I spent all night tonight reviewing my collection of π-Calculus papers and thinking about the ways in which one could implement an extremely simple π-Calculus for Clojure. In part, this involved reflecting on Erlang's general architecture (from the perspective of communicating nodes and processes).

I think I've got a plan for tackling this in what is essentially a "π-Calculus lab" – two JVM instances and a simple distributed messaging/storage solution. This would allow mem to simply focus on the mechanics of communicating processes/functions without getting bogged down in developing a system. That can come later, once various avenues have been explored in the "lab".

I'm pretty stoked to get started on this :-)