It’s an understatement to say that Paul Gerrard is prolific.
He is an internationally renowned, award-winning software engineering consultant, author, mentor, webmaster, programmer, tester, conference speaker, rowing coach, and publisher. Whew! That’s a wide array of experiences and skills. Not only has Paul consulted in all aspects of software testing and QA, he has also agreed to keynote PNSQC 2020 this fall. We got the chance to ask him some questions about testing, poetry, the dreaded “Q” word, and his upcoming conference presentation, to whet your appetite for his keynote.
PNSQC: Since getting into the testing business in 1992, what do you think has made the biggest impact to testing?
Paul Gerrard: One word? I’d say Agile. But that’s a cop-out. More seriously, I think the biggest difference between then and now is that testing has become a career choice. In 1992, in the UK, I think there were just a few thousand people who would call themselves testers. Nowadays there are perhaps 50,000 people with “tester” in their job titles in the UK and — I’m guessing — 1.5 to 2.5 million testers worldwide. Outsourcing, certifications (whether you like it or not) and the maturation of the testing services market have all made contributions to this change. I also like to think we understand the thought process so much more than we used to, so the skill of testing can be more appreciated.
PNSQC: One definition of quality assurance (QA) is “the maintenance of a desired level of quality in a service or product, especially by means of attention to every stage of the process of delivery or production.” Are there paths to QA without testing?
PG: I’m not a fan of the Q word — everyone has a different perception of quality of course. But, taking the question in the spirit it’s intended, the answer is surely no. Testing is an essential activity in every systems project.
I compare the tester to the navigator in a plane. The developer is the pilot. The pilot can’t fly without navigation support; the navigator can’t fly without the pilot. Testing — of documentation/requirements, components, and systems — whether performed by developers, testers or users, provides information on achievement. Without knowledge of achievement, we can’t make meaningful decisions on whether stuff works or is acceptable.
Testing tells us “where we are.” And as the saying goes — “If you don’t know where you are, a map won’t help.” It’s the same with systems: “If you don’t know what’s been achieved, your plan won’t help.”
Can you give us a sneak peek of “Rethinking Test Automation,” your upcoming PNSQC keynote presentation?
PG: My view of automation is that the state of the practice hasn’t moved on in 30 years. Of course, the tools we use are more sophisticated and the systems we test are very different nowadays. But, the principles are the same: the four quadrant models, the pyramid model are sound — though far too simplistic. I still think people try and equate what tools do and what testers do.
I created my New Model for Testing to try and understand how testers think. I use the model to explain everything test-related nowadays. What’s clear from the model of the 10 thinking activities is that test execution tools support only ONE of those activities. Current tools generate lots of data and logging output, but, for now, we still need testers to interpret their outputs. In the future, I think ML/AI will have an increasing role to play. But to test new code, I think humans still have that responsibility. I do hope to trigger some new thinking in the audience. However, you’ll have to wait for that, I’m afraid.
PNSQC: If all software has bugs, aren’t we just digging the hole deeper when we develop automated tests? Who’s to say the tests don’t have bugs? (Who tests the testers?)
PG: Well it’s a recursive loop isn’t it? Who tests the testers of the testers of the testers… ad infinitum. It doesn’t help.
Of course, test automation code can have bugs, so to get it working, there’s a debugging process of the test code. Beyond that, the tests in prepared data might misalign with the requirement of course. But where that triggers false-negatives, for example, that’s self-correcting. Bug reports raised will be dismissed and the tester will have to adjust their understanding of the requirements concerned. False-positives are a different problem, but this is addressed through collaboration and sharing of, for example, story scenarios used to challenge requirements and generate test automation.
PNSQC: This year’s theme is 2020 Vision: Quality Looking Forward. Care to prognosticate on near-future tech trends for the next 12 months?
PG: I think Machine Learning and Artificial Intelligence will demonstrate real use cases and change people’s opinion on the level of hype and skepticism. ML/AI have a somewhat different role in testing than most people envisaged and will make inroads into the test process.
The tools marketplace will further consolidate — in particular, the major ERP, Banking, and other industry domain package vendors may acquire test tool vendors particularly if they have model-based test capabilities. The demand for smarter testers will increase and there may be some shedding of let’s call them less skilled testers. Recognition of the importance of business and personal skills in the testing community will increase and testers will pay more attention to these skills and improve them to add value to their employers.
PNSQC: How have you applied “quality” thinking to other parts of your life such as hobbies or sports?
PG: Since I avoid the Q word, I haven’t. Life is a journey of exploration and learning. In those respects, I apply them to everything I do. I believe if you stop learning, you die from the neck up. This year, I am going to learn much more about ML/AI by working hands-on with tools. I think there are deeper insights to be had regarding the testing thought process and also potential business opportunities too. Watch this space 😉
PNSQC: Is there a particular poem, artwork, or metaphor that sometimes inspires your work?
PG: I write poetry as a pastime — but mostly love poetry which isn’t really appropriate. If I get the time, I’ll write a testing-inspired poem. But don’t get your hopes up that the poem will inspire! The testing metaphor I use quite a lot nowadays is “tester as a journalist”. We are independent, researchers, and reporters. We don’t judge the news – that’s for others to judge whether the news is good or bad.
Thanks, Paul, for your insight and time! We’re looking forward to October when we can hear your full thoughts on “Rethinking Test Automation.”
Don’t miss Paul’s keynote, or the other talks and cutting-edge ideas on Quality Looking Forward in 2020 and beyond. Be sure to register for PNSQC now!