- From: Allan Beaufour <beaufour@gmail.com>
- Date: Thu, 20 Apr 2006 09:40:14 +0200
- To: "John Boyer" <boyerj@ca.ibm.com>
- Cc: "Xforms W3C WG" <www-forms@w3.org>
Hi John, On 4/19/06, John Boyer <boyerj@ca.ibm.com> wrote: > When the label has an empty nodeset, the string result of that is empty string. Spec.ref? It is correct that it is the empty string if the element node that a control is bound to has no text children, but this is the case where the bound node does not exist. > Non-relevance also typically gets styled as non-visible, so in both of those cases the label > of the item would not be seen. Non-relevance could also be styled as disabled, Yes, I agree it is also a question of styling. But it is also a question of semantics. > but for a label to be disabled means that it is greyed out. Spec. ref? > Non-relevance is applied based on a UI binding between the control and the model. We > have direct language in the spec that says if you have no UI binding, then relevance does > not apply. Spec. ref? > So, right now relevance does not get applied to items, regardless of how they are > declared. Not directly to items. But what is an item? The item is just a grouping of a value and a label. So my point, and I think also Erik's, is what the semantics are for grouping a non-relevant value/label. I think it makes sense to define a meaningful behaviour. We may, or may not, have something in the spec. that defines that (the latter I guess). -- ... Allan
Received on Thursday, 20 April 2006 07:41:16 UTC