Niedersachsen testing day
Is there a better way to start writing a blog than reporting about a conference?
Since it was held in my neighborhood, I decided to visit the Niedersachsen testing day this week, which was organized by the local chapter of the ASQF.
And what can I say? Overall it was time well spent.
There were six talks of 45 minutes each.
The first one was about risk-based test case priorizing. An alternative for priorizing test cases was introduced. It was said that the usual categories of 1, 2, and 3 are not sufficient, especially when you want to decide which tests to run when there is not enough time for all (which, as we all know, hardly ever happens…). The speaker actually puts Hans Schaefer’s formula into use. Quite interesting, but nothing really new.
The second talk dipped into model based testing using risk based, requirement-driven and security aspects. The part I found most interesting was about Behavior Engineering, which deals with transforming written language into different graphs. Something we did for years at university (x-bar, anyone?), but I guess those graphs were quite different. This was actually the talk I was looking forward to most and it did not disappoint me. Unfortunately the lecturer had to leave early because she had an appointment with the Microsoft CEO, who somehow seems to be more important.
The following talk was the one giving me a headache. Literally that is, as there was quite some acoustic feedback. The German Testing Board presented its new certificate: CTFL(R)-CAST: Certified Automotive Software Tester, which is an extension to the foundation level certificate. Furthermore some information about the processes at ISTQB were giving. Since the GTB was a sponsor of the conference, I couldn’t really shake off the feeling that this was a promotional talk. The following discussion seemed to reinforce that and reminded me a bit of #STOP29119
And now to something completely different: Lunch! Food was alright, though discussions and small talk were better.
The first talk after lunch dealt with efficiency. Dealing with variants and reducing the number of tests were the main ideas. Some scholastic methods and evidence were given for combinatory feature testing, especially highlighting the ICPL algorithm. Overall this was the best talk in my opinion.
A talk about cooperation of technical systems followed. The emphasis was about the connection of and communication between different systems, in this case traffic control systems, and how these are tested using a model based approach. This one was nicely designed as well.
The last presentation was from one of the sponsors again. But this one was rather interesting as a MBT tool was introduced. And in contradiction to the more theoretical talks before, this was more on the practical side and quite interesting. Besides introducing the tool, testing in the cloud was a key feature. While in the earlier talks the number of test cases were supposed to be reduced, this one focused on increasing the resources via cloud testing, so that all tests can be executed in an acceptable period of time. So that’s certainly a different kind of approach, which of course only works for automated tests (never mind what your stance is towards that topic…).
I had to leave afterwards. Therefore I could unfortunately not follow the open discussion after the talks.
Strikingly, most speakers agreed on requirements and the ambiguity of language as the main reason for software products getting out of hand. Their conclusion was mostly to formalize language. Something I do only in part agree with, but that will be a topic for itself some other time.
It certainly wasn’t the biggest conference held this week with the Agile Testing Days, Öredev and Lean Kanban Central Europe Conference all taking place this week, but all in all it was an interesting experience for me. I got some valuable insight into MBT and in general some interesting input on different topics.
