A possible game changer for invalidating bad software patents | Opensource.com

A possible game changer for invalidating bad software patents

Posted 03 Feb 2011 by 

Rob Tiller (Red Hat)
Rating: 
(12 votes)
Image by : 

opensource.com

submit to reddit

A case now before the Supreme Court could lower the threat level created by bad patents in a meaningful way. The issue before the Court is whether to change the burden of proof on patent invalidity from clear-and-convincing evidence to the preponderance standard. This change would be good for free and open source software.

Today Red Hat joined in an amicus brief in support of this change in Microsoft Corp. v. i4i Limited Partnership. [PDF] On the amicus brief, Red Hat is part of a diverse group that includes Google, Verizon, Consumber Electronics Association, Comcast, Dell, Hewlett-Packard, HTC, Intuit, L-3 Communications, LinkedIn, Lockheed Martin, Mastercard, The New York Times, Rackspace, Shutterfly, Software & Information Industry Association, Time Warner, Wal-Mart, and Zynga.

When companies with such different models, philosophies, and experiences can get together, it suggests that a lot of people recognize that there's a serious problem that must be addressed.

The amicus brief faces up to some of the serious problems in the current U.S. patent system and explains them in a helpful way. At present, a party challenging a patent as invalid must show by clear and convincing evidence that the patent should never have been granted. This is the heaviest evidentiary burden applied in civil cases, and it applies only in a few legal situations. Most of the time, the applicable burden of proof is a preponderance of the evidence, meaning that a fact is found to be more likely than not. Applying the clear-and-convincing standard suggests a background assumption that there has been careful consideration of the merits of a patent application before a patent is granted.

In fact, though, this assumption is unfounded. A patent application is assigned to be reviewed by a single patent examiner. All examiners have enormous workloads. They are required to complete the processing of each application (that is, understand the application, search for prior art, communicate with the applicant, and write up conclusions) in a matter of just eighteen hours or so. The time constraints make it almost impossible to discover all applicable prior art, particularly in the software area, where much of the prior art is embodied in code for which there is no practical search method.

Because of crushing workloads and structural factors, it isn't surprising that a lot of software patents have been issued by the Patent Office that should have been rejected. But the situation is made even worse by the requirement that the patent issue unless the examiner is able to present a prima facie case of unpatentability. In other words, there's something close to a presumption that a patent application should be granted.

Once a patent is inappropriately granted, it is possible, in theory, for a party accused of infringing it to show that it is invalid. In practice, this is quite difficult. When software patents are at issue, the technical issues are often complicated and difficult for a lay jury to understand. Jurors frequently mistakenly assume that the patent examination process was careful and exhaustive, and so have a tendency to assume that a patent must be valid. On top of all this potential confusion, jurors are instructed under current rules that they may only invalidate a patent if they find the evidence for invalidity clear and convincing. Even when there's strong evidence that a patent should never have been granted, it's difficult for lay juries to conclude that the technical issues are clear.

The clear-and-convincing standard protects bad patents. As the amicus brief says, bad patents "do not foster innovation. Instead, they block and impair others from innovating. . . . If the holder of an invalid patent secures – or even threatens to secure – an injunction or royalties, it can block or tax innovation, driving up prices and decreasing the availability of innovative products and services to the public."

If the Supreme Court decides that the preponderance standard, rather than the clear-and-convincing standard, should apply, it will be much more likely that juries will invalidate bad patents. Knowing that, patent aggressors may be less aggressive. The threat to innovation generally and FOSS development in particular should be reduced. Compared to where we are now, this would be a real improvement.

submit to reddit

25 Comments

simonphipps
Open Source Evangelist

It's great to see Red Hat joining in to challenge the basis for a patent troll (and amazing to see it done to support Microsoft), but I am concerned to see you framing this as an assault on "bad patents", as if improving patent quality will help developers somehow. The locus of the problem is not "bad patents" but rather the bad system of law surrounding the legacy understanding of patents.

