I made it run only on every 300th draw (mostly a random choice) and it dramatically reduced CPU usage, increasing the FPS by 3-6 times depending on the machine I ran it on. I let it run for two hours without any problems on my laptop. I have to admit that I forgot to track the memory usage though. :)
Why is the implicit garbage collection needed at all? Shouldn't Processing/Java clean up automatically? Were there problems?
I am about to implement it at home with the Tactile Matrix and the projector, but before that, I would like to know which is the most stable version of the Cityscope repo?
The core features of 'AndorraPlayer.pde' have been moved to 'AgentDemo.pde' located in 'CityScope/Misc/AgentDemo/'
The goal is to polish AgentDemo.pde into an excecutable application that we can distribute to users.
Through the application, the users should be able to independently understand the logic of (a) agent generation and (b) pathfinding of those agents from A to B.
Currently, the UI is a mess of keyboard commands that Ira merely created for debugging. The UI also needs much more text and explanation. Imagine your grandma will download this app. Can she play with it to better understand how we do agent-based modeling of origin-destination data?