Previous Lecture | Exercises | Top Level

Basic Client/Server Model

"Don't write big programs"

Next Section | Contents

As programs get bigger, the number of possible interactions between parts of the program increases dramatically:

In order to keep this complexity increase under control, different ways of chunking parts have emerged. We have already met two techniques:

Chunking into subtasks: One way to do this is to align the tasks into a pipeline. Every element in the pipeline performs a particular task and hands over the result to the next element in line. As we have seen in lecture 4, our monopoly project consists of a two element pipeline, of which we are only going to do the first one: the monopoly engine.
Chunking into objects: Objects provide a way to group data and functions into one structure. We thereby encapsulate and hide the internal communications between the elements within an object and provide a clean interface for other objects to use.

Both techniques work by reducing the number of relationships between objects, like so:

Next Section | Previous Section

Another way to simplify big programs is to streamline the way objects interact:

Next Section | Previous Section

Contents | Previous Section

Previous Lecture | Exercises | Contents

Christian Goetze