I have realized that recently people pay attention mainly to “how to test”, i.e. which python library, which framework, which ALM tool etc.. While we should focus more on “what” or even “why”. It is not only my observation – there are couple twits, posts, articles like “testers do not test anymore”…. Having interviews at my company I give as a starter a really “simple” requirement to be tested: “If A and B and C>10 Then E is True Else E is False” and well there are different answers some better some worse (some people start to writing code for that instead of tests), but what I really expect here is not the answer (because maybe there is not any proper one😊)! I expect questions: what is A? what is B? what is the smallest possible increment for C? Is 4 test cases enough? If this is safety-relevant maybe we need 8 test cases or even more etc. Only one person started correctly among one hundred! And it was an engineer without testing experience (but I guess with proper mindset).
The idea for a workshop is to discuss and create some tests cases for requirements like that and even more complex.
They are going to be created based on real requirements for embedded software for railway vehicles, but I am going to put them into the frame of DeLorean car-time machine from “Back to the future” (this is why “back to the testing” topic), so the requirement could be: “If valid target date has been entered, speed is valid and greater or equal 88 miles per hour then the command to activate the flux capacitor is set to TRUE”.
I am counting here for a nice discussion about test techniques and a lot of “what if” questions 😊
Something to write on – laptop or pen/paper
Sli.do – N930