Patent on Tax Refund System Deemed Invalid under Section 101

H&R Block v. Jackson Hewitt (E.D. Tex. 2009)

E.D. Texas Magistrate Judge Love has recommended that H&R Block's advance-tax-refund patents be held invalid for failing to claim patentable subject matter under Bilski. The claims in question are all directed toward either a "computerized system" or "computer-implemented method." Quoting Nuijten, the court first held that Bilski controls both system and method claims – since a "court should not be 'overly concerned with pigeonholing subject matter once the court assures itself that some category has been satisfied.'"

The claimed invention involves issuing a cash-advance to income tax filers and then retaining a right to receive payment from the government.

The court held that none of the claims satisfied the particular-machine-or-transformation test. The "computerized system" claims did not identify a "particular, special purpose machine" that was more than "an insignificant, extra-solution component of the claimed invention." Likewise, the transformation of "tax return data" into a "spending vehicle" is not the type of transformation required under the test.

"At all steps in the claimed processes, the manipulated data represent legal obligations and relationships. However described, the data and resulting loan represent money. Although tangible in some forms, money is simply a representation of a legal obligation or abstract concept."

Claim 1 of Patent 7,072,862 reads as follows:

  1. A computerized system for distributing spending vehicles comprising:

    • a payment due from a governmental entity;
    • an assignable right to receive said payment from said governmental entity, said assignable right held by an individual;
    • a spending vehicle offered by a third party sponsor to said individual in exchange for at least a portion of said individual's right to receive said payment due;
    • an assignment of at least a portion of said individual's right to receive said payment to said third party sponsor in exchange for said spending vehicle;
    • wherein information associating said payment with said spending vehicle from said third party sponsor is stored in and retrieved from a computer to facilitate processing of said spending vehicle and said spending vehicle is issued to said individual in an amount for spending by said individual of said at least a portion of said payment, said governmental entity is electronically notified to transfer said at least a portion of said payment to said third party sponsor, and said at least a portion of said payment is received by said third party sponsor.
Dennis Crouch

About Dennis Crouch

Law Professor at the University of Missouri School of Law. Co-director of the Center for Intellectual Property and Entrepreneurship.

248 thoughts on “Patent on Tax Refund System Deemed Invalid under Section 101

  1. I can help you out with that one guys:

    The actual definition.

    Also, you can read the stern brief as Ned suggested if you want to know about the answer.

    Oh and broje, traditionally I think we weed out the artistic pieces with 103, there is a section in the MPEP about it. Not that I would necessarily disagree with you, but I’m pretty sure that’s how we take care of them.

  2. Ned:

    What possible interpretation of the “useful Arts” clause of the constitution could the Supreme Court construe that would make business methods non statutory?

  3. Well, Ned, I think everyone can agree that there are some business methods that should not be patentable subject matter, regardless of whether they are novel and non-obvious. The method of advertising by word of mouth comes to mind. Such a patent would impact freedom of speech, wouldn’t it?

    But your point is well taken. there are also apparatuses, even non-computerized ones, that should not be patentable subject matter in my opinion. For example, a metal sculpture comes to mind, or a mobile. Those artistic pieces can certainly be claimed in terms that strike one as being directed to patentable subject matter. But they are fine art pieces, with no practical utility. And throwing a lawn mower in the metal sculpture or mobile might tie it to a particular machine, but would still fail to make it patent eligible subject matter in my view.

    That’s why the issues are so difficult, and why the machine or transformation test is not the answer. And it’s why looking for the answer of where to draw the line with respect to business methods impacts all areas of patentable subject matter. A misstep here could be disastrous.

    But, inevitably, there are people whose minds inevitably run to the worst case scenario of a business method patent, and panic, and bend over backward to throw out the bathwater, with no mind to the consequences. I hope we can all strive not to be one of those people.

  4. Ken, State Street Bank held business methods “useful” because they were not expressly excluded by Section 101. What they did not do and what I expect the Supreme Court will do if they want to exclude business methods per se regardless if they are tied to a computer is decide the case under the “useful Arts” clause of the constitution. Stern’s amicus brief laid it out the argument well. I suggest you give it a read.

    If one does not resort to the law to decide this case, it all then devolves into policy arguments. From THAT point of view, I side squarely with Bilski. If someone invents a new and unobvious business method, why is it important to society that others be legally allowed to copy the invention? THAT is the only valid fundamental question from a policy point of view.

  5. Ken Brooks wrote: Can someone tell me why tax avoidance methods shoud be per-say unpatentable?

    Only an anti business method ideologue can tell you that. In which case do not expect an answer based on the merits of science and technology. Having worked at Bloch during the time the “Rapid Refund System” {as it was first called}, was actually invented I can tell you now that its is tied to a computer. There is no way this system can accomplish what it was designed to do without a computer. So there can be no pre, post, and extra solution activity. The patent is statutory subject matter.

  6. step back wrote:
    “”smart enough”
    “software engineers”
    buried in Dilbert dungeons
    and hungrily gulping down mushroom food”

    Wow. Way to take the high road…

    (BTW, to answer your one valid point — copyright is enouth.)

  7. “There have never been unions of software engineers, [because] these guys are smart enough”

    LOL

    Yes, which is why these guys reside in dark Dilbert cubicles, drink stale coffee all day and mutter to themselves about how they are the most valued people in our society (thanks to their incredible social manipulation skills).

    It is also why these guys are the easiest to convince that they don’t need legal protection –like by creation of property rights for the work they do (via award of patents to workers who are recognized as human individuals).

    LOL

    “smart enough”
    “software engineers”
    buried in Dilbert dungeons
    and hungrily gulping down mushroom food

  8. “Unions will not be needed.”

    Stop right there. There have never been unions of software engineers, these guys are smart enough to survive without unions.
    But! Unions will *ALWAYS* be needed. Why? Because without unions, workers have no possibility to get any representation at political arena: no votes, no participation in political project drafts, no opposition to big corporations, nothing! suddenly you can issue new laws without any respect to actual workers.

    Is that what you want? You are arguing for corporate fascism, which is already taking the control over US!

  9. MM: :”You were the one arguing that “strong” patents resulted in job creation.”
    Ahhh….Okay I see your problem. You are not an Actual Inventor, or Entrepreneur so permit me to educate you on the economic realities of work and business. Unlike a government union job, or one provided by the state in say a socialist system, in America we have to work and compete for a living. The way to create new jobs is to create new successful business that sustain long term growth.

    Now, new businesses that invent new technological processes can get in the marketplace and sustain rapid growth and provide hundreds and even thousands of jobs. This technology includes all manners of software and business methods. By having a patent on new, useful, non obvious business methods, and software, a start up can get funding for offices, advertising to attract customers, and hire employees to serve the customers. Since America desperately needs 15 million jobs we can’t get those new business started too soon. By having a strong and valid patent the company can look forward to least 5 years of growth and prosperity in the marketplace thus providing job security for their employees.

    Unions will not be needed. Nor will government created hand outs. Now just so you know, other entrepreneurs that wish to compete against these business with patents can still do so. All they have to do is use their imaginative, creative genius to design around the patent with a new and improved process that the public will enjoy even more. This creates even more jobs, lower prices and higher quality goods and services that satisfy the consumer. And this Malcolm is how we create jobs in America.

    MM: Seems to me that making it more difficult to invalidate a patent under 103 would make the patent stronger.

    No, Malcolm. Patents are strong based on the genius of the inventor, not necessarily on how a law is written or changed. I personally have no problem with the steep incline at 103. As an inventor I am confident I can create new and often pioneering technological processes that no one would have thought of before. In fact, that’s what makes being an inventor so fun and rewarding. But if down the road some other inventor proves they invented and used virtually the same technology before me, then guess what? My invention is not as pioneering and non obvious as I thought. And thats the way it should be. Let the market place decide.

    Now enforcement and eligibility are two different sides of the coin. So creating a law that prevents an inventor from having their invention even be eligible for patent examination is clearly stiflingly innovation and helps no one from an Actual Inventors point of view.
    I hope you understand Macolmn. I don’t want to change your pro socialists ideology, I just want you to know why we love our free market system in America and want to indeed keep it free and accessible for all.

  10. AI: Exactly how would congress overturning KSR create more jobs?

    MM: You tell me, AI

    Sorry,I can’t read your mind. If you want your position understood you have to be able to elaborate.

  11. Malcolm, I would “enact” eBay along the lines of the following:

    Upon filing of a complaint for infringement of a patent, the court shall, without the requirement of a bond, enter an order to show cause why an preliminary injunction shall not issue or the action not be dismissed in the interests of justice depending upon whether, upon the filing of the complaint, the patent owner has sufficient evidence of infringement to avoid summary judgment of non infringement.

    Upon the final determination that a defendant infringes any claim in a patent, the court shall enter an order enjoining further infringement.”

  12. MM Wrote: “Your ignorance sends chills down my spine. ”

    Umm…..Thats way too much information

    MM Wrote” “I thought you’d finally laid down the sockpuppet after the Supremes crushed your fantastical interpretation of Diehr during the Bilski hearing.

    Malcolm, the SCOTUS has not yet ruled on Bilski so they have not crushed anyones interpretation of anything. Personally I have no interpretation, fantastical or otherwise. I simply agree with the majority in Diehr while it appears you agree with the minority in that case.

    As for the arguments made during Bilski I did not hear any that indicated the court would support the PTO’s narrow interpretation of Diehr as represented by the MOT test.

    Nor did any of the justices, including Stevens, whom I believe is the last living Diehr dissenter still on the court, indicate they believed Diehr is no longer good law. If you indeed heard such an argument being made please reiterate it here and provide your logical reasoning for your interpretation. I would be most interested in hearing it,

  13. AI: Exactly how would congress overturning KSR create more jobs?

    You tell me, AI. You were the one arguing that “strong” patents resulted in job creation. Seems to me that making it more difficult to invalidate a patent under 103 would make the patent stronger. But please go ahead and explain yourself. Eventually you’ll hit magma.

  14. Malcolm, I think you have a point here. We should enact KSR as follows:

    “If the examiner demonstrates that there are fewer than five variations in the prior art, it should legally be presumed that it would be obvious to try each of these even if the prior art suggested that the claimed combination would not work. If there are more than five variations in the prior, there must be some suggestion in the prior art to make the claimed combination.”

  15. MM Wrote: “Yes, Congress should pass legislation to overturn KSR in order to create jobs. Makes a lot of sense.”

    Malcolm, that makes no sense to me. Exactly how would congress overturning KSR create more jobs? If you have a logical reason for your beliefs I would be interested in hearing it.

  16. “Malcolm, I don’t know whether they should pass legislation to overturn KSR, but they certainly should pass legislation to overturn eBay”

    Perhaps they should really clamp down on the obviousness standard eh? Make it near impossible to obtain a patent by upping the obviousness standard? Perhaps Ned, I like where you are going with this. Maybe you could flesh this out a bit.

    Re ebay, perhaps you’re right, perhaps they should just throw injunctions away completely. Strictly monetary damages, for a practicing entity’s lost profits only? Again, I like where you are going with this, do go on and tell us specifics.

  17. Malcolm, I don’t know whether they should pass legislation to overturn KSR, but they certainly should pass legislation to overturn eBay and re-examinations.

  18. AI: Small businesses and start ups create 85% of all new jobs. And without strong patent protection, especially for business methods and computer software these would be providers of jobs have as much chance at succeeding as a new born baby alone in a forrest of hungry, ruthless tigers.

    Yes, Congress should pass legislation to overturn KSR in order to create jobs.

    Makes a lot of sense.

    cold chills down the spine of every American

    Your ignorance sends chills down my spine. I thought you’d finally laid down the sockpuppet after the Supremes crushed your fantastical interpretation of Diehr during the Bilski hearing.

  19. In even considering making processes and methods non statutory subject matter, the Supremes are teetering this country on the precipice of economic and technological disaster!

    I have read that America needs to replace 15 million jobs before we can fully recover from the current economic crisis.

    Where does the SCOTUS think those jobs are coming from? Surely not from the pro socialists, communist sympathizers and anti patent crowd that want to ban business methods and computer software.

    Small businesses and start ups create 85% of all new jobs. And without strong patent protection, especially for business methods and computer software these would be providers of jobs have as much chance at succeeding as a new born baby alone in a forrest of hungry, ruthless tigers.

    In fact, I am reminded of the story the Jungle Book, when Badghera (sp?) explains why ShereKahn the tiger wants to kill the boy ( man cub) Mowgli. He says, “ SherKahn will get Mowgli while he is young and helpless.” Never allowing him the chance to grow up and become another hunter with a gun.

    The same can be said of those big corps that have submitted briefs for banning business methods or software patents. Yes, without the competition they can remain king of the Jungle. Only problem is these well fed tigers just consume and don’t create jobs. Instead they are eliminating jobs by the thousands.

    Ultimately Bilski, Prometheus and process inventions in general should be decided based on ideas that are supported by objective testable scientific evidence, despite the fears of the intellectually constipated that refuse to advance their arguments into the realm of 21st science.

    When even one justices is flippant about processes ( the scientific core ) of every invention and waxes philosophically about horse and buggy mechanics rather than asking questions based on the facts of quantum mechanics it should send cold chills down the spine of every American that values free enterprise, capitalism and true progress of the useful arts.

  20. 6, you’re fun to talk with. Have a good day and let’s continue this the next time a Section 101 issue arises.

  21. ” They got to the result they did by coming up with the wholly preempt holding.”

    Bingo.

    “But every patent claim wholly preempts what is claimed.”

    Does every patent claim defeat the purpose of disallowing claims to disembodied algorithms/math formulas etc? No? Oh, so that is the problem with Benson and that isn’t necessarily a problem with “every patent claim”. Amazing! Truly!

    ” Such an argument is therefore illogical and is almost a non sequitur.”

    This is lawl Ned, not logic 101. In lawl, sometimes we have to look beyond the one principle “all claims preempt what is claimed” and to other principles as well “the purpose of not allowing claims to disembodied math formulas is to prevent someone from wholly preempting the use of the formula”. When we look at that second principle, this ruling isn’t illogical at all.

    “The real question the Court should have asked but did not was whether the claim wholly preempted all ways of converting BCD to binary while the disclosure showed only one way.”

    So you think they should have considered 112 1st WD req? I think that is “a” real question that the court should have asked, and maybe/probably did ask. They just didn’t write about it because the answer is clear and irrelevant to the logic upon which they made their decision.

    “As I demonstrated before, data is real.”

    Maybe. So what if it is?

    “So what does Benson in the end stand for after Diehr? Not much if anything at all.”

    Um, it stands for the exact same thing it stood for the day it issued! JE SUS NED! /facepalm. You may not preempt every use of an algorithm with a claim per 101 reqs. Diehr did not attempt to change that what so ever, and did not change it what so ever. In fact, it reaffirmed it as the proper inquiry to make! Several times and in several different ways!

  22. 6, I know what the court in Benson held. I just disagree with it and I am explaining my reasoning.

    One can view the claim in Benson in two ways, one as a claim to a mathematical formula, or second, a claim to conversion of BCD data to binary data in a computer.

    Let’s turn to the claims themselves.

    Claim 8 reads:

    “The method of converting signals from binary coded decimal form into binary which comprises the steps of”

    “(1) storing the binary coded decimal signals in a reentrant shift register,”…

    Now what are signals and what is a reentrant shift register? Signals are data. A reentrant shift register is piece of hardware located in a computer.

    This is NOT a claim to a disembodied mathematical formula, but to a conversion process that operates on data using physical hardware. As I demonstrated before, data is real. It is physical. I has real potential energy states in terms of real computers and how much energy it takes to process data in one form vs. anther form.

    The Benson court never denied this. They even stated as did subsequent courts that useful applications of laws of nature are patentable. They acknowledged the Benson claims covered just such a practical application. They got to the result they did by coming up with the wholly preempt holding. But every patent claim wholly preempts what is claimed. Such an argument is therefore illogical and is almost a non sequitur. The real question the Court should have asked but did not was whether the claim wholly preempted all ways of converting BCD to binary while the disclosure showed only one way.

    So what does Benson in the end stand for after Diehr? Not much if anything at all. The decision was wrong.

  23. BigGuyBecause all effective methods were implemented years ago on Turing-complete looms.

    So, who’s going to let the Supreme Court in on this?

    LOL. My guess is that Clarence Thomas was independently thinking about the Turing-complete loom issue during oral arguments which was why he didn’t ask any questions.

  24. “The structure is in the software.”

    HAHAHAHAHAHAHAHAHAHAHHAHAHAHAHAHAHAH

    Wait wait,

    LOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOL

    Come on NWPA, that’s rich, but I know you’ve got some better jokes hidden somewhere!

  25. “Transformation counts if it represents something in the “real world.””

    Look for that to be the part of Bilski that gets overturned in Bilski.

  26. “Frankly, anyone with 1/2 a brain should be able to understand that CPU + software is a clever way to reduce the size of the computer. That’s all it is. ”

    If this is true, as you maintain, then this “clever way” was invented a long time ago. 102b on all of this silly CPU+software.

  27. “And, 6, I am of the opinion that Benson is simply an abomination. It should be striken from the record and Justice Douglas’s name should be forbidden to be spoken on the birthday of Von Nuemann.”

    Everyone has an opinion. But since you’re in the minority of ~1, I don’t see why we should pay heed to your opinion. Are you scheduled to become king soon?

  28. “Your knob thing and a new use for an old machine fail because a CPU + software is just a trick so as not to have to build the whole machine. You see how do you reconsile that? You can’t. If I can build a machine to perform some other type of functions, let’s see, all circuits, and then just put the actual processing in software, which I can do, does that mean that then all circuits are then not eligible for patentability?”

    The misunderstanding of the basic principles of basic patent law run deep in this one.

  29. “For those of us that aren’t over 70 years old, we better understand this ”

    “For those of us that aren’t over 70 years old, we don’t understand this ”

    fixt

  30. And, 6, I am of the opinion that Benson is simply an abomination. It should be striken from the record and Justice Douglas’s name should be forbidden to be spoken on the birthday of Von Nuemann.

  31. Frankly, anyone with 1/2 a brain should be able to understand that CPU + software is a clever way to reduce the size of the computer. That’s all it is.

  32. >>WHAT THE COURT HELD

    Bilski. Transformation counts if it represents something in the “real world.” So, we are almost there. We just need that information transformation to count always.

    The little grains of wheat are represented information. But apparently those little grains of represented information that are transformed only count when the transformation is of something that is somethign that Justice Stevens doesn’t believe that angels take care.

  33. “6, I still think the Supremes were trying to prevent a patent on a law of nature, natural phenomenon, or abstract principle. They believed the BCD to binary formula was like a law of nature standing there, side by side, with F=ma. They believed the claim to the BCD to binary in a computer was, in effect, a claim to the algorithm per se because there was no other practical use for the algorithm. So I do not fully understand why you think I am wrong to say this.”

    Looky there, you do understand what they held. I’m not sure how or why you can keep on bringing up irrelevant arguments in threads relating to that holding.

    “But compare F=ma and BCD->binary. Using a computer to calculate the one is entirely different from using the computer to calculate the other. I’ll leave you to consider the difference before we continue.”

    How about you just tell me what the difference is Ned? Because it isn’t jumping out at me. Or was all that nonsense about potential energies what the difference supposedly is? Along with the whole fallacy that since BCD is created by man then that saves the BCD>binary conversion because it isn’t a “law of nature”.

    Outrageous Ned, simply outrageous. I’d love to hear you argue that in court and watch as the judge lols or tries to corral you like a kindergartener into the proper inquiry at hand. Are all algorithms, regardless of what their underlying represented material is, disallowed under the judicial exceptions? The answer is yes, and it always will be.

  34. >>I’m pretty sure you get to the result that the >>PHOSITA considers practically all software to >>be an obvious, suggested use of a general->>purpose computer, and as such unpatentable.

    How much do you want to bet? I’m at least a PHOSITA and what you said is very wrong. In fact, it is nearly the opposite of that. Information processing methods are without question some of the most difficult and novel inventions.

