W3C home > Mailing lists > Public > semantic-web@w3.org > September 2014

Re: genid example from RDF1.1 is bad

From: Stian Soiland-Reyes <soiland-reyes@cs.manchester.ac.uk>
Date: Wed, 24 Sep 2014 15:05:46 +0100
Message-ID: <CAPRnXtk0cCCcC1f0X2cSSMas0-wFK_rF1HXL2G-MpQrY28umuQ@mail.gmail.com>
To: David Booth <david@dbooth.org>
Cc: "henry.story@bblfish.net" <henry.story@bblfish.net>, Pierre-Antoine Champin <pierre-antoine.champin@liris.cnrs.fr>, Semantic Web <semantic-web@w3.org>
On 24 September 2014 02:21, David Booth <david@dbooth.org> wrote:

>> It is clear from
>> Pierre Antoine Champin's interpretation of the RDF1.1 spec that
>> genids are not dereferenceable when in .well-known space.
>
>
> AFAICT, the RDF 1.1 spec says nothing either way about whether Skolem URIs
> should be dereferenceable:
> http://www.w3.org/TR/rdf11-concepts/#h3_section-skolemization
> So I think Pierre's interpretation is reflecting his own assumptions and
> perhaps common experience, but not the intent of the RDF 1.1 spec.

Perhaps what is not too clear from the paragraph is that it is:

> the authority responsible for the domain example.com

that is minting the URIs - which of course is also free to provide
representations of those URIs or not - just as it is when minting all
other URIs within its authority.


Now is this the domain of the producer or the consumer? Well, it's
whoever wants to get rid of the bnodes - so it could easily be that

    <http://www.example.net/resource>

is retrieved by

    <http://other.example.org/sparql>

- where it internally replaces all bnodes from example.net with

  <http://other.example.org/.well-known/genid/a9ecc0b8-53eb-474c-bf21-afef3c3a04b1>

etc. before exposing them to the SPARQL queries.


Alternatively, it could be www.example.net that replaces them at the
point of delivery - in which case it could mint
<http://www.example.net/.well-known/genid/0ac3fdad-4618-4ea6-ae10-c40c6de2e7cf>


Following regular REST principles, an attempt to resolve
http://other.example.org/.well-known/a9ecc0b8-53eb-474c-bf21-afef3c3a04b1
could say 410 Gone (query finished!), 404 (don't know.. eh..) or give
a 200 OK  with a vague hint of its provenance, or if you are very
lucky, all the statements with a9ecc0 as subject or object.


200 Vague response:

<http://other.example.org/.well-known/genid/a9ecc0b8-53eb-474c-bf21-afef3c3a04b1>
prov:alternateOf _:b01
_:b01 rdfs;isDefinedBy <http://www.example.net/resource> .

(of course you won't find anything about a9eec or even _:b01 there)


A client who want to save bandwidth and time, and have read the RDF
1.1 spec, can simply avoid resolving any URIs that have path starting
with /.well-known/genid/ because there is not likely to find any
additional information there. More naive clients will probably waste
more time - but then that's probably true for all the other resources
it comes over as well.



Now - the RDF 1.1 spec does not say if you would be guaranteed to not
find any ADDITIONAL information if you try to resolve the genid. But
this is also true for bnodes, even though it might not seem like that
at first - after all the bnode can't be referenced outside its
document - as conclusion from reasoning can infer equivalence and
additional statements about the bnode.

As you see in my attempt above I have added more information. Is this allowed?


-- 
Stian Soiland-Reyes, myGrid team
School of Computer Science
The University of Manchester
http://soiland-reyes.com/stian/work/ http://orcid.org/0000-0001-9842-9718
Received on Wednesday, 24 September 2014 14:06:45 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 07:42:53 UTC