Slashdot Mirror
Recipient of First Software Patent Defends Them
Arguendo writes "Martin Goetz, who obtained the first software patent in 1968, has penned a thoughtful defense of software patents for Patently-O. Goetz argues that there is no principled difference between software and hardware patents and that truly patentable software innovations require just as much ingenuity and advancement as any other kind of patentable subject matter. The Supreme Court is of course currently considering whether to change the scope of patentable subject matter in the Bilski case, which we've discussed before." Does it weaken Goetz's argument that his description of the software lifecycle harks back to the waterfall days and bears little resemblance to current development practice in open source and/or Internet contexts?
Patents by themselves weren't a problem back in 1968 and shouldn't be problems now either. The issue is how they are used as legal clubs to beat down competition with or simply as a way to make money through litigation alone.
The major problem is that most software patents were not awarded to truly patentable software innovations
I'd like to know this guys opinion on straightforward software patents like "the hyperlink", the virtual "shopping cart", ..., and "patent holding" companies.
We all know that software patents do not help the advancement of useful arts. Period.
patents protect the little guy who can't afford lawyers from big corporations.
I think that software patents are every bit as valid and valuable as every single patent on the wheel.
If you were blocking sigs, you wouldn't have to read this.
The big principle difference is that in the physical world we can discover new things and invent better ways to do existing things. I am yet to see a single software patent that I feel is really a new invention that requires protection. If someone can link me a software patent that they feel was a new invention and that others might benefit from then by all means, I am happy to be proved wrong...
Uh, nope... sorry. Phrasing this as a question doesn't even give the appearance of a NPOV.
You know, there is a difference between trolling and pointing out the flaws in your reasoning. Just saying.
Software is sometimes conceptualized as a machine, and machines are patentable, therefore software should be too.
Software can be implemented in hardware, which makes it patentable even when it's not implemented in hardware.
The following terms are common to both software and hardware manufacturers:
research, competitive analysis, functionality, specifications, operational environment, operating characteristics, interfaces, modules, engineering, implemented, debugged, tested, quality assurance, alpha and beta testing, documentation, installation, training, OEM, component, system, re-packaged, maintenance, warrants, workmanship, guarantees, errors, defects, improved, enhanced, upgraded, and models
and therefore the products of software companies should be patentable just like the products of hardware companies.
Respectfully to Mr. Goetz, my reaction to his points is "So?". Analogies are aids to thought, not legal arguments.
Equine Mammals Are Considerably Smaller
Goetz argument falls short for the following reasons:
1. He assumes that hardware patents are not as controversial as software patents and makes an appeal to equality. Who said hardware patents are off the hook?
2. He claims that the ingenuity required to make hardware or software patents is the same. This is slightly different than a regular appeal to equality in my first observation in that the focus of the ingenuity, according to him, is the creation of the patent; however, patents were designed to secure investments made toward new inventions by granting an exclusive license. Implementing a software patent is considerably much cheaper than most hardware patents (no manufacturing, hardware typically requires software too, shipping, etc). Since the investment towards a hardware based invention may be significantly higher, a patent makes sense. Many software; however, can be implemented at very minimal costs and therefore SHOULD expect very limited ROI or none (no patents). The focus of comparison is on the inventor's investment, not on the innovation. Patents are for large investments that need ROI to recover losses from the invention.
With that in mind add increasingly cheaper tools, a wealth of free information, enormous marketing capability and open source to the software equation and you quickly begin to see why software patents do not make sense. The patent process cannot even keep pace with the implementation of most software because of the low investment and rapid market deployment.
As technology becomes more accessible and cheaper, the cost of investment for new creations will decrease, hence the patents length (its mechanism for delivering ROI) should ALSO be reduced. With software, think approaching zero (in cost -- hypothetically). Without adding a market implementation clause and ROI cap based on actual provable investment, Software Patents do not fit our patent structure.
Trying to install linux on my microwave, but keep getting a kernel panic...
As if life wasn't already complicated enough. Now I have to add "patents" to my list that already includes...
1. Eat ..."Patents"
2. Sleep
3. Fuck
4. Work
(5)
Ah well, I may have to remove "work" from the list so I can have more time.
The choice of implementation for computer functions is a pure economic choice which mainly has to do with cost, speed, and flexibility.
That's absolutely correct. And the reason that people choose software when they can get away with it, is because it generally much cheaper, faster to implement and more flexible. Which in turn results in completely economic effects of patent claims that include software implementations compared to claims that only cover hardware implementations.
The whole argument about software companies also performing state-of-the-art R&D should be irrelevant from a patents-point-of-view. Patents are not a natural right (they're not awarded because someone deserves them for whatever reason), but they are purely economic tools. And blunt and wide-reaching ones at that. Applying blunt and wide-reaching tools without discrimination just because some things are alike in terms of intellectual achievement does not necessarily have the same economic effects. Again: patents are not rewards, they're tools, to be applied with care where it makes macro-economic sense, not to protect certain business models at the cost of a huge overhead for an entire sector of industry.
And according to various economic studies (also includes citations from some political documents, feel free to ignore those), patents indeed don't have overall positive effects on in the software field. In fact, the 2003 report on innovation from the FTC came to the conclusion that patents are not very important to innovation in the semiconductor industry either (although slightly more so than in case of software, and their downsides are slightly less detrimental in that case). So the whole comparison with hardware and software does not necessarily lead to the conclusion that the author of the article envisions...
Donate free food here
Patents themselves were created to increase competition.
Nowadays, with our global marketplace, if your brilliant idea is already patented, just go to a country where it isn't.
If you do get a patent, someone in a country where it doesn't apply would gladly make millions for you.
Software patents are ludicrous. Not only can someone copy an idea somewhere else, they can let your neighbor download it before you can even run next door to slap their wrist.
It's my opinion that most of the software patents that I've seen fail the obviousness test for someone who is qualified to be an expert witness. Most of the software patents that bother me are things like one-click or hyperlinks. There may have been a point in time where some part of their functionality was patentable, but generally they are obvious extensions of prior art. You see this frequently in hardware as well.
The development methodology as mentioned by the OP doesn't change the presence or absence of the vision and insight that something that is patentable should require. Its important not to confuse the what with the how when considering this. It is a pity that for the most part software patents were awarded by people who didn't know the industry well enough to know which of the ideas they were seeing were obvious extensions of prior art. More importantly, the practice of patenting demonstrated but undefended prior art has hurt us all by requiring patents for things that should have been in the public domain in the first place. I sometimes wonder what would have happened if it had been possible to patent movable type or a method of translating human understantable instructions into commands to control the operation of machinery.
