I’ve finally watched the excellent video “Patent Absurdity: how software patents broke the system”, and I highly recommend it. It contains several interviews surrounding the Bilski case in the U.S. Supreme Court, with commentary from some of the biggest names in the software patent debate. I will let the video speak for itself, it really is an excellent documentary. I will just mention that the Bilski patent application is for a mathematical method of hedging risks in commodities trading, which was rejected by the USPTO.

I just wanted to highlight one specific section, which to me is the most concise and well-stated demolition of various software and process patents.  In one instance, one of the attorneys for Bilski was asked whether their “invention” was just “picking up the phone and calling other people”. The attorney responds that it could be reduced to that, but it is much more. He says that:

“If you look at Claim 4 (we have these things called claims which describe what the patent is for), there is a long mathematical formula in there that didn’t exist in nature or anywhere in the literature, that these very inventive folks came up with.”

Really? Let us forget for a moment the wider and important question of whether or not it is a good idea to patent mathematical formulae (unsurprisingly, I think it is a very bad idea). Is it true that Bilski’s patent application discovers a new formula? Apparently not.

Ben Klemens, the author of Math You Can’t Use (which I own but haven’t read, and has now moved up my must-read list), has different ideas. He is the one who I think thoroughly destroys several math-related patents. He explains that a the heart of writing software there must be an algorithm that explains the process that will produce a result. These algorithms can take the shape of mathematical formulae, and you just change the variables according to what you want to achieve. Klemens explains for example that there is a common formula in linear algebra called the singular value decomposition (SVD), which is an “important factorization of a rectangular real  or complex matrix, with many applications in signal processing and statistics.” In plain English, it is a tool that can be used to describe various processes in mathematical terms. Klemens explains that it is possible to assign all sorts of variables to the equation, and goes ahead to prove his point. He breaks down the variables in a given SVD and assigns them different values:

We have the matrix:

x1 = sexuality
x2 = cats
x3 = affection

And then the vectors:

J1 = Jane’s responses
J2 = Joe’s responses

Then you substitute the values and enter them into the SVD equation, and you have just derived U.S. patent 6,735,568 which protects a “Method and system for identifying people who are likely to have a successful relationship”, eHarmony’s infamous patent.

Klemens has in one step completely demolished a patent that should have never been awarded, and the worst part of it is that he claims this happens all the time, that there are clever patent attorneys and patent trolls out there who are pretty much conning a mathematically-illiterate system by taking well-understood mathematical equations, assigning values to the variables, and filing for a patent.

For anyone who supports business method patents, this is a point that requires reply. There are thousands of patents out there that are nothing more than a clever mathematical con. Let us hope that Bilski delivers the right result.


69 Comments

Avatar

Les · April 20, 2010 at 9:34 am

Here is the reply you require.

If a patent had been granted for a method for determining the amount of energy that is releasable from a given mass, the method comprising:

determining the mass; and

multiplying the mass by the speed of light squared.

the patent would not be for maths but for a particular method of determine a particular value.

The patent would not cover all instances of squaring one number and multiplying it by the other. The wild haired professor did not just stumble upon a=bcc and "assign variable names to it". To assert otherwise is simply absurd.

Similarly, eHarmony didn't get a patent for all uses of SVD. Instead they got a patent for determining a measure of compatibility. You are free to use SVD for other things and you are free to use other measures of compatibility.

Just as the Wright Bros. applied known wood, cloth and cables to arrive at a new invention, eHarmony may have applied known elements to solve a problem in a new and perhaps better way.

You can argue that eHarmonies method is obvious to mathematicians…but that doesn't mean ALL software is not worthy of a patent. Furthermore the test is not whether a technique is obvious to mathematicians, the test in this case would probably be was it obvious at the time to ones of ordinary skill in the art of matchmaking…

    Avatar

    Andres · April 21, 2010 at 6:44 am

    Hi Les,

    Thank you, this is a very thoughtful response. I disagree however. For example, I completely agree that usually it is the specific application of the equation what is being protected. So, e=mc2 is not subject to a patent, but the nuclear reactor that can be built can.

    However, what I see happening with various business method patents (and some software ones) is different to that. These patents take directly an algebraic formula which describes processes, and then simply substitute the variables. This is both obvious and not-novel. Patent law was not designed to protect every substitution in an algebraic formula, where is the inventive step?

    But more important than that, what does society gain from awarding such a monopoly? The eHarmony patent does nothing but impose a competitive restriction on competitors who may be doing similar stuff, because when we boil it down, there are not many other ways of addressing compatibility.

Avatar

les · April 22, 2010 at 1:21 am

Andres –

Please cite the particular examples you refer to when you say,". These patents take directly an algebraic formula which describes processes, and then simply substitute the variables. "

Remember, as I pointed out in the Einstein example, a claim too:

a method for determining the amount of energy that is releasable from a given mass, the method comprising:

determining the mass; and

multiplying the mass by the speed of light squared.

is not simply assigning variable names to a=bcc. The inventive step is in realizing that THIS particular formula is the right formula for solving this problem and that the speed of light squared is somehow involved it determining the answer.

This is also where the inventive step is in the eharmaony example and it is likely where the inventive step is in in any other example you might find.

    Avatar

    Andres · April 23, 2010 at 7:46 am

    My pleasure. I hope you are asking only for examples, it would be impossible to point out every single patent that misuses algebra. I’ll list some. The first algebraic patent is our old friend State Street (5,193,056). There are several blatant Fourier transform patents out there: 5,835,392; 6,055,556; 6,434,582; 6,640,237. There are some in trivial uses of Java virtual machines (6,918,122); and there is a recent blatant misuse of SVD in a Netflix patent application (20070150428).

    Your Einstein patent example is describing something entirely different to the patents described. You adequately make the distinction that what is being patented in that instance is not the equation, but the specific application. This is fine, and there are many such physical examples. The difference with SVDs and other algebraic formulae is precisely the triviality of the substitution, there is nothing novel or inventive about eHarmony’s patent, nothing at all, it is simply a trivial substitution of variables.

    We will have to disagree as to what constitutes an inventive step. If you consider that simply going through a formula, finding out that it fits a process, and changing some variables is worthy of a monopoly for 20 years, then we have conflicting views about what is patentable and what is not. Your examples reduce “inventiveness” to clever lawyerly weaselling.

Avatar

Les · April 23, 2010 at 4:26 pm

Well, you needn't worry about 2007/0150428. That APPLICATION went abandoned in February after being rejected by the patent office.

    Avatar

    Andres · April 24, 2010 at 5:10 am

    Why use caps on the word "application"? It is quite clear that I did say it was an application.

    I am glad the application was rejected, that means that someone at the USPTO is still doing their job. But doesn't that prove the point about the invalidity of the eHarmony patent? They are very similar applications of SVD forumulae, so the fact that the Netflix one was denied does not fare well to your argument that SVD and other algebraic patents should be accepted.

    Before I continue, you seem to be involved in a Gish Gallop. I followed your engagement in the End Software Patent News site, and you seem to have a lot of time on your hands. I have to warn you that I will be away most of next week if this is going to continue.

    You seem to be looking for mentions of the eHarmony patent anywhere online and posting to defend it. Would it be fair to assume that you are somehow related to that particular patent, or with eHarmony? I'm just curious.

Avatar

Les · April 23, 2010 at 4:37 pm

In the broadest claim 5,835,392 claims:

1. In a computer system, a method of performing a butterfly stage of a complex fast fourier transform of two input signals comprising the steps of:

a. performing a packed multiply add on a first packed complex value generated from a first input signal and a set of trigonometric values to generate a first product;

b. generating a second product which comprises said first product with a sign inverted, wherein said step of generating said second product further includes packing said first product and said first product with said sign inverted into a single source; and

c. performing a packed add of said second product and a second complex value generated from a second input signal to generate a first result, and performing a packed add of said first product and said second complex value to generate a second result.

