EWD 850

Trip report E.W.Dijkstra, Helsinki, 1 - 4 March 1983

On Tuesday 1 March I left Nuenen at 7:15 and arrived in Helsinki at 15:35; on Friday 4 March I left Helsinki at 13:40 and arrived in Nuenen at 18:45. The flights (Finnair and KLM respectively) were dead on time.

I was the guest of OY Paragon AB, the Burroughs representative in Finland. It was primarily a Burroughs trip. I had a busy schedule. They had arranged

that I should give a lecture on Wednesday afternoon as part of their release of LINC —in today's jargon "a commercial applications generator"—
that I should deliver on Thursday morning the keynote address at ADP 83 —"atk 83" in Finnish— , the yearly data-processing conference in Finland
that I should meet the Finnish computer press on Friday morning. Furthermore they had included
a visit to the University of Helsinki on Wednesday from 11:00 to 14:45 (this visit had been included on my request)
a visit to the Technological University on Thursday afternoon (at the request of Heikki Saikkonen).

On Tuesday evening —after a sauna of 109 degrees C.— I had dinner with Paragon; we started to prepare the LINC release, a preparation which continued next morning. On Wednesday evening the organizing committee of ADP 83 offered a dinner to itself and its foreign guests. On Thursday evening I had to attend the official Conference Dinner. On Wednesday I lunched with people from the University of Helsinki, on Thursday I had to attend at ADP 83 a lunch with the press (in general), on Friday I had lunch with Mr.Runge Karlsson, Chief Operating Officer of Paragon. The Finns took very good care of me, but Finnish hospitality seems such that one had better eat only half of the food one is offered.

Finnish audiences are difficult to work with. I had been warned that they don't respond while you are speaking, which is a pity because, for lack of feedback, your decisions about speed and degree of explicitness of your presentations have to be pure guesswork. After my lecture at the University of Helsinki several people expressed their admiration for the fluency with which I had done all the formal manipulations on the blackboard; my interpretation of that compliment was that I had been a bit fast.

My 20-minute talk on the subject of "The development of programming languages" at the LINC release was for me the most difficult one, since both the audience —about one hundred data-processing managers— and the setting —a sales presentation— were unfamiliar for me. I confined myself to the intersection of what I really believed and what seemed appropriate under the circumstances. I think I succeeded well: during the following cocktail party both organizers and participants expressed their appreciation.

In degree of difficulty it was followed by the one-hour keynote address at ADP 83 under the title "Tacit assumptions that hamper software development", because in the audience (of about 250) the scientists were an almost negligible minority. It was the main reason for my invitation, which I had accepted because I have to give a similar keynote address for a similar audience (only much bigger) next June in Barcelona. It went very well, despite the fact that I had prepared visuals. (Typically my financial foil attracted most of the attention; its text had been:

"A main reason why software is so expensive is that programming is done by cheap labour.
Another reason is that existing programs are considered valuable because their development has been so expensive."

Later they explained to me that I had said something they all agreed with but for which the independence of the scientist was needed to say it so clearly.)

At the University of Helsinki I spoke about "Predicate calculus and its role in Computing science" and at the Technological University about "The detection of termination of distributed computations". Both times I had an audience of about one hundred people and the lectures went very well. At the University of Helsinki I have devoted the first hour of my talk to Yossi Shiloach's Algorithm: it is always a fascinating experience to need at least a full hour for the explanation of a 5-line program solving efficiently an easily stated problem and manipulating no more than three counters! After that very convincing demonstration that a program can indeed be a very compact deposit of our intellectual labours I talked in more general terms about the role of the predicate calculus in computing. I regretted that I could not spend more time at the University for I would have liked to get better acquainted with some of the people I met. Fortunately I met some of them the next day after my lecture at the Technological University.

Other foreign speakers at ADP 83 were Anthony I.Wasserman (UCSF, Medical Information Science) and Niels Bjørn-Anderson (Copenhagen Business School). From the latter's abstract I quote

"Almost all efforts by manufacturers are aimed at making the system "user friendly", by lowering the requirements on the user. As a secretary remarked "it is as if the systems with intelligent terminals are designed on the assumption that there is no intelligence in front of the VDU"."

(I only saw this after Heikki Saikkonen and I had already decided on a joint project, viz. the education of terminal-friendly end users.)

The highlight of my visit was the interview by the Finnish computer press. It was conducted by Pertti Jotuni (?), the doyen of Finland's science journalists. He was a delightful, knowledgeable man, who did not ask a single stupid question. At his request we discussed not only the influence of mathematics on computing, but also the influence in other direction. (Every book I quoted from he knew!)

*         *         *

I had forgotten to take a pair of solid shoes with me, but to my relief (and the amazement of my hosts) I was quite comfortable with my sandals in their -16 degrees C. When I came home, the Dutch spring had broken loose and Arthur Koestler has died.


Plataanstraat 5
The Netherlands
6 March 1983
prof.dr.Edsger W.Dijkstra
Burroughs Research Fellow