Over the past few months I have been chugging my way through the videos of this meeting of computer science luminaries. Gérard Berry is very funny. (I must watch the other videos on his home page.) Gordon Plotkin's talk is quite abstruse even by the standards of this audience. Philip Wadler shows how to run a panel. Gérard Huet and Larry Paulson give great accounts of the early days of interactive theorem proving systems, and John Harrison shed some light on what he does at Intel.
Milner's last innovation in his process algebraic tradition — bigraphs — don't look like gaining much traction from what I saw here.