Slashdot Mirror
Easy Fix For Software Patents Found In US Patent Act
WebMink writes "What if there was an easy, inexpensive way to bring software patents under control, that did not involve Congress, which applied retrospectively to all patents and which was already part of the U.S. Patent Act? Stanford law professor Mark Lemley thinks he's found it. He asserts that the current runaway destruction being caused by software patents is just like previous problems with U.S. patent law, and that Congress included language in the Patent Act of 1952 that can be invoked over software patents just like it fixed the earlier problems. All it will take is a future defendant in a patent trial using his read of a crucial section of the Patent Act in their defense to establish case law. Can it really be that easy?"
No.
... we'll generate enough cash to... oh wait ... so many people in power don't want to fix the patent system... nevermind.
I swear they give me mod points to shut me up.
The letter of the law doesn't matter. The spirit of the law doesn't matter. The constitution of the USA doesn't matter.
Do you have money? Power? That matters. And as long as the powers to be want software patents, that's what you'll have, the way it is right now.
"The agriculture ministry is not in charge of Gundam" - Japanese ministry official.
but good luck with the jury.
Pretty sure submitter meant retroactively.
It's relieving to hear somebody may have found an existing law which would end the patent debacle, but the problem remains that the law is not clear to the average person. Even this paper hoping to clarify what the law says just looks like 57 pages of MSWord to me, and I feel at the whim of the high powered attourneys and politically charged judges to interpret it.
Herman Cain may not have been a quality pick for president, but I'm beginning to think his mantra of "simplify the law so the public understands it" has something to it.
but only the particular means of implementing that goal described by the patentee and equivalents thereof.
The problem is that the means of implementing that goal is described in the vague term "on a computer". Sometimes it expands this into the various components of a computer, but I suspect it's still enough to walk round the functional claiming defense mentioned.
Donte Alistair Anderson Roberts - hi son!
Karma: Chameleon
He said:
>> If courts would faithfully apply the 1952 Act, limiting those claims to the actual algorithms the patentees disclosed and their equivalents,
Clearly the authors were thinking non-obvious stuff with complexity such as a Fourier Transform would be patentable but how about just parts of it? What is the smallest/simplest functional thing that could constitute an algorithm?
If this gets enacted as case law the obvious next step is that Apple will patent the 'for' loop and Microsoft will licence the 'if' statement.
His flaw is the assumption that software which is nothing more than math falls in the realm patents. It doesn't. Any Darwinian monkey would know this.
No, his awareness is that, however silly it may be, the courts have to this point declined to declare math implemented in a software algorithm as unpatentable. Given that reality, he outlines a possible approach to mitigating the resulting damage.
Your assumption is that common sense matter in a court of law. It does not.
No
MouseClass extends ScrollClass, which extends TabClass, which extends SidebarClass, which extends PowerClass, w
Surely you mean retroactively old boy?
Software is Algorithms. Algorithms are methods of achieving a goal. Patents are for novel non-obvious devices or methods.
The claim that software shouldn't be patentable because "it's just math" isn't very strong. By a similar argument all patents could be invalidated because physics is just math and machines are simply applying physical algorithms. Furthermore the "physical things are different from virtual things" meme is going to hold society back in the long run so we may as well start getting rid of it now.
What needs to be reformed in software patents is the distinction between abstract algorithms and applied algorithms (so you can't patent "merge sort", but you could patent a particular formulation of a recursive merge sort).
I do tend to agree that software patents have become about patenting the idea of implementing something and has become divorced from the specific implementation.
You can patent how you implemented something, but not the notion of doing it in the first place.
The infamous "One-click" patent, which as far as I'm concerned amounts to "a method and system for doing one of the operations of a system is capable of doing, but with a single button click by using already configured information" only they've added "when buying stuff".
Because there's a lot of things which people have implemented behind a single button which can easily gather several bits of data, assemble them, and take an action. In fact, it's damned near the Von Neumann model of "input, processing, and output".
We've had buttons before. We've bought stuff before. We've even bought stuff online before and have transaction processing which handles it. We've even allowed you to log into a system and be recognized as a distinct user for which we have information stored.
But as soon as a web-site presents you with a single button next to an item, and clicking it causes it to use the information already known about you (shipping address, credit card info) to process a transaction and initiate shipping ... well, clearly we have performed magic and nobody else could have possibly come up with this idea on their own.
Lost at C:>. Found at C.
While the US legal system is such that entities can drag even blatantly bogus lawsuits out for years, so winning against individuals and smaller businesses just by attrition of legal costs, fine-tweaking the definition of bogusness wont have even the slightest effect.
It can get pretty small.
It just needs to be somewhere few people have looked at. You can define an algorithm for any two numerical values added together. We keep thrashing into the Obvious problem because nobody/one lonely guy in the Midwest USA/ thinks of stuff like adding the Weight Watcher Point Count of your fridge contents with the number of times you ordered takeout to get your average ranking of a recreational gamer/nerd/techie.
See how fun it gets? It's not obvious - but once you hear it you can't "unhear it" so then it sounds obvious 7 minutes later. It doesn't even have to work. It's just an algorithm.
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
The parent post isn't Offtopic (current mod as I write my own comment), it's simply false. And the poster is a fucking idiot. Clearly he hasn't even reached the intelligence of a Darwinian Monkey.
Here is the crucial quote from the abstract:
I'm not sure this would fix the problem. Sometimes, the fact that the algorithm is patented is the problem. Think about it: if a patent didn't cover the idea of generating a sorted sequence from a potentially unsorted one, but only covered quicksort and its equivalents, would that still be a problem? I'd say it's better than covering any implementation of any sorting algorithm, but I'm not sure the resulting situation is really where we want to be.
Please correct me if I got my facts wrong.
All it has to do is win in court with an argument that by all appearances to this layperson has failed already many times in patent cases.
"All it has to do to win is not lose yet again" seems tautological to me. But I'm not a lawyer.
I'd love it if the lawyers here responded that this was in fact under the subtleties of the existing law an entirely new tack that is likely to be a winner in an actual case and establish the given precedent.
I would love it even more if it actually happened. But this (again:) layperson wouldn't be likely to bet on such an outcome within our current system using actual betting dollars.
STOP . AMERICA . NOW
Im sure judges are getting tired of all these patent suits and would happily hang their hat on whatever point of law that would produce the most productive and expeditious result consistent with due process. If a court can narrow the application of a law in a way that reduces their workload, you bet they will.
You know, I've worked in the software industry for a long time.
As much as what computers do is expressible in math and formal logic, the actual act of doing most forms of writing of software (unless it's strictly doing math) is actually nothing at all like math in my mind.
Sure, it's logic and problem solving. But breaking the problem up into manageable chunks and object classes often seems more analagous to modelling real things than math ... and when you abstract away how you actually implement things, "a container which holds things" and modelling the functionality an object class incorporates isn't what I'd call rigid mathematical concepts.
As much as the final code is essentially a manifestation of math, to me the actual process bears no resemblance to math. Largely because things tend to be done as a representation of something which is more based off how a real, physical thing would behave.
Lost at C:>. Found at C.
This is one of the most important things one can understand about the law system. My father started to study law and attended law school, something to occupy himself with in his older years. He regularly blew my mind by telling me about court decisions that not only bypassed common sense, but also morality, and kept going into the realm of truly messed up.
The court room has nothing to do with the real world, real experiences, common sense, morality, or anything a sane person might recognize. It's a separate world with its own strange rules, and the only way we mere mundane mortals can tap into it is through our spirit shaman lawyers.
Love sees no species.
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
No sigs in BETA. Beta SUCKS.
Copyright already solves this problem. If I copy the exact way you do it, copyright grants you protection (like writing the same paragraph as you in my thesis). If I come up with a different way to accomplish the same thing (as in, found a way to copy documents that isn't the same way a Xerox machine works), I should be allowed to. I can't see anything other than personal agenda to disagree with this
"When life gives you lemons, don't make lemonade. Make life take the lemons back!" -- Cave Johnson
... his interpretation of the patent law is copyrighted.
It's good to see, and a pleasure to read, rational discourse on patents who knows what the fuck they're talking about. Almost every single article about patents is so for wrong, calling it a straw-man argument is a joke. Let's have more articles like this on /., please.
Software is Algorithms.
Huh? I can't agree that software is algorithms. Software is an implementation of algorithms, not algorithms. You cannot patent algorithms, but you can patent their implementation...
Every judge that has heard a patent case and not noticed this part of the law should be impeached for dereliction of duty. There is no excuse for a judge not to know the law that she is enforcing. Ignorance of the law is no excuse.
If an inventor claimed to own the function of nanoparticles altering the abosorption of wavelengths of light to increase the efficiency of a solar cell, then he could keep everyone else from using nano-particles to do that regardless of material or structure.
Hence, I can see the logic of the 1952 patent law structure in that you can patent a specific new structure leading to a good end result, but you can't patent "the release of radiation" from a light bulb type structure.
