- From: Daniel Veillard <daniel@veillard.com>
- Date: Fri, 12 Nov 2004 16:25:34 +0100
- To: Norman Walsh <Norman.Walsh@Sun.COM>
- Cc: public-xml-core-wg@w3.org
On Fri, Nov 12, 2004 at 06:56:26AM -0800, Norman Walsh wrote: > The XLink Rec, as written, requires the xlink:type attribute on all > XLink elements[1]. It's the xlink:type that tells the link processor > what to do. When XLink was developed, I think we imagined that a DTD > default could almost always provide a value for xlink:type and authors > wouldn't ordinarily have to provide it. > > But in reality, documents don't always have a DTD, and even when they > do, lots of applications don't read the external subset. That means > that authors can only rely on conformant XLink processors to do the > right thing if the author provides the xlink:type attribute on every > XLink'd element: > > <link xlink:href="http://..." xlink:type="simple">...</link> > > In retrospect, it seems obvious to me that an element that has an > xlink:href but does not have an xlink:type should be treated as if it > had a "simple" link type. Sounds right to me. > It seems like a simple, easy change to make in a 1.1 spec. That > probably means it's impossible :-), but does anyone else have the will > to try? I would feel that it's the right thing to do technically, but politically: - Are we (Core WG) responsible for XLink ? - Do we really want to reopen this, I already see the namespace prefix question raised again ! The scars are not fully healed yet from my perspective... Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ |
Received on Friday, 12 November 2004 15:23:16 UTC