Insights from Munkres' Topology — LessWrong