Between the 14th and 18th November, the three conferences Better Software East, DevOps East and Agile Dev East are taking place simultaneously in the same venue in Orlando, Florida. The conferences are organised around two days of tutorials, two days of talks, and a closing Agile Summit day with keynotes by several international speakers. InfoQ attended the conference to report on its contents.
Although the conference as such started on Monday 14th, some attendees had the chance to come earlier to participate in one of the multi-day training courses available. These started on Sunday and lasted 2-3 days, depending on the course, providing official certifications in subjects such as “Agile Tester” or “Scrum Master”. The contents of the certification courses served as an indication of what the main areas of concerns would be; although several topics were covered, half the courses gravitated around testing.
Monday and Tuesday were centred on half-day and full-day tutorials. The topics were varied but, much like the certification courses, they also had a big focus on testing. The author of this article had the chance to attend some of these tutorials, more precisely “IoT Testing Workshop” by Jennifer Bonine, and “The Tester’s (New) Role in Agile Development” by Rob Sabourin.
Jennifer Bonine focused on the new aspects that testers and test managers will have to be concerned about with the advent of the Internet of Things. As Bonine indicated, these concerns can be mostly classified into three categories: testing gaps, security, and identification of new requirements.
When talking about testing gaps, what Bonine highlighted is that with more and more devices being able to connect and interact with each other, the paths of execution will grow exponentially, and with that so will the possible test paths. Many teams will struggle to decide who is responsible for some of the new paths, which might lead to some of them being unattended. On security, she highlighted that each new device that can be connected to the Internet means one more exposed surface that can be attacked, which implies that the traditional “castle” approach to security (build a “wall” around your system and assume everything within it is safe) won’t work any more; instead, security considerations have to be embedded in every single device. Finally, in what seems to be a trend across conference speakers, Bonine encouraged testers to use their out-of-the-box thinking to help identify how connecting all these new devices together could unearth new value for existing products.
As per Rob Sabourin’s tutorial, he first talked about what testing actually means according to different organisations, identifying three flavours: validation, or checking that teams are doing the right thing (i.e., helping defend the correct requirements); verification, or checking that teams are building things right (i.e., helping clarifying requirements); and exploration, exercising the system in new ways to try to uncover insights. Regardless of the view, Sabourin emphasised the need to change mindsets so that testing is seen as a set of skills and activities, and not necessarily as just a role or job title. This allows breaking down the test tasks into more specific activities that several members of the team other than the tester can perform; for instance, a lot of the testing effort performed by a tester typically goes into setting up test data, which any developer can also do.
After this, Sabourin went on to cover several testing practices, from TDD to Exploratory Testing, while pointing out that some of the common wisdom around them might be misplaced. For instance, TDD is a practice that is usually practiced exclusively by developers, considering the activity too low-level for the involvement of testers; however, he pointed out that the theory behind TDD is that the developer should carry on writing tests “until there is nothing else to test”, but testers are usually better equipped to answer this question, reason for which testers should be involved in the process of writing unit tests. On the other hand, while it is commonly understood that pair programming produces better outcomes, early studies seem to suggest that the opposite is true for exploratory testing: two testers will tend to find more bugs if exploring the same functionality separately than together.
After testing, continuous integration, continuous delivery, and build pipelines seem to be the second most popular topic. This would seem to suggest that, if the topics of Better Software / DevOps / Agile Dev East are to be taken as indicative of the industry, companies are switching their attention to ensuring that products include the right functionalities, and that the delivery of such functionalities is as streamlined as possible, possibly worrying less about the actual technologies used to build the products.