    Your knob thing and a new use for an old machine fail because a CPU + software is just a trick so as not to have to build the whole machine. You see how do you reconsile that? You can’t. If I can build a machine to perform some other type of functions, let’s see, all circuits, and then just put the actual processing in software, which I can do, does that mean that then all circuits are then not eligible for patentability?

    Or a machine that builds molecules. Then software changes to build different molecules. How can that machine infringe when according to you it is nothing more than a new use for an old machine.

    The structure is in the software. The invention is in the software.

  35. “First, I note that F=ma is a law of nature, but BCD is a creation of man. ”

    First that is kind of stretching it Ned. Man does not create numbers, BCD is just numbers. If we go so far as to say that BCD is a creation of man because it is a system of expressing/representing numbers then we may as well say that F=ma is a system of expressing/representing numbers. I’m not convinced on your “bcd is a creation of man”, but even so, that is irrelevant. BCD /= the algorithm for converting BCD > binary. BCD > binary IS AN ALGORITHM and it then falls into those laws of nature/math etc.

    “Third, I note that BCD-encoded data has a lower potential energy than binary data for computer operations involving the manipulation or display of decimals, particularly decimal mathematics performed by the computer’s ALU.”

    Ned I’m afraid you just went out of your mind. There is no way to say that BCD date has “lower potential energy than binary data for computer …” That is going to be a situation specific inquiry, and one that you cannot generalize about. Even if it is true, it is irrelevant in the extreme.

    “Fifth, I note that the conversion of BCD-encoded data to binary-encoded data involves a physical transformation of data from one potential energy to another. As Night Rider has articulated a number of times, data transformation is physical it is subject to the laws of physics.”

    So what? Like NWPA, you fail to mention how that is relevant when you make no mention of this alleged transformation in the claim as in Benson. Again, irrelevant “facts” or irrelevant disputed facts are irrelevant.

    You need to tie these into your argument against WHAT THE COURT HELD somehow.

    “Presumptively, a claim to the transformation of BCD-encoded data to binary-encoded data is patentable because it transforms something physical from one state to a different state.”

    I’m not sure why you bother to “presume” this since a court has told you specifically that it is not, or at least not always, true.

    “I believe that the Benson claim was properly tied to a transformation of BCD-encoded data to binary-encoded data in a computer. ”

    Let’s assume, just for the sake of argument, that you are 100% correct and that the court believed you. You still lose under their reasoning and your failure to see why is appalling.

    “In contrast, F = ma is a fundamental law of nature. Here one has to be careful that the claim is not truly directed to the algorithm itself so that anyone calculating the formula would be an infringer. ”

    Why should we not be careful of that exact same scenario in the instance of other algorithms? This is a question which you will never come up with a good way to answer in any other way than: we should. And that is why the court held the way they did.

    I don’t know why you want to keep starting the argument afresh Ned, if you could cease such behavior it will lead to your understanding increasing. Finish the hard parts. You know that we were already down to the heart of the matter, issues that the court raised and about which you get very non-responsive:

    1 Is a claim that is directed directly to an algorithm permissible?

    2 If a claim is directed to an embodiment that effectively preempts every use of the algorithm permissible?

    3 If so, does that not negate the entire purpose of the answer to the first question? Please explain your answer whilst touching on the issue of form vs substance.

    4 Did Benson’s claim preempt every possible use of the algorithm?

    That’s a few yes/no questions with one simple “explain your answer”. Go ahead and give them a shot because that’s what your opponents in court will make you do. And it is where all your irrelevant arguments are going to fail you.

  36. “Returning to my original post, if you want to present policy reasons against patentability of computer-implemented methods, fine.”

    First, stop using the neologism “computer-implemented methods” when you don’t mean it. A computer-implemented method is something like the rubber-curing process. If it contains no industrial activity, just software and interactions with human brains and eyeballs, it’s not a “computer-implemented method”, it’s pure software.

    OK. Look at the economics: nobody has ever been documented to have been incented to create a computer program by patents, but plenty of companies have been trashed by pure-software patents. There’s piles of economic studies showing what a disaster software patents are.

    Some of the earliest such studies caused software to be excluded explicitly from the European Patent Convention, despite while patents on software continue to be issued with the addition of insignificant post-solution activity, as patent lawyers call it.

    Furthermore, software is covered very strongly by copyright. Copyright has been highly effective in the software industry.

    Yet further, patenting software presents a conflict between disclosure and restrictions on use, because a blueprint *truly* sufficient to construct the software — sufficient to prove that the “inventor” actually wrote it rather than just dreaming it up — is such that it is, in fact, the software itself.

    Yet further, patenting software presents a problem of foreclosing entire fields of development. With rare exception, software patents patent “any method of presenting this result to the user”, which is generally completely impossible to deal with except by getting rid of the bogus patent, or ignoring it.

    Yet further, when a software patent presents an actual algorithm rather than an interface, it *is* pure math and generally attempts to exclude *all uses* of the algorithm, which again forecloses entire fields of research.

    In short, there’s no value whatsoever to software patents (we have copyrights), they almost always violate the principles of the patent system, and they really are the patenting of “abstract ideas”.

  37. “Follow the analogy from there and see where you get. I’m pretty sure you get to the result that the PHOSITA considers practically all software to be an obvious, suggested use of a general-purpose computer, and as such unpatentable.”

    Got it. Because all effective methods were implemented years ago on Turing-complete looms.

    So, who’s going to let the Supreme Court in on this?

  38. “Let’s suppose that the saw has a bunch of knobs. Lets suppose these saws have been the same for 100 years; no changes. Let’s suppose that everyone has always used the knobs the same way. ”

    BEEP BEEP! Analogy fail!

    In computers, they are explicitly designed, and intended by manufacturer’s instructions, to be used with *any possible* knob settings (to maintain the analogy), and the record shows that users make radical changes in the knob settings as a *matter of course*.

    Follow the analogy from there and see where you get. I’m pretty sure you get to the result that the PHOSITA considers practically all software to be an obvious, suggested use of a general-purpose computer, and as such unpatentable.

  39. Michael Lenaghan:
    “”Undecidability” means that an algorithm won’t halt on some inputs.”

    Yes, thus effectively such algorithm is useless. Imagine court executing such test: it just runs, runs and runs, and nobody knows whether or when it is going to finish.

    “It doesn’t mean that you can’t write an algorithm.”

    Right, it doesn’t, but any such algorithm is useless if you want an answer in a definite and finite time.

  40. Night:

    To be clear, I wrote my 10:00a comment before seeing your 9:44a comment. I’m not quite sure how to reconcile what you wrote then with what you wrote earlier, but defining that boundary–the boundary between what can and cannot be patented–has been the common thread to my argument. Effective methods are a way of separating the essential from the physical. What’s more, there’s no subjectivity in the separation.

  41. Night:

    Your proposed system rests on a different foundation. That foundation assumes, for example, that math can be patented. Your system may be logically consistent but I don’t think it would ever be adopted in the U.S.

    From Breyer’s dissent in LabCorp:

    “Patent law seeks to avoid the dangers of overprotection just as surely as it seeks to avoid the diminished incentive to invent that underprotection can threaten.”

    From what you have described so far I don’t think your system provides limits to the dangers of overprotection. (Breyer explains that that’s the concern–the premise–that drives the exclusion of nature, math and abstract principles.)

  42. Well, Michael, that is a good point and the only reason to have exceptions. And that is preclude other uses exception that comes up in all the different areas of patent law and needs to be considered separate from general eligibility issues under 101.

  43. The absurdity of all this is that even in Bilski the claim itself cannot be represented without a physical representation.

    Stevens talks about abstractions as if they exist outside his head. This is really medieval thinking. For those of us that aren’t over 70 years old, we better understand this and prepare for the machine that processes information and can do it better than we can.

  44. Ned:

    “F=ma” is math that represents a law of nature. The fact that it’s a representation of a law of nature adds a complication–so let’s eliminate it. Let’s talk about “f(x) = x + 1”. That’s exactly like the BCD algorithm; it’s just simpler. So in what sense–and in what cases–can I patent that?

    And–imitating a broken record–I have no issue with patenting a wooden device that does “x + 1”, or a metal one, or an electrical one. What I take issue with is that successfully patenting one would prevent you from building the others.

  45. What you will find is that those who do not get it are the same ones that believe that their thinking occurs on the heads of pins by angels. (Or in the armpits of baboons like MM.)

  46. >>If you’re talking about speed or capacity >>you’re talking about physical properties. >>There’s no question that physical things can >>be patented.

    It is information processing. Information has to be physically represented to be transformed.

    A transformation of information takes, time, energy and space.

    Sounds like a physical process to me.

  47. There is no such thing as a Turing machine with tape of infinite length and tape advancing speed greater than that of light.

    If it helps, think of a “Turing machine” as an Intel 4004 processor from the dawn of time. The point is that every algorithm that can run on on a Core i7 can run on a 4004, and vice versa. The same will be true for any chips created 10 years from now. It’s the algorithms we’re talking about. Well, that I’m talking about.

    If you’re talking about speed or capacity you’re talking about physical properties. There’s no question that physical things can be patented.

  48. Anony:

    “Undecidability” means that an algorithm won’t halt on some inputs. It doesn’t mean that you can’t write an algorithm. Given that, my suggestion still holds: if a sequence of steps can be turned into a computer program it’s empirical proof that the original steps represent an effective method–which is to say, a sequence of steps that could be performed with a human with pencil and paper. That’s all that’s needed.

  49. You are making this all too complicated.

    The fact is that F = MA is an invention. It is an approximation of a way of representing the outside world. There is the outside world and then there is us. We represent the outside world to perform information processing tasks.

    Math is not outside of us, but a result of our brains.

    The only question is whether or not the court should allow or not a patent on something that happens to represent very well how the outside world works.

    Really, people, information processors is what we are. Computational intelligence. We represent things with physical neurons and we are physical. Sheesh. This is more about religeon and philosophy than patent law.

    And, yes, I understand theoretical computer science very well. I’m still hoping to prove that NP <> P before I die.

  50. 6, Michael and Night Writer, we are considering the difference between F=ma and BCD->binary when a computer is involved.

    First, I note that F=ma is a law of nature, but BCD is a creation of man. Presumptively, therefore, BCD is patentable subject matter if it physical.

    Second, I note that BCD, when tied to data in a computer, is physical because data is physical. A medium encoded with BCD data is physical.

    Third, I note that BCD-encoded data has a lower potential energy than binary data for computer operations involving the manipulation or display of decimals, particularly decimal mathematics performed by the computer’s ALU.

    Fourth, I note that binary encoded data has a lower potential energy than the BCD-encoded data for computer operations involving manipulation or display of binary data by the computer’s ALU. Binary data operations are particularly useful in so-called RISC machines.

    Fifth, I note that the conversion of BCD-encoded data to binary-encoded data involves a physical transformation of data from one potential energy to another. As Night Rider has articulated a number of times, data transformation is physical it is subject to the laws of physics.

    Presumptively, a claim to the transformation of BCD-encoded data to binary-encoded data is patentable because it transforms something physical from one state to a different state.

    Therefore it seems to me, that even if one can calculate the Benson algorithm on paper, if the claim is properly tied to a physical transformation of data in a computer, it is patentable subject matter.

    I believe that the Benson claim was properly tied to a transformation of BCD-encoded data to binary-encoded data in a computer.

    Consider the Turing machine. If due to its circuits or cogs or whatever, it processed BCD-encoded cards more efficiently than binary-encoded cards, or vice versa, a machine that read a card of one type and punched out a card of another type for subsequent use in the Turing machine would be a patentable machine. Not so? It is easy to see this because the cards themselves are physical. But so is data encoded on a medium.

    In contrast, F = ma is a fundamental law of nature. Here one has to be careful that the claim is not truly directed to the algorithm itself so that anyone calculating the formula would be an infringer. But if the equation were used as part of an otherwise patentable process as we had in the Diehr case, the use of the formula would not render that otherwise patentable process unpatentable.

  51. step back:
    “There is no such thing as a Turing machine with tape of infinite length and tape advancing speed greater than that of light. That too is merely an abstract concept and myth.”

    Get your facts!!! Turing machine is as real as Internet, or you and me. Anyone with Computer Science (should) have programmed at least the Turing machine emulator.
    As for unlimited tape: yes, the stock computers have limited memory, but guess what, you can connect computers! connect them continually as you go, expend the memory as your algorithm proceeds. Ever heard of cloud computing?

    Despite this surreal property, Turing machine is as good as you can get in this Universe with current physics. Even quantum computers are not more powerful, they are just faster.

    Moreover for halting problem to apply you don’t have to have infinite tape: the state space is exponentially large of any memory and it would take exponential time just to analyze simple algorithm running on finite memory.

    Do you have any sense what exponential is? Get a clue, it’s worse than astronomical! Current cryptography operates on numbers where state spaces are larger than there are particles in our Universe as we know it!

    Effectively it is as bad as not doable at all.

  52. Michael Lenaghan,

    Bilski is very interesting indeed. But here is a proof that patents in general case are in a hopeless position:
    1) courts rely on guidance in terms of method/test/procedure (let’s generalize by algorithm) to determine what is patent-eligible and what’s not, useful or not, and other interesting properties. Currently it seems that everybody agree that machine-or-transformation test is just a clue. Actually it is much worse than that, thus read-on.

    2) if you were to create a any kind of test to tell any property of software/algorithm, you would have to derive some kind of method/procedure/algorithm just to do that, like MOT test, right?

    3) now Rice’s theorem says, that the problem of designing an algorithm which decides about properties of other algorithms is undecidable, i.e. there does not exist such algorithm for any property, even for the most trivial ones.

    4) conclusion: in case of software/algorithms there does not exist a test for any property, including patentability or patent-eligibility.

    Now this looks totally hopeless, but the current trends in computer science say there is hope with software *models* (think of abstract diagrams with precise semantics). The crucial property software model is that it is not Turing-powerful, i.e. the expressiveness of the model is limited some way or another, which makes analysis by algorithms of such things feasible and thus analysis methods can be derived for that particular model formalism.

    Overall conclusion is:
    there is no hope what’so ever to find any test for (software) patents, as long as attorneys insist on complete coverage. The most general/liberal constrain could be that patent-eligible algorithm should be analyzable in the first place (there are many formalism and methods to choose from) and then we can search and apply some kind of tests designed in advance (so that everybody knows and agrees what is patentable and what’s not).

    Now this clearly has some consequences to other patents (not just software) too, but I am not an expert there.

  53. Posted by: 6 | Nov 18, 2009 at 10:08 PM

    “First of all what Flook’s majority AND dissent said about benson was exactly what I just told you. The dissent simply put it in confusing new clothing because they were having a hard time comprehending what was going on in flook’s majority decision and were having a hard time enunciating what they wanted to from benson.”

    6, I still think the Supremes were trying to prevent a patent on a law of nature, natural phenomenon, or abstract principle. They believed the BCD to binary formula was like a law of nature standing there, side by side, with F=ma. They believed the claim to the BCD to binary in a computer was, in effect, a claim to the algorithm per se because there was no other practical use for the algorithm. So I do not fully understand why you think I am wrong to say this.

    But compare F=ma and BCD->binary. Using a computer to calculate the one is entirely different from using the computer to calculate the other. I’ll leave you to consider the difference before we continue.

  54. But it seems that attorneys would prefer to see it complete (cover everything) and inconsistent (mess up some parts) rather than consistent (firm and coherent) and open (leave some things out).

    Don’t give up hope; the tide turned with Bilski. Now we’re just waiting to see how far out it goes.

    p.s.: Before Lotus there was Visicalc. Before Visicalc there was The Void. :-)

  55. HierarchyOfPontificationBuckets,
    I don’t really care if a person is pro-FOSS or not (in earlier thread it was RedHat, and there was material cited concerning FOSS, but this is just coincidence). I do care about technical side and coherence, my ultimate quest is to find out should I care about this legal side or not.
    FOSS is treating software as service, not a product, hence contradiction to most companies who try to sell product as a product. If you look at software engineering, it is inherently service: one needs to maintain and improve it constantly.
    Now spreadsheets from MacroHard is nothing new, ever heard of Lotus-123?
    There is always prior art, and before Lotus there were bunch of accountants doing it on a sheet. I’m sorry, there is no software innovation here.

    Malcolm Mooney:
    “Might I ask: what sort of information and/or signal processing is a computer theoretically incapable of doing?”

    There are fundamental limits to what todays computers can do, please lookup Turing halting problem and Rice’s theorem:
    link to en.wikipedia.org
    These are the most basic problems that are unsolvable, there are even harder, needless to say.
    Oh, and please read Church-Turing thesis that Michael Lenaghan is refering:
    link to en.wikipedia.org
    Which says that all computers (regardless of novelty and speed) are functionally equivalent (to Turing machine).

    As lawers you might want to know Gödels theorem:
    link to en.wikipedia.org
    which says that any theory is either incomplete or inconsistent — it would be nice to apply it to rules of law and require the theory to be at least technically consistent (coherent).

    But it seems that attorneys would prefer to see it complete (cover everything) and inconsistent (mess up some parts) rather than consistent (firm and coherent) and open (leave some things out).

  56. Anony Mouse,

    Please keep in mind that is easy for an entity to pose as a “FOSS developer” but in fact be a for-profit undertaking.

    Suppose we organize distributed Team to develop FOSS-Spreadsheet-program. We aim to give our code away for free. We pay people on the Team at levels typical for non-profit entities, or, we don’t even pay them because they find the work gratifying in some way. Now we solicit donations. Several large companies, let’s say the government of France (GOF) and MegaCorp contribute $500,000 each to our effort. We buy infrastructure to organize our Team, give them tools needed for software development, we send everyone our team a smokin’ workstation work $10,000 each, so they can work on FOSS-Spreadsheet-program.

    The team looks at CommercialSpreadsheet – the commercial spreadsheet program published by MacroHard, which the donors are currently spending $5,000,000 each year to license and use. The team carefully copies the user interface features (easy enough), copies the look and feel of CommercialSpreadsheet. They read the documentation and study the program to uncover most of the functionality. They figure out how all of the features work. They spend a year coding up FOSS-Spreadsheet.

    In many ways, Team is like a corporation. They do a decent job. It’s not a facsimilie reproduction of CommercialSpreadsheet, but still missing some of the sophisticated features of CommercialSpreadsheet. They publish their source code and FOSS-licenses. It’s a big hit. Government of France and MegaCorp only buy a fraction of the licenses of CommercialSpreadsheet that they previously bought (for the power users), and the majority of their spreadsheet users now use FOSS-Spreadsheet-program.

    Isn’t that a nifty arrangement? The world gets free software. GoF and MegaCorp save millions. They’ve managed to buy the ideas and creative work of MegaCorp for much less than it cost to produce them.

    Sadly, the plummeting sales of CommercialSpreadsheet take a substantial hit. Laying off 90% of the American software engineers isn’t enough for MacroHard, and the company folds.

    Sure, Team keeps working on FOSS-spreadsheet-program. But without the real thing to copy from, innovation slows considerably. FOSS-spreadsheet-program works, but the public no longer gets new and better CommercialSpreadsheet every year. Over time, the state of the art is not what it would have been.

    Realistic? Yes. I’ve worked for a client who faced direct competition from an “open source” knockoff. The closer the knockoff got to the real thing, the less money they made. They started to file more patents and work harder to find out what customers wanted. They’re still in business and doing well enough (even won some major awards about 5 years ago). The opensource group, though, knows about the patents. The company lets them know about patents that might be infringed. The opensource group tries to avoid them. The opensource product is still pretty nice. Students and academia use it. Professionals buy the real thing for big bucks and produce entertaining things with it. It’s a stable and fair arrangement. Plus, the company knows that they have to build something better than the ordinary and routine. The money they make largely corresponds to their advances and creative contributions, rather than from having coded up ideas and algorithms available in textbooks and other unpatened public sources.

