Dennis has provided detailed analyses of most of these cases; this article entitled “Software Patents are Crumbling Thanks to the Supreme Court, provides a nice high-level overview, including a very telling chart, summarized thus:
The 14 patents the courts invalidated on subject matter grounds in 2013 was a record for recent years (such decisions were rare in the 1990s and early 2000s). And this chart reflects decisions on all types of patents, not just software patents. With 11 software patents invalidated in just the last three months, the courts are on track to blow away last year’s record with software patent cases alone. So this is the most hostile the courts have been to software patents in at least two decades.
Judge Moore was right in CLS, it seems. Next industry on the chopping block: Biotech, in Ariosa (discussed by Dennis below).
[…] An article by Mike Masnick, another vocal opponent of software patents, is titled “Be Happy: Software Patents Are Rapidly Disappearing Thanks To The Supreme Court” and “Software patents dying out in US” is another headline to keep a record of. Some of the most popular lawyers’ sites are prepared to acknowledge this. […]
Reeling from the shock of these crippling judicial blows to the Most Important Industry Ever in the History of the Universe, the stock market hit an all-time record today.
link to money.cnn.com
Perhaps now the 29-30 billion per year lost to litigation will taper down…
Quite an infusion of capital.
Is this not all OBVIOUS
The subject matter invalidation of Alice and subsequent cases seem to me to be misdirected.
The main arguments seem to be that …if it can be performed by humans without the computer then adding the computer/software doesn’t make it patentable…
In other countries this is sometimes referred to as “mere automation” and has never been patentable as it’s not considered inventive to merely automate a known process.
Isn’t this argument really a disguised ‘obviousness’ argument, i.e. I can perform the task without a computer and therefore it would be OBVIOUS to use a computer to automate it.
All of these cases involve methods that are arguably known or obvious and so why this has not been raised as the primary objection seems odd. It may be that the Obvious argument has traditionally been much harder to succeed on. If this is the case then the question must be asked WHY?
I don’t believe there is anything inherently unpatentable about software. The problem that the public sees is that inventions of dubious inventiveness seem to be granted and reach the news and they all happen to be for software-related inventions.
There are plenty of ground-breaking innovations made in software that deserve patents and the public would not object to these if they saw them. Unfortunately, as with all news, only the exceptional cases are newsworthy.
The problem is not software as patentable subject matter, the problem is that there needs to be more rigour in examination.
As noted in the comments, it’s difficult to search for prior art software as it’s typically non-visual. Mechanical inventions are easier to search because you have drawings, chem and biotech because you have equations, genetics because you have sequence codes.
This is not a reason to reject it all. This is a reason to find a better way of searching.
Can anyone explain why these patents weren’t attacked on obviousness grounds?
None whatsoever. I believe that was why math was supposed to be excluded from patents in the first place. The results are obviously derived from previous results… Simple, rigid, logic.
No inventive step.
Ashlin, The main arguments seem to be that …if it can be performed by humans without the computer then adding the computer/software doesn’t make it patentable…
I think you misunderstand — but that seems common to a lot of folk.
The problem is more related to whether the novel subject matter is statutory and/or whether the novel subject matter, if otherwise statutory, is stated at too high a level of abstraction.
Economic principles and math are both examples of nonstatutory subject matter. Invention in the claim must be in the otherwise statutory, or the nonstatutory (e.g., math) must be applied to transform (Alice) the statutory into something new or improved.
When the novelty is otherwise statutory, the claim remains abstract if it is not applied with any specificity. The Rubber-Tip Pencil is case is a good example. A claim for “a new and useful rubber head for lead-pencils” was held unpatentable because a rubber tip was old and the idea of attaching it to the pencils was without any detail.
Ned,
The point above is that the objections to these particular software patents can all be addressed by arguing the inventions are obvious. You don’t need a separate software exclusion or ineligible subject matter attack.
In other countries the rubber tip case would be considered a ‘mere collocation of integers each performing their own function’. There is no ‘synergistic effect’ obtained in the combination and therefore not patentable.
Attaching two things together so they are conveniently located together but each operate independently is not inventive.
To me that case would have nothing at all to do with subject matter, which is after all ‘mechanical’.
