Home Page
07.13.2006 (previous | next)
I Ask You, What is a Patent Troll?

An ancillary issue in current US patent system reform discourse asks “what is a patent troll?” Having reviewed some of the literature, I argue that consistent with the established concept of division of labor, as well as common and necessary practices in the technology industries, non-manufacturing firms and firms that do not compete in markets in which they hold patents should not be considered patent trolls.

The attributes commonly used to call out a patent troll include: abuse of the patent system, litigious, non-contributing of ideas, non-manufacturing, seeking exorbitant licensing fees, spending nothing on development. Still, as Nathan Myhrvold states: “nobody knows what a patent troll is.”

I know what a patent troll is not. Innovators who risk substantial investment in R&D and then patent their work, who do not specialize in production capability; thus, I would not call research universities patent trolls when they license or "spin-off" rather than manufacture inventions. Further, patent holders who seek to license their patents arguably help develop products, and thus should not be vilified. It does not matter if they actually compete in markets in which they seek to license. California Congressman Howard Berman does not refer to these parties when he talks about patent holders that "misuse the patent system." Thus, after setting aside various entities in the "is-not-a-troll" category, what is left that we can call a patent troll?

Paul Misener, Amazon VP for Global Public Policy, spoke before the House Committee on the Judiciary in June 2006 and claimed that the patent system can be reformed without defining what a troll is. However, Misener admitted that the term “vilifies entities that simply are taking advantage of the current, flawed patent litigation system.”

Elsewhere, there seems to be a vague consensus on the intent of the term “patent troll.”

Professor Donald Chisum refers to trolls as those who have patents but “do not produce anything under the patent or even enter into prospective, cooperative licensing arrangements.” Instead they "wait under a bridge” for a company to produce a product and then seek exorbitant licensing fees above the value of the patent or threaten litigation. Generally patent holders are mutually inclined to cross-license patents but a troll has no such interest because it develops no products.

I read Professor Chisum to imply that patent holders who license and exchange patents as not being patent trolls. Such patent holders partake in the innovation-patent process.

Nathan Myhrvold’s view that non-manufacturing patent holders “have long been a part of America’s engine of innovation,” refines Professor Chisum’s position. He states: “it is important to note that the vast majority of patent holders who don’t make products are perfectly honorable." Myhrvold finds the central characteristic of patent trolls as those who manipulate or abuse the intent of the patent system.

Myhrvold fits some definitions of patent trolls "waiting under a bridge," however he points out that his company, Intellectual Ventures, has not filed a single suit, making it (jokingly of course) at least a non-effective troll.

Dean Kamen, inventor and licensor of patents for the Segway scooter, in testimony to the House Judiciary Committee, also found himself fitting the description of a patent troll. He explains his position however:

When I walk into that large company, they've got marketing, they've got distribution. If I show them what I've got, the only thing I have on my side of the table is that patent. The only way to get them to commit huge resources to turn that into the product, is to say to them, 'You exclusively have the right to do this.'

Inline with Myhrvold’s view of trolls holding ill-motivation regarding the patent system, Professor Mark Lemley contributes to the debate on what is a patent troll: “(Trolls) use the patent system not to develop and make products but to squeeze money out of those who do.” Lemley also acknowledges that large corporations are often the target of patent trolls. He differentiates these patent owners from trolls, citing that trolls “pop(ing) up years or even decades later and try(ing) to fit an old patent to a different purpose.”

There is sufficient description of what constitutes a "patent troll" to know what one is without vilifying parties that otherwise risk substantial investments and hold patents as the primary means of commercial and technological exchange. Although some claim resolution of "patent troll" is not important, the increasingly divergent definitions arising out of policy discourse often confuse productive activities for those that patent system reform aims to fix.

posted by Noel Le @ 1:39 PM | Patents

Link to this Entry | Printer-Friendly | Email a Comment | Post a Comment(9)


Comments

I don't know about other industries, but there seems to be literally no quality control with software patents. I have frequently seen patents highlighted on Tech Liberation, Slashdot, etc. that are so bad that a typical Computer Science senior, with no knowledge of them, could reproduce the algorithm off the top of his head during a job interview.

