Minutes: XDI TC Telecon Thursday 8-9AM PT 2008-02-07

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

Following are the minutes of the unofficial telecon of the XDI TC at:

Date:  Thursday, 07 Febrary 2008 USA
Time:  8:00AM - 9:00PM Pacific Time

Event Description:
Weekly unofficial call of the XDI Technical Committee.


ATTENDING

Bill Barnhill
Markus Sabadello 
Giovanni Bartolomeo 
John Bradley
Drummond Reed
Victor Grey


AGENDA

1) DOLLAR WORD PAGE

Per the last meeting, Drummond created an XDI TC wiki page for tracking
proposed dollar words:

	http://wiki.oasis-open.org/xdi/DollarWords 

He recommended a three-step process for proposing updates:

1) Update the page with the proposed new or changed $ word.
2) At the same time, update the page change log so it is easy for TC members
to review.
3) Send an email to the TC mailing list regarding the proposed change.

Bill asked if we planned for the $ dictionary to become a live service vs.
just a paper specification. There was general consensus that this was in
fact now a desirable service as it could be very useful to developers and
implementers of XDI software and services. 

Markus mentioned that it could be hosted together with the Community
Dictionary Service project that has been started by the Identity Schemas
working group of Identity Commons. The CDS is aimed at providing a
community-maintained + space, but the $ space dictionary definitions could
also be hosted there provided they were defined by the XDI TC (and XRI TC)
first.

There was consensus that this bears further exploration and discussion with
OASIS staff, and should be addressed in our charter update.


2) HIGGINS F2F REPORT

Drummond and Markus reported on the XDI RDF presentation they gave at the
Higgins Project face-to-face meeting in Provo, Utah last week.

* Drummond said there is particular synergy with Higgins because the Higgins
Data Model is RDF-based and also seeks to enable an abstraction layer on top
of existing data repositories. Therefore almost all the fundamental concepts
of the two models line up.

* Markus said he has worked out a mapping between key Higgins Data Model
concepts and XDI RDF concepts.

# MARKUS to post this mapping as a XDI TC wiki page.

* Markus explained that from the standpoint of Higgins, XDI will be another
context, i.e., another type of data repository. Markus is implementing an
XDI context provider, i.e., a Higgins code component that exposes a data
repository as a Higgins context. In conjunction with the XDI protocol, this
will enable Higgins deployments to read/write to XDI contexts independent of
network domain.

* Drummond said the other key area of synergy is link contracts. Higgins is
just beginning the work of developing a context-independent authorization
model -- something that was tried but never accomplished in the LDAP world.
Since both Higgins and XDI need the same fundamental semantics for this
model, there is a high potential to cooperate.


3) UPDATED XDI RDF MODEL DOCUMENT

Drummond, Markus, and Paul posted an V8 of the XDI RDF Model document last
week. The latest version is linked off a "cover page" on the XDI TC wiki:

	http://wiki.oasis-open.org/xdi/XdiRdfModel

* Drummond explained this version includes new proposals for XDI addressing,
XDI core predicates, XDI context descriptors, and the X3 compact
serialization format. It also includes many new example XDI documents all
written in the X3 Display format.

* Before the call, Bill posted a feedback document. We spent most of the
rest of the call reviewing this document:

	
http://www.oasis-open.org/committees/download.php/27112/XDI%20RDF%20v8%20Com
ments-Barnhill.pdf 

* Bill suggested that the XDI RDF model should have compatibility with the
CWM RDF tool.

	http://www.w3.org/2000/10/swap/doc/cwm.html 

* Bill said that the XDI concept of dictionaries and instance documents
corresponds to the ontology concept of "T-box" and "A-box".

	http://en.wikipedia.org/wiki/Tbox 
	http://en.wikipedia.org/wiki/Abox

* Bill said that the proposal of using $ as the XDI context self-descriptor
appears to be directly congruous to the RDF practice of using a single
underscore to represent the current document.

* There was a discussion of the use of direct concatenation in XDI RDF
addressing. There was general consensus that it made addressing much
simpler. John said that he had Steve had reviewed direct concatenation and
at this point they did not see an issue with compatability with the XDI ATI
model. Drummond noted that it appears to be the only way to represent
merging of two XDI RDF graphs that maintains the $has relationship between
the two XDI RDF subjects. To address the dependency between XDI RDF and XRI
3.0 syntax, there was general agreement that the XDI RDF specs should be
developed in conjunction with the XRI TC's development of the XRI 3.0 specs.

 
* Lastly there was a discussion of both XDI and RDF queries, and how they
would interrelate. It was agreed this is another area that needs further
exploration by the TC.