Re: RDFa FAQ comments and activities

Simone,

Thanks for the very useful feedback. Some thoughts below.

> - 2.1.1. Can I use RDFa on any HTML version?
> 
> [[ The intention of the RDF in XHTML Taskforce is that RDFa can be
> used with every (X)HTML version. This is, however, still an issue that
> is in state of flux. ]]
> 
> We may add:
> 
> "Actually we support XHTML from version 1.1 [2] with specific module
> for metainformation [3] and attributes [4]. (link to the page with DTD
> declaration and an example 1.1 and 2.0 both). We are also working for
> an HTML implementation."

Yes, good point. Feel free to add this to the wiki.

> I think to new HTML WG [5], we already have Dan and more IWA/HWG from
> Semantic Activity, now I've an high working time, so I can join it
> later, but I may have to ask to my AC).

That would always be useful!

> - 4. RDFa in 10s
> 
> Elaborating Steve's words: [[ Here's one: When you do a search on the
> Web for Pascal, the search engine doesn't know if you mean the
> programming language, the philosopher, or any of a dozen other
> possibilities. Amongst other things RDFa helps the search engine work
> out what a page is about. ]]

I am wary of such an aggressive semantic web pitch. RDFa alone doesn't
do this for you and, in fact, it's questionable whether all of the
semweb technologies (RDF, SPARQL, etc...) do this yet.

We don't need people to buy into the whole semweb vision to use RDFa. We
want people to start using RDFa for a very simple reason: data
interoperability and "sprinkling structure into existing web pages."
That's useful right now.

> I wrote on a document regarding the power of Semantic Web (translation
> from Italian can be more good):
> 
> (...) With exponential growth of Web, search engines are most
> importants, it - should - found in few seconds things that we ask.
> Well, Text or Hyper-Search engines do not answer with a single or sure
> result. It give us a huge number of results hoping that the right
> answer it's on the first page. A Semantic Search Engine, instead,
> computing pages with a Semantic Layer (with RDFa, of course) can
> elaborate good our querys and return only one result, or a little
> range of consistent and correct anwers.

same worry as above: you could replace RDFa with "RDF" in the above
paragraph, and it would still work. In other words, the advantages of
RDFa are not clear: why embed RDF in HTML? Because what we really want
is to augment the *existing* web with structure. So you can right-click
a link and ask "what does this link represent? an endorsement? an
insult? a copyright license?"

> One little note on hGRDDL [6], few editorials:
> 
> On first example DTD can be XHTML 1.0 Strict (Be Strict to be cool
> [7]) instead Transitional and on second example DTD can be the
> specific XHTML+RDFa.

excellent. Please feel free to edit!

I really appreciate your feedback on all of this. Any additional
feedback is always welcome!

-Ben

Received on Sunday, 11 March 2007 18:44:29 UTC