decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books

Gear

Groklaw Gear

Click here to send an email to the editor of this weblog.


You won't find me on Facebook


Donate

Donate Paypal


No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.


What's New

STORIES
No new stories

COMMENTS last 48 hrs
No new comments


Sponsors

Hosting:
hosted by ibiblio

On servers donated to ibiblio by AMD.

Webmaster
Curing the Problem of Software Patents, by Michael Risch
Sunday, June 10 2012 @ 06:44 PM EDT

Introduction, by PJ

I have a real treat for you, and also a serious assignment, if you are up for it. Professor Michael Risch has been good enough to write an article for us at Groklaw on the subject of software patents. He thinks he has some ideas on how the problems they pose could be ameliorated even without changing the law.

Will you please treat him as my personal guest? He is a member of Groklaw, actually, as mrisch, so in that sense you are colleagues. He and I have been in quite a discussion about this topic, ever since I wrote the article about the Prometheus case, where the US Supreme Court referenced two of his articles, Everything's Patentable, which they disagreed with, and Life After Bilski, which was relied on heavily by the court. Over time, I realized I really like him as a person, even though we don't agree on this topic altogether, and one day as we were emailing, I thought to myself, why am I not sharing this with all of you? You know more than I do about software patents, and it's a real opportunity for you to explain your point of view to him and vice versa. If patent law includes a specific reference to those skilled in the art, in the category of software patents that would be you. Surely your voices should matter. In my dreams, he ends up influenced to some degree by Groklaw, and then goes on to influence the US Supreme Court in a way we might actually like. If not, I still like *him*, even if we never agree. He plans to present the ideas in this article as a solution to the problem of software patents at a conference this fall, so he is very interested to hear from you. And he does know how to program, so you can speak freely and specifically, and he'll understand you. So with that introduction, I give you Professor Risch and let the conversation begin.

****************************

Curing the Problem of Software Patents
~ by Michael Risch1

I would like to thank the Groklaw community for having me, and in advance for your comments. I’ve read posts here on and off over the years, and the recent discussion about Prometheus and Oracle v. Google piqued my interest. Before I get into the meat of things, note that I have a lot of experience in software, as I discuss below. So, if you disagree with me (as I suspect many of you will), please don’t assume that it is because I don’t understand software – I do.

Many people who have read my work probably assume that I am pro-software patents. I’m not really; I am just not anti-software patent. I’m agnostic.  I think that there are a lot of bad software patents out there, and we should get rid of those. However, I don’t think that the solution is to get rid of all software patents. I think that such patents can be inventive and productive like any other, and I also think that defining when a patent becomes software (as opposed to hardware) is too difficult an inquiry.

Abstractness and Software

I’ve read a few postings here about how all software is really just mathematics and thus abstract (I’m paraphrasing, of course). And indeed, I’ve written in an article: “In software patents, hardware is almost never the inventive aspect.”  But note this comment I got at PatentlyO from an engineer/patent lawyer in response to the above statement: “ALL software is a specification of hardware. ALL software is run very differently on different hardware and is written for specific hardware platforms. Software is hardware in every sense that matters for patents.”

Here was my response: “Of course you are right that software has to be written and compiled for specific hardware. Indeed, the real ingenuity of most software patents is that they attempt to do something that is really difficult given the available hardware at the time of the invention. But so what? No patent is written at that level of abstraction. If they were, then the Court [in Bilski v. Kappos] would likely not have worried so much. The issue is that patents aren't for specific implementation of specific solutions on specific hardware. It's that patents claim general steps to solve general problems (or sometimes specific ones) on general hardware.”

So, you can see that I’m in sort of a middle-ground that makes no-one happy. Anti-software patent folks think that I give too much credit to the non-abstract aspects of software. Pro-software patent folks think that I too easily see software as abstract. As we like to say at home when the kids have to compromise: if no one is happy, then I must be doing the right thing.

An Example

I thought I would give an example of a patent I find potentially meritorious. Note, of course, that the patent still has to hold up to a novelty analysis, etc., but Patent 7,098,896 [claims] for continuous stroke input on a keyboard is the type of software patent that the system should protect if it is otherwise valid.

This patent, owned by Swype, was filed in January of 2003. It claims at its most basic the idea of moving the finger from letter to letter on a qwerty keyboard. The motion of the finger is then captured (the shape of the line), and matched against a dictionary of shapes for different words. The system also compares for close shapes and suggests words that might have been intended but the finger didn’t quite get to the correct keys.

Is this math? Absolutely – it encodes data about the touch and does a mathematical comparison. But it is so much more; it provides an interface from my finger to the touch keyboard to the words on the screen in an incredibly efficient and useful way. I connect with this patent in part because it is the only way I’ll type on a touch keyboard – the other methods just don’t work with my big fingers, and I stubbornly held onto a Blackberry for a long time believing that I could never effectively type on a touch keyboard. It may be math, but it is math applied to a very specific hardware problem – small touch keyboards.

So, does this impede innovation? I don’t think so – continuous stroke typing was a twinkle in the eye of researchers at the time this was invented. The patent cites the leading work at the time, and that work was not nearly as easy or useful as this product. The idea, once disclosed, was easily copied by competitors, and the only way to ensure early development of it was to protect it with a patent. I suspect the patent helped the company get venture funding, and thus survive long enough to complete development. Furthermore, there are ways to work around the patent to allow for continuous stroke typing – this is only one specific way. Of course, it would be better if they invented this, disclosed it to the world, and then never enforced the patent, but that’s not how the world of incentives works.

Now, you may not agree with me that this is a worthwhile patent, but I believe that it fulfills the basic role we would expect of all patents. If you think that patents in general don’t do that, it’s unlikely that I will convince you otherwise.

Patentable Subject Matter

Now onto the legal perspective; let’s start at the beginning. First, software can be patentable subject matter. Why? The real question is: why not? It is either part of a machine that does something useful, or it is part of a process for a machine to do something useful. While information processing is certainly more complex (and abstract? I’m not convinced) than primitive programmable machines like the Jacquard loom, at bottom there is no principled reason to differentiate as a matter of subject matter. Even if patenting of methods was unclear in the distant past, Congress made clear that machines, new uses of machines, and processes are all patentable.

Of course, one might except “abstract ideas” from patentability, but as I argue in my article Everything is Patentable, figuring out what is abstract or not is nearly impossible, especially in light of the fact that “insignificant postsolution activity” is to be ignored. In my co-written article Life After Bilski, we tried to come up with a way to define an abstract idea, but I still think it is pretty difficult to apply consistently.

Solving the Problem of Software Patents

Does all this mean that I don’t think software patents are a problem? Of course not. Many (most?) software patents are terrible. Indeed, I spend a lot of time finding prior art to invalidate software patents, and more often than not I find something that is pretty close, if not identical to the patent claims. There are many other problems with software patents. I just don’t think that subject matter is the way to deal with them. Instead, I think the better course is the one I propose in Everything is Patentable (and Forward to the Past): more rigorous application of the traditional patentability criteria.  I believe most of these techniques are already available or in reach of current rules.

Some of these solutions are best applied by patent examiners at the time the patent is being sought, but the reality is that we will never have enough money to fully test every patent application. For that reason, I’m a firm believer that for many disputes it is more efficient for the parties that have a stake in the matter to fight about validity at the time of litigation. This increases costs on patent defendants, but if examiners will not get the job done anyway, I would rather they not spend a lot of time trying.

I’ll explain what I mean by more rigorous application.

Utility: To be patentable, a claim must be practically useful – it must provide a specific and substantial benefit to the public. In my articles Everything is Patentable and A Surprisingly Useful Requirement, I argue that this requirement means that the claim must “do something.” As a result, patents that are too abstract likely lack practical utility. They don’t do anything. I argue that Bilski’s first claim for hedging fits this description in Bilski v. Kappos: sell a commodity to one group at one price and to another group at another price. There is no limitation for type of commodity, how to calculate pricing, which groups of sellers, or even who is supposed to benefit. By being so broad as to be everything to everyone, Bilski’s claim really covers nothing specific. This application of practical utility to overly broad, abstract patents is already within the bounds of the law, I believe, even if it is not being applied that way very often.

I realize the Supreme Court has said that algorithms are unpatentable. My view is that this is an untenable solution, because every process patent is an algorithm, mathematical or not. And it can't be that every process is unpatentable. I don't think the dividing line can be software, because so many solutions use electrical signals as part of the product or process. It is easy to rail against software, but it is far more difficult to determine what is and is not software. So I look for a different dividing line. Here is a lengthy quote from the article I wrote, A Surprisingly Useful Requirement, that discusses this issue:

Practical use analysis can help answer whether a seemingly intangible process is patentable or not. Bilski is merely the latest in a line of cases considering this question. For example, in Gottschalk v. Benson, the Court disallowed a computerized method whose purpose was to simply convert numbers from one format to another. Interpreting Gottschalk on subject matter grounds has caused much confusion, but the opinion can be better explained by usefulness. The method at issue did not do anything – it was an isolated series of calculations without any inputs or outputs applied to a real world problem.

Thus, it had no practical use because the process standing alone did not provide a specific and substantial benefit to the public. It surely could have been incorporated into a number of practically useful applications, but where a calculation can be used everywhere, including simply for calculation’s purposes, it lacks practical usefulness.

Similarly, a process to calculate the result of Einstein’s famous e=mc2 lacks practical use, while a nuclear reactor using the same process might have practical use. Thus, if such processes were applied to solve some practical problem, they might be useful. In general, a series of steps that achieves no end is not practically useful, and is therefore unpatentable. . . .

[P]rincipled consideration of a claim’s usefulness is preferable to an ill-defined “abstract idea” test because ... the tens of thousands of intangible method claims that need to be tested by the PTO and courts [might or might not be abstract] and an undefined test has no way to determine which ones are abstract or why.

Novelty: The heyday of software patents involves filing dates before 1998 or so. It is much more difficult to find prior art before that time. Further, there were also PTO limits on using search engines. Current software patents do not have the benefit of limited search resources. As a result, it will be much easier to find prior art for software patents in the future. Furthermore, companies such as Article One Partners provide crowdsourced prior art  searching that has aided finding prior art.

Obviousness: My experience with computer scientists is that they believe everything is an obvious improvement over the past. I’ve read enough engineering journal articles to know that this is not true, but many software patents are obvious.

The Supreme Court case of Dann v. Johnston is significantly underutilized. That case held that it is obvious to implement in software something that was already well known in noncomputerized circles : “[Patentee’s] ‘category code’ scheme… is, we think, closely analogous to a bank's offering its customers multiple accounts from which to choose for making a deposit or writing a check.”. (A little known fact is that the Supreme Court also agreed to hear a subject matter challenge, but ruled only on obviousness). This means that many software patents, such as “slide to unlock” are obvious.

However, a fundamental problem with obviousness is the growth of the Federal Circuit’s rigid “teaching, suggestion, and motivation” test during the 1990’s: “Seeking to resolve the question of obviousness with more uniformity and consistency, the Court of Appeals for the Federal Circuit has employed an approach referred to by the parties as the ‘teaching, suggestion, or motivation’ test (TSM test), under which a patent claim is only proved obvious if ‘some motivation or suggestion to combine the prior art teachings’ can be found in the prior art, the nature of the problem, or the knowledge of a person having ordinary skill in the art.” (See KSR v. Teleflex).

This rigid test made it much harder to find software obvious, because it barred examiners from using common sense, instead requiring that every difference between existing technology and the patent be suggested in writing somewhere. This rigid test helped create a large volume of software patents. The test also meant that finding a software patent obvious required courts to imagine software engineers as superskilled experts so that they would know to combine specific references. As noted below, this superskill hurts the enablement analysis.

Obvious is not so hard today. Today, the Supreme Court’s 2007 decision in the KSR v. Teleflex case has made it easier to use common sense to find patents obvious: “When a work is available in one field of endeavor, design incentives and other market forces can prompt variations of it, either in the same field or a different one. If a person of ordinary skill can implement a predictable variation, § 103 likely bars its patentability.”

This also allows us to reduce the skill level of our ordinary software engineer while still finding patents obvious.

Enablement: Many software patents do not enable the ordinary engineer to make and use the invention as required by the patent statute. For broad claims, this means that the patent must teach not only the specific implementation, but also every implementation available at the time of the filing. Usually, the filing date is very early; otherwise, the patent would be obvious. Such broad patents may enable particular solutions on particular hardware platforms, but limited technology available at the time means that one can’t enable the general, abstract solution on any platform available. However, because engineers were considered superskilled, they were deemed to be able to fill in the gaps to implement broad claims. This should be dialed back a bit by assuming that software engineers are not as skilled as previously thought, and that a specific hardware solution does not teach one how to implement the solution on all the hardware available at the time.

Written Description: More important than enablement, it is often not clear whether inventors are in full possession of their broadest claims, even if a skilled engineer could fill in the gaps. A recent case, called Ariad v. Eli Lilly, reinvigorated the written description requirement. Under that ruling, the ordinary skilled engineer must see enough in the specification to believe that the inventor has actually invented the broadest claim scope: “We held that a sufficient description of a genus instead requires the disclosure of either a representative number of species falling within the scope of the genus or structural features common to the members of the genus so that one of skill in the art can ‘visualize or recognize’ the members of the genus.”

Patentees often do not disclose all of the species of broad software inventions. Usually, the most one can expect is that the inventor solved a specific problem in specific hardware, and does not disclose enough general principles to show that he/she understood all ways to implement the particular claim – especially where there is no source code or pseudo source code.  The reality is, however, that there is usually some new technological breakthrough that makes the patent possible, like cheaper RAM, faster networks, etc. Requiring more detail in the written description might flesh this out, and make obviousness easier to find. Written description was, for example, the downfall of many patents held by Ronald A. Katz (see, e.g., at p. 27): “Instead, the court held that the specification fails to describe the step of ‘visually displaying customer number data’ because the only descriptions of visual display in the specification involve information that was not entered by customers.” In short, Katz tried to claim something that he didn’t actually invent at the time.

Best Mode: Patent reform (the America Invents Act) actually hurts when it comes to software. A major failing of software patents is that they do not disclose the best way of practicing the patented claim as is required by the Patent Act: “The specification shall … set forth the best mode contemplated by the inventor of carrying out his invention”. 35 U.S.C. § 112. Inventors often don’t disclose specific hardware, and they certainly don’t disclose source code.

This disclosure used to be a basis for invalidating a patent in litigation, but the AIA removes best mode as a defense in litigation, including inequitable conduct for intentional failure to disclose a best mode. However, disclosure of best mode is still required in the initial patent application. This means that a) the PTO should be asking if patentees have a best mode to force the issue, and b) the PTO should be aggressively considering the issue in reexamination proceedings.

Definiteness: There’s not much to say here; software patents are often impossible to understand. Examiners and courts should demand more definiteness, as required by the statute: “The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.” 35 U.S.C. § 112.

PTO Claim Construction: The PTO uses the broadest reasonable construction rule to decide whether a patent is valid. This is different than the way claims are interpreted in litigation, which is the way someone skilled in the technology would understand the patent. This is one reason why reexamination can lead to invalid patents in a way that litigation does not. The idea is that if you interpret each claim as broadly as possible, then there will be more prior art to apply to it and it is more likely that the written description is not sufficient to support the broadest claims. Examiners should use the broadest reasonable construction rule to expose broad claims as old, obvious, or ill-described. The problem is that many examiners have not used this rule to their advantage in the past.

I think that each of the above applications is achievable under the law as it stands, or at most a small stretch from current law. By rigorously pushing these points, opponents of software patents can make sure that only the few meritorious ones make it through without eliminating the incentives created by the patent system. Your collective thoughts are appreciated, especially because I’m presenting this as a solution to the problem of software patents at a conference this fall.

Finally, as promised above, here’s a bit about my own background. I started working with computers when I was 10 (and the elementary school’s only computer was an Apple II in the principal’s office). I started writing computer programs when I was 13 and my Dad gave me his Commodore CBM – a brilliant child-rearing decision, I have to think. I recall the dual floppy disk drive costing more than $1,000 (ouch). While I didn’t major in engineering in college, I took courses in math, statistics, computer science, symbolic systems, and neural networks. I’ve read on the subject as well; “Godel, Escher, Bach” is one of my favorites. I currently program a bit in a variety of languages, but not nearly as often as I used to.

_____________
1 Michael Risch is an Associate Professor of Law Villanova University School of Law. Professor Risch’s teaching and scholarship focus on intellectual property and cyberspace law, with an emphasis on patents, trade secrets and information access. His articles have appeared in the Stanford Law Review, Indiana Law Journal, George Mason Law Review, BYU Law Review, Yale Law Journal Online, Penn. Law Review’s PENNumbra, Tennessee Law Review, and Harvard Journal of Law & Technology, among others. His work has been cited by the U.S. Supreme Court. Prior to joining the Villanova faculty in 2010, he was an Associate Professor at the West Virginia University College of Law, and an Olin Fellow in Law at Stanford Law School. Professor Risch graduated from Stanford University with honors and distinction in public policy and distinction in quantitative economics. He earned his law degree at the University of Chicago, where he graduated with high honors and was an Olin Fellow in Law & Economics and a Bradley Fellow in Law & Economics.


  


Curing the Problem of Software Patents, by Michael Risch | 1347 comments | Create New Account
Comments belong to whoever posts them. Please notify us of inappropriate comments.
Corrections here please
Authored by: jesse on Sunday, June 10 2012 @ 07:06 PM EDT
Thank you.

[ Reply to This | # ]

News Pick discussions
Authored by: jesse on Sunday, June 10 2012 @ 07:07 PM EDT
Than you.

[ Reply to This | # ]

Off topic discussions
Authored by: jesse on Sunday, June 10 2012 @ 07:07 PM EDT
Thank you

[ Reply to This | # ]

COMES thread
Authored by: jesse on Sunday, June 10 2012 @ 07:08 PM EDT
Thank you.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Sunday, June 10 2012 @ 07:16 PM EDT
Here is an ideological argument.
Patents are bad. Patent is an artificial monopoly given by the government to a
specific individual or company and allows the use of government force to
maintain itself.
Artificial monopolies are bad. The idea that something is "good for the
public" or "bad for the public" relies on completely arbitrary
definition of "public" which does not exist. Members of the
"public" have the different ideas about what is good for them and what
is bad for them. There will never be a smart-godlike-man that can tell what is
"good for the public" and be right about it.

Patents are bad.

[ Reply to This | # ]

all software IS math - software patents need to die
Authored by: Anonymous on Sunday, June 10 2012 @ 07:17 PM EDT
there's no getting around it.
all software IS math.

it shouldn't be patentable, ever, no matter what. period.

software patents need to die.

if you agree with the arguments for inventiveness, then you must also agree that
inventive math must be patentable too.
gene sequences and such should also not be patentable,
and neither should "business processes"

[ Reply to This | # ]

The first example is bad
Authored by: jesse on Sunday, June 10 2012 @ 07:19 PM EDT
...it provides an interface from my finger to the touch keyboard to the words on the screen in an incredibly efficient and useful way.

The mathematics does none of these things.

1. the interface from your finger to the touch keyboard is hardware. Once the "touch" is converted to symbolic notations it should no longer be in the realm of patents because the symbols are manipulated by the mathematics.

2. The "words" on the display are not words - they are a light pattern that generated by the hardware. These light patterns are identified by mathematics, which again should not be in the realm of patents. The device that translates the symbolic values into light patterns is something that should be patentable.

The hardware should be patented. But not the mathematics. That is nothing but abstract symbolic manipulation.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Sunday, June 10 2012 @ 07:21 PM EDT
Software is both utilitarian and elegant, programmers create both art and value
through programming by making things work with the sweat of their time and
effort.

Since I can program nearly any requirement given enough time and understanding,
by what right does someone come along and say they can own, monopolize
something, an idea that I have developed from scratch using my own expertise in
order to feed my family.

If I am sufficiently capable, skilled in the art to independently develop an
idea from scratch to working code it should not be patentable ever.

Copyright is sufficient to protect and monetize (if that is your requirement)
third party code that I may or may not use to help reduce the time taken to
create my product.

rgds

[ Reply to This | # ]

software has to be written and compiled for specific hardware
Authored by: Wol on Sunday, June 10 2012 @ 07:58 PM EDT
WRONG WRONG WRONG

And it's been wrong pretty much since the dawn of the mainframe back in the
sixties. And the plethora of virtual machines spawned in the seventies.

Given the fuss we've had with the OraGoogle lawsuit these last few weeks, surely
that's obvious? The whole point of that lawsuit was that Oracle were upset that
Google had allegedly subverted their plan to make Java run on any hardware. But
let me set you a little test. Tell me what hardware a Java program is meant to
run on - any Java program - without giving me wriggle-room to argue "well I
can run it on something else completely different". It can't be done.

Let's say I write a Java program. It runs on a JVM. Which I might have installed
on Windows. Which runs in VirtualBox. But I don't have a x86 chip so that's been
installed in Qemu or Bosch or somesuch. Which might be running on a Sparc. Which
is a microcode CPU so the published instruction set bears no resemblance to the
real hardware instruction set. Which itself might be yet another program...

At NO POINT WHATSOEVER can you point at it and say "that level must be
hardware".

And indeed, though I've never used an IBM mainframe, I understand that all the
old 60s code written for OS/360 and its predecessors still runs fine. It may run
on an emulator, that runs on another emulator, that runs on yet another
emulator, but you can take an old mainframe program and run it on the latest
Z-series without any problem whatsoever.

Let me give you another example, from a programmer we all revere :-) Linus
himself is on record as saying that Linux is not written to run on any
particular hardware. How odd! That's an OS, who's primary purpose is to
interface with hardware and provide an abstraction to the programs on it! Yet
Linus is on record as saying it is more efficient, and produces much better
code, if he programs to an idealised abstract hardware! If memory management,
for example, works best with a 3-levels-of-indirect manager, then he programs to
that abstract model and writes a shim underneath it to hide the fact that say
x86 only has two levels.

You mention the Jacquard loom in passing. And totally ignore the point we've
brought up repeatedly in connection with that (and the pianola). Are you arguing
that changing the punch card set that made the loom create a different pattern,
or the pianola play a different tune, is actually a new machine worthy of a new
patent? Because if you aren't, then nor is a general purpose computer with a new
program worthy of a new patent.

Look at programming this way. You're a student in a maths class. Your professor
sets a maths problem on the board, and you write a twenty-page solution in your
notebook. Is that solution worthy of a patent? Now imagine you're a CS student.
Your professor sets a programming problem on the board, and you write a
twenty-page program in your notebook. Is that worthy of a patent? And NO MATTER
HOW a compiler translates that into assorted machine codes, the principle of
equivalence states that the binary is mathematically identical to the source.

Now if you think that sticking a different piano-roll into a pianola creates a
new, different, patentable pianola then you can argue that sticking a different
program into a general-purpose computer makes a new, different, patentable
computer. But I don't think many people here will agree with you.

Cheers,
Wol

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Sunday, June 10 2012 @ 08:13 PM EDT
First, Thank You for sharing you thoughts with us.
If we accept the premise that a novel, utilitarian idea which is implemented as
a stored program on a computing device is subject matter for Patent, Your
guidelines might stem the flood of useless, overbroad patents. Although I would
prefer all software (see definition above) be declared not subject matter for
patent. However we must be realistic. I expect legislation to repeal software
patents unlikely. I despair when I see old ideas applied to new technology
receive approval by the patent office. Slide to open/close is a prime example.

My other complaint is that so many ideas are cloaked in language which totally
prevents any useful teaching, which is key to the value of the Patent system. I
can't point to a specific example now, but there have been patents asserted
against defendants that When I, allegedly one skilled in the art, read, I had no
idea what practical application could be made of the invention. Then to see is
asserted against code which performed tasks which seemed to be only vaguely
related to the description of the invention. The Patent examiner is supposed to
be standing in the place of the intended audience. They demonstrably are not
representing the interests of the global community that are working in the
field. I think that we need to convene a panel of "experts" to review
samples of approved patents, and if this panel finds the descriptions unusable,
and non-specific, the Patent examiner should be placed on probation, subject to
removal if they continue to approve this useless drivel.
I read you article, and found the suggestions a positive approach to improving
the system. However, I fear that they will never be implemented. I have
written my Senators several times, and each time I have been thanked for my
input, and never felt their response suggested they were sympathetic, or that
their staff understood the issues.

