Intellectual property policy for the XML geek

Eduardo Gutentag


This presentation will try to make some rather abstruse terms and concepts that relate to Intellectual Property clear and accessible to an audience that for the most part would rather not think about them. Once that's done, and if there is anyone left in the audience, the author will try to paint an accurate picture of where various standards organizations stand, and the various choices members of the audience face.

Keywords: Intellectual Property

Eduardo Gutentag

Eduardo Gutentag has been involved in the SGML/XML world for over 15 years, and has participated in many SGML/XML related activities at W3C and OASIS (even when it was still called SGML Open), as well as at ebXML. He is currently a Technology Director at the Corporate Standards group in Sun Microsystems. Through many twists of fate he has become is Sun's W3C AC representative, a member of the W3C Advisory Board, a member of the Idealliance Board of Directors and a member of the OASIS Board of Directors, where he currently serves as its Chair. In spite of all that, he's kind of an ok guy.

Intellectual property policy for the XML geek

Eduardo Gutentag [Technology Director; Sun Microsystems, Inc.]

Extreme Markup Languages 2006® (Montréal, Québec)

Copyright © 2006 Eduardo Gutentag. Reproduced with permission.


One of the unavoidable realities of the times we live in is that IP [Intellectual Property] protection is a fact of life, even though many believe it hinders interoperability and hurts the public good. Some standards organizations, through the companies and organizations that make up their membership, are trying to compensate for this through upgrades to their patent policies and practices. This comprises both a tightening of the members' obligations and an exploration into newer models. But it is hard to know, at this stage, whether their efforts are successful or not in the great scheme of things. This would not be so hard were it not for the fact that one of the unfortunate realities of the times in which we live is that the patent system in the USA [United States of America](and increasingly in other countries and geo-political entities as well) is sick and broken.

In the case of the USA it is more or less clear what the reasons behind this breakdown are; an even perfunctory examination of the history of the US Patent and Trademark Office will reveal that at some time after the Reagan presidency the funding model for the USPTO [United States Patents and Trademarks Office] was changed, and instead of being fed off taxpayers monies, its funding model started relying on submission fees.

As its Introduction page says[PATOFF]: "Since 1991--under the Omnibus Budget Reconciliation Act (OBRA) of 1990--the agency has been fully fee funded " (my emphasis).

Following this action, the law of unintended consequences took over: since the more submissions there were the better off the Office was (at least in the sense of ensuring continuous employment to its people), it became counterproductive to discourage submissions because rejected submissions discourage future submissions. Therefore no submission was ridiculous enough, un-inventive enough or trivial enough to be dismissed.

And of course the second consequence was that since the USPTO needed more inspectors to review more patents, because there were more patents to review because everybody was submitting since chances of their submissions being rejected approached nil, well, it had to fund them through the submission fees, so again those new inspectors better not reject submissions if they wanted to keep their jobs.

According to Robert P. Merges [MERGES] , while the official estimate of the ratio between patents submitted and patents granted in the USA is 62%, he quotes the estimate presented by Cecil Quillen (12 Federal Circuit Bar Association, 2002) as being as high as 97%, while that of Robert Clarke, from the USPTO, is presented as a moderate assessment of "just" 75%.

It is obvious that even at 75%, any process that grants that many submissions is completely out of control.

Undoubtedly none of the above is a conscious act on the part of anybody at the USPTO, but it certainly explains some of the absurdities that we have seen in the past years. For instance, US patent [6,960,975]' A space vehicle propelled by the pressure of inflationary vacuum state is provided comprising a hollow superconductive shield, an inner shield, a power source, a support structure, upper and lower means for generating an electromagnetic field, and a flux modulation controller', and further down: 'By creating alternative anomalies and modulating their parameters, the space vehicle's crew would dilate and contract time and space on demand .'

Let us be clear: this is not a patent submission, this is a patent that has been awarded. Now, granted, this may indeed be rocket science, so there may be something very subtle here that I'm missing which may be the basis for granting this patent. But what about all the other outrageous patent applications that we all have heard about from time to time, applications that are obviously trivial, if not downright silly. Some may be urban legends (a patent for the wheel???), but some are definitely not.

