Tiny RDF Schema at openid.net?

Tue Jan 30 21:19:34 UTC 2007

On 30.01.2007 11:56:41, Recordon, David wrote:
>I'm not an RDF/OWL expert, though that looks reasonable to me.  How do
>we deal with Auth 1.x which uses openid.server and openid.delegate
>versus Auth 2.0 which uses openid2.provider and openid2.local_id?
To provide a little bit of background for my initial mail: My objective
is to generate some RDF triples from the OpenID hooks encoded in HTML.
The common practise these days is to create a simple custom transformation
which does the job ("GRDDLing"). However, the existing OpenID format
instructions (link+rel+href) are already compatible with a format 
called eRDF[1], which follows the DC guidelines. We could directly 
re-use eRDF parsers and transformers. What's missing is a namespace
URI for the "openid" prefix in HTML. (It would actually be possibe 
to use eRDF to encode the RDF Schema information directly in the 
HTML version of the OpenID specs as well. There is no need for a 
separate schema file).

With regard to general RDFS/OWL versioning, there are different options.
If v1, v1.1, and v2 form *one* evolving spec with mostly overlapping 
terms (i.e. the semantics -and maybe even the naming- of many v2 terms
is identical to those from pior versions), the trend seems to be to 
stick to a single namespace URI and to use OWL constructs to indicate
deprecated terms before they are removed from the spec. If v2 is really
different from v1 or you want a clean start, the specs should be in 
two schema files with different namespace URIs, maybe with links from
the new one to the old (there are things like "sameAs" and 
"equivalentProperty" in OWL).

Different namespaces also means that the HTML hook prefixes should 
reflect that, i.e. "openid.server openid2.provider" is fine, 
"openid.server openid.provider" would not allow the definition of 
different namespaces URIs for the two terms in eRDF. So, with a new
prefix (openid2), there should probably also be a new spec file.

My suggestion would be to create two files, with the v2 one linking
to the v1 one via owl:equivalentProperty (if the new terms are
equivalent, that is)

Ugh, I hope that was more helpful than confusing..

Ben

[1] http://research.talis.com/2005/erdf/wiki/Main/RdfInHtml

>
>--David
>
>-----Original Message-----
>From: Benjamin Nowack [mailto:bnowack at appmosphere.com] 
>Sent: Monday, January 29, 2007 10:13 AM
>To: Recordon, David; Scott Kveton; specs at openid.net
>Cc: connolly at w3.org
>Subject: RE: Tiny RDF Schema at openid.net?
>
>On 29.01.2007 07:53:15, Recordon, David wrote:
>>I'd be happy to do it; I think we were talking about using 
>>xmlns.openid.net/foo as a format.
>Awesome :)
>
>>I think the next step would be sending a copy of the RDF file for 
>>people here to look over. :)
>
>I've attached a draft which contains already some nice2haves (e.g.
>the OWL and isDefinedBy bits which may be helpful but are not strictly
>necessary), I'm not 100% sure about the prose, and I guess DanC will
>have a comment or two as well.
>
>(The resource/about/ID attributes work similar to HTML's href/id, they
>use the doc's URL as base, i.e. if the file was published at
><http://xmlns.openid.net/auth>, the full term URIs would be
><http://xmlns.openid.net/auth#server> etc.)
>
>[[[
><?xml version="1.0"?>
><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>  xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
>  xmlns:owl="http://www.w3.org/2002/07/owl#">
>
>  <owl:Ontology rdf:about="">
>    <rdfs:label>OpenID Authentication Schema</rdfs:label>
>    <owl:versionInfo>2007-01-29</owl:versionInfo>
>    <rdfs:comment>
>      A basic schema for core OpenID authentication terms.
>    </rdfs:comment>
>  </owl:Ontology>
>  
>  <rdf:Property rdf:ID="server">
>    <rdf:type
>rdf:resource="http://www.w3.org/2002/07/owl#ObjectProperty"/>
>    <rdfs:label>server</rdfs:label>
>    <rdfs:comment>
>      The OpenID Identity Provider to be used for authentication.
>    </rdfs:comment>
>    <rdfs:isDefinedBy
>rdf:resource="http://openid.net/specs/openid-authentication-1_1.html" />
>  </rdf:Property>
>  
>  <rdf:Property rdf:ID="delegate">
>    <rdf:type
>rdf:resource="http://www.w3.org/2002/07/owl#ObjectProperty"/>
>    <rdfs:label>delegate</rdfs:label>
>    <rdfs:comment>
>      The delegated OpenID Identifier to be used for authentication.
>    </rdfs:comment>
>    <rdfs:isDefinedBy
>rdf:resource="http://openid.net/specs/openid-authentication-1_1.html" />
>  </rdf:Property>
>
></rdf:RDF>
>]]]
>
>Best,
>Ben
>
>
>>
>>Thanks,
>>--David
>>
>>-----Original Message-----
>>From: specs-bounces at openid.net [mailto:specs-bounces at openid.net] On 
>>Behalf Of Scott Kveton
>>Sent: Monday, January 29, 2007 7:42 AM
>>To: Benjamin Nowack; specs at openid.net
>>Cc: connolly at w3.org
>>Subject: Re: Tiny RDF Schema at openid.net?
>>
>>With just a quick look at this, it seems like a good idea.  I'd like to
>
>>see it happen somehow.
>>
>>Anybody see any problems with doing this?
>>
>>- Scott
>>
>>
>>
>>
>>
>>On 1/29/07 2:13 AM, "Benjamin Nowack" <bnowack at appmosphere.com> wrote:
>>
>>> 
>>> 
>>> Hi,
>>> 
>>> I was wondering if you guys could be persuaded to host a little RDF 
>>> Schema file on the openid.net site. As far as I can tell, there is 
>>> great support for OpenID among SemWeb folks as it can be combined 
>>> with
>>
>>> things like FOAF for all sorts of cool applications.
>>> 
>>> People recently started to write RDF extractors for the OpenID hooks 
>>> embedded in HTML (openid.server/delegate). As these hooks are in line
>
>>> with the Dublin Core guidelines [1], there are even multiple ways to 
>>> do this. The only thing we're missing for more widespread use is an 
>>> agreed-on namespace URI for the core openID terms ("server" and 
>>> "delegate"). And ideally this would be an openid.net one. So here is 
>>> my request: any chance we could put a little RDF Schema file on the 
>>> openid server? We would of course provide the file (it'd be just 5-10
>
>>> lines of XML), and the actual URL/path doesn't really matter. An 
>>> alternative could be to host it in some other stable URI space, Dan 
>>> Connolly (CC'd) might be able to provide one at w3.org, not sure. It 
>>> would be cool to get your blessing either way, though.
>>> 
>>> 
>>> Cheers in advance for perhaps considering it, Ben
>>> 
>>> --
>>> Benjamin Nowack
>>> 
>>> Kruppstr. 100
>>> 45145 Essen, Germany
>>> http://www.bnode.org/
>>> 
>>> 
>>> [1] http://www.dublincore.org/documents/dcq-html/
>>> 
>>> _______________________________________________
>>> specs mailing list
>>> specs at openid.net
>>> http://openid.net/mailman/listinfo/specs
>>> 
>>
>>_______________________________________________
>>specs mailing list
>>specs at openid.net
>>http://openid.net/mailman/listinfo/specs
>>
>>
>
>