Thank You again

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: tknarr on Sunday, June 10 2012 @ 08:13 PM EDT

Your Swype patent brings up the basic issue, though. Take a look back at patent 5596656 for Unistrokes, the precursor and inspiration for Palm's Graffiti. Once you have the basic mechanism of recognizing a shape and matching it to a symbol, it's not a large leap from your symbols being individual characters to your symbols being entire words. The compression algorithms used by PKZip, gzip and the like are based on doing exactly that. I'm afraid I have to be unimpressed with the Swype patent. It's a great idea, but the patent itself seems to handwave away significant prior art that the applicant surely knew about.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: charlie Turner on Sunday, June 10 2012 @ 08:16 PM EDT
Welcome to groklaw, Professor Risch! Civil discussion on this issue (as well as
almost anything in society today) is somewhat lacking. I hope we all can learn.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Sunday, June 10 2012 @ 08:31 PM EDT
My own experiences taught me that, every software has bugs. So does patents.
I've seen convincing patent turns out to be claimed by ignorant, and so ignorant
that he may not even has consulted document. I always relies on source code and
benchmark to form my decision. Anyway, fixing bugs can be more important than
claim anything new. "If it's not broke, don't fix it."

[ Reply to This | # ]

Than You for your Views
Authored by: Anonymous on Sunday, June 10 2012 @ 08:44 PM EDT
I have more than once tried to state a rule for patentability that gets around
VHDL code for an ASIC or a FPGA. The closest I can come is the patent would be
for the exact configuration of gates and hardware, not the input/output of the
device. I find that insufficiently clear or meaningful.

You acknowledge the huge body of bad patents for software. That is a problem
that is causing real harm right now. How do we fix that?

An issue with patenting software is that computers are finite machines.
Programming languages are finite. Given a particular problem, it is often
required to develop a solution that is very close to the original work. The
breadth of software patents is the devil in the details here. I am very
uncomfortable with anyone owning the input/output of the problem. It is like
someone patenting the mousetrap, not A mousetrap, but anything that can catch a
mouse. With software patents, that is often the effect.

I would like to see you address the techniques for delaying the onset of the
patent time. A patent filed years before it is granted and starting the 20 year
clock.

There is also the problem that you partially address for finding out if there
really is a patent out there that does what I want to do. The current
descriptions and specifications are impenetrable to this engineer. Even if I
wanted to search, I am unable.

Your approach may be workable, over time. I really object to someone claiming
ownership of the thoughts and processes of my mind. I like the idea that if you
can do it with a pencil and a piece of paper, it is not patentable. We create
and innovate based on the work and inspiration of others, science has always
advanced by that principle. A physical device is one thing, but my way of doing
things should be out of bounds. Even if I am not sure how to say that legally.

-- Alma

[ Reply to This | # ]

new uses of processes ?
Authored by: Anonymous on Sunday, June 10 2012 @ 08:45 PM EDT
Congress made clear that machines, new uses of machines, and processes are all
patentable.

So the Congress did not say "new uses of processes"
A computer is a invention on it own meant to move around
bit values and that is all it does and that is already the process.

If you buy a computer you have already payed for all the
inventions you can have in a computer.
It's like having a scissors and then patenting cutting
paper with it;

Oops now i have fallen back to "new uses of machines"
But that is because if you have to explain software to
people who think they know what software is,
you have to switch to other meta fores.

/Arthur



[ Reply to This | # ]

The problem is money and we should fight battles we have a chance of winning.
Authored by: Kilz on Sunday, June 10 2012 @ 08:56 PM EDT
The patent office takes in money for the patent. In the
entire history of this country no part of government that
takes in money has ever been completely done away with. The
government will not let it happen because it allows people
in power to continue to keep their job.
It is much easier to put rules in place to regulate the
intake of the money by making it harder to get a patent.
IMHO the amount of useless software patents is the bigger
problem, and one that has a chance of being changed. Dont
hold your breath waiting for software patents to be stopped.
While I know that the people who say that all software is
mathematics are right and they shouldnt really exist. I also
know that the idea that software patents disappearing in our
lifetime is almost an impossibility. Its like trying to make
congress balance the budget.
INMO it is better to focus energy on whats possible than
whats not going to happen. Professor keep fighting on this
front. We need to make some headway and to me we are loosing
ground every day as people patent more and more of the
things that should never be patented even if software
patents exist.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Sunday, June 10 2012 @ 08:58 PM EDT
Good article, Michael, but I'm not convinced software
patents are necessary. I am convinced that changing the
system may be impossible. A few things could be done to
improve the situation:
1. Reduce the term for software patents to one year.
Software can be monetized quickly, much more quickly than
hardware.
2. Eliminate patent trolls. S/w patents could be sold only
to practicing entities. Trolling _never_ contributes
anything useful to society.
3. Require source code for all s/w patents.

Regarding your 'swipe' example. It's a software solution
for a hardware problem. How to decode 2 or more letters
when the finger slides from one to the other. The fact that
a 'pattern' is deduced is trivial, as is the 'lookup'.
There are codes that resolve to letters, just as if you
typed them individually. The 'shape' business is an
unnecessary extra step. We implemented the 'lookup' feature
in 1986. Isn't it called 'auto- complete'?

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Sunday, June 10 2012 @ 09:02 PM EDT
What on earth is a "non mathematical algorithm"? Is it
something like non wet water?

[ Reply to This | # ]

Copyright vs. patentability
Authored by: josmith42 on Sunday, June 10 2012 @ 09:14 PM EDT

I understand you arguments as this: software patents in and of themselves aren't necessarily bad, it's just that a bunch of bad software patents have been granted. I don't disagree, and your arguments are compelling.

However, there is one argument against software patents you failed to address: my understanding is that software is the only thing in the world that is protectable by both copyright and patents. How can we need patents, when copyright is in full force as soon as you write the code? That one argument always gives me pause whenever I start to think software patents are a good idea.

---
Always a lurker, so you will never see this signature. But you are. What?

[ Reply to This | # ]

You lost me at 'fight about validity at the time of litigation'
Authored by: SpaceLifeForm on Sunday, June 10 2012 @ 09:19 PM EDT
You make an excuse for the USPTO not 'getting it',
and allow damage to occur later.

That is not acceptable.

Software is based on math and logic.

Math and logic were not invented, they just exist.

All software, no matter how creative, is still
based on math and logic, that no one invented.

Condoning 'software patents' is condoning attacks
on creativity, and all homo sapiens should be allowed
to be creative in their own way, without having to
endure an attack by someone that found the patent office.

Sorry dude, you don't get it. Your support of 'software patents' is part of the
problem.



---

You are being MICROattacked, from various angles, in a SOFT manner.

[ Reply to This | # ]

Mr Risch, can you give me an example of code that is patentable?
Authored by: jsoulejr on Sunday, June 10 2012 @ 09:32 PM EDT
n/t

[ Reply to This | # ]

Abstractness and searchability of prior art are big problems with Software Patents.
Authored by: Anonymous on Sunday, June 10 2012 @ 09:36 PM EDT
Thre are specific reasons why software should be treated
differently from other patentable fields.

1) With software, unlike for example a mouse trap, the
patentable idea is abstracted in the code. In order to
search for prior art, you have to decompile the code to get
to the idea. This is usually requires more effort than
"inventing" the idea in the first place in the case of 99%
of the software patents granted. In this context, it makes
no sense to grant patents because the search for defending
patents is more expensive than the invention being
protected. Searching prior art in software patents is more
miss than hit because of this.

2) Most ideas in software are practised but not patented.
The vast amount of prior art and known practice which is not
patentable is not expressed in software patents, but instead
embedded in actual code, in an abstract format which cannot
be readily searched. In addition the majority of software
code is proprietary, and thus cannot be searched. It is
ridiculous that patenting of subject matter that cannot be
properly searched due to abstraction and non-availability of
most of the existing art.

3) The same life is granted to software patents at to other
patents. The purpose of granting an artificial time limited
monopoly (also known as a patent) is to encourage early
implementation of an invention in order that once the patent
lapses, others can also implement the patent before its
usefulness life expires. In other words, patents are not
intended to grant a monopoly for the whole useful life of an
invention. In the case of software, development is very
rapid. The grant of a 20 year patent life is a grant of a
monopoly for an eternity, well beyond the useful life of the
invention as a result. For example, think where software was
20 years ago.

4) Patents are a crude form of IP protection, which is
necessary for many areas of development because there is
nothing better. Software has a better and more refined form
of IP protection called copyright, and therefore does not
need patent protection. Indeed it developed much faster
before patents were allowed. Patents on software in fact
slowed down the pace of innovation - the opposite of the
supposed purpose of the patent system.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Sunday, June 10 2012 @ 09:39 PM EDT

I would like to focus on one area of your discussion that strikes me as a major flaw.

I’m a firm believer that for many disputes it is more efficient for the parties that have a stake in the matter to fight about validity at the time of litigation.

This makes the false implied assumption that both parties in a dispute have similar assets. That's not true. In cases like Oracle v Google, both parties have sufficient assets to cover major litigation costs. The problem is that in software there are a lot of very small operations, often a single developer, who do not have the financial resources to defend against a patent troll or an unscrupulous commercial company.

If you defend patents, then you have to come up with a way for a small business to defend itself without consuming all it's resources.

[ Reply to This | # ]

Nothing is patentable
Authored by: pcrooker on Sunday, June 10 2012 @ 09:48 PM EDT
First, thank you Dr Risch, for writing your thoughts about patenting for
Groklaw, this will be an interesting exchange.

I think Dr Risch presents a well-reasoned argument that software
"inventions" can (though rarely in practice do) contain the same
qualities as their physically-realised counterparts.

"Abstract idea" is an oxymoron - all ideas are abstract. What is meant
here is the difference between mere ideas and their useful expression. This
distinction forms the basis for patent legislation, yet it has become so blurred
as to be meaningless due to computers, software, the Internet, the speed of
change and a globally available, knowledgeable and skilled workforce. Any
invention, software or physical, useful or not, can now be reduced to an
algorithm and easily implemented (by those skilled in the arts, as they say).

Because there is no longer any meaningful distinction between ideas and their
implementation, shall we rely on mere tradition as the basis for the granting of
state-imposed patent monopolies? A tradition that now overwhelmingly hampers
those that the writers of the Constitution sought to help with this provision?

Copyright remains for the protection of the specific expression of one's ideas,
perhaps (though this has yet to play out) but not protection for ideas alone.

How can patents be justified?

[ Reply to This | # ]

Oh dear
Authored by: Anonymous on Sunday, June 10 2012 @ 09:53 PM EDT
Another amateur dazzled by a stupid patent.

I am really sorry, but if you think that keyboard-patent example of yours is an
example of something that should be protected, then you are part of the
problem.

There is NOTHING in that patent that deserves protection, that is innovative in
the least, that advances the state of the art in the slightest, or is any
different than what a modestly competent programmer is expected to do every day.
Nothing. Nothing. NOTHING.

I tried to believe you. I took the time to read the patent. It is junk. If
you think it is patentable, then you are wrong. It is that simple.

And, as far as your hardware person goes, NO. There are very few differences
between different hardware implementations. Hardware people are simply not that
innovative. Computers operate pretty much the same way, both hardware and
software wise, as they did FIFTY YEARS AGO. Just scrambling the codes and
making it LOOK different should not be enough to turn software into hardware.
Massively parrallel SIMD and associative-memory LISP processors are about as
different as it gets, and they simply are not that different to anyone
"skilled in the art."

Maybe your hardware friend needs to stop being so impressed, just learn
software, and find out how wrong he is.

[ Reply to This | # ]

Why patents can be so bad for consumers
Authored by: Anonymous on Sunday, June 10 2012 @ 10:23 PM EDT
You have a great example there in swipe. It is a wonderful idea and you could
maybe argue it is worthy of a patent. However it also points out one problem
with patents. When it was first released it was free demo software and lots of
people loved it and much user testing was done and a great product made. Then
It was taken away because of course they needed to make money out of their
invention. So all those people who had come to use it and love it wanted to
keep using it so you would think now they would be able to buy it right? Nope
the company making it refused to sell it and instead released it only to one
handset maker (and a few others later maybe) to differentiate their products and
force any users who loved this to buy one of these new phones (and landfill
their perfectly good previous phone). I know its the right of the patent holder
to stop everyone using the idea and to discriminate about who they sell it to.
But it's just a very bad side effect of how patents work that they can abuse
this monopoly they have on this idea to keep it away from people who are willing
to pay good money to buy it.

With something that can be sold standalone this is not always a problem as the
item can be sold as is for a fair price. But what about if it can be combined
with something that is 20-50 times more expensive to make that expensive thing
sell more? Imagine if a drug company makes a new life prolonging drug and then
says that it can only be sold as part of a package with a Ferrari (obviously
implying that if you are rich enough to own a Ferrari then you deserve to live
longer than the rest of us). In this situation they will make lots of money
selling Ferrari's by the boat load but its not a very nice use of a Patent's
monopoly power is it? Of course this kind of abuse would quickly get people out
on the streets with pitch forks and touches but if it's just a $400US phone you
have to buy to get this $20 feature which isn't going to kill you if you don't
have it people often ignore this kind of abuse.

The move to restrict access to swipe meant it was heavily pirated as well but it
didn't matter I guess because they got paid so much by the handset makers in the
end. It's just a shame that many consumers have to be steal to get what they
would prefer to pay an reasonable fee for.

Michael

[ Reply to This | # ]

The first example illustrates why NO software should be patented.
Authored by: RTH on Sunday, June 10 2012 @ 10:33 PM EDT
Others are writing about algorithms and maths etc., so I'll tackle the basic
uselessness of software patents. I am assuming that the first example is about
"as good as it gets" for software patents.

First off, NO software engineer (except maybe those employed somehow in the
patent litigation industry) EVER gets ideas about what software to write or how
to write it by studying patents. None. Never. The entire compendium of patents
and the knowledge they encompass is completely, totally useless. Completely.
Without exception. Ever.

So as far as advancing the sciences and arts, the patent process has NO direct
benefit. Absolutely none. (I hope I have stressed that sufficiently.)

That leaves only the possibility of indirect benefit. For example, someone sees
a good idea in practice and decides to mimic it. For the sciences and arts to be
advanced, this easy replication that is hindered by patents must be such as to
dissuade the original creator from creating the first example. I use your
(presumably best-case) Swype patent to show why this argument also fails.

1) What is the original element in that patent? It is the idea that Finger
motion Patterns can Represent Words (FPRW).

2) Given FPRW, everything else follows as night follows day, or at least it will
if, as you say, you have software expertise. How do we know what words? Have a
database of them. What patterns stand for what words? Store patterns with the
words in the database. But people are inaccurate -- so have some error allowance
- maybe links to similar patterned words, or maybe add an error margin around
the swipe and find all patterns that fit in the error margin.

--Interlude-- Did I tell you I stopped reading as soon as I saw the example? Let
me go back and read the rest of your article now... --- And as I thought, the
implementation basically amounts to what I just wrote - off the top of my head
in the time it took to type it out. --end interlude---

3) So what do we have so far: (1) a bright idea, that suddenly struck someone at
some point, and (2) a drop dead obvious implementation strategy. But (1) could
occur to anyone at any time. And this same bright idea probably occurred to a
few dozen of hundred people some time or other without prompting. But even if it
only occurred to a single soul on the planet, would the lack of patent
protection have stopped it from happening? Let's get real. As for (2), given an
idea, how to make it happen WILL be discovered by any reasonably competent
software writer. Ideas that can't be implemented because the way to do it is so
darn non-obvious just don't exist.

4) So what else is there, after the flash of inspiration that no patents will
encourage or deter, and the non-obvious how-to-do-it? The only remaining step is
the sweat of the brow in spending the hundreds or thousands of hours needed to
write the actual software. And guess what? We have copyright to prevent someone
taking that away from me. If I invest that time, I'll have at minimum a
headstart over any competitors; by the time they can duplicate, I'll either have
made some money or learned how to do it even better or both. The only thing a
patent would add is a decade or so of others who would have had the bright idea
themselves anyway being legally prevented from producing a similar or better
product.

In short, patents don't solve any problem for software or software innovation.

Lastly, the surprising thing is that, without arguing that they are also useless
everywhere else too, I can at least note that they are surprisingly less useful
than they are assumed to be. The 'light bulb' moment for me was when I learned
the early history of the steam locomotive. Steam locos need high pressure steam
because low pressure requires such a bulky engine it cannot be put in a vehicle
and made mobile. Watt had a patent on the steam engine, and he refused to make
high pressure variants. So the entire world had to just sit around for ten years
and wait out the expiry of Watt's patent; only when that happened could the
industrial revolution and the railway age really get started.

[ Reply to This | # ]

Where is the analysis that determined software makes a new machine?
Authored by: PolR on Sunday, June 10 2012 @ 10:55 PM EDT
Thanks Dr Risch for sharing this with us. This is a most needed and most welcome
discussion.

I agree that every tools at our disposal must be used to improve the quality of
patents. But I think that in the case of software there is more than quality at
stake. Much of case law rely, to stay polite, fictional principles of computer
science. One avenue to improve the software patents is to interpret the law
according to principles of computer science which are correct.

For example in re Alappat says that programming a computer makes in effect a new
machine. Where is the factual analysis that leads to this determination? It is
not in the Alappat case. I have read the precedents that Alappat invoked in
support of this holding and I don't find the factual analysis there either. The
closest I could find in in re Benhardt where the CCPA ruled that a new
arrangement of memory elements is a change in machine structure, but without
justification for this conclusion.

If you know where the factual analysis is located please tell me. I will be
happy to read it.

Here is a computer professional view on the topic.

Consider driving a car. The engine is running. The wheels are rotating. These
are physical changes but they are not structural changes to the car. If we
decide otherwise no car goes anywhere because as soon as a moving part of the
machine moves, it is a new machine. We need to distinguish changes which are the
normal operation of the machine and changes like replacing the engine which
genuinely alter the machine structure. The CCPA argument in Bernhart lacks the
explanation of why a change in computer memory is a change in computer
structure. Memory is changed billions of times per second. Computations can't
proceed without changing the memory contents. If Bernhart is right no computer
runs any program because as soon the memory is modified it is no longer the same
computer.

What is the function of the CPU? It is to execute the instruction cycle. The CPU
reads from memory one instruction and executes it. Then the CPU reads the next
instruction from memory and then executes it. The CPU proceeds like this
instruction by instruction until the execution of the program is completed.

The list of possible instructions we can use in a program is finite. This list
is predetermined by the designers of the CPU and it is documented in technical
manuals. The CPU has all the structure to execute each and every instruction and
it has the structure to cycle over a program and execute it in its entirety.
This is the function of the CPU. Loading a program in memory doesn't expand this
functionality. It is a required operation to use the existing built-in function
of the CPU.

Why would a computer need more structure to execute a program? The instruction
cycle is all that is needed. There is no operation of loading a program into the
CPU to configure it electrically. There is no arrangement of transistors that
corresponds to the programs functions. The program is executed instruction per
instruction with the same CPU structure.

Besides programs are modifiable, like all other memory contents. A running
program can be modified in memory while it is run. The '104 patent in the Google
v. Oracle case claimed precisely one such programming technique.

Programming languages in the functional family are not implemented solely by a
sequence of instructions like the usual imperative programs. They are
implemented with a combination of instructions and data structures called
closures. The function of the program depends as much on the closures as it
depends on the instructions. The execution of the program requires that the
closures are constantly be modified as the execution progresses. This means that
the memory contents which gives the program its functionality is constantly
changed, like the wheels of the car are constantly rotating.

You can see why I am complaining about fictional computer science. The holding
from Alappat has no factual basis.

[ Reply to This | # ]

Your suggestions on how to get rid of software patents completely.
Authored by: Kilz on Sunday, June 10 2012 @ 11:09 PM EDT
I have read enough posts saying that patents on software shouldn't exist. I am
not going to debate thta because I dont
think they should either. But the person who wrote this
article has a way of dealing with some of the patents.
For those that dont think they should exist here is your
chance to enplane exactly how we can get rid of them completely. Please offer
step by step real world ideas on how
to get rid of them that we at Groklaw can implement that will
have the effect of getting rid of them for good.

[ Reply to This | # ]

Why not use the definition of algorithm used in computation theory?
Authored by: PolR on Sunday, June 10 2012 @ 11:31 PM EDT
In mathematics the term algorithm is term of art. It has a precise definition in a branch of mathematics called computation theory. This is, by the way, one of the mathematical basis of computer science. Mathematical algorithms in the sense of computation theory are the same thing as algorithms in computer science.

An algorithm in this sense is a procedure for manipulating symbols which must meet some special requirements. The procedure must be entirely determined by the explicit rules of the procedure and the symbols being read. No judgment call, no ingenuity is allowed. Algorithms are machine executable. This is why they are so important in computer science.

If we use this definition for the term algorithm we won't end up deciding no process is patentable. We are not using some overly broad definition which is synonymous with process. The term is circumscribed. Not all processes are procedures for manipulating symbols in this manner.

Here are some examples of algorithms. The alphabet has two symbols: 0 and 1. The two algorithms are:

  • Boolean AND
    1. read two symbols
    2. if both symbols are 1 then the answer is the symbol 1
    3. otherwise the answer is the symbol 0
  • Boolean OR
    1. read two symbols
    2. if both symbols are 0 than the answer is the symbol 0
    3. otherwise the answer is the symbol 1
You will have recognized the ordinary logical operations under the ordinary convention the 0 means false and 1 means true. But the algorithm doesn't rely on these meanings. It works exclusively by looking at the symbols. This is why the algorithm is a manipulation of symbols and not a manipulation of concepts.

In digital electronics boolean operations are related to circuit structure. Here are the description of two circuits which manipulates voltages.

  • Circuit no 1
    1. receive two voltages on two input wires
    2. if both voltages are 0.5V then produce the output voltage 0.5V
    3. otherwise produce the output voltage 0V
  • Circuit no 2
    1. receive two voltages on two input wires
    2. if both voltages are 0V then produce the output voltage 0V
    3. otherwise produce the output voltage 0.5V
Can you tell which circuit implements the boolean AND and which implements the boolean OR using voltages? The correct answer depends on a convention. The usual convention is that 0V means the bit 0 and 0,5V means the bit 1.The circuit 1 is AND and circuit no 2 is OR. But what if the engineer decides to reverse the convention? That 0V means the bit 1 and 0.5V means the bit 0? Then the functions of the circuits are reversed. Circuit 2 is now AND and circuit 1 is OR.

This means two things. Bits are symbols and this is not the same thing as an electrical signal, like voltage. Symbols are abstractions, like the difference between a letter and a mark of ink on paper. The second thing is that the algorithm is abstract. It is not the same thing as the description of a circuit. It is a description of how the symbols are manipulated.

I believe the law doesn't think clearly about software because the necessary distinctions are not being made.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Rubberman on Sunday, June 10 2012 @ 11:49 PM EDT
For the most part, I agree with Dr. Risch - I am the holder
(sole inventor) of a software patent (owned by Brooks
Automation) that is applicable to adaptive systems (software
systems that can easily adapt to environmental changes). It
has been utilized in the manufacture of semiconductors,
flat-panel displays, and disc drives by most of the major
manufacturers of such devices (IBM, Samsung, Philips,
Seagate, etc). That said, I have to agree that most software
patents are just plain ridiculous! They are so obvious as to
evade logic as to why they were granted in the first place.
In my situation, this was the solution to a serious problem
- allowing shrink-wrapped systems software to be modified by
the end user to their needs by simple specification, and not
requiring any changes to the base code or code compilation
in any way. It took me 5 years of deep thought and research
to derive the solution that was recognized by US patent
#7185325. The efficacy of this invention was proven at the
Samsung/Sony joint venture LCD manufacturing plant in South
Korea in the early 2000's.

So, in my professional opinion (I am now Senior Systems
Engineer for a tier-one mobile phone manufacturer), most
software patents are bogus, but that doesn't mean that we
should "throw the baby out with the bath water". It means
that the USPTO and other patent-granting authorities need to
apply much more rigorous standards before granting these
patents. I think that about 1% of current software patents
are worthy of that protection. That means 99% should be
revoked, and the sooner the better!

