Re: [Maxima-discuss] float(%e^1000) first gives error, on second try gives "false"

[Camm Maguire]

If maxima calls an error here instead of signal, it should be a cerror
so one can get the #<inf> if desired.  GCL can do the following, but I
have not been able to discover a way for C code to return from a traped
FPE:

>(si::break-on-floating-point-exceptions :floating-point-overflow t)

(:FLOATING-POINT-OVERFLOW)

>(exp 1000000000000.0)

Error: 
Fast links are on: do (si::use-fast-links nil) for debugging
Signalled by EXP.

Condition in EXP [or a callee]: INTERNAL-SIMPLE-FLOATING-POINT-OVERFLOW: 
Arithmetic error when performing (:INSN "fnop ;" :OP ...) on NIL: 

Broken at EXP.  Type :H for Help.
    1  Return to top level. 
>>

Take care,

James Cloos <cloos@jhcloos.com> writes:

> i meant to add that ccl (also on linux amd64) returns 1E++0 .
>
> unless one calls (decode-float), after that it start sigalling a
> fp exception.  if instead one calls (integer-decode-float) the
> error makes it clear that 1E++0 is indeed +∞.
>
> -JimC

-- 
Camm Maguire                                        camm@maguirefamily.org
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah