Summary

As you can see, testing can be quite involved, and overwhelming to a developer, but it is currently, with the exception of formal method modeling, the only way to be sure your code is doing what you say it is doing. Empirical analysis on instrumented code can show you each and every line in your code that is “covered” by a test. This coverage information is good and bad. It is good because you have a number which marches forward or backward that you can use as a gauge of how much quality your code has within. It is bad because, just like any other gauge, developers can tweak the coverage numbers without actually increasing quality. By merely running code within the testing framework, the coverage number will increase but if your tests do not have valid assertions and are looking for the right things, the test is bunk. A pragmatic and honest approach to testing...