Justin Hunter’s presentation at the 12th annual international software testing conference in Bangalore, India, December 2012.
The techniques discussed focus on how to reduce the amount of time spent selecting and documenting test scripts, reduce the amount of tests needed for execution by creating unusually powerful tests and thus increase the thoroughness of software test suites.
The talk explored the significant risks, for users, companies and employees of failing to catch software application failures before release (for example, looking at releasing Apple Maps with significant failures). And discussed how combinatorial (also orthogonal array or pairwise) software testing can be used to create test plans that test a large number of parameters/factors quickly.
This is an 8 part series of videos showing, in detail, various steps that can be used in creating a pairwise software test plan. Many test plans will not require many of these optional steps. Most of the videos have run times of 1-4 minutes.
Video 1 of 8, getting started creating up a new test plan.
Video 2 of 8, marking invalid pairs. This allows a test planner to avoid pairs that are not sensible – for example, if a certain product is not available for payment with a check (perhaps a service that requires a monthly payment). Hexawise will then avoid creating any test conditions that contained the invalid pairs that were identified.
This video provides practical tips for selecting appropriate test inputs for pairwise and combinatorial software test design.
Our experience shows as software testers begin to use pairwise test design strategies selecting the correct input parameters to test is often a bit of a struggle. This video will help testers get up to speed with designing pairwise software test plans effectively.
A basic introduction to pairwise software testing, designing a set of pairwise tests for a simplified banking application, and answering the following questions:
What is pairwise testing?
What challenges does pairwise testing solve?
How do pairwise software test design methods prioritize which specific test conditions are included in each of the tests in the plan?
What two aspects of pairwise test solutions don’t receive as much attention as they deserve? (Variability between tests, and front-loading of combinatorial coverage)
Video 2 of 2:
This video demonstrates:
How should you think about including parameters and values in your inputs so that they trigger specific business rules?
How can you remove “impossible to test for” scenarios from presenting?