Re: Event-Oriented Computing

[Marcus G. Daniels]

>>>>> "GR" == glen e p ropella <address@hidden> writes:

GR> You put ALife into the Swarm kernel that hunts down and destroys agents
GR> that use system calls and I'll put ALife in there that tries to
GR> use system calls but obfuscates the fact that it uses system
GR> calls. 

I expect that much reckless experimentation will end in failure, and
so it might be useful to install in the simulator some pattern
detection for common failure syndromes, e.g. CPU-sucking NOP loops,
suspicious use of system calls, or more conceptual level detection
like for tragedy of the commons.  A library of response strategies
(either automatic or triggered by experimenter confirmation) for
bringing order to the experiment might be helpful for debugging models --
being able to backtrack and not having to start over with a new set
of parameters.  An immune system for Swarm.




                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================