When visuals come under test automation, the outcome rarely justifies the cost. Visual tests tend to be too brittle, to difficult to maintain, and too slow to execute for quick iterative delivery cycles. However, recent improvements in cloud computing and browser capabilities make it possible to change the economics of the test automation pyramid. This presentation shows how trends such as approval testing, cloud functions and automated image analysis can help us automate acceptance/regression tests for visual look and feel in a visual language, rather than xUnit style code, and make such tests easy to write, understand, execute and maintain.
Video producer: http://ndc-london.com/