AR, I think there is a similar doctrine in the United States where a parameter is known to be variable within a range without changing the substance of what’s going on in order to accommodate various conditions. This is considered to be obvious. Thus if it was known that the rubber piece could be attached to anything that had a tip because it was naturally elastic, attaching it to a pencil tip would seem obvious.
One could approach the topic from common knowledge and common sense and arrive at the same result as the US Supreme Court did.
But legally, it also makes sense to say that the broad suggestion to attach a piece a rubber to the tip of a pencil without any detail on how to do that other than its natural properties is quite a bit overbroad if the claim were construed to cover means of attachment not disclosed and described other than simply pushing the tip into the rubber. In a sense, this is the same problem we encountered with Morse’s claim eight.
And just suppose, that the specification disclosed shaping the piece a rubber in a manner that it could effectively attach to the tip of a pencil without falling off, etc., then the claim itself would not include the essential structure or methods actually disclosed. This would present a separate section 112 problem where the claim did not particularly point out and distinctly claim the invention.
only the exceptional cases are newsworthy.
There are literally thousands of these “exceptional” invalid software cases out there. Every Tuesday morning reams more are granted.
it’s difficult to search for prior art software as it’s typically non-visual.
It’s typically not recited in the claims is the problem. Only the functionality is recited.
Can anyone explain why these patents weren’t attacked on obviousness grounds?
“Judicial activism” turned obviousness into an expensive proposition. On top of that, there is no consistent nomenclature used by computer-implementer types and they appear to go out of their way to invent terms that obscure what they are believe they are “innovating” to the greatest extent possible.
At the end of the day, they are claiming either (1) logical information processing steps that are older than the hills (in spite of their “new” labels) and which do nothing more than what anyone would expect them to do; or (2) information itself, which is ineligible.
I don’t believe there is anything inherently unpatentable about software.
I do and so do a lot of other people.
I don’t believe it is the “computer implementor types” doing that.
They are much more straight forward in using the language of the art.
Patents are written in legalese…
It has been pointed out that if the patents really were supposed to “reveal” the how-they-did-it, then the patent would include the source code used, and/or the exact formula used.
These sound like comments against patents in general.
e.g. a machine is not patentable because it is no more than joining of nuts and bolts to make a hovercraft. A new cancer drug is no more than taking compounds A +B + C and combining in solution.
The only difference with software is that rather than physical components/compounds you are using software modules/code.
Therefore, by your logic nothing should be patentable.
People don’t think Software should be patentable because
a) there’s been some software patents issued for trivial inventions. This is a problem with examination process and searching, it is NOT a result of a subject matter issue.
b) the majority of people use software on a daily basis and therefore feel affected by the patent, as opposed to cancer drugs or a hovercraft which also have patents opposed by those people who want them for free.
c) it’s similar to the sense of entitlement people have to products of the mind, e.g. copyright infringement of books, movies, music. People have a problem paying for something that’s a result of mental effort, as opposed to physical goods.
You often hear the argument against software patents that “all software inventions are obvious” or as your comment says, (1) logical information processing steps that are older than the hills (in spite of their “new” labels) and which do nothing more than what anyone would expect them to do”
If this assertion is correct for a particular case then the invention is obvious and not patentable. You don’t need a separate software exclusion. Simply state that these inventions are inherently obvious.
Why should software that is not obvious and therefore doesn’t fall within your definition be excluded from patentability?
The problem is that the software patents are on math, which is supposed to be excluded from patentable subjects.
That is not the problem jesse.
The problem is that you are not recognizing that applied math** is not excluded from patentable subject matter, and that you are only looking at the elephant blindfolded and holding the elephant’s tail.
You are also refusing to heed the warnings given to you as to being careful where you step. That is particularly unwise trailing the elephant.
**for simplicity, I am not even getting into the point that more than just applied math is involved.
The problem you have is not recognizing the math and software are the same.
It is pure math. It even has math proofs do show that fact.
You just don’t seem to like it.
Jesse – I think think that you are extending the proofs further than they were meant to go. The software-is-maths proof shows that a software algorithm can be equivalently represented as mathematics – of course that transformation leaves out the physical interactions caused by the software. Many physicists also believe that the entire universe (and anything within the universe) can be represented in mathematical form.
The “physical interactions” are the property of physics, not the math.
