Well, exam is the 27. so starting to prepare for that sounds like a good idea. And after a nice weekend me and MadsG met monday to discuss what needed to be done. The most important thing to show was quickly decided to be examples of our technology working.

So i set out to write a small graphical tool to show how an agent talks. The tool is now done and gives the user possibilities to: * Load/Save agents from file. Add/Delete beliefs from a knowledgebase. Modifying the dissonance table. Modigying the personality of the agent. And uttering one or more beliefs, and seeing the line that is uttered. Screenshot attached.

My work on restructuring libPathFinder (which is at a stand-still at the moment due to the mentioned exam) is turning out to be a bit more undertaking than what i initially thought. I found out that what the old A-star algorithm did was actually not A-star but rather Dijkstra graph searching. But while restructuring the code to do A-star i started finding different places where a little code-change would yield a different search method, or unit movement. This sounded like a job for C++ templates, and my hope is that when the code is done, it can do shortest path searching in normal matrix like maps with units moving diagonally, 4-directional or even like a chess-piece like a horse. It should also be able to support searching in hexagonal map, without any modification at all. Only a new class for finding adjecent nodes are needed.

More on this later, when time allows it.