[Shannon Williams (2011.08.03.1300 CST)]
I am thinking that there are two distinct architectures for assessing
When simulating, the system simply iterates through a list of
controllers and each controller deposits money and maybe collects
objects from a predetermined list of other controllers.
When assessing- The system first groups all controllers with the same
(to be determined) properties. The behavior determined for a
controller will automatically be applied to all controllers with the
same properties. (This saves time)
In the first versions of the app, I am thinking that controllers would
come in two forms: Individual and corporation (government falls under
corporation). All links between controllers are created and updated
via configuration. So, in the first version each controller has
predetermined set of behaviors that it can work with.
In later versions we can figure out how to generate new behaviors
(find new customers or start new corporations, etc).
How does it sound so far?