As a further red herring I should mention that the use of patents in general doesn't work well against most entities that you would most want to be defended from. A large number of players are using the slowness of the legal response to make their money and then vanish. It works against large corporations and other entities that are stuck at a fixed address which is why they tend to try to gather as much IP as possible related to their business in self defense.
This is an old argument going back at least a century. One line statements of position certainly don't adequately express an appreciation of the complexity of the issues.
The problem isn't necessarily _software patents_ as they are filed today.
The REAL problem is that so many of the current patents being filed are such crap, software or not.
Patents are supposed to cover INVENTIONS. Lots and lots of patents today cover IDEAS. Small, incremental and for a person skilled in the field, obvious, solutions to everyday problems that occur during product development in that area.
"But solutions always seem obvious in hindsight", you may say. Sure, some do. But real inventions are still clever. Most patented crap today isn't.
How was the LZW compression patent different from patenting any other mathematical algorithm?
So, by his argument, if calculus were invented today, it could and should be patented because it "require[s] just as much ingenuity and advancement as any other kind of patentable subject matter"? I'm sure Zombie Newton would be pleased with the royalties, but I can't see how making such things patentable would be beneficial to anyone else.
In practice, there is a world of difference.
When you develop a drug or a new car engine, you have to invest hundred of millions of dollars. Spending a few millions on patent lawyering is nothing.
I can spend a few weeks developing a program in my spare time for an investment of zero (0) dollars, and be infringing on some stupid patent without knowing it. I don't have a few millions laying around to pay some scumbag in a suit.
And that's just one of many arguments against software patents. But that's the main one from a social contract point of view. You can't make laws that people can infringe without knowing it. Lawyers might love them for all the business it brings them, but it's just fundamentally wrong.
Arrrg, I've just been goetzied! If you have a team of writers working on the mechanics of a plot for a weekly show, does that make it a machine? During the devolpment of a script or novel, it's reviewed, edited and checked for continuity. As a software author myself, I find Mr Goetz views on the subject fascinating and would like to subscribe to his newsletter.
100 Years ago: Patents give us an incentive to innovate. If you copy, then you will loose.
Today: Patents increasingly used to stifle innovation. Nowhere is it more apparent than in software. You can write something new, only to find out that mr.big_company has a broad patent that makes your idea theirs.
Future: New ideas in software are gridlocked because of over-reaching patents, and we find that places without a history of innovation have leap-frogged us because the developers there write code without a fear of being sued.
If we want to remain leaders in innovation, we have to release the chokehold on our own necks.
Patents worked when we were competing with ourselves. That's not the case anymore.
One of the major differences that could define this debate is the ease of production of software vs. hardware. While many of the comparisons Goetz uses to make his point are indeed valid, practically only large entities have the power to manufacture hardware, making patents a practically enforceable tool in the market. All the disputes that arise are between companies that have a team of lawyers and an account that can fund them. Meanwhile, software readily reproducible by anyone with a few hundred dollars, intelligence and time (or a few thousand people with less time each). The production of software is not really comparable to hardware in this aspect. It would be as if anyone with a simple soldering iron and scrap metal could produce an original processor. Intel and AMD design concepts would be mixed up and scattered all across the globe. And not even this captures the case of software, because of how easy it is to share and copy software.
Add to this the applicability of real world paradigms to software, which results in things like shopping cart patents, and the fundamental emphasis of patterns and abstraction across problems in software development and patents begin to look less and less reasonable both in terms of enforceability and applicability.
They are a political and economic tool to advance economic growth and development of new technology.
In the software world where there are many people working on the same problem, and finding the same solutions. Patents are of no use whatsoever.
I would like to be proven wrong though. Is there any examples of software that would not have been developed without software patents?
Max M - IT's Mad Science
I think there are two reasons for this: First, there are a lot more people building meaningful stuff with software than with hardware because it is cheap and easy. All you need is a computer for a couple of hundreds of bucks, and if you like you get all the rest for free (as in beer). For everybody who wants to write software, a patent could be in the way.
Second, and more important, many if no most patents lack substance and should not be patents at all, and for software this is even more obvious than for hardware. What is this one-click-buy-stuff about? Should there maybe be patents for storing text documents on a harddisk or using pixels to show text on a screen? I probably wouldn't mind a software patent for reasonable inventions, like a O(n) sort algorithm. But it's so incredibly hard to come up with such inventions ...
I have a past at a global player in the electronics market and I've seen some drafts for patents that in my opinion should never be granted. However, giving the current patent system, the number of patent applications is a value in itself, regardless of the substance behind.
All that said, as long as the patent system is not fundamentally changed, software patents are going to do a large damage to both the economy and the consumers.
All software is math, no exceptions.
You can't patent math.
Well, in the current state of things, you /can/ patent math, and that's something a lot of people are hoping is revoked.
Imagine if something like calculus had been patented, or the quadratic equation?
We'd be fucked, as a species.
For a concise, well written and much elaborated explanation, see "An Explanation of Computation Theory for Lawyers": http://www.groklaw.net/article.php?story=20091111151305785
Now, that's written specifically for lawyers, but it should be clear enough, perhaps more clear because of that, for most of the Slashdot crowd to get it - all software is math, no exceptions, and math should not be patentable.
In a word, no. It doesn't matter how or in what context you come up with something, it's whether it's sufficiently unique or not. And in any case, I am not sure why you refer to "current development practice" since that varies wildly from application to application, and the "waterfall" process is certainly still in full force for many serious programming applications.
Brett
Most people always mix two different things:
1. Inventions
2. Transformations of already known things to computers
- Inventions are for example new algorithms for speech-recognition that were unknown before.
- Transformations are just things that already arer know to the normal world and now will be done with help of computers. This is nothing new, should not be patentable and is called "Business-Idea".
So once and for all:
Transforming things that already work in the real world, and whose parts are already known to computers should not be patentable because they are not new!
Example: Software-Development with Speech-Recognition:
- is already known, just imagine boss that says to you: Pleas write funktion x with y.
