Fluent C

Chapter 10. Implementing Logging Functionality

Choosing the right patterns in the right situations helps a lot when designing software. But sometimes it is difficult to find the right pattern and to decide when to apply it. You can find guidance for that in the context and problem sections of the patterns from Part I of this book. But usually it is much easier to understand how to do something by looking at a concrete example.

This chapter tells the story of applying the patterns from Part I of this book to a running example that was abstracted from an industrial-strength implementation of a logging system. To keep the example code easy to grasp, not all aspects of the original industrial-strength code are covered. For example, the code design does not focus on performance or testability aspects. Still, the example nicely shows how to build a logging system piece by piece by applying patterns.

The Pattern Story

Imagine you have a C program out in the field that you have to maintain. If an error occurs, you get into your car, drive to the customer, and debug the program. This works fine until your customer moves to another city. The car ride now takes several hours, which is not at all satisfactory.

You’d prefer to solve the problem from your desk to save both time and nerves. In some instances, you can utilize remote debugging. In other instances, you need detailed data about the exact software states in which the error occurred, which is very hard to get via a remote connection—especially ...