🚀 go-pugleaf

In article  <126@castle.ed.ac.uk> db@lfcs.ed.ac.uk (Dave Berry) writes:
>[in Icon]...
>What about  (a < b) = (d < e), i.e. comparing results of relational
>operations? ...
>and not to the expected result from boolean logic.  Or does bracketing
>a relational expression force it to produce a canonical truth value (ugh) ?

No.  Icon doesn't have anything resembling a boolean type.  I might
point out that the expression 

    (a < b) = (d < e)

is from _extremely_ obscure boolean logic, not from common high school
algebra.  No one would know what to do with it except digital
engineers, mathematicians and Pascal programmers.  Icon was written
for a larger audience.

And furthermore,

#include <soapbox.bool>

Boolean algebra is an inelegant hack in programming languages.
Programming languages should either restrict relations to conditional
contexts (ick) or find an interesting and useful purpose for
relationals outside of conditional contexts (like C, Lisp, and Icon
have done).  Actually, boolean algebra is an attempt at the second
option, but a failed one.  Why use a trivial domain like truth values
when it can easily be subsumed into a larger, more interesting domain?
-- 
					David Gudeman
Department of Computer Science
The University of Arizona        gudeman@arizona.edu
Tucson, AZ 85721                 {allegra,cmcl2,ihnp4,noao}!arizona!gudeman