You're going to have to explain to me what that's obvious. It certainly doesn't cover all FFTs. I don't know what a packed complex value is. But in light of this patent, you are still able to perform FFTs with unpacked complex values.

Furthermore, you are allowed to even use packed complex values as long as you don't:

generate a second product which comprises said first product with a sign inverted, wherein said step of generating said second product further includes packing said first product and said first product with said sign inverted into a single source.

Still further you can generate the second product that way, just as long as you don't:

perform a packed add of said second product and a second complex value generated from a second input signal to generate a first result

and you can even do that, as long as you don't also

perform a packed add of said first product and said second complex value to generate a second result.

If you ask me, this is one butterfly living in the high plateaus of the Andies that no one has ever seen.

As for the rest….don't worry…Butterflies are free to fly….fly away, high away…

    Avatar

    Andres · April 24, 2010 at 5:22 am

    You seem to have moved the goalposts completely while I was not watching. My argument is, and always has been, that these patents should have never been awarded because they describe trivial algebraic formulae, or try to protect processes which have already been disclosed to the public, and therefore cannot be patented because of prior art.

    What you seem to be saying here, and in the rest of the other comments, is that these are rare and unique patents, and therefore there is no problem that they were granted at all.

    Why is it pro software and business patent defenders always fall back to this clearly illogical and faulty argument? "Oh, THAT patent is fine because nobody will ever infringe it, it is too rare". Then why apply for a patent in the first place? Why go to the trouble of ask for monopoly protection for something as trivial and unique as this?

    The core of the argument, which you seem to have ignored in favour of trying to bury me in copied patent claims, is that these patents are not inventions. They are not novel. They do not advance the knowledge of humankind. They do not promote innovation. That is why we have a patent system.

    If you can tell me how taking fourier transforms and turning them into patents by clever patent attorneys advances innovation, I will let this go.

Avatar

les · April 23, 2010 at 6:48 pm

In 6,055,556 The broadest claim is:

1. A method for multiplying a first matrix {A} with I*N complex elements a(i,n)=(Re(a(i,n)), Im(a(i,n))) and a second matrix {B} with N*J complex elements b(n,j)=(Re(b(n,j)), Im(b(n,j))) to a third matrix {C} with I*J complex elements c(i,j)=(Re(c(i,j)), Im(c(i,j))),

wherein i=1 to I, j=1 to J, and n=1 to N

wherein, a, b, c are complex numbers

said method comprising the following steps:

(i) loading Re(b(n,j)) and Im(b(n,j)) for all n and for all j into a memory array;

(ii) loading Re(a(i,n)) and Im(a(i,n)) for one i and for all n into registers of a register array;

(iii) simultaneously for all j and for all n providing intermediate results

f.sub.1 (i,j,n)=Re(a(i,n))*Re(b(n,j)) by first multiplication units,

f.sub.2 (i,j,n)=-Im(a(i,n))*Im(b(n,j)) by second multiplication units,

f.sub.3 (i,j,n)=Re(a(i,n))*Im(b(n,j)) by third multiplication units,

f.sub.4 (i,j,n)=Im(a(i,n))*Re(b(n,j)) by fourth multiplication units;

(iv) simultaneously for all j and for all n providing intermediate results

ReSum (i,j,n)=f.sub.1 (i,j,n)+f.sub.2 (i,j,n) by first adders

ImSum (i,j,n)=f.sub.3 (i,j,n)+f.sub.4 (i,j,n) by second adders;

(v) simultaneously for all j accumulating over all n said intermediate results

ReSum (i,j,n) to Re(c(i,j)) by first accumulators and

ImSum to Im(c(i,j)) by second accumulators while repeating said steps (ii), (iii), and (iv); and

(vi) repeating said steps (ii), (iii), (iv) and (v) for all i.

This is a very narrow claim. To get around it, all you have to do is not use registers or not use memory. To violate this patent you have to among other things:

(i) loading Re(b(n,j)) and Im(b(n,j)) for all n and for all j into a memory array;

(ii) loading Re(a(i,n)) and Im(a(i,n)) for one i and for all n into registers of a register array;

I think the odds of doing that accidentally are pretty slim. Does anyone even write machine code anymore? How would you even specify memory for one and registers for the other in C++ for example? (that's rhetorical, I don't really care).

    Avatar

    Andres · April 24, 2010 at 5:28 am

    Again you seem to be missing the point. Whether this patent is narrow is irrelevant to the point I am trying to make.

Avatar

les · April 23, 2010 at 6:55 pm

The broadest "software" claim in 6,434,582 claims:

11. A method for computing the cosine of an input value, the method comprising:

performing a plurality of AND operations on bits of the input value, wherein each of the AND operations operates on two or more bits of the input value;

generating N output operands, wherein at least one of the output operands includes (a) at least one output bit which is set equal to an input bit, and (b) at least one output bit which is set equal to the result of one of said AND operations, wherein N is greater than or equal to two;

adding the N output operands to generate a single resultant value which represents the cosine of the input value.

I'm sure there are plenty of ways to calculate a cosine that don't involve this very specific method of anding. You are REALLY going to have to explain why THIS is an obvious way to calculate cosines….

You are still allowed to use adjacent over hypotenuse…if the above is some very efficient method…shrug…so be it…. he invented it, he's entitled to it. deal with it.

    Avatar

    Andres · April 24, 2010 at 5:29 am

    And again you moved the goalpost. The narrowness and/or uniqueness of the patent is irrelevant to the point being made.

Avatar

les · April 23, 2010 at 7:07 pm

This looks like the broadest "software" claim in 6,640,237.

1. A method for determining a trigonometric value for an angle, comprising:

receiving an input angle;

separating the input angle into a first component angle and a second component angle at an angle separator;

establishing one or more trigonometric values of a set of trigonometric values associated with the first component angle and the second component angle, at least one trigonometric value established according to a procedure selected from a group consisting of:

comparing the second component angle to a predetermined threshold to determine the at least one trigonometric value; and

determine the at least one trigonometric value at a sine beta device;

generating one or more intermediate values of a set of intermediate values at one or more multipliers in accordance with the one or more trigonometric values; and

determining an output trigonometric value of a set of output trigonometric values at one or more adders in accordance with the one or more intermediate values.

One again, I don't see what you're worried about. If it works at all, its one very specific what to determine a trig value. Memory is cheap. If I were you I'd just store the tables and look 'em up.

Avatar

les · April 24, 2010 at 6:28 am

Well then, what is your point?

All patents block others from doing things. That's what patents do. If a patent is to a mouse trap, other people who make mouse traps are annoyed. If a patent is to an airplane, other people who make airplanes are annoyed. If a patent is to a light bulb, other people who make light bulbs are annoyed. Why do you think is special about software? Why shouldn't the guy with, what he believes is a more efficient method for calculating FFT, which the world was not aware of before be entitled to the fruits of his labor? How is granting that patent any different than granting a patent to a more energy efficient automobile transmission?

You said:

"What you seem to be saying here, and in the rest of the other comments, is that these are rare and unique patents, and therefore there is no problem that they were granted at all. "

No, I'm not saying that. You are saying that they are all known and not unique or obvious. I'm just showing you that what is being claimed here IS unique and not obvious.

The claims are there. Show me where they were done before they invented by the inventors associated with these patents or explain why they were obvious. I assure you, a patent examiner tried and was eventually convinced that he could not because the claimed techniques WERE new and WERE NOT obvious.

Avatar

les · April 24, 2010 at 7:15 am

"I am glad the application was rejected, that means that someone at the USPTO is still doing their job. But doesn’t that prove the point about the invalidity of the eHarmony patent? They are very similar applications of SVD forumulae, so the fact that the Netflix one was denied does not fare well to your argument that SVD and other algebraic patents should be accepted. "

No, what did that application in is the inappropriate way the claims express the invention. The application appears to have been prepared by the inventor himself, without an understanding of the legal requirements. Claim 1 was:

1. A method of incrementally approximating a generalized singular value decomposition of a data matrix comprising finding the gradient of a cost function incorporating the approximation error and following it.

