One down, two to go.
Went to the logic colloquium. It was a talk by Masahiko Sato
on a vaguely LF-like system, except that it seemed to be
very sparse on types. There's a subtlety involving the encoding
of intuitionistic propositional logic I still don't fully
understand -- somehow A prop is equivalent to A true v A false
(where the disjunction is at the meta-level) and yet the logic
is not classical.