- To do this there are no new things for the computer to do, just use your favorite development tool and speech-recognition software.
best regards,
Jan Kechel
It's not the design practices per se, but rather the software market that does. It is significantly easier to build an independent, popular product today than it was 40 years ago. The main argument for patents is to protect a company's investment in the risk of doing R&D to get a product to market. Pharmaceuticals are one of the last industries where that is probably still a valid concern. For software it is at best, rent-seeking, and at worst a cudgel with which they can beat smaller players or extort big, lumbering giants.
No, really. And I say that as an outspoken opponent of software patent.
His most important sentence is It is obvious that software products are not "software ideas". - and that's what's wrong with the patent system. A hundred years ago, you had to submit a working model to get a patent. These days, you can patent software ideas and business processes.
When we speak about "software patents" we are really talking about monopoly rights on algorithms. And that's as stupid an idea as a copyright on a number.
What the author apparently means when he talks about "software patents" is a complete product, a working implementation of a concept that can be demonstrated to do what the list of claims contains. That's a bit of a different animal than about 99.999% of the software patents issued during the past 10 years.
And I agree that a bit of protection on that would be nice for those who invented it. We can discuss whether or not patents is the right tool, or copyright, or some new form, but that's not important.
Assorted stuff I do sometimes: Lemuria.org
I would argue that the 'ingenuity' required to make hardware vs. software is most certainly not equal. Many software developers might not understand why, but that's because they are not hardware developers. A software analogy is, for example, defining specifications versus defining implementation. A hardware-ish example is the analog / digital divide. Hardware designers know, that nothing actually is digital.
The concept of patenting software has so many logical weak points it's not even debatable in my opinion.
The main difference between software and hardware development is not how clever once can be in either medium, but in how difficult it could become or how much dependency there is with previous works. In hardware patents, there will still be some dependency on previous works, but not so much as those things that are typically covered by copyrights such as software or other language based works.
But if the arguments are "because software can be clever" or "because it can make a bunch of money" then it's about time I started trademarking every clever turn of a phrase I can think of and suing people every time they say it in the same way the music industry wants to sue people who sing at restaurants. (Yes, I know there are some differences, but you can't deny how incredibly stupid it is in general)
I think the end game for software patents are not about the principles of the patents themselves where many are weak to begin with (rife with prior art that is never found or presented prior to granting) but with the weak patent system and how it is being abused, how it is used to harm and monopolize an industry where patents where intended to benefit them. When a practice no longer meets its intended purpose, it is time to change that practice to match its purpose or abandon it all together, whichever is more effective.
When you refer to 'illiterate people', do you include people who can't form complete sentences?
We all bemoan the software patent situation largely because it's a market with nearly no barrier to entry otherwise, but the system isn't good as it stands for physical inventions either. Infringing a copyright is hard to accidentally do, but infringing a patent accidentally represents a huge risk to someone. Even if you didn't think it was patentable, if it hits you, it will be pricey to defend yourself even if right. Exacerbating this problem is that patents are extremely long lived given the general rapid pace of the software market (don't have to invest a lot to 'ramp up' production, delivery to customer is potentially instantaneous, even if it takes you a full year from patent filing, you can probably earn the worth of it in five years or less.
Assuming you do have a unique and genuinely innovative idea, unless you can tap the resources of a large corporation who will assume ownership of your idea, it may be prohibitively expensive to get one as an entrepreneur. Assuming you do have the resources to acquire a patent, the process takes a relative eternity to get approval. If you need to go to market *now*, you have to run under the hopes that your patent will be granted. In the intervening area of time, competitor's may bet you won't get it and drive you out of the market. Even if you eventually will get the patent, that competitor may drive you out of business before your chance to seek damages comes (even if retroactive, the damage is done). Assuming you make it to market and have the patent granted, going after a large company is also likely not to work relative to a large company going after a small company given the legal resources available. So a good shot at patents require the resources more available to large companies, and the patience of a large company that has enough bankrolled that can afford to be patient, and the resources to defend that patent vigorously.
XML is like violence. If it doesn't solve the problem, use more.
He would, wouldn't he?
Goetz argues that there is no principled difference between software and hardware patents. In that respect, he is right. There is no fundamental reason why an invention has to be made of something physical. But that is not really the point.
Patents, like copyrights and trademarks, grant the user rights by taking rights away from the rest of us and granting the inventor or composer a monopoly. In some cases, this may be necessary or desirable to protect the creative individual from the clutches of capitalism, red in tooth an claw. However, monopolies are often unnatural things that have to be maintained by states passing laws to grant the rights, and courts enforcing the rights. This all comes at great cost. This is not something we ought to do unless there are proportional benefits.
If someone makes a piece of music, they might be annoyed if someone else copies and sells it. They would feel entitled to some sort of redress, and rightly so. However, most contemporary music makes almost all its sales in the first seven years or so. This would seem to the the natural scope for a copyright. However, if copyright covers seven years, then why not twenty years; the writer's lifetime, the writer's lifetime plus twenty years, and so on? Suing someone for singing "Happy Birthday To You" is a clear sign that things have gone too far.
The same applies to patents. While software patents may not be harmful per se, we see patents for pieces of coding that are so well-known as to have no known 'inventor' being patented (usually in Texas - not a good sign). This is practically the same as copyrighting a folk tune which has no identified composer. There is also no good way of searching prior art for software just yet, because this was not necessary before. The same sort of argument applies to business practice patents. It surely makes sense to disallow both classes of patents because their presence does not seem to do anyone any good, other than patent trolls and their lawyers. This is surely better than to allow them because we cannot argue "why not?"
He's right. There's no principial difference between software/method and hardware patents. It's always the use of an idea that gets monopolized.
He's wrong. Patents are bad both software/method and hardware. They're all granted monopolies backed by government use of force against people, for gain of some people and loss of some other people. That's not the proper role for government in a good society.
Too much social networking, very little reading of real books.
Constructing a transform table from input/output observations will tell you what the black box does but it will not tell you how it does it. For example let's pretend I am a genius and I have figured out an analytical solution to the three body problem. You observe my black box and create a transform table. However without my insight you are still left scratching your head as to how the black box can perform the transformation so rapidly and accurately when the only known way to approximate a solution is via numerical analysis.
Clean room is a black box by another name. Again by definition you cannot know how the black box performs it's task (the algorithim). Sure you could guess the right answer but if it's that's obvious then why would it be patentable?