The claim is meaningless. How does following a gradient result in an incremental approximation of anything? "following it" was overloaded with meaning in the mind of the application drafter, but it doesn't mean much in the legal sense to the examiner. The claim mentions a cost function out of the blue and doesn't describe the relationship to the matrix (if there was one). The relationship of the gradient to the matrix isn't expressed either. He could have fixed the claims through amendment, assuming the application explained things better…. but he didn't file a response to the first office action.

You said Netflix is involved in the application, but I don't see any sign of that.

Avatar

les · April 24, 2010 at 7:27 am

"Why is it pro software and business patent defenders always fall back to this clearly illogical and faulty argument? “Oh, THAT patent is fine because nobody will ever infringe it, it is too rare”. Then why apply for a patent in the first place? Why go to the trouble of ask for monopoly protection for something as trivial and unique as this? "

We are responding to the mantra of the anti-patent crowd that these patents block all other software from ever being written.

Don't forget, YOU cited these as examples of variable name substitutions on known formula. I don't see any of that in any of them. It is YOU that is moving the goal post.

I'm not burying you in claims. The claims are the important part of a patent, not the title. The anti-patent crowd sees a patent titled Method of FFT and over reacts thinking it covers every method of FFT, when in fact it covers only a very specific technique involving writing some info in memory and some info in registers. That's why I'm showing you the claims…in the hope that it might turn down the temperature on the anti-patent crowds rhetoric and lessen the chance that the baby gets thrown out with the bath water.

Avatar

les · April 24, 2010 at 7:33 am

"The core of the argument, which you seem to have ignored in favour of trying to bury me in copied patent claims, is that these patents are not inventions. They are not novel. They do not advance the knowledge of humankind. They do not promote innovation"

In all cases, the examiners tried to prove that the claimed techniques were known or obvious variants of known techniques. In all cases, they were unable to do so. You continue to say they were known. But you haven't shown that they were known.

Avatar

les · April 24, 2010 at 7:34 am

…by the way, if you find yourself blocked by a patent that claims something that was known….don't worry about it…its not valid.

Avatar

les · April 24, 2010 at 7:39 am

"If you can tell me how taking fourier transforms and turning them into patents by clever patent attorneys advances innovation, I will let this go."

Again, none of these takes FFT and turns them into patents. The ones related to FFT only cover very specific ways of calculating a very specific FFT. Its no different than patenting a particular way of making an airplane or a particular way of making a mouse trap. They do advance the public knowledge of mankind, some more than others, but all in at least a small way.

Avatar

les · April 24, 2010 at 7:46 am

"Why use caps on the word “application”? It is quite clear that I did say it was an application. "

Yes, I saw that you seemed to recognize that it was an application, but the fact that you cited it implied that you didn't fully appreciate the difference. Anyone can file any piece of paper and call it an application. 18 months later it will be published. Applications have no bearing on this discussion. Currently about 2/3 of applications are being rejected.

Avatar

les · April 24, 2010 at 7:53 am

"The core of the argument, which you seem to have ignored in favour of trying to bury me in copied patent claims, is that these patents are not inventions. They are not novel. "

"The narrowness and/or uniqueness of the patent is irrelevant to the point being made."

You say they are not inventions and are not novel. I show you that they are inventions and are novel and you say the uniqueness is irrelevant.

Who's moving the goal post?

Avatar

Roger Lancefield · April 24, 2010 at 8:53 am

Les wrote:

"They do advance the public knowledge of mankind, some more than others, but all in at least a small way"

"Public knowledge" that the public is legally prohibited from using? That's got to be a non-sequitur.

If patents are necessary for innovation, why is it that so many technology businesses *inadvertently* infringe the patents of others? This "accidental" innovation should be the exception rather than the rule, after all, patent system supporters tell us that the time and costs involved in creating new knowledge/inventions are only worthwhile because of the protection afforded by the patent system. But clearly, many businesses are having the same ideas as the "innovators" without even being aware of it. This seems to be a major contradiction in the "patents are necessary for innovation" claim.

If that claim were true, most patent violations would be restricted to deliberate acts of copying, and inadvertent violation would be unlikely (for the same reason that few people unwittingly write great novels). Yet over the years I've heard of many cases in which companies were subject to legal warnings, threats or action from companies with technologies that the recipients knew nothing or precious little about. This suggests to me that patents are, in these cases at least, not at all necessary for innovation and are in fact nothing more than an artificial and unwarranted reward to those who registered their ideas first.

It seems to me that the patent system has turned our global intellectual namespace into a legal minefield, and while it may act as an incentive and/or protection for some, it's also a fact that for many fledgling technology businesses (at least) it presents a troublesome, costly and sometimes insurmountable legal and technical hurdle.

Avatar

les · April 24, 2010 at 10:18 am

“Public knowledge” that the public is legally prohibited from using? That’s got to be a non-sequitur.

1) you can pay a royalty to use it.

2) Patents expire. You're free to use it when it expires.

Avatar

les · April 24, 2010 at 10:20 am

"If patents are necessary for innovation, why is it that so many technology businesses *inadvertently* infringe the patents of others? "

They aren't necessary in all cases. They PROMOTE innovation generally. They are necessary where investments in solutions are large and copies of the solutions are cheap (like in the area of drugs).

Avatar

les · April 24, 2010 at 10:23 am

"why is it that so many technology businesses *inadvertently* infringe the patents of others? "

So many? I'm not aware of any. Well….there may be "inadvertent" infringement with quotes around it….

Avatar

les · April 24, 2010 at 10:25 am

"it’s also a fact that for many fledgling technology businesses (at least) it presents a troublesome, costly and sometimes insurmountable legal and technical hurdle."

Wilbur and Orville were fledgling business owners. They didn't find it troublesome. Its only an insurmountable legal hurdle if you are using someone else's invention.

Avatar

Lauro · April 25, 2010 at 5:19 am

Les

Is there any reason you can't post your replies in one go? You look like a complete nutter (and that's before I digest what you're actually saying)

Avatar

les · April 25, 2010 at 7:45 am

Sorry Lauro –

I prefer to deal with one issue at a time. I find it best to keep the posts short when dealing with those with limited attention spans.

    Avatar

    Lauro · April 25, 2010 at 8:53 am

    Ah yes, ad hominem, that greatest of debating techniques.

Avatar

les · April 25, 2010 at 9:30 am

Sorry?

Did you take something I said personally? I was referring to the lowest common denominator in the internet audience, not to anyone that has posted here so far.

Jeeze, its not like I said anyone looked like a complete nutter or anything.

    Avatar

    Andres · April 26, 2010 at 2:25 am

    Well, it's not an ad hominem if it is true. You DO sound like a nutter.

    I'm away all this week in Geneva so I have no time to answer to a Gish Gallop.

Avatar

les · April 26, 2010 at 6:10 am

I have no idea what a Gish Gallop is.

I have notice that throughout the internet that when someone posts an argument that prevailing opinion makers on a cite have no intelligent response for, they resort to name calling and dismissive labeling.

I'm disappointed Professor.

Avatar

Matthew Slyman · February 4, 2011 at 4:03 pm

http://www.cl.cam.ac.uk/~mgk25/stallman-patents.h

– This says it all. From the website of Cambridge University, England; and further, this article has been parked on the personal web pages of a very influential doctor/ associate professor of computer science. For some reason he's kept the article on his page for the last eight years. Can't think why…

Avatar

Les · February 7, 2011 at 6:10 pm

Mr. Slyman –

Mr. Stallman makes many valid points. However, he also makes many errors. For example, he draws a comparison to distinguish software patents from chemical patents in an effort to show that patents make sense for chemicals, but not software. However, in making this comparison, he says that a chemical patent is to a single chemical, and is therefore ok.

But, this assertion is simply incorrect. A chemical patent will often claim a chemical made of several sub chemicals having a particular frame work or a particular medical function. For example, a cancer fighting drug will be couple to a carbohydrate so that it will be absorbed. Also, the cancer fighting drug will be claimed as a class of drugs…..

