Last week the Competition Directorate of the European Commission and Microsoft again did battle in the European courts. This time it was a hearing in the European Court of Justice to argue the propriety of the fine levied on Microsoft by the Commission stemming from the 2004 determination that Microsoft was violating EU competition laws. Microsoft was appealing the $1.3 billion fine on the grounds that it was excessive and had not been arrived at through proper due process.
In 1998 Sun Microsystems lodged a complaint with the European Commission asserting, among other things, that Microsoft was abusing its dominant position (what those of us from the U.S. would think of as using illegal means to maintain a monopoly) by refusing to disclose necessary interoperability information pertaining to Microsoft's desktop and related workgroup server technology. After extensive investigation, in 2004 the Commission issued a
decision in the case and found Microsoft to have violated competition law. Among other remedies, Microsoft was ordered to make its workgroup server protocols available to competitors on a reasonable and non-discriminatory (RAND) basis and was fined approximately $800 million. Microsoft sought annulment of the Commission decision and fine, but the Commission was upheld by the Court of First Instance in 2007.
Almost immediately following the 2004 decision the Commission and Microsoft began arguing over what constituted adequate compliance with the protocol disclosure requirements of the decision, and in November 2005 the Commission
levied a further periodic penalty of $2.5 million per day on Microsoft for its continued non-compliance. When the fine was first levied in July 2007 it totaled approximately $350 million. Much of the dispute had to do with the licensing practices Microsoft was employing with respect to the protocols under its Workgroup Server Protocol Program (WSPP). [If you would like to know more about the WSPP, here is an article on the history of the program and another on the WSPP agreement itself, both written by Andrew Tridgell of the Samba Project.] The parties continued to argue over compliance, and in February 2008 the Commission fixed the final amount of the penalty at $1.3 billion for continuing non-compliance. It is this fine that Microsoft appealed to the European Court of Justice and which was the basis of last week's hearing.
May 24, 2011 Hearing Before the European Court of Justice
The European Court of Justice met on Tuesday, May 24, 2011, to hear the arguments of Microsoft and the Commission over the propriety of the substantial fine levied against Microsoft. The President of the Chamber (Court) was the distinguished justice James Forwood of Britain. The other two members of the panel were Franklin Dehousse of Belgium and Juraj Schwarcz of Slovakia. Arguing the case on behalf of the Commission was Nicholas Kahn of the Legal Services Directorate. Jean Francois Bellis of the Brussels office of Van Bael & Bellis argued on behalf of Microsoft.
Providing key testimony on behalf of the Commission was Andrew Tridgell of the Samba project, represented by well-known FOSS attorney Carlo Piana. Mr. Piana also represented the Free Software Foundation Europe in the proceeding, and Karsten Gerloff, president of the FSFE was present. Groklaw caught up with Messrs. Tridgell, Piana and Gerloff, and interviewed them about the proceeding. Following is a summary of what I learned from them:
What were the critical arguments addressed in the hearing, and how did they come across?
There were two different arguments addressed in the oral arguments. The first was that the Commission did not apply due process in imposing the fine. Microsoft argued that it was never clear exactly what the Commission wanted, that Microsoft was trying to hit a moving target and the Commission did not provide enough cooperation. Further, from Microsoft's perspective the Commission sat on its hands without responding promptly to proposals from Microsoft. In the end Microsoft espoused that the Commission had effectively tried to extort unjustifiably low royalty rates for the protocol licenses. The sense of our observers was that the Commission response to this argument was very effective.
The second argument addressed was that Microsoft, in good faith, should have known that information for which they were seeking royalties was not innovative, at least under patent law. Microsoft asserted certain patents would be necessarily infringed by implementation of the protocols, and the Commission did not challenge the right of Microsoft to charge royalties for licensing the patents. But a significant portion of the protocols were not covered by patents, and a number of the protocols contained nothing innovative that could even be protected by trade secret. Microsoft's initial licensing approach was all or nothing. In other words, you had to license all of the protocols, both those covered by patents and those not covered by patents, under a single license and royalty rate.
The Commission insisted that Microsoft unbundle the protocols into those covered by patents and those not, permitting a licensee to avoid the patents altogether if they should so choose. Microsoft argued that some of the patents could not be avoided. When Microsoft finally unbundled, they still required a patent license as a prerequisite to obtaining licenses for those protocols not covered by patents. These repeated instances of incremental concessions by Microsoft caused the whole process to be prolonged (recall that this all stemmed from a 2004 decision). Eventually, Microsoft was forced to offer a trade secret-only license for those protocols not covered by patents, but this did not happen until 2007. Even then it was not clear that a license to any patents was necessary to make an implementation that would comply with the protocols allegedly covered by patents because most of the Microsoft patents only pertained to their implementation, not to the standard (the architecture) itself.