I certainly don't want people to give away their "hard work" and throwing out copyright would also throw out the GPL that my comrades here at socilistdot are so fond of. However if the algorithim/invention is so obvious that it can be guessed by observing a black box then I would argue the inventor has not "worked" hard enough to earn a state sponsered monopoly on the idea.
Dissasembly is of course a white box that allows you to copy the algorithim without necassarily understanding it. This may or may not come under the perview of copyright or trade secrets but I will leave that argument to someone with a better understanding of law.
"Stop posting grossly incorrect statements about a field you know nothing about." - Have you checked your arse for bite marks?
And did you exchange a walk on part in the war for a lead role in a cage? - Pink Floyd.
Goetz argument isn't weakened by using the waterfall model. He's just comparing phases of the waterfall model traditional manufacturing so that's hardly an argument at all.
His argument seems to be that since you can implement any algorithm in hardware just as well as in software, both ways are equal and thus patentable (given the current patent laws).
His mistake is that you can not and should not ever be able to patent algorithms (since it's math), only a specific physical machine that "executes" the algorithm.
No one has (yet) claimed a numerical method like Runge-Kutta should be patentable. However, if you find a novel way of implementing Runge-Kutta using sticks and rocks, you can patent that particular physical machine.
(Ok, some crazy patent lawyer or politician has probably claimed algos should be patentable but no "real" people with braincells)
This was a very thoughtful explanation of why the system currently justifies software patents, and it does effectively pull my thinking a little more to the center on the matter; however, the author is justifying software patents from the perspective of someone who was programming when nothing fit the criteria of "obvious" in the field, when everything was new and innovative. The examples he gives, handwriting analysis, speech software, etc, are all very impressive, but I see them as bodies of creative work, collections of software engineering techniques that add up to something special, but also something anyone else should have the right to create if they have that same foundation of common software engineering tools in their minds. This makes software copyrightable, but not patentable in my view. For instance, I've been learning design patterns for OOP this year. The GoF put a great deal of effort into formulating these solutions, and there is nothing "obvious" about them. Should they have been patentable? If design patterns were patented, it would cripple the progress of software innovation, but the GoF published them for everyone to use, and, as a result of their wide-adoption within programming, they have become "obvious," and would no longer fit the criteria for patenting. We work in a field of accelerating cultural-adoption, as more people become programmers, the less esoteric our field becomes and the non-obvious innovations of 20 years ago are the cultural norm of today.
i ~ Celebrating Science, Cyberspace, Speculation
I can see some justification for software patents. However, what I have never seen is why they should last as long as patents on say, drugs. To make a new device or chemical can require years of testing in the laboratory. With software, I typically design, write and test each idea in the space of a week or a few months. As well, if the patent lasts significantly longer than it would take a second inventor to discover the idea and publish it, then the patent is becoming harmful to society rather than helpful. In short, because it is easier to invent software than physical inventions and because most software patents would be discovered soon anyway, I believe that software patents should be much shorter in term than physical patents. I want to know why someone thinks that a computer invention from 1990 should still have a monopoly. If there is a case for software patents, it would be for a much shorter term, like maybe 3 years.
I don't think he was calling the lack of patents socialist. he was calling the slashdotters' naive agenda socialist.
.. Blub falls right in the middle of the abstractness continuum. -- Paul Graham
The winner of the first Mega Millions Lottery says state lottos are a good thing, too.
You are welcome on my lawn.
Something very big just flew over your head.
.
There is no inherent right for anyone to restrict others copying / doing with whatever they see and obtain. Patents are a trade-off, a contract in society, but no right.
The only reason we have patent laws, was because people thought it would promote science. Now we see it is only to promote big business to hinder fair competition, or for small companies to hurt the big players through litigation. Nothing useful is produced in a patent, as the knowledge in a patent is legally dangerous knowledge to everybody. Thus it is not being useful for society. Patent-laws should therefore be revised, so it is again in alignment with the original purpose.
Artificial monopoly means "artificial". You can disregard nature only so long. When we have devices that can copy objects. Will you still restrict copying, when it's about physical matter rather than just information?
Rather than seeing the property of information being cheaply clonable, we should be mature about it, and see how such a useful property about software and information technology can further society - both technologically and spiritually (they go hand in hand).
The world is larger than just your own company..
http://www.debunkingskeptics.com/
My biggest complaint about software patents (ok patents in general) is that you can patent the idea without ever having to invest the effort and time to create the software. For example, AMD owns a patent on location based reminder. But as far as I can tell they do not sell any software that implements the patent. Yet this patent keeps me from making the effort of creating my own software that provides reminders based on location. This whole patent thing is just ridiculous. I'm going to patent "patenting ideas" ;) maybe finally I will have my own patent.
I've always said English was my second language. Had Romeo and Juliet been written in C, I might have understood it.
Any physical system requires significantly more investment to create, requiring raw materials, machinery, etc. and the marginal cost of production is non-zero.
By comparison, software costs a pittance to start writing, effectively just being the cost of a computer, say around $2,000, and has a zero marginal cost of production. This is orders of magnitude lower than any material good, and why software already has a breakneck pace of innovation: because new players entering the game require little startup capital.
The only way I would support software patents is if the economics of software production are taken into account. The best compromise I've devised so far is to reduce the duration of a software patent by an order of magnitude to around 2 or 3 years. This gives innovators a chance to exploit their ideas for a short time, without shackling the rest of the industry with what quickly becomes an "obvious" idea.
Higher Logics: where programming meets science.
If there's no principled difference between hardware and software, that doesn't imply that software should be patentable, it implies that hardware SHOULDN'T be patentable.
If that seems a little extreme in the opposite direction it's probably because there IS a principled difference.
Theoretical arguments aside, empirically software patents are unnecessary. The software industry innovates at a faster pace than any other on the planet. Most of what we create is never patented and doesn't need to be in order to be useful. The only people in the software business who need patent protection are professional litigators. If software patents disappeared tomorrow, innovation would INCREASE because we would no longer be working under the constant threat of a lawsuit.
So by this logic, any algorithm that is patented would need accompanying source (real full code) that would implement the algorithm, thus at the end of the patent period the code itself would be unencumbered and could be used without restriction. Thus effectively requiring them to decide which to protect the software with, copyright or patents.
I like this solution to the problem, the tough part would be actually getting the requirement to include fully operational source in all software patents.
I agree there is no difference in *principle* between a hardware patent and a software patent. There is a difference in *practice* and patents are a *practical* matter. As a mechanical engineer friend of mine likes to say, "In principle there is no difference between principle and practice but in practice there is."
