Search

Scientists from various domains resort to agent-based simulation for a more thorough understanding of complex real-world systems. We developed the Agent Visualization System; a generic system that can be added to a simulation environment to enrich it with a variety of browsers allowing the modeler to gain insight into his simulation scenario. In this paper we discuss how the various features of the Oz language and the Mozart platform aided us in the development of our system. Of particular importance were dataflow variables, high-orderness, the support for distribution and concurrency, the flexibility offered by QTk which was crucial in generating browsers whose structure is only known at run-time, in addition to a miscellany of features that were conductive to our work. We also highlight some of the implementation difficulties we faced and explain the techniques we utilized in overcoming them. [ABSTRACT FROM AUTHOR]