Re: identifying test subjects

On Tue, 26 Sep 2006 09:23:25 +0200, Shadi Abou-Zahra <shadi@w3.org> wrote:

> Hi Charles,
>
> Charles McCathieNevile wrote:
>>> There is an open issue that the Test Subject class does not have a  
>>> (built in) mechanism to identify what it is describing (a handle).
>>  RDF itself provides this, by pointing to a resource. If you want to  
>> use dc:foo or some other form of identification there is nothing to  
>> stop you, but I don't think they should be mandatory.
>>  I strongly object (and would make a formal objection) to a uri  
>> property as an identifier (other than r:resource) for the reasons  
>> outlined by Shadi. I suggets we do nothing, and encourage people to use  
>> rdf:resource
>
> So you prefer rdf:resource over uri:uri. How about test subjects that  
> are not identified by a URI in the first place? Do we want to even  
> consider these for EARL 1.0 (question to the whole group)?

You can use an rdf:resource to identify anything - it is not a URI that  
you necessarily expect to download, but it is an identifier. Which means  
you can use it for a person (some people have an rdf identifier or two  
already), a webpage (either "whatever you get at http://foo or an  
identifier that points to some RDF which describes a page) or a book  
(again, many of these have a URI as identifier from some other RDF). In  
other words by doing the standard thing that RDF always does, we have all  
the flexibility to cover whatever we want.

uri:uri is only useful for things that are actually available on the web.  
Restricting ourselves in that way, by usng something that isn't generic  
RDF, seems like a very bad idea to me.

cheers

Chaals


-- 
   Charles McCathieNevile, Opera Software: Standards Group
   hablo español  -  je parle français  -  jeg lærer norsk
chaals@opera.com          Try Opera 9 now! http://opera.com

Received on Tuesday, 26 September 2006 09:26:52 UTC