Review: Test Automation Day 2016

I like conferences. For me, attending a conference is a great way of breaking out of the daily habit of doing work for clients, for meeting new people and for getting inspired and bringing back home ideas with me to apply in my job. Therefore, I try to attend at least two conferences every year. Also, I’m slowly getting into public speaking, answering the call for papers for conferences that seem like a good fit for a story I like to tell.

Last Thursday (June 23rd) I went to Rotterdam for the Test Automation Day (TAD) for what I believe was the fifth time. It’s one of the major conferences on test automation in the Netherlands, and therefore almost a must visit. If not for the quality of the keynotes and other sessions, then at least to meet old and new friends, colleagues and others working in the test automation field. I responded to the call for papers this year, but unfortunately my proposal was not selected, so I attended the conference as a visitor. I’ll just have to try again next year (more about that at the end…).

I missed the bigger part of the opening keynote by Sally Goble (from The Guardian) due to heavy traffic, so I can’t really comment on that. The next keynote was by Mark Fewster, who talked about test automation health. By this, he means that you shouldn’t just focus on adding more and more automated scripts to your automated test solution, but also keep a close eye on the quality and the performance of your solution itself. This to prevent unnecessary maintenance effort, the possibility of your tests getting flaky and ultimately the demise of your solution itself. A very valuable insight and one that is easily overlooked.

One of the sildes from Mark Fewster's talk on test automation health

One of the sildes from Mark Fewster’s talk on test automation health

After that, it was time for the first break and a chance to talk to some of the other attendees. I then ran into Richard Bradshaw, also (or maybe even better) known as the Friendly Tester. We’d exchanged some tweets in the past and I am a reader of his blog, but I never met him in person. He was there to deliver a talk later that day, and we had a short talk over coffee. It’s always nice to meet people you’ve only known from the web in person.

After that, it was time for the first round of breakout sessions. I opted to go to what I thought was a workshop on agile test automation, hosted by the chairwoman of this year’s TAD, Linda Hayes. I selected this sessions since I like attending workshops much more than presentations that merely broadcast information. Unfortunately, it turned out I misread the programme, as this was a masterclass rather than a workshop, so there were no interactive moments during the session. Also, I was somewhat disappointed by the contents of the presentation itself, as I felt that the concepts Linda explained weren’t new, at least not to me. Still, I got a couple of small nuggets of wisdom from this session as well, some of which I might also be able to apply in my current project, so it wasn’t all bad.

After that: lunchtime! I spent most of the lunch break talking to people I hadn’t seen in a while (in some years more than five years) so that flew by. After lunch, another keynote was delivered by Marielle Stoelinga from Twente University (which I graduated from as well, coincidentally). She talked about applying model based testing to probabilistic software, something quite different from my daily practice of applying tools to test web applications and/or distributed backend systems. Since I still have a bit of interest in the mathematical background of computer science in general and software testing in particular left behind from when I was a student myself, this to me was an interesting subject.

Then, another round of breakout sessions. I first went to listen to Richard Bradshaw, who delivered one of the better talks of the day in my opinion. He’s a great presenter and managed to make his session on expanding the view on applying tools in testing (he talks about ‘automation in testing’, not about ‘test automation’) interactive as well. Very refreshing and something more people should do!

The second breakout session I attended was a short workshop on service virtualization (the other field I’m particularly interested in, next to test automation) using a web-based tool called StubUI. This workshop was delivered by Maarten Metselaar and Derk-Jan de Grood and introduced the concept and benefits of service virtualization. There was far too little time to go into any real depth, but I hadn’t worked with StubUI before, so I learned something new in this session as well.

The closing keynote for the day was delivered by Alexandre Petrenko. He talked about dealing with faults and uncertainty in software in a systematic way. This talk wasn’t one of the highlights of the day unfortunately. The concept – including an especially good bit about applying mutation testing to software behaviour models – was very interesting (although fairly academic), but the delivery of the talk wasn’t really on par with what I think a good closing keynote should be. It was all a bit dry, to be honest.

A slide from Alexandre Petrenko's talk. on mutation testing applied to behaviour models

A slide from Alexandre Petrenko’s talk. on mutation testing applied to behaviour models

Of course, when things get a little dry, there’s only one solution: after-conference drinks! After all talks finished and the conference was essentially over, I spent some more time talking to people, ice cold drink in hand (it was HOT that day). Looking back on the day, I wasn’t all that impressed by the quality of most talks, but meeting and talking to other attendees and some of the sponsors still made it a day well spent. With regards to the quality of the talks: there’s only one solution and that’s to put my money where my mouth is, so I’ll definitely try and get in as a speaker again next year!