- From: Paul W. Abrahams <abrahams@valinet.com>
- Date: Tue, 06 Jun 2000 14:48:11 -0400
- To: David Carlisle <david@dcarlisle.demon.co.uk>
- CC: abrahams@acm.org, xml-uri@w3.org
David Carlisle wrote: > > None of this contradicts what is in the spec now. It presents a different way of > > describing what is there now. > > agreed. One almost suspects that the current wording is deliberately > vague to paper over cracks that arose due to earlier versions of this > discussion (but I wasn't there, so probably I shouldn't suggest such > scurrilous gossip:-) If we assume your hypothesis, then the cracks are now opening up wide. The only way to close them is to clarify what had deliberately been left vague. It's too late to shut the door now to arguing out all these issues. > > There are lots of such [unique ID] algorithms around, > > True although most of them don't lead to the easily memorable names > that using URI gives. That's true too. But perhaps this discussion is somewhat beside the point. I wasn't really arguing (in this context, anyway) that arbitrary unique IDs are the way to go with namespace names. I was merely making the point that if you go by what's in the namespace spec alone, there aren't any truly compelling technical reasons to prefer URI references to arbitrary serial numbers. There are other possibilities, I'm sure, for creating names that are both memorable and unique. After all, most book titles are unique if you ignore "Elements of Calculus" and ones like that. Paul
Received on Tuesday, 6 June 2000 14:48:29 UTC