The fact is that anyone can gain a patent on an idea that any software developer anywhere might spontaneously devise, and then use that patent as the anchor for what is effectively legally-sanctioned blackmail. Improving the technical quality of those anchors for extortion does little to give FOSS developers safety from attack.

So while it's a great thing to challenge the trolls, please can Red Hat use a different "frame" to talk about it? One that doesn't imply there are "good" software patents?

Vote up!
8
Vote down!
0
foo

RH are probably taking this in a "one step at a time" style fashion. Getting making it easy to get rid of most patents is helpful. Maybe the next step is to significantly reduce the length of patent validity to perhaps 1-2 years for software, then get rid of them.

Vote up!
2
Vote down!
0
rtiller
Open Source Evangelist

Simon, I'm certain we could find a lot of things to agree on regarding the shortcomings of the U.S. patent system in treating software. But I would not agree that the best way to work to improve the legal system is an all-out, full-frontal assault. There's a chance to gain some meaningful improvement in the i4i case, which is a good thing, even though it will not resolve ever problem.

Vote up!
6
Vote down!
0
simonphipps
Open Source Evangelist

I know we would, Rob - indeed, I have long been a proponent of both pragmatic engagement (forcing key exceptions to enforceability for example) and incremental change. All I am suggesting is that we heed Lakoff and ensure that when we advocate for change we don't inadvertently strengthen the frame with which copyright and patent maximalists make their case appear inevitable. Hence I'd not speak of "improving patent quality" but rather of "challenging patent abuse", for example.

I'm convinced that in this sort of political struggle the words matter even more than we think - hence my comment.

Vote up!
4
Vote down!
0
rtiller
Open Source Evangelist

Fair enough. I'm with you that words matter, and appreciate the struggle to find the most effective ones.

Vote up!
2
Vote down!
0
stites
Open Source Champion

Simon Phipps said: "Hence I'd not speak of "improving patent quality" but rather of "challenging patent abuse", for example.

I'm convinced that in this sort of political struggle the words matter even more than we think - hence my comment."

The distinction is very important in the problem of finding prior art. IBM and the U.S. Patent Office would like us to form expert committees to review all software patent applications to apply our knowledge of software history to find prior art. This would "improve patent quality". However the amount of work involved makes this impractical just as it makes it impractical for the Patent Office to "improve patent quality" by their ordinary work methods.

The few software patent attacks that we have to face in court makes "challenging patent abuse" practical by searching for prior art. We can do that for two or three cases a year and have a very potent and cheap weapon to counter software patent attacks. It also makes the lawyers nervous to mount a software patent attack on Open Source because they have no way of predicting how effective our prior art defence might be.

------------------------------
Steve Stites

Vote up!
4
Vote down!
0
asmiller-ke6seh

It's not bad law ... it's bad lawyers which create the problem. Remember when there was an attempt to patent the "look and feel" of an application interface?

At least we now know that one cannot patent the expression of a "law of nature", such as a mathematical truism - that, at least, was a step in the correct direction.

We need more motion in the direction of favoring true innovation, and not squashing it.

Vote up!
2
Vote down!
0
Unidentified

The law has power, not lawyers. Lawyers just make a narrative around it in hope that a judge interprets it the same way. When the law lets lawyers argue dangerous and absurd things, it's the law that needs to be reformed.

Vote up!
8
Vote down!
0
NH

The FTC back in 2003, published a report:
http://www.ftc.gov/opa/2003/10/cpreport.shtm

"FTC Issues Report on How to Promote Innovation Through Balancing Competition with Patent Law and Policy"

Clearly part of the recommendations, is preponderance of the evidence. This report is good reading, and in retrospect, seems to be as relevant today as it was then.

Vote up!
2
Vote down!
0
stites
Open Source Champion

From the article: "The time constraints make it almost impossible to discover all applicable prior art, particularly in the software area, where much of the prior art is embodied in code for which there is no practical search method."

Open source has a huge advantage in software patent fights by possessing a very superior method of finding prior art. Just throw the question to the community at large asking if anyone has seen any prior art for the software patent in question during their career. Within a few days we can gather a mountain of prior art that no law firm could ever discover no matter how much time and money they spent.

