September 15, 2003

Intellectual Property Run Amok!

Yet another instance of intellectual property "protections" run amok. This time Microsoft and all the rest of us are the victims:

Silicon Valley - Dan Gillmor's eJournal - Saving the Browser: The Eolas patent lawsuit (Infoworld) against Microsoft, which has all the earmarks of yet another bad patent being wrongly enforced, is one of the biggest threats to the Net in years. If Microsoft and other browser makers have to radically revise their code, the Internet community will lose big.

Ray Ozzie, CEO of Groove Networks, has done some work. And he says there's fairly clear "prior art" that should invalidate the Eolas patent. Here's Ray's analysis. Persuasive stuff.

The saddest thing is the University of California's joining Eolas in the lawsuit. The whole point of a university is to distribute knowledge, not to restrict and charge for it.

Posted by DeLong at September 15, 2003 07:59 AM | TrackBack

Comments

"The whole point of a university is to distribute knowledge, not to restrict and charge for it."

I think there's a typo in your post. You typed "is" but surely you meant "was"?

Posted by: Tom Slee on September 15, 2003 08:31 AM

sandwor.. I mean, software patents. ya hate 'em, right?

there's not much to say on the subject that hasn't been said more eloquently by some of the net's more-motivated antipatent types. suffice it to say, I'm one of the many who thinks patenting math -- to which abstract, non-machine specific software patents boil down -- is a Really Bad Idea.

Posted by: wcw on September 15, 2003 08:36 AM

I'm aware of two main instances of prior art (although I'm sure there are more). Ozzie's Notes 3.0 is one, and the Viola browser is the other. Both appear to FULLY encompass the claims made in the EOLAS patent prior to its filing date.

I have never understand why, in this rather astonishing culture of technology, the patent office has never been able to open patents up to public commentary, on the internet. After granting, each and every patent should be granted in a PROVISIONAL state, and public commentary should be collected. This commentary can be used to fuel additional examination of the patent. A single examination is an extraordinarily brittle system, given that there is an entire legal industry devoted to the pointless obfuscation that is the average patent filing.

Posted by: Ross Judson on September 15, 2003 08:53 AM

"I'm one of the many who thinks patenting math -- to which abstract, non-machine specific software patents boil down -- is a Really Bad Idea."

I don't see why it's any worse than allowing patents of physics or chemistry. If someone makes a jet engine with a novel feature, it's still a mathematical concept implemented in a physical structure. If someone wants to design, for example, a compression algorithm (for sound or text), it's not any more abstract because it's implemented as a set of instructions for a general purpose computer instead of as a machine which implements only that algorithm. I don't see how the creativity or detail of mathematical concepts involved differ from the jet engine case either. Neither would come about without original thinking and research, although in either case you could claim the math existed all along.

The real issue with software patents is that they continue to be issued for things which would be so obvious to undergraduate students that they were independently "invented" 100,000 times within a few years. The 1-click ordering nonsense stands out. And the patent which covered retrieving anything from a multimedia database really stands out.(http://www.base.com/software-patents/articles/compton.html ) Business practices in common use were being implemented in software and were awarded patents to the first person with the audacity to believe they qualified for patents.

Posted by: snsterling on September 15, 2003 09:44 AM

sure, patents that don't pass the obviousness smell test are a problem -- but that problem is not new with software patents. it's explicitly codified in law; if the patent would have been obvious to a "person of ordinary skill" then it cannot be patented. yes, the USPTO has been breaking that particular rule more than usual for software, but all we need is objective and consistent enforcement of existing law.

the problem with math/algorithm/software/basic physics patents is different. let's say I can get patents on the Zermelo-Frankel axioms and, as a result, have certain rights to all derivative works. guess whom I can hit up for royalties now? everyone.

jet engine patents of the old style worked because they were specific; they did not preclude someone coming up with a different implementation of an action/reaction engine. software patents that are sufficiently specific would be fine, too; the problem is that the USPTO is issuing such general software patents that they are, effectively, allowing people to patent the action-reaction engine.

the argument about where to draw the line is involved and I've already ranted way too much. suffice it to say that patenting math, physics, chemistry or software without great specificity as to the implementation remains a Really Bad Idea.

Posted by: wcw on September 15, 2003 10:48 AM

"The whole point of a university is to distribute knowledge, not to restrict and charge for it." Yeah, right? So that's why subscriptions to scientific journals are so damned cheap!?

Posted by: Mats on September 15, 2003 01:03 PM

There's also the the problem that the idea has been used in non-browser technology for many years. It's obvious to begin with (if you're an educated software professional), and it's been done before.

Posted by: Randolph Fritz on September 15, 2003 01:05 PM

Slashdot had an article about demonstrating prior art for this Eolas patent with an old release of Lotus 123 (I think that was it).

Our patent office now is a joke -- the patent clerks are paid to offer patents. Not only do they have no incenctive to turn away patents, they have every incentive to grant them. They get a bigger paycheck if they do! And they are the gate-keepers. It is a clear cut, by-design case of conflict of interest. Patent law is the new in-thing for aspiring law students -- I know two people who are all dewey-eyed over getting a BS so they can go onto law school and become a patent lawyer -- because that is where the money is at.

