Wednesday, August 4, 2010

Munich's Linux migration and concerns over software patents: six years later

The first time that software patents appeared to adversely affect free and open source software was six years ago, at this particular time of the year. News media reported on 4 August 2004 and the following days that the City of Munich had "suspended" its Linux migration project due to legal concerns.

I remember that situation well because I was -- indirectly -- involved with what was going on. Since I started this blog on FOSS patent issues, it's the first "anniversary" of the LiMux incident, and I'd like to describe what happened and why. Years later, some people became confused and also attributed another -- completely unrelated -- delay of that project to software patents. So there are some misconceptions out there, but they can be easily addressed on this occasion.

The key thing to know is that the "delay" announced on 4 August 2004 amounted to a week and a half. On 13 August 2004, Munich's mayor announced that the project was going forward as planned. It's doubtful that those 10 days made any practical difference for the project's time line, given that in Bavaria (the federal state whose capital Munich is) all of August plus the first half of September are the summer vacation period for all schools (and effectively also the slowest part of the year for business).

But those 10 days played a key role in our efforts to oppose the proposed European "directive on the patentability of computer-implemented inventions", commonly referred to as the "software patent directive" that the European Parliament threw out the following summer.

So the benefit we (the supporters of FOSS innovation and opponents of software patents) got out of this was hugely greater than whatever negligible downside.

Two years later, ZDNet reported that "the City of Munich's decision to stall its Linux migration, was the result of a PR stunt by anti-patent campaigners." I'll explain why there was much more to it than just a PR stunt, but in terms of a stalling of the migration project, we're talking about 10 days of a month in which nothing big would have happened anyway due to the vacation season. I just repeated myself because that's really important. And the confusion as to what the city administration might do lasted only a few hours, as I'll explain further below.

A creative solution to an awareness problem

I joined the movement opposing the EU software patent bill in the spring of 2004, more than two years after the European Commission's original presentation of the proposal. Prior to that presentation, there had been a hearing and all sorts of discussions. So it's fair to say the process had been going on for several years by the time I became aware and decided to take action.

The primary problem I identified was a lack of awareness by the media and, consequently, by politicians not specialized in IT and by the general public. Some really interesting things had actually happened, including a landmark European Parliament vote in September 2003 (first reading) and a resolution with which the Dutch parliament called on its government to withdraw its support for software patents. But all of that went largely unnoticed except by a few experts.

Big things going on in Brussels (the de facto capital of the EU) are often underestimated or overlooked, such as a European Commission plan to require all significant market players to provide interoperability. EU processes are lengthy and complex. "Eurocrats" use a language to describe their procedures that average people don't understand too easily.

But in my mind our issue was a really important one. The proposal that was on the table would have resulted in a major strengthening of the legal status of software patents in Europe. Tens of thousands of existing ones would have become much harder to challenge; as a result, many patent holders would likely have enforced their rights far more aggressively if the bill had been passed into a law.

I thought the economic implications of that deserved more media coverage, and a stronger interest by a wider circle of people. The activists who had been working on the issue before me had done a great job in the European Parliament in 2003 (and not only there). They just hadn't figured out how to really draw attention to what was going on.

Whether or not a political process gets attention has important implications. Generally speaking, the more attention there is, the more "populistic" the debate may get, but that's an opportunity for civil society to influence the outcome. An opaque process that is managed by a circle of specialists (including specialized politicians) is more likely to favor the interests of that professional group. In case of patents, it's obvious what the patent profession wants: more patents, stronger patents, broader patents. There are economic reasons for it; reasons related to power; and people who believe in what they do for a living may also overestimate the positive effects of their trade and turn a blind eye to its negative consequences.

If we wanted a level playing field, we had to do something. I wasn't the only one to think so. But in our movement, I was probably the one with (relatively speaking) most PR expertise, having previously managed product launches in the European market (including Warcraft II, Diablo I, Starcraft I). I guess that's why I looked at the PR question as a problem to be solved, as homework to be done.

The Greens got the ball rolling

I can't take credit (nor blame) for the stratagem that ultimately worked out so well. There was an IT consultant who played an active role in the Munich chapter of the Green Party, Michael Fritsch. He knew one of the Green aldermen (city councillors) of Munich very well: Jens Muehlhaus. Michael was also in contact, via a business network, with Jan Wildeboer, then the lead programmer of a company in the web domain business and now a hybrid between a lobbyist and an evangelist at Red Hat.

I knew Jan before, and met him, Michael and Jens at a so-called "kick-off event" for the LiMux project. The city administration tried to meet members of the FOSS community for a little conference and get-together in July 2004.

