W3C home > Mailing lists > Public > www-tag@w3.org > April 2008

Re: Uniform access to descriptions

From: Xiaoshu Wang <wangxiao@musc.edu>
Date: Fri, 11 Apr 2008 18:54:19 +0100
Message-ID: <47FFA5CB.5010408@musc.edu>
To: "Williams, Stuart (HP Labs, Bristol)" <skw@hp.com>
CC: Jonathan Rees <jar@creativecommons.org>, "Michael K. Bergman" <mike@mkbergman.com>, "www-tag@w3.org WG" <www-tag@w3.org>, Phil Archer <parcher@icra.org>

Stuart,

Hold on.  Let's go one step at a time.

We agree that there are legacy data, yes?  Let's make its URI x, whose 
owner is Joe.
Case 1. Joe is lazy.
Then, no LINK, no Conneg. Is this fair?
Case 2: Joe is not lazy.
(a) Joe makes LINK(x)=metadata.
(b) Joes make Conneg(x)=metadata (can easily GET x Accept 
application/rdf+xml).

I think your later use case is assuming Joe is lazy toward (b) not not 
(a).  I don't understand why it has be so.  In fact, Joe should be lazy 
toward (b) because Conneg is in the current HTTP spec but not LINK. 

If Joe doesn't have any preference, see if all you later provide use 
case can be solved with Conneg.  If you give me a use case with the 
assumption that requires Joe to use LINK, then, what can I say?

I said to Harry before, if the LINK is in current HTTP spec, my position 
will be reversed.  It is not about if LINK won't work.  It is about if 
existing system can still make it work.

Xiaoshu

Williams, Stuart (HP Labs, Bristol) wrote:
> Hello Xiashou,
>
>   
>> -----Original Message-----
>> From: Xiaoshu Wang [mailto:wangxiao@musc.edu]
>> Sent: 11 April 2008 15:24
>> To: Williams, Stuart (HP Labs, Bristol)
>> Cc: Jonathan Rees; Michael K. Bergman; www-tag@w3.org WG; Phil Archer
>> Subject: Re: Uniform access to descriptions
>>     
>
> <snip/>
>
>   
>> Can not all your posted problem be not solved with Conneg?
>>     
>
> I think that you are confused about what content negotation is for.
>
> One thing that it is not for is negotiating between [a representation of [a thing]] and [a representation of [a descriptionof [a thing]]]
> ie between a resource's content and its meta-data.
>
>   
>> Treat the following as plain text "http://www.w3.org".  Anyone read it
>> don't know how to use it?
>>     
>
> Well where are the links... how do I know that the references in the representation are references. text/plain doesn't tell me. Ok, we see strings that begin http:// and most probably are links but that is sniffing and guesswork.
>
>   
>> If the legacy data don't mind update to fit in with
>> Conneg, what does it make them to use LINK?  I don't understand the
>> rational behind.
>>     
>
> But that is precisely it, most of the legacy does in fact mind.
>
> Much of the legacy is unmaintained... it is just there. And where's the motivation for the owners of those resources to do the updates - the value is all for other people/agents, not them... and besides... they have done nothing wrong!
>
>   
>> Perhaps, it is me who have failed to understand the other people's
>> problem.  But isn't it possible that it is others who have failed to
>> understand HTTP-Conneg?
>>     
>
> See above...!
>
> <snip/>
>   
>>> Are we agreed that Jonathan is *not* "inventing a synonymy"?
>>>
>>>       
>> No. I don't agree.  Because give me a concrete example of "description".
>> If you can find such an example that cannot be explained in the context
>> of *representation* and *resource*, I will agree.
>>     
>
> Let me try to be clear.
>
> One awww:resource describes another if it makes statements *about* it. It may make those statements as narrative text in a hypertext document, it may make them as assertion on some KR language such as RDF or KIF or whatever. One way that it is apparent that one resource is descriptive of another is that its representations (which as we have repeatedly agreed are distinct from the resource that they are a representation of) make reference to the things being described. Thus
>
>         <http://www.w3.org/TR/> :describes <http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/> .
>
> because, amongst other things, http://www.w3.org/TR/ contains (in narrative form) some biblographic information about http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/. In that sense http://www.w3.org/TR/ is a description of http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/. I'm sure that there are better examples where there is a more significant bibliographic record.
>
> Likewise,
>
>         <http://www.w3.org/2002/01/tr-automation/tr.rdf> :describes <http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/> .
>
> because, amongst other things, <http://www.w3.org/2002/01/tr-automation/tr.rdf> makes RDF assertions about <http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/> .
>
> Are those concrete enough?
>
> If you are looking for description as a class I suppose you could say:
>
>         { ?a :describes ?b} -> { ?a a :Description }
>
> but ?a is only interesting as a description in that it describes ?b.
>
>   
>> But I
>> still cannot see any.
>>     
>
> Ok... so tell me the story of the three resources in your terms.
>
>   
>>  I have said Mark's LINK draft is very much DC.
>> All those might not fit, such as POWER's GRDDL sheet can be easily
>> solved with Conneg.  Find an example that I cannot reduce the
>> problem to these categories, I will be the first to support it.
>>     
>
> Tell me how if find the two resources above that I have deemed to be descriptions of
>         <http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/>
>
> and without touching its content? Actually there is a good answer for the /TR case, but I defy you to find the RDF.
>
>   
>>> Please... that is the point/claim I was addressing. I have
>>>       
>> no idea whether we have agreed on it or not.
>>     
>>> To answer your other question here: "What can
>>>       
>> it[description] be?". I gave my answer that way before, with
>> all the "awww:"'ing and you agreed:
>>     
>>> It is a relation between (in this context) "awww:resource"
>>>       
>> where on[e] resource is descriptive of another.
>>     
>> This is exactly the problem I have.  Isn't RDF is all about *describing*
>> resource with another resource?  You tell me what is the
>> difference between
>>
>> (a) <x> HTTP-LINK: copyright <y>
>> (b) <x> dc:copyright <y>.
>>     
>
> Please think about what you aretrying to say/ask with (a). (b) I can parse as turle/N3. (a) needs explaination.
>
>   
>> Then tell me what is the difference between two types of  *description*?
>> I am too dump or numb to see the difference.
>>     
>
> Ok... we're almost at the atomic level of triples and how they come to be stated or inferred.
>
> An HTTP response to an HTTP GET with a request URI of http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/ which contains the following header (and no content-location: header which IMO would change of the assertion);
>
>         Link: <http://www.w3.org/2002/01/tr-automation/tr.rdf>;rel="http://example.org/#metadata"
>
> conveys the following RDF assertion:
>
>
> <http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/> <http://example.org/#metdata> <http://www.w3.org/2002/01/tr-automation/tr.rdf> .
>
> It is an assertion that you can choose to ignore; it is resource metadata that provides a reference to more metadata about, amongst other things, the referenced resource. It is an assertion not apparent in the representation of the SPARQL rec itself. It is probably moot whether you regard the assertion as being made by: the SPARQL rec editors; the naming authority who assigned the name to the SPARQL rec; 'the Web' itself; or the web master for the relevant site.
>
>   
>> Xiaoshu
>>     
>
> Stuart
> --
> Hewlett-Packard Limited registered Office: Cain Road, Bracknell, Berks RG12 1HN
> Registered No: 690597 England
>
>
>   
Received on Friday, 11 April 2008 17:55:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:55 GMT