« Open Source Fig Leaf | Main | Smug RIM »

November 16, 2005

Brain Compression

Something's amiss. Daniel Ravicher of the Public Patent Foundation (PUBPAT) today took a re-examination shotgun and pointed it Forgent Networks JPEG patent cash cow: 4,698,672, claiming invalidity in light of other, similar patents generated by Compression Labs, specifically 5,451,012, and inequitable conduct for not citing those other patents. Two possibilities: Danny's off base, or attorneys representing more than 80 companies & law firms didn't do their homework.

With regard to the value of '672, in just a very few years, Forgent has scooped $100 million in license fees on a patent due to expire next year; hence Pubpat's Chicken Little routine in their brief to the patent office claiming public harm.

Though Pubpat's claim chart is tantalizing, it seems more than a wee unlikely that the same ground hasn't been already been covered by the 40 companies having already taken a license, or the 40 or so currently facing patent infringement litigation. What's more, Forgent at the beginning of the month wooed Susman Godfrey to represent them as a supplement to Godwin Gruber for '672. Did Godwin and Susman not perform what seems obvious due diligence in a prior art search?

The proof that this is something more than a publicity stunt by Pubpat is whether the patent office bites on the reexam.

Peter Zura on this.

Posted by Patent Hawk at November 16, 2005 10:46 AM | Patents In Business


"Two possibilities: Danny's off base, or attorneys representing more than 80 companies & law firms didn't do their homework."

Let me offer a couple of better possibilities:

- attorneys representing more than 80 companies & law firms did their homework, but didn't understand the technical details of compression algorithms and the likely similarity between these claims.

- litigation to contest '672 would likely have run more than just rolling over and paying the licensing fees.

It will be interesting to see how the USPTO decides on this. But unfortunately, a patent office that routinely grants patents even when prior art is widely available and even when usefulness can't be determined is not a reliable 'proof' of anything.

Posted by: joey5 at November 16, 2005 11:30 AM

It’s exceedingly unlikely, and frankly insulting, to suggest a failure of understanding. In this case, the claims are not obscure. At worst, attorneys hire technical consultants. Hmm… being a patent technical consultant, strike that preamble “at worst.” ;-)

Interesting point about license fees vis-à-vis standing up to litigation, but at an average of around $2.5 million per licensee (40 licensees, over $100 million in royalties), that too seems a stretch.

Posted by: Patent Hawk at November 16, 2005 10:47 PM

Peter Zura has a much saner analysis of this whole thing. I'm glad you linked to him.

So what's with the very pro-IP-maximalism stance here at prospector, and defense of some of these overly-broad patents? Do you really believe that '672 (and others like it) are good for a) society, b) innovation, or c) societal progress? The evidence certainly seems to say the opposite - that these patents impede progress, limit the dissemination of inventions, and generally benefit only the pocketbook of patent holding companies.

Posted by: joey5 at November 17, 2005 10:30 AM

There is a six-part series on Patent Economics in the Patent Prospector, beginning on April 26. The 3rd installment on The Inventor’s Dilemma addresses most directly the pro-patent stance.

Much of my work is prior art search towards patent invalidation of asserted patents in litigation, so am quite familiar with patents that perhaps shouldn’t have been granted. My anecdotal experience, however extensive, does not negate the general argument that patents are, on the whole, beneficial to society, to the point of being necessary for technological advancement. Patent protection is granted in most every country in the world, so there is worldwide consensus.

Patents are valuable, and can put companies’ fortunes at risk. Sentiment towards a patent infringer, or against a patent asserter, can emotionally skew one’s view, but rationally, what goes on with patents is simply a market mechanism for valuing invention, where the courts are involved when the parties fail to agree (which is where courts are always involved). Because of the sums commonly involved, there tends to be a lot of disagreement. Patents are so prominent only because many cases individually involve large sums - a high-stakes game that draws attention.