A visit to LawGeek's Silly Patents page [LAWGEEK] or to Patently Silly [PATSILLY] can be a very informative experience. As can be just thinking about the concept of patentable business processes. A business process is simply a process that is composed of more than one step; not a single one of the steps that make up the business process needs to be patentable (in other words novel, non-trivial, etc.) but the concatenation of the steps can be considered to be novel enough to be patentable, even if it's mind-numbingly trivial.

And then there is the issue of patenting schemas. Not copyrighting them, mind you, but patenting some applications intimately related to a given schema. And you can't avoid thinking: what exactly is being patented? The element names? The attribute names? How many names must you use from a given schema before you can be found infringing it? What about the content model? If a content model is exactly the same as some other content model and only the names are different, are you infringing? And, let's be real, how can anybody prove that you have infringed their schema, given that for any given instance there are at least two different schemas that validate it?

But the truly alarming thing about all this is that other countries are trying to imitate the USPTO - well, of course, there is a certain logic to that, since if they don't do that they weaken their position even more than what it already is today. The USA has become a veritable juggernaut of patents because of the ease-of-grant there . Anything that anybody wants to do anywhere in the world has probably been already patented in the USA. And even if it is sometimes clear that the patent in question could not stand in a court of law (either because the judge would die laughing before being able to issue a veredict, or because of clear previous art in the matter) disputing patent claims and patent-based suits is quite expensive, and most entities in other countries do not have that kind of money.

So if anything can be patented today, and if any patent can be acquired by trolls for the single purpose of charging money for licensing it, then everybody is at risk (including you) and everybody should understand some of these very basic concepts.


  • IPR [Intellectual Property Rights]
    Although this is a somewhat simplistic view, when we say "intellectual property rights" we mean either copyright rights or patent rights.
    We all know, of course, Proudhon's saying "Property is theft". Those who agree with him could be tempted to assert that Intellectual Property is also theft. However, to do so would reveal a rather bad misunderstanding of what Proudhon meant by "property". He did also write, after all, that "Property [is] a triumph of Liberty. For it is born of Liberty ... Property is the only power that can act as a counterweight to the State [...]". So let's be clear: there is nothing wrong with the concept of having rights that cover one's intellectual property; after all, one should be able to own whatever one produces, (although many [if not all] companies, universities and other entities, have clauses in their employee contracts and grants that make them the owner of any individuals' intellectual property that is produced with their equipment or during working hours.)
    But the real question is: what constitutes one's own intellectual property? If I were to say now "Property is like stealing", would that be a true manifestation of intellectual property? Or would that be just a re-phrasing of Proudhon's words?
    In what way is a business process different from saying "Property is like stealing" in terms of originality, non-triviality, etc.?
  • RAND [Reasonable and Non-Discriminatory]
    This term refers to a special kind of license granting that is generally negotiated in secret between the grantor and the grantee; the most common interpretation of what it means is that the payment that the grantor can extract from the grantee must be reasonable, and that the conditions under which the license is granted must not discriminate against someone else. Past these very general observations, everything else is open to interpretation and, therefore, can ultimately be decided only by a Court of law, including whether a very reasonable fee, accepted by all as such, remains reasonable or not when coupled with all the other reasonable fees that the grantee may have to pay to all possible grantors for a given implementation.
  • RF [Royalty Free]
    This term refers to a special kind of license granting that is generally announced publicly by the grantor. When used in contrast to RAND-Z (q.v.) it means that other than a couple of minor conditions (like reciprocity, by which others are required to also grant their licenses on Royalty Free terms) there are no other conditions. When used by itself, it is ambiguous, though it is generally safe to assume that it is being used as if it were RAND-Z.
    Another possible source of ambiguity is that RF is generally taken to mean "for no payment whatsoever", but royalties are a very specific form of payment, so it could be argued that asking for a one time licensing payment is not the same as asking for royalties and therefore such a license is still RF even though a user would have to pay a one-time fortune.
    However, I have never heard of any company or entity making this argument.
  • RF on RAND Terms (aka RAND-Z)
    Sometimes spelled out in its entirety as "Royalty Free under reasonable and non-discriminatory terms" , this term refers to a special kind of license granting that is generally advanced as the only or best or preferable mode for Royalty Free licenses. However, grantees of these licenses must suffer some of the same inconveniences that grantees of RAND licenses suffer. So while they represent a clear concession on the part of the grantors willing to offer these licenses, from the point of view of many potential grantees they are not such great shakes.
    For reasons that completely escape my understanding, some RAND-Z proponents get very agitated when this kind of licenses are referred to as RAND-Z, so one can derive some perverse satisfaction from teasing them with it.

