[Haskell-cafe] Some historical remarks [Was: Why Haskell is beautiful...]

[Jerzy Karczmarczuk]

Le 30/08/2015 03:29, Donn Cave a écrit :
> I was already 4 when Lisp was
> invented, so computer programming pedagogy when I was born - that must
> have been a dry job
Dry job?... I don't think so. It was a glorious period, I think.
Some years later as well.

My first language (learnt and also taught some years later) was Algol60 
[replaced very soon by Fortran]. The relationship between our 
computation needs and the coding styles, paradigms, algorithm 
structuring, etc. was quite obscure at the time, and our main source of 
inspiration was the collection of algorithms in CACM, published in Algol.

I suspect strongly that those - fundamental AND practical - materials 
conditioned the opinion of the community (at least mine: physicists) 
about "what is needed, and what should be taught" (and perhaps also 
"what is elegant"). Moreover, they played (I am not certain) a 
significant role in specifying what should be the "main stream" of 
languages, and gave a strong push to the development of compilers of all 
them -- Jovial, Pascal, C, Java...
If, at the beginning of the '60ties, the Lisp community had published 
more code recipes, more algorithm presentations, more practical 
functional codes, perhaps the evolution of programming languages would 
be different. Perhaps the evolution of computer architectures would be 
also a bit different (more importance for hardware stack architectures)

We have seen some nice psychological paradoxes. Anthony Hearn told me 
that he recognized the importance of Lisp for the symbolic computations 
very early, but he found out that his colleagues abhorred its syntax so 
strongly, that his Computer Algebra package Reduce became Lisp disguised 
in an Algolish language... Now, the Beauty and the Beast, where were 
they?...

On the other hand, around 1963, Martinus Veltman (his Nobel had to wait 
more 36 years...) found out that the "fast", and low level Fortran for 
such needs was useful as well, but as an implementation platform, not 
the front-end. He conceived his computer algebra system Schoonschip 
(name chosen "among others to annoy everybody not Dutch", quote from 
Wikipedia, probably authentic, knowing the personal character of 
Veltman) upon Fortran, and assembly language, but he manufactured a 
*rewriting system*, completely different from any other language! The 
result was that for many years only theoretical physicists used it, 
because it worked, but others (e.g. people from an engineering school to 
whom I proposed a course of Schoonschip) refused to touch the "monster".

Well..., this "monster" evolved to FORM of Jos Vermaseren (also 
initially coded in Fortran, it still exists and is used). It inspired 
Cole and Wolfram at Caltech in their work on the computer algebra 
rewriting system SMP, that you know well.

Oh, you don't?
Yes, you do, only that it is called now "Mathematica".

You don't use it? But you *DO USE* another rewriting system, only 
perhaps you don't know that it is one. It is called TeX...

I repeat again:
Programming languages evolve as the Culture does, with redundancies, 
inspirations, contradictions, and re-discoveries of the same paradigms 
several times. (Some people teach Prolog and say that it was the first 
known language which used logical non-determinism. But Colmerauer began 
to work on it at the beginning of '70, while Snobol was there for almost 
10 years, and Griswold introduced into it the non-deterministic 
string-processing algorithms since the beginning.)

I find it pityful that nowadays some young people learn 2 - 3 
programming languages, and quarrel about which one is "better" or 
"worse"... Recall please one silly discussion between Van Gogh and 
Gauguin, when Van Gogh accused Gauguin that he gets everything false, 
that his world is flat, which is ugly and silly. Gauguin answered that 
the world *IS* flat, and that all these pointilistic details of Van Gogh 
are useless, false, and go nowhere.
Well, both masters are great and they are still with us. Their 
discussion, perhaps invented, is almost forgotten, but it can be used as 
a /memento/. Their colours remain. Be colourful!

==

Jerzy Karczmarczuk
/Caen, France/