What is a good test case? While test automation certainly is a necessity, I believe that this question really is at the core of what we struggle with when testing systems. Structural and random tests have undoubted merits and, good news for academics, lend themselves to automating the generation of tests. Yet, are these really the tests we want to rely on when testing, say, advanced driver assistance systems? In this talk, I will revisit the idea of defect-based testing, argue why only tests based on defect hypotheses can be "good", and present a framework and several examples of how to render defect hypotheses operational.
展开▼
