Wednesday, November 24, 2010

Brussels conference on "Tensions between intellectual property rights and standardization: reasons and remedies"

Last week I mentioned a European Commission and European Patent Office conference on "tensions between intellectual property rights and standards". The conference took place in Brussels on Monday (22 November 2010), and I attended it.

Different speakers pointed out that an increasing number of patents read on today's high-tech products. The European Patent Office has already received 4% more patent applications this year than during all of last year, and there is still about a month to go. While some view this trend as a key indicator of innovation and of awareness for the importance of intellectual property rights, others complain about increasing complexity and a growing risk of disputes.

Most industry standards are FRAND-based and some panelists stated that they don't see a need for a fundamental departure from the current system, which allows companies to opt for alternative models. Others, however, made proposals and demands that would in different ways weaken the rights of patent holders and discourage innovators with IP-centric business models from participating in standard-setting processes.

Major European stakeholders favor FRAND

The European Commission plans to take decisions on two standards-related initiatives before the end of the year. Other legislative initiatives on standardization will continue through 2011. Against that background, it didn't surprise me that some of the speakers made political statements and demands directed at the EU officials in the audience. But the loudest voices weren't the most convincing ones.

Before I address some of what was said, I'd like to share an observation. Speakers from US corporations (Oracle and HP) attacked FRAND from different angles and received support for this from the FSFE, a lobby group whose primary partners are IBM, Oracle, Google and Red Hat. However, there were key European stakeholders on the panels who clearly disagreed with them. Ericsson contradicted Oracle; SAP posed questions (to HP and FSFE) regarding the "openness" of open source software, the different business models that successfully leverage FOSS for financial gain, and the need to balance the rights of IPR holders and licensees.

Ericsson stressed that "the FRAND IPR regime has served the industry very well in the wireless communications area." Its vice president of patent strategies and portfolio management, Gustav Brismark, pointed to the fact that over the years there have been "numerous new entrants", which he views as an indication that the system is pro-competitive. In his experience, SMEs who make themselves knowledgeable about the patent licensing landscape can conclude the necessary license agreements with right holders and participate in the market.

SAP placed the emphasis on choices that are made by right holders on a case-by-case basis. For example, the standards developed by the World Wide Web Consortium (W3C) must be patent-unencumbered or patent rights must be available on a royalty-free basis (with possible restrictions set out in the organization's patent policy). Many of the companies participating in the W3C are, however, simultaneously involved with FRAND-based standard-setting processes.

From SAP's point of view, "market dynamics have worked very well so far", not only because there are different kinds of standard-developing organizations competing but also because certain SDOs such as OASIS create standards on the basis of FRAND as well as under other regimes.

There's nothing wrong with US corporations -- especially if they have significant European operations and partner networks -- participating in European policymaking debates. However, it's a legitimate question to ask who (and how credible and independent) the local entities on their side are. In addition to promoting continuity, which means the burden of proof is on the other side, the proponents of FRAND have the home team advantage: they can point to broad-based support for this approach among European politicians and regulators and a long list of impressive allies among European companies.

I mentioned the panelists from SAP and Ericsson. Some others of this nature and stature, such as Philips, Siemens and Alcatel Lucent, apparently weren't invited to make presentations. But the European Commission is aware of where they stand.

If US companies make demands in Europe that are different from the status quo in the US, one may ask what really dictates a different approach on this side of the pond. Oracle and HP didn't have any specifically European reason for what they proposed -- such as market characteristics. It looks to me like they are just trying to advocate here what they are unable to achieve at home.

Oracle's extrapolation and Java

I thought Oracle's Don Deutsch would comment on the Java situation. Instead of addressing it directly, he referred to Oracle's involvement with over 100 standard-setting organizations and 500 Oracle engineers participating in 600 working groups. But one thing he said raises interesting questions concerning Java.

He claimed that disagreement over FRAND (more specifically, over which terms a right holder is or is not allowed to seek after having made a FRAND commitment) resulted in "many complaints filed alleging abuses in the area." I talk about many patent-related disputes on this blog but I don't see "many complaints" of that kind. They are few and far between. Don Deutsch probably knew that many in the audience were aware of that fact, so he preemptively added that there are "probably many more complaints than we know about because they are almost always handled in a bilateral way, not discussed publicly."

That makes me wonder what all of this means for Java. There are two high-profile disputes at the moment: Oracle is suing Google and a refusal to grant a Java TCK license to the Apache Software Foundation on the terms to which the ASF believes to be entitled.

If it's true that public disputes are the exception and secret "complaints" are the norm, then the number of unreported Java disputes would have to be a lot bigger than the two conflicts that are out in the open. It would be very interesting to hear from Oracle how high that number is. Are we talking about a handful of unknown Java conflicts? A dozen? More?

