This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
As specified, it gets a value when an Attribute Declaration has a value constraint but not when an Attribute Use has one. Either we should 1) Deprecate this in 1.0 and remove it from 1.1 (what use is it, anyway -- why is it here?); 2) Add a Note about its half-brokenness; 3) 'fix' it, pbly by moving it under Complex Type Defn PSVI contributions and folding in Attribute Uses.
Turns out it's there because in 2000 there was a [default] property on attributes in early drafts of the Infoset REC (e.g. http://www.w3.org/TR/2000/WD-xml-infoset-20000726#infoitem.attribute)
Looking at this, I see two relatively easy solutions, one relatively easy half-solution, and one harder solution: 1 remove this property entirely from both elements and attributes (easy) 2 change the relevant paragraph in section 3.2.5 (Attribute Declaration Information Set Contributions) from Also, if and only if the declaration has a {value constraint}, the item has a property: PSVI Contributions for attribute information items [schema default] The {lexical form} of the declaration's {value constraint}. to Also, if and only if the declaration, or the Attribute Use to which the item was attributed, has a {value constraint}, the item has a property: PSVI Contributions for attribute information items [schema default] The {lexical form} of the {value constraint}. (easy) 3 Add a note to the attribute flavor of this property, noting that it doesn't appear when the value constraint is on the attribute use instead of the attribute declaration, and thus may not be useful in some situations. (easy, but not a full solution) 4 (in Henry's words) 3) 'fix' it, pbly by moving it under Complex Type Defn PSVI contributions and folding in Attribute Uses. (harder, mostly because I don't know where Henry had in mind to move it to). Of these, 1 or 2 seems most satisfactory to me. Two questions arise which might help us choose between them: (1) does anyone currently use this property, who would be inconvenienced if it went away in 1.1? (2) do people agree that the change in approach 2 suffices to make the property no longer half-broken?
Proposal: to adopt option (2), with some wording changes. Resolution: adopted at 2006-10-17 telecon.