We now have a simplified model of the program based on data abstraction. If we look at the overview of the subsystem decomposition that we created, we note the new level in the graph and the decomposition into two subsystems. This makes the program easier to understand and easier to explain to others. In the source code window, we now have only four nodes that represent the program. Relationships between subsystem nodes and our ADT nodes are represented by composite (red) arcs. These composite arcs tell us that the two objects depend on each other in some manner, but do not provide any details. Details on the exact dependencies between two objects can be retrieved through an exact interface operation.