Software Debugging

Lesson 1: How Debuggers work

Theory: Scientific method and its application to debugging. Fun fact: First bug in the history of computer science. Practice: Building a simple tracer.

Lesson 2: Asserting Expectations

Theory: Assertions in testing and in debugging. Fun fact: The most expensive bug in history. Practice: Improving the tracer.

Lesson 3: Simplifying Failures

Theory: Strategy of simplifying failures. Binary search. Delta debugging principle. Fun fact: Mozilla bugathon. Practice: Building a delta debugger.

Lesson 4: Tracking Origins

Theory: Cause-effect chain. Deduction. Dependencies. Slices. Fun fact: Sherlock Holmes and Doctor Watson. Practice: Improving the delta debugger.

Lesson 5: Reproducing Failures

Theory: Types of bugs (Bohr bug, Heisenbug, Mandelbug, Schrodinbug). Systematic reproduction process. Fun fact: Mad laptop bug. Practice: Building a statistic debugging tool.

Lesson 6: Learning from Mistakes

Theory: Bug database management. Classifying bugs. Bug maps. Learning from mistakes. Fun fact: Programmer with the most buggy code. Practice: Improving your tools and practicing on a real world bug database.

• Andreas Zeller - Andreas Zeller is a computer science professor at Saarland University, Germany. His research centers on programmer productivity: What can be done to ease the life and work of programmers? Among Linux and Unix programmers, he is best known for GNU DDD, a debugger front-end with built-in data visualization. Among academics and advanced professionals, Zeller is best known for Delta Debugging, a technique that automatically isolates failure causes for computer programs.