Posted by: Patent Hawk at November 17, 2005 11:17 AM

The fact that patent protection is granted in a large number of countries says nothing of the validity of the practice. Most countries also engage in the practice of warfare -- does this mean that bloodshed is beneficial?

The truth is that current stakeholders wield incredible power, and they use that power to gain more power. I don't think you see this because you work too closely with those stakeholders.

In the 'Inventor's Dilemma' you claim that the only ones who complain about software and business method patents are the non-creative types. This demonstrates how out-of-touch you are with our most important present day innovators. The truth is, independent inventors find themselves up against a wall. It is virtually impossible to write a piece of software, for example, without violating a minefield of dubious patents and receiving several cease-and-desist letters from 'inventors' who have *never* written a piece of software themselves, but rather have written a lot of patent applications.

This I know from firsthand experience, and I am not alone. The system is incredibly broken. It rewards the wrong thing (dreaming up ideas and submitting paperwork to the patent office) instead of the right thing: actual invention.

You are right in quoting the famous "invention is 1% inspiration and 99% perspiration," but you ignore the meaning. 1% of the work is coming up with the idea in the first place. The other 99% is getting it to actually work. Ideas are literally a dime-a-dozen. Getting an idea to work in implementation is where the real work occurs. Any real inventor will agree with that.

Getting a patent requires 1% + attorney fees.

Posted by: joey5 at November 21, 2005 8:17 AM

Another one of those who think they 'invent' every day, but in actuality all they do is write some trivial code. The fact that there are too many junk patents in software field does not preclude the notion that there are also many valid patents too.

Stop bitching about software patents !

There is no way a software writer like you can possibly come up with something like RSA algorithm, for example (not to mention Karmarkar algorithm).

There is absolutely no need for those who came up with those brilliant ideas to actually code - there are plenty of cheap coders like youself, some of them are even good.

Posted by: small guy at November 25, 2005 8:57 AM

The discoverers of the RSA algorithm were /all/ computer scientists, mathematicians, and coders. The Karmakar algorithm is a method for solving linear **programming** problems, an art that is almost entirely innaccessible to non-coders.

This isn't about a "cheap coder" like myself, if that is what you are convinced I am. This is about science and progress.

Rivest, Shamir, Adleman, Karmarker -- /all/ brilliant. Are you saying there is no way any of them could have come up with what they did, because they are programmers instead of patent attorneys?

Rivest is still at MIT, teaching *computer science*. His latest invention, Peppercoin, is likely doomed to fail because he has it locked up behind patents. The patent scheme he used with RSA encryption will not work today like it did 10 or 20 years ago -- the industry is too smart; when there are alternate technologies that are not patent encumbered, those that are encumbered lose out, every time. And human cleverness dictates that we will always devise alternatives. Unfortunately, we have to waste a lot of time, effort, and money fighting the system to get those alternatives out there when, if software patents were invalidated, we could just forge ahead without all the lawyer overhead.

BTW, is 'small guy' a psuedonym for 'patent hawk'?

Posted by: joey5 at November 28, 2005 10:00 AM

By the way, 'small guy' is not Patent Hawk. Silly conjecture.

Posted by: Patent Hawk at November 28, 2005 10:23 AM

All I want to say is that the ONLY way for
a small guy like myself to be paid anything at all
for creating some really non-trivial algorithm is through the patent system.
I know it from my personal experience.
And you keep defending those corporate thieves who used to steal each and every idea from small guys. Is that what you want ?

Posted by: small guy at November 28, 2005 1:54 PM

small guy said: "And you keep defending those corporate thieves who used to steal each and every idea from small guys. Is that what you want?"

I'm not defending 'corporate thieves.' I'm attacking idea monopolists. If that describes you, then yes, I'm attacking.

If your idea is so trivial that a rival (whether it be a big corporation or whoever) can simply imitate it with minimum effort, guess what? -- It isn't that valuable or non-trivial of an idea in the first place, and doesn't deserve patent protection. This is *less* true of physically manufactured inventions, but is *absolutely true* of algorithmic discoveries.