    Avatar

    Matthew Slyman · March 9, 2011 at 11:49 pm

    Les,

    I don't know enough about chemical patents specifically, but from studying chemistry and software in some detail, I suggest that chemistry is a more mature field of science than software. A few more differences between chemistry and software:

    In chemistry, if someone patents a process that gets you from A to B via C, you can usually invent a different process to get from A to B via D (particularly in the field of medicine, which involves some moderately advanced chemistry, and therefore yields alternative ways of breaking down or constructing a complex structure).

    The same applies in software to a certain extent, except that pretty much the whole of basic elementary chemistry is now in the public domain (and much of it was never "protected" by patents in the first place, only by guilds of secretive craftsmen protecting trade secrets, which worked fine in the early days of chemistry), whereas on the other hand, certain basic, fundamental and unavoidable elements of computing are not in the public domain.

    The software industry is starting out on the wrong foot, and will get into trouble very quickly if something radical is not done to fix the system.

    Another difference: In chemistry, it's usually not worthwhile patenting a method until you've done some practical experiments (or at least expensive simulations) demonstrating the feasibility of the method. On the other hand, in software, any fool can write an obvious method in opaque language, and ring-fence large portions of potential innovation from anyone being able to use it for the next 25 years. WHY?

    Chemistry is easy to visualise and has specific methods of defining the scope of a claimed "invention". Software is hard to visualise, and there are many fields of software innovation for which there are no established conventions for defining and documenting them. This makes searching for prior art particularly difficult, and makes it easier for patent applicants to pull the wool over the eyes of patent adjudicators (something that many software patent applicants have evidently been doing for some years now.) The way I see it, if there is no effective means of searching for prior art in a particular field of science, then there should be no patents in that field.

    Have an idea for a technique in software design/ development? Patent it. There's a good chance the principles can be made to work (unlike chemistry) – after all, you don't have to go into any detail about HOW your design is implemented. Because you're being granted a patent on PRINCIPLES, and not on SPECIFIC CHEMICALS or on CLASSES OF SPECIFIC CHEMICALS that can easily be defined and understood by all concerned qualified persons.

    Just look at some of the software patents that have been granted, and you'll see what I mean.
    http://arstechnica.com/tech-policy/news/2011/03/p
    – An article about a software patent that covers pretty well all software since the 1970's. Just shows the quality of review going on at the USPTO.

    Another criterion of patentability is supposed to be the non-obviousness of the innovation. If something is obvious, it should not be patentable. Because software has hard to visualise, it's easy for patent applicants to present an "innovation" that is totally obvious to anyone with any expertise or experience in the field, but not obvious at all to the USPTO. For an example of this, check out Amazon's "One Click Purchase" patent. Laughably obvious, but for some reason, the USPTO seems to think it's defensible.

    If a field of science (or a subsection of that field) is proven to be an alternative manifestation of the elementary principles of mathematics (such that much of software can be described in mathematical terms, and all of mathematics can be described in software), in other words, if we can prove that many software "inventions" are merely a DISCOVERIES and not an "inventions", and that there may be no way to work around them (so that many software patents are merely attempts to ring-fence mathematics itself); then in context of the other limitations we're seeing (such as the difficulty of searching for prior art), I think we have a strong case for saying that software should not be considered patentable.

    We simply don't have a mature enough science to be able to patent these things at all. Patents do not advance the state of the art in software. They hold it back, stagnate the market and help entrench the positions of a few industrial conglomerates. The only reason the US government is defending this position is because most of these industrial giants are presently American. If this position changed, they'd change patent policy in a flash.

    The USPTO is almost a laughing stock in the rest of the world. They grant a significantly higher proportion of applications than any other major patent office worldwide. Why? Because Americans are much more innovative and more careful and guarded in their applications, than the Europeans and Oriental folk from whom they recently descended?

    Personally, I feel so let down by the US Government and USPTO's nefarious practices with "intellectual property", that I've got a good mind to develop some world-beating software and refuse to do business with the United States. Personally, I think the whole world should band together, sign a treaty and stamp out this nonsense.

    What to do about existing software patents? In my personal view, all current US software patent holders should be band-boxed together as having applied for those patents in bad faith, and their patents should be cancelled with minimal recompense; since anyone with the expertise to apply for one should have the expertise to know that it's not a legitimate way of doing business. Software patents only do two things:

    1. Entrench the position of abusive monopolies and oligopolies,

    2. Give ammunition to patent trolls (highwaymen and robbers) who mostly have no time for real innovation.

    Recourse might be granted for court-based appeals, for any software patent holders who have an arguable case, and might be granted some form of industry-sponsored compensation.

    If we don't do this, then we're probably heading for a slow-motion train-smash that will be as significant for the software industry as the banking crash has been for the American economy.

    A MORE MODERATE VERSION OF MY PLAN (and perhaps much more of a sensible, happy middle ground) would involve saying that software CAN be patented, but only for 10-15 years, and the moment a mathematician discovers a fundamental proof that the method described in the patent is the best or only possible method, that patent should be invalidated. This, more moderate version of my plan might simultaneously spur the software industry, the computer scientists and the mathematicians into worthwhile and productive work.

    No patents written in language that cannot be understood by a graduate with a relevant degree should ever be granted. Patents like this that are already on the books should simply be tossed out as they do not fulfil the ostensible function of a patent (as it is so framed by the proponents of the patent system). The software industry is full of such patents. Time to scrap them all, start again and limit the number of applications that can be made in a given time period, so as to drive up the quality of applications. Apply for a patent and have it tossed out on the grounds of blatant obviousness or prior art that any fool should have noticed? Have an application rejected on the basis of it merely being an implementation of pure mathematics? Your fees (or those of your company) should increase exponentially for subsequent unsuccessful applications.

    DISCLAIMER: ONLY my personal opinions, with a few facts thrown in for justification. If ever I was in a position to do something about this almighty mess, I would take advice from more qualified persons, and consider these things more carefully before proceeding.

Avatar

Matthew Slyman · March 10, 2011 at 12:01 am

One tip for Les… "Les is more." In other words, why not, instead of posting 20 comments to the same blog (and demonstrating a weird determination to have the last word), simply write one comment pointing to an article of your own, presenting a coherent case that examines these things from another point of view? If, as part of such an article, you can demonstrate an actual understanding of the true points being made by the other side (and not merely blow down shadows of your opponents arguments, or set up "straw men"); you might actually convince some people.

Avatar

Les · March 10, 2011 at 11:11 am

Mr. Slyman

I agree that the USPTO is flawed. The EPO is also flawed. But that does not justify denying true inventors protection of their inventions, even if they are implemented in software. If someone invented a virus scanner that both worked and didn't slow my computer down to a crawl, I'd say that was probably patent worthy, even though virus scanners are implemented in software and even though the class of software known as virus scanners has been known. That would be true, even if in hindsight a software expert such as yourself would say it was obvious after the fact. If its obvious, do it first.

I assure you, that mechanical engineers will say that each and every can opener is obvious too. Everything is obvious once it is shown to you.

I agree it can be tricky to find prior art software patents, but with some experience and the right tools they can be found. I use google/patents to find them on a regular basis. If you as an expert in software can't find a document describing an important software technique, it may be because you are on the cutting edge and you should consider applying for a patent.

If you care to suggest an area of prior art you think can't be searched, I'd be happy to play along and try to find some just to show that it can be done. Heck, your colleagues have developed a program that can find answers to Jeopardy questions. If that can be done, software prior art can be found. (if not, set yourself the task of solving that problem in software and patent the solution!).

As for posting a number of times to the same blog….I'm not sure what the perceived problem is….but if there is one….be careful, your up to 3 posts already.

