Re: WebID definition proposal with any kind of URI

On 16 Nov 2012, at 15:04, Stéphane Corlosquet <scorlosquet@gmail.com> wrote:

> That's a good use case Jürgen, something that was discussed on the mailing list a long time ago. We should consider this use case of non-HTTP URIs.

I am looking for a way of defining these things properly.

> 
> During ISWC Boston, I had a discussion with Bart who was arguing that there are scenarios that don't necessarily rely on dereferenceable HTTP URIs such as an Active Directory which would have its own way of looking up AD credentials in its db (its own way of "dereferencing" an AD URI). I'm cc'ing Bart hoping he can chime in and give more details.
> 
> So maybe there is some need to have a high level WebID protocol definition for URIs in general, and separately define the HTTP URI / Linked Data way of doing WebID protocol as a subset of the generic protocol. Others interested in AC, LDAP, snail mail can work on these in parallel. Think decoupling.

snail mail can be covered in my view with Identity Interoperability.
  http://www.w3.org/2005/Incubator/webid/wiki/Identity_Interoperability

LDAP is possible, but it makes things a lot more complicated to define. So my view
is one can leave this to another later time. There is not point having a specification nobody
implements.

> 
> Jürgen feel free to add this to the wiki as well, or I'll do it later when I get to work...
> 
> More below:
> 
> > PROPOSAL turnguard
> >
> > A WebID is a URI that denotes an Agent, like a Person, Group,
> > Organization, Software or the like. When dereferenced the WebID returns
> > a document that uniquely describes its referent.
> The URI when de-referenced returns a document.
> The document uniquely describes its ( the WebID I suppose? )'s referent
> But you just said above that de-referencing the URI returned a document.
> Your definition is a bit loopy :-)
>   I think you need to define what  de-referencing a URI for any
> URI scheme means. Also you need to explain how that would work with
> telnet and irc uris.
> 
> Yes, we know dereferencing in the case of HTTP, but not sure what the means for other URI schemes. But that is not an issue in the decoupled approach I'm suggesting above, where at the top level we define the WebID protocol not tied to any kind of URI, and separately we define the HTTP URI WebID protocol as we know it today. The other kind of URIs can be worked out elsewhere without blocking progress on HTTP WebID.

yes, I am really getting the feeling that the time of trying to please everybody is over. It did 
not make more people adopt our spec, and there is more of a risk that people looking it
are getting the feeling that it is in fact impossible to implement. 

We should now be pushing for something that is easy to explain, and easy to implement.

We can also have an overview of how things work in general. Think of the picture of
   http://www.w3.org/wiki/WebID

But without the public key stuff. Still at the same time removing the TLS key stuff is making
our project even more abstract.

Think of html version 1. It was easy to implement by developers single handedly. Then
as it grew more features could be added. I think we are clear that the architecture we
are building on is very general and can develop nicely.

> 
> Steph.
> 
> On Fri, Nov 16, 2012 at 3:18 AM, Henry Story <henry.story@bblfish.net> wrote:
> 
> On 16 Nov 2012, at 02:04, Jürgen Jakobitsch <j.jakobitsch@semantic-web.at> wrote:
> 
> > hi,
> >
> > analog to henry's mail for any kind of URI.
> > (henry stated that people => in favour <= of the #-ur(i|l) webid
> > definition should reply to the mail with subject "WebID definition
> > proposal with hash urls". - i think that's a good idea, these threads
> > get hard to read from time to time.
> >
> > PROPOSAL turnguard
> >
> > A WebID is a URI that denotes an Agent, like a Person, Group,
> > Organization, Software or the like. When dereferenced the WebID returns
> > a document that uniquely describes its referent.
> 
> The URI when de-referenced returns a document.
> The document uniquely describes its ( the WebID I suppose? )'s referent
> But you just said above that de-referencing the URI returned a document.
> Your definition is a bit loopy :-)
>   I think you need to define what  de-referencing a URI for any
> URI scheme means. Also you need to explain how that would work with
> telnet and irc uris.
> 
> >
> > comments :
> >
> > 1. so i'm +0.9 for kingsley's suggestion
> > 2. i'd add an example of what uniquely describes means
> > 3. i'd add a sentence that lead from unique description to the realm of
> > authentication (something like : some unique descriptions (like a
> > cert:key, foaf:openid) may be used to authenticate some may not. with a
> > link to webID + TLS
> >
> > wkr j
> >
> >
> >
> >
> > --
> > | Jürgen Jakobitsch,
> > | Software Developer
> > | Semantic Web Company GmbH
> > | Mariahilfer Straße 70 / Neubaugasse 1, Top 8
> > | A - 1070 Wien, Austria
> > | Mob +43 676 62 12 710 | Fax +43.1.402 12 35 - 22
> >
> > COMPANY INFORMATION
> > | web       : http://www.semantic-web.at/
> > | foaf      : http://company.semantic-web.at/person/juergen_jakobitsch
> > PERSONAL INFORMATION
> > | web       : http://www.turnguard.com
> > | foaf      : http://www.turnguard.com/turnguard
> > | g+        : https://plus.google.com/111233759991616358206/posts
> > | skype     : jakobitsch-punkt
> > | xmlns:tg  = "http://www.turnguard.com/turnguard#"
> 
> Social Web Architect
> http://bblfish.net/
> 
> 
> 
> 
> -- 
> Steph.
> 
> 

Social Web Architect
http://bblfish.net/