Which style first ? Re: Associating DSSSL style sheets with documents

[Norbert H. Mikula]

James Clark wrote:
> Where the content provider knows which style they want, they can use a
> fragment spec in the URL to pick out a particular spec from the document.

The scenario I have in mind is putting documents onto a WWW server.
In this case the content provider can not know whether the user (-agent)
downloads for online-rendering, hardcopy or a mixture thereof.

> >What about adding an attribute with a list of catagories
> >of DSSSL engines as possible attribute values.
> >
> >For instance : output (hardcopy|online|....?) hardcopy
> 
> What other categories are there?  If a style sheet is for online use, then
> it has to use the scroll flow object, which means it ought to list the
> online feature in the features element type form. 

Yes, that is an interesting approach. However, if we have multiple
style-specs. in one document, for instance hardcopy and on-line.
Would it not be the case, that you would find the online feature
in the features element type form regardless of what stylespec. the 
user agent is really interested in ?

I don't have the DSSSL specs at hand right now. So I hope it makes sense
what I am saying. If not -> (element my_comment (empty-sosofo)) ;-)

> Maybe a DSSSL engine
> could use this, or maybe it could look to see which flow object classes the
> spec uses.

I don't think that this is a practical approach. If a style-spec. is 
supposed to be used for online-rendering, is it really a must to use
scroll-fo ? What about simple-page-seq. ? I know that this is not
something
somebody would intuitively do, but why not. I could envision a browser
that 
uses simple-page-seq. For large document instances the browser takes
advantage of the explicit information in the document instances combined 
with the style-spec. In other words, chapter starts a new page. The
browser
would have to render only the active page.

I guess it might also foster more reusable style modules.

<FN>It remains to be discussed how the document should be divided into
more Internet suitable sizes and how the user agent would/should deal
with this and to what extend it has/should have/must have an influence 
on the design of stylesheets. I especially have in mind the case 
of XML entity treatment.</FN>

Makes sense ?

However, I can see no point in scanning the whole document just to
deduce
somehow what class of style-spec. it belongs to. It sounds to me like
trying
to figure out the semantics of a paragraph without using markup.

-- 
Best regards,
Norbert H. Mikula

=====================================================
= SGML, DSSSL, Intra- & Internet, AI, Java 
=====================================================
= mailto:nmikula at edu.uni-klu.ac.at 
= http://www.edu.uni-klu.ac.at/~nmikula
=====================================================


xml-dev: A list for W3C XML Developers
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To unsubscribe, send to majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
List coordinator, Henry Rzepa (rzepa at ic.ac.uk)