This boils down to defining a set of methods for what makes two cs1graphics objects "equal". We must be very careful about differences that don't affect the visual output (like the order of corner points in a polygon, hierarchy of layers, and so on).