Avatar

Matthew Slyman · March 10, 2011 at 7:31 pm

You are using the word "invention" very liberally. Traditionally, an "invention" meant a physical object designed to perform a specific function in a greatly improved way. "Inventions" in the traditional sense do not only require imagination or a bit of design work, but also require a physical implementation. This was one of the traditional criteria of patentability. The patent system was originally developed to help "inventors" to amortize the high costs of investment in physical objects, manufacturing tools and machinery, physical distribution networks etc. Such costs are naturally amplified by the difficulties and expenses of marketing the new product that nobody understands yet.

Re: Can opener comparison – shows you don't understand software. A can opener is a physical object (see above). Implementing a can opener inherently requires a roughly even combination of pure science and engineering, and design work.

But software, by comparison, can be written that does not require any real design work at all, does not require any new tooling or machinery, and does not require a physical distribution network. You can implement pure mathematics in a computer program, with no embellishment whatsoever other than expressing that pure mathematics in computer-language format; and apply for a patent for it (as many people and big corporations do).

And guess what? The USPTO will probably grant the patent, and invite you to change the text and make it broader (in the US national interests, they suppose). The European and Japanese patent offices (in contrast) would laugh at you and tell you that you can't patent it, because it's pure mathematics.

Your following paragraph about finding prior art shows again that you don't understand software, and certainly don't understand the threat that software patents represent.

I have indeed used the USPTO's online search tools to search for prior art in the field of software patents. I started doing this about ten years ago (I think their abstracts were one of the first things to go on the internet). I was very much disappointed with what I found. The first thing I found in my search for cryptography related patents was a bunch of novices (who knew nothing about computer science) who had basically patented a "new way" of using the one-time-pad (a technique that has been around since the 1950's at least.) The rest of what I saw was equally disappointing. Stuff that was blindingly obvious to me as a 14-year-old computer user at high school, people had patented (and that too, some years after I was fourteen years old and had those ideas, and discussed some of my ideas with my peers) – they had patented first, not because they were more intelligent or more inventive than me, but because they had more money with which to corner the body of registered intellectual property.

I know how to look for prior art using the existing system. I WRITE INTERNET SEARCH ENGINES FOR A LIVING so I ought to know how to use them. I do know how to use the USPTO's search engine, and the quality of what I find just makes me want to vomit. Literally, stuff that I found obvious as a ten-year-old, twenty odd years ago; has been patented, less than ten years ago.

"Your colleagues"… Whatever. My colleagues at IBM Research… Hmmm… Sounds nice. I like you.

PROBLEM IS, most software patents don't involve the "inventing a can opener" approach, and don't incur the costs associated with that kind of development. Most software patents are all about using an existing can opener (a piece of pre-existing mathematics) in a slightly new way (or using a combination of just a few pieces of prior art, in a slightly "new" way). That sort of thing is not traditionally patentable, and it

Don't forget, patents are not the only form of intellectual property. There are:

* Copyright,

* Registered designs,

* Trademarks

etc. And what about trade secrets?

The above are much better suited to software generally, than patents.

Last point – a question actually. You wrote,

"Mr. Stallman makes many valid points. However, he also makes many errors." – Would you care to enumerate the VALID points Mr. Stallman makes? At least, to demonstrate that you have actually read the whole thing, and fully understood it? I just want to know that you're not just trying to sound conciliatory. I don't want that kind of condescension. I want understanding, please.

Matthew Slyman, M.A. Computer Science (Cambridge University, 1999).

p.s. You probably shouldn't comment again until you understand the issues; particularly if you're paid to do so, or at risk of appearing mercenary in your approach. Otherwise, if you insist on commenting on blogs like this one, and wish to appear relatively normal and professional, you should at least declare any potential conflicts of interest. Who exactly is paying you? What do you do for a living? What qualifications have you got for this? To be direct, are you being compensated in any way for these posts, either directly or indirectly, financially or in-kind?

Avatar

Matthew Slyman · March 10, 2011 at 8:12 pm

@Les: "I have no idea what a Gish Gallop is." Pity, because if you knew how to use Google, you'd be able to find out in less than 20 seconds, as I did. It's strange, how you don't know how to use Google, but you do know how to use the USPTO's search tools for prior art… It's like you have a doctoral degree from Harvard, but failed your high-school diploma. Something's not right…

To finish my unfinished sentence (before you call me out on it) – "That sort of thing is not traditionally patentable, and it " … "should never be." Can you not imagine how it might impede scientific progress, if people were allowed to patent every possible COMBINATION of two or more basic and obviously unpatentable ideas; without making any real & personal investment in research and development? That's basically where we're heading in software, because of people like Les.

Still waiting for Les's enumeration of Stallman's valid points…

Avatar

Les · March 11, 2011 at 1:14 am

If you can get useful results from the USPTO search "tool", my hat is off to you. I use that as rarely as possible. I recommend google/patents instead.

As for Mr. Stallman, I read his treatise over a month ago and don't have time to read it again right now….but as I recall his complaints about the examination process were spot on and he is correct in asserting that sometimes the results of the examination process are less than ideal.

I'm a bit confused with regard to your position. I thought you said, in an earlier post that it was impossible to search for prior art in the area of software and therefore the category of "software" should be excluded from patenting. Now you seem to be saying that prior art can be found easily.

"Re: PROBLEM IS, most software patents don’t involve the “inventing a can opener” approach, and don’t incur the costs associated with that kind of development. Most software patents are all about using an existing can opener (a piece of pre-existing mathematics) in a slightly new way (or using a combination of just a few pieces of prior art, in a slightly “new” way). That sort of thing is not traditionally patentable, and it "

That can be said of 99.999% of all inventions,not just software. There are very few patents to things like –cold fustion–. And even those claim assemblies of known components….fish tanks, water, platinum electrodes etc.

Your assertion that software is cheap and therefore not patent worthy is simply unsupportable….I'm sure Microsoft's programmer salary budget is in the many hundreds of millions.

Your assertions about the original intent of the patent system is also not supportable. The intention is/was to promote progress (at least that's the stated intent of the U.S. system). While the cost of new drug development makes an easy to understand example of the need for the system, the cost of an invention is not a factor in whether something should be patentable.

I dare say, the cost in developing a new/better speech recognition package is far higher than the cost in designing a new can opener.

With regard to your assertions about tradition…why would you lock us into what was done when people were writing with feathers? Of course software wasn't patented in the early industrial age. Software HADN'T BEEN INVENTED YET.

No one is paying me to make these posts. Who do you think has that kind of money? Where should I direct my C.V.?

Avatar

Matthew Slyman · March 11, 2011 at 2:44 am

I won't believe you're capable of using "Google/Patents" until you show me that you can find out what a "Gish Gallop" means.

"I’m a bit confused with regard to your position. I thought you said, in an earlier post that it was impossible to search for prior art in the area of software and therefore the category of “software” should be excluded from patenting. Now you seem to be saying that prior art can be found easily."

– Seems like you're deliberately confusing yourself, because I've been pretty clear. Searching the body of "prior art" represented by prior patent applications, might turn up a few matches (as it did in my case), but because of the patentese mumbo-jumbo that's peculiarly endemic within the field of software and business-process patents, you can rarely be more than 50% certain that you've covered yourself legally. So, more particularly in light of the fact that those who look can be hit with triple damages, what's the benefit of looking? Even if you are diligent, a "patent" can still be held incognito, lurking until the litigation target is fat enough to accomodate its hungry parasite.

It appears you are deliberately trying not to understand any of my arguments. Have you not seen the profit margins in the software industry, compared with (say) manufacturing? The manufacture and distribution of physical goods is what the patent system was originally designed to help support.

"99.999% of all inventions"… Care to tell me where you're getting your statistics from? I thought so… Who can argue against this kind of nonsense?

With regard to the application of tradition, I think you will find that:

* Tradition (or "precedent") is the foundation of all legal systems that take "case law" as their authoritative guide.

If we're going to be breaking from tradition, then we ought to have some jolly good justifications for doing that. We ought to be looking long and hard about the original justifications for the creation of the patent system in the first place, and seeing whether those justifications fit with the situation we're heading into. Still fit for purpose?

Judge by the results.

