Re: HTML in RDF

Hello Karl

That was the old (malformed) example, based on a misunderstanding of Turtle
semantics. The newer example was structured so that rdf:value attaches
element content to an element (doesn't have to be rdf:value of course),
rdf:List maintains element content ordering; hopefully I "got" Turtle now.
Attributes are unordered and just attached to the element node directly via
regular properties.

The idea I have is to capture the entire portal as RDF, including form
posts. That requires, at first, an HTML <=> RDF bi-directional
transformation. I'll make a taxonomy akin to the DOM one to make querying a
bit more interesing. Then the deam would be to have an editor directly work
on top of the RDF datamodel so only RDF => HTML will be necessary past that
point. This editor will enable different kinds of annotations to the text
which can be exposed to the user in the rendered HTML ("D", or "Ajax").

Of course RDFa could also be used to capture some information, but I'd like
to maintain a full RDF model all the time. Not sure all the opportunities it
opens, but I'm pondering...

I'm not sure I get the parser ordering issue. Whilst RDF graphs are not
intrinsically ordered, it would be very disappointing to me if it was
somehow not useful to semantically capture ordering. That's the reason for
rdf:List, right?  I get the open world assumption, but

Regards

Bent


2007/10/29, Karl Dubost <karl@w3.org>:
>
>
>
> Bent Rasmussen (27 oct. 2007 - 03:37) :
> >       h:p rdf:value (
> >         h:text rdf:value "It has to be possible!"
> >       )
> >       h:p rdf:value (
>
> …
>
> >  Ordering is preserved where necessary.
>
> I'm worried because specifically a graph is not done to preserve
> ordering. So that will break going from parsers to parsers.
>
> What is the benefit to express everything in RDF?
> What is the specific use case you have in mind?
>
>
> --
> Karl Dubost - http://www.w3.org/People/karl/
> W3C Conformance Manager, QA Activity Lead
>    QA Weblog - http://www.w3.org/QA/
>       *** Be Strict To Be Cool ***
>
>
>
>

Received on Monday, 29 October 2007 08:51:36 UTC