At CAST 2011 Henrik Andersson and Leo Hepis led a workshop session on the relevance of context during testing planning.
Henrik introduced the workshop stating that he has to adapt to different contexts quickly as he is a consultant in software testing and gets paid by the hour. The session was constructed around a simulation on context.
Henrik gave the definition of context being “everything that can change in a product I am considering.” During the session we were going to experience context changes that will help us reveal context on a meta level.
In the simulation we collected different questions around the context of the product. We had a remote operated robot to explore hidden chambers in pyramids in a traditional waterfall company given a full year of preparation. The following shows our results on that.
We debriefed some of the several groups that had similar missions. Some created mind maps to get there plan down, while other collected single sentences.
Then we got new context information. We were informed that our product would be used in a televised setting on December 12th 2012 only by 8 year-olds. We added new considerations to our previous brainstorming, which I colored in blue.
We went through another round of debriefings hearing from the other groups struggling with such changes in requirements.
In a third round we got some other deadline (on year earlier) and updated our plan again based upon that new information. The new thoughts are colors yellow in the following flip chart picture.
We went through a final debriefing in the workshop.
Henrik and Leo finished the workshop with a list of different triggers for context. The different triggers were splitter into our own context, the context of the product, and the context of others. They had provided a set of context triggers for each category. We put our thoughts into the different categories.
Henrik was going to pass the final list around. If you would like to see the final result, drop him a line. I can provide you his email if you like to.
