On Tue, 14 Oct 2003 15:15:22, Brian Smith wrote:
>Found out what it is, and it's not what I first thought it to be.  Seems
>that a type cast from double to Int32 is what's causing the problem
>(adding about 3k of code and apparently a global or two).  What I don't
>know is how to get around it.  Ideas?

If the numbers are in a reasonable range (approx 1 <= x < 2^31,
etc.), and you don't care about rounding accuracy, it's not hard to
extract the exponent and mantissa bits from an fp double and do
a shift, and with much less than 3k of compiled code plus globals.

There are a large number of web pages which document the IEEE
binary floating point format.  Here's a random one from google:
  http://pheatt.emporia.edu/courses/cs220f03/hand16a/hand16a.htm  


Ron Nicholson
HotPaw Productions
 <http://www.hotpaw.com/rhn/palm> 

-- 
For information on using the Palm Developer Forums, or to unsubscribe, please see 
http://www.palmos.com/dev/support/forums/

Reply via email to