Namespaces are not broken, and don't need fixing.They aren't broken in the sense of being technically defective and unworkable. But the very existence of the active and sometimes vociferous debate in this forum, with many conflicting views being expressed along with a great deal of confusion, indicates that something is seriously wrong with the namespace spec.
The fundamental problem with the namespace spec is that it creates misleading expectations. That was the whole point of the quacks-like-a-duck discussion. It uses URI references to name namespaces, but (unless I'm missing something) makes no use of any properties of URIs beyond their being presumably unique character strings. The namespace spec would work just as well if namespace names were required to be serial numbers generated by some algorithm given in the namespace spec itself.
The namespace spec provides no statement whatsoever, except for an odd disclaimer pertaining to schemas, about what kinds of URIs are appropriate, what resources they might identify (and a URI by its definition identifies something), and to what uses the namespace name might be put, if any. It provides no rationale for the choice of URIs as namespace names.
It does state that a namespace name, to serve its intended purpose, should have the characteristics of uniqueness and persistence. Uniqueness is clear enough and obviously a Good Thing. But persistence? What does it even mean for a URI reference (as contrasted with a resource) to be persistent? That statement is especially puzzling in the context of the many statements made about the namespace spec (but not within it) that for the purposes of the namespace spec, the resource identified by a namespace name - indeed, its very existence - is immaterial. That URIs should be persistent seems to contradict the statement that it doesn't matter what they identify. Character strings are persistent independent of what they happen to mean. The character string "Today is July 4, 1776." is the same string that it was (if the American Founding Fathers thought of such things) at the time of the American Revolution.
Any revision of the namespace spec should make it absolutely clear what is expected of a namespace name in terms of what it identifies or might identify and why URI references are used for this purpose. The treatment of the case where distinct identifiers identify the same resource should also be spelled out explicitly. (Here my opinion lies with the majority: only the identifier matters, not the resource identified.)
So the following facts about namespaces should remain true after anyThese all seem reasonable enough to me. I won't replicate them here.
re-issue of the namespace spec.
Paul Abrahams