Take this(http://www.infoq.com/news/RedHat-Sued-Due-to-Hibernate-3-O) for example. It's absurd on its face, but moderately better than this(http://www.techliberation.com/archives/040151.php). There is nothing innovative about ORM as a concept as it is just a way of interfacing two different paradigms: relational databases and object-oriented programming. All this patent describes is the idea of using a file to map the attributes from one paradigm to another. That's trivial for any programmer to do. The next step is to generate an object using that mapping. Again, nothing new and exciting, or even very difficult for many experienced programmers, even if you are talking about dynamically generating the machine instructions programmatically (which is possible with Java 6 and .NET).

Some things, such as the RSA algorithm were/are sufficiently innovative to warrant patents. However, these are not software ideas, but patents from other areas of work. In software development, there is little actual work here that is innovative. Most of it is in fact rehashing of old ideas and methods.

Posted by: MikeT at July 14, 2006 9:50 AM

Although I appreciate your comments, they go beyond my original topic of "patent trolls."

I understand your point about the importance of patent quality, perhaps here you were indirectly addressing patent trolls; but I take a less drastic view. That USPTO administration is a critical issue for software patents doesn't reflect on how well software patents serve their function and role.

I'm not clear what you mean about the RSA algorithm: "these are not software ideas, but patents from other areas of work."

Posted by: Noel at July 14, 2006 11:21 AM

The RSA algorithm is a cryptographic patent, not a software patent. People make the mistake of assuming that because a very unique idea is implemented in software that it is a software patent. You could implement RSA directly in hardware if you had the hardware engineering and math background. Most software patents are, for lack of a better phrase, "software engineering process patents." They are attacks on the process of making software, not a specific patent on a specific algorithm.

The Hibernate example that I provided is basically a patent directly on an idea, or rather a set of ideas merged together. RSA is not. It's a specific, detailed algorithm that (probably) cannot be reproduced without reproducing the fundamental algorithm. The changes between a RSA corporation implementation and a Microsoft implementation would probably be trivial. At its heart, it can only be implemented one way, lest it become a true variant of RSA.

There is a very fuzzy area between algorithms and ideas because code is a representation of an idea. It would be possible to effectively wipe out software patents without wiping out patents on codecs and encryption algorithms. In fact, it would be very good for the industry to wipe out software patents as the industry got its start by systematically violating patents in the first place. The PC industry's origins would not be possible today with people like Nathan Myhrvold working against it today under our laws.

My point is that the systematic lack of quality in software patents is the main reason that patent trolls exist in IT in the first place. The only way to fix it, and forgive me for being a code monkey here, is to scrap the entire patent system in **this** area and take a sober look that reflects on the role of patents in the past and how they have worked in practice with the various forces driving the industry.

Another issue that this industry faces that others don't is the open source movement. The flaming partisans aside, many of us who do a little OSS programming on the side, do it for the fun of it. You can't do this in other areas of engineering. People don't collaborate and compete from their homes against Monsanto in biotech or Honda. This is a major, major issue that must be taken seriously. The primary, non-governmental solution to Microsoft's browser monopoly comes from Firefox, for example.

The trolls are, I think the monstrously mutated canaries in the coal mines.

Posted by: MikeT at July 14, 2006 11:43 AM

*...systematic lack of quality in software patents is the main reason that patent trolls exist in IT... The only way to fix it...is to scrap the entire patent system in **this** area and take a sober look that reflects on the role of patents in the past and how they have worked in practice with the various forces driving the industry.*

I suspected your main issue was patentable subject matter. However, there must be a far less drastic remedy than "Napsterizing" hundreds of thousands of software patents, pulling the rug out from under an entire industry and rewinding American innovation policy back to the cold war era. Unless the Defense Department will fork over many billions of dollars to subsidize the technology industries, there's no way early Internet era regulatory policy would help the industries right now.

Basically, I'll agree with professors Mark Lemley and Julie Cohen that the patentable subject matter issue is "for the history books." Some scholars, such as Professor Kenneth Dam, foresaw that software patents would pose some new obstacles for the USPTO and that the resolution would be found in what today's real policy debate focuses on: patent system reform.

Posted by: Noel at July 14, 2006 12:20 PM

"Napsterization" is, I think, a misunderstanding of what I want. I am not advocating eliminating the ability of software companies to patent bonafide research. There is in fact little bonafide patent-worthy innovation in software development today. It is so minimal that protecting it would create the potential to protect a hundred times more bad patents because very few people in the legal profession have any serious training as software engineers or computer scientists.

