- From: Tim Bray <tbray@textuality.com>
- Date: Tue, 18 Mar 1997 10:05:45 -0800
- To: w3c-sgml-wg@w3.org
At 12:51 PM 3/18/97 -0500, Gavin Nicol wrote: >>I just did some tests and it appears that the fragment identifier (the text >>following the hash on the URL) is even passed to the server We assume he means 'NOT even passed to the server' >Right. We looked at the problem of subdocument addressing in >DynaWeb a number of times. You cannot use fragment specifiers >in URL's if you want server-side processing. In XML, we do *not* want to build the server-side/client-side distinction into the link syntax. We want to attach a URL and a TEI xpointer and let the server & client figure out what to do. If we want to force server-side processing, the '?' syntax is available. There is nothing in the current URL RFC that allows a resource/fragment partitioning without also forcing the process model. In my view, this is purely a bug. Either we introduce a new syntax, or we overload '#' and let the market deal with it. -Tim
Received on Tuesday, 18 March 1997 13:28:58 UTC