TODO Hackathon:
-
Implement a timer for the agents inside the Java server -
Debug the Java server(OK for now) -
Save the output files in the Java server -
Implement a game visualizer -
Implement a GameSimulation object in order to easily run many simulations in parallel -
More game loop steps should be done between the commands asked to the AI. As per now, 4 steps per second create collision glitches. -
Implement an 'Organizer AI' (@stefano).
-
Create a python script for pool simulation using the openshift cluster.
-
Set NO winner upon equality inside java client. -
Finish game gracefully at the end of the commands in replay mode -
When there is an agent's timeout, the python agent needs to be notified and interrupt the my_ai function -
If time allows it, remake the physical system -
Debug pipe problem where the pipe hangs when there is too much data