[ Reply to This | # ]

Obvious infringement
Authored by: Anonymous on Sunday, June 10 2012 @ 11:54 PM EDT
What about a patent where its formal description does not
contribute to public knowledge? I can imagine some patents
where a video or photo of the software in action would serve
as a sufficient description, and the patent only gives an
"obvious" implementation algorithm.

From just this video or photo, one skilled in the art would
likely use the same algorithm as stated in the patent to
develop an independent implementation. But, the obvious
solution will be an infringement of the patent.

Could/should patents like this be rejected for obviousness?

[ Reply to This | # ]

A cure for the wrong time.
Authored by: jvillain on Sunday, June 10 2012 @ 11:57 PM EDT
The problem with patents is the same as it is for copyrights. When they were
invented they were designed for a world that just doesn't exist any more. Let me
use my music copy right example first and then we can apply it to patents but
not in a David Boies I have no clue way.

When music copy rights first came out the world was very, very different. No
cars,planes, telephones, computers etc, etc. Cities and towns were much smaller
and the distance between them much much larger. This is important.

If I was in New Orleans and I wrote a song about the only people I had to worry
about copying was the people in the same town or the odd rare traveller. The
idea that some one from San Francisco or London was copying it was virtually
impossible because they had no access. Even if you popped into a speak easy and
heard me play it once the best you could get from it was a rough interpretation.
Maybe the chord changes and a few lyrics.

The idea that some one in Sydney Australia created the same song at the same
time was flat out impossible. In this environment the copy right system has a
chance to work the way it is supposed to.

Move forward to today and music moves around the world and to all corers of the
world in less than a heartbeat. I have access to millions of songs. I have the
ability to create music on line with people any where in the world in real time
if I have the bandwidth. If we create a song how much of it is truly going to be
novel compared to what you would have seen at the beginning of copyright?

Shift gears to patents. Marconi and Bell fought it out on opposite sides of the
ocean with the fastest form of transport between them being ocean ships. The
idea that they were using each others work or were collaborating usefully is a
stretch.

Leap forward to today where there can be hundreds of teams working on the same
thing borrowing from each other or people who are not evolved but have shared
their knowledge and all located in different countries. Yet only one gets the
patent in the US for example. Is that fair? Is that right?

The answer that you might come up with is that possibly no patent should be
issued. But answer this. How many patents issued these days in any field not
just software are pure inventions with no input from the outside? Probably close
to none.

[ Reply to This | # ]

Posner calls patent system "dysfunctional" while dismissing Apple/Google case
Authored by: Anonymous on Monday, June 11 2012 @ 12:01 AM EDT
http://gigaom.com/mobile/famous-judge-spikes-apple-google-case-calls-
patent-system-dysfunctional/

[ Reply to This | # ]

If he practicaluse is patentable, why not patent that instead of software?
Authored by: PolR on Monday, June 11 2012 @ 12:08 AM EDT
I think you example patent is poorly chosen. A process for a human for entering
data by swiping on a screen is not the same thing as the execution of a program.
It reminds me of the patent in Diehr. It included the step of actually curing
the rubber. Both patents are doing more than claiming software because they are
actually claiming all the steps required to actually achieve the practical
result.

But a lot of patents don't do that. They mention the utility as an intended
purpose but if we run through the elements of the claim only the machine or
software functions elements are recited.

I believe this distinction is crucial. It makes the difference between claiming
mathematics because it could be used in a practical manner and claiming the
actual use of mathematics. When combined with the notion that in pure software
the invention is the mathematical algorithm and not some machine or process by
which it operates, a lot of the silly patent will get weeded out and the ones
which are genuinely drawn to patentable subject matter will remain. These
patents may then be analyzed for novelty, obviousness etc.

An algorithm in this sense is a procedure for manipulating symbols which must
meet some special requirements. The procedure must be entirely determined by the
explicit rules of the procedure and the symbols being read. No judgment call, no
ingenuity is allowed. Algorithms are machine executable.

There is a difference between how humans manipulate symbols and how machine do.
When presented with the word FOX a human knows which kind of small furry animal
this is. A computer will see the letter F followed by the letter O followed by
the letter X. Algorithm are the only way computers can manipulate symbols
because, in part, they are the kind of procedure which don't require the machine
to act based on meaning. The machine can act based on the bits alone.

The task of a programmer is to find a procedure that works which won't require
the machine to understand meaning. This procedure will be the program written in
source code. The program is a written description of the algorithm.

What does it mean to utility? In most circumstances utility is totally related
to meaning. Therefore utility has no structural relation with the program
execution. Meaning has been abstracted away by the task of writing a program.

Many software patents are like attempts patent a novel by reciting an outline of
the plot claiming this describes the structure of a stack of paper covered with
marks of ink. But the exact theory is more like claiming the structure of a
computer, or the process by which it operates, by reciting an outline of the
meaning of what is written in its memory. But meaning is not part of the machine
structure and it is not part of the execution of the program. The machine alone
cannot deliver the utility without some external party able to read the symbols
and act on the meaning.

Software is distinguished from a practical use in the same manner the
calculations of the engineer about a rocket are distinguished from the rocket.
Are we going to patent the rocket? Or are the calculations patentable because
they refer to the rocket even though we don't actually make the rocket?

[ Reply to This | # ]

Apply the arguments to this test case
Authored by: Anonymous on Monday, June 11 2012 @ 12:23 AM EDT

Markus Kuhn wrote an excellent article describing a specific, simple software patent which turned out to be harmful. He discusses the particular situation of software patents required for ensuring interoperability of systems (by implementing standards).

It is possibly the simplest to understand example I have seen. Non-programmers should be able to understand his description of the technical details.

Here's the challenge: which arguments touted in this article and comments handle Kuhn's example sensibly?

Read it here: M arkus Kuhn: A One-line Software Patent

One of the comments discusses a creative licence to solve(?) this problem used by the Bluetooth group.

Let me add one question. Has anyone ever seriously considered expropriation (eminent domain for our American friends) for intellectual property which is needed for critical public interactions among systems?

[ Reply to This | # ]

Software might be math but it doesn't matter
Authored by: Tolerance on Monday, June 11 2012 @ 12:26 AM EDT
I was an engineer and I'm trying to be a lawyer. I just know people are going to
give you a hard time about phrases like "Of course you are right that
software has to be written and compiled for specific hardware", because of
concepts like "Write once, run anywhere" beloved of Java enthusiasts,
or the ancient p-Code.

More importantly, for your engineer interlocutor to say 'ALL software is a
specification of hardware' is absurd - it's precisely the reverse of the truth.
Consider pseudocode, the point of which is to remove the software idea away from
hardware.

But none of it matters, ultimately. I suspect the way out of this for pro-patent
engineers is to get SCOTUS to abandon the long established rule that maths and
abstractions in general can't be patented. They're already half-way there.

It's not hard to show that software is a subset of mathematics. Turing
completeness certainly shows that an algorithm (by the old definition, a program
which eventually halts) is isomorphic to an object in maths (a suitably defined
Turing machine).

Someday the advent of computers which are not equivalent to a Turing machine
(e.g. quantum computing) might extend the isomorphism to non-algorithmic
processes for real hardware, but we're not there yet.

This means anything outside quantum computing has a complete analog in math - it
means that computer algorithms are indeed necessarily analogous to mathematical
formulas. That's a result straight out of the theory of computation. Of course
not all programs are algorithms in the old sense; they don't all halt, for
example. Still, it's not guaranteed that computers will always be classic
Turing machines. The idea of a Turing machine has already been generalized. But
it's all still math!

Technically speaking, there is an isomorphism between any program executed in a
computer (Turing machine) and a corresponding abstract mathematical procedure
(algorithm). This arises from the founding work of Church and Turing, who in
turn were looking into the work of Kurt Godel. All this was settled before the
second world war. Things haven't really changed since.



---
Grumpy old man

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Monday, June 11 2012 @ 12:27 AM EDT
So, does this impede innovation? I don’t think so – continuous stroke typing was a twinkle in the eye of researchers at the time this was invented. The patent cites the leading work at the time, and that work was not nearly as easy or useful as this product. The idea, once disclosed, was easily copied by competitors, and the only way to ensure early development of it was to protect it with a patent.
Yes, it most certainly does impede innovation. One only has to ask how long it might be expected for that particular solution to arise had Swype not disclosed, or even devised, it. Do you really think it would have taken 20 years?

This is no longer 1790 and the population of the U.S. is now somewhat north of 4 million -- there are more professional programmers now (about 1 million) than there were educated Americans at the time the patent system was devised. And the number of "amateur" programmers who are in possession of both the necessary tools and the basic skills to tackle the problem is likely an order of magnitude or more larger (SourceForge has about two dozen projects dedicated to special keyboard input needs).

The U.S. patent system originated in a time of virtually unlimited natural resources and a dearth of skilled tradesman to exploit them (it was largely a recruiting tool for attracting scientists and artisans from the Old World). Our society no longer has the luxury of unbounded resources, and can no longer afford the extravagance of inhibiting for decades the wealth of talented, educated developers that it does have.

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Monday, June 11 2012 @ 12:44 AM EDT

What usually inspires change isn't the discovery of a new principle. It's the changing economics which make ideas which were "lying around waiting to be used" financially practical.

I have written software for automated test systems. The way I would do things today isn't how I would have done them twenty years ago. What has changed is that CPUs, RAM, and disks have become simultaneously both faster and cheaper. Designs that were uneconomic then are the cheapest way now. Cost guides design choices, and uneconomic choices are briefly examined and discarded. The fact that "nobody did 'X' before" is meaningless if nobody did it because it was too expensive. I don't spend time and money on ideas that are impractical today but which might be practical in future. It's conceivable that someone might come up with a useful new idea in test equipment that I would never have thought of, but after 30 years in the industry I haven't seen it happen yet. I have come up with ideas that nobody else has, but I don't kid myself that other people wouldn't have come to the same conclusion given the same motivations. New ideas like that just don't occur at the frequency that we see patents being filed.

However, economics isn't a consideration for someone filing a patent. That's someone else's concern. You simply need to file your patent and wait for Moore's law to make it practical for you. In other words, patents have become like buying a lottery ticket.

Perhaps another way of looking at the patent problem is like comparing it to the spam e-mail problem. If e-mails are very cheap, someone can spew out millions of them in the hopes that a very few will pay off. Software patents are very cheap, in that it seems very easy to file thousands of patents that vaguely cover broad concepts on "hot" new product areas (e.g. mobile phones). Like spam e-mail, spam patents can be profitable even with a low success rate. Also like spam e-mail, spam patents impose costs on society which far exceed the (unjust) benefits which accrue to the spammers. Those costs are not borne by the spammer though, but are externalized to society as a whole and become a net economic loss.

The people who claim to have some brilliant idea to make some subtle reform to patent law to "solve the spam patent problem once and for all" are like the people who claim to have a brilliant idea to make some subtle change to e-mail protocols to "solve the spam e-mail problem once and for all". Neither address the fact that the spammers are working actively to subvert the system because they stand to make a great deal of money by doing so.

Why should software patents exist? If you ask almost any major software patent holder, he will almost invariably tell you that he has these patents "for defensive purposes only". In other words, if you believe the companies who are actually filing the patents, they will tell you that software patents are only necessary because other companies have software patents. The only people who consistently claim that software patents benefit society are patent lawyers and patent trolls.

Software patents are an artificial government regulation which are intended as a form of economic central planning intended to "promote innovation". The people who actually do the innovating however say that software patents do nothing of the sort, and indeed are a hindrance to innovation. So why not simply abolish them?

[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: jimrandomh on Monday, June 11 2012 @ 12:53 AM EDT

I find it interesting that you used Patent 7,098,896 as an example, because I personally have a small notebook of novel ideas about how to do text input, mostly orthogonal to slide typing, and a codebase with a decent start on implementing them. I abandoned the project unpublished soon after discovering this very patent. While you say that "there are ways to work around the patent to allow for continuous stroke typing", I concluded at the time (and still believe) that this is untrue - that making a touch-screen keyboard not infringe this patent would require crippling it so badly that users would never accept it. This would strongly argue against your statement that this patent does not impede innovation.

What is obvious when you are starting from scratch, and what is obvious when you have been thinking about touch-screen input for a month, are very different things. The patent and legal systems are entirely incapable of evaluating the latter criterion. I believe that stroke typing falls in the latter category ("obvious if you think about the topic long enough"), and that the apparent novelty of the idea comes only from the lack of widely-available touch screens on the market at the time it was patented.

[ Reply to This | # ]

Compromise
Authored by: ailuromancy on Monday, June 11 2012 @ 01:23 AM EDT

Do you reach a compromise with a mosquito or a leach? If someone comes to you house, tells you it looks fragile or combustible and that you should pay some protection money to prevent something from happing to it, what sort of compromise would you consider reasonable?

I suspect the patent helped the company get venture funding

I suspect the patent cost lots of money, provided no revenue, helped drive the company into bankruptcy where the patent was bought by a troll who is now trying to extort money from anyone who profits from a device with a touch screen.

Patents are an abysmal disaster for anyone with a novel invention. It takes years to get a patent, then more years to sue the pants off the person who actually turned the idea into a profitable business.

The way to profit from an invention was to manufacture and sell. You could become a market leader before anyone else had time to copy your invention, and by that time you could have a second generation product ready.

These days, the way to profit from an invention is to let someone else to all the work, then threaten him with a dozen invalid patents.

If you haven't worked out "why not" for software patents yet, then I remember that you "spend a lot of time finding prior art to invalidate software patents". What a complete an utter waste of your life. Wouldn't it be so much better if you could do something constructive instead?

A decade ago IBM read patents to look for ideas. Nobody else had the resources. Now not even IBM looks for new ideas patents, partly because they are a huge pile of dross, and partly because of the triple damages.

There are now two good reasons to read a patent: Because someone is suing you, or because the patent has expired and implementing it means defending yourself will be cheaper when someone does sue you.

People have no problem creating products without reading patents. Why do we have a patent system at all?

[ Reply to This | # ]

What I mean by software is mathematics
Authored by: PolR on Monday, June 11 2012 @ 01:31 AM EDT
First let's get some misconceptions out of the way. Software is not the
description of a machine structure for the reasons I presented in a previous
comment to this same article. Therefore the math of software is not a
mathematical description of such a structure.

Next the mathematical part of software is its execution by the CPU. It is not
the source code and it is not some end to end process involving the computer as
a component. Diehr's rubber curing patent would survive the software is math
argument and so will many similar processes which involve more than just running
the program. These patents may still be invalid for novelty, obviousness or
other reasons.

But there is a category of software patents which claim nothing more than
running the program, feeding it with input and outputting the answer. These
patents are the ones which IMHO could be invalidated by the software is math
argument. There may also be some patents involving software which amounts to
nothing more than reciting the algorithm and say apply it. These patents may be
invalidated by a Mayo v. Prometheus kind of argument on the basis that an
algorithm is an abstract idea.

There is in mathematics a special kind of algorithms which can compute
everything which is computable. If we build a circuit to compute such an
algorithm we get a general purpose computer. In the ordinary computers we use,
the universal algorithm is the CPU instruction cycle. Some programming languages
rely on software implementation of other universal algorithms such as a byte
code interpreter.

Software is the input we give to a universal algorithm to obtain the computation
we desire. And the execution of the program is the computation of the algorithm.
When a patent claim reads on nothing more than the execution of the universal
algorithm then the claim is directed to mathematical computation.

There is no need to make a new algorithm for each different function. For a
simple example consider the function of doubling a number. We may add the number
with itself. We may multiply it by two. We may use an algorithm for addition and
put a constraint on its input, that the number must be added with itself. Or we
may use an algorithm for computing a multiplication and put a constraint on its
input, that one of the numbers must be 2. We have two algorithms for the same
function. If we look at it in terms of circuits, we may use a circuit for
computing addition, or a circuit for computing multiplication. We don't have to
make a circuit dedicated to doubling numbers.

This example shows that functions may be implemented by putting constraints on
the inputs of algorithms and/or circuits designed to do something else. We don't
have to make a new algorithm or circuit every time. This is what happens in a
general purpose computer. We use a universal algorithm and we put constraints on
its input. We use the instruction cycle of the CPU and its input is the program.

[ Reply to This | # ]

Swype example is obvious, the mathematics old
Authored by: Winter on Monday, June 11 2012 @ 01:39 AM EDT
"It claims at its most basic the idea of moving the finger from letter to
letter on a qwerty keyboard. The motion of the finger is then captured (the
shape of the line), and matched against a dictionary of shapes for different
words. The system also compares for close shapes and suggests words that might
have been intended but the finger didn’t quite get to the correct keys."

The underlying model is well known. This is simply an application of statistical
systems like the (very) old Hidden Markov Model in recognizing temporal
sequences. It is the basis of all modern speech recognition and the Blast
algorithm used in genomics research. An alternative could be to use Artificial
Neural Networks (not as good for time series). Also well known from the 1980s.

These two types of data driven recognition methods are the staple of
"intelligent" recognition systems since the 1980's.

Many Secret Service type of eavesdropping systems since the 1950s use timing,
noises, and emissions of typing equipment to determine what was typed, eg, the
old Tempest system did that for telex.

What is actually *not* obvious? Maybe selecting the swipe movement of the
fingers?

Any CS engineer can build such a system when she is given the question
"Recognize words and ase the finger movements".

Why is that so mind blowing new that this should enjoy patent protection?


---
Some say the sun rises in the east, some say it rises in the west; the truth
lies probably somewhere in between.

[ Reply to This | # ]

Here is where you are wrong
Authored by: Anonymous on Monday, June 11 2012 @ 02:02 AM EDT
<blockquote>For that reason, I’m a firm believer that for many disputes it
is more efficient for the parties that have a stake in the matter to fight about
validity at the time of litigation. This increases costs on patent defendants,
but if examiners will not get the job done anyway, I would rather they not spend
a lot of time trying.</blockquote>
"This increases costs on patent defendants," The problem is that this
cost is there basically independent from the merits of a case. It is like
saying "This increases costs on SPAM recipients." There is no
protection against being a patent defendant, and if you actually do research
avoiding patents, you are liable to triple damages because of willfulness,
anyway.

[ Reply to This | # ]

The 896 Patent
Authored by: Anonymous on Monday, June 11 2012 @ 02:13 AM EDT
Swipe is not entirely novel and it interferes with other work. For example, the
grafiti input method for Palm pilots (which also ended badly for Palm) is very
similar in nature to 896.

But to show that Swipe is not novel, when I first started using an Android
phone, I quickly realized that the problem with touchscreen input is that keys
are often missed in favor of other keys which are neighboring... indeed, it's a
problem with typing in general -- only moreso in the case of touchscreen input.
But I, as a non-expert in the field; a mere user, realized that in addition to
simple spell checking, input should be validated by proximity on the keyboard.

This is completely independent of sliding one's finger from key to key as that's
not the only way line patterns could be created. Indeed, by recording the
coordinates of keys until the space or return key is entered, "lines"
could be drawn in memory or not at all since it's not really lines which are
being used for comparison as much as the points described as
"direction" changes.

I believe it is completely obvious as I am not exactly a professional in the
arts. And isn't the test for novelty one in which the "invention"
must be one where a person skilled in the art would not come up with that
solution on his own?

In my view, the statement of the problem presents the obvious answer. The
problem is obvious and so the answer becomes obvious. But then you might
say/ask "Then why hasn't it been done before?" I can't say except
that I believe some spell checkers already take much of this into account...
just not all spell checkers.

[ Reply to This | # ]

Harmful Patents
Authored by: sproggit on Monday, June 11 2012 @ 02:27 AM EDT
Michael, thank you for taking the time to research and write such a
thought-provoking article for Groklaw. Whilst I can't promise to agree with all
your views on the subject, I'm grateful that you've taken the time to express
them in such a thought-provoking and educational way. I hope that PJ can
persuade you to continue to contribute.

So down to business. Having read the article and the top-line comments I see
that as a community we're already snuggling into familiar trenches of thought
and ideology. Rather than pick up on specific aspects of your piece to further
that paradigm, I thought it might be nice [to paraphrase something we quote
often enough] to advance an argument that may be "non obvious to one
skilled in the art"...

Instead then, let's consider this using political doctrine as a metaphor.

IMPORTANT DISCLAIMER: I am not an historian nor any form of student or history.
I am using the following as illustrative example. If I go slightly off-point,
please try and stay with the principle of the illustration and not get lost in
the detail...

In my counterpoint I am going to argue that there is a very fine dividing line
between the "state-sponsored monopoly" that is a patent and
"state ownership" for anyone who stands outside of that relatively
tiny portion of society that is comprised of government post-holders and patent
holders.

In the Communist ideal, a society cedes personal ownership to the State for the
common good. The State, in turn, determines how everything is organised and run.
Not wanting to compete with itself, the State may well have a single
company/factory producing tractors and another single source for televisions.
Designs are simplified and standardised to bring efficiency of mass production,
drive down unit costs and eliminate wastage [both effort and materials].

In the "Free Market Ideal" [for the sake of my response the mirror to
the Communist model], competition isn't merely permitted, it is actively
encouraged. The philosophy of free markets is that competition fosters
innovation by encouraging producers to continually refine and improve their
products and services to better serve their markets and customers.

The world has seen numerous examples of communist and free market economies
since the close of the Second World War. It is not lost on me that the United
States, a staunch advocate of the Software Patent Monopoly is also a staunch
advocate of free markets and opposed to communist ideology. I would like to
argue that, since 1945, nations that embraced free market economics have been
more successful than those which adopted the communist model. I would be happy
to measure "success" in a variety of terms, including affluence of
average citizens, productivity per capita, average lifespan, per-capita income,
etc, etc.


I find it somewhat perplexing to think that many people who believe passionately
in the democratic freedoms of what we often call "the West" are so
quick to embrace the idea of state-sponsored control through patents. Perhaps
it's because even those who argue in favour of patents don't see how that patent
system could be abused by a corrupt state... To me this just looks like adopting
opposing arguments at the same time... Is this what we want?

But more than that.

Suppose I have an idea for a truly innovative tractor. If I live in a free
market society I can build my design and take it to market. In a communist
state, if I tried to do anything like that, I'd be classified as a subversive
and told to stop. If I failed to obey the request I could be arrested. How is
this any different [in principle] from what we have today with patents? In much
the same way, if I am a "lone inventor" and have an idea for a
significant enhancement to a commonly used piece of software [say for example
the person who thought of adding tabs to web browsers] then in a software-patent
encumbered marketplace, I would be unable to get my idea to market because the
underlying browsers would likely contain restrictive software patents.

In other words, great ideas and innovation are stifled not because they are not
worthy, but because the inventor does not have the freedom to innovate: the
moment they go to market with a product they are jumped on. Patents *might* be
helpful when an invention is totally new, and works in a brand new field of
endeavor. Patents *don't* work in any market where innovation is incremental.
Software is, intrinsically, an "incremental" marketplace.

I would argue that this was *not* what the authors of the 1790 Patent Act or
House Resolution 41 had in mind when they drafted the original statute.

Whatever it's merits, the Patent Act has been peverted to the point that, rather
than fostering innovation, it is now having the exact opposite effect and being
used to create and maintain monopolies in many different fields of endeavor.
Patents have become a weapon to be used by those with established market share
and financial resources to "keep out" new market entrants and startups
through a "barrier to entry".

That was not the intent of the Patent Act, but, sadly, that is how it is being
used today.



[ Reply to This | # ]

Curing the Problem of Software Patents, by Michael Risch
Authored by: Anonymous on Monday, June 11 2012 @ 03:26 AM EDT
If one takes as starting point that software patents are here to stay (which is
probably realistic given that a change in legislation would be needed), these
are pretty useful suggestions.

The main one that I'd quibble with is the trade-off between examination and
litigation... IMO it would be good for everyone if the standard of examination
was raised, and with it the patent filing fees.

The suggestions around practical use and specificity are on the right track but
can we make them even tighter? Basic inventions seem to get proceesed into a
bunch of legalese the goal of which is to make the claims as broad and
impenetrable as possible. That is clearly wrong, as it hampers examination,
litigation, and use of the patent.

