- From: Martin Gudgin <martin.gudgin@btconnect.com>
- Date: Wed, 24 Apr 2002 03:13:31 +0100
- To: "Ray Whitmer" <rayw@netscape.com>
- Cc: <xml-dist-app@w3.org>
----- Original Message ----- From: "Ray Whitmer" <rayw@netscape.com> To: "Martin Gudgin" <martin.gudgin@btconnect.com> Cc: <xml-dist-app@w3.org> Sent: Wednesday, April 24, 2002 1:58 AM Subject: Re: Summary of Issue 194 - encodingStyle > Martin Gudgin wrote: > > >----- Original Message ----- > >From: "Ray Whitmer" <rayw@netscape.com> > >To: <xml-dist-app@w3.org> > >Sent: Tuesday, April 23, 2002 6:47 PM > >Subject: Re: Summary of Issue 194 - encodingStyle > > > > > ><SNIP/> > > > >>Actually, some places you need to know a namespace binding are clear, > >>and others are not. I am not sure that it is ever clear that you need > >>xml:base. Especially not at the raw infoset level. > >> > > > >Actually the infoset carries a base URI property for Document, Element and > >Processing Instruction Information Items. So xml:base would modify the > >infoset. > > > >Gudge > > > ><SNIP/> > > > Yes, obviously it does modify the infoset, and I didn't claim otherwise. > encodingStyle also modifies the infoset. Only in that it is itself an Attribute Information Item > You could also trivially > produce an expanded infoset with an encodingStyle value on every node, > as is done with xml:base. Yes, but [base URI] is actually an infoset property ( and may for some infosets not be the value of an xml:base attribute ) > But it still says nothing about what effect, > if any, setting xml:base has on the interpretation of particular values > or nodes, > just like encodingStyle may affect the interpretation of > values but there is no obvious interpretation of where it applies and > where it does not. > The application probably decides that ultimately for > both. > xml:base does not automatically apply, modifying the infoset, > just because there happen to be URIs in content. The same content may > be interpreted in different ways by different applications. > > This is the same with namespaces. You do not really know which content > happens to have qnames in it that will rely on the declaration for > interpretation. For this reason, it does not directly affect the > infoset of such since that can only be determined by applying a schema > or other processing, as with the use of an encoding. Agreed. The interpretation of attribute values and element content is at a higher level than the infoset. And interpretation of any given value in the infoset may be determined by the some other value in the infoset. I think I've lost track of where we're going with this... Gudge
Received on Tuesday, 23 April 2002 22:13:05 UTC