The Greens played an important role in this. They've always been very FOSS-friendly, and during the fight against software patents no other party in the European Parliament cooperated so closely with our movement (although others also lent important support, and in terms of gathering votes, some MEPs from the large political groups were absolutely key). The Munich chapter of the Green Party had also been at the forefront of the push for the use of free and open source software by the city administration.

The concern was already there

The aforementioned stratagem -- to raise the issue of software patents in connection with the Munich Linux project -- was a Green idea, and the Greens made it fly. Before I talk about the "implementation" of the idea, let me mention that Wilhelm Hoegner, then the head of the data processing office of the city administration of Munich (sort of the city's "CIO") and a member of the Social Democratic Party, was already profoundly concerned about software patents. After his presentation at the event, someone asked Mr. Hoegner whether the plan to switch to open source solutions was still going to work out even in the event that Europe legalizes software patents.

Hoegner's uneasiness over software patents was visible even before the questioner finished speaking. First, Hoegner pointed out that he didn't have a profound understanding of the legal issue itself. Then he confirmed that it would be "indispensable" to analyze the possible effect of the proposed EU software patent directive on open source, and he said that any related "mistake" would be "a catastrophe for Munich's Linux migration project, and for open source in general".

So the concern was there even prior to our stratagem. It was only after that meeting and the statement I quoted that Jens (the alderman), Michael (the IT consultant and Green Party activist), Jan (an activist who now works for Red Hat) and I had a conversation, and Michael presented his plan. He proposed that the Green Party in the city council should pose an official question about the potential impact of the proposed European software patent law on the migration project.

His thinking was that Munich's decision to turn down a bid from Microsoft and switch to Linux (which wasn't based on a short-term cost calculation but a very political decision for the long haul) had already received a great deal of media attention, so it could serve as a vehicle to generate awareness for software patents. In fact, Michael said the objective should be for the EU software patent bill to receive a similar level of media coverage as the LiMux project. That was ambitious. LiMux had been discussed in the general press around the globe, while the EU software patent bill had received a very limited degree of coverage by a very few IT-specialized websites.

Jens (the alderman) wanted to proceed with caution, but he understood the idea. He was willing to propose it to his group in the city council on a conference call, and he felt that it could be done in a way that wouldn't call into question the Green Party's commitment to the LiMux project. None of us -- Jens, Michael, Jan and I -- wanted to harm that project in the slightest. We felt that it was indeed endangered by the legislative proposal in question, but we wanted it to happen one way or the other. At the same time, we wanted to seize this opportunity to achieve a PR breakthrough for the software patent problem.

Michael insisted we make a list of existing European patents that could, at least
potentially, be enforced against the city of Munich for using Linux. He wanted to lay the emphasis on tens of thousands of existing European software patents that the directive in question would make more easily enforceable. If the directive provided a strong legal basis for those patents, the patent holders might decide to instigate a flood of lawsuits in order to derive economic value from their rights.

The FFII, the non-governmental organization that started the resistance against software patents in Europe, took care of this. Holger Blasum, at the time an FFII core activist, was also at that LiMux meeting and agreed to take care of the required initial patent analysis.

The implementation: no immediate breakthrough

That meeting was on a Friday afternoon, and over the weekend Michael drafted two written questions for the Greens to send to the mayor. My contribution to that drafting process was limited. I provided a few paragraphs I thought they could use as a basis for a brief description of the software patent issue. That's nothing unusual for outsiders to do. In any parliament that kind of input is provided all the time by people who are familiar with the details of a subject.

Since this issue is so special, the Greens appreciated my offer to handle the PR part of it: sending information to the press. I wasn't acting as a spokesman for their party, but I made myself available to journalists who had questions.

There were severe time constraints. After the late July LiMux meeting, there was going to be only one more week before the vacation season. If there had been any delays, this would have had to wait until late September.

The Green Party approved the plan; the questions were formally presented to Munich's mayor; and I issued a press release on 30 July 2004.

But very disappointingly, there was much less media coverage than I had hoped. The bomb didn't go off properly. The mere fact that the Green Party posed a couple of critical questions reflecting abstract concerns just wasn't enough to have major impact. A statement by the mayor could have helped, but at that point it wasn't likely to happen until after the vacation season, if at all.

The real thing

Unexpectedly, there was a dramatic turn a few days later. As part of its effort to communicate with the FOSS community, the city administration had joined a public mailing list. On 3 August 2004, Mr. Hoegner (the city administration's "CIO") posted the following message to that list (my translation):
Subject: [LiMux] Motion/question of Greens in city council concerning patent law
Date: Tue, 3 Aug 2004 19:26:42 UT
From: Wilhelm Hoegner

Hello, dear participants in this project,

You have certainly read in the specialized Internet media that the Greens of Munich have introduced a motion and a question in the city council with respect to software patents.

