>You're right about it only using integers. Hexadecimal values were used primarily for indicating the value stored in bytes, which, of course, are integers. How 'bout taking two fixed length values? Use BINTOC() to covert the integer portion, then raise the fractional portion by an appropriate power of 10 to covert that to an integer. Then use BINTOC() again and concatenate the results.

Hey Geo,
So ok this would work but one thought... ick ;)
We're looking at N 15,4 values, so binary would be a larger expression to store. But I took your thought and came up with this to do the same kind of thing but into hex:

original = 10.23
valuestored = transform(orignal * 10000, "@0")
originalval = val(valuestored)/10000

This allows the storage of numbers with 4 decimal places like we wanted. What do you think?