Aunc.3407
net.followup
utzoo!decvax!harpo!duke!unc!smb
Thu May 6 15:46:17 1982
Re: Dijkstra

I thought Dijkstra was in his usual form in that article: opinionated,
arrogant, and sometimes painfully right.

The idea that language shapes thought isn't new, and it's fairly obviously
in programming as well.  The best example I can think of is APL vs.
almost anything else:  if you're used to, say, Pascal, you won't be able to
program decently in APL; in Pascal, one deals with array elements, whereas in
APL, one deals with arrays as wholes.  There's also a functional style
of programming in APL that makes for much more modular programs; again,
Pascal doesn't encourage this nearly enough.  (UNIX programmers tend to
do well at that aspect; a good APL function is like a good filter.)  The
problem with APL isn't its control structure -- if that's what's making
your program messy, it was almost certainly written in Pascal but
transcribed in APL -- rather, it's just too easy to write extremely
cryptic programs.

No one who's read other stuff by Dijkstra should be at all puzzled by
his emphasis on mathematics.  Whether or not he's right -- and I think
that his position has some merit -- the attitude that anyone can just
sit down and code up an operating system or a compiler is, in
Dijkstra's opinion, responsible for the sad state of the software
industry today.  As "A Discipline of Programming" makes very clear, he
feels that programs *are* mathematical constructs; further, viewing
them in that fashion leads to better, more correct programs.


               --Steve Bellovin

Disclaimer:  my dissertation is based on predicate transformers and the
programming language described in "A Discipline of Programming"....

-----------------------------------------------------------------
gopher://quux.org/ conversion by John Goerzen <[email protected]>
of http://communication.ucsd.edu/A-News/


This Usenet Oldnews Archive
article may be copied and distributed freely, provided:

1. There is no money collected for the text(s) of the articles.

2. The following notice remains appended to each copy:

The Usenet Oldnews Archive: Compilation Copyright (C) 1981, 1996
Bruce Jones, Henry Spencer, David Wiseman.