And within the limits imposed by the physics, anything in the universe should be representable… but that mapping is defined by the physics. Not the math.
My simple example is the measure of temperature and the math of the conversion…
F = 9/5*C + 32
Simple formula. Describes a line from -infinity to +infinity.
Simple math.
However, the physics of reality say it is wrong. Temperature doesn’t exist below –273.15 centigrade. (There is also an upper limit but that depends on the pressure of the media the temperature is being measured…)
As an abstract math formula, that is irrelevant.
Now within physics (until quantum mechanics) there is no such thing as a discontinuous function. All physical transformations are continuous.
Math doesn’t have that restriction. Simple number theory is discontinuous.
Computers use circuits that emulate boolean algebra expressions. I say “emulate” because the physics of the circuit requires the function to be continuous… Thus the emulation is limited to only certain physical ranges (for the given circuit). That emulation is then used to implement another totally abstract math operation – the instruction cycle.
Math operations are always discontinuous. There is no such thing as “time”, only sequence – one operation preceding/following another.
In reality, time is another continuous function. A computer CAN emulate reality… but not fully due to the discontinuous nature of the computations. It can come close… Sometimes close enough to substitute for a mechanism that actually performs continuously. (It happens to be one of the reasons we don’t have reliable fusion at this time – maintaining a stable plasma cannot be done in real time – even though the math CAN describe the situation… Trying to go through all the conversions from reality to math to reality is far, far too slow)
To me, the thing that separates the math from the physics is the conversion step. The math is independent of the conversion; but reality depends on the conversion to/from the symbolic form that can be used in math.
Within the computer, everything is just a number. Data, programs, input, output… just numbers. The only way for the math embedded within the computer to interact is by some form of translation.
That conversion must be done by something interacting with the physical world. And that device is certainly patentable. The keyboards, the display, the analog to digital conversion and the digital to analog conversion (and yes, even the disk drives have such conversion as well). Even the pencil and paper provides an external storage – but what is stored must still be converted to a symbolic form before use, whether that be a computer, programmer, or a mathematician.
To the computer, the physical conversions are usually referred to as reading or writing. What is done with that data (or where the data comes from) is irrelevant – it is just symbolic data (numbers actually), and that is the realm of math.
Interesting article here by Eli Dourado, who references the Lee article:
link to cato-unbound.org
Software patents have characteristics that make them particularly susceptible to litigation. Unlike, say, chemical patents, software patents are plagued by a problem of description. How does one describe a software innovation in such a way that anyone searching for it will easily find it? As Christina Mulligan and Tim Lee demonstrate, chemical formulas are indexable, meaning that as the number of chemical patents grow, it will still be easy to determine if a molecule has been patented. Since software innovations are not indexable, they estimate that “patent clearance by all firms would require many times more hours of legal research than all patent lawyers in the United States can bill in a year.
This is an important point and right at the heart of the matter. Any reform that allows software (and functionally claimed hardware) to continue to be protected by patents must include (1) a requirement that applicants in these “arts” adopt a pre-determined lexicon for describing, in the claim, the “precise elements” (to coin a term invoked by Dennis Crouch) responsible for the allegedly improved “functionality”; and (2) Examiner’s at the PTO must be given guidance as to how the massive prior art of logical information processing steps (computerized and not) fit into that lexicon.
>>in the claim, the “precise elements” (to coin a term invoked by Dennis >>Crouch) responsible for the allegedly improved “functionality”;
That is ridiculous. If a person of ordinary skill in the art knows how to write code/design a circuit to perform the function then it should not have to be in the specification.
So, let’s see according to this nonsense (and notice how they always have this authoritarian tone when making this suggestion and would never provide an actual example in real life as it would render them absurd) I would have to put in all the course work for ee/cs and maybe up to a ph.d. work. So, let’s see how about a patent application that is 100,000 pages long?
This is another attack on the whole system with these assertions that have no backing in patent law. In fact, these type of suggestions are why patents are so bloated and expensive.
It is scandalous that people have the audacity to make these types of suggestion. Even in Ned’s favorite Haliburton they did not make such a requirement.
I would have to put in all the course work for ee/cs and maybe up to a ph.d. work.
No, you wouldn’t.
If a person of ordinary skill in the art knows how to write code/design a circuit to perform the function then it should not have to be in the specification.
