W3C home > Mailing lists > Public > www-rdf-interest@w3.org > October 2002

Re: DC/RDF and DC/XML (was RE: The Tragedy of RSS)

From: Charles McCathieNevile <charles@w3.org>
Date: Fri, 4 Oct 2002 09:59:16 -0400 (EDT)
To: Jon Hanna <jon@spin.ie>
cc: www-rdf-interest Mailing List <www-rdf-interest@w3.org>
Message-ID: <Pine.LNX.4.30.0210040948310.7513-100000@tux.w3.org>

On Fri, 4 Oct 2002, Jon Hanna wrote:

>
>> However, I question whether the use of DC-in-RDF [1] elements in a
>> document which is not an RDF document can really be said to be a use
>> of RDF per se, especially given that the element names are identical
>> to those used in a 'pure' XML encoding of DC [2].
>
>Out of interest what do people think of the fact that the same namespace
>names are used in both RDF and non-RDF based DC?
>
I think it is an intersting issue to deal with. There are lots of things we
use RDF for (DC, EARL, Annotea, are three that leap to my mind) where common
use cases allow processing according to an XML schema (which is generally
easier to do). There are some use cases where we want the full power of RDF.

For example:

in EARL, I can assert that resource X fails to meet requirement Y. If I have
a collection of EARL data, I can look for all the elements that have
<earl:fails rdf:resource="Y"/> as a child, if the EARL is written out
according to a particular schema - which doesn't stop it being RDF. This is a
one-line-hack in XSLT (I can't remember the syntax so I won't give an invalid
code example)

On the other hand, it is helpful in a Quality Assurance Process to be able to
say things like X almostPasses requirement Y. In RDF we can do that -
almostPasses is a subProperty of fails, but different from totallyFails.

Using an RDF query I can find the things that pass or almostPass. (I am
assuming already a normalisation so that the fact that something almostPasses
leads to the version canonicalised to the XML schema including the
information that it fails). These I can send to one department, while the
totallyFails stuff or the notImplementedYet stuff goes to a different
department.

As far as I can tell, the namespace identifies the semantics of the element -
but those in turn depend on the environment. If you are asking for how the
element behaves in an XML schema you want something different to what you
need when asking how the element works in RDF.

My first reaction is that this isn't a problem if you can query the namespace
and get something in RDF Schema or XML schema by content negotiation. Of
course this is making implementation assumptions that go beyond what is
standardised for namespaces...

Cheers

Chaals
Received on Friday, 4 October 2002 09:59:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:56 GMT