"No one is paying me to make these posts." – I simply don't believe this. You're either deeply indoctrinated, or else you've worked (at some point or other in the past or present) in an industry that benefits from this kind of litigation. Don't forget that I asked whether you benefitted DIRECTLY OR INDIRECTLY from any of this. Have you got a stake in this system? It sounds very much like you have. I'm smelling the astroturf, and the smell is coming VERY strongly.

To be plain, I have a stake in software NOT being patentable. I'm a genuine innovator, struggling to make ends meet and feed a family. As soon as my business gets big enough, there are going to be crocodiles looking for a slice of the action. If I copied their work, they should sue me for copyright infringement. I don't do that sort of thing. I'm the real deal. As a software developer who has studied software patent law in general terms, I can say for myself that US software patent law inhibits grass-roots innovation, reduces free-market competition, and increases abuse. I can innovate and work all I like, and spend a high proportion of my time trying to avoid other people's "intellectual property"; and my lunch (and my family's lunches) will still be stolen by good-for-nothing litigators who never innovated in their lives (and by their client patent holders, who as the supposed "inventors", wouldn't even have guessed that my innovations infringed on their own patent.)

CV – link on this thread please (maybe it will redeem you in some small degree, by providing some insight into who you are). I don't want it in my in-box. I'm not detecting any depth of understanding here (some of your arguments are just too stupid to bother responding – "HADN'T BEEN INVENTED YET" – oh dear. No recognition of the fact that software, under the original system, hadn't just "not been invented", but would never have been patentable without some seriously dodgy reinterpretation of what should be patentable, reinterpretation that doesn't bear any scrutiny. Sorry but you don't appear to have understood Stallman's message, though you claim to have done so.)

I'm going to call it a day because you're obviously winning the debate. twenty-five posts to six… I can't compete. You win.

Avatar

Les · March 11, 2011 at 7:00 am

With regard to:

" As a software developer who has studied software patent law in general terms, I can say for myself that US software patent law inhibits grass-roots innovation, reduces free-market competition, and increases abuse. I can innovate and work all I like, and spend a high proportion of my time trying to avoid other people's "intellectual property"; and my lunch (and my family's lunches) will still be stolen by good-for-nothing litigators who never innovated in their lives (and by their client patent holders, who as the supposed "inventors", wouldn't even have guessed that my innovations infringed on their own patent.)"

This is an effect of EVERY patent. The other guys trying to make flying machines said the same thing about the Wright Brothers patents.

Re:

"I’m a genuine innovator, struggling to make ends meet and feed a family. As soon as my business gets big enough, there are going to be crocodiles looking for a slice of the action. "

I hope for your sake, that you get over your aversion to the patent system and patent your innovations. Otherwise, if the product is good, someone will copy it, if only in the big picture sense, who wont have invested the time and energy you did and they will be able to under cut you and drive you out of the very business you created.

Bon chance mes ami

Avatar

Les · March 12, 2011 at 4:33 am

"“99.999% of all inventions”… Care to tell me where you’re getting your statistics from? I thought so… Who can argue against this kind of nonsense?"

Prove me wrong. Find 1 patent that doesn't claim a combination of known components (other than a software patent, which at least has the chance of including a new algorithm).

"If we’re going to be breaking from tradition, then we ought to have some jolly good justifications for doing that. We ought to be looking long and hard about the original justifications for the creation of the patent system in the first place, and seeing whether those justifications fit with the situation we’re heading into. Still fit for purpose?"

But we are not breaking with tradition. In the U.S. the "traditional purpose" is: To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.

Surely software is a useful art, despite what you might think of Grand Theft Auto or Mario Bros II.

Avatar

Les · March 12, 2011 at 5:18 am

By the way, I've now had an opportunity to look at the patent complained about at the link you provided:

http://arstechnica.com/tech-policy/news/2011/03/p

The patent is amazingly detailed:

http://www.google.com/patents?id=P4MpAAAAEBAJ&amp

225 pages of figures alone.

The claim language is broad. But, the related application was filed in 1990 and apparently, the claimed method was new then…so… the inventor was entitled to broad protection. The patent appears now to have expired….so…if you were chomping at the bit to implement a version of this technique for displaying search results…have at it.

Avatar

Matthew Slyman · March 15, 2011 at 5:59 am

Hypertext, new in 1990?!? Or are you suggesting that referenced diagrams, such as those found in common encyclopaedia for over a hundred years, were new in 1990? Such diagrams having been in general use for hundreds of years prior to 1990, and computer software merely being a new medium for displaying such, should we not consider it rather obvious to implement such a diagram in computer software? Taking an existing technique and saying "my version is computer-based" doesn't make it innovative.

Just look at the references section of that patent – it's huge, and contains a large number of very reputable sources. With a little experience of what such articles typically contain, and a little knowledge of the history of software, and having skimmed through the patent we're discussing (I'm not going to read it in detail, no thanks; since the parts I have read have taught me NOTHING); I find it very hard to believe that this patent adds any new and non-obvious innovations to the body of science contained within the referenced articles.

Specifically, I reckon that if I looked hard enough, I could find BBC Micro, Apple, Xerox or Amiga software that invalidates at least parts of this patent through prior art. I'm sure I can find an early piece of learning software, or networked multimedia authoring software, or even just elements of early GUI based filing systems, that cover parts of this patent.

The bits of the patent that I've read look either sloppy or wilful, on the part of the applicant. To me, the patent looks to have been written by the hand of a hotshot businessman who was trying his luck at a get-rich-quick scheme.

The fact that the patent is 270 pages long, and has all sorts of ancillary details, is no indication of ingenuity on the part of its authors. It's merely an attempt to get maximum value-for-money out of the patent application process (which is not exactly cheap). Don't forget that the patent is subdivided into 78 (seventy-eight) individual "claims", each of which appears very broad in application, to me.

If the patent has expired, as you suggest (which would be rather odd, considering that I understand US patents last for 25 years, meaning that the patent in question would expire in 2015), this does not prevent people from using the case as an example in demonstrating why software patents are bad. The principle applies just as well in general, whether or not this particular patent has now expired. As I originally suggested, and as the authors of the Ars Technica article are suggesting, and as everybody except you appears to acknowledge; this particular patent has a history of being used as a weapon by non-innovative "patent assertion entities" against ordinary hard-working businesses trying to do genuine work and feed their families.

You just don't seem to get it – and I think it's not just an incapacity to understand that causes your dogmatic obstinacy. It looks increasingly like wilfulness too.

How do you think the engineers would do, if the second law of thermodynamics had been patented? Or if its application to any form of mechanical machinery had been patented? Those are the sorts of patents we're increasingly seeing in computer software – and the sort of scope that the lawyers are trying to extort from the intellectual property system on behalf of their clients. The lawyers are increasingly trying to make the specific design/ implementation section of patents irrelevant to the actual scope of the patent. That's why anyone who knows anything about software, is complaining about the apparent negligence of the USPTO, and the ever-growing, seemingly exponentially increasing, body of non-innovative over-general patents being applied for by massive corporations with deep pockets, who use the patent system to prevent start-ups from threatening their position as "top dog" with genuine innovation.

Avatar

Les · March 15, 2011 at 10:08 am

1) Your understanding is incorrect. The general rule is U.S. Patents application for which was filed after some time in 1995 (I think) expire 20 years after their filing date or earliest priority date.

2)I don't think the claims of that patent read on hypertext. They read on displaying search results in some sort of relationship. I think the intention if for a most relevant result to be placed in the center with less relevant results displayed further away…but with similar results displayed near each other…

I'm out of time right now…. so I will respond to other issues later (if I see any).

Avatar

Les · March 16, 2011 at 10:11 am

1-continued) In the present case, the application was filed before 1995, so the life of the patent was 17 years from the 1993 issue date, which means it expired in October of 2010.

Avatar

Les · March 16, 2011 at 10:17 am

"and computer software merely being a new medium for displaying such, should we not consider it rather obvious to implement such a diagram in computer software? Taking an existing technique and saying “my version is computer-based” doesn’t make it innovative."