    I don’t mean to say that there shouldn’t be opensource imitators or competition. I only mean to point out that there are real consequences of letting them “borrow” freely.

    You can really continue this hypo full circle. Team members grow up a little. They get married, they have kids and other things that require money. They decide that FOSS-Spreadsheet is so good they really should just sell the next version. They turn into a for-profit company.

    Now, TeamB comes along and decides that people shouldn’t have to pay for FOSS-spreadsheet-program. They start to imitate it …. and so on and so on and so on.

  57. We could make up some communities within areas of expertise, perhaps establish some certification process and just vote: electronic voting must be cheap nowadays, quick and efficient and we get to listen to interested majority… wait, is it already patented?

    Diebold software is kept as a trade secret, IIRC. Have fun digesting that, teabaggers.

  58. Red Monkey:
    “This should come as no surprise. The whole point of law is to make sure everybody is treated fairly (because people who feel unfairly treated can do nasty things). If the law relies on technically unsound arguments but nevertheless manages to keep enough people happy most of the time, it’s working just fine. A fool’s paradise is still paradise.”

    It’s not that surprising, but why waste so much time on this pseudo-intellectual discussion (and admitting it) when we can just vote? We could make up some communities within areas of expertise, perhaps establish some certification process and just vote: electronic voting must be cheap nowadays, quick and efficient and we get to listen to interested majority… wait, is it already patented?

  59. As an aside, within the past biweek I’ve done 101s in four cases that fall into the “is tied to a particular machine but does not impose a meaningful limit or does not involve more than insignificant extra-solution activity”. Each case had claims fall into the 101s that did not contain the words “method” or “process” in their preambles.

    Hammertime for hogwash.

  60. Moreover you have cited no evidence against my claim.

    I am prepared to produce expert witnesses testifying to prior art. The youngest, no doubt, would be happy to demonstrate a practiced hand.

  61. “Lack of novelty. Rejected”

    Rebuttal to your failure to show all elements. I don’t recall my hand having ever turned around and hit you upside the head. Moreover you have cited no evidence against my claim.

  62. Malcolm, I learned that I chose the right word. Does that qualify? :-)

    snark·y

    1. Rudely sarcastic or disrespectful; snide.
    2. Irritable or short-tempered; irascible.

  63. Malcolm:

    It sort of sucked having to wade through your snarkiness. Oh well; life is a learning process.*

    * Pat. Pending

  64. Anonymouse

    “Attorneys don’t care about the technical details/truths or how things are implemented. All they care about is who gets compensated for what and whether it is “fair”.

    This should come as no surprise. The whole point of law is to make sure everybody is treated fairly (because people who feel unfairly treated can do nasty things). If the law relies on technically unsound arguments but nevertheless manages to keep enough people happy most of the time, it’s working just fine. A fool’s paradise is still paradise.

  65. Michael L your summary is correct

    Thanks. It sort of sucked having to wade through your windbaggery to figure that out, though.

    overlooks the point that the tool has uses beyond 101.

    I “overlooked” that “point” because it’s irrelevant to my question. I also “overlooked” mentioning the weather outside my window, for the same reason.

    It’s a little surprising that naming the process used in a process claim could be considered “jumping the shark.”

    Anyone who brings up “Shewhart cycles” in the context of a discussion about medical diagnostic patent claiming is begging for a smackdown. Get over yourself, dweeb.

  66. Hierarchy:

    You’ve missed a fundamental point. The basic work that a Core 2 Duo chip does can be done by any device capable of computation. “Capable of computation” has a precise mathematical definition–and a very low threshold. Anything you can compute on one you can compute on any other. And they don’t have to work in the same way; they don’t have to emulate each other’s hardware, for example.

    Why does it matter? It matters to Transmeta if you’re thinking about competing against Intel by producing a chip that can run anything a Core 2 Duo can. It matters to Intel if Transmeta can sell a Core 2 Duo-compatible chip that erodes their market (and their profits). So if there’s a patent on the Core 2 Duo does it block Transmeta or not? The issue has nothing–nothing–to do with how a Core 2 Duo is implemented. That’s completely irrelevant.

    Indeed, you say that a computer is just a device, without any math in it. I understand what you mean, but you miss the perspective that the device was created in order to process math–i.e., information. That matters. That’s why I can build one chip that does what another one can do even if they’re built on completely different principles.

    Software patents represent exactly the same issue–without the benefit of ever having a physical manifestation.

  67. “How can the overall process be “new” if it has an existing name? ”

    A process of smacking Michael upside his head comprising:

    turning 6’s hand around
    striking the backside of Michael’s head.

    (add brass knuckles to get around bilski lol)

    Assuming you’ve been smacked upside your head at least once in your life there is an example.

  68. Malcolm:

    You meant Prometheus, not Mayo. Otherwise your summary is correct but overlooks the point that the tool has uses beyond 101. It isn’t just a 101 filter.

    It’s a little surprising that naming the process used in a process claim could be considered “jumping the shark.” How can the overall process be “new” if it has an existing name? I think the fact that it does raises the threshold for any process claim which is simply a feedback process.

  69. Michael: “What I’m trying to do is explain the boundaries in a consistent way. For example, you can’t patent E=mc^2. But what if that math was hidden inside (i.e., implemented as) a piece of hardware? How would you identify it as math and not hardware?”

    Michael, why does it matter? Would we look to a method of refining gasoline, let’s say improved cracking, the method includes some numbers/ratios/etc. We pull out the math, and then somehow say patentability of the method cannot be based on the math? Or we say others can use the math in any way they want?

    Here’s another way to see it. The device you pose does not have any math per se. There are no theories, there are no numbers, there are no algorithms. There is a machine. It does things. It does them a certain way. We can describe that way with math (as in a claim), but the math is not performing the actions of the machine. The machine is just a complex set of moving/changing parts. Just like the Newtonian formula for gravity does not pull a feather to the ground. The formula is a way to describe what we can observe and measure about matter.

    As has been said here many times, all patent claims have abstract concepts within, whether labeled as math, ideas, laws of nature. But the claim, if valid, must be limited to a real world physical thing/process (that meets all the requirements for patentability). A valid claim does not cover only pure math. It must cover real things acting in real ways that can sometimes be measured and described with math, but a real world machine/process nonetheless.

    “Here’s a concrete example. Transmeta was working on a chip that could emulate other chips in microcode. Would a patent on a Core 2 Duo prevent Transmeta from emulating it? (Again, we’re talking patents here, not copyright. And I’m assuming that Transmeta hasn’t stepped on other Intel patents in the process.) By my test, the answer is “no”. My test–which is empirical–lets you identify the “math” buried within the Core 2 Duo.”

    The question is unanswerable. One does not obtain a patent on a label like “Core 2 Duo”.

    Whether an emulation infringes a patent would depend on the claims of the patent and what the emulation does. An emulation-CPU would presumably operate like the CPU being emulated. If there were claims on how the emulated CPU functioned, e.g., certain ways that information is shared on a bus or certain ways of using the CPU’s cache, and if the emulation-CPU did this, then that could infringe. If there were a patent on a hardware feature of the emulated CPU and the emulation-CPU did not have that feature, then it would not infringe. It’s really impossible to say.

    I think the point you’re getting at is whether it matters if a CPU is emulated or real. It would depend on the claims. The mere fact of emulation shouldn’t alone excuse what might otherwise be infringement.

    Anyway, the emulation aspect of the hypo is unnecessary. Why not just say the AMD-X chip? Let’s say the Core 2 Duo has several instructions for virtualization and the instructions cause certain processes to occur in the CPU (let’s say some movements of memory or changes in protection). Now lets say the AMD-X is a different device with different architecture. Nonetheless, AMD looks at those instructions and says, “good idea, let’s do the same thing”. The AMD-X is modified to have instructions that when invoked do the same type of memory manipulations. You might say the hardware is different (arguable), but what does it matter?

    Same with the Transmeta. If they manipulate memory in the same way, what matters if it is an emulation?

  70. Michael, you really need to shrink your windbag by an order of magnitude. I’m not interested in what the court said in the Mayo case and you pretty much jumped the shark with your reference to a “Shewhart cycle”.

    I think the cliff notes version of your novel is “Yes, the Mayo claim passes my “effective test” for 101 eligibility because it recites a step that can’t be carried out with pen and paper (the determining step). But there might be a separate utility issue under 101 with this claim.”

    Do I have that right?

  71. Michael Lenaghan,

    I am in the same boat as you, except IANAL.
    I had my share of discussions on this blog, interesting people, really.
    But here is what I learned:
    Attorneys don’t care about the technical details/truths or how things are implemented. All they care about is who gets compensated for what and whether it is “fair”.
    In case of patents, I tried my own (technical) reasons why software patents are incoherent with this and that, found out a few tricks how attorneys make claims to be compatible with previous cases, and there’s always at least one response: where would we be if not the patent system? would this company have had created that Raven if not for patent system promised compensation?

    So I have to agree: creating (novel) software takes time and patents is one way of getting compensation.

    Law is dominated by attorneys and it is quite easy to make them believe that somebody has spent awful amount of brilliant thinking into production of this and that, and there was no other way to achieve but through “arts”, thus their model deserves a merit to be compensated with a patent.

    And the definition of “arts”, “useful” and “fair” in this soup is historically shifting with majority.

  72. Malcolm:

    I believe that Prometheus would fail a similar analysis, although the analysis might add an element which is present in LabCorp but not quite as visible. … I think the additional element gets added to the analysis because the claimed process is more “complex” than that described in LabCorp.

    I think I understand those issues a bit better now. In both cases the overarching goal of the processes seems to be “feedback”–basically, a Shewhart cycle. Prometheus seems, at first glance, to be more complex simply because LabCorp. doesn’t cover the full cycle. But in both cases the steps of the process are fully constrained; they have to be applied in a specific order.

    I think the “obviousness” arises out of those two things: feedback is common in any iterative process and the feedback steps have to be performed in a particular order.

    That in turn explains why the process becomes pre-emptive if the claimed steps consist of effective methods and nothing more. It implies that you’re trying to capture those steps “in the abstract”. Since the sequence of steps is both known and constrained, capturing claimed steps in the abstract leaves you with no other path to the goal.

    Now, Prometheus also specifies thresholds. I’m not sure I fully understand the issues involved, but it seems to me that you can’t capture the fact that water freezes at 0 C. You can make use of that, but you can’t own that. On that basis I think Prometheus could make use of the thresholds but not capture them as part of the patent.

  73. Good luck playing World of Warcraft on that Turing-complete loom, guys.

    That kind of comment has appeared a few times on this thread. It confuses two issues.

    First, what we’re talking about is how to detect the spring in the mousetrap–i.e., the math that can’t be patented. For example, the force that makes an apple fall on your head is the same force that pulls a star into a black hole. The two examples aren’t equivalent *things*–but the underlying law of nature is.

    Second, no one has made the claim that a specific piece of hardware can’t be patented. So, yes, of course you can patent a Core 2 Duo.

    What I’m trying to do is explain the boundaries in a consistent way. For example, you can’t patent E=mc^2. But what if that math was hidden inside (i.e., implemented as) a piece of hardware? How would you identify it as math and not hardware?

    Here’s a concrete example. Transmeta was working on a chip that could emulate other chips in microcode. Would a patent on a Core 2 Duo prevent Transmeta from emulating it? (Again, we’re talking patents here, not copyright. And I’m assuming that Transmeta hasn’t stepped on other Intel patents in the process.) By my test, the answer is “no”. My test–which is empirical–lets you identify the “math” buried within the Core 2 Duo.

    Finally: Anyone with any imagination should be curious to see what World of Warcraft would look like running on a loom. Maybe it would come out looking like a Greek tapestry. :-)

  74. I wish I had time right now to go into information processing patents. Michael there are many good academic papers that were written about these topics at about the time of Benson.

    Ong wrote about how asymtopic behavior of methods should be considered for patentability.

    Even some of the great minds in computer science at the time wrote about these topics.

    Why for example that information processing patents did not arise naturally out of the new computers. One answer is that the software wasn’t where the money was and the big companies like IBM didn’t want patents for information processing methods.

  75. Malcolm:

    Your “test” is not much of a “test” if you can’t apply it in a credible fashion to a simple, genuine (and very typical) claim.

    Slow down, cowboy. I had to read up last night in order to answer your question more precisely. As I said, I’m new to this game.

    The claim in Prometheus as given constitutes more than simply an “effective method”. That’s because it contains steps that interact with the physical world. (An example of claims that wouldn’t–and that would be filtered as purely effective methods–are Benson or E=mc^2. Note that the effective method test covers both rather than labeling them as different things. Effective methods are an empirical way of defining “math” in all its guises. They’re “empirical” because we can write a computer program that would completely capture the claim.)

    The fact that Prometheus isn’t filtered as an effective method doesn’t mean that it qualifies on all 101 counts; the effective method test qualifies it as a “process”, for example, but doesn’t address “new” or “useful”. While you could try to expand the meaning of the test to cover those issues I think it makes more sense to define them separately since they apply to all four categories. In other words, as a statement of logic 101 might be expressed as:

    new AND useful AND (process OR machine OR manufacture OR composition)

    “Effective methods” are meant to only filter (or define) “process” in such an expression leaving “new” and “useful” to other tests and definitions.

    I mentioned that Prometheus isn’t filtered because it “interacts with the physical world.” In my first attempt at an answer I tried to add qualifiers about the nature of that interaction, but as I understand it those qualifiers should not be treated as 101 issues–and that’s what you were asking about. (From Prometheus: “…it is improper to consider whether a claimed element or step in a process is novel or nonobvious, since such considerations are separate requirements set forth in 35 U.S.C. ss 102 and 103, respectively.”)

    However, effective methods are not simply a 101 filter; they support subsequent analysis in 102, 103 and 112.

    For example, claim 13 in LabCorp clearly states that the “determine” step can use any test–even an unpatented one. The rest of the claim can be described as an effective method. The combination of unpatented test plus effective method would not, in my opinion, be eligible, because it would be (and I may be using the wrong term here) pre-emptive. In other words, it would be trying to capture the essential “math” of the process.

    I believe that Prometheus would fail a similar analysis, although the analysis might add an element which is present in LabCorp but not quite as visible. In my opinion Prometheus is unpatented devices/tests plus an effective method–combined in an obvious way. (See link to en.wikipedia.org for details!)

    I think the additional element gets added to the analysis because the claimed process is more “complex” than that described in LabCorp. Having said that, it seems that it should be possible (i.e., eligible) to combine devices with an effective method in a non-obvious way. In such situations effective methods could still help to define what is and is not protected by the claim–again, the essential “math” of the process.

    As a side note I have to say that, personally, I find the transformation logic in Prometheus tortured. That may arise out of an attempt to define “useful”, something I haven’t thought about. But they seem to use it in order to define “process” and that… Well, it certainly doesn’t work for me.

  76. >>Here’s a start for you: how about you >>define “abstract”? Somehow it doesn’t include >>software, right, or other instructions for >>processing information? So what does it >>include? What does it exclude? What’s on the >>border? Learn to contribute.

    MM, you are such a coward. I have challenged you on the meaning of this word many times. You either run or start spitting out your meanness.

    Please define the word for us, MM. Particularly in the context of your arguemnt that software is abstract. Please read the dissent of Newman in Bilski and be enlightened.

  77. Michael (or POIR),

    “Let’s say that I discover some sequence of steps you can do yourself at home that comes to the exact same determination. I write about it in a book. Should that book–or the steps you perform at home–be covered by patent? What if it took me a long time to come up with the steps? What if it cost me a lot of money? Do either of those things have any bearing on the question–or the answer?”

    The book? No, that’s just copyright. The process? Yes, unquestionably. (I am ignoring for the moment whether public policy in medical care should override the patentee’s right to enforce)

  78. > The law is not interested in computer science theory.

    It should be. When you guys try to write fiction into law, like making pi equal to 3.2, it drives us nuts. If you don’t remember, incidentally, the law that guy was seeking was effectively a patent on his circle-squaring formula. At least they had the good sense to write into patent law that things like perpetual motion machines were to be thrown out without looking at them.

    I mean, look what happens when you try to draw lines in the air. Sadly, Bilski has proven ineffective because the courts can find a “transformation” anywhere they want to. I wish they would at least require that the transformation be novel (i.e. part of the invention), because right now, I’ve seen all kinds of silliness over whether a transformation matters or not. You can find *some* kind of transformation in any activity, even if it’s merely the chemical and electrical changes in your own brain as you try to cope with the absurdity foisted upon you.

  79. My example is probably not as good as some above, but I was thinking of the computer itself. The electronics are just as old and obvious as software, with an even smaller universe of arrangements of mere transistors. And, frankly, in the average computer, there just aren’t that many different types.

  80. I wrote to Michael L: That’s from your 11:31 comment. Assume “determining” is defined in the spec as “inserting a needle, taking blood from a subject, and exposing blood to an agent which allows measurement of level of metabolite”. Is method patentable under 101 or not, according to your test?

    And Michael L responded: Here’s the simple answer: Not unless a physical device used in one of those steps is the thing being claimed.

    That’s not a simple answer. That’s a nonsensical answer. We’re talking about a method claim.

    Here’s the nuanced answer: Patent law seems to contemplate that the whole can be greater than the sum of its parts. For example, I might successfully patent a mouse trap that

    Again, we’re talking about a method claim, Michael. I’m asking you to analyze the Prometheus claim (which was successfully patented, applying the Bilski test) under your “effective method” test, with the assumption stated above. Patentable or not, and why or why not?

    Your “test” is not much of a “test” if you can’t apply it in a credible fashion to a simple, genuine (and very typical) claim.

  81. “But now that you “programmed” it to have 50G of unerasable stuff stored in its one and only hard drive, it is no longer capable of receiving and storing essentially 100G of new info.”

    What kind of hard drive is that? If you bought one that was unerasable then you need to go get your money back 😐 You got gyped.

  82. APOTU said:

    “Even though I get step back’s point, … and the computer is still good for doing everything it was good for doing before it was programmed.”

    Nonsense.

    Say your new computer had a 100G hard drive.
    Before you programmed it, your computer was capable of receiving and storing essentially 100G of new info.

    But now that you “programmed” it to have 50G of unerasable stuff stored in its one and only hard drive, it is no longer capable of receiving and storing essentially 100G of new info.

  83. “-especially the dissent.”

    First of all what Flook’s majority AND dissent said about benson was exactly what I just told you. The dissent simply put it in confusing new clothing because they were having a hard time comprehending what was going on in flook’s majority decision and were having a hard time enunciating what they wanted to from benson.

    Second, you rely far too much on the people that don’t understand what is going on and did not make what is going on be the things that are going on. You need to pay attention to what is going on and the people that made it that way. That is, the majority.

    Third, we’re talking about Benson itself, not some commentary by a subsequent court on that opinion.

  84. “How is this case going to affect examiners?”

    It’s going to change what it takes for a claim to pass muster under 101, and then you’re going to have to do a 2nd non-final on every case you’ve already worked with a method claim for the next 6 months. Mmmm, delicious voluntary unpaid overtime!

  85. “It probably wouldn’t be hard to prove that the loom is Turing-complete, in which case it can do anything a computer can do, given enough thread and the right punch cards.”

    Good luck playing World of Warcraft on that Turing-complete loom, guys.

  86. I have hope, the Court rules as I believe they will, to pull software off the table with process patents because from the arguments that made it clear in my opinion, such patents (and others), have no limits, as it is today.The Court, I know seeing the patent system and its impact, knows well the danger to open the gates wide.

  87. “After all, any effective method could be completely reduced to a physical machine.”

    Way too long; didn’t read all of it. This statement needs clarification since methods and machines are in different statutory categories. A claim is one or the other. A method claim and a machine claim are considered separately. There’s no useful legal analysis benefit to be gained in arguing that a particular method claim happens to be arguably equivalent to a particular machine claim.

  88. Night Writer: The computer hardware operates according to a rather complex equation. The software is a (really big) number that gives the result we want. No, I’m not kidding. We can, in fact, do the math. It’s just a matter of defining the right functions and giving the correct input.

