On Wed, 25 Aug 1999, Kenichi Okuyama wrote:
> >>>>> "MD" == Mike Davis <[EMAIL PROTECTED]> writes:
> MD> Is there a method (or algorigthm) for dealing with 64 bit integers
> MD> using only 32 bit data types?
>
> CodeWarrior, have flags to supports long long ( or was it longlong?
> ) keyword, which stands for 64bit fixed point variable. GCC, as
> default, also supports keyword "long long" for 64bit fixed point
> variable.
If that is true, I feel kind of stupid. I scanned the header files and
did not see anything for 64 bit except for floating point. I'll look
again. What I want to do is use numbers like 2^60th as an integer.
Oh wait a minute! I was looking in the Palm headers and you said this
was a GCC type. That's why I did not see it.
If this 64 bit fixed point type just a binary representaion of an
integer or does it have some other format?
Thanks
> Are you talking about the way to handle 64bit fixed point without
> using these?
>
> If you do not have to take care of compatibility with other
> platforms, I recommend using "long long".
>
>
> Even if you can't use these keywords for compatibility, yes there
> are algorithm for dealing these. It's basically exactly the same as
> how you calculate by hand, only, instead of using decimal, you are
> using 4294967296-mal (^o^) for add/subtraction, and 65536-mal for
> multiply and divide.
>
> If you are looking for a good "code" ... I'm sorry. I do know the
> good book, but it's written in Japanese.
>
>
> regards,
>
>
----------------------------------------------------
Shoot-to-Win
Protect the 2nd Amendment
----------------------------------------------------
