Lists Home |
Date Index |
> From: Bill Lindsey [mailto:email@example.com]
> I'm not convinced that using xsi:type does base the
> processing model on W3C XML Schema's notion of types.
> It simply makes use of the declaration of an element's
> having a type of a given name.
> If you can make the argument that W3C XML Schemas says
> the type or its name *is* the the W3C Schema type definition,
> I'll gladly change my position.
> I was (perhaps mistakenly) approaching it as though W3C XML
> Schemas says "Here's how you can assert that an element
> is of a named type when that type name doesn't match the
> element name."
> And while W3C XML Schemas says "here's how you can use
> type names to validate documents with W3C XML
> Schemas", I don't think it claims ownership of either the
> notion of types or the meaning of type names any more
> than it claims ownership of element names.
XML Schema has a very explicit notion of what is a type, and that notion is
quite unique to XML Schema. Other specs (e.g. the XML Namespace spec) use
the term "type" to mean something completely different. The xsi:type
attribute is quite specifically coupled with XML Schema's notion of types.