    One of my classes, long ago, even had us write out the actual number corresponding to our program in hexadecimal. And we had to give the exact state a machine would be in after running a given program. In short, we actually calculated the result of those equations given certain data.

    And that other guy is not kidding about the loom. You could do computations with it, just like you can on a computer. It probably wouldn’t be hard to prove that the loom is Turing-complete, in which case it can do anything a computer can do, given enough thread and the right punch cards.

  89. Hierarchy:

    I don’t care what box you put things in. I don’t care if you label something as a math or information transformation.

    You don’t have to care about the boxes, of course–but if you don’t define the boxes well (or at all) I think you end up with an ad hoc system that doesn’t make much sense. A system, say, that contains both State Street and Bilski. A system where Bilski oral arguments seem surprising.

    If someone invents a new way to identify those likely to have a heart attack using a physical device, and the outcome is possible via processing information represented as physical bits, I dont’ care what category you put it in, it’s of a technical nature and seems like the kind of thing that merits a patent (assuming new unobvious etc.) It’s the kind of thing that we should give the inventor the option of preventing it from taken by others.

    Let’s say that I discover some sequence of steps you can do yourself at home that comes to the exact same determination. I write about it in a book. Should that book–or the steps you perform at home–be covered by patent? What if it took me a long time to come up with the steps? What if it cost me a lot of money? Do either of those things have any bearing on the question–or the answer?

    The problem I have with your test(s) is that it would exclude things that I think should be eligible. It’s the bottom line that matters.

    Yes, I think that explains how you get State Street. But I think you have to make an effort to understand or explain how you then get Bilski. Or the oral arguments for Bilski.

    If I understand your suggestion, to me it makes as much sense as saying we can’t allow patents on circuits because they can be described abstractly with a set of equations. It’s like saying we can’t allow patents on chemical processes because they can be described with chemistry notation and laws of thermodynamics. It’s like saying no mechanical patents because the device can be described with mathemtical measurements.

    No. All of those things are clearly eligible. What I said is that an effective method becomes an empirical test for pointing out what isn’t covered by the patent. It’s like being able to identify the spring in a new mouse trap–even if it doesn’t look much like a spring at first blush.

    You might be able to model a chemical or mechanical process on a computer. In that case, what you model is a clear illustration of the spring in the mousetrap. (By “spring” I mean “the law of nature that can’t be patented in the mousetrap.”) But if you want to catch a mouse you need the actual spring, not a model of it; if you want the effect of the chemical or mechanical process, you need the actual process, not a model of it. The spring in the mouse trap has various properties–composition, construction, etc.–and while the model of those properties can’t be patented a physical embodiment of it certainly can be. As can the actual chemical and mechanical processes.

    I understand that what I’m describing isn’t where you’d like to draw the line. But I think that it might be a reasonably accurate description of where the line is being drawn. Or where they’re *trying* to draw it.

  90. I don’t know if anyone has mentioned this, but H&R Block filed a continuation after the patent issued. The continuation just received a notice of allowance yesterday (11/17/09) for a set of system claims which passed muster under Bilski as interpreted by the USPTO.

    I anticipate that the issue fee will be paid shortly.

  91. Noise:

    I was really being more facetious about Google Apps than anything else, mainly because it can be difficult to construct an airtight hypothetical.

    Even though I get step back’s point, the legal concept still bothers me that a general-purpose computer magically transforms into a special-purpose computer by programming it. After all, the hardware itself doesn’t really change physically by programming it (yeah, yeah, stored electric charge and aligned magnetic domains and all that), and the computer is still good for doing everything it was good for doing before it was programmed.

    On the other hand, ours not to question why….

  92. If Sotomayor writes the majority opinion in Bilski, I think she will re-iterate a doctrine requiring unfettered flow of “ideas” which doctrine is already well established in copyright law under the name of the idea/expression dichotomy doctrine.

    Pure math is an area where idea and expression are inextricably merged and thus, by trying to preempt all use of the math by way of patent claims(or copyright claims) you are in effect trying to stop the unfettered flow of ideas (or of “information” as Sotomayor phrased it during the Bilski orals).

    Admirers of existentialism out there of course understand that every item is an expression of itself and of its idea. (Consider for example the French painting: This is not a pipe.) The question becomes one of when idea and expression inextricably merge with one another.

    But hey, this is just another reading of the Bilski oral tea leaves and I could be completely wrong.

  93. Ned and the other guy that responded thanks for your interest but I’m afraid that the concept in the claims at issue is too complex for me to make a good explanation without simply giving away the claim. It might could be done, but I’m too lazy to do it.

    Ned, I’m surprised and let down that you have let your views on benson fall back into nonsense that had nothing to do with what they held or their reasoning in the case. I had thought you had made progress in understanding that they weren’t taking issue with it being on a computer, on a plane, a train or with green eggs and ham. All they took issue with was the claim precluding every possible use of the SPECIFIC algorithm which was being used in the claim (and which benson may have invented). That’s it, the end. Nothing else was on their mind.

  94. NWPA: Please define that word for us in patent law MM.

    Sorry, benchwarmer. When you start stepping up your game and actually start thinking twice before posting your conclusory bullcrap here, I’ll stop shredding it and start answering your questions.

    Here’s a start for you: how about you define “abstract”? Somehow it doesn’t include software, right, or other instructions for processing information? So what does it include? What does it exclude? What’s on the border? Learn to contribute.

  95. HighbucketIf someone invents a new way to identify those likely to have a heart attack using a physical device, and the outcome is possible via processing information represented as physical bits, I dont’ care what category you put it in, it’s of a technical nature and seems like the kind of thing that merits a patent (assuming new unobvious etc.)

    Do eyeglasses count as a “physical device”?

  96. Michale,

    I don’t care what box you put things in. I don’t care if you label something as a math or information transformation. If someone invents a new way to identify those likely to have a heart attack using a physical device, and the outcome is possible via processing information represented as physical bits, I dont’ care what category you put it in, it’s of a technical nature and seems like the kind of thing that merits a patent (assuming new unobvious etc.) It’s the kind of thing that we should give the inventor the option of preventing it from taken by others.

    The problem I have with your test(s) is that it would exclude things that I think should be eligible. It’s the bottom line that matters.

    One can write a claim that fairly covers an invention of machine computation such that others in the field can read the claim (plus the spec) and have a fair idea of whether they’re infringing. This claim can be limited to “real world” use and thus not prevent others from studying teaching pondering improving printing and otherwise working with the math that partly describes the patented physical device/process. At the same time, this claim can maintain for the creator the financial benefit that is supposed to go with creating and sharing with the public. You create a product and sell it using my claimed method or device, I can stop you. You give a lecture describing my method or device, no problem. You improve my invention, great, that’s desirable and you can patent the real world embodiments.

    If I understand your suggestion, to me it makes as much sense as saying we can’t allow patents on circuits because they can be described abstractly with a set of equations. It’s like saying we can’t allow patents on chemical processes because they can be described with chemistry notation and laws of thermodynamics. It’s like saying no mechanical patents because the device can be described with mathemtical measurements.

    Patent claims always involve some level of abstraction. A “board” is an abstraction. “Hydrogen” is an abstraction (is it an ion, is it liquid or gaseous, what temperature, etc.) The presence of abstraction in a computer implemented process is just the nature of patents. The issue for all patent claims is when is the claim “too abstract”.

    Anyway, thanks for your comments. It’s interesting to see how people who aren’t examiners or attorneys view the issue, and I think you approach it like a computer scientist would. Although I think the “effective method” test is a bad idea, many bad ideas lead to good ideas…

  97. MM wrote: >>abstractions

    Please define that word for us in patent law MM. And no turning your tail and running.

    Michael you have good points that I will answer when I have the time.

  98. Night Writer:

    >>you believe that it’s *right* to exclude math from patents

    I don’t believe that.

    Sorry, I’m afraid I can’t parse that unambiguously; I can’t tell whether you’re saying that math *should* or *should not* be excluded from patent eligibility. (If you’re saying that it shouldn’t be then our premise is different and it’s not surprising that our conclusions are too.)

    And, I believe the term “mathematical algorithm” has no clear definition. Many journal articles were written at the time of the Benson decision heavily criticizing Benson and the use of this term.

    Three formalisms were developed: Godel’s recursive functions, Church’s lambda calculus, and Turing’s Turing machines. They initially had different goals, but they were all shown to be equivalent (in the technical sense of the word). The Church-Turing Thesis states that any algorithm can be defined in one of those formalisms. (Since they’re all equivalent it doesn’t matter which one you pick.) There is no known math formalism that can’t be reduced to one of those systems. Likewise, there is no known computing system that can’t be reduced to one of those systems. So there’s no known sense in which “mathematical algorithm” can’t be precisely defined.

    Look, if it were the case that some particular kind of math couldn’t be reduced to one of those formalisms then you’d be saying (at the same time) that that kind of math couldn’t be used on a computer. There simply is no such known math.

    The Church-Turing Thesis is a thesis, not a law–but, again, there are *no* known counter-examples. And it’s believed no such counter-examples do exist.

    If you think of information being represented then transformed by the methods, then all of this makes sense. It become very simple to spot the ridiculous arguments. Indeed, the machine or transformation test has it nearly right in that a transformation counts but only if you happen to be representing something that is physical. What? That is an absurd attempt to exclude business method patents. Just nonsense.

    Again, I don’t know whether you meant to include or exclude math. If you meant to include math as patent eligible, fine, but that’s a different premise leading to a different system. You can’t, however, exclude it and maintain your everything-is-physical approach at the same time because you end up with logical inconsistencies. I mean, in short form you’ve said: math is excluded; everything physical is included; every transformation is physical; math is a transformation; therefore math is physical; therefore math is included. Contradiction. There is no way to fix that without excluding information. Because there’s no way to distinguish “math transformations” from “information transformations.”

  99. NWPA Ever consider that in Deener that a person could practice the method of tranforming the wheat without any tools?

    Yes, we’ve considered that hundreds of times here in these threads.

    Why is that doesn’t bother us, but the thought of a computer doing something we do seems to drive Justice Stevens into a mental frenzy

    It’s not the thought of a computer doing something we do that drives Justice Stevens “into a mental frenzy.” It’s the thought of claiming “problem-solving” abstractions, on a computer or not, that drives the “frenzy.” And it’s not really a “frenzy.” It’s just irritation and annoyance, particularly when the thoughts are banal crap like “analyzing” a person’s prior purchases and recommending a particular color of lollipop for that person to suck on.

  100. I wrote to ML: That’s from your 11:31 comment. Assume “determining” is defined in the spec as “inserting a needle, taking blood from a subject, and exposing blood to an agent which allows measurement of level of metabolite”. Is method patentable under 101 or not, according to your test?

    And ML responded: Here’s the simple answer: Not unless a physical device used in one of those steps is the thing being claimed.

    That’s not a simple answer. That’s a nonsensical answer. We’re talking about a method claim.

    Here’s the nuanced answer: Patent law seems to contemplate that the whole can be greater than the sum of its parts. For example, I might successfully patent a mouse trap that

    Again, we’re talking about a method claim, Michael. I’m asking you to analyze the Prometheus claim (which was successfully patented, applying the Bilski test) under your “effective method” test, with the assumption stated above. Patentable or not, and why or why not?

  101. I think there is a baby in the bathwater with respect to software patents

    As I said before, your favorite software app is like a baby with a billion identical twins that will be replaced the minute it swirls down the drain and you’ll never even miss it. The only people who’ll notice will be the gamblers who “invest” in the software patent lottery. Many here will weep for them.

    I won’t.

  102. Just throw in there any of the SCOTUS exceptions.

    Ever consider that in Deener that a person could practice the method of tranforming the wheat without any tools? Why is that doesn’t bother us, but the thought of a computer doing something we do seems to drive Justice Stevens into a mental frenzy to try and come up with a way to mischaracterize information processing.

  103. >>you believe that it’s *right* to exclude math >>from patents,

    I don’t believe that. And, I believe the term “mathematical algorithm” has no clear definition. Many journal articles were written at the time of the Benson decision heavily criticizing Benson and the use of this term.

    Further, the gate to eligibility should be open to all methods that transform informaiton. The SCOTUS can then apply their exceptions.

    If you think of information being represented then transformed by the methods, then all of this makes sense. It become very simple to spot the ridiculous arguments. Indeed, the machine or transformation test has it nearly right in that a transformation counts but only if you happen to be representing something that is physical. What? That is an absurd attempt to exclude business method patents. Just nonsense.

    You have to understand the word abstract as used in patent law and engineering to figure out the rest of the puzzle. See Newman’s dissent in Bilski.

    Understand the word abstract and understand that transforming information is a transformation and suddenly, ding, ding. All done.

  104. well.

    after reading all this. And being a patent attorney for more than 10 years.

    I think there is a baby in the bathwater with respect to software patents and have yet to read a good way of distinguishing the two.

  105. If the honorable Supreme Court Justices can conflate “patentable” with “patent eligible subject matter”, why not everyone else on this esteemed Trainwreck?

    Someday, Troll’s own blog may be more popular than this one and then the bitterness will end.

    But I doubt it.

  106. Ned:

    [T]he risk of heart attack case is Arrythmia Res. v. Corazonix…

    Thanks, I’ll take a look later. And I’ll tell you whether I think my test passes or fails and why.

  107. Night Writer:

    Information obeys the laws of physics and has had a profound effect on how we view the physical world. If read up on information theory it will help you to understand.

    Right, shoot for insults; that will certainly add substance to your argument!

    You’re arguing for a reductionist view. You’re free to do so, but that perspective isn’t consistent with the exclusion of math from patent eligibility–because, by your argument, math (and the processing of math) is just as physical as information (and the processing of information). So you’re arguing that the patent system needs a new foundation. That may be an interesting discussion, but it’s also a completely different one.

    If you believe that it’s *right* to exclude math from patents, then you should a) define what math is and b) define what information is and c) explain how the one doesn’t capture the other. But I have to tell you, a lot of very smart people have already figured out how that book ends. And I peaked at the last chapter. You can’t get there from here; there is no way to separate “information” from “math”. Not unless you somehow cripple the definitions to take them outside what is normally meant by those terms.

  108. [PS, Jules, if an attorney has spent more time practicing software arts than practicing law, are they an attorney or are they a software engineer/designer/developer? If such a person wrote this blog post, is it an attorney beating down Michael, or is it someone else?]

    A balanced approach is the best way to go, in my opinion. The two-pronged, or should I say two-fisted, approach of practical+legal experience should win the day. Throw philosophical experience in there for the trifecta. When I said “attorney,” I was referring to that specific class of attorney that leans towards the ignorance side of things, i.e. less practical experience, more legal experience.

  109. “‘Otherwise the boundaries–and there certainly are and should be boundaries–will seem completely arbitrary and ad hoc.’

    Welcome to the real world…”

    I pick BigGuy for my dodgeball team.

  110. Michael, the risk of heart attack case is Arrythmia Res. v. Corazonix. See a discussion here: link to patentsusa.blogspot.com

    The risk of heart attack claim was patentable because it was based upon physically measured signals that were converted to data as a step in the claim.

    The recent Prometheus case was allowed under the MOT because is was based on a real blood sample.

    Benson has been confined to its facts; but still I think that case to be wrongly decided as it was a claim to an improved computer, not to a disembodied mathematical algorithm.

  111. Lionel:

    Michael, the whole point of this discussion is 101. There are many who argue that software patents should be unpatentable under 101.

    Yes, I believe the bulk of what I’m talking about is 101. Based on my understanding, though, “loom + cards” is not a 101 issue; the presence of the loom gets it past the first gate.

    The essence of my proposal deals with how to tell if a claim is loom, loom + cards, or just cards. The cases Malcolm asked me to look at were all “just cards”. Given my proposal they’re are all pure 101 failures. I think my proposal justifies and explains why–and does so with an empirical test.

    I would argue that if someone invents a new software problem that produces or is used to produce a physical result, it should be patentable under 101. And as others have pointed out, it’s no different than a process of operation in many cases and there is no argument that a method of operation passes 101.

    We would need a more specific example. I have no issue with Benson, Flook, Diehr or Bilski. By Diehr I would agree that software can be part of a patentable process; by Flook I would agree that software isn’t sufficient to make something patentable.

    Tonight I’ll see if I can think of a better example. Or perhaps you might come up with something interesting. I think the example should be consistent with the above decisions, and I think it should be consistent with the premise that math and laws of nature cannot be patented. It should help us to better understand where those lines should be drawn–and why. So: it should probably involve a machine which is purely physical and yet can be completely represented as a computer program. Or, in other words, as an effective method.

  112. (I just typed up a longer response which got lost, so i’ll try to summarize)

    Michael,

    My first example, I’ll agree, was bad. I thought that you were stating that your method removes extra solution activity as a result of the initial test, and i hadn’t realized that the test was also being supplemented by the “extra-solution activity” corrollary. Therefore, it’s a bit of a non-issue, except that i will say, that it can be far more difficult to determine what is extra solution than my extremely basic example.

    The heart attack risk comment may be referring to the case, Arrhythmia Research Technology, Inc. v. Corazonix Corp.

    Anyway, the decision is mostly concerned with 101 analysis over the patent eligibility of this data transformation in view of Diehr. One of the statements was that 101 doesn’t consider which features are old or new when considering patentabilty.

    Anyway the novely analysis on this case might be enlightening. (I.e. If EKG measurements are known, and if analog to digital conversion is known, it would leave the analysis method to provide a basis for novelty.)

  113. Lionel Hutz,

    Every computer-executed program produces a physical result. At minimum it coverts electricity into heat energy. :-)

  114. “Otherwise the boundaries–and there certainly are and should be boundaries–will seem completely arbitrary and ad hoc.”

    Welcome to the real world…

  115. Noise above Law (NAL):

    I am glad that at least you understood what I was getting at.

    There is no such thing in the field of “used” and useful computers as a generic stock computer. That is merely an abstract concept and myth.

    There is no such thing as a Turing machine with tape of infinite length and tape advancing speed greater than that of light. That too is merely an abstract concept and myth.

    The proof of point regarding the generic stock computers is when I ask YOU to dump your personalized computer that has that special combination of software and data that you value and replace it with a blank-slate store-bought computer every morning.

  116. “Physical world” so, the information processing that goes on in your head when you transform 3 into 6 isn’t physical? Angels are helping you are they.

    The fact is that what goes on in our heads and makes us the special animals is information processing. Computational intelligence is what it is called now.

    The fact is that these computer machines are doing what we do: information processing just as the physical machines of the iron age (read Stevens and Douglas) processed iron (and wheat.) And in Deener, guess what? Not one new machine. Just an arragement of the machines and method of how to process the wheat.

    Information obeys the laws of physics and has had a profound effect on how we view the physical world. If read up on information theory it will help you to understand.

    101 is very simple when you see that information processing transforms information. It answers all the silly riddles and makes MM look like the baboon he is.

  117. Mark,

    Thanks for the link. The reference therein to Dembiczek was interesting although the case before the CAFC appeared to have nothing to do with SW or 101. I did not realize the Commission engaged in such intellectual dis honesty.

  118. Night Writer:

    The whole 101 issue is simple. Transformation of information is a transformation as much as transforming wheat. Done.

    Er, that isn’t true. Information–by definition–is an abstraction. Just like numbers are. You can “transform” 3 into 6 by addition. Does that mean you can patent “3” or the process that turns it into “6”?

    What’s more, the process of transforming “3” into “6”–or the process of transforming a list of unsorted words into a list of sorted words–can be done entirely in your head. Does that mean you can (or should be able to) patent processes that can be done entirely in your head?

    You can’t transform wheat in your head. You can think about transformation all you want, but in the end to actually transform it you have to come back into the physical–and observable–world.

    This isn’t sophistry. The distinctions matter. I think in order to come up with a consistent theory you have to take those distinctions into account. Otherwise the boundaries–and there certainly are and should be boundaries–will seem completely arbitrary and ad hoc.

