One of my classmates was telling me that it is an open question whether every 3-manifold can be triangulated. This was rather surprising. He said that the question as far as he remember is settled only for 4-manifold where answer is negative. If this is the case, can somebody shed some light why this problem is so hard?