I do agree that Harlan's code looks good. I haven't tried to compile and run
it, but it looks like a good approach.

Bill
------------------------------
Jim Thomlinson wrote:
I am with you on the brute force requirement, but there are a couple of
tricks to minimize the permiutations and combinations. By sorting the list of
input values you can determine to stop testing certain combinations knowing
that certain solutions can not be possible because they are going to be too
large. That is where the code that I posted is very good. I had some other
code that did almost exactly what you were suggesting but it was far slower.
From what I have seen Harlan's code is hard to beat. That being said the list
you are searching should be at most 25 or 30 entries.