CSE 275, S07, Prof. T. J. Peters

For program_3, restrictions on data sets:

0. Project onto the plane z = -100.  (For the sample data
already provided, a good choice would have been to project
onto the plane z = -90.  If you did project the posted data
on another plane (say z = 0), you could see unexpected artifacts.)

1. This criterion applies between each pair of convex polyhedra:

	There will exist a plane of the form z = c, such that
	one polyhedron has all its z-values less than c and the 
	other polyhedron has all its z-values greater than z
	(Note:  This implies no ties in z across polyhedra, 
	leading to a simple sorting criterion across polyhedron.)

2. Each convex polyhedron will consist of a collection of triangles,
so there will be shared co-ordinates across those joined triangles,
leading to ties in multiple z co-ordinates *within* each polyhedron.  
(Unfortunately, there may have been some mis-understanding of this 
circumstance in class, since I moved quickly from polyhedra to the 
polygons forming them.).  However, then sort the polygons in each 
polyhedron according to the following criteria:

	Find all polygons who have the min-z value for the convex polyhedron.
		Within that collection of polygons with a specific min-z value, 
			Sort by max-z value, with the smallest max-z value
				going to the top of the list
				For a given min-z value, if there are
					ties in max-z, the sorting can be
					arbitrary.  (Why?)
	Continue to sort rest of polygons by next smallest min-z value,
		returning to the top of the algorithm and stopping when
		all faces have been sorted.
		
3. There will be no cycles across the faces.