A few quotes on Algol 68

[Jose E. Marchesi]

Hi people.

Reviewing my notes I found several interesting quotes on Algol
68. Thought you may enjoy them.

If you find an interesting quote, please send it to me!
I am eager to expand my collection.


  - "Why is it that Algol 68, despite almost invariably becoming the preferred
    language of anyone who takes the trouble to learn it, has failed to become
    widely used?" - Algol bulletin 49 p.6

  - "As some areas of industry are beginning to realise an ALGOL 68 prorammer
    can actually enjoy wha the is doing and, as a consequence can be very
    productive.  ALGOL 68 is a language for those who do not believe that life
    is infinitely long; it is for those who wish to avoid the experience of
    trying to dry a swamp with a teaspoon.  There is no denying that (like
    religion?) ALGOL 68 does require some initial intellectual effort, but
    this is not as great as some - who have not tried - would maintain."
    J.R. Oliver and R.S. Newton in "Practical experience with ALGOL 68-RT".

  - "It has been said that Algol 60 was an improvement over most of its
    successors.  Algol 68 would have been an improvement over most of its
    successors, had it had any." - Lambert Merteens

  - "A programming language that makes it really hard to write bad programs
     probably also makes it bad to write good programs." - Lambert Merteens

  - "Algol 68 is extremely bug-resistant.  For example, in one program of
    about 1000 lines, only about four errors penetrated to run-time and these
    were uninitialised-variable errors.  Each of these variables were intended
    to have been initialised as a side effect of a procedure call.  By far the
    greatest majority of programming errors were caught by the mode-checking
    at compile-time.  As has been noticed elsewhere, the reason for this is
    that a mode corresponds to a number of statically verifiably properties of
    an object.  These properties may not be explicitly written in the program
    source code, which usually discusses only the representation of the object
    on the machine, and may exist only in the mind of the programmer.  The
    programmer must himself ensure that the primitive operations he defines
    preserve the properties he is concerned with.  Thereafter when the
    compiler checks mode compatibility it is implicitly ensuring that these
    properties hold, even though it knows nothing of the properties
    themselves." - H.J. Boom.

  - "The difference between Algol 68 and other languages concerning coercion
    is twofold.  On the one hand, which coercions are to be applied is
    determined by the syntax and not by the semantics of he language.  This
    allows for a much more systematic treatment of the problem than .e. in
    PL/I.  On the other hand deciding which coercions have to be applied is
    much more context-dependent than in other langauges.  This increases the
    complexity of the algorithms involved." - Goos, in "Some problems in
    compiling ALGOL 68".