Re: SemWeb Non-Starter -- Distributed URI Discovery

On Apr 10, 2005, at 12:41, ext Andy Powell wrote:

>
> On Sun, 10 Apr 2005, Charles McCathieNevile wrote:
>
>>>  I think it should be feasible to issue this sort of request:
>>>  GET /food/blah HTTP/1.1
>>> Host: example.com
>>> Accept: application/rdf+xml
>>
>> Patrick replied
>>>  Not to just jump in and jump out calously, but this has been
>>> explored quite a bit for quite some time and content negotiation
>>> is simply not the correct mechanism for this.
>>>  C.f. the FAQ section of http://swdev.nokia.com/uriqa/URIQA.html ...
>>>  Patrick
>>
>> I'm not so sure. I can see the problem if you see the world through 
>> URIQA
>> glasses, but I think outside of that it is not a bad answer.
>
> I tend to agree.  And I find the following answer in the URIQA FAQ
> somewhat confusing...
>
> --- cut ---
>
> Why not use a MIME type and content negotiation to request a 
> description?
>
> Content negotiation is designed to allow agents to select from among a 
> set
> of alternate encodings. The distinction between a resource description 
> and
> (other kind of) resource representations is not based on any 
> distinction
> in encoding. In fact, a given description (which is itself a resource) 
> may
> have several available encodings (RDF/XML, XTM, N3, etc.).  Thus, if 
> you
> use content negotiation to indicate that you want a description, you 
> can't
> use it to indicate the preferred encoding of the description (if/when
> other encodings than RDF/XML are available).
>
> --- cut ---
>
> ...since (I think) it uses 'description' in the rather specialised 
> sense
> of 'concise bounded description'?  The first sentence would be better 
> if
> it read "Content negotiation is designed to allow agents to select from
> among a set of alternate *representations*"?

No. What I was saying was that, by not using conneg to ask for
the description, you can use conneg to ask for a particular
encoding of the description.

E.g. what if your sw agent prefers N3, or XTM, or TriX, and all
are supported by the given server. If you've used conneg to get
the description itself, what do you use to ask the server for
a prefered encoding?

In short, using conneg to provide URIQA-like functionality is
a misuse of conneg. It's a hack. And it precludes using conneg
for what it was intended for.

>  In which case, given that a
> description *is* a representation (and this is true even for a CBD), 
> it's
> hard to understand why content negotiation couldn't/shouldn't be used 
> to
> negotiate a 'description representation' or even a 'CBD 
> representation'?
>
> The problem (as outlined in the final sentence of the FAQ answer 
> above) is
> really with limitations in the current set of MIME types isn't it?
> Couldn't much of what URIQA provides be achieved either by creating a 
> new
> type, 'cbd' (concise bounded description), wtih sub-types being used to
> indicate particular 'encodings' - cbd/rdf+xml - or by concatenating cbd
> into the sub-type - 'application/cbd+rdf+xml' ??

Uggg, no please! We have enough pain with trying to figure
out how to interpret all the +xml MIME types to add yet
another dimension of complexity into the mix.

A CBD is not a MIME type. And thus MIME should not be
used to ask for CBDs.

>
> Sorry, I appreciate that this is probably going over old ground and 
> that
> I'm not familiar enough with the history of this discussion.  But the
> answer in the FAQ doesn't seem, to me, to be a compelling reason to use
> URIQA rather than content negotiation.

I guess I need to update the FAQ then... sorry it wasn't clear enough
for you (and hopefully the above comments, and comments elsewhere
in this thread, fill in some of the gaps).

Patrick

>
> Andy
> --
> Distributed Systems, UKOLN, University of Bath, Bath, BA2 7AY, UK
> http://www.ukoln.ac.uk/ukoln/staff/a.powell/      +44 1225 383933
> Resource Discovery Network http://www.rdn.ac.uk/
>

Received on Sunday, 10 April 2005 17:40:10 UTC