(via Slashdot)
Robin Milner talks about program proving, Caml, petri-nets, CCS, pi-calculus, bisimulation, whether CS is a "real" science and many other interesting things.