Very Computer

Hello.
I'm running ESIX 4.0 on a 386/40 clone. I don't have a math coprocessor.
I was wondering why the DBL_MIN, in /usr/include/limits.h
is different from MINDOUBLE in /usr/include/values.h.

/usr/include/limits.h:
...
#define DBL_MIN         2.2250738585072014E-308  /* min decimal value of a "double"*/
...

/usr/include/values.h:
...
#define MINDOUBLE       4.94065645841246544e-324
...

It really erked me when I saw this, because I'm working on a parser for
floating point numbers. And I don't know how to set the limits on the
exponent and the significand in lue of this anomaly.

To further obfuscate the matter, the following compiled passed okay:

------------------------------------------------------------------------------
void main( void )
{
/*
reminder :
MINDOUBLE       4.94065645841246544e-324
DBL_MIN         2.2250738585072014E-308
*/
        double f ;
        f = 5.0e-324 ;
        exit( 0 );

Quote:}

-------------------------------------------------------------------------------

The compiler should do some complaining about this; in spite of the above
inconsistancies! n'est pas ?

Could somebody body enlighten me. I must be missing something ... or drunk!

Thankyou again,

Erica C. Ramsey