Patents aren't real property, they are a social bargain designed to maximize utility. Judged that way, most software patents are a failure.
Goetz makes a good point about the need for research and investment in *some* software projects, and he almost puts his finger right on the problem:
There is six phases in the life cycle of software products: Definition, Design, Implementation, Delivery, Maintenance, and Enhancements. Let’s look a little closely at these phases and you will see how closely they resemble characteristics common to all manufacturing companies.
So why don't we reward companies that put effort into maintenance and enhancements with a monopoly? That would be even better -- in the software world. You see, software *is* different; it has the same topology, but not at all the same shape.
Patents are supposed to reward originality. What is originality? Well a patent is supposed to be non-obvious. Well what does that mean? Well, I suppose that means the patent must demonstrate fresh insight. But what does "fresh insight" mean? It's just another way of saying "originality".
I think what it all comes down to is that invention is about *Definition*. A real invention doesn't just restate known solutions; it restates the problem. The example I like to use is the mechanical watch. Before the use of the balance wheel (in essence a wheel shaped pendulum), the problem of making a compact timepiece more accurate could be stated this way: "How do balance friction and force more precisely in order to obtain more precise performance?" After you hit on the notion of the balance wheel you *still* have this problem, but you've just got much further down the precision road by briefly considering an entirely *different* problem: "How do I get the benefits of a pendulum in something compact that does not require the timepiece to be held in the same orientation relative to gravity?" By setting aside the problem everyone else is working on, you can arrive at the essence of what a pendulum does: balancing the acceleration of a known mass around a pivot (a wheel instead of a lever) against a known force (a spring of known elasticity instead of gravity).
The majority of the trade-offs you choose in making a timepiece more mechanically precise aren't like this. They're just different *designs*, not inventions, because they take a different stab at optimizing a problem with no perfect solution (which probably would require a largely frictionless device driven by an endless source of energy).
The process of definition is *different* for software than hardware, because computers are *general purpose data transformation engines*. In software engineering, our lever or the inclined plane is the Turing equivalent language. With this one machine we can literally do *anything* that is computationally possible. It *defines* computationally possible. This shows right off the bat the far greater flexibility the intellectual tools of the software designer are over the mechanical designer or civil engineer.
If you want an analogy for the relationship of hardware invention to software invention, think of the relationship of civil engineering to building architecture. In civil engineering, innovation is highly risky and very, very rare. A civil engineer prefers to be able to adapt a known and trusted design pattern to a problem. Where there is invention, it is driven by necessity. When there is no *room* to put conventional bridge piers on the site, he'll be forced to come up with a cable stay bridge.
Now what about architects? They design useful things all the time. Why can't *they* claim patents on th
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
effort and ingenuity. I say OK, let's patent software, but first allow for patenting of mathematical theorems and ideas of proofs. (A little disclaimer - I am a mathematician :) , and this is sorta joke, but look where such reasoning leads to). I think that what should be patentable is some working device, not a general idea and not a knock-off.
and always has been regarding this matter.
The very nature of software makes patent ludicrous. I can come up with seven different ways to writes something and have it compile down to doing the same work. It's like patenting sentences.
The Kruger Dunning explains most post on
The hammer/executable is neither patented nor copyrighted. But making a copy of the hammer/executable would be patent infringement. So the effect would be the same (you can't copy this unless I say so) except the duration of protection would be about 20 years instead of 70+ years.
That sounds reasonable. The parts of the executable that are considered "machine" rather than "expression" should be copyable verbatim after the patent expires.
Comment removed based on user account deletion
The effort of reverse engineering is as much as writing your own BIOS. It's worse in fact, since you have to find out all the bugs too, if they're relied upon. See Wine for an example.
The ONLY reason for such is interoperability.
And copyright WAS NOT intended to stop people being able to interoperate. It's why we don't copyright dictionaries or language definitions.
Comment removed based on user account deletion
Comment removed based on user account deletion
Comment removed based on user account deletion
You didn't answer the original posters question:
"Where would we be if Taylor had patented his infamous infinite series, for example? I don't really see the difference between this and a Taylor expansion in fact, but maybe I'm missing something?"
Just because it is ingenious, doesn't mean it should be patentable. In fact, you could even say, that's a reason why it should NOT be patentable. Indeed, where would we be if every discoverer / inventor in the world had put restrictions on the knowledge we now take for granted.
Riding on the back of Giants.
http://www.debunkingskeptics.com/
Neither reality nor logic support that statement. Lots of patented machines seem obvious in retrospect. The hard part isn't making it work once you see the design. The hard part is developing a design that solves an unsolved problem. These designs need protection because they take a lot of time, money, and ingenuity to develop and many of them would never be developed, much less published, if there weren't an incentive like a temporary monopoly.
A machine that can easily be cloned without examining somebody else's invention is obvious and ineligible for patent protection. But a machine that a person skilled in the art can't think of when faced with the same problem is not obvious and society will reward somebody for inventing a solution and making it obvious to the rest of us.
...the problem is that software patents don't work like actual patents.
I made a rant about this some time ago and put it on my mostly defunct blog. I will summarize:
Actual patents have industries making an choice to use them. 'Should we license this tech for 4 cents an item from a competitor, and save 10, or should we attempt to research our own way, or maybe we should research on top of their patent so we can cross license our addition back for the original patent'.
It is an actual, calculated decision.
Software patents, OTOH, fall into two categories. The first is 'patents that should not be issued'. These aren't even actual 'software' patents, they're 'patents on how computers get used', idiotic things like RSS feeds and one-click shopping, stuff that isn't original in any way, shape, or idea, and should never have exists.
But let's forget about those, and talk about reasonable patents, like on MP3 encoding and stuff like that. That specific one is 'Here is a process to strip out audio information from a file and still have it sound mostly the same to people'. It's certainly patentable, it required a lot of scientific research, and it doesn't attempt to cover all ways to strip audio information and reduce the size of a file, just a single specific one.
The problem there is programmers don't choose to use a patented process and pay royalties. Ever. If they hit a (reasonable) patent, they just spend a little more time and effort, and perhaps a bit more CPU and filesize, and work around it.
It's only when they don't know they're using a patent that they just one. And then, a decade later, when it becomes popular, suddenly everyone realizes it's patented.
