Post-mortem analysis of student game projects in a software architecture course

In 2008, a game development project was introduced in a software architecture course at the Norwegian University of Science and Technology. The motivation for introducing the project was to let students learn how software architecture practices and processes can improve the final product in an inspiring and practical way. In the project, students organized in groups had to establish functional and quality requirements, design the software architecture of the game, evaluate the architecture, implement the architecture and test the architecture. After completing the project, all groups had to perform a post-mortem analysis of the project to reflect on the positive and the negative issues related to the project. This paper summarizes and describes the results of this post-mortem analysis along with the students' experiences from performing a post-mortem analysis of a game development project. The results show that there are both positive and negative effects of teaching software architecture in the context of a game development project. Students found it motivating to learn about software architecture through game development, but some students found it hard to apply the theory when developing the game. Most students were very positive to learn about new game technology as a part of the course and it was very stimulating to create an actual product. The main complaints were shortage of time, that many found the evaluation of architecture (ATAM) worthless, and that the project demanded too much documentation. Most students commented positive on doing a post-mortem analysis as a part of a game development project.