- From: Daniel Veillard <veillard@redhat.com>
- Date: Sun, 13 Feb 2005 06:26:33 -0500
- To: Elliotte Harold <elharo@metalab.unc.edu>
- Cc: www-tag@w3.org, public-xml-core-wg@w3.org, public-xml-id@w3.org
On Sat, Feb 12, 2005 at 08:58:30PM -0500, Elliotte Harold wrote: > 3. Even if there is an xml:id layer, and the xml:id is inherited by > several different elements, some high performance streaming solutions > may stop as soon as they've located the first instance of the ID, and > never notice that there are duplicate values further along in the document. I'm personally against relaxing that checking myself. If you use xml:id on a receiving side, this mean your application is interested in providing ID lookup, and IMHO errors should be reported to the application. You should have the capability to desactivate xml:id if gathering the ID table is too much for your memory constraints, but loweing the spec level of checking is IMHO not a good idea. Daniel -- Daniel Veillard | Red Hat Desktop team http://redhat.com/ veillard@redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/
Received on Sunday, 13 February 2005 11:26:35 UTC