demers

I can't explain why this is happening, but when expr is asked to calculate values beyond 10^9, it starts putting out weird numbers. Try 10^10, 10^11 with the patch below.

a-bug-with-expr-and-pow

You're running into one of the limitations of computers. Most of the time there is only a certain number of bits available to represent numbers. This is the case in Max. The maximum integer that can be represented is 2147483647. After that it wraps around to -2147483648 (the lowest integer that can be represented).

If you really need numbers bigger than that, you need to look into another solution, like programming your stuff in mxj/Java (see for example, java's BigInteger class which can support numbers of arbitrary magnitude). 

For values outside the range of 32-bit integers, ie -(2^31) .. (2^31)-1, you will need to use floats.

However, 32-bit floats also have precision issues.

Welcome to digital arithmetic. You might want to read up on the subject.

> I can't explain why this is happening, but when expr is asked to

> calculate values beyond 10^9, it starts putting out weird numbers.

Use floats. You will have a limitation to a 23-bit mantissa in 

Inouk Demers schrieb:
> I can't explain why this is happening, but when expr is asked to
> calculate values beyond 10^9, it starts putting out weird numbers.
> Try 10^10, 10^11 with the patch below.
>
> Any solutions?

Use floats. You will have a limitation to a 23-bit mantissa in 
resolution, but its certainly usable...

-- 
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com

I certainly will read up on the subject. I had no idea until now, so thanks for the explanation.

A bug with expr and pow?