W3C home > Mailing lists > Public > public-rww@w3.org > October 2011

Re: Using Twitter as an IdP space for WebID

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Sat, 29 Oct 2011 13:40:12 -0400
Message-ID: <4EAC3A7C.7040505@openlinksw.com>
To: Michael Brunnbauer <brunni@netestate.de>
CC: public-rww@w3.org
On 10/29/11 1:30 PM, Michael Brunnbauer wrote:
> Hello Kingsley,
> On Sat, Oct 29, 2011 at 12:04:38PM -0400, Kingsley Idehen wrote:
>>> this is a good Idea but I would call this TwitterID and not WebID.
>> No, its an InterWeb scale identifier that's verifiable using the WebID
>> verification (authentication) protocol.
> http://www.w3.org/2005/Incubator/webid/spec/ says that "The WebID Profile
> document must expose the relation between the WebID URI and the Identification
> Agent's public keys using the cert and rsa ontologies, as well as the cert or
> xsd datatypes.".
> I cannot find anything about key fingerprints in the draft or about extracting
> them out of a HTML page or via the Twitter API.

Of course it isn't in there. The definition is a work in progress 
(IMHO). I say this (again) because nothing in WebID needs to be 
constrained by syntax specifics.
> You seem to use the twitter API in a hardcoded way to verify the fingerprint.
> So you will have to add code for every other social network. This does not look
> like a "protocol" to me.
>> <http://twitter.com/kidehen#this>  is a URI associated with an object
>> accessible from Twitter's Web accessible data space.
> wget -q -O- --header "Accept: application/rdf+xml" http://twitter.com/kidehen |head -1
> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">

My relying agent makes a Linked Data Object from the URL above. The end 
result is a Linked Data Object that has the association. I have OWL and 
owl:sameAs in use. In reality, I've just spewed out implementation 
details (which are ultimately always distractions relative to the concept).

The concept:
Verify an identifier via a trust logic oriented relation that is exposed 
via a de-referencable URI. That's it.

Use of Modulus and Exponent is one way. A Fingerprint is another. A 
relying agent can implement either or both. We are integrators, and we 
implement both.

Here is cURL output based on the proxy linked data URI we construct 
inside the relying agent:

curl -I -H "accept: application/rdf+xml" 
HTTP/1.1 303 See Other
Server: Virtuoso/06.03.3131 (Linux) x86_64-generic-linux-glibc25-64  VDB
Connection: close
Date: Sat, 29 Oct 2011 17:36:31 GMT
Accept-Ranges: bytes
TCN: choice
Vary: negotiate,accept
Content-Location: /about/data/entity/xml/http/twitter.com/kidehen
Content-Type: application/rdf+xml; qs=0.95
Content-Length: 0

>>> What is the connection to linked data ?
>> The aforementioned URI is de-referencable. It resolves to a structured
>> representation it the URI referents description.
> Huh ? Are you talking about the HTML document I get with the command above ?

See output above.

>>> Or is WebID about promoting the use of X.509 certificates in general ?
>> No, it is about verifiable identity via identifiers via exploitation of:
>> 1. existing InterWeb infrastucture
>> 2. trust logic and structured data representation.
> If I use the word WebID, I mean this draft: http://www.w3.org/2005/Incubator/webid/spec/
> I think the majority on this list uses the term in the same way. Correct me
> if I am wrong.


WebID canonical definition is a WIP. Implementation details != canonical 
definition that's devoid of implementation details style specificity.

>>> You are using http://twitter.com/{your-twitter-handle}#this
>>> Are you again creating URIs in other peoples namespaces ?
>> How can I create a URI in other people's namespaces associated with
>> their data spaces? I am a URI to identify a chunk of network accessible
>> data in a manner that's distinct from the representation of said chunk
>> of data.
>>>   Maybe you should
>>> omit the hashtag in the subject alternative name.
>> Of course note, a chunk of data (i.e., a data object) is endowed with
>> the following fundamental characteristics:
>> 1. an Identifier (a URI)
>> 2. Representation (a graph pictorial in EAV/SPO triple form)
>> 3. an Access Address (typically a function specific identifier e.g., a URL).
>> I hope this clears up matters for you :-)
> Not at all :-) If twitter does not use the #this hashtag to identify a person
> or account, you should not do it.

Sorry, I don't agree. I think you are missing some critical subtleties 
that are inherent to the concept of identifiers.

URIs identify things.

> Regards,
> Michael Brunnbauer



Kingsley Idehen	
President&  CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen

Received on Saturday, 29 October 2011 17:40:50 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:39:55 UTC