OpenEuphoria

Re: math.e and misc.e

• Posted by Juergen Luethje <j.lue at gmx??e> Aug 04, 2007
• 468 views

CChris wrote:

> Juergen Luethje wrote:
> > 
> > Jason Gade wrote:
> > 
> > <snip>
> > 
> > > Um, why not just use 'nan'?
> > 
> > I also have thought of 'nan', but I didn't suggest it because I don't know
> > whether it is actually supported by the Eu interpreters and translators on
> > all platforms.
> > 
> > > It should be relatively easy to generate.
> > > <a
> > > href="http://www.rapideuphoria.com/cgi-bin/guru.exu?refDoc=on&libDoc=on&otherDoc=on&keywords=nan">http://www.rapideuphoria.com/cgi-bin/guru.exu?refDoc=on&libDoc=on&otherDoc=on&keywords=nan</a>
> > 
> > Regards,
> >    Juergen
> 
>  From %EUDIR%\SOURCE\global.e, line 205:
> 
> global constant NOVALUE = -1.295837195871e307
> -- An unlikely number. If it occurs naturally,
> -- there will be a slight loss of optimization
> -- as we will not know the value of a variable at compile time.
> -- (NaN could be used, but it's not 100% portable).
> 
> I'm afraid ther ar some pesky intercompiler differences that just prevent it
> from being used, even though it was the very purpose of NaN in IEEE specs.

That's a pity. Thanks for the info.

Regards,
   Juergen