If that's not the case, then Oracle shouldn't blow those controversies over FRAND out of proportion.

ICT convergence and software-specific rules

Oracle's Don Deutsch was on a panel on "ex-ante disclosures" of essential patents and most restrictive terms, an approach that doesn't rule out royalty obligations. Right holders seeking royalty payments would just have to state beforehand the maximum amount of license fees they will charge later. FRAND could still serve as a set of rules governing which terms and conditions are acceptable. Some of those demanding an ex-ante disclosure regime may, however, view it as a first step away from FRAND and toward the elimination of different restrictions (or of all of them in the long run).

In his closing statement, Don Deutsch limited his support for ex-ante to "certain market segments". He said that "it particularly fits the IT industry" and added: "We've certainly heard that the telecoms industry has a different history. The industry segments have different characteristics and therefore they have different requirements."

There appears to be little support from the broader ICT industry for proposals to depart from FRAND. That is probably the reason why Don Deutsch and some of his allies limit their demands to only a segment of the ICT field.

But Oracle did feel forced to acknowledge (without using that particular term) the convergence that is taking place. That fact, of course, is one of the reasons for which software-specific rules would not be sound policy in this particular area.

HP presented an unsuitable categorization

The panel on "open source, freely available software and standardization" was the one I was primarily interested in.

Scott K. Peterson, a Hewlett-Packard open source lawyer, started his "FOSS license analysis" with a reference to fragmentation ("it seems daunting to have to analyze open source licenses because they are so numerous, hundreds of them possibly"). But he believed that a categorization into "permissive" and "copyleft" licenses would provide clarity in connection with the impact of patent-encumbered standards on open source.

In some other contexts, the distinction between permissive and copyleft licenses is key. In this case, it misses the point.

Open source licenses are predominantly copyright licenses. A few of them have patent clauses; others don't. Copyleft can indirectly affect patent issues, but in connection with standards, the key distinction is between "patent-agnostic" licenses and those that have certain kinds of patent provisions.

To the founder of the movement, Richard Stallman, they are so different that he calls on everyone not to "lump them together" under the term of intellectual property. HP's Scott Peterson blurred that distinction completely with his unsuitable categorization.

This resulted in some errors, and one of them was particularly embarrassing in front of an EU audience: he listed the EUPL as a copyleft license and, therefore, as one that faces a particular challenge with patented standards. However, the official EUPL blog already agreed with me last month that it is possible to obtain licenses on patented standards in connection with software distributed under the EUPL. That is the case because the EUPL doesn't contain any patent provision prohibiting such inbound licensing in any way.

FSFE's reality distortion field still active

The president of the FSFE (Free Software Foundation Europe), Karsten Gerloff, spoke after HP's Scott Peterson and credited him for a "masterful introduction" to the "thorny issue of FRAND licenses, restriction-free licenses, and copyleft."

At least there's a little bit of progress related to terminology. Karsten said "restriction-free". I previously criticized the "royalty-free" movement for narrowing the issue down to only one kind of condition patent holders may impose. So it's more accurate and more honest to admit that "restriction-free" is the goal the FSFE pursues. But this means that patent holders would have to waive the entirety of their rights, and whether or not one supports the patentability of software, it's understandable that right holders will oppose such ideas.

On the question of whether free and open source software licenses can handle patent-encumbered standards, the FSFE still doesn't specify its concerns about FRAND. It still wants to make people believe that all FRAND is irreconcilable with FOSS and especially the GPL -- even though the most important GPL'd software is Linux and all of the three leading Linux distributors (Red Hat, Novell and Canonical) have licensed patents. To ignore fact, one has to operate inside a powerful reality distortion field.

HP and the FSFE were trying to advocate the same concept but gave inconsistent portrayals of the legal situation. Scott Peterson did not support the FSFE's claim that the GPL is "incompatible with FRAND licenses". Instead, he said that "even though the text of the license is not the problem", one would not get a certain "set of permissions" that is key to FOSS dynamics. He previously outlined three key "FOSS characteristics": free redistribution, ability to modify source code, redistribution of modified software on FOSS terms.

By admitting that "the text of the license is not the problem" (which he didn't say about the GPL per se, but the GPL was part of a list of licenses he discussed), Scott Peterson probably made some people in the audience wonder whether the FSFE is a reliable source of information or makes up problems only to gain a political advantage.

But even the "dynamics" HP described aren't fundamentally at odds with inbound patent licensing. Scott Peterson once again failed to distinguish between copyright and patents. The program code shared under a FOSS license can have all those benefits regardless of a need to obtain patent licenses.

The combination of HP and FSFE was interesting to watch: two speakers demonstrating two different ways to reach the same false conclusion. But they were happy that they agreed on the same anti-FRAND result. No matter why and how they arrived there.

If you'd like to be updated on patent issues affecting free software and open source, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents.