This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Section 3.4 in the Datatypes spec lists all facets that are already present on the derived types. For the 3 built-in list types, whiteSpace = collapse should be listed. whiteSpace should also be removed from the "may also specify" list for these types.
(In reply to comment #0) > whiteSpace should also be removed from the "may also specify" list > for these types. (Most easily done by making whiteSpace fixed when set to collapse. This was done relatively recently for some datatypes.)
It's not clear to me why whitespace should be removed from the 'may also specify' list. Is it illegal to specify whitespace=collapse when restricting a list datatype?
(In reply to comment #2) > It's not clear to me why whitespace should be removed from the 'may also > specify' list. Is it illegal to specify whitespace=collapse when restricting a > list datatype? Of course not. Just should be specified fixed="true" (as we did a few months ago for most of the primitives). That way it gets the lead-in "these facets must not be changed from the values shown" rather than "these facets may be further restricted in the derivation of new types" (which really suggests that it can be *changed*, not just *specified*). And it *should* be listed as having a value, not just available to be given a value in a subsequent derivation (which is what the "may also specify values for" list is about).
The WG decided on wording to address this issue: http://www.w3.org/XML/Group/2004/06/xmlschema-2/datatypes.b6734.html
The change mentioned in comment 4 now having been integrated into the status-quo documents, I'm marking this issue RESOLVED / FIXED. Sandy, as the originator, will you please verify that the problem has been fixed to your satisfaction and CLOSE the issue (or REOPEN it if not)? Thanks. On Friday, silence will be taken as consent.