Cart
[Home] [By Thread] [By Date] [Recent Entries]
> From: Micah Dubinko [mailto:MDubinko@c...] <snip/> > I just posted a reasonably detailed overview of the situation. > > http://lists.w3.org/Archives/Public/www-forms-editor/2002Jan/0005.html > > .micah That looks like that specifically addresses the issue of whether default namespaces apply to QNames in attribute values. I think Evan was asking a more general question regarding the use of QNames at all in attribute values (unless I'm misreading his post). Issue R-85 [1] is an interesting one, though. I personally think that the XML Schema approach, here, just lends itself to confusion. It should probably not use the default namespace and should require an explicit prefix. However, the schema approach is not really logically inconsistent with XPath or XSLT. The XPath and XSLT rules apply to Name productions, which don't have to be namespace qualified. The XML Schema reference, though, is explicitly regarding QName productions, which must be namespace qualified. It is not unreasonable to have default namespaces apply to QNames in attribute values, but not Names. However, that does lend itself to confusion and I'd be in favor of requiring explicit prefixes on QNames in attribute values in all cases to minimize such confusion. [1] http://www.w3.org/2001/05/xmlschema-rec-comments#pfiQName
|