Since patents are (supposedly) descriptions of technical inventions, they should
be couched in technical language that is clear enough for an engineer to either
implement the patented invention or test their own design against the patent to
ensure that it is not infringing.

This would probably have the side-effect of facilitating patent workarounds (the
transformation into legalese was once explained to me as having the purpose of
blocking workarounds) but that is not a problem as something that is truly
innovative will by its nature have limited workarounds.

The other key change that would make a huge difference in the flood of junk
patents is to have a requirement that a patent remains "pending" until
the patentee has demonstrated an actual implementation (preferably in the form
of a prototype product) incorporating the invention.

I think that's the way the patent system works in South Africa (where I am
from). If you can't produce an implementation winside 5 years of filing the
patent, the patent lapses.

That would help tremendously with the patent logjam and also deal with the
problem of non-practising entities, all at once. The patent system exists to
promote the development of technology, and "sitting on" a supposed
invention is a hindrance to that progress.

Leading on from that thought, that is a test in its own right that could be
applied in both examination and litigation - is the patentee acting to hinder or
promote progress? If the patentee is not using the invention to contribute to
the progress of the "useful arts", they should not have any right to
act as a gatekeeper for that technology or inhibit others from doing so.

One last thought about the patent-pending phase... during this phase the
patentee should not have any monopoly, i.e. if someone else who has no knowledge
of the patent implements the same invention, that implementation cannot be
considered to be infringing.

[ Reply to This | # ]

Who should administer the cure?
Authored by: stegu on Monday, June 11 2012 @ 03:33 AM EDT
I can agree with a lot of what you are saying, and I definitely think your
suggestions would significantly improve the situation. However, your proposed
solutions assume that a great deal of common sense and self-criticism should be
instilled in the processes for patent examination and litigation. Currently,
there seems to be absolutely no such tendency, because there is no incentive for
it. As long as the key players that are involved in creating patents (the patent
office and the filing inventors) are mainly interested in patents as a cynical
way of making money from nothing, and as long as the courts and the lawyers see
endless litigation over stupid patents as a source of income rather than a
fundamental failure, the system will never reform itself from within.

I think that we might need at least a threat of completely revoking the patent
rights for software-based inventions, to instill a sense of impending doom in
the PTO and the "inventors" filing all these junk patents. Without
that imperative, the system will most likely continue on its runaway path.
People are not good at seeing the big picture and aiming for long term benefits
on their own accord, certainly not as long as they can make money from being
short-sighted. The patent holders, the patent lawyers and the patent examiners
are not going to see the error of their ways before it is much too late.

I guess my question is: who should force this fundamental change on the actors
in the farce that the patent system has become, and how? Courts seem to act very
slowly, and they are quite naturally reluctant to take actions that would
effectively invalidate 95% or more of all software-related patents.

[ Reply to This | # ]

If software is abstract, riddle me this:
Authored by: Anonymous on Monday, June 11 2012 @ 03:35 AM EDT
1) Why does poorly written software make my life an existential hell?
2) Why does that so-very-abstract compression algorithm save me oh-so- very real disk space and network bandwidth?
3) Why does heavy computation of all that abstract stuff make my PC go all hot and whiny?
4) Why don't we have working examples of other abstract ideas, like jetpacks and flying cars? (Oh wait, you mean those are ABSTRACT abstract ideas, as opposed to "only" abstract ideas as covered by the Swype patent?)
5) Why do I need a computer to run my games at all? Shouldn't I be able to run it all in my head?!?

BTW, I commend mrisch for stepping into the lions den like this. I too have undertaken such foolhardy endeavors, albeit only in the comments section, in the past (and am doing so even now!)

[ Reply to This | # ]

What does society gain
Authored by: Anonymous on Monday, June 11 2012 @ 04:06 AM EDT
What does society gain through software patents?

Inventions are not disclosed. The things are written in bizarre lawyer-speak and
are ridiculously general and vague. No source code is included. They are almost
completely useless from a practical point of view when it comes to figuring out
how software works. In any case nobody dares to read them for fear of triple
damages.

They don't offer an incentive for invention. You invent something and you'll get
sued by someone else who wrote a patent earlier that looks vaguely like your
invention. You'd be safer opening a burger restaurant than trying to invent new
technology these days. At least the patent on burgers has expired and nobody is
going to sue you over it.

What is society supposed to gain by allowing these licensed monopolies? I see
lots of talk about tweaking the system to try to minimise problems. But the
basic issue is that there is no point to these things. No advantage at all as
far as I can see to having them. Patents are nothing but a work scheme for
lawyers.

[ Reply to This | # ]

Warning: may contain elderly auto analogy
Authored by: Ian Al on Monday, June 11 2012 @ 05:25 AM EDT
Note this comment I got at PatentlyO from an engineer/patent lawyer in response to the above statement: “ALL software is a specification of hardware. ALL software is run very differently on different hardware and is written for specific hardware platforms. Software is hardware in every sense that matters for patents.”

Here was my response: “Of course you are right that software has to be written and compiled for specific hardware... The issue is that patents aren't for specific implementation of specific solutions on specific hardware. It's that patents claim general steps to solve general problems (or sometimes specific ones) on general hardware.”
In fact, there are no software patents. 'As a general rule, where software constitutes part of a best mode of carrying out an invention, description of such a best mode is satisfied by a disclosure of the functions of the software'. (Fonar Corporation v. GE). Just because the patent declares that a system comprising a computer with a memory is the best mode does not excuse the patent holder from patenting the functional elements of the best mode rather than an invention for which software is the best mode. The USPTO should look for prior art for the claimed functions amongst all the likely implementation modes and not just software artifacts. If the invention can only be found in a software solution, then it is almost certain to fail many of the tests set by the Supreme Court.

Software source code is not written to target the computer hardware. I have two computers that can run Windows; an Intel Atom based computer and an AMD Athlon four core computer. Windows has to have hardware drivers and BIOS to suit the OS and to provide the link to the specific underlying hardware. However, the only reason that both my computers can run Windows software is that the processors share a common implementation of a set of mathematical algorithmic machine code instructions. The software written to run on the OS has to be compiled to run on the shared X86 instruction set, but the source code is hardware agnostic. If the patent is on the programmer's source code, then it is on a set of functions that are unrelated to the underlying hardware. None of the 'software patents' I have seen make any reference to specific hardware, quite the reverse. In Oracle v. Google the 'patented' Java software runs on mainframes, desktops, laptops, mobile phones, tablets, televisions and MP3 players (I might have made the last one up!)

von Neuman came up with an electronic implementation of an abstract mathematical idea of Turing for a universal device that could execute any math algorithm. The mathematical set of algorithmic instructions in computer processors continue to be just that; a set that can execute any program of math algorithms. Two early uses were to aim guns and to pay employees. It does not change the fact that every bit of software is a math algorithm manipulating binary symbols. It is only us applying abstract ideas to that process that makes computers appear to do something else. Computers only do math. If there is patentable subject matter in there, somewhere, then the patent has to show what it is. A magic trick appears to be what it isn't. What computers appear to the courts to do is just a magic trick employing the software arts.

The last case we considered on Groklaw was for the Java system. Neither the source code nor the compiled code has any relationship to the underlying hardware. The same applies for Dalvik applications running in Android.
I thought I would give an example of a patent I find potentially meritorious... Patent 7,098,896 [claims] for continuous stroke input on a keyboard is the type of software patent that the system should protect if it is otherwise valid.

This patent, owned by Swype, was filed in January of 2003. It claims at its most basic the idea of moving the finger from letter to letter on a qwerty keyboard. The motion of the finger is then captured (the shape of the line), and matched against a dictionary of shapes for different words. The system also compares for close shapes and suggests words that might have been intended but the finger didn’t quite get to the correct keys.

Is this math? Absolutely – it encodes data about the touch and does a mathematical comparison. But it is so much more; it provides an interface from my finger to the touch keyboard to the words on the screen in an incredibly efficient and useful way.
This is not a software patent. Software is (today) probably the 'best mode of carrying out the invention'. The description of the best mode 'is satisfied by a disclosure of the functions of the software'. However, any compentent engineer could implement the invention much more cheaply and efficiently with an asic or an FPGA. No processors would be harmed and the invention implementation would be much better. There has to be a chip to encode the keyboard into computer readable key codes. That is the best mode implementation of the invention and not the processor receiving the keycodes.

I use the automatic gearbox for a vehicle as an example of an invention for which there may be several 'best modes' including a stored program control computer.

If the chosen mode is mechanical, then the patent is on the arrangement of gears, actuators, clutches and the control mechanism. The invention is the arrangement of the components and there may be some novel components that warrant a patent in their own right. You cannot make a stored program computer into an automatic gearbox.

You can replace the control mechanism with a computer that electrically interfaces with the actuators. What is patentable is the whole device and not the components. If the original control mechanism was patentable in its own right then it would be the precise control function that would be patented.

Doing the precise control function with a computer should be equally patentable. The software/software function would not be patentable, only the computer/electrical interface/controlling function combination. Most automatic gearbox control mechanisms will not be any more patentable than the gears they manipulate so that even the special purpose computer assembly would be the patent equivalent of a standard nut or bolt.

Both the keyboard invention and the gearbox invention have significant post process activity. The Supreme Court said in Flook and repeated in Bilski that a lack of significant post process activity made the invention abstract ideas. Flook changed the alarm limits in a real world industrial process and the court found that this was a far cry from the significant post process in Diehr that produced precision, cured, rubber components. Although the Supreme Court said they disapproved of an exclusive machine-or-transformation test, they still leave the test as useful for determining patentable subject matter.

In essence, I agree that Benson, Flook, Diehr, Bilski and Mayo provide an existing framework to exclude invalid 'software patents', I see no readiness in the courts to acknowledge the Supremes repeated and pointed demand that this framework be respected. The USPTO never challenge the patent assumption that a system comprising a computer and memory is the best mode and never look for other modes for prior art.

In Fonar Corporation v. GE the court said 'It is well established that what is within the skill of the art need not be disclosed to satisfy the best mode requirement as long as that mode is described. Stating the functions of the best mode software satisfies that description test. We have so held previously and we so hold today'. The patents in Lodsys, Oracle v. Google and in the software part of the 'mobile phone wars' are monopolising a considerable portion of the functions used by the software programming art when the patent has no real world post solution activity such as your keyboard and my auto gearbox.

The only patent that was not invalidated in the Oracle v. Google case was a patent on resolving references to functions with text labels just before the program was executed as opposed to much earlier in the source code compilation process. That is a software art function that can only be about the software and not about some patentable invention.

br

---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid!

[ Reply to This | # ]

"ALL software is a specification of hardware"
Authored by: Anonymous on Monday, June 11 2012 @ 07:33 AM EDT
Does that mean I can patent a piece of music on a CD or in an MP3 file?

[ Reply to This | # ]

Here is the problem.
Authored by: Anonymous on Monday, June 11 2012 @ 07:45 AM EDT
The small inventor does not have the means to avoid a court fight by waving around a portfolio, and a court fight against a large company, even if they don't find some funny counterclaim in their portfolio, will render him bankrupt. Never mind that he actually was the one creating new value.

For the large company, patents are not a principal source of income, they are just a somewhat ambivalent bartering mass exercised in the course of doing business.

Since the patent system realities have perverted to a state where inventing things poses existence-endangering financial risks for the non-manufacturing small inventor outweighing the incentives, and since it favors companies specializing in patent litigation rather than invention, it does more damage than good.

Software patents here are much worse than mechanical engineering patents since

  • The protection duration outweighs the time of usefulness.
  • Basically any non-trivial piece of software can be considered infringing on existing patents, whether meritous or not.
  • In the practice of software, this means that you are usually prohibited from reinventing the wheel, because most forms of the wheel are covered by some patent, or at least the usual patent portfolio will provide enough vaguely matching claims to render you bankrupt in court if you don't fall in line, long before a verdict will be made.

    So if a "settlement" is offered, you better bow to the blackmail.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: globularity on Monday, June 11 2012 @ 08:27 AM EDT
    Well for me patents in general are bad I am an abstract thinker with a sound
    background in Chemistry, Electronics Mechanical systems and well up on
    industrial history. I would argue that less than one in ten patents is anything
    other than work product, given a similar problem somebody of ordinary skill in
    the respective art would have come up with a solution, maybe the same maybe
    better.

    Now down to the software patents.
    Obviousness, Abstraction of communication systems has been taught for many years
    yet time an time again we get patents issued with claims like, using wireless or
    using the internet as if this is an inventive or distinguishing feature it can
    and has been proven that any layer in a communication stack can be replaced with
    something else of equivalent function. This goes much further than computers it
    is a general result of abstraction of function From a mechanical perspective it
    is as silly as saying using a 5/16" bolt or using an M8 bolt both have the
    same function.

    Another absurd feature of the US patent system in general which causes problems
    for areas with vaguely worded patents is allowing the plaintiff to sue based on
    what they claim the patent to mean. The USPTO issued the patent on what the
    USPTO thought the patent meant not what the applicant intended. What the
    applicant intended may not have even been patentable. This anomaly encourages
    vague poorly worded patents and could have someone being sued for infringing
    something different to what was patented which also may be something different
    from what the inventor invented. This is called an update anomaly in database
    systems. It is probably called an employment scheme in legal circles.

    Allowing applicants multiple bites at the cherry is another problem. Imagine
    someone invents something and gets their patent application rejected so they get
    a chance to reinvent it with the benefit if hindsight. Get it right first time
    would be an incentive for precise carefully worded patents of narrow scope at
    the very least it should be sent to the back of the queue with no reset of the
    clock.

    Mentioned was the "inventions" enabled by new technology at best they
    can be called a liability, new technology enables new functionality new
    possibilities are opened up, this is low hanging fruit which should not be
    locked up by patents, of more value are novel ways to use mature technology.

    To give one example, a US patent was issued on maintaining constant surface
    speed during as metal cutting process by reading the cross slide position and
    telling the variable speed drive on the spindle to change speed. Machinists had
    been aware of this mechanically for a century prior it was the availability of
    electronic measurement and variable speed drives which automated the function.
    Hardly inventive, due to the late application years after the technology was
    available, mid 80's there was abundant prior art in use yet a patent was issued.
    It was so obvious no self respecting engineer would have thought of patenting
    it, much of software is like that which is why there is so much derision of the
    USPTO from writers of software.

    Imagine the fuss if authors started patenting plot elements or lawyers patenting
    defense strategies. Why should society be burdened with these artificial
    straitjackets. Software was written before patents, Authors do not need patents
    to write stories and lawyers do not need patents to come up with come highly
    creative arguments.

    ---
    Windows vista, a marriage between operating system and trojan horse.

    [ Reply to This | # ]

    It just does not make sense.
    Authored by: HP on Monday, June 11 2012 @ 08:36 AM EDT
    To operate software the principle of
    input -> processing -> output
    has to be followed.

    For software to be patented the processing part has to be defined correctly:
    Input will be modified by processing in a well defined, correct way to create output.
    The well defined and correct way for that matter is the patent application.

    The prove of correctness for the processing step can only be done by mathematical methods.

    Hence, processing in it's very own is math, albeit written differently than we did write math in school.

    Conclusion: Software can not be patented.

    What frightens me is the fact that software patents would require complexity and uniqueness which is contrary to the need to prove correctness. Patented software then will be put into production with hard to prove or even unproven functionality.

    Are the implied risks for the public from unproven algorithm worth any software patents?

    ---
    hp

    [ Reply to This | # ]

    There is really nothing new here...
    Authored by: Anonymous on Monday, June 11 2012 @ 08:45 AM EDT
    I wrote this elsewhere 2 years ago, that a very big part of
    the overall problem is the USPTO and how they are doing (or
    NOT doing their Jobs). This is all straightforward stuff.
    What this is all about is *politics*, it is a political
    problem that the USPTO is organized and executing in a
    Kafkaesque manner.

    Of course there are also deeper Philosophical issues about
    patents and about the theme of granting these kinds of
    monopolies and if this is truly beneficial to society. BUT I
    don't actually believe that is really the problem at hand.
    The problem we face right now is a political one.

    Fortunately some Judges in this system seem to be wisening
    up to whats going on here and beginning to take a
    significant harder line. I hope that trend continues.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 08:49 AM EDT
    Thank you for this article - it is an interesting contribution to the patent debate.

    On one point I can agree with the author: it is difficult to delineate "abstract ideas" from tangible inventions without ambiguity. Yet I cannot fully agree to the proposed solutions, and also believe that the problems specific to software patents are insufficiently discussed.

    Let me first address the problem analysis. Unltimately, the justification of patents as a legal construct as well as the attitude toward patents by engineers and companies is driven by economic analysis. At a fairly extreme end of the spectrum is the position of Boldrin and Levine ("Against Intellectual Monopoly") who argue that economic analysis speaks against patents as well as copyright at all. While am am personally reluctant to fully subscribe to their position (in particular with regards to copyright), I believe that this type of economic analysis is sufficiently compelling to require an answer, especially by someone who is out to fix the patent system.

    Specifically with regard to software patents, however, economic factors weigh heavily against patents. In most tangible products, there is at least a rough correlation between value of the product and the number of patents which are touched. This is simplest to see for drugs: often, there is only one (or a small number) of drug components. These are relatively easily searchable and negotiable among a fairly limited number of players. The situation is more messy, but not fundmentally different for complex industrial products such as cars. Many patents concern specific functions such as breaks or transmissions, so that there is again some very approximate proportionality between the patent surface and the number of functional components. With software, things are very different because of the strong trend to integrating previously separate functionality into new software products or collections (think smartphones). There is almost no correlation between the physical value of a product and its patent exposure and the tendency to bundle and to extend the software stack vertically makes it very difficult to assign a fair value to any one patent. The value of many of today's patents is closer to the cost of defending against a claim or the potential loss of income due to an injunction than about quantifyable value of the invention as such. I very much doubt that the (clearly sensible) suggestions by the author would alter this assessment fundmentally.

    The second problem is that of interfaces, software interfaces as well as human-machine interactions. Both have the property that they are easy to work around, technically speaking, but with an obvious cost to society. Interface patents are also an extreme example of the situation that patents protect the established market player at the expense of newcomers, which, in its own right, is undesirable. An interface patent is valuable only if you already have a dominant or at least strong position in the market. So I think the interface problem needs to be addressed explicitly, which the article does not do.

    Turning to the author's suggestions: Using "utility" as a criterion for patentability seems elegant, but would create new problems as the cited continuous stroke input patent illustrates nicely. Suppose I was a researcher in machine learning (which I am not, but I know a thing or two) who has a nice generic classification engine running. Now I see the drawing on the front page of the patent which makes the interface component of the idea very clear. Now the hard part is of course the classification engine, but that's what I already have. Now, is the interface part patented, so that I would violate the patent now matter how my engine works? But then it would be a pure interface patent, with all the associated problems which I strongly believe that, as a society, we should find undesirable. Would I need to understand the suggested implementation to see if my engine is doing something similar? But my engine is older than the patent, so what if the underlying algorithm is the same? Would that then invalidate the patent? But if so, I admit to not having had the idea of hooking up my engine to the screen input device. I got the idea by looking at the front page of the patent and then it was clear to me. I don't think the idea of "utility" helps here. My engine was not practically useful - I maintained it to do fundamental scientific research on machine learning. So clearly I wasn't doing anything useful in the sense of "utility" and did not patent my engine. On the other hand, the idea embodied in the figure of the front page of the patent itself should not be sufficient to obtain a patent, it lacks a written description and a best mode, which the author also emphasizes. So what if someone had taken my engine (published under a free license on my website, say) and filed the patent? What if someone had done an independent implementation of the same algorithm and filed the patent, later I use my engine to the same effect? What if someone had done an independent implementation of an independent algorithm, and I then use my algorithm with the same user-facing effect? What if the patent writer had read my papers and concluded that there is an algorithm that works so-and-so and can solve the problem, but hadn't actually bothered to go to my web page and try it out again, and then I am the first one to actually get a truly working implementation?

    I don't think that any of these questions have a clear-cut answer except if one argues against software patents in a broad sense altogether. Any other argument will have a huge gray area even with a supposedly "meritorious" patent (I agree, it's better, especially better written, than a lot of software patents I have had the displeasure of having looked at). But that brings us back to the initial question: is the cost associated with living in a system where even supposed best-practice examples do not seem to lead to clear answers given the highly incremental nature of software development maybe much larger than the economic benefits (unless you are a patent lawyer).

    [ Reply to This | # ]

    Utility for who?
    Authored by: Anonymous on Monday, June 11 2012 @ 09:11 AM EDT
    The utility argument needs some software develop process
    influence e.g. from use-cases.
    Who is the actor for whom the utility is considered? Is
    utility to another software developer not utility?

    It would be hard to argue that a new design for an diesel
    engine cylinder head shouldn't be patentable, but without the
    rest of the engine it has no utility.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 09:44 AM EDT
    I appreciate your arguments, but I must strongly disagree.

    I have been writing software and programming computers for
    over 30 years and continue to do so even in my retirement.
    I have never applied, and never will apply, for a software
    related patent. They are just plain wrong. Writing
    software is just a means to solve problems using a computer.
    There are no great discoveries. Just ones own thoughts and
    reliance of those who have been kind enough to share their
    ideas through books, conversation, or source code.

    Consider your first example. What was patented? It
    sounded like a general way to figure out what someone meant
    to type. If it had been for a very specific method such
    that other methods were indeed possible, it might have
    merit. But, it basicaly excluded any and all other
    specific methods that someone might come up with. It is
    just too general. How does excluding anyone from solving
    the same problem promote innovation and the progress of
    science?

    How many patents are there for shovels? Is there just one
    such that every shovel manufacturer has to get a license to
    make any kind of shovel? I seriously doubt it.

    My own opinion is that software should never be patentable.
    It already has copyright protection. That is enough.

    Writing code is no different that writing any thing else.
    It is a language used to express an idea. So what if it
    causes bits to change? When an effective speaker gives
    a speech, they cause people to have new ideas or change
    their thinking pattern and modify their behaviours. Why
    isn't that patentable? How about legal arguments?

    Software patents also prevent free speech and thus violate
    the first ammendment of the US Constitution. They in effect
    prevent people who write software from using their code and
    thus prevent them from freely expressing their ideas using
    a computer.

    On a good day I probably write original code that infringes
    at least one software patent, if not more. I'd say that
    is true of any experienced programmer who is writing
    software regardless of why they are writing it.

    Why is doing something so ordinary as writing software deserve such protection
    that it puts the common programmer
    in jepardy of being sued? How do you reconcile that? You
    can not. No one can because writing code is excersizing
    free speech.

    Finally, you do realize that your "changes" will never
    happen.

    All this patent bru-hah-hah is only about lots and lots of
    money, making people rich from the labors of others, and
    keeping the wealthy wealthy. Nothing new about that. All
    the oratory and speeches will never change a thing.
    Agruments for software patents are only there mollify the
    disenters and to reinforce status quo. Arguments against
    are ignored or cast as being to extreem because those who
    are getting rich from software patents have too much to
    loose.

    Sadly, regardless of your, my, or anyone elses arguments,
    nothing is going to change for the better with respect to
    software patents. They will continue on as they are or
    just get worse especially if it means puting even more
    money into some wealty persons pocket. History has shown
    us that over and over again that those who can make a
    difference are loath to do so because they have too much to
    loose. You know the old saying: "Those who do not learn
    from history are doomed to repeat it."

    [ Reply to This | # ]

    Economics still don't work
    Authored by: Anonymous on Monday, June 11 2012 @ 10:00 AM EDT

    There's a large community of software developers that invents and distributes software at extremely low cost: Open Source software is usually free and mobile apps cost less than a coffee. This community creates enormous social value.

    These developers do not benefit from the patent system incentivizing disclosure of inventions, but they are harmed by patent suits because they do not have the resources to defend. Software patents form an existential threat to them.

    Professor Risch, if there was a way to remove or protect this community from the rich man's game of patents, I think your suggestions would fix software patents. Otherwise, your suggestions still form an existential threat. Are software patents worth the social cost?

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 10:01 AM EDT
    Thanks for providing a thoughtful and detailed look on how
    some software patent concerns could be addressed under
    existing law.

