After a discussion on the Agile-Testing mailing list, I decided to give up the proposal to rename the Agile School of Testing. Erik Petersen put it in such a good way, that I fully agree with, so I decided to quote him here:
The schools as defined by Brett in soundbites are:
sees testing as rigorous and technical with many proponents in academia
sees testing as a way to measure progress with emphasis on cost
and repeatable standards
emphasizes process, policing developers and acting as the gatekeeper
emphasizes people, seeking bugs that stakeholders care about
uses testing to prove that development is complete; emphasizes automated
I see no evidence in those descriptions that the Agile school has a monopoly on examples. All of these schools choose examples to demonstrate that a system appears to deliver their interpretation of functionality at a point in time, with differing degrees of attention to context and risk. I believe the Schools idea was originally intended to describe groups who tended to favor their ideas (dogma?) over others and focused mainly on functional testing, and when the original Agile School was named, it claimed to be replacing all the other schools. This has since changed considerably, and with new techniques such as Mocking and Dependency Injection and a focus on refactoring (CRAP metric anyone?) I would argue that Agile is much more about design and development aimed at simplicity (YAGNI), of which automated testing is only a part, rather than a specific School of (functional) testing. As I have said before, schools tend to manifest themselves in organizational culture and IMHO are relevant for discussion purposes only. Testing can involve many ideas, some of which are typically associated with schools, and depending on context and risks, testing can draw from all of them. My 2 cents.
Part of the problem is what the earliest schools have become, not what they started as. The original articles on waterfall in the early 1970s stated that just going from dev to test in one step never worked and needed to iterate between the two, but that got lost. In the mid 70s, Glenford Myers in amongst all his “axioms of testing” said that tests need to be written before being executed (because computers were million dollar machines and time was money, no longer such an issue) but he also said stop and replan after some execution to focus on bug clusters, and that also got lost. We need to be open to new ideas and weigh them against our current ones, based on their value and not the perceived baggage they bring from a particular school . Enough of the examples! [grin]
So in a sentence, I agree with Lisa’s posts and Markus’s later post about combinations of techniques (quote “Thinking in black and white when thinking of the schools of testing leads to misconception”), but Markus please ditch the Agile school rename attempt!
Beside the remaining very good comments on this topic on the list, the idea of Schools in Testing simply does not care Agile Testers. Here is an excerpt from the discussion I had with James Bach on the topic:
When we speak of schools in my community, we are speaking of paradigms
in the Kuhnian sense of worldviews that cannot be reconciled with each
Basically context-driven thinking helps Agile testers as well, but they don’t adapt a Kuhnian sense of worldviews towards testing. Mainly I am considering whether there is such a thing as a Agile School or not. Bret Pettichord felt like there was, but currently I am not convinced about it. I am glad that I learned a big lesson from great contributors on my renaming approach, and I finally ditch this attempt.