If the circuit is what you are trying to claim, it needs to be in the specification.
Yes I would.
“If the circuit is what you are trying to claim, it needs to be in the specification.”
One of these authoritarian vapid responses. A functional claim refers to what a PHOSITA knows.
Unbelievable to me that the anti-information processing crowd including Lemley are trying to push this point as if it makes any sense at all.
So, according to you MM if I claim a door rotatably connected to a frame then I better put in all the possible hinges.
That is your logic. Ridiculous. Counter to all known patent law principles.
Lemley should say, “I am reality.” Because from what I’ve seen of Lemley he puts a lot stuff out there but does not respond to anyone’s questions.
This functional claiming nonsense is just beyond believe.
“A functional claim refers to what a PHOSITA knows.”
That’s nice. We still require drawings and description around these here parts.
I didn’t realize you had been promoted to the SCOTUS.
Night, you are not making any distinction between claiming old and well-known structure in a combination claim and claiming new structure.
Why don’t you start with reading O’Reilly v. Morse in its discussion of Wyeth v. Stone, where in essence, the patentee tried to claim all machines that will cut blocks of ice having invented just one such machine.
The Halliburton case concerned what the Supreme Court thought to be a functional claim to novel structure. They first identified that there was nothing new in the patent but the new structure disclosed in the specification. All other apparatus was old, and the method disclosed was not new but for the apparatus.
In this context, the Supreme Court held that the functionally expressed novel element essentially violated section 112 because the patentees had not invented all apparatus for carrying out the recited function. You and I have talked about this in terms of written description support, enablement, or indefiniteness. The Supreme Court really didn’t care which part of section 112 that was the problem, if not all three.
So please stop trying to divert attention away from the issue by errecting strawman. The functional-point-of-novelty doctrine is all about claiming novel structure not in terms of what it is but in terms of what it does or the result it achieves.
Ned, if the structure is new then you would have to describe.
That doesn’t mean you can’t functionally claim it. LazardTECH for how claim scope should be policed.
Night, novel structure can be described as a combination of old, well known elements all of which are functionally claimed, or which include functional explanations of otherwise structural elements.
As I said, the vice occurs only when the claim is functional at the point of novelty such that the combination claimed is really old with the invention lying in the one element and not in any combination.
I think what you are saying goes to 103 and scope of enablement Ned.
Indeed, Night, it goes to section 112 primarily as a statutory basis for why there is a problem.
It should at least say what is in the “improved” door…
EXACTLY JESSE!! The patent application should focus on what is new and satisfy the 3 prongs of 112.
This functional nonsense is fabricated by Lemley.
Now if the “improvement” is not subject to patents then it should not be patented…
Might be reasonable as “trade secret” though…
“chemical formulas are indexable, meaning that as the number of chemical patents grow, it will still be easy to determine if a molecule has been patented”
So are sequences of 1’s and 0’s but nobody is on board with indexing them, despite their alleged importance.
MM, software as a category is and must remain essentially unpatentable. However, innovations in computer technology, and larger machines, systems or conventional processes might be claimed in terms of the larger machines, etc.
For example, an improved servo system in a disk drive where the novelty was in an improved algorithm might be claimed as an improved disk drive servo system. This would confine the search appropriately.
an improved servo system in a disk drive where the novelty was in an improved algorithm might be claimed as an improved disk drive servo system. This would confine the search appropriately.
In theory I can see something like this possibly working if a reliable lexicon and system for claiming and searching algorithms is put in place.
In practice, I have doubts. And I also predict the same folks who insist that the old status quo was “the only way” will fight tooth and nail against any attempt to set up such a system. That’s not even a prediction, at this point. More like an observation.
The reactions to the Alice decision are interesting to those of us who like to keep track.
In the immediate (and not unexpected) aftermath, proponents of software patents fell immediately into two major camps: (1) end of the world for computer innovation and massive portfolio destruction with a subsequent shareholder/stockmarket plunge; and (2) “no biggie because Alice is easy to claim around”.
Now, five to six months down the road, when we can see the first real signs of genuine progress being made in shutting down the “do it on a computer” gravy train (unfortunately no sanctions against the patentees for asserting the patents, but that’s surely coming), the same camps are still around.
