Although thats the first time I've had to use the float command, whats really strange is I'm doing the samekind of thing earlier in my proggie - and it works fine.
But here.. it just stopped dead !!!.
Anyway, thanks - I can stop banging my head against the wall now. (appologies for posting in the 3d thread, but it is part of a 3d program and I needed a quick answer - honest !)
what it is doing is calculating the first 2 values as integers(10/8) then passing the result value to diff# so the result is rounded down to 1 before diff gets it.