I will review (mostly) old results and (mostly) recent conjectures relating modular representation theory to geometry.