There are probably not many German computer scientists who have done so much for the expansion of the Internet in Europe and know as many details of the protocol wars as Daniel Karrenberg. Because there was so much to do there, he first moved to the Netherlands for a year in the 1980s. Because the German political climate was sometimes too dogmatic for him, he stayed there and still sees failures in the creation of an Internet market on this side of the Atlantic: In our series we speak with the RIPE architect Daniel Karrenberg.
What is missing: In the fast-paced world of technology, there is often the time to rearrange the many news and backgrounds. At the weekend we want to take it, follow the side paths away from the current, try different perspectives and make nuances audible.
More on the “Missing Link” feature section
This is another episode of our conversation series with pioneers of the German Internet:
As the first European and the second engineer ever, Daniel Karrenberg received the Jon Postel Award, probably the highest honor for the master builder of the Internet. Karrenberg ported Unix to an early Apple, brought Usenet to Europe and built the Eunet before it became a provider. He is best known as one of the architects of the first local IP address allocation office in the world, the RIPE, with dedicated views. In an interview, he cites the main fears that Europe will again fail to create its own market – this time for the cloud, and that at some point he could become too stubborn.
heise online: In the laudation of the Jon Postel Award, which you received in 2001, it said “without Daniel Karrenberg’s Work the internet would be a different place today”. When you look back, what do you think is your most important contribution?
Daniel Karrenberg: You will fall into the house with the barn doors. I am someone who does not believe so much in the outstanding achievements of individuals. I tend to believe in processes that involve different people and that you can contribute to. Sometimes it works better and sometimes less. You have to see it every decade. My first contribution was to help the Internet protocol break through together with others. That was in a phase when it was not yet clear whether this internet fuss would be useful and also the right way in the long term. Like many colleagues, I leaned far out of the window. That was in the 80s. In the 90s I got involved in the Genesis of RIPE …
heise online: the European IP address registry …
Daniel Karrenberg: Yes, it started with an exchange about which routers worked, which modems had to be plugged in and how to order international connection lines. I contributed my little stone to it. The RIPE NCC was founded at the same time. That was the 90s. In the 2000s, I then threw myself more into research on the Internet. Not on a political-economic level, but on a purely empirical measurement level.
heise online: Elder Statesman….
Daniel Karrenberg: Some say gray eminence (laughs).
heise online: Why did you study computer science, that was still fairly new at the time?
Daniel Karrenberg: That’s right. I can still remember the reactions from people to whom I told that I was planning to study computer science. The most common question was, ah, IT, and then do you want to go to the newspaper or the radio? (laughs) Or to watch TV, of course that was the non-plus-ultra. My actual reason was that I did physics and math in school, but also a few other things. On the one hand, I still took pedagogy to compensate, and because I didn’t feel fully occupied yet, I voluntarily added a history course on top of it. I found the great thing about it, there were people who wanted to do it, they were interested. It was the same with Russian, which I also learned on the side. To this day I am someone who would like to be more broadly based.
Dortmund instead of Silicon Valley, Doc Peter instead of Ballmer or Gates
heise online: What attracted you to computer science?
Daniel Karrenberg: I thought IT – or data processing as I understood it – was good because there is always an application. You always have to deal with users and with a use case. That seemed varied to me. If I do physics, I thought, my life will only be varied if I keep changing jobs. I didn’t like mathematics because of the job description, because I didn’t want to calculate death rates in an insurance company. Science wasn’t my goal at first either. I also had some programming experience because I was lucky that we had a dedicated IT person from Feldmühle AG at the school. He taught us Fortran in the afternoon. Friends of mine in Bochum were able to use a basic computer at their school that was connected to the city’s data center. You can’t say that anymore today, but then we also printed out the residents’ registration file. There wasn’t much with access protection (laughs), no one had any pupils playing on the screen.
heise online: There weren’t any computer games either, right?
Daniel Karrenberg: No. Computer games were programming and just such jokes as the registration office file. At that time, data processing was already moving into companies like Opel in Bochum, which I heard from my father-in-law. There were great fears and therefore a need for explanation. That’s when I saw the connection to pedagogy. It looked like a holistic thing to me. Because friends from Bochum went to Dortmund to study computer science a year before I did, I took a look at Dortmund, Aachen and Bonn. I thought Dortmund was great because the department was just being set up. The dean, Volker Claus, ran around and said, oh, Mr. Karrenberg, there are the assistants back there who will be doing their freshmen next year, why don’t you go to the cafeteria with them. From then on, the matter was clear.
heise online: During your time in Dortmund you co-founded EUnet, one of the first two providers in Germany. Was its real origin really a European Usenet?
Daniel Karrenberg: Yes, that’s true. In retrospect, it all always seems so logical – but I got there like a virgin to a child. I studied my first semester in the winter of 1978/79 and also worked in the university computer center. I didn’t want to be a mess for my parents and had already applied for the job in advance. They were looking for someone who knew about the Pandas software for IBM mainframes on the notice board and I happened to be familiar with it. So I trotted over there, presented the certificate from my computer science group and said, by the way, I can do it. My job was to advise the users of this software package. Immediately after the first semester, I had to do community service after I had to go to the administrative court to check my conscience. During the civil service in the community hospital in Herdecke, I was then back in the IT department. It was funny, it was an anthroposophical hospital that took such devices as the devil’s stuff. But they had a mini-computer. When I returned to university two years later, I already knew that mainframes weren’t for me.
heise online: So you practically infected yourself with Unix among the anthroposophists …
Daniel Karrenberg: Exactly. That’s why I’m no longer back to the data center, but to the IT computer operating group at the University of Dortmund. Then I had to look after a mini-computer again and Unix ran on it, namely at night. During the day for the ‘real work’ the DEC (Digital Equipment Corporation, d. Red) operating system was running. The computer was a PDP 11/60. At night we – we were a group of student assistants – were allowed to do our experiments with Unix version 6. In the daytime research project, microcode was supposed to be written and the participants kept complaining that the compiler was so bad. We looked at it and it quickly became clear that it wasn’t commercial code, a student had done that over the summer.
Together with a few colleagues, including Klaus Eckhoff and the later Daimler researcher Stefan Hahn, we took on the problem and decided to make a microcode compiler that runs under Unix. The effect was that Unix no longer only ran at night …