--------------------------------
Steve Stites

Vote up!
2
Vote down!
0
Paul Johnson

In my blog I show that any method for processing information can be reduced to a formula in the lambda calculus. Since the lambda calculus is part of mathematics, this means that any method for processing information can be reduced to a mathematical formula, and hence according to Bilski it is not patentable subject matter.

So if we reduced this particular patent to a formula and got a professor of mathematics to provide expert testimony that this was a mathematical formula, couldn't we get it invalidated that way?

Vote up!
2
Vote down!
0
Paul Johnson
Vote up!
2
Vote down!
0
Unidentified

How about this: When a patent is invalidated, the patent office has to refund the money it charged to the patent holder, PLUS pay for the litigation expense of BOTH sides.

That way, there's an actual disincentive for perceived rubber-stamping. It also appeals to a sense of fairness: The patent office isn't SUPPOSED to grant bogus patents.

Vote up!
0
Vote down!
0
simonphipps
Open Source Evangelist

If it's pragmatic reforms you're looking for, take a look at the article I wrote about that a while back.

Vote up!
0
Vote down!
0
JJS

I learned a lot about patents when I responded to one of those ads for inventors to get their inventions patented. I came up with a throw-away idea (swing set anchors--I made some out of rebar for my daughter's swing set) and went through the process. (End result: Too much risk/money for a nebulous reward.)

For a physical device, you have to include a schematic, drawn to scale, with the patent application. For software patents, there is no equivalent, which right away gives software patent holders broader rights than other patent holders. The software equivalent of a schematic is a working prototype. (And make the patent applicant supply the equipment to run the prototype.)

This also gives courts something tangible to work with. The comparison between a product charged with infringing and the prototype is much more understandable than the algorithmic gobbledygook we see currently.

While in a perfect world, there would be no software patents, now that the genie is out of the bottle, we need to at least get rid of bad patents.

Later . . . Jim

Vote up!
4
Vote down!
0
stites
Open Source Champion

Jim said: "While in a perfect world, there would be no software patents, now that the genie is out of the bottle, we need to at least get rid of bad patents."

That implies that there exists such a thing as a good patent. What is a good patent?

---------------------------
Steve Stites

Vote up!
0
Vote down!
0
NH

Good question "That implies that there exists such a thing as a good patent. What is a good patent?"

One approach to an answer, is understanding the patent system doesn't recognize developments in software outside the Patent Trademark Office. Often prior art as these developments, though exist, aren't given serious consideration, not without a push anyway.

Patents are being designed to not only protect the main idea, but to include other technology areas in support of the main idea. The language of claims are taking more ground then the rights of a patent for the main idea, would include

Giving one idea of "bad" patents, would be that claims language is too complex for a first read understanding. That technology areas not directly part of the main idea, but only in support, are included into the main idea within the claims language.

Giving one idea of a "good" patent, are those that claim language on first read gives a clear understanding of the main idea, and other technology areas in support, are clear as well not written in the claim language.

I'm sure what I offer here fails in the law department, I hope that I'm making my general idea clear though.

Vote up!
0
Vote down!
0
Rufus Polson

That would be a "well written" patent. Doesn't make it good.
I can imagine two basic measures. One is, a patent is good if it is on something which is genuinely new and genuinely would be difficult for other people to independently invent, thus there is an actual non-obvious innovation being patented. That would be a technical measure of a "good" patent. This is rare nowadays, especially in software, but it is possible.
But the purpose of patents supposedly is to spur innovation and technological progress, ultimately benefiting the economy as a whole. So in a broader political sense, a "good" patent would be one whose granting spurred innovation and progress rather than, or at least more than, retarding it. I would argue that in our modern social/economic environment there is no such thing as a "good" patent in this sense, certainly not in the software realm and probably not at all. Maybe there was back when they started granting patents, but things have changed.

Vote up!
2
Vote down!
0
stites
Open Source Champion

Rufus Polson said: "That would be a "well written" patent. Doesn't make it good."

