Re: Constraints and CSS

Hi Christian,

I think you proposal for Constraints in CSS works very well. I am only  
afraid that it might be difficult to be applied for other kinds of  
constraints. Another point which worries me is that we would not be able  
to validate such constraints within a tag set. This would be different if  
we would choose to rely on XML data types, which could express all  
constraints we discussed at the f2f (including max width).

I would propose to put this text also in a wiki - esp. because many people  
are already in their christmas holiday, I guess ;) Is that o.k. with you?

Regards, Felix.

On Mon, 19 Dec 2005 23:08:58 +0900, Lieske, Christian  
<christian.lieske@sap.com> wrote:

> Dear all,
>
> During the F2F in Abingdon, I took away the action item to come up with
> some text about "external" constraint information.
>
> Looking forward to your feedback.
>
> Best regards,
> Christian
> ---
>
> A look at vocabularies which can work with Cascading Stylesheets (e.g.
> XHTML
> and XML) reveals that they already have a means for capturing at least
> some
> of the constraints we have been discussing. To be specific, the CSS
> property
> "max-width" constrains box widths to a certain range.
>
> Accordingly, with CSS we have the case were we may not want to
> enforce an ITS-specific representation of max-width but rather use
> what is already part of the native XML application (in the sense of
> the combination of  XHTML instance and CSS). Of course, we need
> to make sure that the semantics and domain (in the sense of allowed
> values) of the ITS data category for constraints, and the are
> compatible.
>
> Once we principally have come to the conclusion that the CSS
> "max-width" covers the ITS data category, we need to take care that
> the CSS representation makes sense to ITS aware applications. This
> may require:
>
> 1. a mapping declaration (see our discussion at
> http://esw.w3.org/topic/its0504ReqPurposeSpecMap)
> 2. possibly special locator processing
>
> With 'locator processing' I refer to the situation in which a linking
> mechanism is
> used in order to associate a vocabulary with a stylesheet. Examples:
>
> A. Via special element: <LINK REL=StyleSheet HREF="style.css"
> TYPE="text/css">
> B. Via processing instruction: <?xml-stylesheet type="text/css"
> href="style.css" media="all"?>
>
> Obviously, in these case, processing expectations related to the mapping
> declaration are
> different from the ones where "internal" styling (ie. either with the
> "style" element or the "style"
> attribute) is used: Actual values are not related in the XML instance
> but rather externally.
>

Received on Tuesday, 20 December 2005 06:12:47 UTC