the test bots are supposed to launch, but they don't set up listening ports yet; that's the next step

this is also not exactly the same folder hierarchy as the framework will eventually use.  the folder containing the musebots should really be at the same level as the folder containing the conductor


===

- i sort of want the option of killing the bot either from the bot or from the conductor
	- in fact, i want to be able to trigger anything from an individual bot that i can from the conductor
	- that way the conductor interface would be optional
- how important is the individual-port model for the bots?

===

HAAAAAAAAAAAAH!!!!!
udp SEND ports can only be bound to one application at the same time
but udp RECEIVE ports just wait around for anything to send to it
it's the binding of the SEND port that is special and unique; hence the send needing to be supplied with the host as well
	(the host isn't really hosting the send; but the send needs to know the host that is hosting the receive it's aimed at)
RECEIVES ARE TARGETS; SENDS NEED TO BE AIMED

the rules are a bit different when all the ports are inside the same application