xml-dev - Re: Success factors for the Web and Semantic Web

[Uche Ogbuji <uche.ogbuji@fourthought.com>]

> > I don't know what this means.  What is this "language"?  Is it an
> > XML vocabulary/schema?  What is this "correct"?  Is it a matter
> > of authority to use a namespace?
>
> Say I created a namespace for a language at http://infomesh.net/testns/
> Then, on the Web, there are two different XML documents:-
> 1. <roota xmlns="http://infomesh.net/testns/"/>
> 2. <rootb xmlns="http://infomesh.net/tesns/"/>
> Here, the root elements, (and possibly the enitre definitions of the
> langauges!) are different, but they use the same namespace! Which one is
> definitive? Dereference the namespace, and you might well find out (in this
> case you won't becasue I didn't write a Schema: but I could have done).

You are bringing up a problem that has nothing to do with technology.
Two companies call their breakfast cereal "Cheerios".  How do you find out
which is the "correct" one?

You wait and see who wins in court.

Your idea of making namespaces referenceable does very little to address
this matter.  Not all namespaces are URLs.  And even for those that
are, there are no guarantees.  What about proxies?  Domain hijacking,
etc.?

> > > 2. A Namespace is still not definitive. Tim's example was of a URI
> > >      for Zip codes. There could be hundreds of them: no one is going
> > >      to use the same one!
> >
> > I'm still not understanding you.
>
> O.K. You define a zip code in a Schema as a 6 digit number. Then I make a
> different Schema that says in fact a zip code is an 8 digit number. There
> is nothing to say that either one of us is correct.

I see.  I was thrown off by your "no one is going to use the same one!".
So are you saying that it is very unlikely for _any_ two people to use the
same Zip-Code to URL scheme?  I must still be misunderstanding you.

> > > 3. There is no current way to ascribe structure to extended RDF.
> > > (Action: create a way!)
> >
> > What is this "extended RDF"?  I'm not even sure I can guess what it is.
>
> RDF that will require additional structres not set out in the Model and
> Syntax spec, i.e. structures provided by XML Schema.

So is RDF using RDF schemas "extended RDF"?  I find that hard to buy.  I
think XML schemas fit right into RDF without any extension beyond that
provided by RDF M&S itself.

> > >      1. Create XML to RDF extraction tools.
> > Happenning all the while.  Examples include Jonathan Borden's work,
> > and that of Dan Connolly
>
> I am well aware that there is a lot of work in this area :-) I have been
> doing some myself ((X)HTML to RDF).

You might want to look at Dan Connolly's work in the same area (XHTML ->
RDF).  He has some great stuff

See

	http://www.w3.org/People/Connolly/smart-home.xsl
	http://www.w3.org/People/Connolly/home-smart.rdf
	http://www.w3.org/People/Connolly/events/events-smart.rdf

And you really should just check out

        http://www.w3.org/People/Connolly/

> > >      2. Create notepad databases of RDF
> > Can you explain this?
>
> Stuff like Blogspace but in RDF. http://www.blogspace.com/ c.f. FRED
> http://infomesh.net/sbp/#fred

I'll have a look.  Could this be similar to the demo right on Connolly's
home page?

> > See the work of Dan Brickley and others.  One of my Fourthought
> > colleagues is doing some work in this field as well.
>
> Do you have a URI for that?

Brickley's RDF visualizer is at

http://www.ilrt.bris.ac.uk/discovery/rdf-dev/rudolf/rdfviz/

There are other such tools listed at

http://www.w3.org/RDF/

As for Chime's work, I'll ask him, though I doubt it.  He's only begun to
take it up in earnest very recently.

> > Well, from your 1, 3 and 4, I'd have to conclude that the semantic
> > Web is well on the way, but I think your list is not exhaustive.
>
> Of course! Not by any means. In fact I asked people to each come up with
> ten uses of the SW they'd like to see. Could you give me your 10?

Remember that I'm still trying to wrap my head around the SW concept.  I
promise I'll contribute as soon as I'm able.

But I thought your list was prerequisites for, not uses of the SW.
Perhaps I misunderstand you again.


-- 
Uche Ogbuji                               Principal Consultant
uche.ogbuji@fourthought.com               +1 303 583 9900 x 101
Fourthought, Inc.                         http://Fourthought.com
4735 East Walnut St, Ste. C, Boulder, CO 80301-2537, USA
Software-engineering, knowledge-management, XML, CORBA, Linux, Python