khafra comments on Bayesian Inference with Partially Specified Priors