Minutes: XDI TC Telecon Thursday 1-2PM PT 2009-02-26

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

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

Date:  Thursday, 26 February 2009 USA
Time:  1:00PM - 2:00PM Pacific Time (21:00-22:00 UTC)

ATTENDING

Giovanni Bartolomeo
Mike Mell
John Bradley
Markus Sabadello
Drummond Reed
Nick Nicholas

REGRETS

Bill Barnhill (in England)


1) XDI MIME TYPE AND XDI CLIENT-SIDE PROCESSING

Part of our XDI Serialization spec is defining one or more MIME types for
XDI documents. This has some powerful uses for triggering client-side
processing in browser scenarios. We reviewed the new wiki page on this topic
at:

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

Drummond ran through the scenarios listed there. Markus ran through the
final section detailing the MIME types already used in the XDI4J project.

John asked how the XDI client would know how to return the response to a
client. Drummond explained how it would happen with XDI - by resolving the
XRI of the "sender" (the website).

John then pointed out that the XDI request from the website could describe
that the response can be returned to the site any way the site would like
it: HTTP post, OpenID AX (attribute exchange), SAML, etc. There was
consensus this could make it more attractive to sites to adopt because it
meant fewer changes to their back-end systems.

John also noted the query must be open so that it can apply to any user.
This can be accomplished using $$ variables as Markus has demonstrated in
the XDI Query utility. 

	http://graceland.parityinc.net/xdi-querier/XDIQuerier 

We also noted that mechanism could even be used for authentication to the
site in a manner similar to information cards. We discussed how it could be
used directly in conjunction with information cards as a way to extend them.
The card essentially could serve as a security token wrapper around the XDI
messages.


2) $HAS SEMANTICS (MORE ON +X/+Y/+X+Y)

We started discussing Giovanni's message posted to:

	http://lists.oasis-open.org/archives/xdi/200902/msg00034.html

Giovanni's first point was that what =giovanni and =markus represent in
statements...

      =drummond/+friend/=giovanni
      =drummond/+friend/=markus

...are not instances of +friend. They are instances of the "individual"
class, i.e. class identified by the equals symbol (=). What we can read from
the statements =drummond/+friend/=giovanni and =drummond/+friend/=markus is
simply that two instances of the individual class +people, namely =giovanni 
and =markus, are referenced by =drummond/+friend. Actually =giovanni and
=markus would exists also if they were not +friend, and if they were
instances of +friend this would be not possible.

There was consensus Giovanni was right: the predicate +y in +x/+y/=z does
not mean that =z is a +y

Giovanni gave the example @uroma2/+dept/*telecom. This is different than
=drummond/+friend/=giovanni because *telecom is not a global identifier. So
the * and ! generally identify local instances of a global class.

In general, Giovanni is in favour of the chain pattern
instance-class-instance-class-instance-class:
@uroma2+dept*telecom+member=giovanni, etc.

Drummond agreed that pattern works well - in RDF that same pattern is called
"striping".

John noted that this is really ATI (Authority/Type/Instance) under the hood
because chains authority/type/instance to create a new authority.


3) NEXT CALL

Thursday 3/5 1-2PM PT (21:00-22:00 UTC)