Coming into software with an exploratory testing mindset is like coming to a multi-layer canvas with lots of information and an open ended task: find what we may have missed! This is the assignment for us all in software teams in our quest for quality.
Framing the search of how our system falls short of expectations is easier when we are able to see software from its user’s perspective. However, useful tests aren’t a collection of end-to-end tests we automate, but great tests to leave behind will decompose the testing problem differently. In this talk, we learn about using architecture as a filter in decomposing tests and look at an example of taking control over the API responses to test a React frontend.
Users don’t know or care if the problem is in the frontend and services your team provided if it fails to meet their expectations but you care. Granularity of feedback matters. Recognizing the same problems in incomplete scope - half-done features or only in frontend or APIs - is a skillset the software industry needs to be building.