The innovation you speak of occurs in cross-discipline areas, such as cryptography. "Napsterizing" software patents wouldn't affect those areas of real innovation. It would affect things like the patents I provided you above, which are obvious on their face.

I have yet to see anything which actually shows that any successful software developer has relied on patents for a meaningful percentage of their revenue. If anything, they rely on being first to market with the best features and largest group of people who rely on their products for their own livelihood. That's how Oracle, for example, is so successful. PostgreSQL is open source and provides most of the same practical features that Oracle does, yet Oracle has a tremendous base of loyal developers and DBAs pushing it. They could thrive without even a single patent in their portfolio, AND PostgreSQL would continue to nip and their heels and provide them an incentive to not rest on their laurels.

I fail to see how ending software patents would actually damage or destroy American IT. The industry has been so productive precisely because there was such a cavalier attitude toward patents. Companies take ideas from each other all of the time, but the ones with the best products still tend to win and win very handsomely.

As I said, if patents were enforced in the past, you wouldn't have a commodity PC today nor would you have AMD and Intel competing in the same product space, driving down costs to buyers. The very industry that you want to protect, wouldn't exist anywhere near its current level if had been "protected" back when it was developing. Commodity software and hardware exist because there have been almost no patent restrictions on making compatible products.

Again, this is an issue that is unique to this industry. Programming in particular is not like any other industry, and you have to protect its IP under different conditions than you do elsewhere. I am completely pro-patent on most issues, just not on this one because the industry is an exception to the rule.

Posted by: MikeT at July 14, 2006 1:21 PM

I’m not going to argue about “what is a software patent.” I’ll settle on defining patent trolls for the time being. However, I will point to pages 9-12 in Ronald Mann, "Patents and Business Models for Software Firms" (April 5, 2006), ExpressO Preprint Series. Working Paper 1239, which provides good discussion of different ways of categorizing software patents.

READER COMMENT: “There is in fact little bonafide patent-worthy innovation in software development today. “

I’m trying to follow here. Are you saying that there is so little innovation in the software industry, and hence, that software patents should be eliminated? This sounds like an assumption of the open source movement, as they’ve given up making money from their ability to come up with new software inventions.

READER COMMENT: “I have yet to see anything which actually shows that any successful software developer has relied on patents for a meaningful percentage of their revenue.”

A good way to measure the revenue you speak of comes from licensing revenue of software patents. I apologize for not having industry figures, but I’ll dig this up and post when possible. If any readers are from AUTM or LES, please email me.

READER COMMENT: If anything, they (software companies) rely on being first to market with the best features and largest group of people who rely on their products for their own livelihood. That's how Oracle… is so successful.

Who would steal an Oracle system? What would they do with it? There’s simply little danger of misappropriation. This is the issue I raised in my Coca-Cola trade secret blog.

READER COMMENT: The very industry (PC, semiconductor) that you want to protect, wouldn't exist anywhere near its current level if had been "protected" back when it was developing.

I’m baffled when reading comments such as this. The important lesson is that IP policy should enable sustained and future innovation. Whats the point of rewinding the clock, and saying that current IP policy should be shaped to re-invent the Internet, the PC or the semiconductor chip. The Internet has already been invented. The software industry currently exists. Consumers can choose between AMD and Intel chips. We surpassed the technological dark ages decades ago.

READER COMMENT: I am completely pro-patent on most issues, just not on this one because the (software) industry is an exception to the rule.

I don’t see how software should be disadvantaged from other kinds of innovations by not receiving the protection of patents. Even though innovation may occur in the interdisciplinary fields that you mention, even if you don’t work on pure software inventions, others may and they should be able to enjoy benefits of the patent system that you do. I also recognize that innovation in software may occur differently than in other industries; but the concepts of patent quality and patenting standards also apply to software.

Posted by: Noel at July 14, 2006 3:43 PM

Noel,

It is obvious that you and I approach this from two completely different perspectives so I will say my last piece now.

The very reason that I say that software patents are inherently bad is that the bulk of them deal with common processes or trivial apporaches. The very reason that I cite interdisciplinary patents, such as cryptography patents, as ones that should remain is to point out that in software development, those are the ones that are non-trivial and patent-worthy.

The way that the industry works is that cost to production is lowered through competition among compatible products. Supporting multiple platforms, hardware or software, is extremely expensive and time-consuming. The very reason that the PC was a success is that it standardized quickly around a common compatible base that was unencumbered by patents. The reason that the Internet works is that protocols are open, well-documented and accessible.

