This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Comment from a reader/user who misread the type hierarchy diagram (appearing in both DM and FO). See the thread starting at http://markmail.org/message/or2ialckrngn5xdk <quote> My mistake was I drew the wrong conclusion from the Type Hierarchy diagram: http://www.w3.org/TR/xpath-datamodel/#types-hierarchy. There is a line from xs:QName to xs:string on the diagram, which I was incorrect in thinking was xs:QName -> xs:string. In reality it is of course two lines merging that represent xs:QName -> xs:anyAtomicType, and xs:string -> xs:anyAtomicType. IMHO the edges of the graph on the diagram should state their direction. UML has the transparent arrow head symbol to represent the direction for this relationship. Something similar here would be a good idea. </quote> Another point about this diagram: I think it's really confusing trying to present two type hierarchies (schema types and item types) on one diagram. It would be much better to split them into separate diagrams.
Adding arrows may be a good idea, but I think showing the connection of schema types and item types in one hierarchy is important. The original diagram called the distinction out using shapes and colors... Maybe we could improve the visual clues, but I think it is important to show that atomic types are both while complex types and node types are just one...
After considerable reflection, the WGs decided not to make any changes to the type hierarchy diagrams (in either the Data Model document or the Functions and Operators document) with respect to the 2007 Recommendations for XQuery 1.0 and XPath 2.0. The WGs do not believe that it makes sense to revise the diagram retroactively for the Recommendations. Therefore, I'm marking the bug RESOLVED/WONTFIX (for the Recommendations) and trust that you are willing to mark the bug CLOSED. However, we have taken positive action for the XQuery 1.1 and XPath 2.1 drafts by splitting the diagram into three diagrams (one for "item", one for "xs:anyType", and the last for "xs:anyAtomicType"), with arrows "pointing" from derived types "up" to the types from which they are derived. Hope this helps.