On Thu, 20 Sep 2001, Jesus M. Castagnetto wrote:
> Just a short 2 cents:
> double is referred to number that is a floating point
> with double precision, i.e. twice as many bits are
> used to store as it is used for regular floating point
> numbers (called float).
>
> In fortran (and other langs), float was 16 bit and
> double 32 bit (IIRC). Some newer langs and scripting
> ones use the double width for either.
I'm sure Hartmut already new that, he just pointed out that 'double'
doesn't make sense if there's no 'single' precision floating point
numbers.
There's just one kind of it, and the natural name for it is float.
One additional note: java uses 32bit for float, and 64bit for double.
Btw:
Curious: we seem to have discussion about something (nearly?) everybody
agrees on ;-)
> Just my 2cents.
--Jeroen
Jeroen van Wolffelaar
[EMAIL PROTECTED]
http://www.A-Eskwadraat.nl/~jeroen