    Something's going over my head though: how does this help?
    To me the two main issues are (i) nonpracticing firms
    attacking large companies and (ii) large companies shutting
    smaller companies down that can't afford to litigate.

    Applying precedent and law in the way you suggest might
    _eventually_ help with both of these issues by thinning the
    patent pool; that looks like a very long and astronomically
    expensive cure. Would you mind elaborating on why that's
    preferable to addressing the issues with new legislation (or
    a significant new ruling)?

    [ Reply to This | # ]

    Minimizing impact of software patents
    Authored by: Anonymous on Monday, June 11 2012 @ 10:27 AM EDT
    If we are going to continue to have software patents, how can we minimize the
    damage that they cause to competitiveness while providing some reward to the
    inventor? While I would rather not have software patents, I don't think they are
    quickly going away.

    Some suggestions:
    1. Eliminate injunctions as a tool for software-related patents. No injunctions
    from either the ITC or Federal courts. This will encourage monetary settlements
    rather than "swinging for the fences" (a baseball term implying a high
    failure risk/high return approach).

    2. No royalties on software provided as FOSS. Let the proprietary companies
    figure out how to deal with royalties in their proprietary software.

    3. Royalties calculated on actual value added to entire solution, not in an
    abstract sense. Especially in light of the Motorola/Apple case, junk patents
    won't be worth filing. Also, the royalties should be on a decreasing scale, as
    they typically become a smaller part of the value of a solution over time. For
    example, had Google infringed on one of Oracle's patents that improved
    performance, the royalty might have been based on how much a faster processor
    and bigger battery would have been needed to get a comparably performance
    device.

    4. Some lower cost mechanism for determining royalties is needed short of
    clogging Federal courts and requiring expensive lawyers.

    5. Eliminate treble damages for willful infringement.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 11:13 AM EDT
    Thanks again, Michael, for your paper. It has released a
    plethora of comments, and I'm glad to see some of our
    regular contributors are a lot less long-winded than before
    (a complement, not a criticism), or should I say 'more
    succinct' {:-)>

    Let's step back a little. Many years ago, when one could
    physically look up patents in a big library(!), I researched
    a particular class of patents, the nature of which is
    irrelevant. They were mechanical and electrical devices.
    The point is this: ALL were easily duplicated by anyone (me)
    skilled in the art. The drawings and description were very
    easy to follow. Now, years later, I still see these patents
    referenced by newer patents of similar devices. Of course,
    these newer devices are more complex, but still capable of
    being duplicated. The patent author did his job, and the
    examiner only had to check each reference for infringement,
    and, if appropriate, grant the patent. THIS is how the
    system is supposed to work. I believe the anti-patent
    positions are well-argued, but the whole patent/anti-patent
    discussion has gotten WAY out of hand, and way too complex,
    even for those of us skilled in the art. My challenge is
    this: How do software patents fit into the original patent
    scenario as described above? If they don't fit, what
    arguments can be made for continuing to patent software?
    (My conclusion is: none. S/w patents are a con game, a
    lottery, benefiting no one.)

    Albert The Un-Registered

    [ Reply to This | # ]

    Why would the problem be limited to the merits of the patents?
    Authored by: PolR on Monday, June 11 2012 @ 11:36 AM EDT
    It appears to me that this proposal is targeted at improving the merits of the patents being granted by a better application of the current mechanisms of patent law. But how did we get to the conclusion that the problem of software patents is too many patents without sufficient merits? Is this really the right problem to solve?

    Before State Street software patents we few and far between. But software as in industry thrived. From an historical perspective I see no reason to believe patents are needed of their lack has limited innovation in any way.

    The practice of sharing software in source code form has a long history. It predates the terms "free software" and "open source". These terms refer to the codifications of this ancient practice accompanied with a legal framework for the copyright licenses. However sharing of source code has helped innovation and the advantages of sharing give a strong incentive to disclose the invention. Source code is disclosure. Working source code which is actively maintained by developers is disclosure which is guaranteed to match exactly what works in practice and which is kept current as the invention is improved. No patent can match this.

    Many business and legal people think of FOSS as a fringe movement. But when we look at which technology has been developed and disclosed by means of sharing source code it is clear that this is not fringe.

    • The UNIX operating system, especially BSD Unix
    • the X-Window graphical user interface
    • The TCP/IP protocol stack
    • The world wide web, including protocols, browsers and web server
    • Several programming languages: C, Perl, Python, PHP etc
    These technologies are of paramount importance and they are among the most important innovations in software.

    Here we see an argument on how patent help innovations based on a data entry method by swiping on a touch screen. Why do we need patents to get such innovations when we can get the Internet and the world wide web by FOSS methods?

    Here we have a proposal which:

    • Acknowledge that the patent system is dysfunctional, ie issue many bad patents.
    • Brings up a proposal to cure the dysfunction.
    • Success is uncertain. If the system is dysfunctional now, how can we be sure we will succeed in making things better?
    • If successful the improvement is that the newly issued patents will be more meritorious according to the standards of patent law and this is believed to make the patents more meritorious period.
    • We don't know how long it will take to get the patent system in proper working order of even if we ever get there.
    • It is unclear how the current inventory of issued poor patents will be handled and how long they remain into effect.
    I agree this plan will lead to some improvement over the current situation, although I question how big an improvement this will be. But why should this be the situation we aim for? Patents are not just about providing incentive to innovators. They also impose liabilities and litigation risks on everyone else. In the best of scenario, this proposal means that:
    • The playing will remain tilted in favor of those who can afford the legal costs of patent litigation.
    • Licensing requirements for patents will remain incompatible with FOSS licenses.
    • The development models which has led to the innovations mentioned above will bear all the inconveniences of patents with none of the advantages.
    The argument in favor of software patents should not compare with absence of intellectual property protection and absence of disclosure. It should be based on whether there is an incremental improvement over the situation that preceded State Street taking into consideration the effect of liabilities and risks on FOSS developers.

    [ Reply to This | # ]

    This quote is simply outright false:
    Authored by: Anonymous on Monday, June 11 2012 @ 11:36 AM EDT
    : “ALL software is a specification of hardware. ALL software is run very
    differently on different hardware and is written for specific hardware
    platforms. Software is hardware in every sense that matters for patents.”

    Your patent lawyer friend is simply wrong. Software is not a hardware
    specification. Software is not written for specific hardware platforms (hardly
    ever, anyway). Software is not run significantly differently on different
    hardware; to the extent that it is run differently, the hardware hides the
    difference from the software programmer. Software is not hardware in any sense
    that matters for patents.

    Your problem, Mr. Risch, is that you have been listening to people who don't
    know their facts. Bad facts make bad law -- false "facts" make worse
    law.

    [ Reply to This | # ]

    A political problem with a financial reason but could just a bad joke.
    Authored by: David665 on Monday, June 11 2012 @ 11:56 AM EDT
    From a political view point I always thought the problem was directly related to
    funding.
    The patent office has to increase patent filings to have funding. Thus award
    more patents enticing others to file. Patents that get challenged also collect
    fees so a proper combination of accepting bad patents will not only increase
    fees for new patents but increase the challenges so a perfect balance will cause
    a corresponding increase in revenue allowing ... more patent examiners.

    Since this is about software GOTO top.

    The cycle is complete. :-P

    [ Reply to This | # ]

    Discoveries and inventions.
    Authored by: Anonymous on Monday, June 11 2012 @ 12:06 PM EDT
    How does patent law distinguish between the two?

    As you know, computers are machines that take specific inputs called software
    and produce specific machine outputs. The inputs, though large, is ultimately
    nothing more than a pattern of 0's and 1's.

    Theoretically, then, ALL POSSIBLE software can be derived randomly. Therefore a
    set of inputs that outputs something useful is nothing more than a discovery.

    [ Reply to This | # ]

    There are TEN major problems with software patents.
    Authored by: Anonymous on Monday, June 11 2012 @ 12:16 PM EDT
    It's useful to separate out the ten separate problems with software patents.
    I've sorted them from most broad-rangingly problematic to smallest; even the
    smallest of the ten problems is enough to wreck the nation's computer industry,
    but fixing the larger problems would solve the smaller ones.

    (1) At least outside pharmaceuticals, patents harm industry, period. (The
    economic literature is clear on this.)
    (2) Software is fully abstract mathematics and as such patenting it preempts the
    use fundamental laws of nature and is invalid according to all precedent.
    (3) Software is fully implementable in pen and paper and as such patenting it
    patents mental activity.
    (4) Software does not "create a new machine", so all of the software
    device patents are invalid.
    (5) If you consider a "process" patent for software, this can only be
    violated by RUNNING the software, so that ALL the patent lawsuits actually filed
    over software "method" patents are invalid: it is not possible to file
    primary infringement lawsuits against a seller or distributor of software, only
    "inducement" of infringement.
    (6) Software processes are not "industrial", so in general the process
    patents are invalid as business-method patents anyway.
    (7) Actual software patents are generally on obvious mathematics, not difficult
    mathematics.
    (8) Actual software patents are practically always on unoriginal mathematics in
    the prior art.
    (9) Actual software patents *always* fail to disclose the implementation
    (where's the source code?)
    (10) Actual software patents frequently attempt to patent all possible solutions
    to a problem, not a particular solution.

    If you allow mathematics to be patented, you still have to fix problems (7)
    through (10). But you shouldn't allow mathematics to be patented -- precedent
    says mathematics is NOT patentable -- which will fix (2) through (10). Fixing
    (1) entirely requires actually changing patent law to reduce patenability.

    [ Reply to This | # ]

    "crowd source" patent examination
    Authored by: Anonymous on Monday, June 11 2012 @ 12:19 PM EDT
    Because patent examiners have limited resources and must err on the side of
    granting the patent when unsure, the burden of proving validity (and related
    costs) thus shifts to the accused infringer.

    One possible fix is to "crowd source" patent examination where a
    voting system is used to determine validity.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: miltonw on Monday, June 11 2012 @ 12:34 PM EDT
    I realize the Supreme Court has said that algorithms are unpatentable. My view is that this is an untenable solution, because every process patent is an algorithm, mathematical or not. And it can't be that every process is unpatentable.
    You make the claim about software that "it can't be that every process is unpatentable" but that's not a good starting point. That isn't a given, that isn't proven. If you use that as your starting point, you are already half-way down a road that is causing lots of problems.

    You give, as evidence that software can't be unpatentable, one "valuable" software patent. But that doesn't excuse all the harm that is done by software patents.

    Then, you try to deal with all the problems that software patents cause. Realistically, no one is going to implement all the fixes you propose. So, in the real world, following your idea, we're going to be left with software patents and all the harm they cause.

    In the alternative, we don't allow software patents, we invalidate all those patents and all that harm goes away.

    Would there be some harm because some "valuable" software isn't protected? Maybe. But there are lots and lots of very valuable ideas that are not protected by patents, and the world survives, companies that created those ideas survive.

    We don't need to protect every "valuable idea" with patents. Really, we don't.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 01:42 PM EDT
    It took me a while to find it buried in about the fifth
    paragraph, but I think we agree: the most pressing problem
    with software patents today is that most (all?) of them are
    totally bogus non-inventions, stuff that was already known
    in the prior art. Your primary solution, if I understand
    you:

    >it is more efficient for the parties that have a stake in
    the matter to fight about validity at the time of
    litigation.

    You go on to say that the patent office can't be expected to
    do its job, i.e., weed out such bogus patents, at least as
    long as it continues to be (under-) funded in the ways that
    it is.

    It seems to me that these observations lead to a very simple
    legislative improvement:

    ABOLISH THE PRESUMPTION OF VALIDITY.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 01:47 PM EDT
    I'm retired. Over 50 years in the computer software and hardware business.
    Physicist with minors in math and EE. Have done things that were patented. They
    were all physical objects, not "writings". That is, the patents spoke
    about actual physical object. In the olden days one would bring an actual
    physical object to the patent office. Nowdays one needs only to bring a
    description of that object.

    So it is with a software "patent". One brings a CD, DVD, tape,
    or paper with the software on it. The physical object is simply a means to bring
    the software to the patent office. Does it contain "the software". No,
    it does not. It contains representations of the software, descriptions of the
    software, explanations of what the software does. The software is in no way a
    physical object. I have heard over and over the concept that "software
    turns a general purpose machine into a specific physical object." This of
    course is silly. It does no such thing. The general purpose machine remains a
    general purpose machine. The software is simply a set of instructions ... ideas
    ... abstract ... non-physical. Protect it via Copyright? No, that makes no
    sense. Protect the description of the software, the source code or the
    specification of the software? Certainly. Those are documents no different than
    a novel or a short story.

    In my view what is needed is a new "thing". Not a Patent, not a
    Copyright, but a specific form to protect intellectual property, i.e. ideas.

    --- Retired Hacker.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 01:59 PM EDT
    I also agree with you that arguing about the abstractness of
    software (i.e., whether "software is math") is a waste of
    time, but for totally different reasons. I think it's both
    politically impractical and largely irrelevant. Nothing in
    the statute says that "you can't patent math", as the term
    "math" is understood by the folks who say that "all software
    is math".

    But you are dead wrong about "software is written for
    specific hardware", and mostly wrong about "software is
    executed differently on different hardware" (well, it is, at
    a low enough level, but that's not the level that's being
    patented).
    Any computer-science major will tell you that software and
    hardware are "logically equivalent": they set of
    calculations possible to perform in software is the same as
    those that can be done in hardware. Furthermore, any
    hardware circuit can be simulated in software, and any
    software program can be implemented in hardware. The choice
    of hardware vs software comes down to economics (cost and
    development time vs reliability and performance). You
    *could* implement Linux in a single hardware circuit, but it
    would be a big effort. You could emulate a Pentium in
    software, but it would run too slowly to be useful.

    Drawing a subject-matter line between "software" and
    "hardware" is not always easy, but it's not the end of the
    world. It'd be a big improvement compared to the insane
    stranglehold occupied by software patents today.

    Bottom line: you correctly identified the most pressing
    problem (non-novel patents), and you acknowledge that the
    PTO cannot ever be expected to do the job of enforcing
    standards of novelty, but then you go on to suggest tighter
    enforcement of those same standards??? You mean at
    litigation, after a patent issues? Sounds like a money-
    making proposition for patent lawyers, but why on Earth
    would that be a good solution for anybody else?

    [ Reply to This | # ]

    Interesting, but not convincing
    Authored by: dwheeler on Monday, June 11 2012 @ 02:43 PM EDT
    This is a very interesting article. My thanks to Michael Risch for writing it,
    and to PJ for posting it.

    However, it is fundamentally not convincing to me. The proposed example of a
    "good patent" is an example of the PROBLEM. If this is the best
    example of a "good software patent" then we need to eliminate software
    patents immediately (and we do). This example is a proposal to patent
    understanding text written continually. This is just handling cursive writing,
    but with the letters placed on a keyboard. That doesn't seem innovative to me,
    it's spectacularly obvious to anyone skilled in the art. If fact, many
    16-year-olds can probably think this up. What's more, I see no evidence that
    there is a social advantage to granting a 20-year monopoly to this and related
    ideas. There is good evidence people will write and implement this code even if
    no patents exist. In fact, it will probably be implemented more widely WITHOUT
    software patents.

    This essays leaves the real issue unaddressed: Patents exist to "promote
    the progress of science and useful arts". There is no evidence given in
    this essay that software patents actually promote progress. Indeed, no evidence
    has, to my knowledge, ever been shown (other than "proof by repeated
    bluster"). Hence, software patents are unconstitutional.

    Another problem: Leaving the software patents for COURTS to decide is completely
    wrong-headed. Software developers cannot wait to develop software based on the
    speed of court cases. The Microsoft/Novell case is still going on, long after
    it matters. And if the patent office cannot do its job, why have one? Why not
    abolish the patent office entirely? If we got rid of software patents and
    business patents, the PTO would no longer be overwhelmed, and instead be able to
    actually do its job.

    Now I would agree with much of the rest of the text: If it is impossible to fix
    the patent system to meet its constitutional mandate, but we must instead accept
    that software patents will continue to harm humanity, then various mechanisms
    (like this) to enforce the dead letter of current rules would certainly help.
    If we must be stuck with the bad system, clearly forcing actual novelty in
    software patents would be an improvement. I'd even be happy to help those doing
    so. But let's be honest: these would be improvements because they would reduce
    the number of software patents, which are the problem. When "0" is
    the right answer, making the (nonzero) number smaller is a good thing. But that
    approach still just works around the edges, helping instead of solving the real
    problem.

    I'm not against all patents. A process that involves a permanent change of
    material (e.g., rubber processing), does make sense to patent. But when there
    is only processing of information into other information, those should be out of
    scope of the patent office, whether there is a computer involved or not.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: rsteinmetz70112 on Monday, June 11 2012 @ 02:44 PM EDT
    One of the most effective ways of limiting the damage from bad patents would be
    to require an actual implementation of of the patent. That is actual code that
    can be run on an actual computer with appropriate peripherals. This would enable
    litigants to see what the inventor was actually thinking and doing, not what
    some patent attorney wrote because it sounded better or broader.

    The patent office used to require models of inventions, and stopped when they
    ran out of room to store them. It doesn't take much room to store some software
    and some specifications.

    ---
    Rsteinmetz - IANAL therefore my opinions are illegal.

    "I could be wrong now, but I don't think so."
    Randy Newman - The Title Theme from Monk

    [ Reply to This | # ]

    TSM assumes a super-UN-skilled practitioner
    Authored by: Anonymous on Monday, June 11 2012 @ 03:04 PM EDT

    In your description of the (bad) TSM test, you say it assumes a "super-skilled" engineer who can mentally cross reference all prior publications. That is not really what it assumes. It assumes as super-dumb computerized robotic engineer who can cross-link all prior art like a human- language-understanding search engine but cannot actually think. And that assumption of complete lack of skill in the art is what made that test so fatal.

    By requiring prior art to be obvious to a lawyer/computer wholly unskilled in the art, the test turned a key legislative requirement on its head and allowed patents on anything so obvious to those skilled in the art, only a complete idiot or unskilled freshman would think it was novel.

    This problem with the test was in no way limited to software, it was just the area of technology lacking enough precedents and examiner experience to prevent its application.

    Now the second assumption, that an excessively abstract description would be enough to teach those skilled in the art to practice the invention, does presume superskilled practitioners and thus directly contradicts the first assumption made by the Federal Circuit and the USPTO at the time.

    Such contradictory assumptions, both in favor of the same side really shouldn't be allowed to stand, as a matter of simple judicial logic, one must choose a level of assumed skill and then stick with it for both obviousness and enablement requirements, across all or most cases in an area of technology (an "art").

    [ Reply to This | # ]

    Everthing is Patentable - False Premise
    Authored by: Anonymous on Monday, June 11 2012 @ 03:42 PM EDT

    There's a lot that can be discussed, but I think we have to take this kind of extraordinarily complex discussion in baby steps to be able to properly understand.

    The problem with starting with a false premise is that one ends up with a lot of conclusions that stem from that premise that are incorrect.

    "Everything is Patentable" speaks from the premise that the existing State of Law does not support. Not everything is patentable as math is clearly not currently patentable.

    Perhaps the statement is speaking from the perspective of what the Law should be instead of what it is. In which case:

      Why should the concept of Gravity be patentable? It exists all around us impacting all of us equally. If a person jumps to their demise, should a "patent holder" really be rewarded with funds from the persons estate because the person made use of Gravity?
      Why should the concept of the Earth revolving around the Sun be patentable?
      Why should the light we see that is attributable to our nearest Galaxy be patentable?
    Humans were not involved in creating any of that. Yet they are discoveries and by the specific limited (ignoring such things as math is not patentable) wording of Patent Law the people who discovered said items have obviously earned a potential Patent.

    However, if one feels that such should not be patentable, then starting from a position of "Everything Is Patentable" is misleading at best.

    As a result, here are two very valid questions for Michael Risch:

      Why start from an easily provably wrong position compared with current Patent Law?
    And if it's a question of "what should be":
      Then why start from the assumption that "what should be" actually is and completely bypass the valid questions surrounding such things as why the concept of Gravity should even be patentable in the first place?
    There's a claim to be balanced in between patentability and non-patentability. When one starts from such an easily provable false infrastructure - balance does not honestly exist.

    I suggest Mr. Risch rethink whether or not there's a valid starting point with defining what should be patentable and what shouldn't be to begin with.

    The harm that results from patents will be brought out into the open and weighed against such benefits patents provide so they can be kept in mind while subject matter is considered for patent protection. Only then, with the harm kept in mind as the initial definitions which form the infrastructure of everything else, can we ever hope to gain a measure of sanity in the madness that exists today.

    A madness that seems to be extending to normal people. Just today I read an article of a woman who named her baby Benjamin (Franklin anyone?) who was upset that someone else she knew also choose the name for the other woman's baby. Some people have even appeared to have trademarked their child's name. Heaven forbid such a future exists where that kind of behavior seems normal.

    RAS

    [ Reply to This | # ]

    Swype Patent 7,098,896
    Authored by: Anonymous on Monday, June 11 2012 @ 04:53 PM EDT
    Does the patent cover the idea, or the actual method of implementing the idea?

    If someone else came up with a method of "slide-typing" which was not
    the method of implementation described by the patent would it infringe, or is
    the patent on the idea of "slide-typing" which means that no one else
    could do it, regardless of method of implementation?

    [ Reply to This | # ]

    On the subect of definiteness
    Authored by: Anonymous on Monday, June 11 2012 @ 05:02 PM EDT
    Definiteness: There’s not much to say here; software patents are often impossible to understand. Examiners and courts should demand more definiteness, as required by the statute: “The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.” 35 U.S.C. § 112.
    I hope the examiners are aware of the advice the USPTO [website] gives to patent applicants regarding how to write the patent (which I assume is advising on 35 U.S.C. § 112):
    The specification must include a written description of the invention and of the manner and process of making and using it, and is required to be in such full, clear, concise, and exact terms as to enable any person skilled in the technological area to which the invention pertains, or with which it is most nearly connected, to make and use the same. [Emphasis added]
    If, as is asserted, "software patents are often impossible to understand" then clearly (sic) they are not written in "such full, clear, concise, and exact terms" as required and as such the patent is invalid?

    In which case are the USPTO [examiners?] guilty of negligence in granting the patent, especially if it is then used for [legalised] extortion (no better word for it)?

    [ Reply to This | # ]

    Two arguments against software patents, on principle
    Authored by: Imaginos1892 on Monday, June 11 2012 @ 05:17 PM EDT
    First argument:

    A modern digital computer is a state machine consisting of a
    large number of electronic circuits designed to express several
    billion abstract entities called "bits". These bits are not part of
    any circuit and do not exist in any physical sense; they are
    expressed as certain patterns of voltage or current within the
    circuits. Each of these abstract bits can be set to 0 or 1, and
    the aggregate values of all the computer's bits at any instant
    in time is called a "state". The computer's entire purpose is to
    manage those states and the transitions between states, which
    are accomplished according to rules encoded into the CPU,
    called the "hardware instruction set".

    Computer software is simply a way of describing certain specific
    states, and "loading" or "installing" software is merely a
    matter of setting the bits to those states.

    Thing is, every possible state, including every program that has
    ever been or can ever be written, is already inherent in and
    anticipated by the design of the computer hardware. The entire
    field of programming and software development is just a semi-efficient
    way to select states that result in the computer's peripheral
    hardware devices doing something interesting and/or useful.

    And that's the point! Programmers are not INVENTING new
    states, they are DISCOVERING uses for states which are
    already inherent in the computer. Finding those states, and
    separating them from the many similar states that do ALMOST
    what you want, can be difficult and time-consuming, but it
    is NOT invention! It is selection and exclusion.

    ---------------------------------

    Second argument:

    A computer's CPU chip is an engine for manipulating symbols
    expressed as binary numbers according to mathematical rules.
    These are the only operations it is capable of. Any program or
    software to be run on that computer must first be rendered down
    to those simple mathematical operations, or the computer will
    be incapable of processing it.

    Anything that consists entirely of mathematical operations is
    math, and can be nothing other than math.
    ---------------------
    "That's not an argument, that's just contradiction!"
    "No it 'tisn't!"

