Re: infinity and beyond

[christos%astron.com@localhost (Christos Zoulas)]

In article <201111020226.WAA25172%Sparkle.Rodents-Montreal.ORG@localhost>,
Mouse  <mouse%Rodents-Montreal.ORG@localhost> wrote:
>>>>    double x = INFINITY;
>>> I think it's a defensible proposition that it _should_ break on the
>>> VAX, because VAX floating point does not have [infinities]
>> I think it should compile, otherwise we are not going to be able to
>> build libm/complex/cproj*.c.
>
>Well, I haven't looked at the code.  But perhaps it's the libm code
>that really needs fixing here?  If it tries to use infinities, I think
>someone needs to give it a good hard look to see if it's going to work
>on the VAX at all, for any definition of INFINITY.
>
>Of course, you may have already done that, but you don't say anything
>that clearly indicates so to me.

Well, this is as good as it gets (using DBL_MAX or HUGE_VAL). If anyone
has better ideas, I am all ears. At least it is better than before, where
the 1.0e+39 would get silently truncated.

christos