What this boils down to is tightening up the enablement and written description requirements of 112 of the Patent Act, specifically with regard to the way functional claims are judged. This is not a particularly new idea, and quite a few people (including those that could be called 'pro-patent') have been saying for years that this is a reasonable way to address problems with lots of different kinds of patents, including software patents, that doesn't involve technology-specific changes to the law. I certainly have mentioned it on Slashdot many times (e.g. this comment from 2010).
One issue is that the use of functional claiming has been in pretty steep decline for a few decades now. So I'm not sure just how many software patents this approach will catch, but it could be that software patents are one of the last holdouts for functional claiming. Alternatively, the courts could begin to interpret software patents as using functional claiming even when they don't use the traditional "means-for" construction. This is already done in some cases (i.e. "means for" is not a necessary magic phrase), but it could become more common.
The bottom line is that this is a call to restrict patenting to only that which the inventor actually discovered and actually described in such "full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains to make and use the same." In other words, to enforce the statute as written.
What would a non-Darwinian monkey be?
At any rate, you're wrong. While some functions of software are functionally mathematical/algorithmic in nature (sorting, searching, recursion, iteration, encryption, compression, etc.) most software developers would be hard pressed to describe or express their work as a whole in mathematical terms. Requirements definition, human factors / usability, prototyping, supported interfaces, etc. are approached from a form and function perspective, not a mathematical perspective.
If you aren't part of the solution, then there is good money to be made prolonging the problem
...because physics is just math ...
No, it's not. Mathematical language is used to describe physical ideas and theories much as English or French are used to describe news events.
Except physics isn't math. We can model physics with mathematics, but that doesn't mean they are literally the same thing.
Conversely, algorithms *ARE* math. This is entirely provable (a good text on the theoretical foundations of computer science might even describe such a proof), but probably not readily comprehensible to people who do not have a very thorough understanding of mathematics and mathematical proofs (which is, I'm afraid to say, most people... including even a lot of computer professionals).
File under 'M' for 'Manic ranting'
Ah... but the point being raised here is that you currently *CAN* patent algorithms... as long as the algorithms are considered sufficiently complex to be allegedly "non-obvious". That's the crux of the problem.
File under 'M' for 'Manic ranting'
Copyright doesn't solve the problem because they would simply protect the "text" and not the actual invention itself, which is why patents are used for software (in conjunction with copyright).
Wait? Aren't methods no longer patentable? Well, business method patents anyway. But a method, a way of doing something, should be equally disallowed. And in the case of software, it's still a method... a way of doing something... as directed by a set of instructions readable by a computer/processor device.
Any time I see a phrase like this in a legal setting, I cringe.
I realize that "formal proof" in law means "something that convinces a judge/jury", but there is a formal meaning in computer science for equivalence of algorithms. You can't prove two arbitrary algorithms equivalent, because that would enable a solution to the Halting Problem.
You can sometimes prove that two particular algorithms are functionally identical (e.g. different methods of sorting), but if their implementations differ (bubble/merge/heap sort) then that's a case that the law should permit.
To a Lisp hacker, XML is S-expressions in drag.
the unfittest monkey
"The ability to delude yourself may be an important survival tool" - Jane Wagner -
That's entirely possible. The problem is that the question is not really asked. Err let me rephrase that, the problem is that in all those patent cases, all those attempts to ensure they are not violating other people's patents, in all those cases where they are protecting their own patents, the questions being asked aren't the same questions leading to this answer.
The status quo for a normal company is not how to I force someone to let me use their patent or how can I violate their patent. It's does my product encroach the patent, what do I need to do to stop that. It's does that product violates my patent, I need to stop it. Almost every billion dollar company will have their own patents or licenses to a patent and will be looking into how they can protect their products, not how they can weaken their positions on them. It is not the normal forward looking face of those companies.
Except physics isn't math. We can model physics with mathematics, but that doesn't mean they are literally the same thing.
Now you are being too literal.
Algorithms cannot be implemented without physics. I dont think that anyone is seriously suggesting that the movement of baryons and leptons (classical machines) is any more or less special than the movement of just leptons (electronics and digital machines.)
The confusion arises because the modern algorithm is implemented on top of so many layers of abstraction, and many of the arguments on both sides rely on making the distinction between those layers, just as you inadvertently are. The abstractions are just framework and were intended to be meaningless. We dont see patents ending with "...constructed of wood" or "...constructed of stainless steel" but we do see "...on a cell phone" and "..on a network." That is the actual problem, that old ideas are considered new simply by changing the materials, not that algorithms shouldn't be patentable.
I argue that it is only algorithms that were supposed to be covered by patents. Nobody got to infringe on the steam engine patents just by using different materials, and nobody got to re-patent the idea when stainless steel was invented.
