Challenge 0 - Due Friday,October 5
A First Attempt at Blocking Out a Multiagent
Simulation
Do as much of the following as
you can without getting bogged down. Please give it an honest effort. If you
can write the simulation, then do so. If not, think about how you might go about
it. In a sense your response to this challenge will help me get to know you
better so that I can tailor your individual experiences to better correspond
to your skills and interests.
Some of you have never programmed before and some of you are experienced programmers. That doesn't matter; I have different expectations for each of you. If you have never programmed before, see if you can isolate what is going on with each agent. What does each agent sense? What does it think? How does it act? How do the different variables affect its behavior? This is the first step in analyzing any social situation: blocking out a program to explain what is going on. Play the role of an ethologist or ethnologist. Try to figure out how each agent relates with its neighbors, how things change through time and space, and finally how all this behavior is visualized on the screen. Now put your ideas to the test. If you think you understand what is happening, then try to build your own model of Wator from the bottom-up. Consider yourself as a director of this little movie. What scripts and instructions would you give to each actor? What is each actor's goal and motivation? How would you get the action on the (computer) screen? What directions would you give your cinematographer, set designer, editor and writer. As a first step, sketch this all out using whatever language you think is appropriate to tell the story? This intermediate language is called a "pseudocode." If you have programming experience, see how close you can get to writing code to make it happen. Finally, try writing a Borland C++ Windows application for this simulation.
Be prepared to talk about your progress early next week.