Re: LONG, LONG math - aka 64bit math....

[Dave McKenzie <davemck@xxxxxxxxxxxxx>]

There are modules in SLIC that implement Colin Plumb's bnlib (BigNum lib), 
which handles arbitrary-precision decimal numbers.  Presumably they're used 
for the Java BigDecimal class.  Arbitrary-precision means the numbers can 
contain any number of digits.  However a note on alphaworks says that IBM's 
implementation is limited to 999,999,999.  That doesn't mean a number with 
nine digits--it means a number with 999,999,999 digits.  Considering that the 
number of elementary particles in the universe has something like 85 digits, 
that should be big enough to count anything physical :-)

The SLIC modules are named bn, bn32, bninit32, bn_prime, lbn32, lbnmem and 
sieve.  There's also a module named legal, which contains nothing but the 
string "bnLib 1.1 Copyright  (c) 1995 Colin Plumb".

BTW, most references to bnLib that I found on the net point to:

ftp://skip.incog.com/pub/bnlib-1.1.tar.gz

but that doesn't seem to be there anymore.  I found it at:

http://munitions.iglu.cjb.net/software/libraries/mathematics/bnlib-1.1.tar.gz

--Dave


On Wednesday 04 July 2001 08:44, you wrote:
> Yes, but what do they do for BIG numbers?  <grin>
>
> Happy Independence Day, to all the US list attendees!
>
> > -----Original Message-----
> > From: Leif Svalgaard
> >
> > the machine has always had 2048-bit ADD/SUB instructions
> > (ADDLC/SUBLC).
+---
| This is the MI Programmers Mailing List!
| To submit a new message, send your mail to MI400@midrange.com.
| To subscribe to this list send email to MI400-SUB@midrange.com.
| To unsubscribe from this list send email to MI400-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: dr2@cssas400.com
+---