To a lawyer a good patent is one that cannot be successfully disputed in court. From a lawyer's viewpoint how well a patent is written is everything, nothing else matters.

I think that to everybody else the rest of Rufus Polson's comment is a much more meaningful explanation of a good patent. I also agree that a software patent rarely (actually never to my thinking) meets his definition of a good patent.

-------------------------
Steve Stites

Vote up!
2
Vote down!
0
NH

@Polson

You are right about being well written, that point begs the question, of how does one tell the difference if not well written when given, a well written one doesn't mean a good patent.
You've mentioned that "a "good" patent would be one whose granting spurred innovation and progress", and we're going to have to agree to disagree on that.

Earlier in the Patent Office history, the seeking of prior art didn't go far beyond the inside walls, and many patents have been granted with little or no "real world" references to the vast prior art that exist.

Claimed innovation and progress is at what cost, I believe is a very good question. Written well is a good patent to gain a clear understanding of just what is being claimed and all if any innovation and progress will result. Our ideas of a "good" patent are nothing more then different approaches.

NH

Vote up!
0
Vote down!
0
Cam
Community Member

As an outsider (a software engineer who is rarely involved in patent issues), my impression is that there are many people who make a living from patents as they stand. Because of this they are resistant to rule changes that would affect their business.

It will continue to be very hard to overcome the opposition unless there is a clear benefit to society from the change. Making it easier to invalidate a patent is a clear example - it devalues all patents from a gilt-edge commodity to a somewhat more dubious one.

I think we do need changes to the system, and there may be a role for patents. But we need to focus on how we can add value to society, to businesses and to the academic world rather than supporting a bureaucratic and legal system that is a parasite on innovation.

Vote up!
0
Vote down!
0
NH

Agreed.

The problem with patents is the claim language, as I gave in very simple terms and probably not the best, ideas of both "good" and "bad" patents, that in a Court of law, even in the best of terms, not going to end software patents.

The focus must target the claims of patents, where much of the litigation is created from, as earlier I mentioned, that claim language is taking far too much ground. Language that too often includes technology areas that otherwise would only support what is to be patented. Not part of it.

A Court would not differentiate between "good" and "bad" software patents, simply because those terms are subjective, have no independent well defined meaning beyond a given observer. However, a Court would differentiate between two software patent types, if shown that a common pattern of claims beyond the main idea, run through one of them as a group of enough patents that support the pattern.

Having shown the pattern as doing harm, through but not limited to litigation, having no value beyond taking such action in the market the patented, are claimed to add value. I believe things are on their way in such a direction today, and those in support of patents continue their efforts will have the SCOTUS do away with software patents for sure.

Vote up!
0
Vote down!
0
AC

Eliminate software patents. There are nothing more than make-work for lawyers, and an impediment to progress.

Vote up!
2
Vote down!
0
Unidentified

a good patent is one that is expected to survive a challenge in court

Vote up!
0
Vote down!
-4
Unidentified

All other patents are unconstitutional IMO IANAL.

Good patents are needed to protect inventions that take a huge amount of effort and investment to create, and virtually no effort or expense at all to copy. Without patent protection, there is no incentive to develop the invention.

Software never falls into this category. In my experience having written a lot of code, It takes nearly as much effort to develop an independent copy as to create the original. Thus copyright by itself is adequate protection.

The proof of this is that the software industry was more vibrant and innovative before software patents. Software patents have IMO obviously impeded Progress of Science and useful Arts and are thus unconstitutional.

Vote up!
2
Vote down!
0

Rob Tiller is vice president and assistant general counsel for Red Hat, where he manages patent, trademark, and copyright matters. He is a frequent speaker and writer on open source legal issues. Before coming to Red Hat, he was a partner with the law firm of Helms, Mulliss & Wicker, PLLC, where he specialized in commercial and IP litigation. He is a graduate of the University of Virginia School of Law, and a former clerk for Justice Antonin Scalia of the U.S. Supreme Court, and Judge