If every piece of code worked perfectly the first time, programming would be an easy job. Sadly, that is not the case. In any large project, you will probably spend more time testing, tracking down bugs, and figuring out how to fix them than you spend writing the code in the first place.
Dealing with bugs is an unavoidable part of writing software, but there are things you can do to make the job easier. At every step of the process, you make decisions which influence the risk of creating bugs later on. Experienced programmers understand this and intentionally write their code in a way that minimizes the opportunities to make mistakes. This is sometimes known as "defensive programming".
In this article I make this practice a little more systematic by presenting a taxonomy of "bug opportunities" -- situations in which it is possible to make a programming mistake. I categorize them into five levels, based on the time and effort needed to identify the mistake. And I give examples of each level, and suggests strategies for turning the more severe levels into less severe ones.
A bug opportunity is not the same as a bug. With enough caution, even a minefield can be successfully navigated. But you will make mistakes from time to time, and the more opportunities you have for mistakes, the more mistakes you will end up making. A key to writing bug-free software is therefore to avoid situations in which you might make a mistake; or when that is not possible, to choose your risks carefully so that bugs will be found and corrected as quickly as possible.
Most of the examples in this article are in Java, and I assume familiarity with the Java language. None of the principles are Java specific, however, and they apply just as well to almost any other language.