People would have invented their own method of doing that, but now it's too late, everyone is using that format, and we're all screwed.
Software patents are all either a) too broad, and shouldn't be issued, or b) submarine ones. The actual legitimate situation of real patents, where engineers choose to use the patent, or use something else, is entirely nonexistent.
If corporations are people, aren't stockholders guilty of slavery?
Hehe, before you claim others to be misinformed, I suggest actually trying to see the argument I and the other poster was trying to convey.
Since I think this is a bit important, I'll give you the spoonfeeding. although someone else could probably do a lot better to say it more clearly:
I think you're are a little misinformed, it's the other way around. No one has an inherent right to force an inventor to reveal his invention to the world.
Nobody is forcing an inventor to share any invention with the world in any way. If your inventor wants to stay bitter in a cellar all his life and make stellar inventions, he is free to do so. He doesn't have to show anything to anyone, or give them anything, but he will not receieve much from society either (so he better have a millionaire dad or something).
However, there is absolutely no natural right for said inventor to restrict what others do with what he distributes to them - wether it be information or physical devices.
It is in fact laws like the patent-system that force their way on people, not the other way around. Laws should not deviate too much from the natural order of things, or you will experience such things as the fall of civilizations (which has happened many times before..)
That's what the world was like before patents. Trade secrets were employed to prevent competitors from stealing ideas to maintain monopoly and high profits.
So what?
Today, no company that actually makes something, not those patent-holding companies, can actually read patents, for fear of being sued for triple damages for knowful violation. Of what use is such patents. Where are the benefits of it? To stop other companies from making business, ie. stop competition with artificial government-sanctioned monopolies. Oh yeah, so much for promoting sciences and arts too.
There still are trade secrets.. Do you really believe anybody put anything really worthwhile in patents? They are unreadable crap. Better use as toiletpaper than anything else. The REAL stuff is still in people's heads, in public domain / free software or trade secrets..
But the inventions died with their inventors.
This is a problem with the inventor's altruistic mentality, not the system. Today, the world is entirely different. Information flows much more freely. Any discovery today is usually rediscovered many times, in different parts of the world. What is limiting innovation today though, is a creaking patent system designed for the previous century. Rather than promoting innovation, today the patent system is raising the bar to invent and implement products in a world where a 12 year old kid can make his own computer program. Your glorious patent system could squash such 12 year old kid like RIAA could squash an old lady with Kazaa. RIAA is out of sync with the reality of things, and so is the aging patent system.
If the govt were to remove patent protection, the inventors would go back to trade secrets. And if that won't work, they probably won't release any products using that invention -- they don't want to do free R&D for some big company.
While a small company could have problems, they can still operate under the radar, as they still rely on now anyways. They still have the edge and know-how, and can be bought up, with or without patents. Wether some industries, like medicine, should be awarded patents is up to debate, but the software industry is a totally different industry altogether, with very low barrier to entry. One of the few things blocking progress is the aging patent system.
So good luck! The enterprising part of us can easily reverse engineer most products, or figure it out for ourselves thank you! We don't need litigous bastards to set toll booth on our profession of software engineering. Computer science is built on freely and open information, as is any science. Without this free access, society's progress will be much poorer, and education only beneficial to the already rich
http://www.debunkingskeptics.com/
"Does it weaken Goetz's argument that his description of the software lifecycle harks back to the waterfall days and bears little resemblance to current development practice in open source and/or Internet contexts?"
No, but it does weaken your synopsis to suggest that current development practices don't resemble waterfall. Not sure which Ivory Tower of Google in which you reside, but corporate America is hardly a progressive bastion of methodology. Waterfall is the rule, rather than the exception even today.
Sorry, it's definitely been earned.
You are confusing your concept of "earning" something, and what is granted by the government.
You can "earn" a salary, because it is a contract between you and your company that you work for x hours a day.
You can "earn" license fees, because it is a contract between the society and the government that patents grants you a time-limited monopoly on said implementation (not idea, so it should be possible to implement the same idea using a different implementation anyways).
However, all such "earning" is only because of the existing contracts in place! If we change the contracts, we change the whole playing field. As such, there is no natural right to earn anything at all! You can scream and shout for your supposed "earnings" and "rights", but unless there is a contract in place, it is all in vain..
Example:
You could work for a month, the company refuses to pay you at the end of the month.
The natural way to solve this dilemma, could be to demand half a month advance pay, or a full month.
Because of contracts in society though, you can sue the company instead, but many times since the company is bankrupt, you will be the last to get any money anyways.
So by not following the natural order, you may in this example, risk not getting your money.
You could add a contract in society, a sort-of insurance, which pays your salary, if the company should go bankrupt however. This could solve this dilemma by adding more complexity to the system.
Essentially, we decide how many contracts we want to have, and what they should be. If patents are restricting innovation, then maybe the whole system should be revised on a per-industry basis? All of history shows that the more complex a culture becomes, the faster it falls when it falls..
Not that I really believe that will happen now, but I do think we can do much better at utilizing the power of free information flow that internet is challenging us with, than we're currently doing.
http://www.debunkingskeptics.com/
Right or wrong (and by a matter of degree and your point of view), we all owe a lot to Marty for making it possible for us to code. Marty broke software free from the monastery and the monks of IBM. He created the entire software industry as we know it. (Of course he is human, and indirectly helped destroy mainframe software by selling ADR to Ameritech, which then sold it to Charlie "The Craw" "Not Wang, Wang!" Wang and CA.)
Marty is one cool cat, and let us play ping-pong in his luxurious office. Even if he was there working! As far as I am concerned, what Marty wants, Marty Goetz!
My user name was a mistake. Input wasn't restricted, my bad.
Comment removed based on user account deletion
> I am yet to see a single software patent that I feel is really a new invention that requires protection....
The one example I am aware of is the write-anywhere feature of WAFL -- the NetApp filesystem. This unique and brilliantly conceived feature enabled cheap snapshots, which in turn enabled the development of a suite of data protection products that made building a billion-dollar company from scratch possible.
I'm not sure whether patent protection had anything to do with NetApp's success though. There are clones of this technology out there now--and lawsuits to go with them--well before the patents are set to expire (ca. 2015?). But the cloning didn't happen until NetApp's revenue got big enough to generate envy. That kind of analysis would seem to indicate that the original motivation for patents--to protect small inventors--is essentially not operative anymore.