Due to the questions that have to be checked into, the bidding process for the LiMux base client, which had originally been slated for late July, has been stopped. The [city] administration will first try to estimate the legal and financial risks involved before this can continue.


Wilhelm Hoegner
Now that was more concrete than just two abstract questions posed by a group of aldermen. This was an explosive email. I discussed it that same evening with Jan Wildeboer. He had made me aware of it, and I agreed that it was an important development. In my opinion, that decision by the city administration was going to have repercussions one way or the other, so I wanted to take the bull by the horns and ensure that this would be put into the political context of the EU software patent bill in a way that non-specialized audiences (including non-specialized journalists) would understand.

There was a risk that it could have reflected negatively on the Linux migration project without much of a positive effect for our resistance efforts to that EU bill. I wanted to minimize the downside for LiMux and maximize the upside for us.

With that in mind, I put out a press release on the morning of 4 August 2004 via a German-language news wire service. Its headline (my translation): "EU software patents: Munich puts Linux project on ice for now, EU-level politics of federal government share responsibility"

So this wasn't at all about claiming that the project was going to be stopped. It accurately reflected the temporary stalling that Mr. Hoegner's email had announced:
Yesterday evening, the chief information officer of the city of Munich, Wihelm Hoegner, announced on a mailing list that the bidding process for the "LiMux Base Client", which had originally been slated for late July, cannot commence for the time being. After an advisory from the Greens, the city administration firstly has to analyze the legal and financial risk.
The press release then went on to explain why software patents are a threat to FOSS as well as SMEs (small and medium-sized enterprises).

Only 25 minutes later, the first report appeared on a large Web site. It became "the Day of the Ringing Phone" for me, with countless questions from reporters.

The mayor's statement -- and subsequent activism

In mid-afternoon, Christian Ude, the mayor of Munich, made his official statement. He was personally very committed to the LiMux project, and we all knew that. It was a prestige project to him, sort of the digital equivalent to the annual Oktoberfest opening ceremony. His party, the Social Democrat Party, was also completely in favor of the migration project for political reasons.

What we didn't know, however, was the mayor's stance on software patents. In order to defend the migration project, he could have downplayed the impact of the patent bill. That would basically have served the interests of the proponents of software patents. Fortunately for us, he walked a fine line and decried the software patent directive, yet made it clear that the city administration would stand by its strategic decision in favor of Linux despite the patent concerns, and continue the project after an assessment of the risk.

So if there had been any doubt (even though my press release had made it perfectly clear that the project was only stalled temporarily), the mayor dispelled it just hours later. And another nine days later, the city administration then announced, again, the continuation of the project. Since this was during the vacation season, it likely didn't delay the actual migration by even one day.

But we got a lot of mileage out of this. Subsequently, everything related to the software patent process got a lot more media coverage. Some general news media took an interest in this, and IT media were interested in reporting more frequently and in more detail than before.

The Munich city administration put together a position paper and demanded a rewrite of the legislative proposal. The mayor, Mr. Ude, repeatedly voiced his opposition to the original bill. He did so in public, and we know he made a number of phone calls to German politicians to influence the process. In other words, he became a fellow activist, although he obviously didn't coordinate his activism with ours.

We saw the lasting impact of the LiMux story all throughout the further process. It was mentioned over and over. The impact of this was particularly strong in Germany, but there were significant effects throughout and beyond Europe.

Too real to be considered a stunt

To what extent was it a PR stunt? Not very much. The two questions the Greens posed to the mayor were part of a PR plan, but the Greens had a genuine concern and they had come up with the idea all by themselves; I just lent limited help, and so did the FFII and Jan Wildeboer. Mr. Hoegner's email that announced the suspension of the bidding process was no stunt at all, and my press release had accurately reflected the information. All I did was to put it into the context of the ongoing legislative process because most people didn't know too much about it and needed explanations. Even without me, the stalled bidding process would have made headlines.

There was never a risk because the mayor and his party, and his coalition partner in the city council (the Greens), were 100% committed to the Linux migration project.

More than a year later the LiMux project announced a delay, but it was for purely technical and administrative reasons, not due to software patents.

We ultimately defeated the EU software patent bill, and after all those years there's still no indication anyone ever tried to extort the Munich city administration with patents -- imagine what a troll could do with an injunction shutting down an entire city administration! So this worked out very well for FOSS.

If you're interested in more detail, I've uploaded again my e-book, "No Lobbyists As Such: The War over Software Patents in the European Union" (the LiMux story starts on page 130), and its German version, "Die Lobbyschlacht um Softwarepatente", to

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.