One of the Commission’s strongest arguments was that Microsoft could not extract value (charge a royalty) for strategic value, only for innovative value. In other words, to the extent Microsoft designed or controlled protocols to achieve some strategic goal, they could not charge for those protocols unless they actually embodied some degree of innovation even if this innovation was only a trade secret. Further, if the protocol contained no trade secret and did not infringe any Microsoft patent, Microsoft could not charge for the protocol at all.
Ultimately, the definition of innovation is central to this case. Whether Microsoft is to be permitted to charge running royalties depends a great deal on whether the information in the protocol documentation is innovative. The key to the definition is to ask "innovative for whom and when?" Microsoft asked the court to take the view that what matters is whether the information would have been innovative in 1998. However, when you are assessing the value for potential licensees in 2007 under the Microsoft Workgroup Server Protocol Program (WSPP), the key is whether it is innovative for the recipients of the licensing information at that time in late 2007, when the license was offered. By 2007 any pieces of Active Directory which could have been seen as novel had been published (by Microsoft and others) a long time ago. The only information that wasn't yet public was the low level encoding information of how the messages were formatted on the wire. The comments from Judge Forwood showed that the court understood the significance of this.
Another key point in the hearing had to do with comparable licensing arrangements. Microsoft argued that no one releases the sort of protocol information they were being asked to release at such a low cost. As an example they pointed to ISO and IEEE RAND licenses. FSFE/Samba refuted this, arguing that releasing interoperability information at low or no cost is, in fact, the norm, not the exception.
Also on the topic of licenses is the interpretation of what the term “reasonable and non-discriminatory” means. Often RAND is seen by the free and open source community as being something that is bad, as RAND is usually interpreted as meaning that per-copy royalties will be charged. One of the most significant outcomes of this hearing is that the Commission and the court took the 'non-discriminatory' part of RAND as meaning that it must allow for use by FOSS projects released under the GNU GPL. This result came about because Samba based solutions are really the only significant
remaining competitor for Microsoft in the work group server market. So it would have been fairly meaningless for 'non-discriminatory' to mean
a license that excluded Samba.
Interestingly, in its closing remarks Microsoft never mentioned the value of its technology.
The sole intervenor for Microsoft was ACT, and their testimony went to the issue of comparables. But as was pointed out in rebuttal, all of the instances ACT cited were tied to patents which was not true in the instant case. The final remarks for the intervenors supporting the Commission were delivered by ECIS's Lawyer, Thomas Vinje, and he pointed out that all the comparables cited by ACT also only charge a nominal royalty.
What was the proceeding itself like?
Formal. Chamber President (Judge) Forwood (Britain) ran the session and did almost all of the speaking from the bench. He is a very experienced presiding judge and was well read into the case. Forwood is to be complimented for his understanding of both the technical issues and the due process issues of this case. The other two judges were Dehousse (Belgium) and Schwarcz (Slovakia). Forwood's term ends this year, and some have speculated that he wanted to wrap up this case while he was still sitting.
It was the sense of our contributors that Microsoft had more trouble responding to the judge's questions in this proceeding than in past proceedings. Also, they observed that most of the decision had likely already been reached by the panel of judges before the hearing. The primary purpose of the hearing was to test the strength of arguments that had already been presented. Given that so much of the process occurs before the oral argument, it is impressive that the Commission lawyers have taught themselves so much about the technical side of this case, which gets pretty complex at times.
What is the secret to Tridge's effectiveness in the courtroom? [Asked of Carlo and Karsten]
It's a gift. First, he has a deep understanding of the technology. More importantly, he has the ability to explain it in simple terms. This likely doesn't happen by chance but from preparation, preparation, preparation. And it is not just Tridge preparing. Carlo, Jeremy Allison and Volker Lendecke all made significant contributions to drafting the testimony that Tridge provided. When presenting, Tridge does so with confidence, but also with a sense of humor. This time, compared to previous testimony before the Court of First Instance and/or the European Commission, Tridge's testimony was much lower key.
The decision in this hearing, no matter what it is, does not relieve Microsoft of its obligations under the 2004 decision. That decision will remain in effect until the Commission decides that competition has been fully restored.
It is worth noting that competition suits such as this are a powerful tool in changing behavior, but they grind incredibly slowly.
Here are some links to the FSFE website where they provide additional information about the hearing: