>> I might be missing the point on this (as I sometimes do), but isn't the
>> shortest distance from point A to point B a straight line?  So we could first
>> calculate the equation of the straight line from A to B.  To choose the next
>> point in the path, we would first calculate the perpendicular distance to
>> this straight line of each point under consideration.  We would then choose
>> the point that had the shortest perpendicular distance.
> 
> Or, what if each length between points was determined by cost, or
> speed, or some other factor (ie., assembly cost), could you easily
> find the lowest cost, fastest route, or the most economical solution?

The problem is, Mark, that although your suggestion works to a point, it's
quite possible that you might find a dead-end just before the end, which
means that the shortest path may actually go away from the end point at some
stage. Or as tedd says, you may want to check for some other attribute.

Pathfinder actually checks for two different attributes - distance, the
shortest path, and speed, since each link between two points has a speed
attribute. So you can either find the shortest route between two points or
the fastest.

I'm just tidying Pathfinder in FB3, and once this thread dies, I'll release
it as open source with any improvements offered.

Phil.