The problem lies in the analogy: pieces of code to parts of a machine. Seems correct, at first glance. But more properly the analogy should be pieces of code to the physics used to build a machine. After all, software has no physical parts. In physical engineering, one must consider the properties of the different materials and find a way to construct them to achieve a purpose, and that is certainly patentable. But in software engineering, there are no materials - and all platforms are equivocable; an algorithm is an algorithm is an algorithm, regardless of the machine/processor you run it on. When you patent an algorithm, you're not patenting, say, the wheel, you're patenting SIMPLE HARMONIC MOTION. And there precisely lies the problem. Or as we all know in our heads, more succinctly, "you're patenting math". But explaining how that is to a judge and/or jury is where you run into difficulty. Copyrighting, on the other hand, is a different thing altogether. Although patenting brush strokes and geometric figures would be absurd, copyrighting a book or a logo is obvious. Likewise, every software company would be broke were software "pirating" not even a term. Then again, so would they be if there were such a phrase as "algorithm pirating" - the equivalent of a software patent. Now if they were to start patenting books, or, rather, topics to write about... I could go on, but why?
My basic argument against patents has always been that with 6 billion people on earth, it almost always benefits more people to allow the completely free exchange of ideas. As you mentioned, there are many things that "you'd never think of in a million years", but those same things might be completely obvious to a very smart person. The problem with patents is when a novel, useful innovation is completely obvious to at least *two* smart people; at that point the patent system is broken. Simply because one of those people filed a bunch of annoying paperwork first shouldn't prevent the other person from using that idea in any way that they desire, including sharing it with the entire world for free. The exclusivity of ideas has to actually have some basis in fact for it to be applied in law, and I don't think there are very many ideas that are truly unique to a single individual. Patents, if they exist, should be treated much more like copyrights. In strict legal theory, this is true, and a patent is only infringed if every claim of the patent is found to be infringed by another invention. Unfortunately, simply the threat of legal action is often enough to stifle innovation. It's much like if Disney had obtained a copyright on "cartoon animals" or "animated fairytales" and could happily sue any other company producing ideas that fit within those broad categories. Most patents are issued in similarly over-broad terms, and that is the heart of the problem. The other problem is that the business model of R&D, creation, profit is going the way of the music industry. With rapid prototyping, computer simulation, and cheaper CNC machines, the cost to develop new innovations is being decreased dramatically. Does it really make sense to issue a 25 year patent for a new widget that was drawn up in autocad in a week, stress tested with finite element analysis for another week, and then popped out of a rapid prototyper the next day and tested and finally sent to production on a CNC machine at the end of the month? If nothing else, the length of patents should at least reflect the realities of a free market with amazing new tools for rapidly creating new inventions. Often, people don't even realize they need something until they encounter a problem, and at that point it is pretty obvious how to solve it. Just because someone had the same problem within the last 25 years should not give them an exclusive right to the solution.
Comment removed based on user account deletion
Plenty of people have used calculus and other mathematical discoveries as examples of things that would have been greatly harmful to society if they had been patented. But something directly relevant to the discussion is the P ?= NP question. Suppose that someone developed an algorithm tomorrow that solved NP problems in P time, and patented it. Suddenly, one entity would have effectively complete control over many aspects of society. Not only would they be the only entity who could effortlessly break any encryption scheme, but they could also solve almost all the hard problems of physics simulation, including simulating nuclear weapons or global weather patterns or the future positions of solar system bodies, including asteroids or comets that could potentially impact the Earth. Obviously the governments of the world would step in and quickly violate the patent for themselves and start throwing everyone who used Tor or freenet in jail, but where would that leave the rest of the normal people in the world? This is a completely theoretical argument, but it underscores what the belief in the patent system means; that any advancement, no matter how big or small, is worth locking up for only a few individuals to enjoy regardless of the relative costs or benefits involved in creating the advancement or using it, or the power that it can give individuals over everyone else.
Comment removed based on user account deletion
Comment removed based on user account deletion
Yeah, ok, Martin Goetz defends software patents. That's like Neil Armstrong defending moon missions, Adam Osborne defending the Osborne-1, or George W. Bush defending Iraq -- there's no other possible answer from this guy. That doesn't make him wise, or correct.
But reality sets in and there are many, many questions. First of all, every software patent issued before the mid-to-late 1980s should be invalidated, or at least, subject to re-evaluation. Why? One acid test of patentability is supposedly "is this obvious to one skilled in the art?" Well, the PTO didn't have any software people until some time in the mid 1980s. So no prior software patents could possibly have passed that test, simply because the examiner was not one "skilled in the art".
Not that any of this is unique to software patents.. the whole idea of many business method patent approval is possibly worse. I mean, really, if you look at these, there's a whole class of prior art that's existed for decades, perhaps millenia .... at the phrase "on the internet" to it, and violia, it's a frackin' invention. NOT. That's an insult to actual inventors, and the idea that the USA is (or at least once was) an innovater on the global scale. This was nothing more than a gift to large companies with huge piles of cash to go around patenting all of these prior inventions. That's not even a guess, it's a fact... many big companies have whole departments with the express goal of hacking the patent system to their advantage.
I was involved in the other side of this, back in the 1980s. Back then, IBM had decided to start using their patent portfolio as a means to esentially claim cross licenses to all of humanity's inventions, and/or big piles of cash, from, well, everyone else in the computer business. So they went, quietly, after each and every PC company.. there were actually fairly reasonable hardware patents covering much of the IBM PC.. some clever enough, some stupid, but either way, it was hard to make a PC without this. Under the IBM plan, you licensed one, two, or three+ patents.
My involvement was when they came after Commodore for the Amiga. Now, it was pretty obvious... no one involved had a remote inkling of IBM's patent portfolio. No one working on this level of new stuff bothered studying IBM, anymore than a modern zoologist needs to study dinosaurs to understand today's lions, tigers, or bears. Probably less so. But they hit us up with all kinds of stuff. I read the patents, about 35 they threw at us, and wrote up why our stuff was different -- it always was. Or, why the patents were drek. For example, IBM got a patent on "cut and paste between buffers in a text editor", dated 1984. They showed us how we violated that by running cut and paste in MicroEmacs... the very same set of keystokes did the same thing in TECO Emacs back in 1979. Probably earlier.. but that's when I used it. I'd say about 80% of the patents fell into that category... really, really obvious prior art. The others were decent enough, we just didn't violate them.
