Defining IEEE NaN and Inf at compile time

[Jon L White]

re: I would like to invite comment on this: should compile-time
    evaluation of Infinity and Not-a-Number terminate compilation,
    or not?

I think it should not.  In 1987 I added IEEE 754 arithmetic capabilities 
to Lucid Common Lisp, as an extension to Common Lisp, and forgot about 
special treatment for compile-time extremals (infinities, NaN's, and even 
denormalized numbers).  Fairly shortly after introduction of the 3.0 product,
we received at least one complaint about that lacuna, and we subsequently
patched the compiler/loader accordingly.


-- JonL --