Normally one might have used the phrase "the gloves are coming off" to describe the what happened on the last two days in court, but that would have implied the gloves were somehow on the days before. So it is probably better described as: putting on the brass knuckles.

The morning of day four started with a very strong presentation by the interveners, including Andrew Tridgell, president and founder of the Samba Team, who spoke on behalf of the Free Software Foundation Europe, you might already have seen some media echo. He essentially popped the blue bubble filling the heads of the Microsoft representatives and (occasionally) the screens.

The theory of the blue bubble is that some network protocols are more equal than others. Microsoft described them as "closely coupled" before in the interim measures hearing, and now to all Microsoft Active Directory server to server communication. Indeed, the bubble of Microsoft claims to "super protocols" has been shrinking over time and while shrinking, the bubble turned blue.

They claimed that no other vendor could provide "intra-bubble functionality" because identical logic on all hosts would be the prerequisite for their multi-master-replication. Tridge popped that bubble easily, also clearing up a couple of other misinformations Microsoft was so kind to provide to the court.

Later in the day, Microsoft's distinguished engineer, the infamous Mr Shewchuck tried to go mano a mano with Free Software's competent engineer Mr Tridgell, and didn't look too good.

Microsoft then (involuntarily) managed to make one of the best cases against software patents I've witnessed in a while -- essentially proving that software patents are mainly good to build and enforce cartels and that they are fundamentally contradictory to interoperability, which could only be achieved through compulsory licensing. Unfortunately most people in the room may not have realised this as clearly.

Day five then began with the final pleadings by Microsoft and its interveners, followed by the Commission. A clear victory by points for the Commission in my eyes. Funnily enough, Mr Shewchuck then tried to get back at Tridge for having disassembled him the day before, and failed miserably: Wanting to portray Samba as much more powerful than it is and implying the destructive nature of Samba, he referenced to a recent presentation by Tridge in New Zealand about the "Samba Vampire."

Tridge told him to be more careful with secondary information he finds on the internet and briefly showed a couple of slides from that presentation that did not leave a very good impression of the competence of Microsoft distinguished engineers.

You can see, it was getting hairier. Indeed, Microsoft gave a new meaning to the word obnoxious. They were constantly trying to attack anyone and everyone, trying to throw sand into the wheels of the Court and the eyes of everying present until it became very hard to bear.

While it is far outside the scope of this blog and my capabilities to summarise 16hrs of intense legal battle, let it suffice that Microsoft representatives were repeatedly making outrageously false statements to the Court while maintaining a straight face.

If you ask me for my overall impression, I would say that the European Commission and its interveners did well. There are no guarantees and how the judges will decide is exclusively their knowledge, but I am positive that all of us on the Commission side presented a strong case.

There is hope that Microsoft will not get away with taking public protocols, putting them into a blue bubble of secrecy, and using this in combination with their desktop monopoly as an instrument to take over all of information technology.

My heartfelt thanks go to Carlo Piana, FSFE's lawyer on the case, Andrew Tridgell of the Samba Team, who came all the way from Australia to be our top-expert, and everyone else who participated in the interveners team on the side of the Commission.

Todays opening of the second part of the Microsoft antitrust case at the European Court in Luxembourg made a furious start with several hours of Microsoft presentation. Mr Forrester opened with the heartbreaking example of an Irish TV guide called McGill, which was the first to list the program of an entire week across several channels, and which was crushed by the monopolistic tv channels. To me it sounded a little bit like Microsoft crushing Netscape after discovering that their proprietary internet approach did not work.

Microsoft then brought in John Shewchuck, who went into a very long Powerpoint presentation about how Directory Servers were only 2% of the servers typically used in companies, and thus not so important. He also went into a long and vague description of the general functioning principle of clients and servers, which had little to do with the actual substance of the case, but made it seem like Samba was the functional equal to a Microsoft Server.

He also explained that "It is hard to sell products unless you can ensure interoperability" -- go figure. Indeed that is precisely what Microsoft competitors criticise: Microsoft likes to adhere to interoperability as long as it feels it can take over market share from competitors, to then modify the protocols to lock that market away from competition. This is essentially the substance of the entire case.

Mr Shewchuck also explained that the reason for the Microsoft Magic was that one server knew what kind of response and behaviour from another. This is what creates the "Service Boundary" that is defining where the "Tightly Coupled Protocols." Only with some self-restraint is one able to curb cynicism and avoid the statement that we should be glad Microsoft did not declare the entire internet its service boundary: After all, knowing what kind of input brings which kind of response, and how servers should respond to certain inputs, is the basic principle of any protocol.

Microsoft has no idea what their software does

When Mr Forrester took back over, statements became outright hilarious. He explained that after the decision, Microsoft had worked extremely hard to fulfill the terms: According to him, Microsoft had 210 developers working on studying over 10 million lines of code. They also "chased down" retired engineers to find out what they had done and why. In short: Microsoft has no clue of the code and design decisions in its systems. They cannot tell people about protocols, because they don't have that information themselves.

Overall, Mr Forrester claims that more than 35 thousand working hours were spent to create 12560 pages of documentation, expanding continuously. From a software engineering point of view, this is worse than anyone would ever have dared to joke about.

We will sue you

As Mr Forrester went on to explain, this documentation was then put into a kind of digital encyclopedia, which is sorted by a method that Microsoft holds a software patent on in both the European Union and the United States (from memory: us patent no 5,968,211). Without a software patent license, the documentation of their protocols cannot be used -- because Microsoft will sue anyone who dares to do so without license.

Software patents indeed played a central role in the closing remarks. Apparently it is Microsofts proof of innovation in this field to hold 4 granted and 3 applications for software patents in the EU, as well as 27 granted and 12 filed elsewhere. He also made another thing plainly clear: "I doubt that each of these could be engineered around!"

Translation: You may be able to force us to document what we do, but we will make sure you cannot read it without being potentially liable to software patent infringement. And even if you figure out how it works, we may try sueing you for software patent infringement in case you dare to interoperate with us without our permission.

Migration to Microsoft, anyone?

The intervening parties concluded with Centrify, a March 2004 United States startup that apparently wrote "Anti-Samba": Its main functionality appears to be integrating other clients into Windows, providing a migration path into Windows. Indeed, they demonstrated live how to integrate a Red Hat client into a Microsoft Active Directory Server Domain -- and of course Red Hat was running text console only, to imply that only Windows is nice and user-friendly.

Lunch, at last

Finally we heard about how the European Commission ruling was in conflict with the TRIPS treaty, as it was an alledged case of compulsory licensing. This once more demonstrates the connection of "intellectual property" as antithesis to competition and the spread of these monopolies being a direct reaction to antitrust laws, as explained in "Information Feudalism" by Peter Drahos and Braithwaite. I recommend reading it.