The second basic principle of the Context-Driven school of software testing states that there are good practices in context, but there are no best practices. I believe that there are, although not the traditional ones. I propose a simple test by asking "Is there ever a time I wouldn't do or care about X?" If the answer is no, then X is an "always-good" practice. Practices such as risk management, prioritisation of work, using appropriate skills, choosing appropriate techniques, making informed decisions, and understanding our context, all seem to fit this definition. What varies is the degree of formality with which the practice is performed. In this paper, I explain: why the concept of "best practice" does not need to be inconsistent with a context-driven approach, how to evaluate your specific context, and how to choose specific good practices along the formal-to-informal continuum.
Best Practices and Context-Driven Software Testing
Building a Bridge