn't tell that by looking at the obfuscated language in which the patent is written, but that is what it does. Evidence of the open account as a business process can probably be found by looking in records written in Babylonian cuneiform. However, the USPTO doesn't search records in Babylonian cuneiform to look for evidence that a business process has been used before.
A second example is the patent for which Verizon sued Vonage. Any programmer with a brain would know that to hook up a VOIP system to a telephone switch would require some information and that the best way to organize the information would be in a database. The rest is legal obfuscation. If patent applications had to be written to a level of plain understanding, most would fall away because of obviousness.
Finally, patent applications should be required to include a working model or implementation or some kind of other evidence that the applicant wasn't just patenting the problem but actually had a solution. There was a discussion some months ago on Slashdot about a patent that covered all ways of doing something with graphics. The recipient clearly had discovered something that needed to be done in a graphics system. What the patent claimed was all ways of doing what was needed. IIRC, what the patent actually provided may have been one way of doing it, or might not have provided an actual solution. This is the software equivalent of patenting the electric light without figuring out how to make one that worked, a process that took Edison a lot of time and effort.
Many lawyers have a difficult time understanding technology. Some became lawyers because they weren't good at math and science. When those lawyers become judges, they don't suddenly gain an understanding of technology. They stick to their comfort zone. That is the reason the CAFC required documentary evidence to prove what prompted a PHOSITA to do something obvious. If it wasn't obvious to a lawyer, why should they treat it as obvious to a PHOSITA? Fortunately, the Supreme Court saw otherwise. Perhaps the courts will come to their senses on software and business method patents as well.
I can understand why people object to obvious patents, or to patents with not demonstrable implementation, but for all the whining people do about the evils of "software patents" I don't think people understand how similar the situation is with plain old mechanical patents. There are just as many bad patents on physical devices; why is the concern here only about software?
And as the parent suggests, I think a shorter term for software patents is a great idea. 17 months is probably a bit short -- I doubt many good ideas could make it from proof-of-concept to market in under 9-12 months, which doesn't leave a lot of protection time left on the patent. But certainly there's some single-digit number of years that would provide a workable balance.
For that matter I think the term of a patent could be variable in general. We'd want limits on the valid term range, but based on the patent type and things like regulatory barriers to market entry (for example in drugs, where drug X must get FDA approval which takes 2 years, but drug Generic-X can use the previous approval -- the government provides a barrier to market entry that is unique to the applicant and doesn't apply to other in the industry) we could certainly pick a more suitable term for almost all patents.
Which makes perfect sense considering that nothing of any real consequence was accomplished in mathematics or computer science prior to 1980, when the US Patent Office was not considering software patentable.