XDI protocol bindings (was RE: [xdi] Groups - XDI RDF Model V5 (xdi-rdf-model-v5.pdf) uploaded)

["Drummond Reed" <drummond.reed@cordance.net>]

Bill,

You sent the message below six weeks ago and I have been meaning to read up
on WS-RT (Web Service Resource Transfer) ever since. I finally got the
chance to do it today. A few points:

* First, I agree very much that XDI as a protocol should be specified
abstractly, and bound to different transport or transfer protocols as
needed.

* Secondly, I like your suggested set of bindings. I assume the reason for
each binding after HTTP(S) is to take advantage of libraries that already
exist to do RDF/SPARQL, SOAP/WS-RT, and SOAP/XMPP. These all make sense for
that reason. I'm assuming we will learn a great deal from the HTTP(S)
binding, so tackling the others after that will be much easier.

* I submit that one of our highest priorities should be making XDI/HTTP(S)
very lightweight. In other words, one of the key benefits of using the XDI
RDF data model (and which may also apply to the XDI ATI data model) is that
it can be very easy to create XDI resources and begin exchanging/sync'ing
them with XDI messages over HTTP(S). No RDF, SOAP, or WS-* libraries would
be required. Then XDI will "upscale" into those richer environments without
losing any fidelity, i.e., you can interact with an XDI resource via any
binding of the XDI protocol that the service endpoint supports and they will
all give you identical results.

Does this all make sense?

=Drummond 

> -----Original Message-----
> From: Barnhill, William [mailto:barnhill_william@bah.com]
> Sent: Thursday, May 24, 2007 7:00 AM
> To: Drummond Reed; Sakimura, Nat（全社サーバへ移行済み）;
xdi@lists.oasis-open.org
> Subject: RE: [xdi] Groups - XDI RDF Model V5 (xdi-rdf-model-v5.pdf)
> uploaded
> 
> 
> I understand and agree that HTTP binding will likely be the one most used
> initially, but I'm against tightly coupling the XDI protocol to an HTTP
> binding. I think we should stay focused on an abstract protocol and cover
> that in the XDI protocol document, each binding in a separate document,
> with
> http being the first binding.  At it's heart the XDI protocol seems to me
> to
> be very RESTful (Create, Put, Get, Delete), with the caveat that fragments
> of documents are addressable not just the documents. I imagine that the
> abstract protocol will focus on XML document exchanges to enable those
> operations, and I'm not convinced we need to invent another protocol from
> scratch.  One existing protocol that's caught my eye for this purpose is
> WS-Resource Transfer (WS-RT).  While designed to be encapsulated in a SOAP
> envelope it could be used outside of SOAP. It is backed by IBM, HP, BEA
> and
> others, and I believe I read that they intend to submit it to OASIS.
> 
> Btw, with HTTP binding do you mean a RESTful binding or a SOAP Binding? I
> think they should be treated as two separate bindings, and even if WS-RT
> is
> not used for the abstract protocol it could be used for the SOAP binding.
> This could also eliminate the need for an XMPP binding, since XMPP has
> SOAP/XMPP.
> 
> What I envision for bindings is:
>    RESTful using HTTP methods to exchange XDI RDF documents
>    Semantically-RESTful using SPARQL and SPARL-Update Language
>    SOAP using WS-RT
>    XMPP using WS-RT through SOAP/XMPP
> 
> I haven't yet, but will read the RDF model V5 before the meeting in time
> to
> send comments to the list, if I have any, and will be at the meeting.
> 
> Thanks,
> Bill
> --
> William Barnhill                       Phone: (315) 491-6765
> Associate                               Email: barnhill_william@bah.com
> Booz | Allen | Hamilton             i-name: =Bill.Barnhill
> "Delivering results that endure"
> 
> -----Original Message-----
> From: Drummond Reed [mailto:drummond.reed@cordance.net]
> Sent: Wednesday, May 23, 2007 7:34 PM
> To: 'Sakimura, Nat（全社サーバへ移行済み）'; xdi@lists.oasis-open.org
> Subject: RE: [xdi] Groups - XDI RDF Model V5 (xdi-rdf-model-v5.pdf)
> uploaded
> 
> Nat,
> 
> In the binding of the XDI protocol to HTTP, I agree completely. This XDI
> RDF
> model proposal still focuses just on the abstract protocol, and not any
> bindings yet.
> 
> =Drummond
> 
> -----Original Message-----
> From: Sakimura, Nat（全社サーバへ移行済み）
[mailto:sakimura@spmd.nri.co.jp]
> 
> Sent: Tuesday, May 22, 2007 12:05 AM
> To: drummond.reed@cordance.net; xdi@lists.oasis-open.org
> Subject: RE: [xdi] Groups - XDI RDF Model V5 (xdi-rdf-model-v5.pdf)
> uploaded
> 
> Don't we need to specify the status code of the http response for both
> success and failure?
> 
> Nat
> 
> > -----Original Message-----
> > From: drummond.reed@cordance.net [mailto:drummond.reed@cordance.net]
> > Sent: Tuesday, May 22, 2007 1:55 PM
> > To: xdi@lists.oasis-open.org
> > Subject: [xdi] Groups - XDI RDF Model V5
> > (xdi-rdf-model-v5.pdf) uploaded
> >
> > This is an editorial update to the XDI RDF proposal to update
> > references and add some additional explanatory text. No substantive
> > changes from V4.
> >
> >  -- Mr Drummond Reed
> >
> > The document named XDI RDF Model V5 (xdi-rdf-model-v5.pdf) has been
> > submitted by Mr Drummond Reed to the OASIS XRI Data Interchange (XDI)
> > TC document repository.
> >
> > Document Description:
> > PDF of V5 of the XDI RDF model proposal.
> >
> > View Document Details:
> > http://www.oasis-open.org/apps/org/workgroup/xdi/document.php?
> > document_id=24064
> >
> > Download Document:
> > http://www.oasis-open.org/apps/org/workgroup/xdi/download.php/
> > 24064/xdi-rdf-model-v5.pdf
> >
> >
> > PLEASE NOTE:  If the above links do not work for you, your email
> > application may be breaking the link into two pieces.
> > You may be able to copy and paste the entire link address into the
> > address field of your web browser.
> >
> > -OASIS Open Administration
> >