Mastering Object-Oriented Python - Second Edition

Defining processes

We must design each processing step as a simple loop that gets a request from a queue, processes that request, and places the results into another queue. This decomposes the larger problem into a number of stages that form a pipeline. Because each of these stages runs concurrently, the system resource use will be maximized. Furthermore, as the stages involve simple gets and puts into independent queues, there are no problems with complex locking or shared resources. A process can be a simple function or a callable object. We'll focus on defining processes as subclasses of multiprocessing.Process. This gives us the most flexibility.

For the simulation of a stateful process such as a game, we can break the simulation down ...