sbechtel hat geschrieben:The purpose of this task is not clear to me. Just writing some test objects with some combination of traits and a dummy call to measure are not that enlightening. In fact this only shows that objects with some combinations of the stackable traits are runable. To really use the framework one would need to use the generated data which is not the case and to show what is possible and how things work is done in the test cases... Could you please motivate the task a bit?
I am not sure, what I am supposed to motivate, but I will do my best
This task actually is supposed to make you actually use
your framework. Writing programs that actually tax the measure method in a memory or a time consuming way, like the example which we gave you. This example allocates a large array, so the value that is serialised for a MemoryEvaluation should obviously be in the right ballpark.
Also this should make it easier to answer the last question, as you can easily explore what impacts the mixin order has to the test results.