What’s missing? For starters, any evidence of a dramatic decline in the production of new software or improved software, and any evidence of any effect on the stock market. Let me know if I’ve missed an important study somewhere.
It seems clear now that both of these camps were wrong. There are major and permanent changes taking place and patent law in the US is never going to be the same again, particularly when it comes to software patents, broad medical diagnostic patents on “correlations” and patents on methods of forming abstract relationships between actors and/or objects (e.g., certain business methods and social interaction methods).
Interestingly, some of the other major legal shoes poised to crush any software patents that might possibly survive Alice have not yet dropped …
“Massive portfolio destruction….”
Which companies have these massive portfolio’s of business method patents? The people who said this must have someone in mind. It might be interesting to name them publicly that we might know who has an agenda.
“… social interaction methods.”
If correct, 100+ of Facebook’s patents (and pending patents) would be effectively wiped out.
Is that what you want MM?
Might that not affect that company’s stock price / value?
not likely.
Facebook is oriented on people interaction. The patents are more just there for defense purposes.
Facebook is more likely to fail due to poor handling of privacy and PR.
The Bonfire of the Inanities, anyone? Rejoice I say, Rejoice!
“Next industry on the chopping block; biotech.”
Well said. You should care what the rules are, because it could be your ox that gets gored next.
My guess is that the SCOTUS very much would like to eliminate 99.9% of patent. They think that only patents to something like a t.v. or telephone should be eligible for a patent. Their view of technology is from Disney movies. And they have zero understanding of science.
Oh come on, they just now busted farmer Bowman (iirc) for “infringing” a patent on self-replicating bio tech.
I think what he was busted for is planting seeds that he believed to have the glyphosphate resistant trait, spraying the fields with glyphosphate, and replanting the next year with glyphosphate resistant corn. But what do I know; those are just the adjudicated facts.
“is planting seeds that he believed to have the glyphosphate resistant trait, spraying the fields with glyphosphate, and replanting the next year with glyphosphate resistant corn.”
Is that not a self-replicating bio tech? I’m confused.
No, that’s the absence of innocence in any of Bowman’s actions.
Sounds more like how corn was developed by the American natives…
By spraying their crop with glyphosphate?
Yes. The survivors, as long as there are any, will be tolerant.
Granted, it would normally have taken many years due to the time it takes for each generation.
But that is also why the seed corn currently sold is deliberately sterile. It creates a monopoly – which can be disastrous. If the source of the seed corn gets attacked by ANYTHING, then there will be NO corn for the next season… ever.
“that’s the absence of innocence in any of Bowman’s actions”
Are you trying to write a song? Enigma already beat you to that one. I think they got farmer bowman to star.
link to youtube.com
You should care what the rules are, because it could be your ox that gets gored next.
Neither you or anyone else has any reason to believe that a claim describing your new machine or component in objective structural terms that distinguish it from the prior art machines or components is susceptible to being “gored” by Alice. At least, you haven’t provided any such reason.
What you’re really doing is fearmongering.
News flash: your claims aren’t like “everybody else’s” claims.
That’s why we’re having the discussion.
>>News flash: your claims aren’t like “everybody else’s” claims.
That is not true. Look at mechanical claims and electrical claims. And, guess what? The patent system is supposed to adapt to new technologies not try to classify them as “abstract” or too spooky to work with. Or the latest craze as having no physical form with no structure and yet somehow existing.
except when it is specifically excluded.
Do you guys think this would make a suitable theme song for the War on Software Patents?
link to youtube.com
link to youtube.com
It seems to me that what is being discussed in the literature as “software patents” really involves at least three different kinds of inventions that really should be treated separately: business method inventions, software inventions involving a new algorithm hidden inside a computer program, and technology inventions that happen to be implemented in software. Admittedly the lines between these classifications are not bright, but bright lines are never perfectly clear in the law, and I think if we treat these three categories separately, they would be brighter than the so-called line drawn by Alice Corp.
Business method inventions, which I define for now in line with the AIA definition of CBM patents, should in my view never have been patentable. Three justices take this view in the concurrence in Alice Corp. In my mind, “promoting the progress of the useful arts” is not achieved by granting a monopoly on a financial services technique, an insurance product, or anything else that is not technology.