  119. Of course the Loch Ness Monster exists, and the longer you spend by the side of the loch the better the chance that you will see it for yourself.

  120. “And, yes, by patentable I generally mean 101, though I’m not well enough versed to know where some of the distinctions lie. For example, I suspect that my “loom + card” example is not a 101 issue but, rather, an obviousness issue. Or some variant thereof.”

    Michael, the whole point of this discussion is 101. There are many who argue that software patents should be unpatentable under 101.

    I would argue that if someone invents a new software problem that produces or is used to produce a physical result, it should be patentable under 101. And as others have pointed out, it’s no different than a process of operation in many cases and there is no argument that a method of operation passes 101.

  121. “My understanding is that Flook took a known device and a known algorithm and put them together.”

    How did that happen?

    Did Flook write the algorithm on a piece of paper, burn it, and then sprinkle the ashes on the known device and the new machine magically appear?

    Did Flook concentrate really hard on the algorithm and then projected his thoughts of that algorithm onto the known device and the new machine sprang into being?

    I think the answer to both of those questions is an emphatic “NO.” Flook created new structure that enabled the machine to operate differently. The fact that the new structure was allegedly based upon some known algorithm is not relevant.

    The operation of most mechanical devices can be described using known algorithms/formulas. Does that make them unpatentable as well?

  122. “Problem with hypo: nobody is arguing that methods of transforming wood aren’t patentable under 101.

    Better strawmen, please.”

    Then what are you arguing Mooney?

  123. “Might I ask: what sort of information and/or signal processing is a computer theoretically incapable of doing?”

    Ignor-ance is bliss … or in this case, ignor-ance, my name is Malcolm Mooney.

    Do you think your 1984 TRS-80 would be able to run Windows 7? Do you think your current day computer would be capable of functioning as an edge server in a distributed network? There are many, many, different types of computers, each having their own physical capabilities. Do you believe that the moon is made of swiss cheese and ghosts as well? How about the Loch Ness monster or bigfoot?

    Most software isn’t just about adding A to B or multiplying C and D. Instead, most software is about governing interactions between multiple computers.

    Of course, you wouldn’t know this because you are ignor-ant.

  124. Ned:

    Well Michael, I think the court in Diehr really said it all: if the claim viewed as a whole is directed to patentable subject matter, it is patentable even if the only thing new is the program.

    That could be read in a way that makes it inconsistent with Flook. Specifically: the mere addition of a new program obviously cannot make something patentable subject matter. I believe my proposal is consistent with Benson, Flook, Diehr and Bilski. Or, at least, that was my intent. :-)

    For example, what if the number represent the risk of a heart attack? The Feds said that was patentable.

    The Feds seem to be in the process of un-saying a lot of things right now. I’m not familiar with the patent you’re describing, but based on that description I’m not sure it would be approved in today’s climate–or upheld in today’s court. We’ll see. Exciting times and all that.

  125. Well Michael, I think the court in Diehr really said it all: if the claim viewed as a whole is directed to patentable subject matter, it is patentable even if the only thing new is the program.

    If the claim as a whole is directed to the algorithm, numbers in and numbers out, with the recitation of the machine only being incidental, meaning that the machine is merely a tool to do the calculations, then the claim is not patent eligible.

    The rules seem simple, but their application to real claim is difficult, not so?

    For example, what if the number represent the risk of a heart attack? The Feds said that was patentable.

  126. newish3700examiner:

    One problem I see is that I’m not positive that the test works for every type of extra solution activity. For instance, what if a claim says something like,

    “A computer-implemented process comprising:
    -connecting a power source to a computer;
    -dividing a digitally recorded number by 2 using the computer.”

    This method is still pre-emptive because the connecting step is extra solution activity and there doesn’t seem to be any real purpose defined use for the method, but it does recite a non-“effective” transformative step which is also arguably tied to the effective step.

    So it could be said to actually pass MOT and your test. Therefore, if this is the replacement test, then there would still need to be a further consideration of what constitutes extra-solution activity.

    I’m sorry, but I think I’m missing the heart of your question. From what I *do* understand–perhaps incorrectly–the essential part of your claim is an effective method and nothing more–despite being a machine. Therefore it fails my test (and Flook). Therefore it doesn’t qualify. (If the divided number were used as a step in a larger process it might qualify, as in Diehr–but clearly that isn’t what your claim does here.)

    Secondly, the courts would probably be opposed to implementing such a strong bright line test, and it seems unnecessary to impose such a restriction to patentablity for data transformation devices tied to specific purpose computers.

    While i’d grant that that the courts invalidated Benson because it was pre-emptive, they might have been reluctant to do so if the method was tied to a specific purpose computer for making a concrete diagnosis,
    i.e. wherein Benson’s method is employed on an automobile computer and includes an additional step wherein the binary data procured from the transformation of some recorded data is compared to a predetermined threshold in order to indicate a specific diagnostic condition of the car.

    I don’t claim that any court would ever listen to anything I say. :-) But clearly the courts are looking for a test that would be widely understood. The Federal Circuit has proposed a test–but I think it’s problematic and potentially inconsistent. My test may also be inconsistent; I’m just not yet aware of where or how. Perhaps you can think of a more challenging example?

    To get back to your main point, though: back to the mouse trap of my earlier post. I believe that an algorithm should be as protected as the use of a spring in a mouse trap. The use of a spring in a mouse trap certainly doesn’t prevent the use of all springs everywhere, and probably doesn’t even prevent the use of *any* spring in *any* mouse trap, but may prevent the use of springs in mousetraps that are “too similar” to the patented one. I have no problem with that at all. What my test does is to identify the “spring”–even when it’s disguised as (implemented in) hardware. And it does so in a completely objective, completely empirical way.

  127. The whole 101 issue is simple. Transformation of information is a transformation as much as transforming wheat. Done.

  128. Ned:

    Circling back to your machine controlled by cards, if the particular program encoded on the cards caused the machine to make a item better or in less time, I think that patentable. I hope you agree.

    On the face of it I have to disagree. My understanding is that Flook took a known device and a known algorithm and put them together. The act of joining the two was not sufficient, in and of itself, to create a machine that could be patented, even though the resulting machine (presumably) performed better than previous machines.

    While I certainly agree that it’s possible for the whole to be greater than the sum of its parts, I don’t think that the configuration of a machine that was expressly designed to be configurable could ever qualify as that “extra something” that makes a new machine–at least, not on its own. In other words, given the configurability of the machine, the loom + cards seems to have a higher threshold than the device described in Flook. If Flook’s modification doesn’t qualify the configuration of a configurable machine certainly couldn’t qualify either.

    Again, I’m not trying to talk about “what should be”; I’m trying to provide a consistent theory for (I think) “what is”.

  129. plurality,

    If I may speak for step back, the issue of whether Google Apps is or is not patentable subject matter is besides the point.

    Use the machine at hand and nothing else.

    SPQR 451, how long should we wait for a response?

    step back, it might be a long time before SPQR 451’s “same machine” is actually brought back to being the same machine that allows him to communicate with you (long time = never irl). I have proposed the same test for 6 awhile back. Funny, 6 ran from that challenge too.

  130. Malcolm:

    That’s from your 11:31 comment. Assume “determining” is defined in the spec as “inserting a needle, taking blood from a subject, and exposing blood to an agent which allows measurement of level of metabolite”. Is method patentable under 101 or not, according to your test?

    Here’s the simple answer: Not unless a physical device used in one of those steps is the thing being claimed. In that case (by Diehr) an effective method could be used as part of the claim. If no claimed device were part of those steps then (by Flook) an effective method is not sufficient to make it patentable.

    Here’s the nuanced answer: Patent law seems to contemplate that the whole can be greater than the sum of its parts. For example, I might successfully patent a mouse trap that uses known materials, known devices, known laws of nature but assembles them in some unique way. Presumably the same would have to hold true in your example, too–though it clearly doesn’t as stated. (None of the cases I’ve looked at so far have fallen into that nuanced zone–but still, the possibility is there. Or should be.)

    Even in the presence of such nuance, though, I make this claim: my proposal makes clear the essence of what cannot be patented–like the spring in the mouse trap. Indeed, it provides an empirical test for such “essence”.

    I still kinda like my “business method” test:

    “A method whose primary purpose is to achieve a profit or an increase in profitability, wherein the increase is achieved substantially without the creation of a new composition or the material transformation of a commodity for sale”

    Let me state my proposal a little differently.

    From a mathematical perspective the class of “effective methods” and the class of “computable methods” are exactly the same. In other words, any algorithm that can be executed by a person with a pencil and paper can be executed by a computer–**and vice-versa**.

    That fact supports an empirical test for effective methods. Although I tried to answer your question from the perspective of a person with a pencil and paper, the ability to write a computer program amounts to exactly the same thing. (There are deep theoretical underpinnings there; as I said earlier this is why and how we get “general-purpose computers”.)

    That removes the whole question from the world of the subjective and makes it objective.

    You claim x. If I can write a program that does exactly x I’ve proven that your x is an effective method–regardless of its original form. Likewise, if in order to reproduce x I have to break it down into y (an effective method) and z (some physical device or interaction) I’ve defined the “z” that’s covered by the patent–and the “y” that’s not.

    That brings us back to your first question: the point at which the effective method (or computer program) interacts with the physical world defines the essence of what is (or is not) being claimed. I believe that gets at the heart of what the machine-or-transformation test intends but potentially fails to achieve.

    An empirical test would have some value, no? :-)

  131. Another instance where the judge might have done better run the claims first through 35 USC 112 to access support in spec under paragraph 1 for the claims, and whether the claims were “definite” under paragraph 2. In fact, the judge confessed he hadn’t done any claim construction prior the motion for summary judgment based on “patent-eligibility” under 35 USC 101.

    That being said, the judge was on target to tank all of the “computerized system” claims which are problematical as covering both process and machine classes (a hybrid claim). The judge cites to Ex parte Lyell for this proposition, but he could also have cited the Federal Circuit’s decision in IPLX Holdings v. Amazon.com which agreed with Lyell that such hybrid claims straddling process/machine classes are invalid, but under 35 USC 112, paragraph 2, as being “indefinite” because it’s unclear when infringement of the hybrid claim occurs.

    I’ve got more trouble with the judge’s determination that the method claims of the ‘829 patent are “patent-eligible” but the method claims of the ‘425 patent are not. Contrary to what the judge says, the method claims of ‘425 patent don’t simply gather data, but in fact may manipulate data. For example, “estimating at a computer said taxpayer’s income tax refund amount” could require a calculation based on existing data. Put it this way, I don’t see why the method claims of the ‘425 patent are not “patent-eligible” when the method claims of the ‘829 patent are. This case again illustrates how subjective and nonsensical the Bilski “machine or transformation” test is. Again, start with 35 USC 112 to make sure you understand the meaning of, scope of and support for the claim.

  132. Well, that’s very gracious, Night, I must say. of course, Mark Nowotarski is not the only one. I think it’s a respectable position to hold (regardless what you think of some of the others you are in bed with, on that line).

  133. step back: “and no fair backing up your stuff on the internet”

    What, are you saying that Google Apps is not patentable subject matter?

  134. MaxDrei,

    I did follow the link. Perhaps it was cheap jib. But then few will likely follow the link and it looks as though Mark Nowotarski is advocating for no information processing patents. But, I admit. It was a cheap shot.

  135. …and Night, I really did think that Europe actually does have a vast lead over the USA, at least when it comes to such things as telecoms and writing computer games. But put me right on that, do.

  136. NWPA, I don’t follow you. When Nowotarski types verbatim the title of a press release from the Free Software movement, that doesn’t necessarily mean that he is personally espousing the EU Commission line, does it? Would you agree that your contribution deserves the description “cheap jibe”?

    Perhaps you could take the link and read the Press Release, mindful that the EU Commission predicates its initiative on the patent-eligibility provisions of the EPC ,which classifies programs (as such) for computers as not eligible for patent protection. The Free Software people say that’s not good enough. They see all those Beauregards etc issuing from the EPO. They want a ban on all inventions that are implemented by computer (I think).

  137. Yes, Mark Nowotarski, we need to catch up with Europe and their vast lead over the U.S.A. in information processing.

    And we all need to think more like MM. We all just need to accept our fate as slaves to capital and learn to let others worry about the bigger problems. Go home and have a nap.

  138. continuing the line of thought of Red Monkey:

    the statutory category of composition of matter should be nullified as obvious over the periodic table of elements, as all compositions are merely rearrangements of finite states of matter that are thus old in the art.

    The fallacy of that line of thought emerges slowly.

  139. Would you say a cement mixer isn’t eligible for patentability because it wasn’t big enough to mix the amount of cement you want?

    I would also point out that almost all methods are carried out by old machines. Fans, ovens, test tubes, etc.

    Information transformation takes time, energy, and space. Information transformation obeys the laws of physics. Transforming a list from being unsort to sorted takes time, energy, and space. A good method to transform the list can be billions of times faster and consume a fraction of the energy that a bad method to transform the list.

    Silly, MM-trollbot, blogs are for people.

  140. “A computer (by design) inherently encompasses a finite set of states (immensely vast) each one of which represents a unique arrangement of the contents of the computer’s main memory.”

    So by this logic, your post is the same as everybody else’s, because each post amounts to a particular arrangement of letters from a finite set?

    Also, by this logic all inventions are anticipated since each one is merely an arrangement of elements selected from a finite number of elements?

  141. “Also, we shouldn’t say “patentable” so loosely. “Patentable” means a claim passes 101, 112, 102 and 103. I think ML’s post uses “patentable” to mean passes 101.”

    If the honorable Supreme Court Justices can conflate “patentable” with “patent eligible subject matter”, why not everyone else on this esteemed Trainwreck?

    Recipe for gobblygook:
    take broad and narrow rules of law;
    mix indiscriminately;
    apply luddite tendencies;
    ignore the open gate invitation rationale;
    ignore the Congressional intent of later code sections explicitly talking about business methods;
    bake for six months;
    return with simple ruling that Bilksi claims an abstract idea and that MOT is not allowed as a brightline test, and that while being THE clue, MOT is merely still a clue and not the final answer.

  142. “This does not mean that no computer is patentable under 101.”

    about says it all. stop the mental wanking now please.

  143. “When any program is run on the computer, useful or not, the computer can only enter one or more of the range of states already inherently designed into the machine.”

    Well, for a modern computer, there is a non-infinite but extremely large range of possible states. So software design, in some sense, involves selecting combinations of those states such that the computer performs a particular function. On what basis is a non-obvious selection of such states excluded from patentability?

  144. “So, no, the computer never becomes a new machine no matter what program is being run on it.”

    SPQR,

    How about you putting your desktop where your mouth is?

    Let’s have you personally test out your hypothesis by doing the following:

    Every morning, before you start using your it’s-only-a-generic computer, we take the computer that was on your desk from the night before back to the store you bought it at (hopefully it’s an easy to carry laptop) and we replace it with a brand new –and according to you– absolutely identical computer. Now you may begin working –according to your logic– as if nothing at all has changed.

    We can of course simplify the above procedure if your generic computer has a restore-to-default option which reformats you hard drive and returns all other programmable memories (including flash BIOS) to their as-bought-from store states. Then you won’t have to bother getting a new “generic”, finite state-and-never-different computer machine from the store every morning.

    Are you ready to do that?

    (BTW, totally restored to as-bought state means we wipe out every day all the new “abstract” software and abstract data you may have stored on your computer the day before –and no fair backing up your stuff on the internet, or back up auxilliary disk, etc. you are restricted to using only that one computer machine with its finite memory, nothing else. Otherwise the experiment is contaminated.)

  145. Addressing the question of whether or not a computer+software=new machine.

    A computer (by design) inherently encompasses a finite set of states (immensely vast) each one of which represents a unique arrangement of the contents of the computer’s main memory.

    When any program is run on the computer, useful or not, the computer can only enter one or more of the range of states already inherently designed into the machine.

    So, no, the computer never becomes a new machine no matter what program is being run on it.

    The so-called transformations (argued above) that the computer experiences are nothing more than certain sets of inherent states that are made possible by the original design of the computer. And based on the premise that the machine is programmable to reach each of the vast range of inherent states of which the computer was designed to perform as it’s original specification.

    Stated another way, the physical computer programmed with software “A” is the exact same physical computer as that computer programmed with software “B” because the physical computer inherently can be programmed into all possible states reachable by it’s design parameters, i.e., the size and extent and arrangement of it’s main memory.

    The so-called “transformation” that non-computer engineers see is a mis-understanding of the inherent nature of the programmable computer.

    This does not mean that no computer is patentable under 101.

    A new computer having a new physical processor, or a new physical memory can be a new machine that is statutory under 101 based on new physical features of the new programable computer.

    But, every new computer inherently is designed to be programmed to the full extent of it’s design capacity based on the designer’s goals.

    However, this programmable capability does not turn the machine into a new machine simply because a user wrote some software it was designed to be able to run.

    Enough of this transformation malarky, please.

  146. “A computer-implemented process comprising:
    -connecting a power source to a computer;
    -dividing a digitally recorded number by 2 using the computer.”

    This claim is not really a computer-implemented process because part of the process requires a step of connecting a power source to a computer. This is actually a method for dividing a number. Such a method is tied to a machine (a power source) and therefore seems to pass M, but fails under the corollary of not meaningfully limiting (i.e. extra solution activity) since arguably all computers require a power source.

  147. to 6:
    not sure what your temperature sensitive device implies, but the decision about infringement of a snapping shelf for a fridge may be relevant, what do you think?
    I remember it was discussed here, and the patentee was Saint-Gobain.

  148. I still kinda like my “business method” test:

    “A method whose primary purpose is to achieve a profit or an increase in profitability, wherein the increase is achieved substantially without the creation of a new composition or the material transformation of a commodity for sale”

  149. Mooney – I was arguing that a method executed on a computer tranforms the computer’s memory and should therefore satisfy the MOT test, not that it, in and of itself, settles the issue of patentability under Bilski since there are still the gotchas of fundamental principle and post-solution activity. You really side-stepped the gist of my post. But your retort that not everything of value is patentable is fair enough. Me bad for going down that road because the value of software and whether a computer-implemented method tranforms computer memory are really separate issues.

    But then you draw an analogy to a recipe. The implication is that a recipe, written on paper, fails patentability under the printed matter doctrine and therefore so too should software. I don’t want to go rounds on the printed matter doctrine, but a method for cooking something is patentable, even if the sheet of paper on which the recipe is written can’t be claimed. A method of baking Twinkies comprising… is not just a set of instructions but a claim to a patent eligible method. By analogy, a claim to a method comprising using a microprocessor to calculate X and then determine Y from X should be a claim to a patent eligible method, though if you follow the printed matter doctrine analogy the Beauregard claim would not be.

    I’m not interested in getting into protracted arguments on this. My position is this, I think there are a lot of valid arguments on both sides. I see form exalted over substance both in saying that putting an otherwise unpatentable method on a floppy disk makes it patentable, as well as calling a special-purpose computer patentable but a general purpose computer programmed to perform the same task unpatentable. On balance, I personally believe the better policy is to accept software as patentable, and I would prefer Congress to step in and make that clear. But I can go either way, I just want a rational policy that I can explain to my clients instead of having to say “well, it’s a crap shoot, do you feel lucky?”

    What we have in software is simply a different beast than anything else and it needs to be treated as such. By analogy, classical physics had waves and it had particles, and then an understanding of light emerged that said light was neither wave nor particle but has properties of both. In my estimation, that’s what software presents to the patent bar. It is neither 15th century horseshoe making nor “just math.” No amount of analogizing to prior patentable and unpatentable subject matters really is adequate to come to terms with software.

    Returning to my original post, if you want to present policy reasons against patentability of computer-implemented methods, fine. But please if we must apply the MOT test, let’s admit that a general purpose computer is really a specific machine, and executing software on the general purpose computer really does transform the memory.

  150. Michael,

    It’s an interesting idea, and i think it’s at the very least a somewhat viable supplementary test to the MOT test, if not a full replacement, but I’m not sure I can agree that the test is comprehensive.

