http://www.w3.org/Bugs/Public/show_bug.cgi?id=3241 ------- Comment #2 from mike@saxonica.com 2007-04-20 16:18 ------- >It's not clear to us what exactly one would be seeking to accomplish by defining anyListType or anyUnionType. Two things, I guess: (a) QT languages currently provide no access to information about types other than their names and their place in the type hierarchy; so the proposal would provide more information to users of these languages without inventing new mechanisms for access to metadata. (b) Even if access to metadata properties were available, it wouldn't fall into the type system: one wouldn't be able to define a match pattern in XSLT that matches all list-valued attributes. But the comment was phrased more as something to consider in order to maintain symmetry and tidiness in the type hierarchy given that anyAtomicType is being added at the same time. There won't be any push-back if it's dropped.Received on Friday, 20 April 2007 16:18:48 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 23 October 2007 06:13:41 GMT