johnswentworth comments on Some Rules for an Algebra of Bayes Nets