Software inventions involving a new algorithm hidden inside a computer program, I think might be worth patent protection because such protection would promote the progress of the useful art of programming techniques. However, the EFF people may be right that because they are hard for an innocent competitor to detect in their own products, such protection might be outweighed by the drag that they place on innovation by others.
Technology inventions that happen to be implemented in software, in my mind, should clearly be patentable subject matter. In the CBM arena, the PTO defines these as inventions that “solve a technical problem using a technical solution,” a definition very much in line with the Europeans. However I think it should be refined to require further that the invention have a physical effect, or be representative of a physical effect (such as a simulation program). These inventions are much easier for a competitor to detect and avoid (or design around), because of the physical effect. And with processors becoming so fast, most new ideas that in the past might have been implemented in hardware are being implemented in software instead, and this should not by itself be a reason to deny them patent protection.
In the very first Federal Circuit decision after Bilski, Research Corp. v. Microsoft, a patent on a halftoning method was upheld as being “non-abstract”. Judge Rader wrote the opinion and based it on the fact that a halftoning method is technological, not abstract. Subsequent Federal Circuit opinions which invalidated what I would call business method patents, distinguished Research Corp. on that same basis. I personally believe Judge Rader was on the right track.
In light of my views above, of the 11 recent invalidations mentioned in the article cited in the main blog post, the only one that troubles me is the one about on keeping colors synchronized across devices. That one I believe was wrongly decided, because the invention was, in my view, “technological”. I am comfortable with all the other invalidations cited.
I wonder what others in this brain trust think of the above way to organize the subject matter eligibility analysis of so-called “software patents”. (Please, as Dennis has requested, constructive comments only.)
>use of these profiles were merely mental steps that could be done by a >human being and were therefore not eligible for patent protection
I think they were all wrongly decided, but I agree that the above is one of the worst. So, the judge is saying merely being able to do what a human brain can do is nothing. So, a machine that can perform the same information processing tasks as a human is not worth a patent. That is what the judge is saying.
Outrageous. Counter to current science and every principle of innovation (and patent law).
Personally I think these opinions are right out of a Marx movie with no sanity clause.
Night, Personally I think these opinions are right out of a Marx movie with no sanity clause.
Typical Night post — They quality of argument is, well…..
So, if a program could do the judge’s job as well as the judge, then even that according to the judge would not be patent eligible. Something is very wrong with that.
if a program could do the judge’s job as well as the judge, then even that according to the judge would not be patent eligible.
Something is very wrong with that.
Exactly what is “very wrong with that”?
Lots of incredibly useful stuff is ineligible for patenting, for one reason or another, including stuff that would improve the performance of judges. There’s more to it than that.
Well gee willikers, tell us why it isn’t wrong. So far what we have from the judiciary is that it is wrong ’cause we say so.
So far what we have from the judiciary is that it is wrong ’cause we say so.
I’m not aware that the judiciary ever spoke to the facts you raised. Do you have a case in mind?
You made the assertion that “something is very wrong with [a judge finding ineligible] a program [that] could do the judge’s job as well as the judge.” I didn’t. Tell us what that “something” is. Presumably you know or else you wouldn’t have made the statement.
The “something” is that making an intelligent machine is an invention. The “something” is that we live in the information age and the innovations that are happening now are processing represented information.
That is the something. Processing represented information is why we separate ourselves from the animals, but according to at least one judge it isn’t patent eligible. I would say that is absurd.
Furthermore, any of these statements that go something like merely performing the steps of a mental process are not patent eligible are absurd.
So, if a person comes in and can do the judge’s job better is that meaningless? Is that mere? I don’t think any reasonable person would think so. So, what we have is a machine that is performing what the judge holds most dear and what the judge is paid for. It is simply outrageous and absurd to say that isn’t patent eligible.
Night, he didn’t say “meaningless”.
He said not patentable.
Jesse: tell us what he decision was based on. He applied a narrow judicial exception that states the claims are nothing more than an abstract concept with no value to the application.
I agree, NWPA.
That still shouldn’t make it patentable.
If it were, then you should have patented having a baby.
ALL animals process information. You are REALLY living in left field if you think animals don’t process information.
You must have missed the communication capabilities done by gorillas and chimpanzees. Their language might have been limited to a three year old, but it is still communication with a language.