Work of the Devil?

Some developers (irrespective of whether they work for large companies or not) view patents as not just something that can be an irritant, but as something to which negative ethical or moral attributes can be attached. In their view, nothing good can come out of patents. And no upright person should ever obtain a patent.

By contrast, some people believe that anyone who disparages the beneficial effects of patents on society, or who does not appreciate the superiority of RAND over any other IPR mode, should be well, not precisely drawn and quartered, but not considered an upright person for certain.

Patents themselves have become an object of controversy, rather than the merits of a given patent or claim. And it is a rather heated controversy. It approaches religious fervor. And at times it also approaches religious logic.

For instance, in an excellent article on the absurdity of some patents, Richard Stallman [STALLMAN] goes from describing how certain class of software patents would have prevented the publication of novels in the 1880's to a veritable leap of faith at the end of the article by proclaiming "The way to prevent software patents from bollixing software development is simple: don't authorise them." As a clarification, let me add that I am not stating that I agree or disagree with the above statement, only that it constitutes a non-sequitur with respect to what precedes it. Yes, some patents are absurd; yes, some patents are unfair. It does not follow that the only way to avoid that is to prohibit patents. Many more arguments are needed to arrive at that conclusion.

On the other hand, there are innumerable essays and articles written from the opposite side that run through the whole gamut of arguments in favor of patents, basically proclaiming that if it weren't for patents there would be no inventions or progress. It is common to encounter assertions like "[...] the standardization of a patented invention can yield procompetitive benefits, stimulate innovative research and development, [etc.]" [MARASCO] or "RAND licensing also facilitates adoption of standards and promotes further technological innovation" [ACT] with not one single shred of evidence to support these assertions.

There are in fact some very reasonable and likable people on both sides of the argument, so it's only fair to give them both a fair hearing. Now, of course, a fair hearing does not necessarily mean neutrality and impartiality. All it means is that no insults are permitted...

Viva RAND!

So what do reasonable people have to say in favor of Reasonable and Non-Discriminatory licensing terms?

Amy Marasco (then Vice-President and General Counsel of ANSI) did a presentation to the FTC in 2002 [MARASCO]. In it she quotes the FTC Chairman Timothy Muris saying: "The goal of patent and copyright law, as enunciated in Article I section 8 of the Constitution, is "To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." IP law, properly applied, preserves the incentives for scientific and technological progress - i.e., for innovation. Innovation benefits consumers through the development of new and improved goods and services, and spurs economic growth.[...]."

And then she proceeds to assert "Accordingly, the standardization of a patented invention can yield procompetitive benefits, stimulate innovative research and development, and make the patent holder’s intellectual property more accessible to consumers through competing products."

So the logic goes:

  1. the US Constitution asserts that allowing authors and inventors to be the sole owners and users of their writings and inventions promotes progress.
  2. therefore standards that include that allowance promote progress, stimulate R&D, and make the invention more accessible to all.

This is a recurring theme: allowing inventors to recoup their R&D costs and to make some money on their inventions through RAND licensing can only result in more money being invested in R&D and more inventions, which equal progress, and it also results in access to the inventions by more people at lower costs, which equals progress, which equal benefits for all.

But arguments asserting that IP protection spurs innovation seem to forget the existence of the so called 'patent trolls' (sometimes also called 'patent terrorists') who acquire patents by any means rather than develop them. Patent trolls are clear instances of IP-protection-seeking entities that are doing so for reasons completely unrelated to innovation, or to furthering the public good.

It is probably a good idea to remember that in fact RAND is a relatively late development. The most authoritative recounting of its development can be found in what Carl Cargill, Director of Corporate Standards in Sun Microsystems, had to say on this [CARGILL] :

"In the mid 1970's, a series of questionable practices, and a Supreme Court case, rocked the U.S. standardization arena, and Congress began to investigate the standardization activities of the private sector. ANSI, representing a federation of all the private sector standardization participants, rushed to correct the process errors that allowed the violations to occur. It was here that the process requirements that included the patent policy of ANSI first began to appear. Participants who submitted technology for standardization had to commit to licensing the necessary patented technology included in the standard on a Reasonable And Non-Discriminatory Basis (RAND)."

