Comments on: A Gibbs sampler for a class of random convex polytopes

In this comment we discuss relative strengths and weaknesses of simplex and Dirichlet Dempster-Shafer inference as applied to multi-resolution tests of independence.