On Sun, 4 Jun 100, John Cowan wrote: > I agree that the relationship is one-one (and even onto), but I deny > that it amounts to identity. Confounding the namespace with the document > that describes it is a map-territory error. Right on. We are identifying a namespace resource! We are not identifying a hypertext resources which describes a given namespace -- this is a seperate problem entirely. It is unforunate that the current spec allows this; I now agree that the xmlns attribval *should* be a URI, (Tim has convinced me of this much)... However, I think it should not be any old URI from any old scheme. Fellas, we need a URI scheme specifically for namespaces. Let's deprechiate duck quacking. Best, ClarkReceived on Sunday, 4 June 2000 15:24:18 GMT
This archive was generated by hypermail 2.2.0 + w3c-0.30 : Tuesday, 12 April 2005 12:17:24 GMT