This explains perhaps why in the eyes of many, the establishment of RAND obligations is actually something that saved the day, and the mere mention of changing the practice is received as the equivalent of burning the flag, for two reasons: one, because RAND is seen as a protecting mantle, and second because the mere mention of not charging for licensing is seen, by many though not all, as a slap in the face of the principle that monetary incentives benefit all because they spur invention.

For a good example of this position, see the paper produced by the Association for Competitive Technology [ACT]; for a good example of an extreme view of this position, see the summary of that same paper [ACT], especially the sentence: "History dictates that firms have no incentive to unduly burden a standard with restrictions that could suffocate adoption, but seek only measured restrictions that foster greater use."

The mystery of how a restriction can lead to expanded use is left unexplained.

Viva RF!

Not surprisingly many of the arguments advanced in favor of Royalty Free are in fact only anti-RAND.

One very convincing argument is actually advanced by [CARGILL] as a footnote to the paragraph quoted above: "The lack of a clear, equitable, and easily understood definition for the term RAND is one of the major sticking points in ICT industry. Small companies and individuals fear gouging by big companies; large companies fear diminution of their IPR portfolios."

What Cargill is referring to is the fact that there is actually no clear definition of what constitutes "reasonable" in terms of fees, nor what it means to not discriminate with respect to the terms; nor is it clear that "reasonable" refers only to the fees, or that "non-discriminatory" refers only to the terms. And surely what for some people is clearly non-discriminatory (like Field-of-use clauses, which restrict under what implementation conditions the license is granted, see [ACT] page 4), for others it is clearly one of the most discriminatory conditions that could be imposed.

One very good summary of strong arguments against RAND can be found in what Robin Cover writes under the section "Why RAND is Not Appropriate for Internet Open Standards" [COVER] ,starting with:

By what means are we supposed to be comforted by such a promise [of RAND licensing]? The notion that RAND ("reasonable and non-discriminatory") terms and conditions protect the public needs to be debunked. RAND puts the patent owner in control of a government-backed monopoly. RAND offers no certain protection against license fees that constitute oppressive global taxation, and it may be judged highly discriminatory. Hence one writer's characterization of RAND as an expanded acronym: "half of 'RAND' is deceptive and the other half is prejudiced."

One thing that is not mentioned in [COVER] (in fact, I don't remember seeing it mentioned anywhere) is that it is rather difficult to know whether some terms are discriminatory or not given that the negotiations regarding the granting of RAND licenses do not happen in public. In fact, they are quite secret, and in many cases secrecy is one of the terms of the license being granted. So not even a court can know whether a given term is discriminatory, unless it has access to the terms offered by the licensor to others.

Of course, the influence of open source has increased enormously in the last few years, and therefore arguments regarding RAND's incompatibility with open source and open source communities' needs have become increasingly louder. In particular, the argument that open source developers do not have the knowledge, the wherewithal or the time to deal with obtaining RAND licenses, is quite arresting.

At first blush it may be disturbing that many of the arguments that can be categorized as "pro-RF" can actually be seen, or are even presented as, "anti-RAND". But not all arguments in favor of RF are necessarily negative. Some can be in the form of questions. For instance, at the W3C Advisory Committee meeting in December 2005 I argued [GUTENTAG] in the form of a few pointed ones in order to show that companies may in fact be better off offering Royalty Free licenses than RAND ones, such as "How much money do you expect your company to make from IP royalties related to internet/web activities this year?", vs. "How much money do you expect your company to make from internet/web related products or services this year?" The point being actually a positive one: it may be more profitable to license patents for no fee in order to stimulate a market than to charge for them and thus put obstacles in front of it. This so-called "network effect" is in fact one of the most effective arguments I can think of in favor of RF.

Many, or perhaps most arguments in favor of RAND emphasize the stimulating effect that RAND is supposed to have on development, inventions, research and progress. But though presented as conclusions, these are in fact axioms, because as far as I know there is no established body of study that supports them with figures.

Conversely, there is no established body of study that supports the opposite axioms, that RAND licensing stiffles competition and slows development and invention.

On the other hand, there are empirical contradictions to some of the assertions that are made in favor of RAND. For instance, that article from the Association for Competitive Technology contains the following assertions ([ACT], page 10):

"Standards organizations have a history of incorporating patented technology into standards, and the standards- setting process would suffer greatly if they began to exclude solutions that contained patented technology. Many standards containing patented technology have been introduced to the marketplace and have experienced widespread implementation. As such, there is no evidence to support the suggestion that patented technology, when incorporated into standards, deters adoption and use."

