[[ Whoops; I saw Scott's message after I sent mine ]]
At 7:57 PM -0500 11/10/04, Scott Hollenbeck wrote:
The reason that this sort of reference has been a problem for the IETF is that it's not a stable part of the standard. The content can change over time, with no community control.
Some people here consider that a feature, not a problem. :-)
Plus, the primary purpose of the URI in this context is to uniquely identify a namespace -- and no more per recent IETF practice.
Not sure how to parse that.
I know that some view the ability to dereference a URL as one of the strengths of the http URL approach. I can guarantee, though, that it will raise issues that we really don't need to get involved with when it comes time for broad review of the specifications. As I told Tim in our off-list conversation, I'm not going to argue that the working group can't use http URLs to identify namespaces. You're just going to have to come up with a VERY compelling reason for their use -- and the argument that they can be used to provide a pointer to additional information is probably NOT the one you want to use for just the reasons I provided above.
Well, so far we haven't found other "very compelling" reasons.
--Paul Hoffman, Director --Internet Mail Consortium
