This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Currently, the value space of anySimpleType is defined to be the union of the atomic datatypes (which is the value space of anyAtomicType) and the set of all lists of atomic datatype values. This means that it contains lists which (by virtue of having one or more items all of whose lexical representations contain whitespace) do not have lexical values. There is some sentiment for allowing everything as values for anySimpleType (including you, me, and my pet cats), regardless of whether or not they occur in the value space of any primitive datatype. There is also some sentiment for doing the same to anyAtomicType. I believe that if we do that for anySimpleType, we should also do it for anyAtomicType. (This would also imply that lists are anyAtomicType values!) If we choose not to do that, then we should also not allow those lists which have no lexical representations.
This appears to be a near duplicate of bug 3243, and should be discussed together with it and with bug 3025.
Any wording proposal for this should be coupled with resolution of bug 3243 and bug 3025. Accordingly, I'm marking this needsDrafting to agree with those issues.
A wording proposal intended to resolve this issue is available at http://www.w3.org/XML/Group/2004/06/xmlschema-2/datatypes.b3025.html (member-only link).
The wording proposal mentioned in comment #3 was adopted by the XML Schema WG on its telcon today. Accordingly, I'm marking this issue resolved. Dave, as the originator you can signal your assent to this decision by changing the status of the bug to CLOSED, or your dissent by reopening it.