What this arguments forgets is that while there may be no evidence to support that suggestion, on the other hand there is plenty of evidence to suggest that non-patented technology (or at least technology that can be implemented with no worries with respect to licensing) is behind some of the most successful software innovations, bar none, such as the whole Internet and Web infrastructure, and can stimulate as much if not more adoption and use than patented and RAND technologies.

What do Standards Organizations Do?

So, given all the above, it would be interesting now to contrast, compare and analyze what some standards organizations do regarding their Patent Policies.

One of the first things that must surely strike anyone who spends some time studying what standards organizations do and write regarding patents, is that they all follow the most important rule of all: above all, cover your derrière. No organization will try to determine whether a given patent has merit, or whether it applies at all to the specification in question, or, least of all, whether the claims that someone claims are essential for the implementation of a specification are indeed essential or just irrelevant...

For instance:

The IETF's RFC that deals with its Patent policy [IETF-3979]says in Sections 3 and 4: "The IESG disclaims any responsibility", "will take no position on the validity or scope", "The IESG will not make any explicit determination", "The IETF takes no position regarding the validity or scope of any Intellectual Property Rights or other rights that might be claimed", etc.

And the Patent Policy document at the OASIS site says in section 12 [OASIS]"OASIS assumes no responsibility to compile, confirm, update or make public any assertions of Essential Claims", "OASIS takes no position as to the validity or invalidity of such assertions, or that all such assertions that have or may be made have been referred to."

And many, many others.

To be fair, it could be argued that of course these and other organizations, with the very limited resources they have, would be irresponsible if they were to take upon themselves verification and determination.

On the other hand, one can only wish the opposite, because what we are left with is a system where the ultimate authority resides with the Court, which, in most cases, means either a clueless jury or a clueless judge.

One of the things that organizations have started doing over the past few years is actually tightening their disclosure requirements, or rather making them less ambiguous.

For the uninitiated, a disclosure is the act of announcing that you have a patent that covers something. It does not necessarily mean that you must also say whether you will license it or how you will do so. Sometimes it also does not mean that you have to disclose which particular claim of the patent covers which particular clause of a specification. In extreme cases it does not even mean that you have to disclose the patent number. But in all cases a disclosure obligation is needed as protection against the submarining of patents - that is, the manipulation of a working committee to produce a specification whose implementation would necessarily infringe on one of your patents, without telling anybody that you have such patent until the specification has become a standard and everybody is implementing it (for an example, see Robin Cover's discussion of the Rambus case [COVER-RAMBUS]).

A short comparison

So how do the various standards organizations compare to each other in terms of their patent policies? The following table should give you a good idea.

Table 1
IEEE Mushy: from no enforcement to RAND through RF
IETF RAND (But RF is allowed)
W3C RF (but with exclusion)
OASIS RAND, RF on RAND terms, RF like in W3C (but with almost no exclusion) - to be chosen by the TC.

A detailed comparison would tax any reader's patience, but the following is of note: in spite of the fact that many of the members participating in standards organizations would very much want to see RF obligations being imposed as part of the standards organizations' patent policies, this has not happened yet in its purest form in any of the organizations listed above. Although the W3C is mentioned many times as an example of such, the truth is that a compromise was worked into its patent policy that allows for a company to wiggle out of its RF obligation if certain conditions are met. So even in that case the most one can say is "yes, it's RF, but...". In the case of OASIS, there is obviously the RAND option for TCs to work under (although as of this writing there is not one single RAND TC being started or having started at OASIS); the other two options, both RF, permit a very short wiggle room to leave the TC and not incur on subsequent participation obligations if one does not like what is going on there.

A New Model

So these are the choices that a patent owner has had until now:

  1. Not to license at all - allow no implementation
  2. License for a fee and unspecified conditions
  3. License for no fee but unspecified conditions
  4. License for no fee and no unspecified conditions

But there are two other alternatives that are increasing in importance; they are certainly viable alternatives, and developer communities should be aware of them.

