The email below is published with the permission on Robert Cailliau. It was written in response to a note on the history of Notis by Tor G Steine]
From: Robert Cailliau <robert@cailliau.org> Date: Thu, 21 Sep 2017 23:21:46 +0200 To: Håkon Wium Lie <haakon@wiumlie.no> Subject: Re: notis Bl@@$* Hell!!! Yes. Actually David Bates and I wrote a text formatter called REPORT (the manual is attached! Obviously the authors are listed in alphabetical order). Tim used it to write his reports during his stint at the PS Division (in 1980 I think). We first implemented it on a PDP-11 in PL/11 if my memory is correct. Then in 1976 we were "forced" to adopt ND machines. But we soon became enthusiastic. I ported the ETHZ PASCAL compiler to the ND-10 (that's another long story). Then we ported REPORT to the ND10 by rewriting it in PASCAL. Or maybe we used N-PL? Can't remember. ND soon became interested in "Office Computing" and we gave them REPORT. They got it for free of course, as it was done at CERN. They added a lot to it, sold NOTIS with NOTIS-TF as the basis, which was nothing more than a souped-up version of REPORT, and made a fortune form NOTIS. I do not regret it, and I am also aware that they put a lot of effort into it, extending it greatly and making it easier to use. Nevertheless, for many versions of the commercial product, I was able to crash NOTIS by typing a specific sequence of characters with accents. This sequence went through a bug in REPORT that we never fixed. It was a very strange sequence, and the bug was of the type where you have a long sequence of if-then-elses checking for everything but forget one strange possibility. Each time ND distributed a new version of NOTIS I tried that sequence, and if it crashed I knew the core was still the REPORT engine. :-) In the end they re-wrote the entire system and the bug was gone. Now, remember this was 1976 (the manual bears the date 1977 but obviously the first versions were for our own use, not the general CERN public). Look at how the directives were set up: everything was relative. There was no silly H1, H2, H3, … If you took a chapter with several sections and sub-sections, you could move it elsewhere as a section of another chapter. In HTML you would have to make the H1s into H2s, the H2s into H3s and so on. I'm pretty certain Tim threw the HTML tags together without much thought (and rightly so: HTML tags were not the top priority) but I'm still not certain to what extent Tim used the REPORT directives. Also, there was some SGML being used at CERN at the time, and some of the DTDs probably had tag sets that were inspiration for HTML. I also wrote a distributed programming language and its compiler. Doing that I came to greatly admire the simplicity and power of the ND hardware instruction set. It was a RISC machine "avant la lettre". Only recently did I find out that set had been designed by Ole-Johan Dahl and Kristen Nygaard, the Simula guys… You can understand why I was never able to write programs in C or any C-like language. They have a revolting syntax and bad semantics. Well, I do write JacvaScript, but it really upsets me all the time. Sigh. Say hello to Tor & company! We had a great time working with ND. Robert. > On 2017-09-21, at 22:37, Håkon Wium Lie <haakon@wiumlie.no> wrote: > > Hello Robert, > > This week, the Museum of Science and Technology organized a seminar in > honor or Norsk Data's 50th birthday: > > https://tekniskmuseum.no/besok-oss/helgeprogram/1408-norsk-data-jubileumsarrangement-18-sept-2017 > > I was there to talk about Opera and ND. In my presentation I ceded to > Notis: they were right, WYSIWYG was wrong. I used a screenshot found > in a book to illustrate my point: > > http://www.wiumlie.no/2017/talks/09-18-nd-norsk-data#20.0 > > Afterwards, someone told me that the codes I had shown were invented > at CERN. Really, I said. By whom? Robert Cailliau, no less! > > The former ND people I met, Tor G Steine and Ragnar Sturzel remembered > you well. I didn't know you actually wrote a text formatter for ND?!? > > Tor says hello. He allowed me to publish his notes on the subject: > > http://www.wiumlie.no/2017/notis/tos > > In Norwegian, I'm sure you understand. > > Cheers, > > Håkon Wium Lie haakon@wiumlie.no www.wiumlie.no +47 90192217