You would not believe the Real and Integer issues that can be encountered these days, in all kinds of tools. Reals especially lose precision in normal ranges, due to the way they are represented.
The easiest way, imho, to deal with these issues, is, when you reach some threshold, switch your math routines to manual ones. This used to be a standard problem at ACM competitions. The fibonacci problem was one of them. You start with numbers that are out of bounds for a real or integer of any precision or size, and manipulate them. So, you take each digit, perform the operation, and move to the next one, just like when you're doing it by hand. The good news is that with RR it's significantly easier than many other tools because there isn't all the (st00p1d) conversion from string to integer to string. You just do it. One trick in this regard - sometimes you can save yourself some headaches if you reverse the order of the digits. Adding a position to the front of a string CAN be slow, but adding one on the back GENERALLY is not. You could also use blobs, which have arbitrarily large size, and perform the same operations. Note that in the case of blobs and other binary types, you can perform arithmetic operations by using logical binary ones, as well (AND, OR, XOR). All of this is left as an exercise for the reader. _______________________________________________ use-revolution mailing list [email protected] Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-revolution