  1. License for a fee and specified conditions (ex ante disclosure)
    One of the most disturbing things about RAND, as was indicated before, is the fact that the party that asserts that the terms of a license are reasonable and non-discriminatory is the party that imposes those terms (in other words, the licensor itself).
    Until now the suggestion that standards organizations should ask their members to disclose their terms in public, perhaps even as early as when they disclose their claims on a specification, were met with quite vocal protests. This, it was said, would create a situation in which anti-trust laws in the USA would be violated. But in a surprising development, Deborah Platt Majoras, Chairman of the FTC, stated publicly last year that she does not believe that ex-ante necessarily violates anti-trust regulations [MAJORAS].
    It is clear that in situations where RAND is unavoidable (and for now there are such situations, as when many members of a standards consortium would just walk away and create a new organization if RAND were to be excluded from the possible licensing modes), demanding ex ante disclosure is becoming more of a possible position that it was just a short year ago.
  2. Not to license at all - allow all implementations (non-assertion covenant)
    There has been lately a strong movement for companies to unilaterally issue non-assertion covenants; these are declarations that a company will not seek to sue or otherwise prosecute anybody who infringes that company's patents when implementing a given specification.
    There are many variations of these, some more liberal than others (for two extremes in terms of liberality, see [RSA] and [SUN]), but the basic thrust of this mode is extremely important for, among others, open source developers, since it allows them to just develop implementations of the standards in question, without having to worry as to whether they should obtain licenses, how many they should obtain, how long does it take, should they employ a lawyer, etc., etc. And while a non-assertion covenant obviously benefits open source communities, it also benefits others, freeing them from the anxious awaiting for the arrival of batallions of men in dark suits ready to take them to court.

Implementations of the New Model

Although there is no known example of a standards organization that has embraced ex ante as part of its patent policy, there has been plenty of talks in many organizations, so perhaps things will soon move forward in this front.

On the other hand, there is now a very good example of a standards organization that has embraced non-assertion covenants as the only possible patent policy for its members: UN/CEFACT.

UN/CEFACT is the organization within UN/ECE that, together with OASIS, developed ebXML a few years back. Shortly after the completion of ebXML the issue of what its patent policy actually was emerged, and it became clear that it was either a general UN policy that was so general as to be unacceptable to commercial enterprises participating in it, or it simply did not exist, since the one that UN/CEFACT thought they had in place had not been approved by the UN's OLA [Office of Legal Affairs].

Fortunately, a short time ago the UN's OLA approved, and UN/CEFACT ratified [CEFACT], a patent policy that, although called a "waiver", in effect is identical to any non-assertion covenant in that whoever participates in UN/CEFACT agrees to not assert whatever IPR they may possess regarding UN/CEFACT specifications. While this is far from being a perfect policy, and some issues remain regarding whether the interests of the participants are effectively protected, this is undoubtedly a victory for those who like non-assertion covenants, because the only effective way for NACs to be effective is to be many. The principle of the network effect applies here too.

But what about those nasty, nasty, patents? (A call to arms, as it were)

Let's return for a second to the issue of those nasty patents that are so irritating when we see them granted, like patents on stylesheets, or patents on schemas, or patents on how many mouse clicks one must do in order to order something online.

Large and small vendors apply for patents daily, to the tune of thousands per year. Some of those patents are legitimate, in the sense that they are clearly innovative, brilliant even, and as long as one accepts the current rules of the game, namely that legitimate work and investment should be rewarded, there is nothing to complain about.

Some companies are quite proud of their patents. Some companies' venture capital depends on they being able to boast a certain amount of patents submitted. Some companies make millions, if not billions, yearly on fees collected for their patents. So why do we get so agitated? Everybody wins, after all: people invent, people make money, people participate and share...


No. In some cases patents are probably deterrents. Statistics can show how many implementations there are in spite of patents; statistics cannot show how many implementations where never done because of the patents surrounding a standard.

An obvious path for developers to take is to push standards organizations to provide even more protection to the developers than what is currently being provided, and to have more and more standards organizations adopt ex ante policies, and more and more organizations adopt non-assertion covenants instead of licensing practices.

And patent owning companies can certainly take the path that teaches that interface patents should never be enforced; one thing is to charge for the use of an engine connected to the electrical grid, or to charge for the electricity that comes through the plug, but what is not acceptable is to charge for the use of the plug itself. Or, as Greg Papadopoulos said [PAPADOPOULOS]

One answer to all of this would be to create a Lessig-like Commons. Essentially, accept that all of society and the whole of industry is better served when ubiquitous, royalty-free, interface standards belong to everyone (e.g., TCP/IP, radio modulation, etc.).