    One problem I see is that I’m not positive that the test works for every type of extra solution activity. For instance, what if a claim says something like,

    “A computer-implemented process comprising:
    -connecting a power source to a computer;
    -dividing a digitally recorded number by 2 using the computer.”

    This method is still pre-emptive because the connecting step is extra solution activity and there doesn’t seem to be any real purpose defined use for the method, but it does recite a non-“effective” transformative step which is also arguably tied to the effective step.

    So it could be said to actually pass MOT and your test. Therefore, if this is the replacement test, then there would still need to be a further consideration of what constitutes extra-solution activity.

    Secondly, the courts would probably be opposed to implementing such a strong bright line test, and it seems unnecessary to impose such a restriction to patentablity for data transformation devices tied to specific purpose computers.

    While i’d grant that that the courts invalidated Benson because it was pre-emptive, they might have been reluctant to do so if the method was tied to a specific purpose computer for making a concrete diagnosis,
    i.e. wherein Benson’s method is employed on an automobile computer and includes an additional step wherein the binary data procured from the transformation of some recorded data is compared to a predetermined threshold in order to indicate a specific diagnostic condition of the car.

    Thoughts?

  151. Michael, I’m not going to get into details of your examples, but in my view, Section 101 as two distinct requirments: utility within the “useful Arts” as that term was understood in 1789, and a claim properly characterized in one of the four classes. Over the years, the Supremes have never really addressed what was included as a Useful Art, but they have said that natural phenomena, laws of nature and abstract ideas cannot be patented. Presumably, these are not within the Useful Arts; but clearly also, they are not within the four categories.

    That said, useful applications of laws of nature and natural phenomena can be patented. Ditto abstract ideas, such as mathematics that solve mathematical problems.

    Now, what I take from Benson is that the Supreme court did not want to authorize a claim to mathematical formula where the computer was merely used as a tool, like a pen and paper or a calculator, used to help a human calculate the formula. However, if instead the computer is improved or made different by use of the algorithm, that is a recognized utiliy, an improved machine, that should make the claim patentable either as a method or as a machine.

    The problem with Benson from a realistic point of view was that BCD is the heart and sole of many computers and conversion between BCD to binary is required for efficient machine operation. The invention in Benson was to the improved computer, not to the calculation of an abstract mathematical formula.

    Many types of software additionally get the computer to behave differently. An example would be a spreadsheet or a word processor or the like. These are not merely numbers in and numbers out, but way of interfacing the computer with the human in an efficient manner. Clearly useful. Clearly patentable.

    Circling back to your machine controlled by cards, if the particular program encoded on the cards caused the machine to make a item better or in less time, I think that patentable. I hope you agree.

    The central problem in these business method cases is not the four categories, but utility. Are business methods “useful Arts” as that was understood at the founding. They clearly are useful. They clearly produce concrete and not abstract results. But do they have the kind of utility envisioned by the Constitution?

    That is the real question before the Supreme Court in my view.

  152. ML If the devices are part of the claim, then by my premise the the question of patent eligibility rests on the eligibility of the devices.

    That’s from your 11:31 comment. Assume “determining” is defined in the spec as “inserting a needle, taking blood from a subject, and exposing blood to an agent which allows measurement of level of metabolite”. Is method patentable under 101 or not, according to your test?

  153. Classen:

    Claim 1

    A method of determining whether an immunization schedule affects the incidence or severity of a chronic immune-mediated disorder in a treatment group of mammals, relative to a control group of mammals, which comprises:

    immunizing mammals in the treatment group of mammals with one or more doses of one or more immunogens, according to said immunization schedule, and

    comparing the incidence, prevalence, frequency or severity of said chronic immune-mediated disorder or the level of a marker of such a disorder, in the treatment group, with that in the control group.

    This seems to be more of the same so I won’t go into it. Perhaps what I need to do is find counter-examples for things that were successfully patented but shouldn’t have been by my premise. In other words, a different kind of contradiction.

    ===

    Since so far I’ve been agreeing with things that shouldn’t be patented, let me give an example of something that I think should be–and some potentially surprising implications.

    Secure web transactions generally depend on SSL. For example, when you log into your bank via the web your communication is encrypted end-to-end by SSL.

    SSL is computationally very expensive. While you can execute the entire algorithm on a general-purpose computer, high traffic sites often use special SSL co-processors in order to free up the main CPU for other tasks. Under my premise such co-processor boards are patent-eligible. The algorithms they execute aren’t, but the selection, arrangement, etc. of hardware components is.

    Now that entire board was designed just to execute an algorithm very quickly. Clearly what it’s doing–and even the way it’s going about doing it–could be completely simulated in software. Because, from that perspective, *everything* it’s doing is just the execution of an effective method. It’s a computer, a special-purpose computer, but a computer, and all computers do is process effective methods. The software emulation would not, under my premise, be an infringement of the patent.

    Of course, the emulation won’t run anything like the hardware did, so you’d get no benefit by running an emulation. But you could, and without infringing on the patent. (As always, I’m talking about patent infringement, not copyright.)

    I think that’s potentially surprising, but until I find a contradiction I think it’s consistent with the meaning and intent of current patent law.

    I think the machine-or-transformation test might lead to a different result–but I think the machine-or-transformation test potentially leads to contradictions. Any algorithm can be implemented as a completely physical machine. The machine-or-transformation test doesn’t speak to whether or not different machines implementing the same algorithm infringe. If they are seen to then a physical implementation of an algorithm blocks all uses of the algorithm–in violation, I believe, of Benson.

  154. Bilski:

    Wikipedia’s Summary of Claims

    (1) initiating a series of sales or options transactions between a broker and purchaser-users by which the purchaser-users buy the commodity at a first fixed rate based on historical price levels;

    (2) identifying producer-sellers of the commodity; and

    (3) initiating a series of sales or options transactions between the broker and producer-sellers, at a second fixed rate, such that the purchasers’ and sellers’ respective risk positions balance out.

    Again, I’ll initially assume that the way the transactions are initiated, recorded, etc. are not part of the claim (and that I can therefore add them to my effective method without altering the eligibility of the claim).

    The above doesn’t give me much to go on, but the steps might be something like:

    (i) Scan securities
    (ii) If security is [condition 1] or [condition 2] or […]
    (iii) Scan list of purchasers
    (iv) For each purchaser
    (v) Lookup purchaser’s phone number
    (vi) Call purchaser
    (v) Offer purchaser price x for security, where x is > historical low and x < historical high (vii) etc. Again, these are a series of steps that can be executed by a person, a pencil, paper--and devices that aren't being claimed. (Believe me, I know I muffed the details of the steps completely--but that doesn't matter to the line of argument. If you disagree, please advise which step couldn't be written down. By, you know, writing it down. :-)

    As with Prometheus, if devices are being claimed, then patent eligibility rests with them--and the effective method is not sufficient to make them eligible.

    But note that a general-purpose computer is not enough to claim eligibility--because (by definition) a general-purpose computer isn't eligible on its own, and (by my premise plus Flook) the effective method can't make it eligible.

    As before, a purpose-built device might be eligible on its own merits--but the claim would stand on the merits of the device apart from the effective method and there would be no patent on the effective method itself (i.e., the abstract representation of the algorithm).

  155. I can understand why the tax claim in the opening post is unpatentable under Bilski, but what I do not fully understand is why the examiner raised a fuss in the first place since, before Bilski, business methods were patentable per se. Was the problem the preamble that tried to label this business method as an apparatus?

  156. 6 said:

    I’m about to allow a claim that is a method of making a device wherein a limitation of the properties of a device’s structure depends on a wherein clause that depends on the temperature of the surrounding environment. The claim is definite it seems to me like, but it is really a grey area. You could perform the method in a cold environment and the properties of the device formed would not be sufficiently the same for the method to infringe, whereas you could perform the exact same method in FL on a hot day and the device formed would be the same exact device as you made in the other environment but you would infringe because the properties of the device would then meet the claim limitations.

    Then if you took the device made in the cold environ to FL on a hot day it would then have sufficient properties to meet the structural claim limitations of the device being formed within the method claim.”

    I am still a bit unclear of what is being claimed here. Is the claim directed to the environment at time of formation or use? The reason I ask is that your example included taking a formed structure to FL on an hot day. That would be completely irrelevant if the temperature related only to formation.

    That said, I know that one way to make structures operate over a larger temperature range is to temperature cycle them after formation. This way, stress are relieved, gases emitted, etc. etc. The structures before and after may seem the same, but they are not in fact the same.

  157. Prometheus:

    Claim 1

    A method of optimizing therapeutic efficacy for treatment of an immune-mediated gastrointestinal disorder, comprising:

    (a) administering a drug providing 6-thioguanine to a subject having said immune-mediated gastrointestinal disorder; and

    (b) determining the level of 6-thioguanine in said subject having said immune-mediated gastrointestinal disorder,
    wherein the level of 6-thioguanine less than about 230 pmol per 8×108 red blood cells indicates a need to increase the amount of said drug subsequently administered to said subject and
    wherein the level of 6-thioguanine greater than about 400 pmol per 8×108 red blood cells indicates a need to decrease the amount of said drug subsequently administered to said subject.

    I assume for a moment that the devices used to “administer the drug” or “determine the level” aren’t part of the claim. In other words, I assume for a moment that both steps use known devices or tests and that (b) is the essence of what Prometheus is trying to claim.

    If that’s true than an effective method is:

    (b) (i) “determine the level” which could be “read level from paper”
    (b) (ii) if level is “less than about 230 pmol per 8×108” then “increase the amount of said drug subsequently administered”
    (b) (iii) if level is “greater than about 400 pmol per 8×108” then “decrease the amount of said drug subsequently administered”

    The amount of drug to be administered subsequently–ie, the amount to increase or decrease–would be contained within the steps of the effective procedure. (It would have to be enumerated *somewhere* for the above procedure to work, and if it can be enumerated somewhere it can be enumerated on a piece of paper and made part of the steps to be followed.)

    If the devices are not part of the claim, then all that is being claimed is an effective method, and by my premise there is no patent eligibility.

    If the devices are part of the claim, then by my premise the the question of patent eligibility rests on the eligibility of the devices.

    But, more specifically, if the devices are not eligible then (in keeping with both Flook and my premise) the use of an effective method is not enough to make them eligible.

    There is one other variation to consider, and that is if the effective method itself were implemented in a device (rather than, as it seems above, by humans).

    By my reasoning that device could be patent eligible–leaving aside issues of obviousness, etc.–but the patent, if issued, would cover the device as a whole, not the effective method itself. In other words, someone could build a different device which used the same or similar effective method, and the use of that effective method–the essence of the algorithm, as it were–would not be infringing.

    To my mind that’s like saying another device could use electricity, a motor, a pump, tubes, etc. without infringing.

    Also important to note: it wouldn’t matter if the steps of the effective method were loaded from a disk, if they were burned into ROM, or if they were implemented as a series of gears and levers. By my premise the steps themselves cannot be patented, even if a specific physical manifestation could be.

    ===

    Just to be a bit clearer: I did a bit of hand-waving to cover “determine the level”. I initially assumed that how that was done wasn’t claimed; that made it OK to add it to my effective method, even though effective methods are supposed to be pencil & paper stuff. The argument at the end is meant to cover what happens if you don’t buy that: then patent eligibility rests on the device that does the “determining”–but, as I argued, simply tying that device to an effective method isn’t enough to make it patent eligible.

  158. “Each of those fields represents our efforts to codify our understanding of the world around us. Presumably the law has some connection with reality”

    Science makes such an effort, but the law’s purpose is to achieve some policy goal. It may well be that allowing software to be patented is the lesser of three evils.

    Despite the uncertainty and cost of patent litigation, there is also a non-negligible uncertainty and cost in both copyright and trade secret litigation. At least in patents you start with a claim. In both trade secret and copyright, a court needs to figure out what the protected subject matter is without the benefit of a claim, and at least in the case of trade secret, society is denied the benefit of disclosure.

  159. Could you maybe take some actual claims (Bilski’s claim, Prometheus’ claim, Classen’s claim, the In re Gardner claim) and apply your test? Just curious to see how the test works in practice in its creator’s hands.

    Yes, I’ll try. (Just keep in mind that I have about three days’ experience with all of this; I think I understand some of the broad strokes, and I’m aware of some of the nuances, but I certainly don’t have the lingo and details down pat. And no doubt there’s much I’m missing altogether.)

    First, I’ll start with an informal definition of an effective method. An effective method:

    1. Is a finite sequence of steps
    2. That can be performed by a person
    3. Using no more than a pencil and paper

    There are more rigorous definitions–and there are formal definitions that don’t involve people at all. But I personally like the human aspect because I think that touches on part of what is meant to be excluded: things that can be done mentally by people.

    I should note that I’m not touching on the efficiency with which the above can be done. For example, there are algorithms for factoring numbers. It would take a human an impossibly long time to factor a very large number–it takes *computers* a long time, public key cryptography is built on that fact–but the point is that it *could* be done, and that’s all we’re talking about. After all, the goal is to try to draw the line between form and substance. (As I said, I agree that a specific implementation in hardware is patentable. But that patent couldn’t prevent someone from creating new hardware that executed the same basic algorithm more efficiently.)

    I should also note that I started thinking about all of this because of an excellent post on computational theory at Groklaw. That post started from effective methods and tried to establish a chain of reasoning leading to “it’s all math” and “math isn’t patentable.” While I appreciate the reasoning I didn’t think it amounted to an effective and useful test.

    Second, I’ll try to state my premise: a) a patent may contain (make use of) an effective method; b) the effective method itself cannot be patented. Since an effective method may be used but not patented to me it follows naturally that c) a patent must consist of more than just an effective method. (To my mind that line of thinking corresponds exactly to the use of math or a law of nature within a patent. Or, more specifically, to my understanding thereof. :-)

    I’ll tackle the cases in separate posts. Since I probably can’t do all of them let me know if any are particularly interesting.

  160. ML: I believe that “Effective Methods” (or “Effective Procedures”) are not patentable. An “effective method” is a term of art from computation theory, and basically it’s a finite series of steps that can be carried out by a person with no more than pencil and paper. (There’s a little more to it–but not much.) All algorithms are “effective methods” and all “effective methods” are algorithms.

    A patentable machine may certainly include an effective method, but I believe that a machine which consists *only* of an effective method cannot be patented.

    Could you maybe take some actual claims (Bilski’s claim, Prometheus’ claim, Classen’s claim, the In re Gardner claim) and apply your test? Just curious to see how the test works in practice in its creator’s hands.

  161. There’s something undeniably different about a programmed computer as compared to an unprogrammed computer, and that, IMHO, is a transformation.

    Well, yes. The computer is now capable of processing information immediately whereas before someone needed to program it first. The hammer in my hand is likewise transformed into a more useful hammer when someone picks it up from the ground. Big whup. Can I patent the instructions: “Pick up the hammer on a Saturday at 1:00 pm with your left hand if it’s raining in New York, less than 50 degrees, your name begins with a Y, and your favorite movie is RINGMASTER; otherwise pick it up with your right hand?”

    I mean, why not? Promote the progress. Better I should keep this algorithm a trade secret?

    Might I ask: what sort of information and/or signal processing is a computer theoretically incapable of doing?

  162. HierarchyOfPontificationBuckets wrote:
    “A person in a library can help lead to these kind of advances, but the person in the library can’t perform these things. ”

    Yes they can. From the theoretical point of view, they are effective methods, so by definition a person can do them. From a practical point of view: Do you really think it is impossible for a doctor in a library to diagnose diseases or for any person in a library to recognize faces or teach an autistic child to recognize facial expressions?

    If you’re trying to point out the difference between what a human can do in practice or in theory, there are much better examples: such as rendering complex graphics in a video game or testing extremely large numbers for being prime.

    If that was your point, were you trying to suggest that the line should be drawn at the limit of human performance?

    HierarchyOfPontificationBuckets wrote:
    “I don’t think that various interesting and academic characteristics of computers should exclude these kinds of advances from patent eligibility.”

    It is the ‘academic’ characteristics of computers that allow the very practical results of virtual machines and interpreters, among other things. I don’t think you can draw a clear line between what is theoretical and what is practical. If you make a legal argument to a judge using a correct legal theory, and persuade the judge, haven’t you achieved a very practical result?

  163. There are many many things a computer+software can do that no brain can do.

    If you’re talking about what software–i.e., algorithms, i.e., effective methods–can do then, strictly speaking, that isn’t true. (If it is true it would come as quite a surprise to logicians, mathematicians and computer scientists!) If you’re talking about the characteristics of a particular device–i.e., its speed or the ability to detect certain colors or some such thing–then yes, that’s true, but we’re now talking about a physical device, not software (i.e., not algorithms, i.e., not effective methods).

    The law is not interested in computer science theory.

    The law is as interested in computer science theory as it is in math, science, medicine, etc. When those fields have something relevant to teach, the law listens. Or tries to. Otherwise what exactly forms the basis for legal theory? (Each of those fields represents our efforts to codify our understanding of the world around us. Presumably the law has some connection with reality. :-)

    In any event, what I was talking about lies at the heart of computer science theory, not the fringe. It defines, for example, the fact that any known algorithm can be expressed in any known language (as long as that language meets a very low threshold). Or to put it another way: that a general-purpose computer can execute any known algorithm. That is *exactly* the theory that makes general-purpose computers possible (and useful).

    How about this. I’ll limit my so-called software claims (computer configured to implement a particular algorithm) to only implementations where every step is performed by a computer. I won’t claim any abstract algorithms. Now you are free to use pencil, paper, brain, speech and conversation, or anything else other than a computer to practice the algorithm. You can publish it in a book. You can discuss it in computer science class. You can even build an assembly line structured to operate in a way that map to the algorithm.

    Practically, you can code it and study it in private and as long as it is only for private study no one will know or care, though technically you would be infringing. It would be nice if there was some kind of “fair use” exception in patent law like there is in copyright law …

    If I understand your example correctly it sounds to me like you’re trying to apply the same rule used with “laws of nature”. That is, you’re suggesting that the algorithm itself can’t be patented, but its use within a particular “process” can be. I agree with that–but I also agree that the algorithmic part of what you’ve done can be lifted out and re-used without infringing, exactly as a law of nature can be. (Or, to be more specific, with exactly whatever limits would be placed on a law of nature lifted out of a process, etc.)

    That’s just what Diehr says. The fact that a process contains an algorithmic element doesn’t prevent it from being patented. But Benson says that if it *only* contains an algorithmic element it isn’t patentable.

    What I’m adding to the discussion is a way to distinguish form from substance. Under my “effective methods” test–and unlike the “machine or transformation” test–an effective method is an effective method, regardless of whether it’s implemented in hardware or software. Or to put it differently: anything that can be described entirely as an effective method is not patentable, regardless of its form.

  164. Perhaps, but you don’t want to claim writing information to memory, you want to claim the functionality of that writing don’t you? And you want to do it without making any reference to the change that took place. The change therefore is “incidental” at best. And that margenalization may very well kill your claim. After all, it was you (the applicant) that margenalized the transformation, not me or the courts.

    Good point 6.

    There’s something undeniably different about a programmed computer as compared to an unprogrammed computer, and that, IMHO, is a transformation.

    In a method claim (e.g. a method of making rubber) that says a computer is programmed to conduct a certain method step, the actual method step is not the programming of the computer itself but rather the certain method step conducted by the programmed computer (e.g. using the programmed computer to monitor the temperature of the mold). In other words, there is no transformation within the context of what is sought to be claimed. Of course, if the claim recites a method of programming a computer, perhaps an argument can be made that the programming itself does in fact constitute a transformation. The real question for computer-implementation cases is whether a computer programmed to perform a specific method step constitutes a sufficient “tie-to” of a particular machine sufficient to pass muster under the m or t test.

  165. Znutar: “If there’s no difference, why did you spend all the extra money on software when you bought your last computer?”

