2306 2005-09-26 15:28:39 +0000 Union of anySimpleType 2006-03-21 00:39:44 +0000 1 1 1 Unclassified XML Schema Datatypes: XSD Part 2 1.1 only PC Windows XP RESOLVED WONTFIX resolved P1 normal --- 1 ht cmsmcq www-xml-schema-comments oldest_to_newest 6456 0 ht 2005-09-26 15:28:39 +0000 Does the spec. need clarification wrt the allowability of anySimpleType as a member of unions. Part 1 3.14.6 Simple Type Definition Properties Correct rules it out, but 2.5/6.1 of Datatypes might be read as allowing it, because it appears to imply that Atomic+List+Union exhausts the space of possibilities. 7373 1 cmsmcq 2005-12-09 20:45:42 +0000 The WG discussed this item on 9 December 2005 and agreed to treat it as a requirement for 1.1 that we clarify whether anySimpleType may or must not be used as a member of a union. But 2574 raises the same question for version 1.0; it may be relevant to this issue because if 1.0 has a clear answer to this question, some WG members will prefer not to change the answer in 1.1. There is a minor error in comment 1 -- it is clause 3.1 of Schema Component Constraint: Derivation Valid (Restriction, Simple), not Simple Type Definition Properties Correct, which appears to rule out anySimpleType. 3.1 The {member type definitions} must all have {variety} of atomic or list. (http://www.w3.org/TR/xmlschema-1/#cos-st-restricts) This clause must change, however, in any case, since the resolution of bug 2044 (retain union-level facets) requires that member types of variety union be allowed. (See bug 2044 comment 2, see also bug 2333, which is the Structures part of that issue.) 8807 2 cmsmcq 2006-03-21 00:39:44 +0000 At the face to face meeting of January 2006 in St. Petersburg, the Working Group discussed this issue. While there was some sentiment for giving it a high priority, in the end the Working Group decided not to take further action on this issue in XML Schema 1.1. The rationale for the Working Group's decision was simple: the obscurity regarding anySimpleType appears not to be causing serious difficulties in practice (our collection of a thousand schema documents found 'in the wild' has no instances of unions which have anySimpleType as a member), and it was not felt useful to delay Datatypes 1.1 while the Working Group discussed this issue further. Accordingly, I am marking this issue as RESOLVED / WONTFIX.