    [ Reply to This | # ]

    Summing up a bit...
    Authored by: mrisch on Monday, June 11 2012 @ 05:41 PM EDT
    Thank you all for your comments. We're at well over 500, but
    a lot of them are Anonymous, so I don't know if it is 500
    people telling me I'm wrong or 5 people telling me I'm wrong
    100 times each. :)

    I've tried to keep up on the point by point comments, but
    I'm running out of time for the day, so I thought I would
    summarize a few of the themes and respond.

    1. Patents are bad in general. This is an old debate. I
    don't think so, but that's a whole other discussion.

    2. Patent terms are too long. I tend to agree, especially in
    software where commercialization is easier and fast moving

    3. Software patents can hurt open source and small
    developers. No dispute there - but also true of other
    patents. There is a trend away from individual inventing in
    this country, and the irony is that software is one of the
    few areas really open to individuals anymore. I'm sure most
    readers here would be willing to jettison that if it meant
    getting rid of software patents.

    4. Software is all math, and thus not patentable. Yes, I get
    it. You don't think I do, but I do. I just think we are
    talking past each other. Here's why:

    a. I'm not sure we're even talking about the same thing.
    Some of you have said that the Swype patent is not the
    software you're thinking of. If that's true, then we may
    actually agree. Software that just moves information from
    one place to another is probably not patentable - I would
    argue as a matter of practical utility.

    b. Even if software is all really math in theory, I'm
    talking about applied math. We've had applied math as
    patentable in the past - some of you have said the rubber
    process in Diehr. There's also the MacKay Radio case - a
    mathematical equation applied to an antenna, etc.

    c. I have read hundreds and hundreds of computer science
    journal articles dating from the 1980's until now. These are
    written by theorists - the people some have said will teach
    me how I'm wrong and that all software is math. Not one of
    those articles has said, "well, what we've done here is
    math, so it was always theoretically possible and you could
    do it on any hardware anyway and so we didn't really do
    anything important." Even when they present mathematical
    proofs, they then show how it applied in a particular
    environment and how the solution was new and different from
    what others had written. I'm not saying that these folks
    support patenting, only that they recognize the practical
    consequences of their work.

    d. You can derive practical benefits and have computers
    that are in practice different from each other, even if in
    theory they are all the same.

    e. Many of you have criticized my view that hardware
    matters. Yes, in theory you can run linux on an HP
    Calculator. But c'mon, really. Is each of you really saying
    that you've never had to deal with constrained resources?
    Moving big data over a small network pipe? Conveying a lot
    of information on a low resolution screen? Performing
    calculations with slow processor and little RAM? Yes, Java
    runs everywhere - knock yourself out on that old XT. Of
    course hardware matters.

    5. This last point is the fundamental problem of software
    patents - stuff that is impossible on hardware suddenly
    becomes possible when hardware improves. The hard part is
    figuring out whether the inventive solution was obvious
    based on improvements in hardware or not. This is the debate
    we had with Swype - the big idea was the finger gesture
    represents word (as someone put it). Is that just a simple
    idea suddenly made possible by a hardware improvement in
    2003? Or was it a brilliant idea that was possible back when
    graffiti was patented in 1993, if only someone had thought
    of it? Most of you seem to lean toward obvious - and I'm not
    disagreeing. My only point is that the debate should be
    focused on this point: is the invention meritorious, and not
    focused on whether it is this software and therefore
    unpatentable in its own right. Further, we have to better
    define what we mean by obviousness - Do we require a
    nonobvious idea? a nonobvious implementation? both? There's
    not a lot of agreement on this in the patent community.

    6. Some of you have pushed back on the idea it is more
    efficient to fight patents outside the patent office.
    Perhaps as the number of suits grows, this is true. However,
    note that examiners let all these bad patents through in the
    first place. I agree that making it more costly to get
    patents would help. Note that a majority of patents expire
    long before their terms are up due to failure to pay
    maintenance fees.

    7. A couple of you have commented on some fundamental
    problems with software patents, like a lack of prior art. I
    agree. This particular problem is getting better, and I'm
    not sure we should parcel out what types of things get
    patented based on prior art availability.

    8. There has been a lot of talk about software creating a
    "new machine." My general response to that is below (though
    some of you have already posted disagreement there):
    http://www.groklaw.net/comment.php?
    mode=display&sid=20120610180253648&title=Well%2C+since+my+cr
    edibility+is+on+the+line...&type=article&order=&hideanonymou
    s=0&pid=985355#c985779

    As I note in that post, structure is irrelevant for the most
    part, capabilities are what matter. As I also note, I think
    the focus is misplaced since a process is patentable whether
    you create a "new" machine or not.

    9. Many of you note that a problem with patents is that they
    don't teach anything and they are too broad. I agree on all
    counts. The way we deal with this in patent law is through
    enablement and written description rejections. That said, I
    wonder whether many readers think some patents are broader
    than they are. Some of the comments about the Swype patent
    have implied that it: a) covers all shape pattern matching,
    b) all touch keyboards, or c) even all swiping. I don't
    think it does any of these things. I think that it does
    cover a fundamental way to do a particular task, there are
    other ways to do swiping - indeed, other ways have been
    suggested in the comments. The Oracle v. Google experience
    should show that software patents can be narrow and non-
    infringed - quite easily, apparently.

    10. Some of you have implied that because I'm a "fan" of
    certain technology I must have lost objectivity about the
    quality of the patent. I assure you that's not true. I spend
    a good part of my days looking for prior art to invalidate
    patents, including the Swype patent. I can't talk about it a
    lot, because I am under a confidentiality agreement. So, I'm
    well aware that even the coolest app that someone claims to
    have invented may not be new and may be obvious. In fact,
    that's always my first assumption. But this also means (as
    I've noted above) I have read hundreds and hundreds of
    articles about the state of the art at a particular time -
    and that art is often well ahead of what is going on
    commercially. While I would like to think that everyone
    out there is using some patented process long beforehand, it
    is sadly not true - usually because the hardware available
    at the time doesn't make the solution commercially feasible
    (note the hardware matters theme).

    That said, in reading the literature, I can tell when lots
    of people are thinking about the same ideas at the same time
    even if they haven't coalesced. And I can tell when people
    are thinking the opposite of a patented solution and the
    idea really is new. The ideas are almost never really new -
    but sometimes they are.

    This is all a long winded way of saying that I appreciate
    your comments. They are very helpful for me to hone my
    arguments and make sure that I'm precise about what is at
    issue.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 05:56 PM EDT
    Thank you for your views. I believe that, if properly
    implemented, your suggestions would reduce problems caused
    by new software patents.

    I see a couple of remaining unresolved issues that I suspect
    require more than broader application of existing criteria.

    (a) The current presumption of validity is fairly unjust
    because roughly 90% of software patents seem to be invalid,
    even under current rules. Most small businesses (<100M)
    don't have the resources to fight off lawsuits over invalid
    patents.
    (b) The USPTO's current incentive is to grant patents ASAP
    and the inventor's current incentive is to make patents as
    broad as they can get away with.
    (c) Patent duration is reasonable (or a little-short) for
    manufacturing industries. It is too long for software.
    (d) Software patents are completely unreadable.
    (e) Injunctions are problematic. (too many patents in a
    given product)

    Suggestions?
    (a) Remove the presumption of validity and, prior to suit,
    require a public 'validity exam' - and identification of any
    entities to be targeted within the next 3 years. The cost
    should be ~2M USD, of which 1M USD would be available as
    prize money for anyone submitting a brief leading to
    invalidation. Just imagine a world where someone who was
    notified of a potential Oracle-style lawsuit popped some
    champagne.

    (b) Change the process to something where the USPTO benefits
    from rejecting patents:
    Current:
    Submit->fee X.
    If rejected, resubmit.

    To:
    Submit->fee X.
    If rejected, resubmission requires fee 3X and appeal
    requires bond 3X. All appeals require 'beyond a reasonable
    doubt' and patent examiner is presumed to be correct.
    (iterate as needed->main point is that inventors need to be
    motivated to not waste the USPTO's time.)

    (c) 10 years from submission date probably makes more sense.
    5 might be a bit short. 20 years from acceptance is too
    long.

    (d) Infringement should require 'beyond a reasonable doubt'
    Not understanding the patent is a reasonable doubt. Most
    reasonable patents can be explained with great clarity.

    (e) If a particular patent contributes less than 30% of a
    product's value, don't allow injunctions.

    Unresolved issues?
    Free software. I don't see a way to simultaneously prevent
    M$ from giving away reverse-engineered software to kill
    competition and allow Linux to distribute a similar free
    version of the same software.

    --Erwin

    [ Reply to This | # ]

    Nonobvious problems
    Authored by: Anonymous on Monday, June 11 2012 @ 08:02 PM EDT
    Unlike most here, I'm generally in agreement with the author that software
    patents are not bad in principle, it is just that they have been very poorly
    implemented. For example, the RSA public key cryptography algorithm is very much
    useful and nonobvious and deserving of patent protection.

    There is a patentability criterion which is not discussed in the article and
    which would eliminate many bad software patents. The criterion is this: it must
    be the *solution* which is nonobvious, not the problem. If the problem is not
    obvious, but, once stated, the solution is, then the solution should not be
    patentable. I.e. once you realize you have a problem, you should be able to
    solve it without worrying about a patent simply because someone else anticipated
    a few years ago that you might have this problem.

    An example: (Sorry, all from memory from years ago, so no citations, and may be
    grossly inaccurate.) Someone anticipated the Y2K problem well in advance, and
    patented a way of solving it. (Sorry, there were no details in the article I
    read about the 'solution'.) Then post-Y2K he attempted to sue a whole bunch of
    organizations, including the New York Times, on the basis that they'd solved
    their Y2K problems so probably had used his method.

    (If the problem *and* solution are nonobvious, then a patent can be issued, but
    it applies only to the solution. Other solutions to the same problem are
    non-infringing.)

    [ Reply to This | # ]

    Things, however, are not as they seem...
    Authored by: Anonymous on Monday, June 11 2012 @ 08:32 PM EDT
    re Swype, which I use and like...

    As described by you, Swipe is "mouse gestures", e.g. draw a shape,
    compare shape to database of shapes. Do shape action.

    You credit Swype as "creating an interface between your finger and the
    touch keyboard" which is, IMHO, wholly over generous. The touch screen and
    the touch tracking elements therein, and the operating system "create the
    interface", Swipe does a (reasonably novel) job of reinterpreting the
    input.

    This is a classic example of the conflation of apparent action and actual
    function, or as I like to call it "giving credit to the line" (which I
    will now explain).

    Circa 1997 I was involved as an employee of a subcontractor of a contractor to
    the FAA to produce the new backup satellite data network for air traffic
    control. I was fascinated to discover that internally the various
    "regional" traffic zone managers are competing with each other for
    best-in-show performance. As such they were -alarmed- to discover that managers
    in other regions would be able to see "their" local faults real time.
    Such minor failings were to be kept secret so that they could, I don't know,
    spin them? Hide them? Whatever...

    So one of the actual managers came up to me and said "can't you just draw a
    line down the map here, and here, and have the system only show errors between
    your own lines?"

    See, good software makes the user action seem like the function. I could easily
    draw the requested lines, but creating a security barrier and information
    filters to "back that line up with functionality" would take a whole
    redesign of the system. The "regular Joe" user of things like
    power-point cannot tell the lines from the functionality. As such they
    constantly miss-attribute the functionality of a system at such a high level
    that its neigh-on impossible to make them understand that what they think is
    happening is not what is happening at all. Plato's cave, blind men and
    elephants, and all that.

    Lawyers, especially Patent Lawyers, are just "regular Joe(s)" and they
    typically take a rewrite at the patents that makes them useless. (That's what I
    experienced when I was forced to write up one of "my inventions" that
    was then not even attributed to me in the final patent paperwork since the final
    patent filed was a disjunct set to the actual functioning of the device.)

    Note that I am -not- an "all software is math" proponent. I know all
    software can be "described with math", but planetary motion can be
    described as math without making planets math... To a mathematician everything
    is math.

    In that same sense, to a patent lawyer all software is machine because they
    -must- see it that way in order to be patent lawyers else-wise they would never
    be able to take that software patent bite. yes?

    We -wouldn't- have to change the laws to fix software patents because honest
    people know that software is already unpatentable. Not just because of the
    "math thing" but because the "machine" is the
    "computer" and not the software.

    There are lots of complex machines out there. My personal favourite to cite is
    the sewing machine and its buddy the "surger". Over time these two
    close-cousins have gotten very complex even at the home-use level. Go look some
    high-end sewing machines and surgers. They will auto embroider and sew-in button
    holes and buttons.

    People got patents on all those features and add-ons and abilities over the
    years. Reversable fabric movement. One touch button holer. Variable pitch
    movement. Sideways movement. Multiple thread pathways for multi-color
    embroidery. All of it was perfectly good patent fodder.

    Now if someone were to build a big housing over a particular machine and that
    housing pushed those buttons and levers on that machine, that housing would be a
    new machine. And also perfectly good patent fodder as well. An automated housing
    that extended the machine in a real way would be an actual invention.

    But if someone were to come in to court with a patent that said I have invented
    "sewing multiple buttons and button holes in a row to close a garment"
    or "embroidering a flowers on t-shirts to make people feel happier"
    they would get thrown out (and then beaten in the coffee room) for being
    idiots.

    So with computers "the good bites" of the patent apple were already
    taken up when the people who made the machine and all its parts did that making.
    More patents issued for the network controllers. Memory systems. Modems.
    Keyboards. Touch screens. Batteries. All of it. And each one of these devices
    came into existence with a set of functions. "Communicating with another
    comptuer" is the base definition of "a client which communicates with
    a central server". The extra words "client, server, and central"
    are all immaterial. Computer A talks to Computer B using the built in functions
    of the network controllers, and the built in functions of every device between
    the two computers.

    So when "real programmers" read software patents we -never- see
    invention in the software, we see "oh, he used the network card as
    designed".

    In sort we see "oh this is a button hole made by pushing the button hole
    button".

    It's nothing. The patent issued for that feature already and its toll was paid
    when I bought the ethernet card for my computer.

    All software patents double-dip on the "actual" machine's actual
    patents.

    I call this purpose patenting. The Desqview patent on using the virtual memory
    system of a i386 to intercept the memory writes to a display to allow windowing,
    for example, take the "this cpu can intercept any memory reads and
    writes" and says "I'll do that for the video memory" and calls it
    novel.

    Embroider blue flowers on green stems on a red shirt by loading the right color
    thread and shirt and then pressing the feature button. Same thing.

    People in the patent tolling business mistakenly believe that the computer can
    -see- the whole program. That "loading the program into memory" makes
    it part of the machine. But the CPU, the actual part of the computer that -does-
    anything cannot see past its current operation. It literally doesn't know what
    the next opcode will really bring until it goes and gets it and does it. (With
    pipelining and predictive execution it can guess like up to three operations
    away, but really, that's not a program's worth of knowledge.)

    Network cards say "oh you want me to send this data" and "here
    some data arrived".

    Transcriptase doesn't know about cells, it just follows along and copies DNA.
    ha. ha.

    Now there is a very small set of computer-like devices called FPGAs (and CPLDs
    etc.) where loading in an image more or less "makes a new machine" out
    of the chip. These are the only, and still arguable, cases where hardware plus
    software makes new machine. Even these, however, are constrained by their
    mechanics and have already had all those mechanics submitted to patent-land.

    If the court could be made to understand that software doesn't -ever- create a
    machine, and that the machines have already been patented for all the possible
    functions that they can perform because software can not -ever- make a machine
    do something that it wasn't designed to do, then the software patent
    "controversy" would just go away.

    Back to Swype... The touch screen measures. The database bits hold data. There
    is some fuzzy comparison. The a word is chosen or a subset of words is
    displayed. Choices are made and the string of bytes is sent to the input field.
    Or in other words, hardware did a lot of reading and the computer did a lot of
    comparing and copying. All of that is already patented (and most of it is
    expired patents as well).

    It "just -looks- like a line".

    Swype is a great program. It deserves it's copyright and trademark protection.
    It's not patent material because the -machine- doesn't do anything the machine
    isn't designed to do.

    Swype has a -purpose- and -value- but it doesn't describe a transformation or
    machine.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Monday, June 11 2012 @ 09:15 PM EDT
    Mr Risch says:-
    Some of these solutions are best applied by patent examiners at the time the patent is being sought, but the reality is that we will never have enough money to fully test every patent application. For that reason, I’m a firm believer that for many disputes it is more efficient for the parties that have a stake in the matter to fight about validity at the time of litigation. This increases costs on patent defendants, but if examiners will not get the job done anyway, I would rather they not spend a lot of time trying.

    The simple reality is that often these battles end up as a David vs Goliath battle without the feelgood twist of the little guy being triumphant. All this does is allow the big corporations with an army of lawyers to beat the little guy. If the little guy is the one holding the patent, the bigger one can ignore it knowing that the patent defendant cannot afford to put up a fight. Similarly, if the larger company holds the patent, even if the patent is rubbish they can easily put a small developer out of business.

    Obviousness:My experience with computer scientists is that they believe everything is an obvious improvement over the past. I’ve read enough engineering journal articles to know that this is not true, but many software patents are obvious.

    The next problem is for the software developers. How is the system supposed to work? Must a software developer review every software patent before writing each line of code? What if a patent is lodged covering a work in progress where the 'offending' code is written but the product is not yet completed? The real problem is that more often than not, software patents are being 'violated' because they are obvious, even the ones that Mr Risch may believe are not. How many software patent cases have there been where a software developer has read a patent, thought 'That's a good idea', and implemented it vs those who have simply solved the same problem with the same solution, oblivious to the fact that someone had got a patent on it? Finally, a bit of advice...believe the computer scientists you speak to and those who actually develop software over what you read in articles written by self-proclaimed experts.

    Any system that relys on requiring people to perform impractical tasks and on unbalanced court battles is going to be fundamentally flawed. It is never going to work and will only serve to hamper the innovation it professes to support. Programmers and engineers tend to be practical and logical people. If something is broken, fix it. If it's beyond repair, you throw it out. Software patents are beyond repair. This is why there are not too many fans of software patents here.

    [ Reply to This | # ]

    The SIMPLE cure
    Authored by: Anonymous on Monday, June 11 2012 @ 09:17 PM EDT
    I think that the law does have to change. I have been mentally cycling something
    I call "The Software Implemented Machine Patent Limitations and Exclusions
    (SIMPLE) Act" that I'd submit to congress had I such power.

    It would say essentially the following:

    -- Software is Patentable as "an extension to a machine". (don't yell
    yet. 8-)

    -- Implementations of Software can only violate that patent if that
    implementation is the -only- software running on the machine at the same level
    of abstraction. That is, if the software plus the hardware actually creates a
    "new machine". (this gives us CPU microcode as part of the larger CPU
    patents, and FPGA images as part of an industrial machine etc.)

    -- No Implementation of software that is, or can be, run on a general purpose
    computer can be held as infringing on any software patent.

    -- A general purpose machine is any machine that is running, or can -usefully-
    and/or -meaningfully- run more than one patentable or unpatentable element or
    different patentable or unpatentable elements at different times.

    -- As a mater of law, no software running on, or which can be run on a general
    purpose computer, can be held to infringe any patent.

    -- Indicators of General Purpose Computer include, but are not limited to
    machines exhibiting one or more of the following gross characteristics when they
    exist at the same level of computing as the patented agency:

    --- The presence of more than one family of functionality. (If it makes calls
    and plays games and takes notes etc, then those "and"s indicate a
    general purpose device.)

    --- The presence of an "operating system" element.

    --- The presence of any any discrete "applications" distinct or
    optional with respect to -any- "operating system" or
    "application" element.

    --- The presence of loadable software modules or "optional"
    "components"

    --- The applicability of more than one "patent hierarchy" or "top
    level patent" to the execution domain of the computing device. A
    "patent hierarchy" is the set of patents defined by one patent and all
    the prior art disclosed in that patent, and those patents and their prior art.
    If patent A is disclosed as prior art to patent B and patent C; patents B and C
    represent two separate hierarchies unless C also discloses B as prior art.

    (in other words, if patent B defines a machine, and patent C defines a machine,
    the set of patents define more than one machine if C and B are unaware of each
    other. This bullet item needs work.)

    --- The presence of interchangeable or removable (in normal operation) discrete
    data modules that contain "code" the machine will run, including hard
    drive, flash drives, rom cartridges, etc. [Basically if you don't need to crack
    the case to reprogram the device then the device is intended to be regularly
    reprogrammed and so is "general purpose".]

    -- The complaining party shall bear the burden of proof to demonstrate that and
    allegedly infringing device is single purpose.

    -- It follows from the above that a valid patent for the software portion of a
    patented device does control the disclosed material when it is used in general
    purpose context or device .

    =====

    Basically you should be able to make the simple declaration that your device
    runs an OS (windows, linux, mac os, DOS, JVM, etc) or that it "also plays
    angry birds" or "has a scripting language" and then walk away
    with a JOML.

    But people with complex code-heavy specific industrial processors (as in that
    rubber case etc), and people inventing CPU's with microcode components, should
    be able to protect their (rubber processing) machine.

    [ Reply to This | # ]

    Every process is NOT an algorithm
    Authored by: swmcd on Monday, June 11 2012 @ 11:44 PM EDT
    You write
    every process patent is an algorithm, mathematical or not.
    I don't understand the distinction that you are making here. All algorithms are mathematical.

    If you have some category of processes that you consider to be algorithms but not mathematical, then you need to find a different word for them, or else you won't be properly understood.

    Furthermore, I do not agree that every process is an algorithm.

    If the only effect of carrying out a process is to encode the output of an algorithm in the state of some machine (e.g. a computer) then the process IS an algorithm, and should not be patentable. Examples of patents that were (regrettably) granted for algorithms include

    • CadTrak
    • the BCD conversion patent (eventually overturned)
    • patents on compression algorithms
    • patents on encryption algorithms
    In contrast, consider Diamond vs. Diehr. At the end of that process, you have more than just the result of a computation sitting in a register in an industrial controller. At the end of that process, you actually have a batch of cured rubber. There is an algorithm involved, but the process is more than that.

    If we could get the patent office and the courts to understand and respect this distinction, we could eliminate very many software patents.

    [ Reply to This | # ]

    Suggestion: Information manipulation is abstract
    Authored by: Paul Johnson on Tuesday, June 12 2012 @ 04:13 AM EDT
    One issue that the article raises is that of determining when the subject of a
    patent is "abstract". I want to address that specific point here.

    Having read the Supreme Court cases on this, I think the distinction the court
    is trying to make is between the manipulation of matter or energy (which is not
    abstract) and the manipulation of information (which is).

    Thus a design for a new type of touchscreen that senses finger touches in a
    novel way involves the manipulation of matter to create the screen and the
    manipulation of energy to make it work. Both of these should be patentable. But
    the algorithms that determine where the finger has travelled and what words it
    might be spelling are manipulating information.

    The "pencil and paper" test comes somewhat close to this, in that a
    process that can be followed using a pencil and paper is restricted to
    manipulating information. If you read "pencil and paper" to include
    arbitrarily complex and lengthly processes that would in practice be infeasible
    for a real human to carry out then you basically get the same answer (and the
    person with the pencil and paper will sit next to the person "skilled in
    the art", who is also an abstract fiction with impossible capabilities and
    limitations).

    It might seem possible to save such a patent by describing a physical machine
    (i.e. a configuration of matter), with registers and logic units, that would
    implement the algorithm. And it is true that a patent could be written that
    would cover such a machine because it would be written in terms of the physical
    matter that would compose such a machine. But by the same token it could not be
    read to cover a general purpose computer executing software without also reading
    into the patent the information-manipulation underlying the design of the
    patented machine. If information manipulation is not patentable then such a
    reading is inpermissable. So if, for instance, the physical machine claimed in
    the patent included a buffer register for the finger position that was connected
    to a sensor on one side and other circuitry on the other side, then the
    plaintiff in a patent suit would have to point out an equivalent hardware
    register physically connected in the same way in the alleged infringing device.
    Merely pointing out that the alleged infringer buffers the infoming sensor
    information in some other way would not be sufficient.