If we take this logic as the rule, then the Wright Brothers plane should not have been patentable because birds had been flying for thousands of years.

Even if your assertion that the referenced diagrams were used in encyclopedias were true (which I would dispute) that does not make this invention obvious. Even if a human graphical designer can arrange a diagram in a particular way, that does not mean that getting a machine to do it is obvious. Even if humans can play Jeopardy, that does not make it obvious to get a machine to do it.

Avatar

Matthew Slyman · March 16, 2011 at 9:13 pm

Thank you for your clarification about the expiration of that patent – which is a valuable contribution to this particular dialogue.

The more general points still stand though – the patent in question represents a clear abuse of the patent system.

Let's suppose that your previous message is correct – so they patented spider diagrams, or semantic networks, at least, wherever they were implemented on a computer. That would be no more admirable than attempting to patent hypertext in 1990. If that's what they were patenting, why on Earth didn't they make that clearer? How is it that I can read significant portions of that patent, and still think they're describing common hypertext, or regular file systems, or searches within computerised filing systems? Is it not dangerous for the patent to have been drafted in such an overly broad manner, as to make it potentially threatening to such a broad swath of software?

Your Wright Brothers example doesn't make any sense at all. Red herring. Fish can swim underwater. Does this mean that submarines and Cousteau's aqualungs should never have been patentable? Of course not. There were mathematical, technical, manufacturing, practical and design innovations that had to take place before people could follow where fish had gone before.

Since you admire the Wright Brothers so much as to mention them more than once, I suggest you take a look at this:
http://en.wikipedia.org/wiki/George_Cayley
Now, George Cayley (an Englishman) was acknowledged by the Wright Brothers as their inspiration – Cayley invented most of the calculations applied by the Wright Brothers – but most Americans seem to have been taught at school that the Wright Brothers went straight from looking at birds to doing calculations to making their first Flyer. Not so.

NOW – does the story of Cayley/Wright not just strengthen your case? The Wright Brothers just took Cayley's glider and stuck an engine on it, right?

Cayley was prevented from developing sustainable powered flight himself because the steam engines of his day were simply too heavy. The engines and fuels for powered flight simply didn't exist in his time. He couldn't pursue that investigation – he was ahead of his time. That wasn't his fault, but that doesn't change the fact that many new innovations were required to go from Cayley's work to what the Wright Brothers did.

The Wright Brothers' innovative steps were (from my perspective) to take Cayley's work and move it on to the next level –

1. Powered, sustainable flight.

2. Various changes to aircraft configuration, particularly in centre of gravity and lift configuration, that enabled their craft to operate as something more than gliders.

Powered flight might generally have been anticipated by Cayley, fifty years before the Wright Brothers; but it was the Wright Brothers' innovative use of materials, manufacturing techniques borrowed from the bicycle industry, and clear refinements in Cayley's methods that enabled the transition from mere gliding (at a steep angle) to continuous powered flight. This transition was enabled by a number of specific mathematical, technical, manufacturing, practical and design innovations that the Wright Brothers made.

They didn't just take a couple of very simple known techniques, combine them, and attempt to patent the combination, without producing any actual products themselves. That's the key difference between the Wright Brothers and the modern patent we're discussing.

The patent in question makes no discernible contribution to mathematics, science, design or anything else – they merely combine a few simple, known techniques and say,

"We're patenting this combination of techniques. We think it hasn't been done before. If anyone combines these techniques in the next 17 years (as we think they might, given the way the industry is obviously heading), we'll sue them, or at least threaten to sue them." Anyone who knew anything about computing and diagrams in 1989 would say that any and all of the techniques described in that patent were quite obvious. How come the patent office didn't know it?

Case in point – not every modern patent applicant is the modern equivalent of the Wright Brothers. Too many of them are abusing the heritage of the Wright Brothers, and using it as an emotive excuse to extort money from the real innovators.

Avatar

Les · March 21, 2011 at 9:58 am

"why on Earth didn’t they make that clearer? How is it that I can read significant portions of that patent, and still think they’re describing common hypertext, or regular file systems, or searches within computerised filing systems? Is it not dangerous for the patent to have been drafted in such an overly broad manner, as to make it potentially threatening to such a broad swath of software?"

Ah! This is an important issue. I see this time and again. Software guys read the title of a patent and think that everything that falls under the title is covered by the patent. This is not remotely the case. What is covered by the patent is clearly recited at the end of the document in THE CLAIMS.

The specification must explain how to make and use the invention so that one of ordinary skill in the art could do so, without undue experimentation.

Some are more worried about where that line is drawn than others. In the present case, they appear to have thought that a great deal of explanation was required.

The present patent has 78 claims. Claim 1 reads as follows:

What is claimed is:

1. A computer-based method for aiding a user in assembling a customized body of information from a larger body of available information segments, the method comprising:

displaying a set of labels, each label providing an abbreviated indication of information content of a corresponding one of said available information segments in said larger body,

enabling a user to point to individual labels in said set using an electronic pointing technique,

for each label to which said user points, displaying to the user, for previewing, information content of the corresponding segment,

enabling a user to choose to include in the customized body of information, selected ones of said available information segments, while excluding from the customized body of information other available information segments, and

assembling said customized body of information in response to choices of information segments made by said user.

Google makes all 78 available here:

http://www.google.com/patents/about?id=P4MpAAAAEB

As you can see, the first 3 steps sort of sound like mouse over…which I suppose in turn has a hypertex aspect to it…so…the specification when on a bit about hypertex…but I don't remember any mousing over in the 90's… do you?

Now as I look at claim 1 today, I don't see what you are calling a spider diagram…and I'm now not sure why I had that impression the other day…

Maybe I skimmed quickly and came to claim 28 and assumed it was representative…claim 28 reads as follows:

28. A computer-based method for aiding a user in accessing a body of stored information which includes segments of related information, the method comprising

displaying a set of labels, each label providing an abbreviated indication of information content of a corresponding one of said segments,

[b]said labels being displayed in an organized model reflecting relationships among information contents of said corresponding segments,[/b]

enabling a user to point to individual labels in said model using an electronic pointing technique, and

for each label to which said user points, displaying to the user, for previewing, the information content of the corresponding segment.

That is what I was referring to and which you are calling spider diagram.

RE: "Your Wright Brothers example doesn’t make any sense at all."

I agree. However, it is the logical extension of your assertion that "Taking an existing technique and saying “my version is computer-based” doesn’t make it innovative."

Flying was an existing technique. Birds did it. According to your logic making a version based on sticks, rope and canvas doesn't make it innovative.

I believe the Wright Bros first patents were to a glider that could be controlled. That is, they patented a particular wing configuration that could be warped in flight by the pilot, thereby providing a modicum of control and steering. Others then complained that they were blocked from making a working aircraft….until someone came up with flaps and slats….gee that's tough…. you should have solved it faster (Hence, the encouragement of progress)

Avatar

Matthew Slyman · March 22, 2011 at 9:24 pm

"Software guys"—this is a stereotype, and it is strange how you feel so confident to paint me in with that stereotype when you clearly don't know me. Perhaps your messages are not meant for me but rather for others who might be reading this thread? The implication of your stereotype, of course, is that anyone who disagrees with you must be a mere "software guy" who has expertise in nothing but in coding a little software. For your information, I read through the claims of that patent, as well as the abstract and a good portion of the rest of it. I found nothing innovative, and much to threaten the makers of prior art (the genuine innovators) with lawsuits they might ill afford.

Where is your link to your CV? Yes, I mean the resume for the real person/people who is/are going by the screen name of "Les"? You promised us, didn't you?

You seem pretty desperate to cast me as some provincial and ill-educated person, while casting yourself as the expert in this field; but you also seem rather ill-acquainted with the very stories you cite most vehemently.

Cayley gave away all his inventions free of charge—he never applied for any patents, as far as I can gather, most particularly not for his innovations in aircraft technology. He found himself to be wealthy and self-satisfied enough, that he shared his innovations purely for the good of mankind. His example might be followed by many modern inventors, to their personal satisfaction.

