Sean McGrath has written an article entitled A study in XML culture and evolution. It starts out well, making interesting observations about the differing culture between "document" people and "data" people. "Data" people believe in unique ids (names) for data, "document" people are satisfied with uniqueness among all the fields (addresses). Good point.

Being a "document" person, he doesn't see the need for unique ids for his data. Fair enough. But then he tanks, making some sort of leap to the conclusion that since his data doesn't need names, his XML doesn't need namespaces. Huh? I'm hoping Sean was mis-edited, or maybe just had an off day. He seems otherwise to know what he is talking about. To confuse names for data with namespaces for names of attributes seems pretty basic to me.