Bob,

>I was never able to get it to say that 2^2=4 with precision. Increasing the number of decimals just caused it to return 4.000000000000000000000000027.

>>I don't think the .Log() method is the correct approach to work with an
>>Ahnentafel.

Floating point numbers are rarely ever "accurate" especially when you bring functions like log() into the mix, because the constants used in the calculation can't be expressed in 32, 64 or even an infinite number of bits.

>Actually, I've now found a DLL version of the GNU arbitrary precision math library at http://www.egenix.com/files/python/gmp-3.1.1-win32-1.zip. All of the docs still seem to be in C++, however, so I still don't know how to use it :-/

If you look in the VFP code in the two samples I've been referencing you'll see the syntax for the DECLARE command needed to tell VFP what the interface to the function looks like. Once you've done that you can call the DLL functions like any builtin function or UDF.