It was all for show, anyway... once a patent is granted, it's assumed valid -- you have to defeat it in court. Of course, IBM would never let this happen, as they have thousands of licensees, and, if a pile of patents were suddenly overturned, those licensees would be at IBM's door. With their legal teams. So, you go and fight these, claim you don't violate them, claim they're bogus... they can hit you up with another pile of 35. And another, and ... well, you get the idea.
Salient point: you could not possibly be an actual software engineer in 1984 without knowing full well that such things were commonplace. They simply hacked the system, pushed through anything that COULD be patented. And IBM's people, I'm certain, just lied through their teeth about their own personal knowledge of prior art. Without any actual software people in the PTO, they evaluated patents largerly based on prior patents... they didn't even bother with prior art. And there are thous
-Dave Haynie
our whole patent system in the U.S. should be completely dismantled.
Many economists and I agree. Patents have a negative impact on the economy.
Falcon
Should there be a Law?
An alternative to pharmaceutical patents
Falcon
Should there be a Law?
The hammer/executable is neither patented nor copyrighted. But making a copy of the hammer/executable would be patent infringement. So the effect would be the same (you can't copy this unless I say so) except the duration of protection would be about 20 years instead of 70+ years.
That is not reasonable. Patents have a negative impact on economics.
Falcon
Should there be a Law?
a very capitalistic thing. Not having patents or other IP protections would mean that effectively, the property belongs to The People - society as a whole. Quite a socialistic/communistic idea, wouldn't you say?
No, patents are government granted monopolies which real capitalism and free markets oppose. Of course the corporate aristocracy supports patents. But otherwise capitalism is the private ownership of the means of production and property. Patents are neither.
Falcon
Should there be a Law?
I'm not sure what kind of -ism would go with government protecting the interests of big business. Some kind of state-capitalism, I guess. Some people consider this part of fascism.
Actually Mussolini called his system, fascism, the corporate state. At first he started out as a communist/socialist but then Mussolini came to believe in a third way, between communism and capitalism. What's ironic is that FDR's New Deal was also the third way.
Falcon
Should there be a Law?
Comment removed based on user account deletion
Software patents are a great idea, but the execution is so completely flawed that I'm convinced we'd be better off without them.
Patents, whether hardware or software patents, are bad ideas. Not only do they have negative impacts on economics but they also prevent others from using their own invention. When 2 or more people invent the same thing at the same tyme, with patents only the person who got the patent can use or distribute the invention. Why should I spend thousands of hours and millions of dollars to invent something when I can lose the ability to use it because someone else beat me to patenting it?
Patents are not needed.
Falcon
Should there be a Law?
Quit thinking that "market" is the answer to every question.
I agree but a free market is the answer to the problem of patents. With no government enforced monopoly, ie no patents, there's no problem.
Falcon
Should there be a Law?
Patents are neither the means of production nor property. At most patents tell how something is made. And unlike property what government can grant it can take away too. Patents are granted, there is nothing natural about them. With property on the other hand government can only take if just compensation is given.
Now that only applies to the USA, other nations have other laws. But in the US patents are not rights, in the US government can not grant rights, it can only enforce or deny them.
Falcon
Should there be a Law?
t is to a) encourage innovation and b) encourage those innovations to be disclosed.
I don't disagree. What I said is that you yourself said software can be decompiled, so others can see how a program works, ie it is disclosed. Yet you say patents are needed for disclosure. Can programs be decompiled or is it impossible to decompile them? If they can be then they can be disclosed.
You're right, (b) is more or less taken care of. Unless of course I decide to use my software to help my business and don't ever tell anyone about it. In that case a software patent serves exactly the same purpose as a regular patent.
Now that brings up something I want to do. I've said a number of tymes right here on slashdot that I want to start a photography business, selling online as well as IRL. Unfortunately I can't afford to buy commercial closed source software, which can cost thousands of dollars. So what I've been thinking of doing is starting with open source software to build a system for the business. Now if I spend much tyme programming I also want to be able to make money by selling the software. I want to be a photographer so I want to sell not support the software. However if I use GPLed software I can not close my source, so I thought I could find and use BSD licensed software. Then, while I could not close the code I used I could close my own code and so prevent others from selling my code until I release it.
Falcon
Should there be a Law?
Allowing patents on ideas most definitely is a bad idea. What next, thought police? Where will the Precogs needed come from?
Falcon
Should there be a Law?
Comment removed based on user account deletion
I didn't know this thread was about convincing somebody that software patents can be ingenious. I am not about to argument *for* that position though, even if they are not of the "silly" kind.
RSA was patented. Where was the benefit to society. It only held off the use of RSA until it got opened again. Someone wanted to hold RSA away from the public, and patents was a perfect tool to wield that power, and delay crypto software for the masses for a while. Are we sheeple or what?
Software != Mathematics. However, when you are patenting a hardware device, you are patenting the implementation of it. Software patents however, seem to try to abolish all kinds of implementations on the *idea* itself.
Even if piece of software is ingenious, it seems to me just a way to avoid competition, and control the market. Nobody else can build on such "ingenuity", which also hinders innovation in many cases. It may bring innovation, just to circumvent the patent, but it seems a costly way to ensure innovation to me.
About cost. Patents are only costly for the small players. For bigger companies, they're just part of expenses of controlling the market. The cost of a patent is not really that much, and an individual can easily affort several patents without being really rich.
Then you have to think about the cost of being sued, for knowingly or unknowingly breaching a patent, and the overall work of ensuring you are not breaching patents, without reading them, because they are legally dangerous to read (triple damages!). Indeed, the whole patent thing seems more about control, power and squashing competition, rather than fostering innovation and contributing to sciences and arts. The real cost of patents is in the judicial system, where the smaller fish have very difficult time, especially if they are not a litiguous patent-holding company that produces nothing of they're own (then they are impervious to breaching patents, except maybe business patents).
There is nothing standing in the way of patenting mathematical formulas, if the legal system / politicians decide it. It will just further make things complicated and raise the barrier for everybody but the bigger players in the mathematical field too.
Patents will of course expire, but 20 years is a long time in the software industry. It seems silly to have to wait 20 years for every RSA. For most of these problems, you can usually come up with a competing and better implementation yourself in shorter time, however software patents seek to hinder that!
I just don't see any use for software patents, and maybe also many hardware patents are way overboard. They've just gone too far..
http://www.debunkingskeptics.com/