I have nothing against the Wright Brothers patenting their Flyers, and certain of the techniques they invented. Only, the Wright Brothers patents are worlds apart from what we're seeing the field of software engineering.

Your "points" are making less and less sense. You claim to understand my points and to take some of them into account, but you evidently do not understand even the most basic reading material you claim to have studied.

Perhaps the only thing I can do at this juncture is direct you to some more reading material:
http://en.wikipedia.org/wiki/Software_patent_deba

Why not edit that page to reflect your point of view? 😉 Suggest you simply delete the section entitled, "Arguments against patentability", and replace that section with a brief and simple paragraph stating that such arguments might exist. Under the section entitled, "Arguments for patentability", I suggest for you to write an essay about the Wright Brothers.

Avatar

Les · April 1, 2011 at 9:53 am

Software guys is not a stereo-type, its a classification. I didn't paint you will it. I said I have had this conversation with software guys and at the beginning, they think that because a patent is titled method of performing FFT that the patent covers all FFT.

I don't know where you picked up the "mere" part either. I didn't say or imply mere.

I don't know what you mean by makers of prior art. If the art is prior then by definition it is not covered by the patent or it invalidates the patent…if as you say the patent might "threaten" the makers of prior art.

No, I don't believe I promised you my CV.

Again, I have no idea why you think I think your are provincial or ill equipped. I said or implied no such thing. If I thought you were ill equipped, I wouldn't bother talking to you.

I don't think I claimed to study any reading material either…. I did read, that one page you recommended…

I can't figure out how to edit a wiki…. and I'm sure if I did figure it out, some anti patent person would delete my points in a few hours. The authors of the wiki are clearly biased. The pro-patent portions is slanted and derogatory in its choice of words and characterizations…Including things like this:

Inventions can only be patented if they are non-obvious. This reduces the chance of patents being granted on mere algorithms with no technical effect[citation needed] or the granting of "trivial" patents with no inventive step.[9]

as an alleged pro-patent argument.

So, clearly anything I would say there would be pounced upon and deleted.

Avatar

Matthew Slyman · June 2, 2011 at 8:50 pm

Microsoft joins pre-emptive patent protection program:
http://arstechnica.com/microsoft/news/2011/06/mic
It might be interesting to get a more in-depth editorial on this blog.

Microsoft is now openly acknowledging, by their actions, no less; that the software patent system is getting out-of-hand.

The system, as it stands, rewards patent trolls so much that non-productive patent trolling is not only a lucrative business model, but moreover, trolling is characteristically and disproportionately lucrative within the software industry in particular. It's hard to argue that the patent system is serving its intended purpose (most particularly, for that industry). It's obvious that the ecosystem is broken, if you just look at the kind of fish we're seeing in the pond, and the way they behave.

These days, there's basically no way for small software start-ups to make it big with broad-based software (open-source included), without some form of partnership (either formally agreed or else in a mere symbiotic relationship) with one or more of the bigger fish in the pond, basically either Oracle, Microsoft, Apple, IBM or Google or some combination thereof. Even Google would have been nothing but for their initial partnership with Yahoo. All software firms need to participate in the patent protection racket in order to avoid falling afoul of some predator, big or small; once they get big enough to be considered a worthwhile host for a parasite, or a worthwhile target for a hostile take-over bid backed up with patent litigation to weaken the target.

Apple's recent defence of their iOS APP developers from Lodsys, using veiled threats to invalidate Lodsys' "patent" (without actually going ahead and doing so), underlines how broken the system is. It takes a company the size of Apple to outgun patent trolls like Lodsys. But Apple and Microsoft rarely go ahead and trash a troll's patent unless they really have to do it for their own survival. Why not?

1. The less patents they trash, and the less noise they make, the less people will notice how broken the system is.

2. By leaving the patents in force, and paying license fees to Lodsys for an obviously spurious patent, Apple gains something by helping Lodsys to threaten their competitors (any "app" developers not working within the iOS part of the ecosystem).

There's not a leg to stand on, for those who say the system is not broken or at least imbalanced. In case anyone didn't already notice from my previous remarks, I'm in favour of software patents; I just think 20+ years of blanket coverage for an excessively broad interpretation of some minor "innovation", is far too much of a reward for the effort most patent successful applicants put in.

(Les; you know that we're all waiting for your opinion on this, hoping to hear from your inside-track perspective… Tell us why Microsoft is wrong?)

Avatar

Matthew Slyman · June 2, 2011 at 9:10 pm

…Personally I think the system would be roughly balanced if the criteria for the approval of software patent applications were massively toughened up, so that we would see a 100x – 1000x reduction in the number of patent applications, a reduction in the proportion of U.S. patent applications being granted (bringing the USA at least into line with Europe and Japan), and a reduction in the proportion of software patents that subsequently become involved in litigation (bringing that proportion down to levels typical of the industries for which patents were originally created). Here is a simple, measurable way to verify that the software ecosystem will start to get back in balance. I can't see how anyone can argue with this, even if they support patents as a general principle.

Avatar

Les · June 3, 2011 at 9:40 am

Sorry to disappoint Matt – I have no beef with MS looking for prior art for patents it thinks are not valid. That has been my argument all along. If you think a patent isn't valid, prove it. Most of the posts on this topic of of the "oh commmon" eye rolling variety. Its those that I object to.

My experience is that Patent Examiners are hell bent on rejecting applications. So, I don't think MS will invalidate very many, a few for sure…but not many.

FYI…this is not the first attempt at this sort of thing….there was a web site a few years ago….bountyhunter.com or something like that…they tried to get people to submit prior art against patents in litigation…the deal was, if the art was appreciated by the side that wanted it, you got some money….if not, then not…maybe the provided art had to win in court… I forget…

Anyway…they seemed to go belly up after a few months….

Avatar

Les · June 3, 2011 at 9:42 am

Just did a little Googling…. it was BountyQuest.com

http://www.infringementupdates.com/2006/08/patent

Avatar

Les · June 3, 2011 at 9:48 am

By the way, I do object to your disparaging remarks and name calling re: patent trolls. Don't forget, the worst (in your view) of the non-practicing entities purchase their patents from deserving inventors, most likely inventors that were unable to interest any practicing entities in their patents or did not have the financial resources to enforce the patents themselves. While I hope they give the inventors a fair price, and I would object if they didn't, I think the NPE's you so disparage perform the service of sticking up for the little guy.

Avatar

Les · June 13, 2011 at 4:56 am

Matthew –

I wondered how you might respond to the following:

"The software industry offers a similar story from the other side of the coin, where the absence of patents was linked with the development of a monopoly. The U.S. courts had allowed themselves to slide into the view that patents on computer software and methods of doing business were similarly against a range of vaguely defined so-called "public policy" notions from 1972 Supreme Court Benson decision up through the 1980’s and into the 1990’s. That’s when we got Microsoft. And it was only after the 1994 appellate court Alappat decision that the single biggest Microsoft competitor—Google—was able to come to market in the U.S., relying on strong patents and trade secrets."

Which was copied from here:

http://www.hoover.org/publications/defining-ideas

Avatar

Matthew Slyman · July 24, 2011 at 8:27 pm

Wow – four people called "Les" responded since my last call here! Amazing! What are the chances of that happening? Any statisticians out there???

http://www.bbc.co.uk/news/technology-14254744

I thought you would appreciate some more reading material. Oh, and one last thing:

"My experience is that Patent Examiners are hell bent on rejecting applications." — Have you applied for any patents yourself then? Or had a stake in such patents being granted??? I asked you on March 11th:

"Have you got a stake in this system?" — I think it's about time for you to answer this question directly. No more evasion. No more Gish Gallops, trying to scroll my question off the top of the list… Just answer directly, okay?

HAVE YOU GOT A STAKE IN THIS SYSTEM?

Leave a Reply to Matthew SlymanCancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.