- From: Richard Tobin <richard@cogsci.ed.ac.uk>
- Date: Wed, 3 Jan 2001 15:21:34 GMT
- To: james.anderson@attglobal.net
- Cc: www-xml-infoset-comments@w3.org, w3c-xml-core-wg@w3.org
> 1. why don't attributes have a parent? They have an [owner element] instead. This corresponds to attributes not being considered [children] of their elements. > 2. why are expanded entity references replaced by their string > equivalent and then marked with start and end markers? Presumably you mean whay do we do it that way rather than having entity information items containing the content. The idea is that the 90% case is to access documents with all the entities expanded. It should be *possible* to reconstruct the entity structure, but it must be *easy* to access the logical structure. Both these issues are discussed in the XPointer - Infoset liaison statement, http://www.w3.org/TR/NOTE-xptr-infoset-liaison. -- Richard
Received on Wednesday, 3 January 2001 10:21:41 UTC