It all boils down to one simple question: are modern patents fostering innovation? If the answer is no, they must be reformed. And I certainly think the answer is no. Not only are they not fostering innovation, they are stifling it.

Posted by: Timothy Klein on September 15, 2003 01:11 PM

Slashdot had an article about demonstrating prior art for this Eolas patent with an old release of Lotus 123 (I think that was it).

Our patent office now is a joke -- the patent clerks are paid to offer patents. Not only do they have no incenctive to turn away patents, they have every incentive to grant them. They get a bigger paycheck if they do! And they are the gate-keepers. It is a clear cut, by-design case of conflict of interest. Patent law is the new in-thing for aspiring law students -- I know two people who are all dewey-eyed over getting a BS so they can go onto law school and become a patent lawyer -- because that is where the money is at.

It all boils down to one simple question: are modern patents fostering innovation? If the answer is no, they must be reformed. And I certainly think the answer is no. Not only are they not fostering innovation, they are stifling it.

Posted by: Timothy Klein on September 15, 2003 01:16 PM

A 1998 patent on applets?

You have got to be f*cking kidding me!

Is there a mechanism to challenge a patent on the ground that there was prior art all over the fricking place, that anyone with a computer could have found in about 10 minutes?

What the Hell is wrong with people?

Posted by: Ken on September 15, 2003 03:13 PM

snsterling wrote, "I don't see why it's [patents on math and other abstract fields] any worse than allowing patents of physics or chemistry."

I just tried a google search and couldn't find it, but I'm pretty sure that original patent law held that patents could be applied only to *things*, not ideas alone. I assume that the law evolved that way for a reason.

Secondly, mathematical laws aren't "invented," they're *discovered*. And you shouldn't be able to patent discoveries, only inventions. I know that sounds flakey, but most mathematicians view new math as discoveries, not inventions.

A similar but even stronger case can be made against biotech patents of naturally occuring genes--since the gene is already there, how can you patent it? (I.e., the gene is discovered, not invented.) Turns out the patent is not for the gene itself, but for complementary DNA that is made when the gene is used. Sophistry on the behalf of rent-seekers, say I.

Posted by: Stephen J Fromm on September 15, 2003 07:24 PM

(Philosophically I would say that -all- "inventions" are really discoveries of phenomena which existed all along.)

You are right about patent law excluding ideas, but I think you have the wrong sense of it. Patent law demands that one develop something which actually does what you say it will. For example, you can't go to the patent office with the idea that using the Bernoulli Effect you can make a winged airplane. You have to submit the designs for the actual working airplane which resulted from the idea. I suppose this is reasonable since it rewards the end result of the research with all the bugs worked out and not the idea stage, however clever those ideas might be. Here's a link which describes this issue as relates to the transistor. Shockley was able to show he was the first to have the idea of how to build a point junction transistor, but he never produced a working design. http://www.pbs.org/transistor/background1/events/patbat.html

What would you say about a company which hired twenty researchers and after five years came up with a superior video compression algorithm using several novel non-obvious techniques. This is certainly a working product as opposed to the idea for a working product, and if one wanted to they could build a dedicated machine to implement the compression described in the patent. However, since a general purpose computer can emulate any information manipulator, is it ok if Microsoft reverse engineers 100 person-hours of investigation and rewrites a compressor using the exact same algorithm step by step? That doesn't sound right to me. As long as the result of their work is sufficiently specific and non-obvious, it seems to me much more like a thing which was invented then an idea which was discovered.


Posted by: snsterling on September 15, 2003 09:20 PM

oops, person years is what I meant. And some grammar problems. Yuck.

Posted by: snsterling on September 15, 2003 09:28 PM

snsterling:

The patent case you invent up there doesn't sound OK to me at all. Firstly, when it comes to software, the company has ample methods already to keep their invention to themselves. Eg, they will keep their source code secret, they will maintain a copyright on source code, and they will have complete control of any API they wish to publish. They have no reason to publish how their video codec works. Indeed, if you look at how video codes have been implemented, this is precisely the case.

So what does the patent do, then? It gives a company that already had a very significiant business advantage an even bigger cudgel. And what will they use that club for? To make sure they have a complete monopoly on the software. And the length of patents is sufficently long that, in computer time, it is a practically infinite monopoly.

This is not a healthy marketplace. Add to that the further fact that our current patent office is completely incapable of sorting out which software patents might be OK, and which ones are a total joke. The only prudent course of action, barring gigantic reforms in patent and copyright law is to simply banish software patents. Sigificant safeguards are already in place to give software makers competitive advantage as it is.

Posted by: Timothy Klein on September 16, 2003 12:12 AM

I've just recently read that famous Marx quote about conflicts between the order of society and new means of production. (And I can't remember where it's from, bah.) In any event, I think Comrade Karl called that one.

Posted by: Randolph Fritz on September 16, 2003 02:11 AM
Post a comment