If the idea is valuable and non-trivial, trade secret (+ contract law/NDA) is sufficient to protect it. If you wanted to use the idea yourself to build a product, you could do so without fearing that someone else will figure it out -- because it is non-trivial after all. If you preferred not to build a product, you could shop your idea around to companies that would -- as long as it is non-trivial, you should be able to describe its power without revealing how it is implemented. And as long as it is valuable, you can play the interested parties off one another to arrive at a price that is market-based, rather than monopoly-based.

Now, once it gets to market, perhaps it can be reverse-engineered after some period of time. That period of time is your *NATURAL* advantage, your *NATURAL* monopoly on the idea. Its length is determined by the nature of its non-triviality, not by arbitrarily-set 20-year periods granted by the whims of government bureaucrats.

Your statements seem to assume that ideas are inherently without value. I claim that [non-trivial] ideas are inherently valuable, and that free markets are more effective at determining their value than are monopolies.

Why are you afraid that your ideas can't thrive in a patent-free system? I'm confident they can, as long as they are non-trivial (patents aren't supposed to protect trivial, obvious ideas anyway).

Posted by: joey5 at November 29, 2005 7:58 AM

Hey, joey

Have you ever reverse-engineered any compiled code ? You know, writing flow-chart of an algorithm from compiled executable ?
It is rather time consuming when the code base is large, but pretty trivial with small pieces of code.

Some of the most brilliant algorithms, like RSA algorithM can be implemented in just a few lines of code. I remember somebody put 2 or 3 lines of Perl code implementing RSA algorithm on a T-shirt, in attempt to demostrate the absurdity of export restrictions imposed by US government.

You arguments are just ramblings of a computer(and not onlY computer)-illiterate person...
No futher duscussion...

Posted by: small guy at November 30, 2005 12:39 PM

Reverse-engineering a compiled piece of code is not trivial. Heck, it isn't even easy to grok someone else's source code half the time. Some even propose that open-sourcing your code gives you an even greater advantage over your competitors because they will be tempted to use your code as the basis for one of their products. And that gives your engineers, who architected the thing in the first place and understand its every detail, an advantage that no competitor will likely mirror unless they go through their own architecting process, i.e., write their own from scratch. But that is neither here nor there.

You are a poor judge of character and skill. I attended one of the top 3 Engineering schools in the country and received a Master's degree in Computer Science several years ago. I am also published by one of technologists' most esteemed publishing houses. None of those credentials make me a smart person, but they give you a little background in where I'm coming from. But I guess what would really impress you is if I had several patents under my belt. I don't. I've had several opportunities to go that route over the years, but have evaded them with skill. I'm sorry you feel that you have to rely on insults instead of debating these important issues on their merits. How old did you say you were?

Now, as to RSA. The real question about RSA encryption is not whether it was genius. It was. But it is by no means so genius that if Rivest, Shamir and Adelman had all been in a plane crash before getting together to come up with it, that we would be today without it. Indeed, assymetric encryption was a natural development that would have occurred no matter who the players were, and using large primes to achieve this was a natural path to explore. You like to keep holding it up as the smartest thing that anyone ever 'invented,' when the truth is it was 'discovered.' And discoveries are always repeatable. They are eventualities. Maybe it would have taken another year. Maybe another 3 or 5 years. But assymetric encryption was a logical next step, and it's discovery was unstoppable. Such is science and progress.

BTW, most software licenses already prohibit reverse-engineering. Remember what I said about trade secret + contract law being sufficient protection in these cases? Case in point. If someone exposes or 'steals' your trade secret through reverse engineering, you simply extract damages through the normal litigation processes.

If, on the other hand, they can show that they came upon the same discovery through independent means, well, good for them. And good for all of us. It wasn't so non-trivial after all, eh?

Posted by: joey5 at November 30, 2005 4:12 PM