Ein Reisebericht

Edsger W. Dijkstra hat mit seiner Frau im Frühjahr und Sommer 2001 Europa bereist und darüber einen Bericht verfasst, in dem er auch über die Software-Pioniere-Konferenz schreibt:

Then, more than a month later, on 28/29 June, we had in Bonn the sd&m Conference "Software Pioneers". The acronym sd&m stands for "software design & management" and is the name of a Munich software house with (I was given to understand) about 900 employees. I quote from the brochure:
"Today's young computer scientists and software developers don't know much about the roots of their field. [...] This holds as well for some of the young sd&m software engineers.
The sd&m Conference 2001 aims at filling this lack of historical awareness by presenting ground-breaking contributions to the development of software as a field of its own, authentically presented by the pioneers themselves."
(Had the composer commissioned to write something for the occasion been Modest Moussorgsky, he would have composed "Speakers at an Exhibition".) The second reason for the sd&m extravanganza turned out to be that it marked the retirement of Prof. Dr. Ernst Denert, CEO and founding father of sd&m AG. He and Manfred Broy (from the Technical University Munich) had organized the conference, and they had done so very well. The conference (for 1200 people) was in the old building for the Bundestag in Bonn, the speakers were housed in Gästerhaus "Petersberg" in nearby Königswinter. For us this was very convenient , we could make the trip between Nuenen and Königswinter in slightly over 2 hours. Not knowing how easy the place would be to find, we left in time and arrived on Wednesday afternoon.

Gästehaus "Petersberg" was a 5-star Hotel that used to specialize in foreign heads of state. (It was the "Reagan-slept-here" type of hotel.) On the day of arrival I got stuck in the elevator, Ria got stuck the next morning and, later that day, Tony and Jill Hoare would do so together. We had there dinner twice, and breakfast three times, and the food was excellent.
For the record I reproduce the speakers, their topic and the size of their time slot. Broy, who acted as master of ceremonies, had the program under perfect control: everything happened on time.

Here is the schedule for the first day:
(15 min) Ernst Denert & Manfred Broy: Welcome and Opening 
(45 min) Ernst Denert: The pioneers' influence on sd&m
(35 min) Friedrich L. Bauer: From the stack principle to Algol
(35 min) Ole-Johan Dahl: Simula
(35 min) Niklaus Wirth: Pascal
(35 min) Fred Brooks: OS/360
(75 min) Alan Kay: Graphical User Interfaces
(45 min) Rudolf Bayer: B-Trees and Codd's Relational Data Model
(35 Min) Peter Chen: Entity / Relationship Modelling

All this was followed in the evening by a reception, a gala dinner, a (very nice) performance of classical music and a very elegant procedure in which the parting CEO handed the baton to his successor. [Sorry, the passing of the baton was on the next day.] Ria and I were highly impressed by the tastefulness with which these festivities had been organised. When for the benefit of the younger generation loud music was introduced so as to make all conversation impossible, we could fortunately return to our hotel.

The schedule the next day was as follows:
(35 min) Edsger W. Dijkstra: Structured Programming
(35 min) C.A.R. Hoare: Assertions & Program Verification
(35 min) David L. Parnas: Modularization by Information Hiding
(35 min) John Guttag: Abstract Data Types
(35 min) Michael Jackson: Jackson Structured Programming
(35 min) Tom DeMarco: Structured Analysis
(35 min) Michael Fagan: Inspections
(35 min) Barry Boehm: Software Economics
(35 min) Erich Gamma: Design Patterns
(25 min) Manfred Broy & Ernst Denert: Closing and Farewell

I did not make notes and after more than two months, most of what has been said I have forgotten, but I vividly remember Alan Kay's performance because it was so abysmal. He kept us 75 minutes waiting for his lecture to start; instead he showed us one long "demo" out of his "laptop", most of it so insipid that I blamed him for insult of audience. My impression is that, except for his own performance, he did not attend the conference. Rudolf Bayer's talk was a little bit larger than the others, but that was probably because he also wanted to do justice to the contributions by Codd, whom reasons of health had prevented from participating. He was one of the few who seems to be able to give a technical talk without suffering from "PowerPoint".Unkel Fritz, who was in great shape, described the German compiler efforts of the late 50s, a topic perhaps too big for 35 minutes. Ole-Johan, Niklaus and Brooks spoke about topics they have been identified with for a very long time; each of them coped with this problem in his own way. Ole-Johan and Niklaus aimed at a more technical audience than Brooks, but that was to be expected. I did not understand what Peter Chen was telling, but that was to be expected too, for I never understood "modelling".

Friday morning I filled the first slot (known in Marktoberdorf as "The Victory of Mind over Matter"); my message was that "Structured Programming" had been more about reasoning than about eliminating goto's, easily fitted in 35 minutes. I was followed by Tony Hoare, who gave an absolutely beautiful talk. One of my American colleagues remarked afterwards to me that Tony's (legendary) mastery of English made him "green with envy". I wish the younger generation understood the trouble Masters take to give their best: while sitting in the audience the previous day, Tony had been editing his text very carefully! I missed the talk by Dave Parnas because I had accepted to be interviewed; later I heard that this had been unfortunate as Dave had prepared to expose me to his disagreements with me. Thus an opportunity for fireworks was missed.

To John Guttag and Michael Jackson I listened with pleasure, and to Boehm, whom I don't know that well, with interest. DeMarco was a slick salesman, but his saving grace was that he told his audience so himself (and warned that slick salesmen were dangerous). I was surprised by the final attention he paid to "Extreme Programming". Fagan was a salesman too, but a less charming one and Gamma was the last speaker.

That evening (or was it the next morning?) the Munich employees returned with the specially rented train with which they had come. The speakers and their wives returned to the hotel, where we had the privilege of enjoying good food in good company. The next morning we found each other again at a leisurely breakfast, for which the tables had been set outside on a terrace overlooking the Rhine: a perfect closing for a lovely party.