When you’re on a plane, do you ever think about the software that is needed to fly and maintain an aircraft effectively? Software testers do. The importance of testing software, especially when it correlates with the safety of people, is unmatched in the software development process. In regulated industries, any defect that arises from an improper testing process can lead to a lawsuit and hefty fines.
Martin Hynie has been testing since 1998 and is the manager of software testing team for MXI Technologies in Canada. His team tests software that helps maintain military and commercial aircraft. I had a chance to get a quick interview with Mr. Hynie about his approaches to software testing and his biases when testing.
- Do you use automated and/or manual testing?
- Do you use detailed test plans in your exploratory testing?
- What techniques and/or methods do you use that make you unique as a tester?
- What are your biases that you are aware of when you are testing?
Mr. Hynie’s team uses a significant amount of test automation at many different levels. He went on to talk about what scripting languages he uses and the purposes of test automation. It is important to note that his team does not use test automation to validate what the product is doing, but rather if the code is working correctly by design. In addition, Mr. Hynie’s team does more manual testing than automated testing.
It was interesting to hear that Martin Hynie and his testing team don’t spend as much time on test plans as they do on context-driven testing. He explains that the time saved by not creating detailed test plans is spent helping product management.
Mr. Hynie’s team spends a lot of time on value propositions for their testing to present to product management. Getting product management involved early allows Mr. Hynie and his team to eliminate any shallow agreement that would take place otherwise. As with many regulated industries, communication—and how teams communicate to other teams—is of the utmost importance.
Mr. Hynie said that his biases change over the duration of a project. For example, at the beginning of the project he wonders if they are building the right test. Since the software he tests determines if plans are properly maintained relate to life or death situations, his biases have to be tackled consistently.
How do your answers with your software testing team compare to Martin Hynie’s?