Re: [CSS21] Unclear applicability to XML

On Tuesday, August 30, 2005, 10:58:22 PM, Ian wrote:

IH> On Tue, 30 Aug 2005, Chris Lilley wrote:


>> If that satisfies the XSL WG then it satisfies me.

IH> I was under the impression that these comments were "from both the CDF WG
IH> and the SVG WG"; are they from those working groups, or from you?

The point about XSL was added in by me; I have also alerted them to this
discussion via the XML CG. I feel I have done my best there and if they
do not send their own comments then I guess the XSL WG is by default
satisfied.


>>> I don't really understand why you keep on insisting that CSS might be 
>>> specific to XHTML browsers.
>> 
>> Because the spec sometimes implies that and other times does not; 
>> because you sometimes imply that and other times not.

IH> With all due respoect, I think you would have an either time understanding
IH> both the CSS specs and my own e-mails if you read what was written, 
IH> instead of trying to read what might or might not be implied.

I agree, so change the spec to state clearly what it is aimed at and
then I can stop trying to guess.


>> Perhaps 'document' needs to be more clearly explained, as we discussed 
>> on the phone yesterday there are several different meanings for 
>> 'document'.

IH> It would seem reasonable, however, to use the normal English meaning of 
IH> the word in the absence of a specific definition.

It would if people did not also frequently use it in the meaning of
'resource representation'.


IH> One reason we are reluctant to define the term is that CSS2.1 can be 
IH> applied to any random XML, including (in a non-SVG UA) one containing 
IH> elements from the SVG namespace. For example CSS has been applied to Atom,
IH> RSS, MathML, Docbook, XHTML, XUL, random SGML, proprietary XML, CML, SVG,
IH> etc. Just because CSS2.1 can be applied to SVG files, however, doesn't 
IH> mean it can be used to render graphics, and in fact CSS2.1's interaction
IH> with an actual SVG UA is not particularly useful if your statements in 
IH> www-svg regarding its moribund status are to be taken seriously.

IH> In short, the scope in terms of what we want it to be primarily useful for
IH> is document formats that are typeset in typical "document-like" (as you 
IH> put it) ways. However, it can be applied to any XML, SGML, or other format
IH> that uses a tree of elements.

It can be applied to them, yes; the question is whether that produces
any sort of useful result. Based on the properties in CSS2.1, is
primarily aimed at technical documentation, where a lot of the
properties can be applied; its hardly useful at all for SMIL, for
example, has a fairly small overlap with the properties used for
SVG, and nothing useful for CML or X3D. Which is fine, if the spec says
that it is not aimed at being useful for those cases.

>> Well, it would help show that its applicable. I'm sure Norm Walsh would 
>> be able to give you a simple example to style.

IH> It's applicable in the same way that it's applicable to HTML. I seriously
IH> do not see that there is any difficulty here. Anyone competent enough to
IH> use Docbook and CSS is quite capable to work out how they work together.

Does Docbook use more than one namespace?

-- 
 Chris Lilley                    mailto:chris@w3.org
 Chair, W3C SVG Working Group
 W3C Graphics Activity Lead

Received on Wednesday, 31 August 2005 01:32:46 UTC