It would be nice to make encumbered standards impossible by requiring real searches for previous art before granting them and by applying a severe test for “unusual, new and unique”. It would be nice, when a standard is indeed encumbered and RAND, for its terms to be declared "ex ante". And it should be the case that interfaces should be freely reverse-engineered for the sake of interoperability and it couldn't be protected by patents (including data formats and their encoding).

But this may not happen soon enough. Obviously the existence of interface-covering patents that must be infringed in order to read or write certain documents that conform to some schemas is quite disturbing, because this implies that while we assume that schema writing and validation is trivial and non-patentable, the USPTO and other Patent Offices around the world may have a different opinion, and isn't that a cog in the wheels of our peaceful life?

And what other things are there out there that we assume are trivial and non-patentable and interface-like but which, as we speak, are actually being patented by some incompetent moron in some dark office in the heart of the Empire?

What about the little things that we all do: stylesheets, schemas, and all those immensely clever developments and development tools that will never be part of a standards developing organization's output? Are those vulnerable to attack by either some unconsciounable vendor or by some troll whose only raison d'être is to make our life miserable? What about the so called Web 2.0, or AJAX, or Simple Web Services, which are actually nothing more than business processes? Are we developing ourselves into a trap? Are patents being submitted and approved as mashups proliferate?

Wouldn't it make a lot for sense for everybody to go out and submit as many inventions as possible and have a thousand patents flourish as a defensive counter-measure, and as an offensive reductio ad absurdum of the whole system? Patent pools everywhere? Patent ponds? Patents oozing out of the pavement? Dripping from the faucet? Growing amidst the cracks?

Would that flood the system? Would it exacerbate the situation? Would it result in multiple defensive pools, offerings, groups and alliances? Would it overhelm everybody and result in cries for real reform? Perhaps.

And the problem with that is....?


[6,960,975] Space vehicle propelled by the pressure of inflationary vacuum state

[ACT] Association for Competitive Technology, "Open Standards Patent Policies and Open Source Software Implementations", October 2004

[CARGILL] Carl Cargill, Director of Corporate Standards, Sun Microsystems: "Intellectual Property Rights and Standards Setting Organizations: An Overview of Failed Evolution ", presented to the Department of Justice and the Federal Trade Commission on March 27, 2002.

[CEFACT] "UN/CEFACT Intellectual Property Rights Policy", (ECE/TRADE/CEFACT/2006/11). 17 May 2006

[COVER] Robin Cover, in a long, exhaustive and excellent article on "Patents and Open Standards", OASIS

[COVER-RAMBUS] Robin Cover, "Patents and Open Standards -- DDR/SDRAM", OASIS

[GUTENTAG] Eduardo Gutentag, AC representative from Sun Microsystems, panel presentation, "The W3C Patent Policy in Practice",

[IETF-3979] Internet Engineering Task Force, RFC 3979

[LAWGEEK] LawGeek's Silly Patents page

[MAJORAS] Deborah Platt Majoras, September 23, "STANDARDIZATION AND THE LAW: DEVELOPING THE GOLDEN MEAN FOR GLOBAL TRADE", Stanford University, California

[MARASCO] Amy Marasco, "Standards-Setting Practices: Competition, Innovation and Consumer Welfare", Testimony before the Federal Trade Commission and Department of Justice, April 18, 2002

[MERGES] Robert P. Merges, "Empirical Patent Studies and Current Policy Issues", a presentation to the Federal Trade Commission hearings on Competition and Intellectual Property Law and Policy, 2005,

[OASIS] OASIS (Organization for the Advancement of Structured Information Systems) Intellectual Property Rights (IPR) Policy

[PAPADOPOULOS] Greg Papadopoulos, Sun Microsystems' CTO, Blog entry, "INTEL: IT'S TIME TO SHARE"

[PATOFF] United States Patents and Trademarks Office, Introduction,

[PATSILLY] Patently Silly

[RSA] Statement regarding IPR, submitted by RSA Security, regarding the SAML v1.0 through v2.0, 27 April 2006

[STALLMAN] Richard Stallman, "Patent Absurdity", 2005,,12449,1510566,00.html

[SUN] Sun OpenDocument Patent Statement, submitted by Sun Microsystems, Inc., September 29, 2005

Intellectual property policy for the XML geek

Eduardo Gutentag [Technology Director, Sun Microsystems, Inc.]