- From: Daniel Veillard <Daniel.Veillard@w3.org>
- Date: Mon, 26 Apr 1999 06:00:48 -0400
- To: "Joel A. Nava" <jnava@Adobe.COM>
- Cc: www-xml-fragment-comments@w3.org
On Fri, Apr 23, 1999 at 02:33:57PM -0700, Joel A. Nava wrote: > Daniel, are you saying that a fragbodyref be limited to > only being a URL? Oops mistake, not URL, but URI, and as James pointed out URI Reference seems the appropriate term. > As far as the spec goes, a reference is very general, and > fragbodyref examples, and prose on the subject show them > to be URIs, URLs, and other types of references. Since > it is already allowed, I don't see the need to restrict > a fragbodyref to URLs only. Restricting it to a known kind of reference allow to anticipate on the processing needed to handle the fragment on the receiving side. In the case one use a packaging mechanism, having a strutured naming scheme is important for the evolution of the specification. Expressing fragbodyref references as URIs allows to bypass the packaging mechanism, and provide a standard way to use Fragment in the absence of a Packaging format. This is useful when the server has no idea on the client capacity to decode say Mime multipart messages. Daniel -- [Yes, I have moved back to France !] Daniel.Veillard@w3.org | W3C, INRIA Rhone-Alpes | Today's Bookmarks : Tel : +33 476 615 257 | 655, avenue de l'Europe | Linux, WWW, rpmfind, Fax : +33 476 615 207 | 38330 Montbonnot FRANCE | rpm2html, XML, http://www.w3.org/People/W3Cpeople.html#Veillard | badminton, and Kaffe.
Received on Monday, 26 April 1999 06:00:56 UTC