- From: Roy T. Fielding <fielding@apache.org>
- Date: Wed, 30 Oct 2002 21:13:12 -0800
- To: gtn@rbii.com
- Cc: <www-xml-linking-comments@w3.org>
> No doubt you have more than just opinion behind your comments, but it is > not > apparent from your comments. Can you give more substantive technical > backing > such as the above? Try placing a URI with " in it within an anchor href. And explain to all of the people who parse URI in plain text that they should not stop at the first delimiter. > In particular, your comment about being media-specific is interesting, but > your alternate "media independent" identification methods (search terms, > paragraph text, and regular expressions, all of which boil down to > queries of > a form) have serious issues as well, most obviously, lack of uniqueness First-match is unique and well-understood by users. > (something that has plagued critlink in the past). ID, as you note, is a > reasonable approach, but given your (assumed!) criteria, is also > necessarily > media dependent... being tied to a media type that has some intrinsic > notion > of what ID is. No, what the ID means is media-type specific. The actual string of text in the fragment is the same regardless of media type. ....Roy
Received on Thursday, 31 October 2002 17:38:08 UTC