- From: Roy T. Fielding <fielding@apache.org>
- Date: Thu, 2 Oct 2003 01:12:01 -0700
- To: Patrick Stickler <patrick.stickler@nokia.com>
- Cc: <uri@w3.org>
> I don't see why all of the above cannot be addressed using > http: URIs. I hesitate to even enter this discussion, but whether or not every identifier in the universe can or cannot be addressed by using an "http" URI, it is only reasonable to state that opinion once and move on. If folks want to define a hundred new schemes, it won't change how the Web works, nor is the Web actively harmed by new schemes that are never deployed on the Web. I have no problem letting the market decide such issues, providing that no standards organization defines any scheme (including "http") as the one and only right way to identify things. There is one other item that needs to be mentioned, which seems to have been ignored by all of the proposals for meta-schemes: any URI, no matter how abstract its referent or how obscure the scheme, can be placed in the context of a dereferencing system that supplies representations of whatever is supposedly identified by that URI. All it takes is one HTTP proxy and the willingness of someone to do the mapping. As soon as that happens, people will start using that URI to indirectly identify various aspects of those representations. In other words, claims that the addition of a new scheme will somehow "solve" the "car" versus "description of a car" problem is ludicrous, at least for any scheme that is ever sufficiently useful to justify deployment. Cheers, Roy T. Fielding, Chief Scientist, Day Software 5251 California Ave., Suite 110 Irvine, CA 92612-3074 fax:+1.949.679.2972 (roy.fielding@day.com) <http://www.day.com/> Co-founder, The Apache Software Foundation (fielding@apache.org) <http://www.apache.org/>
Received on Thursday, 2 October 2003 05:10:13 UTC