    Spending money on software doesn’t prove your point. Nobody is saying that software does not have value or that software has no function. People pay for non-patentable stuff all the time.

    I spent the money on software because the software is copyrighted. If it was free, I wouldn’t have spent the money. Unless perhaps I was paying for support down the road.

    This doesn’t change the fact that my computer is still just a computer and software is just instructions that allow the computer to do what it is designed to do, just like a recipe allows a chef to do what a chef does.

  166. Sure, but what a person+library can do is irrelevant, because no one is advocating patentability of that.

    Are you sure? I mean, why not? Don’t law firms have their own libraries any more? I’m sure people still look at books. Get a patent, file a lawsuit = PROFIT.

    It’s sort of ridiculous to say “nobody is advocating” a patent on “person + library” when people are advocating patents left and right on “computer + doing anything”. It’s all part of the same problem, which is why this mess is going to be resolved fairly quickly (as these things go).

  167. “Do you understand that the word “new” has different meanings?”

    Yes, and you’re relying on the wrong one to have your position make any sense.

    The fact that the word has many meanings doesn’t mean that you can pick and choose whichever one suits you at any given time or in a given context. The context we are in now is patents. Your “definition”, such as it is, is not the definition we’re looking for in the patent context. Don’t worry, USSC will help you out with this concept here shortly I have little doubt.

    “If I take a lump of a ferromagnetic material and heat it above the Curie temp it’s no longer ferromagnetic but it’s not a new lump,”

    Actually it is.

    “I’m going to jump on “new,” too. A wad putty that I’ve press my thumb into is not a new wad of putty, but it is transformed. A claim to making thumbprints in putty is patent eligible. ”

    You probably lack utility. Ineligable.

    “the substrate is not new, but it is transformed. ”

    Actually it probably is new. It’s atoms are rearranged aren’t they? I could be wrong, it has been a long time since I dabbled in magnetism and I would have to brush up.

    “Writing information to memory transforms the memory. It may not be a very big change, but it is undeniably a change, else no information has been stored. You can trivialize it all you want, but it is a change of state of each bit that is affected. ”

    Perhaps, but you don’t want to claim writing information to memory, you want to claim the functionality of that writing don’t you? And you want to do it without making any reference to the change that took place. The change therefore is “incidental” at best. And that margenalization may very well kill your claim. After all, it was you (the applicant) that margenalized the transformation, not me or the courts.

    “nonesense to argue that there’s no transformation”

    Well that depends entirely on the claim. You’re making certain presumptions (that we’re writing info to a disk or something) that aren’t necessarily always present.

    “So why are people still making the rediculous argument that a CPU + instructions in memory is the same as a CPU + without instructions? ”

    In the words of the USSC: You can’t argue [otherwise].

    Also, because it is.

    “If there’s no difference, why did you spend all the extra money on software when you bought your last computer?”

    Because I’m too lazy, and/or don’t have the time and inclination to properly utilize the computer on my own.

    “There’s something undeniably different about a programmed computer as compared to an unprogrammed computer, and that, IMHO, is a transformation.”

    Yes, one requires less effort to use. That’s not really a transformation :(

    “You’re not an abstract integer are you?”

    Yes, I am.

  168. I just don’t see how people can simply ignore the bottom line. Computer+software is practical, real, physical, capable of being built in ways that are new and unobvious and useful. Why is the reconfigurable nature of the computer’s logic a strike against it? To me, your reasoning defies logic.

    I wanted to cover that question separately because I think it’s a “what should be” question rather than a “what is” question.

    I don’t know the history of patents well enough to give your question a realistic answer. All I can do is give you my perspective based on what I understand and believe.

    First, patents are a big stick. In order to encourage progress they provide a government-sanctioned monopoly within a system that by default assumes monopolies discourage progress. So right from the start patents try to balance two opposing goals. While we might debate exactly where the line should be drawn hopefully we don’t disagree there *is* a line to be drawn. (Of course, saying “it’s expensive to do x” is not enough. Building an airline is expensive, as is building a skyscraper, as is building a cruise ship, but those things should clearly not be covered. If patents were the only incentive to large-scale investment, why would people ever bother with such endeavors? Clearly there’s more to it than just the cost of setting something up.)

    Second, I think the law tries to lay out some fundamental rights that we all have. I think patent law is set up to avoid stomping on some of those rights. For example, I don’t think patent law wants to get inside your head, to tell you what you’re not allowed to think (or to force you to say what you do think). Rather, it seems to focus on the observable–e.g., what we do and how we go about doing it. Algorithms–i.e., effective methods–start down that slippery slope of getting inside our heads because (by definition) they embody exactly things that can be done with some very minor (and purely mental) aids.

    But I know I’m way off on a branch here. This isn’t something I’ve read much about or thought much about. As I said, I was trying to find a theory that could make what I was reading consistent (and, hopefully, more predictable).

  169. I’m going to jump on “new,” too. A wad putty that I’ve press my thumb into is not a new wad of putty, but it is transformed. A claim to making thumbprints in putty is patent eligible.

    If I take a lump of a ferromagnetic material and heat it above the Curie temp it’s no longer ferromagnetic but it’s not a new lump, but it is transformed. A claim to heating a ferromagnetic lump above the Curie temperature is patent eligible.

    If I use an electromagnet to magnetize portions of a substrate, the substrate is not new, but it is transformed. A claim to selectively magnetizing portions of a substrate is patent eligible.

    Writing information to memory transforms the memory. It may not be a very big change, but it is undeniably a change, else no information has been stored. You can trivialize it all you want, but it is a change of state of each bit that is affected.

    You can go off about the transformation being post-solution activity, I suppose, but it is nonesense to argue that there’s no transformation. There are plenty of ways to be dismissive of a method claim these days, even though the claim otherwise passes either prong of the MOT test. So why are people still making the rediculous argument that a CPU + instructions in memory is the same as a CPU + without instructions? If there’s no difference, why did you spend all the extra money on software when you bought your last computer? There’s something undeniably different about a programmed computer as compared to an unprogrammed computer, and that, IMHO, is a transformation.

  170. “”Is a computer reconfigured by software a “new computer”, maybe not,”

    You just shot yourself. ”

    Which of the 8 definitions of “new” are we using? Are you “6”. Maybe, maybe not. Are we talking about the meaning it has as a pseudonym on PatentlyO, or the meaning it has as an integer. You’re not an abstract integer are you?

    You can read, but can you comprehend, specific integer formerly known as “6”?

  171. “”Is a computer reconfigured by software a “new computer”, maybe not,”

    You just shot yourself.”

    I see you can read, but comprehending seems lacking. Do you understand that the word “new” has different meanings? Are you “6”? Yes and no. No, you’re not the numeric abstract concept of a specific integer. Yes, you’re the person who goes by the pseudonym “6” on PatentlyO. Are you “6”? Maybe, maybe not. Depends on what meang of “6” you use. Is something “new”? Depends on which meaning of “new”.

    Is something a “new computer”. It depends on what meaning of “new” you’re using. The patent meaning, as in section 102? My Webster has “new” as an adjective with seven meanings and it also has meaning as an adverb.

  172. “There are many many things a computer+software can do that no brain can do

    And vice versa.

    There are many things a human with a library of books can do that no human can do. Can you patent the books? Can you patent the person with the books? How about an old robot with new books?”

    Sure, but what a person+library can do is irrelevant, because no one is advocating patentability of that. What a compuer+software is relevant.

    Here are some examples, as if all the others haven’t been enough.

    “Disease-Matching Software Could Save Children … Software tools are being developed that can search and compare patient data at hospitals across Europe to find children with closely matched conditions. The doctors can then study how the matched patients at other hospitals were treated and whether that treatment was successful. The information will greatly improve doctors’ ability to choose the right path for their own patient.”

    “Software Helps Doctors Calculate Heart Attack Risk … Pioneering computer software is helping doctors to decide how best to treat patients admitted to hospital with suspected heart attacks. An international consortium of researchers, led by the University of Edinburgh, has developed a programme that enables doctors to swiftly assess the severity of a patient’s condition. The new ‘risk calculator’ is already being used in British hospitals. Doctors using the new system take key data from patients at their bedside, and input it into the specially-devised programme. Key facts — such as a patient’s age, medical history and blood pressure — are recorded by doctors, as well as information derived from on-the-spot blood samples and kidney tests. … Significantly, it will be able to predict the likelihood the patient suffering a heart attack, and even their chances of dying in the next months”

    “Research into techniques of facial biometrics, carried out by scientists at Universidad Carlos III de Madrid (UC3M), has resulted in a system that is able to recognize the facial “DNA” of every individual by determining his/her most noteworthy facial traits, with a of 95% rate of precision.”

    “Computer scientists from Nanyang Technological University in Singapore are working on the development of an efficient and intelligent facial expression recognition system. The system is capable of locating the face region using derivative-based filtering and recognizing facial expressions using boosting classifier. The portable device is being developed to help autistic children understand the emotions of surrounding people.”

    See sciencedaily.com for detail on these and many others. Without resort to artificial categories, these are the kinds of things the patent system has historically encouraged and rewarded and should continue to do so.

    A person in a library can help lead to these kind of advances, but the person in the library can’t perform these things. I don’t think that various interesting and academic characteristics of computers should exclude these kinds of advances from patent eligibility.

  173. Having recently received my associate’s degree in claim drafting from the University of PatentlyOComments (under the tutelage of Professor ActualInventor), allow me to show you claim drafting neophytes how this whole mess could have been avoided.

    “1. A computerized system for distributing spending vehicles on an airplane comprising: …”

    Ta da! Statutoryish as can be!

  174. “the loom is patentable and the combination of loom + specific cards is not”

    How can adding an element to a structure that falls within the ambit of 101 somehow remove it from the ambit of 101? I would think if loom is within 101 then loom+anything is also within 101.

    Also, we shouldn’t say “patentable” so loosely. “Patentable” means a claim passes 101, 112, 102 and 103. I think ML’s post uses “patentable” to mean passes 101.

    To clarify:

    By “loom + card” I mean if the loom itself had already been patented. In that case I don’t think the addition of a specific set of cards would be enough to qualify as a newly patentable machine.

    And, yes, by patentable I generally mean 101, though I’m not well enough versed to know where some of the distinctions lie. For example, I suspect that my “loom + card” example is not a 101 issue but, rather, an obviousness issue. Or some variant thereof.

  175. Malcolm, the quote states “software+computer”.

    Also, you abuse the word “new”. You “new brain” and “new computer” use “new” in the identity sense. A person might have a “new” outlook on life. Are they a “new” person. Yes and no. A model might have a “new look”. Is the look new? Yes. Is a computer reconfigured by software a “new computer”, maybe not, does it have a “new state”? Undeniably. Is a brain with new information “new”? Yes and no. Is a computer with its gates reconfigured into a combination of states that has never existed in that computer or any other computer “new”. No, it’s the same object that people would refer to as “the computer” both before and after its state has changed. And Yes, it’s a “new” computer relative to state-configurations of all other prior computers, as provable by the externally observable new properties of the computer.

    A brain that processes new information can be physically altered. Information (actually, processing of) can lead to the growth of new axons, change in levels of neurotransmitters and receptors, etc. Physiological change in the brain in response to information is well known. Information leads to neural activity, and neural activity is like a feedback loop; the activity can physically and semi-permanently change the brain. When a brain sleeps at night and consolidates new informatin into the form of a long term memory, has the brain changed?

    The issue of whether a reconfigured reconfigurable machine is “new” is a language trap. In one sense of the word “new”, it is, in another sense of the word “new”, it is not.

  176. Thank you, Hierarchy, for taking the time to read & respond.

    I fail to see why people believe that because a computer-as-configured-by-software (cacbs) can be described with mathematics (e.g., Turing Machine) that somehow it should not be patentable.

    I’m not trying to build a consistent theory of what “should be”; I’m trying to build a consistent theory of “what is.” I’ve read various things now. Given what I’ve read I believe that the patent act, Benson, Flook, Diehr and Bilski are reasonably consistent. But I also think the underlying principles aren’t well defined. For example, although I think I understand the broad strokes of the Supreme Court decisions I don’t think the “machine or transformation” test in Bilski is a good summary of their intent. In particular, I think it’s possible to come up with examples that pass the test without making clear what is or isn’t patented (or patentable). For example, using the “effective methods” test a specific hardware implementation of the BCD algorithm in Benson would be (could be) patentable, but the BCD algorithm itself wouldn’t (couldn’t) be. Using the”machine or transformation” test a specific hardware implementation of the BCD algorithm would be (could be) patentable–but that wouldn’t describe why a different hardware implementation of the same algorithm wouldn’t infringe. And yet, that’s exactly what Benson was trying to prevent: the patenting of the algorithm itself.

    We can describe or model any physical thing/process with math and physics.

    I think the “effective methods” test makes that line clearer. A model implemented in software would not be patentable. But of course a model implemented in software can’t effect any change in the real world, either. The attempt to affect the real world would cross back into the physical and run the potential of patent infringement.

    But let’s make this more concrete (and more controversial). I can design and patent an 8086 chip. The patent will, without question, protect the chip itself. By my test (and theory), however, I can build an emulator of that chip entirely in software–and the emulation would not infringe on the patent. (Again, I’m leaving aside any copyright issues on the microcode.)

    I know you’d disagree. But I think my theory reaches consistent conclusions about where the line should be drawn–and I think the the most obvious alternate theory reaches contradictions. For example, if you believe that a chip can be patented, and if you believe that the patent covers the essence of what the chip does in all forms, then I think you believe that an algorithm can be patented in all forms. (After all, any algorithm can be reduced to hardware.) That conclusion contradicts Benson, and given that I’m trying to develop a theory that’s consistent with the Supreme Court decisions that’s a contradiction.

    See, you analyze software and only mention some of its properties, but not others. A thing can have many characteristics. The existence of one characteristic (susceptible to math/algorithmic description) does not negate other characteristics (real, physical, useful, “technical”, etc.)

    Let’s look at it another way. Suppose we just put a curtain around the “brains” part of your loom. No one can see how it is changed to operate in different ways. If someone said, look, I can configure this loom to do new unobvious useful and wonderful things, would the public want to know what is behind the curtain? Would the public really care about the specific details of how the loom is configured (i.e., it can be reconfigured by a card)? Would it be helpful if people disclosed the contents of their cards so that others could learn, improve, the contents of cards in general?

    One more point. The loom+cards versus loom(alone) is a distinction without a difference. If I built lots of looms, and instead of building looms that work with cards I just built many different “hard” looms, each a machine with slightly different parts arranged in slightly different ways, and each of these specialized looms was functionally equivalent to one of your looms+specificCards, why would we treat them differently for patent purposes? Why would we encourage and reward the inefficient wasteful building of “hardwired” looms, and yet treat the card-configured looms as not worthy of being owned by the ones who create the cards? Why would we only reward loom builders and not the people who make them even more useful than when they leave the factory?

    You wrote your original response before I got to the third part of my argument, the one that described my proposed test. I think I covered exactly this point in my last paragraph: I think “effective methods” help to define where, how and why that line should be drawn. As I said, effective methods are effective methods regardless of form.

  177. There are many many things a computer+software can do that no brain can do

    And vice versa.

    There are many things a human with a library of books can do that no human can do. Can you patent the books? Can you patent the person with the books? How about an old robot with new books?

  178. My statement, which is true, is that people are being conveniently (and in a way dishonestly) selective about properties of software+computer when they make arguments about why it should not be eligible.

    Software is instructions (not patentable, just like instructions that can be processed by a human brain are not patentable) read by a computer (not changed into a “new” machine by reading, just like brain not a “new” brain when new information added).

    Who’s being selective and in a way dishonest?

  179. This outcome seems perfectly logical under Bilski. For an analysis on why such a finding is expected under Bilski, see “Predictability and Patentable Processes: The Federal Circuit’s In re Bilski Decision and Its Effect on the Incentive to Invent” 11 Columbia Science & Technology Law Review 1 (available at link to papers.ssrn.com).

  180. Malcolm: “I just don’t see how people can simply ignore the bottom line. Computer+software is practical, real, physical, capable of being built in ways that are new and unobvious and useful.

    “I just don’t see how people can simply ignore the bottom line. Brains plus thoughts are practical, real, physical, and capable of being developed in ways that are new and unobvious and useful.”

    Yes, let’s patent everything. More patents means more stuff to invest in! It’s like creating wealth from scratch. How can this not work to everyone’s mutual benefit?”

    Malcolm, were you absent the day they taught common sense? Nowhere do I suggest that these things alone justify patent eligibility. There is a difference between taking something into account in making a decision, and using that thing as outcome determinative.

    My statement, which is true, is that people are being conveniently (and in a way dishonestly) selective about properties of software+computer when they make arguments about why it should not be eligible. These properties I’ve mentioned are factors that can and should be taken into account and not ignore (after all, they’re factors courts have used).

    We could go into much greater detail about how computer+software is qualitatively different (in ways relevant to patentability) than a legal status, a brain, an abstract algorithm, etc, but that’s been done enough times and the differences are pretty obvious. There are many many things a computer+software can do that no brain can do, and which are exactly the types of things and benefits that patents were traditionally aimed to encourage.

  181. “but first I’d like to see if there’s any common ground to build from.”

    By doing what, exactly?

    By creating an example that starts from something basic and builds to a conclusion. No doubt there will be disagreement between the first step and the last; I’d be interested in seeing where that disagreement lies and what’s behind it.

  182. Michael: “I believe that “Effective Methods” (or “Effective Procedures”) are not patentable. An “effective method” is a term of art from computation theory, and basically it’s a finite series of steps that can be carried out by a person with no more than pencil and paper. (There’s a little more to it–but not much.) All algorithms are “effective methods” and all “effective methods” are algorithms.”

    The law is not interested in computer science theory. How about this. I’ll limit my so-falled software claims (computer configured to implement a particular algorithm) to only implementations where every step is performed by a computer. I won’t claim any abstract algorithms. Now you are free to use pencil, paper, brain, speech and conversation, or anything else other than a computer to practice the algorithm. You can publish it in a book. You can discuss it in computer science class. You can even build an assembly line structured to operate in a way that map to the algorithm.

    Practically, you can code it and study it in private and as long as it is only for private study no one will know or care, though technically you would be infringing. It would be nice if there was some kind of “fair use” exception in patent law like there is in copyright law …

  183. “Why would it be different from determining if an element of an accused structure is in fact a “wheel” as recited in the claim? It seems like it’s the same thing, only harder to do. ”

    Because, Red Monkey, there’s a difference between a legal opinion and a legal conclusion; a binding decision by a judge, agency, etc. If you have a claim that says “a legall binding contract”, arguably, the only way to know if there is infringement is to get a decision in court and exhaust all possible appeals. Then you can truly say it is “legally binding”. There are things that look like contracts, there are informal non-binding agreements, there are contracts that look good but for some policy reason they can be held unenforceable, etc. Even the easier alternative of, during infringement trial, proving the legal requirements for a contract have been met, seems tricky.

    So, back to infringement. If you have a claim and say I infringe because my process formed a contract, and I go and say, no, it’s not a contract, the burden on you, the patent holder, is to prove infringement. You must prove that I have practiced each element (ignoring doctrine of equivs). You must prove that I have formed a legally binding contract. Do you have to get a declaratory judgement? Do you have a “trial within a trial”, where you attempt to convince the fact finder that the requirements for a legally binding contract are met? This scenario, when you think it through, can get very strange. What if your claim has an element for “a patent”? Is that a valid patent? A paper patent that is expired?

    I understand that questions of infringement must be answered for “technical” limitations also, it just seems that there are too many problems with determining if “legal” limitations (e.g., legal relations in the form of rights/duties, legal categorizations, etc.) are met by an infringer.

    Here’s some perspective. If you want to know if an agreement is a legally binding contract (a simple yes/no question), there may be thousands of court decisions, factors like jurisdiction, and all the other things that bear on the issue. If you want to know if someone is producing a claimed “widget” or is “manufacturing a board with a convex shaped side”, the issue is (assuming Markman has occurred) mostly depends on facts and English language, which the judge/jury is presumably able to evaluate.