"His name was James Damore."
But try actually doing anything with an algorithm or maths equation. You can't unless you implement it.
Software is suppose to do something - to fill some needs or solve some problem.You must take your algorithms and build them in such a way that they run on certain hardware and interact with certain interfaces and I/O devices.
A sort or search algorithm is fine, but to really be useful (sellable) software it has to interact with some form of data storage and probably interface with other software.
Now, I'm British and so I'm not familiar with the US patent system. Are people really trying to implement abstract algorithms, or bits of software that use them?
Since when did we need a lawyer to write source code ?
AccountKiller
Software patents are like patenting the plans to make a house. Anyone who makes a house (even using alternative plans that make something like a house) would be in violation of said patent.
Every time I start to have faith in humanity, I ruin it by driving to work between 7 and 8 am.
Once you have specified your algorithm, it can be run on any general-purpose computer out there. And general-purpose computers have already been invented (and patented). But the exact same algorithm can run on any general-purpose computer whatsoever - even ones that haven't been invented yet, or ones that rely on entirely new physical principles, such as optical computing.
Which seems much closer to what Congress actually had in mind, as http://ijlit.oxfordjournals.org/content/14/3/257.full?ijkey=rF2MI0t8NYrGuJJ&keytype=ref#p-88 has pointed out previously:
http://www.eco-action.org/dt/affluent.html
"Above all. what about the world today? One-third to one-half of humanity are said to go to bed hungry every night. In the Old Stone Age the fraction must have been much smaller. This is the era of hunger unprecedented. Now, in the time of the greatest technical power, is starvation an [institution]. Reverse another venerable formula: the amount of hunger increases relatively and absolutely with the evolution of culture. This paradox is my whole point. Hunters and gatherers have by force of circumstances an objectively low standard of living. But taken as their objective, and given their adequate means of production. all the people's material wants usually can be easily satisfied.
The world's most primitive people have few possessions. but they are not poor. Poverty is not a certain small amount of goods, nor is it just a relation between means and ends; above all it is a relation between people. Poverty is a social status. As such it is the invention of civilisation. It has grown with civilisation, at once as an invidious distinction between classes and more importantly as a tributary relation that can render agrarian peasants more susceptible to natural catastrophes than any winter camp of Alaskan Eskimo."
However this could be fixed in our society with a basic income (and/or other changes):
http://www.basicincome.org/bien/aboutbasicincome.html
Every age has its challenges. Twenty thousand years ago, there was no malaria (a side-effect of agriculture), no threat of nuclear war or bioterrorism or nanotech or robotics run amok, communities and familes were probably stronger overall than in industrialized countries, people ate more vegetables and so had little cancer, diabetes, gout, or heart disease, people got a lot of sunlight and so autism and allergies were probably very rare, people who "worked" did so directly for themselves and their families and communities without some complex bureaucratic supervision alienating them from what they were doing, education was very hands-on, religion was likely more a direct experience connected with nature and community for most people, the planet seemed like an endless vista for growth with free land everywhere relative to the number of people, addictions as a "pleasure trap" were harder to get stuck in, etc. etc.. Oh, sure, there were bad things about those times too. My point is not that such times were uniformly "better" (where would we be without twenty-thousand years of dog-breeding to create "man's best friend"? :-), just that the issue is more nuanced than you suggest -- some things have improved greatly, but other things have gotten worse in some ways for a large percentage of the population. Increasing addiction rates are just one sign of social stress and a dysfunctional economic system:
http://en.wikipedia.org/wiki/Rat_Park
http://www.paulgraham.com/addiction.html
http://www.drfuhrman.com/library/article16.aspx
Consider as an example of a conflict between an old way of living an modern society:
"Christian Missionary Deconverted by Tribe"
http://www.youtube.com/watch?v=dr3q6Cid1po
And:
http://www.ishmael.org/origins/Beyond_Civilization/
"Civilization, in effect, represents an attempt to improve upon tribalism by replacing it with hierarchalism. Every civilization brought forth in the course of human history has been an intrinsically hierarchical affair--in every age and locale, East and West, as well as every civilization that grew up independently of ours in the New World. Because it's intrinsically hierarchical, civilization ben
A 21st century issue: the irony of technologies of abundance in the hands of those still thinking in terms of scarcity.
Well that's just the thing, you can't actually patent plans for a house, but you can copyright them. The whole point of the article is that the patents need to be revisited because they are patenting general ideas, and not actual ways in which the ideas can be put to use, which is another reason why you can't patent plans to a house and think that all other houses are in violation.