Banquet Speech at WG 2.3 March 3, 2010 Lachen, Switzerland I am honored to be delivering the banquet speech at meeting number 50 of WG 2.3. This is a special honor because I will be speaking about a colleague, friend and scientist who did much to set the tone for WG 2.3, Edsger W. Dijkstra. Dijkstra passed away nearly 8 years ago; the membership and the observer list of WG 2.3 has changed much since then, and it is appropriate to reiterate his message. Dijkstra had many loves, chief among them his wife Ria Dijkstra whose company we are fortunate to have this week. He cared deeply about doing important things perfectly, such as a desire to dress appropriately. Apparently, for a Texas professor, the appropriate attire is a Stetson hat, a Bolo tie, and the foot ware, Birkenstock sandals, a proper answer to cowboy boots. He tried to create the the perfect blend of fountain pen ink by mixing inks from different vendors in various proportions to achieve the right color (brilliant black), checking the resistance of the ink to fading under intense sunlight (by taping his writings to his office window under Texas sun), and determining how best to fill up a fountain pen for an airline flight (either completely empty or completely full). He felt that being a speaker at WG 2.3 was an honor and a privilege, and the audience has the duty to remind the speaker of his or her obligation. This meant frequent interruptions, a series of questions until a point was clarified, or even the pronunciation of an obscure word until it is perfectly delivered. Most of us, being expert speakers, know how to deflect uncomfortable questions from a pesky audience member, such as "I will have to take it up offline with you", or "let me take the next question from another source". Not so in WG 2.3 meetings; the pesky member would often be Dijkstra and he was not so easily deflected. I owe the following anecdote to Michael Jackson. Once upon a time there was an observer from a computer company who started giving a sales pitch for a product. After two minutes, Edsger politely indicated that he wanted to ask a question. His question was: "Is it your belief that we are all idiots?". The talk ended very soon afterwards, with the speaker nearly in tears. Dijkstra was quite impartial in his treatments, bestowing his favors --caustic comments-- on friends and foes alike. Some of his typical questions for the speakers were: * What is your criterion of failure? Many speakers never got the question; since their research were funded, they were already successful. * What is the meaning of the color scheme in your slides? Most speakers did not know that they had a color scheme. * What is the average age of your slides? This helped the audience avoid reruns of old shows. He believed very strongly in effective use of the language. English is the Lingua Franca of computer science, he declared; so, every computer scientist should strive to be fluent in its usage. His own English was inadequate when he started his career, but he worked hard at it. I urge you to read his EWD reports, all available at my department web site, and observe his transition from a hesitant user of the language to an accomplished one. He felt, and I concur fully, that you can turn an excellent paper to an outstanding one if you can argue your case effectively. He believed in precision, not just in research presentations but in everyday life: his own, and that of his associates. I recall his getting his first mobile phone at my insistence. He came to my home in the evening, quite pleased that the salesman had setup the phone almost completely. He still had to setup his voice mailbox using a password. I offered to help. I had to call a certain number, and as I started receiving instructions on what to do next, I told Edsger: "Edsger , listen to this", and handed over the phone. He listened for about 10 seconds, and then asked me: "do you want me to listen, or do you want me to listen and act upon the instructions? Be precise; you are a computer scientist". We are here not just to listen to his instructions, but act upon them. Thank you.