Object-oriented decompositions

Object-oriented decompositions of systems better are able to cope with change. Each subsystem has a well-defined interface which communicates with remaining system.  All of these interfaces define all form of interaction which is required for proper functioning of whole system, but internal implementations are left to sub-system itself. This is since they manage to encapsulate those items which are likely to change (for example functionality, sequence of behaviour and attributes) within an object and hide them from outside world. The benefit is that the outside cannot see them, and thus cannot be dependent on them and does not need to get changed if these items change. Also, object-oriented decompositions are nearer to the problem domain, because they directly represent real-world entities in their structure and behaviour. The abstraction primitives built into reuse have a massive potential of reuse as commonalities among similar objects can be factored out and then, the solutions can be reused.

                                   Figure: OO Decomposition (OOD)