ACM ICPC World Finals 2011
Despite being the only real computational geometry problem in an unusually geometry-free problem set, this problem is not too hard. Without loss of generality, one can assume that in an optimal solution there are two verties of the polygon touching one of the two sides of the trash chute, so we’ll try all such pairs of vertices and figure out the resulting chute width (which is easy).