This entry will sum up on my last semester and reflecting on some of the comments from the last blog entry.
After the project during the fall semester 2012, where there was a huge work pressure due to individuals enforcing a constant day to day work pace and due to the agile Scrum being abused. I had too little time to read for my courses and also caused a depression in my exam period.
This semester I grouped up with some other students, who all are interested in doing a special in game development. While this group were less quality focused, it provided a much larger flexibility, but also more respect towards individuals and their relationships. It is not a secret that relationships requires time and care, but many students do not embrace a relationship or just haven't found one, while they are studying.
This flexibility caused us to work more loosely than what I have encountered in any other project. Mainly we worked from 9 - 16 , sometimes less. In the start of the semester, one was first meeting at noon. We didn't like it, but instead of punishing him, we tried to work with him as best as we could, and eventually he also joined us in the "morning".
This semester, I was sharing the leadership with a second student. On pair work him and I worked really well together, meanwhile we kept a good teamwork a motivation overall. We took a bit extra work if necessary, to avoid cumbersome parts for the to others in the team. While I knew one of them and the other guy knew both of them, it was easy for us to make sure we worked together as a team, without anyone getting an annoying task.
In our semester we combined two projects, to make dynamic map generation and dynamic strategy selection based on utility maps. Where we defined utility maps as a combination of heatmaps representing the environment. (as compared to utility maps from Machine Intelligence, we use both the high values and the low values)
By using Google Maps, we could retrieve satellite images and environmental data to represent the level with roads, water, grass( anything else) data and altitude data to give the 2D game a third dimension. Furthermore we made a weak map segmentation to locate houses, and use them as obstacles in the generated map.
To create an example and test our work, we made a small Pac-man game.
Our game was tile based, as it would simplify the calculations for the pathfinder; however the movement was not title based, but utilized every pixel. It can be seen in the two images here:
Image of the actual game
Image illustrating tiles in the game, which is used for pathfinding (obstructed / non obstructed)
While we focuses on the production in the game, we didn't reflect on the scientific requirements for this semester. Hence we received a lower grade than we expected, but all in all it was a great project, which we learned a lot from and will be able to use at later stages. We consider to fix some bugs in the game and then release it for free.
Within a year from now, I expect that BZSmod will be released, since I will have a lot of time to work on the project. I have not decided what will happen if we don't make a release before the summer 2014.