    Our hypothetical plaintif might also argue that an FPGA implementation of his
    algorithm infringes, on the grounds that the programmed FPGA is a physical
    implementation created by manipulating matter and energy, and the programmed
    FPGA does indeed contain a register which is connected in the way described in
    the patent. This would be true, but it would also be comparitively easy to work
    around: FPGA implementations are rarely pure hardware: modern FPGAs usually
    contain one or more CPU cores, so that an implementation can be a combination of
    hardware and software. It would then be possible to avoid such a patent by
    shifting a part of the implementation into software. For instance it might be
    feasible to have the CPU read information from the sensor and then pass it to
    dedicated processing hardware on the FPGA.

    ---
    These ideas and others like them can be had for $0.02 each from your friendly
    local idealist.

    [ Reply to This | # ]

    A Different Solution
    Authored by: Chromatix on Tuesday, June 12 2012 @ 04:36 AM EDT
    Mr. Risch,

    Lots of people on this forum have already delighted in telling you that you're wrong, and how and why. I'm not going to bother extending that, but instead to offer an alternative solution. Needless to say, I'm also in the anti- software-patent camp, I just believe in using constructive criticism when possible. Accordingly, I'm going to try to show you how Swype could have turned their invention into a legitimate patent.

    So let's get one thing clear: software is mathematics and therefore unpatentable; the same goes for business methods (such as 1-Click). If the entire substance of the patent requires a computer, or is an algorithm that can (notionally) be carried out by a pencil and paper exercise, the patent is de facto invalid. That's my logical starting point.

    This means that the Swype invention cannot be described in terms of the algorithm that performs the pattern recognition (which would, in any case, not be novel due to prior art in Graffiti). Instead one must go back to what it *is* - a computer *peripheral*. The fact that it is so tightly integrated into a telephone-computer (thank you Iron Sky for that terminology) is irrelevant - the integration is an obvious and logical development beyond the fundamental invention of a computer input peripheral.

    So the patent would have to describe a "device for text input" with distinguishing features such as "no moving parts", "word-at-a-time orientation", "compact size compared to standard keyboards, without impairing usability by a standard-size human", and "flat, smooth surfaces permitting easy cleaning", this last being useful for medical and industrial applications.

    It would have to be written so that it is merely an alternative front-end to a teletypewriter, as used in telegraph service many decades ago. The fact that teletypewriters were almost immediately used as input/output devices for early computers (once they became capable of handling text) would then show that attaching this "new teletypewriter" to a computer is another of those obvious developments of the idea.

    And of course it would have to materially distinguish itself from Graffiti, according to the same criteria as you have helpfully outlined.

    This is not wholly unprecedented. Consider IBM's famous Model M keyboard, built around a patented keyswitch mechanism, known as the "buckling spring". It is still considered the gold standard for keyboard design, to which all other high-quality keyboards aspire. The Cherry MX keyswitch is probably the closest non-IBM equivalent.

    And here's a radical proposal to correct the bad-patents problem: make the USPTO liable for the legal costs of all patent lawsuits which involve invalid patents issued by the USPTO. The huge ratio between the cost of filing a patent and the cost of litigating it out of existence should concentrate minds wonderfully. This liability shall not extend to legal costs incurred after the USPTO itself has invalidated a patent and communicated such a finding to the patent holder.

    [ Reply to This | # ]

    Software Patents -- A Differing Legal View
    Authored by: Anonymous on Tuesday, June 12 2012 @ 04:43 AM EDT
    Bias disclosure: I stopped being agnostic on the topic of software patentability several years ago. Furthermore, there is enough libertarian in me to believe that government- granted monopolies interfere with free market principles and are fraught with potential for favoritism and abuse; to wit our present patent system that overwhelmingly favors the enormously wealthy and entrenched interests.

    The Supreme Court has decided several cases where it might have resolved the issue by holding that software is categorically excluded as non-patentable subject matter. The Court has waffled or ducked on that issue every time, on at least two occasions suggesting that Congress speak to the issue. For that reason, I agree with Prof. Risch that yet another case arguing for a categorical exclusion of software patents on grounds of section 101 non-patentable subject matter would likely be a fruitless quest. The Court has made plain that it doesn't want to go there on section 101 grounds ever since Gottsch alk v. Benson>, 409 U.S. 63 (1972).

    At the same time, the Court has never held that software is patentable, which sparks a suspicion that a majority might be persuaded to categorically exclude software on other grounds if given a compelling argument. The argument hastily assembled from research notes laid out in this post has the advantages of: [i] not being addressed to section 101; [ii] being far easier to understand than section 101 arguments raised thus far; and [iii] cutting to the very quick of objections to software patents, the patent thickets drawn around entrenched products and the dampening effect of software patents on software innovation.

    Caveat, I have not updated my citations in several months.


    Prior Art and Obviousness

    Fortunately, there is a strong legal argument for categorical exclusion of software patents based on prior art and obviousness grounds set forth in a line of cases not yet examined by the Court in the context of software patents.

    The Patent Clause includes a statement of its purpose: "[t]o promote the progress of science and useful arts[.]" Accordingly, the Court held in A. & P. Tea Co. v. Supermarket Corp., 340 U.S. 147, 153-154 (1950): Footnote 1

    The function of a patent is to add to the sum of useful knowledge. Patents cannot be sustained when, on the contrary, their effect is to subtract from former resources freely available to skilled artisans. A patent for a combination which only unites old elements with no change in their respective functions, such as is presented here, obviously withdraws what already is known into the field of its monopoly and diminishes the resources available to skillful men. This patentee has added nothing to the total stock of knowledge, but has merely brought together segments of prior art and claims them in congregation as a monopoly.
    Just so, a software patent also "only unites old elements with no change in their respective functions [and therefore] obviously withdraws what already is known into the field of its monopoly and diminishes the resources available to skillful men." Every software patent issued subtracts from what skilled artisans can create using known methods on their computing devices.

    By analogy, one might have legitimately obtained a patent for the invention of the claw hammer to drive and pull nails. But a subsequent patent granting a monopoly on the use of a claw hammmer with a slightly longer handle and a slightly heavier head to drive and pull longer and thicker nails would almost certainly fail the A. & P. Tea test. Software does no more than the legal equivalent of driving and pulling different sizes of nails, the equivalent of different music being played on a computer from separate CDs. Each melody played from a CD is not a new invention within the meaning of the patent laws even though each recorded song causes a different sequence of bit register states to be created. These are not different "processes" within the meaning of the patent laws. Likewise, different software programs processed by a general purpose computer do not create new processes. The processes are in the computing device, not in the software. The invention lies in the computing device, not in the binary notation, the literary work the device processes.

    The "old elements" combined in software are binary notation indicating desired bit register states, processed by the device as instructions to change its bit register states accordingly. The computing device and its processes are not modified by the execution of software. To the extent that someone might be tempted to argue that binary notation is patentable we need not linger on the section 101 patentable subject matter inquiry; binary notation is unquestionably prior art. Pingal was known to have used binary notation circa 5th-2nd Centuries B.C.E. B. Van Nooten. Binary Numbers in Indian Antiquity, 21(1) J. of Indian Philosophy, p. 31 et seq. (1993). Binary notation is very simply a numerical writing system for recording (and processing) numbers having the base 2, marvelously suited to the requirements of modern computing devices.

    (To take the contrary position, then one must of logical necessity extend patentability to individual digitally recorded audios, videos, and books which likewise are stored and handled by the device as a stream of binary bit states. A bit by any other name is still a bit and regardless of the level of abstraction used to aid the editing of bits (such as programming languages, interpreters, and compilers) all of these marvelous software castles are erected by the computing device from binary notation written by software authors via their programming aids.)

    To be sure, I am aware of the Federal Circuit's In Re Alappat holding that processing a different program on a general purpose computer creates a new machine. But a combination of factors have raised a huge suspicion in my mind that the holding would be swept away if the issue were revisited in today's changed climate in regard to software patents. E.g., the dissent's argument on that issue are compelling and went unanswered by the majority; Chief Judge Rich is gone and the present Federal Circuit is not nearly so software patent-friendly as that court was in 1994; many of the case law underpinnings of the Alapatt majority's opinion were swept away by unreversed portions of the Circuit's decision in Bilski; and I don't believe that the Alappat holding would pass the giggle test if presented to today's Supreme Court.

    Beyond the prior art barrier, software patents also raise the closely related issue of obviousness because they only combine familiar elements (manipulation of binary bit register states) according to known methods, doing no more than yielding predictable results. (When unpredictable results are obtained, the trade generally refers to the defect as a "bug".) In KSR Int'l v. Teleflex, 550 U.S. 398, ___ (2007), Slip Op. at pp. 11-12, the Court reiterated:

    For over a half century, the Court has held that a "patent for a combination which only unites old elements with no change in their respective functions . . . obviously withdraws what is already known into the field of its monopoly and diminishes the resources available to skillful men." Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., 340 U. S. 147, 152 (1950). This is a principal reason for declining to allow patents for what is obvious. The combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results.
    (Italics added).

    The combination of bit register states according to known methods to yield predictable results is all that the processing of software can accomplish. Otherwise, the software could not be processed. There may be significant pre- or post-data processing activity that is patentable but software is of necessity prior art and obvious within the meaning of the patent laws.


    Supporting Property, Contract, and Takings Law

    The immutable fact that software patents "subtract from former resources freely available to skilled artisans" also raises issues involving property law, contract law, and the Fifth Amendment's Takings Clause that have not yet been examined by the Court in the software patent context. A hypothetical set of facts (actually played out countless times) will simplify illustration of these issues.

    Lantern Corp — like Oracle, IBM, and Apple — manufactures and sells computing devices. Lantern also licenses software and frequently obtains software patents whose claims read on the use of the devices they sell and have already sold. Lantern sells a computer to A. Software Author. Later, Lantern applies for and is awarded a patent that restricts what software works Author may develop and license to members of the public.
    On that set of facts, the sale transferred clear title to the chattel computing device from Lantern to Author. Lantern retained no rights to restrict what Author can do with the computer. He owns it. It is his.

    But Lantern's new patent restricts what Author can do with his computer, casting a legal cloud over his title to the machine. The patent also breaches the contract to transfer good title and also raises the issue of Lantern's breach of the implied covenant of good faith and fair dealing, given Lantern's long history of obtaining patents that restrict what can be done with its already-sold computers, a cause of action that in at least most jurisdictions allows claims for general, special, and punitive damages.

    As between Lantern and Author, this set of facts has the potential to raise the question of whose property right is superior. Author's ownership and right to the enjoyment of his computer to the full extent of its capabilities is a fundamental Natural Law property and liberty right. And the "father" of the U.S. Patent system and one of our nation's Founders, Thomas Jefferson, has an eloquent explanation of the reasons there can be, in his view, no Natural Law right to ownership of a patent. Graham v. John Deere Co., 383 U.S. 1, 9-10 n. 2 (1966), citing Letter to Oliver Evans (Jan. 1814), VI Writings of Thomas Jefferson, at 298 (Washington ed.).

    I will not dwell on this point long because bedtime approaches. But history teaches that during the period of the Constitution's drafting and for decades thereafter, a patent was regarded as a revocable "privilege," not a property right, the antecedent of today's definition of a "license" in the Administrative Procedures Act, 5 U.S.C. § 551(8). As best I can tell so far, the line of cases describing the patent right as a property right trace to seminal dicta unsupported by any authority in Consolidat ed Fruit Jar Company v. Wright, 94 U.S. 92 (1876).

    Moving on from Author v. Lantern Corp, the same set of facts raises the issue of whether the USPTO's grant of the patent monopoly to Lantern amounted to a taking of Author's property and/or liberty interests without the due process and just compensation required by the Fifth Amendment.

    As between Author and the Feds, the fact that software patents subtract from what can be done with the prior art of the computing device rather than advancing the progress of the useful arts should weigh heavily here.

    The computer is Author's chattel property and upon purchase he had a reasonable investment-backed expectation that the government would not restrict his use of it, particularly by the grant of an invalid patent. At the moment of purchase, he acquired a bundle of property rights protected by the Fifth Amendment, such as “the right to possess, use and dispose of it.” R uckelshaus v. Monsanto Co., 467 U.S. 986, ___ (1984) (alleged taking of trade secrets) (italics added), quoting United States v. General Motors Corp., 323 U. S. 373, 323 U. S. 377-378 (1945) Bd. of Regents v. Roth, 408 U.S. 564, 561-572 (1972) (“the property interests protected by procedural due process extend well beyond actual ownership of real estate, chattels, or money”).

    Likewise in regard to the liberty interests:

    Although the Court has not assumed to define "liberty" with any great precision, that term is not confined to mere freedom from bodily restraint. Liberty under law extends to the full range of conduct which the individual is free to pursue, and it cannot be restricted except for a proper governmental objective.


    Bollin g v. Sharpe, 347 US 497, 499-500 (1954) (reaffirmned in later cases). There is no “proper governmental objective” in issuing patents whose claims "subtract from former resources freely available to skilled artisans" and grant would-be monopolists the exclusive use of those resources. A. & P. Tea, supra.

    Yet another of the property rights (and liberty interests) Lantern sold to Author was the right to practice with his computer “any existent knowledge from the public domain [along with] any materials already available", John Deere, supra at 6, and to do so without concern that the USPTO would grant any “patent for a combination which only unites old elements with no change in their respective functions.” KSR Int’l, supra; Bolling, supra. Cf., Golan, supra (finding no protectable First Amendment right to publish works removed from the public domain by amendment of the copyright law; branding the John Deere Court's relevant discussion as dicta).


    In my opinion, laws that are so vague and/or ambiguous that their reach is unpredictable are bugs, not features. Moreover, legal slippery slopes should be avoided whenever a bright line rule that is understandable to all would serve the purpose. The software industry thrived when software patents were rare and the only shield for warding off copycats was copyright. I am aware of no persuasive reason the industry would suffer in any way should all software patents be piped to /dev/null.

    I will spare the reader a rousing conclusion, because family life beckons. However, I do wish to express my disagreement with Prof. Risch's position that "for many [patent] disputes it is more efficient for the parties that have a stake in the matter to fight about validity at the time of litigation." In my view, that position cedes the relevant legal ground to the exclusive use of the fabulously wealthy, as does the vagueness and ambiguity of the patent laws and their jurisprudence.



    Notes

    [1] A. &. P Tea was followed in, inter alia, Graham v. John Deere Co., 383 U.S. 1, 5-6 (1966) ("Congress may not authorize the issuance of patents whose effects are to remove existent knowledge from the public domain, or to restrict free access to materials already available"), citing A. & P. Tea; Bonito Boats v. Thunder Craft Boats , 489 U.S. 141, 146 (1989) ("Congress may not create patent monopolies of unlimited duration, nor may it 'authorize the issuance of patents whose effects are to remove existent knowledge from the public domain, or to restrict free access to materials already available[]'"), quoting John Deere. Golan v. Holder, ___ U.S. ___, No. 10-545 (18 January 2012), Slip Op. at 19 (copyright case; distinguishing John Deere) ("In Graham, we stated that 'Congress may not authorize the issuance of patents whose effects are to remove existent knowledge from the public domain, or to restrict free access to materials already available.' But as we explained in Eldred, this passage did not speak to the constitutional limits on Congress’ copyright and patent authority. Rather, it 'addressed an invention’s very eligibility for patent protection.'” (Italics added; citation omitted.)

    — Paul E. "Marbux" Merrell
    Retired lawyer

    [ Reply to This | # ]

    Wrong metric
    Authored by: Anonymous on Tuesday, June 12 2012 @ 04:56 AM EDT
    Basically you are defending software patents because there are cases where they benefit the inventor.

    That's the wrong metric. The correct metric for a law is that it has to benefit society.

    The benefit of patents law for society arise when they enable progress that would otherwise not happen. Namely when inventions happen and are publicly disclosed that would otherwise not reach the public.

    This is the case for the small independent non-manufacturing inventor, not for a large manufacturing company that already gains financially from making inventions, by being able to use them in products.

    The small independent non-manufacturing inventor does not have the means to entertain a patent portfolio protecting him from countersuits. As a result, patent laws only provide him very limited bargaining power. Instead, he is better off working with non-disclosure until sealing a deal.

    Since in the area where they could make an actual essential difference, software patents are doing more harm than good, there is no point in the effort for maintaining the system.

    There also is the problem that patents are supposed to offset public and private interests by granting a limited-time monopoly in exchange for fully applicable disclosure.

    The lifetime of software patents is 20 years. There are only very few protected elements of interest to the public after such a long running time (things like Patricia trees, some forms of B*-trees, sorting heaps, and other stuff). Those few that are actually of interest after the whole protection duration are not tied to particular products and technology, but rather define the state of computer science.

    Placing them under patent protection hampers education and research.

    You defend software patents as being a good idea in principle, for some people. But for those for which they really could and/or should be an enabler, they turn out to be a rather badly mixed blessing.

    If one wants to help society more than damage it, the general "patent" umbrella is not workable. One would need to create something quite different, and the hubris of "I can create something working better for the general good than free market" is rather seldom supportable by solid results.

    [ Reply to This | # ]

    • bad argument - Authored by: Anonymous on Tuesday, June 12 2012 @ 02:30 PM EDT
    What is Missing
    Authored by: wouterdb on Tuesday, June 12 2012 @ 06:20 AM EDT
    A few short comments, that I think have not been addressed:
  • If you are a programmer, show us the code. There are many levels of skill, and it would help the debate if we could judge your skill level.
  • The quality of documentation in software patents is far below standard. It is not written to document, but to obscure. The example patent you present is a nice example. Bad typesetting, no standard UML notation, only a sequence diagram of the program, no executable code and subtle omission of all constants. These constants are what sets the system apart from mathematics, they represent the engineering challenge. I would recommend reading 'Software Architecture in Practice' by the SEI
  • You do not mention the time range in which a patent should be valid. Computer science moves very rapidly. Five years are already a very long time. A patent taking longer than this will almost certainly be meaningless. Taking into account the slow rate at which the justice system works, can there be meaningful enforcement.
  • Would a protection schema such as the Semiconductor Chip Protection Act be more suited for software. It protects the exact program, but allows reverse engineering and explicitly leaves function unprotected. It is granted automatically, like copyright. This mode op protection allows further improvement on the concept. Anyone with the right the technical skill can re- implement the same idea. If he does a better job at it, society benefits. This is how software works.
  • [ Reply to This | # ]

    Patents: An idea whose time has expired
    Authored by: Anonymous on Tuesday, June 12 2012 @ 08:01 AM EDT
    There was a time when patents were believed to provide social utility. It was a time during the infancy of industrialization, a time during which research and production were difficult and industrialists needed all the help they could get. It was a time long before computers became universal and made devising complex methods easy.

    Those conditions no longer apply today.

    Nowadays, if one party doesn't "invent" something after dipping in to the international sea of ideas, then another will, because the sea is out there, and it is deep, and there are millions of parties contributing to it freely and continuously.

    Under such conditions, creating limited monopolies on inventiveness is no longer helpful to society, indeed it is downright harmful. Find a person who today claims that patents are helpful (except defensively) and I'll show you a person who benefits from disallowing other people to come up with similar ideas by themselves.

    The many clever people here on Groklaw have pointed out numerous reasons why software patents lie somewhere between untenable and absurd, quite apart from being a disaster in practice. But I think the matter goes far beyond that. Today's world is one in which innovation is very strongly assisted by computers, and as a result, the entire process has altered beyond recognition since the days when patents seemed like a helpful idea.

    Today patents are no longer helpful. They are simply a means to gain additional profit at the expense of society and progress. Their time should be brought to a close.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: JonCB on Tuesday, June 12 2012 @ 11:04 AM EDT
    Don't know if you're still reading mrisch but, assuming so,
    i would like to thank you for bringing your point of view. I
    disagree with your initial assumption (Sorry, anyone who
    states that software is not math is wrong as a simple matter
    of fact. Whether you believe that or not changes nothing)
    but your conclusions are interesting.

    If i read you right, one of your suggestions is that a
    patent application should clearly state the intended inputs,
    "transformations" and expected outputs of the claimed item.
    I'm actually kind of good with this. If patents actually
    clearly stated their inputs and outputs maybe they'd be
    worth more than the paper they probably got printed on 3 or
    4 times.

    I do think that there needs to be a better way to defend
    yourself against patents. One that doesn't cost several
    orders of magnitude more than the patent application itself.
    Do you think there's any potential in creating a "This
    method of doing things does not infringe on any active
    patent" certificate that costs effectively the same amount
    as the initial patent application and has the same
    assumption of truth as a patent(i.e burden of proof of
    invalidity is on plaintiff to supply overwhelming evidence,
    can apply to USPTO to get re-examined)?

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Tuesday, June 12 2012 @ 01:06 PM EDT
    Some ills of software patents:

    1. Obtaining Patents cost a lot of money. They cost even
    more paying the lawyers to write the application than they
    cost to actually apply (and perhaps more than creating the
    invention itself.)

    2. It is ridiculous to provide a twenty year protection in
    an industry where rate of obsolescence is few months
    (especially at a time when the indutry is gradually moving
    to cloud based service model and where innovations can come
    up overnight.) Patents are a hindrance to this natural
    growth of the industry and may result in unreasonably
    prolonging the life of a product.

    3. Software is different from other engineering and
    mechanical inventions. The latter are generally the kind
    that can revolutionize a given mechanical process. Software
    is generally evolutionary in nature. Its utility does not
    depend as much on the newness of a specific technique as it
    does on the unique combination of known algorithms and
    methods. Such methods of innovation should not be protected.

    [ Reply to This | # ]

    Software is Data
    Authored by: feldegast on Tuesday, June 12 2012 @ 01:15 PM EDT
    Software on a computer is nothing more than data processed
    by the CPU, if we patent data then we will soon start
    patenting movie/book plots and other insanities, after all a
    book plot generates mental images, voices and environments
    in the brain doesn't it?
    How is a book different from a computer program? sure it is
    not as well defined mathamatically but the book is processed
    by the brain (computer) remembered, and commented about by
    the reader. A computer is (if i recall) a device that
    accepts inputs, produces outputs and stores
    information....sounds like a brain to me.
    so if we patent algorithms, we might as well go all the way
    and patent thoughts!

    If you don't want thoughts patented, then please put your
    foot down and stop patents, remove the insanities of
    patenting mathematics which they are not supposed to do
    anyway and programmers might actually be able to get on with
    their job of creating rather than litigating.

    ---
    IANAL
    My posts are ©2004-2012 and released under the Creative Commons License
    Attribution-Noncommercial 2.0
    P.J. has permission for commercial use.

    [ Reply to This | # ]

    An Analogy For A Computing System
    Authored by: Anonymous on Tuesday, June 12 2012 @ 01:25 PM EDT
    There's a system, still in use today, called Relay Ladder
    Logic (RLL). Back in the pre-computer days, banks of
    electrical relay panels controlled a machines operation.
    They took inputs from limit switches, push buttons, etc. and
    produced output to control motors, solenoids, etc. Timers
    were available, and 'analog' was accomplished by banks of
    relays, converting values to groups of relays. Logical
    operations like AND, OR, NOT, XOR, etc were available.
    Later, gate arrays were configured to process bit data from
    inputs, by following a 'program'; a list of instructions.
    The was the Programmable Logic Controller (PLC).

    Relay panels are an exact analog to digital computers, and
    can, in theory, do everything a computer can do; they are
    basically digital processors, and massively parallel at
    that.

    There is no "software" in a relay panel, unless you called
    the whole panel configuration 'software'. They are purpose-
    built. PLCs, however, are easy to reconfigure ('program').
    Are we then to assume that a PLC 'program' is patentable?
    It does something new and unique, you say? Take input from a
    new kind of sensor? Sends output to a new kind of device?
    Those things can't make the 'program' patentable. Wait, you
    say, it does something unique in its processing. Isn't the
    purpose of a general purpose machine to be able to do
    ANYTHING within its capabilities? It's illogical to assume
    you can patent ONE of those possibilities. Yet that's what's
    being done right now. If you support this sort of
    'reasoning', then general purpose machines are useless; they
    are being turned into purpose-built machines, by virtue of a
    patent. Find a way to make the machine do something OUTSIDE
    its capabilities. THAT would be patentable.

