Fri Jan 2 18:01:08 EST 2015

Too much

So lots of ideas came in today, but this is really too much.  Let's
split it up in two distinct parts:

- Write a simple round-robin scheduler in the DTC forth using proper
  tasks.  Possibly compliment that with static stack depth analysis to
  perform better allocation.

- Create partial actor evaluator that compiles actor networks to DAGs
  based on a known set of input events.