Chris Bizer wrote:
> Sorry, having styling hooks in core but not having a mechanism to
> dereference these hooks doesn't make sense.
>
> I also hope that somebody knows what to write into the introduction of the
> paper as we are constantly loosing the ability to express and discover
> cross-browser *display knowledge* and move to some hacky browser
> configuration file language.
I tend to agree with you that the more it goes and the less we express
things in Fresnel.
But do you really consider hooks + a ref to external stylesheets
containing CSS rules to be actual display knowledge? If we had done
everything in RDF (e.g. having CSS instructions as RDF properties), then
it would obviously have been display knowledge. Here, I'd say we have a
convenient mechanism for easily assigning CSS/SVG styles to parts of the
representation a posteriori, but this is not display knowledge.
So we have selection knowledge and formatting knowledge. But we don't
have styling knowledge. Styling hooks are probably going to be useful,
but they don't fix this. So I consider them as a nice feature, but not a
fundamental one.
I'd like to come back to this styling knowledge issue in the future,
once we've gained experience using Fresnel. Right now, I'd rather leave
this out of core.
--
Emmanuel Pietriga
INRIA Futurs - Projet In Situ tel : +33 1 69 15 34 66
Bat 490, Université Paris-Sud fax : +33 1 69 15 65 86
91405 ORSAY Cedex FRANCE http://www.lri.fr/~pietriga
Received on Tue Jul 05 2005 - 13:05:11 EDT