> However, note that my comment on Glenn's question regarding a stdlib > addition of a tree type still applies I agree with all that. Having a tree-based mapping type in the standard library is a different issue entirely.