Miker

17th level Hacker

Niklaus Wirth

Here’s some quick notes from the talk that Niklaus Wirth gave at the Computer History Museum on Oct 20th. The topic was “How I Became Interested In Computer Languages”.

Professsional career really started in Berkley. Thats when his real experience with computers started. Was always into modeling, and a lot of qualites cary over. Using tools, experimenting, persevering in spite of crashes. Liked physics and chemistry, and didn’t really pay attention to languages.

When model airplane parts became available, he was forced to study electronics in order to use them. Got theory at University and practice at home. Eary playing with computers took a long time, and ran into people who were working on a translator for computer languages, called a “compiler”. Only one person really understood the program, and it was messy, so he decided that the practice of building compilers need structure.

The early version lacked abstraction, it did not abstract away details of the computer hardware. Still it was easier than previous methods.

At the time thing were beautiful in Berkley, golden years. Before Flower Power. He joined in a seminar about Algol. Based on mathematical rigor. There was a clear way to figure out if a text was a valid program. Decided to write a compiler for it. The compiler was short and very simple. Met a witty and engaging person who tossed about some interesting ideas. Made Wirth think about a reduced Algol that had the same expressive power as the original. Wijngaarden.

Became a professor, and people were still trying to figure out what to teach about computers besides languages. IE. Data structures. The challenge was to introduce rigor into semantics as well as syntax. Although Algol was interesting it didn’t suit practice very well.

Implemented Algol-W at SLAC. That was a stepping stone to his professional career. He made frequent trips, even overseas, which was not common at the time. There were heated discussions all over about the future of Algol. Algol always remained an academic exercise, never accepted into industry.

Programming became more and more complex, projects became very large, and even large companies drifted dangerously close to collapse. They were not able to deliver on their promises. This is what led to structured programming. Now one person didn’t have to understand the whole system. This was very far ahead of its time (and maybe still is).

Thing were getting bad with the Vietnam war, and he was offered a job in Zurich to continue work on Algol. He took it. Without having to work with comittee, he was able to work towards what he wanteed. The stuff he wanted stemed from having to teach languages and keeping the compiler clean. The idea was to write the compiler quick in Fortran, and implement the new lanuage, Pascal. Eventually, they had to scrap it. The decision was made to write Pascal in Pascal. The Pascal compiler was translated into another language for which there was a compiler to bootstrap.