  184. I just don’t see how people can simply ignore the bottom line. Computer+software is practical, real, physical, capable of being built in ways that are new and unobvious and useful.

    “I just don’t see how people can simply ignore the bottom line. Brains plus thoughts are practical, real, physical, and capable of being developed in ways that are new and unobvious and useful.”

    Yes, let’s patent everything. More patents means more stuff to invest in! It’s like creating wealth from scratch. How can this not work to everyone’s mutual benefit?

  185. Let’s suppose someone comes along and figures out that if you use the knobs in a certain order and with particular settings, the saw suddenly performs the same task twice as fast, with cleaner and more reliable cuts. This person spent a year working on his technique of using knobs. Everyone saw operator agrees that no one thought of this before and the outcome of the result is surprising. Should society wish him to keep his method secret, and it is only used in one saw mill and the rest of the world continue to waste trees etc. at a higher rate?

    Problem with hypo: nobody is arguing that methods of transforming wood aren’t patentable under 101.

    Better strawmen, please.

  186. Finally, I’d like to propose a test for what is patentable–and some implications of that test.

    I believe that “Effective Methods” (or “Effective Procedures”) are not patentable. An “effective method” is a term of art from computation theory, and basically it’s a finite series of steps that can be carried out by a person with no more than pencil and paper. (There’s a little more to it–but not much.) All algorithms are “effective methods” and all “effective methods” are algorithms.

    A patentable machine may certainly include an effective method, but I believe that a machine which consists *only* of an effective method cannot be patented.

    I think that’s a clearer boundary than the “machine or transformation” test. After all, any effective method could be completely reduced to a physical machine. But any physical machine that can be completely reduced to an effective method shouldn’t be patentable in the first place. Otherwise you’d be patenting things that a human could do using no more than pencil and paper. (Again, we’re talking about the effective method, not the hardware of the device itself. Looms are patentable; we’re talking about the cards here.)

    A corollary to this is that any part of a patentable machine which is an effective method is not covered by the patent itself. That’s like saying that a patentable process may take advantage of a newly discovered law of nature but may not patent the law itself.

    To bring this full circle, that’s why cards aren’t patentable: they represent an effective method, something a human could “do” using no more than pencil and paper. (The “do” here is the specification of a pattern, not the production of the weave.) And I think this also makes the “loom + specific cards” question clearer. You could, of course, produce a loom that “hides” the cards within a physical implementation inside the loom. In that case the “card” part of the patent can still be extracted and used without infringement. (We’re talking about patents here, not copyright. That’s another story.) An effective method is an effective method regardless of its physical form.

  187. “The analysis for determining infringement takes on a whole new world if it requires determining a legal status”

    Why would it be different from determining if an element of an accused structure is in fact a “wheel” as recited in the claim? It seems like it’s the same thing, only harder to do.

  188. “the loom is patentable and the combination of loom + specific cards is not”

    How can adding an element to a structure that falls within the ambit of 101 somehow remove it from the ambit of 101? I would think if loom is within 101 then loom+anything is also within 101.

    Also, we shouldn’t say “patentable” so loosely. “Patentable” means a claim passes 101, 112, 102 and 103. I think ML’s post uses “patentable” to mean passes 101.

  189. Michael, just a final thought. Why is a card different than a nob? What if I think of that card as just another physical “part” of the machine, like a special kind of cog. I put in a new cog, the machine works better. I put in a new card, the machine works better. Again, the distinctions that matter should be practical ones, not theoretical ones. Here, there is no practical difference as far as the usefulness of the machine/loom is concerned.

  190. “but first I’d like to see if there’s any common ground to build from.”?

    This is like a scientist trying to find common ground with a creationist by getting him to admit that burning bushes can’t talk.

    No one would say that turning the knob to adjust the angle creates a new separately patentable machine–would they?

    My friend, stick around and you’ll find that people here will say literally anything to defend their position. It’s pretty much a habit for attorneys.

  191. Michael,

    “But now imagine a table-mounted circular saw where the angle of the saw is adjustable. (In other words, the saw can cut wood at different angles.) No one would say that turning the knob to adjust the angle creates a new separately patentable machine–would they? What if the angle were set by holes punched into a card…?”

    Let’s suppose that the saw has a bunch of knobs. Lets suppose these saws have been the same for 100 years; no changes. Let’s suppose that everyone has always used the knobs the same way. Let’s suppose someone comes along and figures out that if you use the knobs in a certain order and with particular settings, the saw suddenly performs the same task twice as fast, with cleaner and more reliable cuts. This person spent a year working on his technique of using knobs. Everyone saw operator agrees that no one thought of this before and the outcome of the result is surprising. Should society wish him to keep his method secret, and it is only used in one saw mill and the rest of the world continue to waste trees etc. at a higher rate? Should we encourage him to share his technique and give him a limited right to exclusively use his method? Should society make it known to others that it can be rewarding to work on finding better ways to use these unchanged and old machines?

    I fail to see why people believe that because a computer-as-configured-by-software (cacbs) can be described with mathematics (e.g., Turing Machine) that somehow it should not be patentable. We can describe or model any physical thing/process with math and physics.

    See, you analyze software and only mention some of its properties, but not others. A thing can have many characteristics. The existence of one characteristic (susceptible to math/algorithmic description) does not negate other characteristics (real, physical, useful, “technical”, etc.)

    Let’s look at it another way. Suppose we just put a curtain around the “brains” part of your loom. No one can see how it is changed to operate in different ways. If someone said, look, I can configure this loom to do new unobvious useful and wonderful things, would the public want to know what is behind the curtain? Would the public really care about the specific details of how the loom is configured (i.e., it can be reconfigured by a card)? Would it be helpful if people disclosed the contents of their cards so that others could learn, improve, the contents of cards in general?

    One more point. The loom+cards versus loom(alone) is a distinction without a difference. If I built lots of looms, and instead of building looms that work with cards I just built many different “hard” looms, each a machine with slightly different parts arranged in slightly different ways, and each of these specialized looms was functionally equivalent to one of your looms+specificCards, why would we treat them differently for patent purposes? Why would we encourage and reward the inefficient wasteful building of “hardwired” looms, and yet treat the card-configured looms as not worthy of being owned by the ones who create the cards? Why would we only reward loom builders and not the people who make them even more useful than when they leave the factory?

    I just don’t see how people can simply ignore the bottom line. Computer+software is practical, real, physical, capable of being built in ways that are new and unobvious and useful. Why is the reconfigurable nature of the computer’s logic a strike against it? To me, your reasoning defies logic.

    [PS, Jules, if an attorney has spent more time practicing software arts than practicing law, are they an attorney or are they a software engineer/designer/developer? If such a person wrote this blog post, is it an attorney beating down Michael, or is it someone else?]

  192. Is a ice melting machine 112, 2 just because it doesn’t work in Florida but does in Michigan?

  193. That is the court’s way of trying to get business methods out of 101 without taking out all computer methods.

    Pretty lame really.

  194. Is “price” tangible? Does price equal to money? Price is definately not a legal obligation, though one may argue it is an abstract concept.
    So money is not a price? But when I paid for the price using money for something…..

  195. that property must be in a certain fluctuated state shall we say for all the limitations to be met

    Rather difficult to discuss in the abstract, 6, so you’ll pardon me for flailing. It sounds as if there is a step of controlling the temperature that is critical for obtaining the recited device. If so, the step should be recited in the claims, no? If not, then we are back to square one where it’s difficult to answer your question without more info.

  196. “the loom is patentable and the combination of loom + specific cards is not”

    This seems more than a bit silly. What’s the point of this exercise again?

  197. Someone Von Neumann invented a short hand for a machine. That was a CPU plus instructions in memory rather than build the entire machine.

    The computer you use is CPU + software.

    6, 103 arguments come almost directly from just signing over a check and getting partial money from a check cashing outfit.

  198. A machine that processes information and can do the work that people are paid for. How can that not be eligible for patentability.

    It was eligible. Who invented the computer anyway?

  199. So, if I build a robot that can infringe other machine patents by its actions, then how can that be? My robot not being eligible for patentability (or assume the physical parts have patent terms have expired.)

    My robot could perform any method and how would you allege infringement?

  200. A machine that processes information and can do the work that people are paid for. How can that not be eligible for patentability.

    Hard for people to get their heads around that.

  201. “Sounds like a product-by-process claim or at least a product-by-process issue and by definition that means a lurking 112 problem. If I can only infringe the “invention” (i.e., what was invented, not the gobbledygook in the claim) if the method is carried out at temperature X, then temperature X should probably be recited in the claim unless one skilled in the art would never consider other temperatures to be relevant.”

    It isn’t product by process, the same structure is formed whether or not it is formed in alaska or in florida. There are properties of the product that fluctuate according to the temperature around the device (even after it is formed, although also during formation) though and that property must be in a certain fluctuated state shall we say for all the limitations to be met.

    I don’t know why you guys are going on and on about using 103 in the OT case. 103 may very well be hard to apply unless you have some evidence at your fingertips that you aren’t sharing.

  202. 4. Building on the previous post, by the Church-Turing thesis, *any* computer algorithm could be implemented as a series of cards fed into the loom.

    I propose:

    4a. The fact that the cards can be interpreted as an algorithm and the weaves as output of the algorithm changes none of the above statements: the loom is patentable, the cards are not, and the combination of loom + specific cards is not.

    I don’t know whether or not any of this is surprising, but it is completely uncontroversial from the perspective of computing theory. Indeed, Turing machines–which are capable of executing any algorithm–are simpler than the loom I’ve described.

    Note that we aren’t talking about the loom itself–how fast it is, how much it can “remember”, how easy it is to define the input or interpret the output; those things are all in the realm of the physical, and those devices would all (I think) be potentially patentable. No, we’re just talking about the information contained on the cards.

    By contrast, if you believe that algorithms should be patentable you must accept that cards fed into a loom to produce a weave should be patentable–or, at the very least, that the combination of loom + a specific set of cards should be patentable.

    But now imagine a table-mounted circular saw where the angle of the saw is adjustable. (In other words, the saw can cut wood at different angles.) No one would say that turning the knob to adjust the angle creates a new separately patentable machine–would they? What if the angle were set by holes punched into a card…?

  203. Hello, and welcome Michael.

    I wonder if your position is the unpatentability of software. Is there any room for the law in your discussion? There’s a lot out there, but you might start with 35 USC 101.

    “but first I’d like to see if there’s any common ground to build from.”

    By doing what, exactly?

    Best of luck with your endeavors. Don’t let the attorneys here crush your spirits.

  204. Hello, all.

    I’ve been reading up on patent law, especially as it applies to software. I’d like to propose a test for what should or should not be patentable–together with some implications–but first I’d like to see if there’s any common ground to build from.

    1. Imagine a loom (a Jacquard loom) that produces output based on cards fed as input.

    I propose:

    1a. The loom is patentable.
    1b. The cards are not patentable.
    1c. The combination of a loom with a specific set of cards is not patentable.

    2. Imagine that the loom produces output based on cards fed as input, but imagine too that the operator can vary the output by making some additional inputs via levers, buttons and knobs. For example, the user might push a button to choose the color(s) of the pattern, or turn a knob to choose the thread count of the pattern.

    I propose:

    2a. The fact that the user can modify the output within certain parameters changes none of the above statements: the loom is patentable, the cards are not, and the combination of loom + specific cards is not.

    3. Imagine that the loom has the ability to record some state, and to act on that recorded state. For example, there might be a way to say: “after three stitches of color 1, do two stitches of color 2.” The mechanism for recording might be moving a ball bearing from one bin to another, or it might be punching a hole in a piece of paper, or it might be turning a wheel.

    I propose:

    3a. The fact that the loom can record some state changes none of the above statements: the loom is patentable, the cards are not, and the combination of loom + specific cards is not.

    That’s the foundation I’d like to build from in one or two subsequent posts.

  205. As an attorney for whose clients this is a live issue, the interesting thing about this case is that there is a chance to compare claims that were and weren’t decided to satisfy the MOT test.

    Compare the failing ‘862/1 claim (in the original post) with the passing ‘829/1 claim:

    A method of providing at least a portion of an income tax refund amount to a taxpayer through a spending vehicle, comprising:
    arranging, among an electronic tax preparation system and a third party spending vehicle provider, to offer at least one spending
    vehicle to said taxpayer, said spending vehicle redeemable only at a retailer designated by the arrangement;
    preparing a tax return by or for said taxpayer using said electronic tax preparation system;
    obtaining tax return data from said electronic tax preparation system, said tax return data comprising an anticipated income tax
    refund amount from said taxing authority payable to said taxpayer;
    selecting by said taxpayer said at least one spending vehicle from a plurality of spending vehicles offered by said third party
    spending vehicle provider;
    executing at the time of preparation of said tax return using said electronic tax preparation system an agreement related to said
    spending vehicle in which said taxpayer agrees to assign at least a portion of said income tax refund amount to said third party spending
    vehicle provider and said third party spending vehicle provider agrees to accept said assignment and provide said selected spending vehicle
    to said taxpayer and to accept an associated risk for non-payment of said portion of said income tax refund, wherein said agreement does not involve issuing a loan or credit to said taxpayer and wherein said taxpayer and said spending vehicle provider obligations occur upon submission of said tax return to said taxing authority;
    transferring electronically said at least a portion of said income tax refund amount to an account for said third party spending
    vehicle provider; and
    issuing said selected spending vehicle to said taxpayer, said selected spending vehicle having a prepaid value greater than said assigned at least a portion of said tax refund and said taxpayer
    redeems said selected spending vehicle at said designated retailer to purchase products or services.

  206. NWPA–

    Yes, 103 instead of 101.

    Big Guy–

    If the claim wasn’t so ridiculously drafted, I would say “process”. As it is, who knows? And who cares? Why discuss 101 when 103 should be so incredibly easy? The claim could be written properly as a process, which makes using 103 instead of 101 all the more attractive.

  207. 6: I’m about to allow a claim that is a method of making a device wherein a limitation of the properties of a device’s structure depends on a wherein clause that depends on the temperature of the surrounding environment.

    Sounds like a product-by-process claim or at least a product-by-process issue and by definition that means a lurking 112 problem. If I can only infringe the “invention” (i.e., what was invented, not the gobbledygook in the claim) if the method is carried out at temperature X, then temperature X should probably be recited in the claim unless one skilled in the art would never consider other temperatures to be relevant.

  208. Red Monkey,

    This is pure speculation, but even though negotiable instruments and cash are tangible, there is a legal dimension to them. If you have a claim that recites a “promissory note” type of negotiable instrument, there’s a legal qualification on that element. To determine whether one infringes the claim, one must determine whether a piece of paper is indeed a “promissory note”. Also, does that element mean something that has on its face the properties of a promissory note, even if it is executed by “Elmer Fudd” to “Daffy Duck”?

    I don’t know if this is why legal relationships are being maligned (in claims), but there is a problem with determining infringement. The analysis for determining infringement takes on a whole new world if it requires determining a legal status (e.g., is something a legally binding contract, if something is a “tax” or “tax refund”, etc.) of allegedly infringing activities/things.

  209. you can criticize this patent, but many of your software-related systems and methods are next in line

  210. 6. I don’t understand FL. But your example reminds me of forging steel. The temperature of the forge makes a different steel does it not? But, again, since I don’t understand FL, I can say nothing about your claim.

  211. “money is intangible? ”

    How ironic history can be. In the old days money was quite tangible. Then someone invented negotiable instruments because it was inconvenient to carry all that tangible money around, especially with all the bandits roaming around. Then came EFT, because those tangible checks kept getting lost in the mail. Fiunally, the tangible money that spawned all this has itself become intangible.

    I suppose this is what comes of dropping the gold standard at Bretton Woods.

  212. People don’t understand that the word “system” must be simply a fill in word for product, apparatus, comp o matter, or process. Ban the word imo, it does nothing but confuse people.

    In any event, here’s a little story that’s kind of funny.

    I’m about to allow a claim that is a method of making a device wherein a limitation of the properties of a device’s structure depends on a wherein clause that depends on the temperature of the surrounding environment. The claim is definite it seems to me like, but it is really a grey area. You could perform the method in a cold environment and the properties of the device formed would not be sufficiently the same for the method to infringe, whereas you could perform the exact same method in FL on a hot day and the device formed would be the same exact device as you made in the other environment but you would infringe because the properties of the device would then meet the claim limitations.

    Then if you took the device made in the cold environ to FL on a hot day it would then have sufficient properties to meet the structural claim limitations of the device being formed within the method claim.

    If anyone can follow what I’m saying, speak now if you think it would be a 112 2nd concern. The case goes out in a few days.

  213. From the recommendation:

    [regarding the ‘862 patent:] “The computer is a component of the claimed system that simply stores and retrieves data associating a payment with a spending vehicle.1 The remaining components of the system are parts or elements capable of holding and assigning rights, offering spending vehicles, and receiving payments. In light of the specification, it is clear that these “components” are individuals and organizations. Thus, the ‘862 claims describe a “system” consisting of a computer and certain financial relationships between entities.”

    This seems consistent with the preamble, which recites a “computerized system”. What does that mean? I don’t think it means a “computer system”, because if it did, that’s what it would say. It leave open the possibility that some parts are not “computer”. This is probably what H&R intended with the claim.

    Same with the ‘452 patent; “A computerized system…”. The last two elements are to “a loan” and “an income tax refund”. Other elements recite things that are “data” “in a/the computer”. But the last two elements omit this, and the implication is that the loan and the refund are not data and are not in a computer.

    The claims of the ‘829 patent PASSED the MOT test: “The Court finds the claims of the ‘829 patent are sufficiently tied to a particular machine and therefore pass the machine-or-transformation test. Accordingly,
    the claims are directed to patentable subject matter.”

    On the practical side, a preamble that a method is “computer implemented” does not require that it be implemented in the entirety. A single action by a computer would allow one to fairly say it was implemented by a computer.

    To me, the focus on the financial or legal relations in the claims is just a return to the ban on so-called business methods.

  214. “101 is not the right means to invalidate this claim. This should be invalid under 103–not 101.”

    Seriously? Which of the four statutory categories does this claim fall under?

  215. “Indeed. Shame on the examiner for allowing such poppycock.”

    I checked the file history. To the examiner’s credit, she tried to kill it under 101. The applicant added all the language involving the computer, and argued that adding “technology” to the claims made it patentable subject matter. At some point, I suppose, if the applicant wants a cr@p patent, give them a cr@p patent. At least this one appears to be on its way to invalidation at summary judgment. Us defendants love those issues of law…

  216. The patentee might as well have called this a “paradigm.” I don’t see how elements like “a payment due” or “an assignable right” define either structural features in an apparatus or steps in a process. You don’t need Bilski to put this abomination to rest.

    Indeed. Shame on the examiner for allowing such poppycock.

  217. 101 is not the right means to invalidate this claim. This should be invalid under 103–not 101.

  218. The “computerized system” claims did not identify a “particular, special purpose machine” that was more than “an insignificant, extra-solution component of the claimed invention.” Likewise, the transformation of “tax return data” into a “spending vehicle” is not the type of transformation required under the test.

    Credit where credit is due: this is how the analysis should work, Bilski or not. Intangible, abstract garbage belongs in the trash can.

    BigGuy: The patentee might as well have called this a “paradigm.” I don’t see how elements like “a payment due” or “an assignable right” define either structural features in an apparatus or steps in a process.

    B-b-b-ut it hh-has a c-c-computer!!!!

    /Diehrbot off

  219. “Quoting Nuijten, the court first held that Bilski controls both system and method claims – since a “court should not be ‘overly concerned with pigeonholing subject matter once the court assures itself that some category has been satisfied.'”

    Say what?

    The patentee might as well have called this a “paradigm.” I don’t see how elements like “a payment due” or “an assignable right” define either structural features in an apparatus or steps in a process. You don’t need Bilski to put this abomination to rest.

Comments are closed.