The specification for this test may look frighteningly long, but the students were given a short introductory document the night before the test. This definitely helped. A substantial number of students got most of the way through Part IV, with nine out of 139 student scoring full marks. I decided to allocate 32 marks, capped down to 30, because I was worried that the test was too long. In hindsight there was no need for this, as the average mark was around 78%.
You will notice the ‘mystery’ Part V, which I decided to include because we had had so much fun working it out ourselves when designing the test! We had no expectation that any student was going to get that far, but one did come perilously close to completing it within the three hours.
I have made only one significant modification to the original version of the test, in order to fix an edge case in the construction of the cartesian product required in Part IV.
There really is only one useful source of additional information, which is the text book upon which the test was based: Concurrency – state models and Java programs (2. ed.)., 2006, by Jeff Magee and Jeff Kramer, published by Wiley, ISBN:978-0-470-09355-9.