    BTW, novel and unique mechanical devices are still being
    patented. The present system handles these quite well.
    Novel and unique processes are also being patented. Swiping
    a touch screen, regardless of how unique the software is, is
    not a process like curing rubber. It's like putting a
    chemical into rubber. You can't patent the act of putting a
    chemical into rubber. Processes take certain specified
    input products, and output certain products, otherwise, they
    don't work. Software systems can take many inputs, and may
    produce many outputs, one, or none. You can't patent
    actions. Software does not fit into the present patent
    system, mechanical, electrical, or process (chemical,etc).

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: mipmip on Tuesday, June 12 2012 @ 01:38 PM EDT
    Thank you for offering your view here on Groklaw. I see problems to your
    solution on the practical side:

    * The interesting thing is, as you note, many of the rules that would invalidate
    patents are already there, just not rigorously applied. You short-circuit here
    and propose doing just that. But there is a reason this isn't done already. The
    patent system is inherently pro-patent, it is geared to allow as many patents as
    possible, and give as much power to them as possible. Any rigidness the supreme
    court may want has been and will be eroded by patent lawyers sitting in patent
    courts having an interest in a "powerful" patent system. Look at the
    past, I'm not a patent historian but the bits I have heard seem all to fit an
    avalanche of small pro-patent change over the years from lower courts.

    * Software, being a production with practically no entry barrier, is done by
    everyone now, not just a circle of people posting in specialized curriculums.
    Enumerating knowledge is a tall order even in the latter case, but impossible
    when millions of developers do the same as a hobby or for profit. Internet
    search only helps when you can put the invention into a few snappy words, code
    itself is not really searchable, not even if there were only one computer
    language in use. This makes prior art in software a toothless weapon.

    * You find invalidating patents afterwards a workable solution. Again there is a
    disparity between all the small developers writing software not having the funds
    to defend against invalid patent claims and big companies holding patents and
    having the means to force their view of the patent on those developers. Look at
    Microsoft and their game of intimidation with undisclosed patents. It is even
    working against big companies, how easy it would be against single developers
    and small bussiness.

    It is like getting laws accepted after a perfunctory check and only if someone
    with money doesn't like it, is the law put before congress.

    * The patent system has lost (if it ever had at all) its original purpose of
    advancing the field by publicizing inventions. The reason the system is here now
    is to put a prize tag on inventions, to make them a commodity that can be sold
    or invested in. A valid reason. At least for bankers ;-). You said somewhere you
    believe in the original purpose of the system, but even you have to concede that
    the reality is different.

    * Finally let me say your plan has at least a chance of getting partly
    implemented. So while being sceptic I can at least hope for a small improvement.
    Good Luck.

    [ Reply to This | # ]

    Software patents are a chicken egg problem.
    Authored by: Anonymous on Tuesday, June 12 2012 @ 01:52 PM EDT
    No, not that chicken and egg problem. But this chicken and
    egg problem:

    People take chicken eggs and eat them or perhaps use them
    in ways that chickens may not approve of. The chickens do
    all the work to create the egg and, as far as we know,
    never really complain about the egg being taken or how it
    is used.

    Now consider what it would be like if chickens did complain
    very loudly about their eggs being taken and used for
    someone else's benefit in a way the chickens did not approve
    of. Would you listen to them? Suppose the chickens said
    "Stop taking our our eggs and using them to suppress
    others! Even if you pay us, we do not want you taking our
    eggs for that!" Well, with chickens being stupid and all
    that, they probably would get ignored. Of course if they
    did talk, they might get patented, however.

    Well here you are. Consider that we as programmers are the
    chickens and the software we create from our ideas and
    effort are the eggs. If the actual producers of the
    software are saying "Stop taking our software and patenting
    it!" should people just continue to ignore them? This is
    not the case of those doing the work wanting their stuff
    patented. This is the exact opposite.

    The pepole who are benefiting from the software patents
    are NOT the people creating the software. In fact the
    people are creating the software are saying "STOP PATENTING
    OUR SOFTWARE!!!" What has to happen for that to sink in?

    Do lawyers and politicians truly think programmers and authors of software are
    so stupid they can be ignored?

    Perhaps the answer is for software authors to create a
    programmers union and go on strike until software patents are abolished.

    [ Reply to This | # ]

    Difficulty in determining...
    Authored by: Marc Mengel on Tuesday, June 12 2012 @ 03:48 PM EDT
    You wrote:
    "I don't think the dividing line can be software, because so many solutions use electrical signals as part of the product or process. It is easy to rail against software, but it is far more difficult to determine what is and is not software."
    Actually, it is exceedingly easy to tell what is software. Electrical signals are not software, neccesarily. Software is a number -- A series of bits -- recorded in some kind of media.

    Note that by this definition, the music on a Compact Disk is software, as is the contents of a DVD of _The_Last_Starfighter_. As is the pattern of holes in a Jaquard Loom card chain.

    But I think that's correct.

    [ Reply to This | # ]

    Problem of obviousness and novelty
    Authored by: Anonymous on Tuesday, June 12 2012 @ 05:14 PM EDT
    Simple solution:

    Prior to publishing the details of the patent, publish the description of the
    invention (eg for Swype: "convert a finger sliding over a flat keyboard [ie
    a keyboard represented on a touch screen] into the words that would be typed by
    the finger having pressed the keys slid over with error correction, with a
    method to offer selection of words that may all match a given "slide"
    eg sliding between l-o-s-t should be recognised as "lost") and allow
    submissions to be made as to a solution before the patent is published (say 6
    months min).

    If any of the submissions is close enough to warrant infringement of the patent
    if granted, then the patent would be declared obvious or non-novel and not be
    granted (and the submitter given a [financial] reward). All submissions would
    be treated as trade secrets held by the USPTO which could be used as prior art
    for future patent examination; the submitter would get a "reference
    receipt" should they decide to patent their submission - a patent can't
    prior art itself!

    [ Reply to This | # ]

    Donald Knuth: Mathematical Ideas, or Algorithms, Should Not Be Patented
    Authored by: Anonymous on Tuesday, June 12 2012 @ 06:09 PM EDT
    An oldie-but-goodie.

    Donald Knuth: Mathematical Ideas, or Algorithms, Should Not Be Patented

    This Groklaw article is from June 2009, covering Donald Knuth's letter to the European Patent Office back when Europe was considering whether they should allow or disallow software patents.

    Any readers who haven't seen it before should definitely give it a read.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Tuesday, June 12 2012 @ 06:38 PM EDT
    I'm different than a lot of the people here. It seems to me that, if I can take
    a chunk of boolean logic (math), and implement that in discrete hardware or in
    an FPGA, and validly patent that, then I should be able to patent the exact same
    boolean logic carried out in a general-purpose CPU.

    And don't bother to say that the FPGA will produce its output as voltages
    leaving the chip, whereas the CPU will only change bits in memory. I could add
    a write to a physical output device (say a DAC), and now it's a voltage leaving
    the CPU + DAC, and so that argument is answered. But that doesn't really change
    the patentability question - it's just a trivial postprocessing step, which
    legally does not alter patentability. Therefore, the patentability of pure
    software cannot rest on "voltages leaving the chip". (For that
    matter, even writes to external memory are voltages leaving the chip.)

    If you're going to be consistent, if you claim that "software is not
    patentable because it's math", then you have to also state that all boolean
    logic that is implemented in digital hardware is not patentable, because that's
    also just math. The two are equivalent.

    The real problem with software patents are not that they should be unpatentable
    because it's math, but that they should be unpatentable because the current
    patent regime is unconstitutional, on two grounds. First, software industry
    practice says "don't read the patents because of treble liability".
    But that means that software patents don't actually teach anything to the
    industry.

    The second area of unconstitutionality is that software patents, even if they
    were read, would still not teach, because they are not written in language that
    is comprehensible to one "ordinarily skilled in the art". "The
    art" in this case is computer programming, not law.

    MSS2

    [ Reply to This | # ]

    I would like a try at a short bit of logic.
    Authored by: dacii on Tuesday, June 12 2012 @ 07:03 PM EDT
    I haven't read (there are too many) all the comments. So forgive me if I repeat
    someone else's comments. I believe that software is no more than a description
    of a process. It is a set of instructions for a (micro)processor to carry out.
    The processor knows nothing other than its instruction set. The instructions
    are encoded to that instruction set. Ergo the processor is reading(running the
    program) based on it's language. Without a processor, nothing happens with
    software. So in general software is nothing more than a description of a
    process for the processor to read and act accordingly. I have seen that you may
    be able to patent a process. But can you patent a description of a process?
    English is a language. If you read a patent then are you not doing what the
    processor does when executing the software? So by extrapolating, just by
    reading the patent, you could be violating the patent. Or worse yet by
    translating it into another language, maybe your native tongue Spanish, and
    typing it up for others to read to avoid, you yourself have violated the patent.
    Instruction sequences should not be patentable. Just my personal opinion.

    [ Reply to This | # ]

    Prior art: KolaWare?
    Authored by: Anonymous on Tuesday, June 12 2012 @ 09:34 PM EDT
    The KolaPad appeared on the market around 1984. It was a mass market digitizer board used by many computer systems, but then digitizer boards were not that new at the time. These KolaPads had keyboard overlays which allowed the user to cursor over the keyboard template to select the keys to be entered into the computer. What is the difference between Swipe having a keypad displayed by a newer backlit display and an older digitizer with a plastic template? The plastic template of course. The Swipe system doesn't add any value except for not having to exchange templates, as they are now generated in software. Swipe didn't invent the backlit display did they? Generating keyboards in software is not new either. So what exactly is novel here? I don't see the added value of Swipe, as it is nothing more than taking old ideas an applying them to newer hardware. Its the evolution of older ideas, not some new amazing insight deserving to the ability to kick everyone else out of the open market. This patent fails to show me anything new, just the reapplied and recombined technology from 1984.

    [ Reply to This | # ]

    Do you believe IDEAS should be patentable?
    Authored by: Wol on Tuesday, June 12 2012 @ 10:21 PM EDT
    Ignore the fact that they are currently, explicitly, excluded subject matter.

    And if you think they should be patented, according to Groklaw you are beyond
    redemption until you repent ... :-)

    Let's get Gedanken to do an experiment for us. She'll bring me two computers,
    completely identical apart from the fact that one of them implements a software
    patent, and the other one doesn't. Now, I *daren't* switch the patented computer
    on, because then it will implement the patent and all sorts of bad things will
    happen to me ...

    All of you who think patents on software are a good idea, please tell me how I
    can tell which computer is safe to switch on! If it's possible, I can do it, I
    can do any test you care to think up to decide which one is safe - apart from
    switching it on to try it and see if it works.

    I'll just kill off the obvious solution right now - running a scan over the hard
    disk to see if the patented software is on it. All that scan will give me is a
    magnetic topography, and unfortunately I have no way of knowing which topography
    implements the patent and which does not. All I have is a load of north and
    south poles, and no way of knowing what they mean. The patent doesn't discuss
    things in terms of poles.

    Can anybody else come up with any way whatsoever of telling me which machine is
    safe to switch on?

    Because if the machines AREN'T physically identical we have a possibly - and
    reasonably so - valid hardware patent. But if they are physically identical we
    have a patent lawyer's wet dream - two identical machines, one of which is
    infringing and the other isn't, with no way of distinguishing them. Which means
    anybody can be sued for any patent violation with no way of being able to prove
    innocence.

    I'll leave you with the problem of telling which computer is illegal, it's time
    for me to go to bed. It's 03:20 BST here ...

    Cheers,
    Wol

    [ Reply to This | # ]

    Is software a process?
    Authored by: Anonymous on Wednesday, June 13 2012 @ 06:13 AM EDT
    The claim that software is a process may not be accurate
    because software is fundamentally a series of machine
    understandable code which is processed by the processing power
    that comes with the hardware of the computer. The hardware is
    patentable, therefore what is the need for patenting software
    (which is but a series of binary code and is well covered by
    Copyright laws.)

    [ Reply to This | # ]

    I don't think that makes too much sense
    Authored by: Anonymous on Wednesday, June 13 2012 @ 08:30 AM EDT
    Patents were made to foster innovation. So far I see they only foster
    reinventing the wheel every time.

    Your example with the typing on a touch screen. Yeah, the idea was great. But
    when I started typing on a touchscreen, without any investigation, the first
    thing I thought was that word corrections should be based on keys that were
    close to the keys I've hit. They found a very good math model to achieve that it
    seems. Btw if this is the same soft keyboard I am thinking about, it has to
    improve for my native language.

    Anyways, do you think they invested in coming to this idea much? I would think
    such idea would pop-up in some developer's head. Make the app and patent it to
    take advantage over competition. But I wouldn't call that patent to be in public
    interest. The idea that every idea you have is owned by you and should give you
    enormous return without being competitive against others is broken.

    Perhaps you are thinking from what is/seems fair in a particular situation. I
    don't argue, maybe that particular patent gave the guys a deserved advantage, I
    don't know. But that's not my point. My point is that patents bring in much more
    disadvantages than advantages. And these are disadvantages mostly to the small
    guys that new ideas come from but big bad guys benefit from them (because small
    guys cannot enforce) and prevent the small guys compete (because small guys
    can't defense). I also don't argue that patents are really, really, really
    always bad. Perhaps there is a very small minority of good patents but that
    can't be an excuse to cause so big harm to the public.
    If digging a hole somewhere is beneficial to the public you don't make a law
    that digging a hole everywhere is allowed.

    And the rational is pretty simple. We accept that a few times some inventor that
    deserves his patent will not benefit too much of it (although I question if
    patents have ever helped here). But for this price the market works so much
    smoother that the inventor just because of that is much better than he/she would
    have been enforcing his/her patent rights.

    Sorry, no time to write it better organized.

    P.S. wrt "big bad guy" meand corporations which are by definition evil
    by working for their interests ignoring public interests. There's a lot in the
    Internet about that.
    Regards.

    [ Reply to This | # ]

    Regarding Skill in the Art
    Authored by: Anonymous on Wednesday, June 13 2012 @ 01:54 PM EDT
    Sometimes things are only obvious after the fact. Here is Henry Briggs to John
    Napier:

    "My Lord, I have undertaken this long journey purposely to see your person,
    and to know by what engine of wit or ingenuity you came first to think of this
    most excellent help unto Astronomy, namely logarithms.

    But, My Lord, being by you found out (i.e. discovered), I wonder nobody else
    found it out before, when, now known, it is so easy"

    A full version is related by: Garry J. Tee, Department of Mathematics,
    University of Auckland, New Zealand at:

    http://web.cs.dal.ca/~jborwein/briggs.htm

    [ Reply to This | # ]

    Programming is just math? Software patent examples
    Authored by: Anonymous on Wednesday, June 13 2012 @ 04:37 PM EDT
    I only got through calculus, but I do not recall an "If" in math.
    Most of programming is just math, but the "If" allows decisions to be
    made. I think this is a profound difference between math and programming.

    Before I get in too much trouble, I want to say that I think most if not almost
    all software patents should be declared invalid. Almost all are obvious. An
    example of super obvious is US patent 4,197,590. This is the infamous XOR
    computer graphics patent.

    But how about US patent 5,454,069. This patent takes 3D images, such as from a
    CT scanner and makes them suitable for use in a rapid prototyping 3D printing
    device. It does this by finding parts that would float off and adding supports
    to hold them in place without user intervention. The patent causes a physical
    change in the output of the rapid prototyping 3D printer, but it is just
    software. There are other methods for finding floating objects and adding
    supports. So does that mean that only one patent should have been granted for
    the automatic support generation? I think not, because that would be patenting
    an idea. Do you think that there should not have been any patent, because it is
    just math? I say the program is making decisions and that puts it outside of
    pure math.

    How about US patent 7,986,859 (I am no long anonymous). This patent is for
    taking the raw (Bayer Filter) image from a digital camera and converting it to a
    displayable RGB image, with a minimal amount of moire effect. The program to do
    this could be written without "If" statements. The method to do this
    did not seem obvious, because it had not been done before and documented. The
    field has a lot of literature and patents, so it seems that prior art would have
    been easy to find. I was on the fence as to if this would receive a patient.
    So the software provided a benefit, was not obvious and there was no prior art.
    So should it have been patented?

    It is obvious that I think there is a place for software patents, but they
    should be very limited. I know others will disagree. Those that do disagree,
    how would you protect innovative software? If you say copyrights, please tell
    me how copyrights protect something like the 5,454,069 patent. The
    "C" code was included with the patent. So just rewriting it in
    another language maybe enough to get around copyright.

    OK, end of my unpopular rant.

    [ Reply to This | # ]

    In re Alappat is an Abstraction inversion
    Authored by: Anonymous on Thursday, June 14 2012 @ 11:42 AM EDT
    I was rereading some of PolR's posts and I have just realized that the error in In re Alappat is an abstraction inversion!

    An abstraction inversion, is the error in computer science, of thinking of a low-level implementation detail, in a high level context. When abstraction inversion is part of a design, it fixes the low-level details into a high-level design. This makes it impossible to change the low-level implementation details without introducing bugs. This is why abstraction inversion is viewed as a error (or anti- pattern,) in computer science.

    Although the phrase "abstraction inversion" is relatively new, the underling concept is old. I remember 30 years ago when I was taking my first programming 101 course, I was told not to concern myself about how bits were physically represented on the machine, because bits can be represented in many ways and electrical engineers were constantly changing them. We wanted out programs to continue to work whenever the EEs invented a new way to implement bits. We were computer programmers not EEs.

    The error in Alappat is precisely looking at programs in a low-level implementation dependent way. Just as abstraction inversions result in bugs in computer programs, this abstraction inversion results in a faulty legal opinion.

    The error is also suspicious. I don't believe it is a sincere error. If it were a sincere error, the courts would make the same error when attempting to determine if a given idea is the same idea as expressed in a given patent, and they don't. If the abstraction inversion were applied consistently, every time EEs changed how bits are implemented in a computer, then every software patent in existence would become re-patentable as a new idea! As a new pattern of electrical activity! The courts do not make this error in practice. The error is used only to cover up the fact that abstract ideas are being patented. Which means it is not an error at all, but a deliberate misinterpretation.

    begin digression

    Which leads me to a problem with Google's search engine! We I realized that the error in Alappat was an abstraction inversion, I naturally wanted to know if some else had had the same idea. No matter how I try to tell Google that I want to search for the phrase "abstraction inversion" as a phrase, with the words together, Google insists on breaking the phrase into pieces, finding pages containing the words "abstraction" and "inversion" separately! Yes, I have tried "advanced search" with the "this exact word or phrase" field! Does anyone know if anyone else has characterized Alappat as an abstraction inversion before?

    end digression

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: rebentisch on Friday, June 15 2012 @ 07:33 PM EDT
    "My experience with computer scientists is that they believe everything is an obvious improvement over the past. I’ve read enough engineering journal articles to know that this is not true, but many software patents are obvious."
    Obviousness in patent law dogmatics is a last resort general clause style filter. But the obstacle is that "obviousness" cannot be falsified because it could always be argued that it was not obvious back then. Obviousness to the "actual" person skilled in the art means something different than obviousness from a patent perspective. It is not empirically founded but a dogmatic device, and the person skilled in the art is a legal fiction, it is not you and me.

    Why are software patents generally "obvious" aka "trivial" aka "bad"? The reason is that the object awarded legal protection ("software invention") is not "scarce", so an incentive system (=the patent system) fails to deliver adequate results.

    [ Reply to This | # ]

    Curing the Problem of Software Patents, by Michael Risch
    Authored by: Anonymous on Saturday, June 16 2012 @ 12:25 PM EDT
    I think the main problem with the patent system isn't the ideas behind but that
    some people have started to abuse it. And those who didn't are being driven to
    join the brawl for no other reason than protect themselves.

    You quoted "Software is hardware in every sense that matters for
    patents." Let's have a deeper look at that. Software is a formalized way to
    write recipes (as in cooking). Software development is a translation process (as
    in language interpreter). We take ideas from customers and turn them into
    detailed instructions to be executed by a moron. A human could do it but since
    it would be boring, we use a computer. Note that software isn't mathematics -
    it's language (as in English). It's very limited (because the computer is so
    very dumb) but we're more authors than mathematicians.

    Now, my questions are:

    1. Can you patent the work of an online interpreter?
    2. How about the words that come out of her/his mouth?
    3. What about the words that she/he hears and translates?

    My answers:

    #1: No. The process of translating something is a craft. Crafts as such are not
    patentable.

    #2: Since the interpreter doesn't add anything of value to the input (they
    should stay true to what is given them), the output can't be more or less than
    the input. It's the input in another form but by itself, it doesn't make the
    input more or less patentable.

    #3 This leads to the question: Can you patent a recipe?

    And the answer here is, sadly, yes. Which leads to the question: Can you patent
    any recipe?

    No: The recipe has to be, well, "outstanding" in some way.
    Unfortunately, there is no universal standard what "outstanding"
    means. When I teach an newbie, they will think my methods are outstanding but
    they are just by relative comparison of what they know and what I know.

    Since "outstanding" doesn't seem to be a good tool to make a decision,
    what else do we have?

    Damage. One of the roots of patent law is the idea to bring more justice to the
    world. It was invented for the specific purpose to protect the work of
    "helpless" inventors so they wouldn't be ripped off all the time.

    This makes sense if you, say, develop a new medicine. As of today, you need to
    spend around one billion dollars for a new medicine. Without patent law, it
    simply wouldn't make sense for corporations to take the involved risks, so it
    makes sense to apply it here.

    But software patents are in a completely different league. They cost $20'000 or
    less to produce and they can cause hundreds of billions of damage when they are
    used as a weapon - which is the sole and main purpose of software patents.

    Since software patents are used solely to damage society as a whole (forcing
    companies to invest in them, costs of defending yourself against a lawsuit no
    matter if it's justified or not, costs of canceling abusive patents, costs in
    lost revenue when you can't sell your product because a competitor wants more
    market share), they need to be outlawed.

    If you fail in this task, then no new computer related products will be sold in
    the USA by ... well ... 2013 because everyone will be suing everyone and all
    money will flow into courts and to patent lawyers. No more software development
    will happen because it's just too expensive/dangerous. Maybe someone will find a
    patent to "display text remotely" and take down Google or the whole
    Internet (at least the part that runs in the USA).

    The lawsuits between Apple and Samsung should be a warning shot. These only
    exist to give one company a bigger market share in a saturated market. Better
    product doesn't count nor how "inventive" it is. It's just the
    logical, inevitable conclusion what happens when all players follow the rules
    that we made.

    Or Oracle vs. Google. Oracle came in demanding several billion dollars and got
    nothing. But if they had a different judge, the outcome could have been
    completely different. Is that what we want?

    This seems to damage companies like Swype. But does it really?

    First of all, someone could copy their idea. But customers would only buy the
    copy when it was better for most of them. Do we really want to protect something
    that most people think worse?

    Copying an idea usually leads to a new, very similar but still different idea.
    The "thief" adds his own wisdom to it. We call that process
    "learning". Isn't that something that we should support?

    If Swype can't patent their idea, they can still make a product and sell it. If
    they fail, some money will be lost. But if they succeed, no other company can
    sue them for billions of dollars for no other reason than to slow them down.
    Isn't that better than the other way around?

    We say that the Internet Year is only three months. New ideas spring up so
    quickly by now that the old, sluggish patent system can never keep up. Do we
    want to slow down progress (might be a good idea) *and* feed the trolls? How
    much is an idea worth protecting if it's outdated in 9 months?

    How much more money will we make/safe if software patents are outlawed and all
    this money goes back into R&D?

    Aaron Digulla

    [ Reply to This | # ]

    Why are physical patents good, but software patents are bad?
    Authored by: Anonymous on Monday, June 18 2012 @ 05:20 AM EDT
    When Eli patents his cotton gin, how many gins can he produce by himself? Not
    very many. He need to buy supplies and hire people to reproduce his invention.
    As a result of this invention lots of people benefit.

    When Steve patents touching your finger to a piece of glass, how many times can
    he reproduce that software? Virtually infinite at essentially no cost. As a
    result of this invention, Steve gets 100% of the revenue and society gets
    nothing.

    Patents on software are completely inappropriate. They are not good for society.

    [ Reply to This | # ]

    Groklaw © Copyright 2003-2013 Pamela Jones.
    All trademarks and copyrights on this page are owned by their respective owners.
    Comments are owned by the individual posters.

    PJ's articles are licensed under a Creative Commons License. ( Details )