Chris Bizer wrote:
> Hi Emmanuel,
>
>
>>Yes. So both have their advantages and drawbacks. Yours is easier to
>>manipulate by "dumb" tools such as XSLT processors because it is very
>>concrete. Mine is more abstract, thus less verbose and more "presentation
>>knowledge oriented", though this is a loose claim and I am not sure it has
>>any actual benefit.
>
>
> You are touching an interesting point. I don't see the intermediate format
> as a way of representing "display knowledge" but just as something that has
> to be practical for further processing.
>
> I think RDF (abstract syntax, not concrete syntax <= for the XML folks) is
> what we want to use for representing knowledge. XML should just be used as a
> practical tool for representing and transforming output.
Yes. I agree with you, and considering this your way of handling
additional content seems more appropriate.
This "vision" of the intermediate tree is also what has triggered my
last question: why do we want to model class/type information
(associated with the lens selector) in the intermediate tree, and if we
need a reference to the lens that generated a particular fragment of the
tree, why not simply reference this lens directly (and thus unambiguously)?
--
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 Wed May 04 2005 - 09:40:28 EDT