A lot of us geeks have a semi-emotional attachment to this issue because it restricts our hobby as well. There is no other area of production where someone can make a thing that competes with a multi-billion dollar company in his or her bedroom and then collaborate to distribute it for free without any real cost to them to others.

The patent troll issue is, as far as I can tell, one that is mostly an issue for software patents. I am not against allowing software companies to get patents for things they put it in code, provided that the patents are for non-trivial, specific algorithms such as the RSA algorithm or the MPEG4 codec. What you have now is a patent system that is so broken that even if a software engineering concept was sufficiently novel to be worthy of a patent, it would be a needle in a poisonous, radioactive haystack not worth going after.

That is where things break down. What most software patent supporters end up supporting are things that would NEVER be patentable in other industries. That ORM patent I cited is sufficiently generic as to be the equivalent of the USPTO granting a patent on the very concept of a hybrid engine, rather than the Prius engine. It's a way of locking down an entire product **category** for our field.

What I am saying is get rid of the engineering patents. These are the process patents that cover everything from aspects of language design to ways of designing user interfaces. There is a minefield of prior art here that somehow the USPTO seems to navigate perfectly and blindly every time it issues a software patent.

I am very conservative about this. What's worked in the past is likely to work in the future. Technology rarely advances radically within a discipline. Evolution is the norm. UNIX is still the dominant design foundation, even after 30 years. That should tell you something about how things work here :)

And you're right. In **theory**, patent quality and standards apply to software patents, but in practice they don't. I have yet to see a pure software patent that didn't leave me shaking my head in disgust.

Posted by: MikeT at July 15, 2006 12:30 AM

Well, I appreciate your knowledge and expertise. Several closing comments...

I don’t think we disagree that much: there are many bad patents that threaten innovation and the standard for non-obviousness has been lax in some areas of patents. However, being conservative, and in my mind, practical, I try to involve myself in the patent reform debate rather than calling for the abolishment of a whole class of patents.

I do take issue with one of your points however. Although you clearly have great technical understanding, I would not approach innovation policy from a “hobbyist developer” point of view. Even if creating code enables you to create something that “competes with a multi-billion dollar company… and then collaborate to distribute it for free without any real cost to them to others,” these considerations pale next to more serious aspects of the technology industry, such as improved standard of living, entrepreneurship, job/wealth creation, economic growth and industrialization.

Posted by: Noel at July 19, 2006 5:44 PM

Benefits the borrower

Posted by: LOWGAUGE at March 19, 2007 10:07 AM








 
IPcentral WebLog

Blog Main

IPcentral Blogosphere Archives

Search the Blog

Recent Posts
  - IP and Marginal Cost
- Academics and Copyright
- More on Jammie Thomas from DOJ
- More Studies of Downloading
- Facebook, MySpace, and Network Externalities
- Copyright and the University: An Academic Symposium
- Tyler Cowan on Chinese Movie Piracy
- More WHO Antics--Roger Bate Reports
- Patents, Meds, and the Developing World: Clips & Links
- Jermaine Dupri's Gripe with iTunes
Archives by Month
  - December 2007
- November 2007
- October 2007
- September 2007
  - (see all)
Archives by Subject
  - Academia
- Access: Commons, Fair Use, Orphan Works, Public Domain
- Accounting
- Analog Holes
- Antitrust
- Art
- Aspen
- Big Tent
- Biotech
- Books
- Comments from Readers
- Counterfeit
- Digital Americas
- Digital Europe
- Digital Europe 2006
- DMCA
- DRM & Watermarks, etc.
- Economics, Game Theory & Public Choice
- Enforcement & Remedies
- Free Culture Movement
- Games
- General
- Infrastructure
- International
- Internet: P2P, Search Engines...
- Legislation and Legislators
- Liberty and IP
- Markets: Business, Investment & Innovation
- Media: Video, Music...
- Patents
- Pharma
- Physical Property
- Prices, Terms, and Licensing
- Privacy and Security
- Radio
- Software
- Spectrum & Wireless
- Standards
- Supreme Court
- Tax-Funded IP
- Telecom
- Theft of Service
- Universities
Links
 

Site Feed

  - Atom
- RSS 1.0
- RSS 2.0
We welcome comments by email - look for a link to the author's email address in the byline of each post. Please let us know if we may publish your remarks.


 
Home Page