Minutes: XDI TC Telecon Thursday 1-2PM PT 2008-11-06

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

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

Date:  Thursday, 06 November 2008 USA
Time:  1:00PM - 2:00PM Pacific Time

ATTENDING

Mike Mell 
John Bradley 
Drummond Reed
Giovanni Bartolomeo 
Markus Sabadello

REGRETS

Bill Barnhill



1) PREP FOR XRI TC F2F

Drummond reported he is writing up a wiki page summarizing the case for
global cross-references as discussed on the previous two calls. (Editors
Note: this is now posted at
http://wiki.oasis-open.org/xri/XriThree/GlobalCrossReferences.)

He also posted a proposal called "Super Simple SEP Selection" at
http://wiki.oasis-open.org/xri/XriThree/SuperSimpleSepSelection. We
discussed that briefly and reactions were generally positive.

A major focus of the F2F will be to clarify these issues so that XDI 1.0 can
move forward with these dependencies on XRI 3.0.


2) $DO

In the Higgins F2F meeting in Boston last week, questions came up about how
to handle multi-part operations and transactional integrity in XDI. We
discussed a high-level proposal for how this can be accomplished using the
proposed $do logical operation.

The basic idea is that any non-atomic XDI operation (e.g., anything outside
of a single $get, $add, $mod, $del) would be a $do, and that only the $do
subgraph would support multiple operations. In addition the $ self-context
of the $do subgraph is where the semantics for controlling transactional
integrity would be placed. Following is a brief example in X3 Simple:

______________________
=example
	$is$a
		=
	$do
		/
			$	<-- subject for transactional integrity
control predicates -->
				$do$all	<-- example trans. int. predicate
-->
					$true	<-- all operations must
succeed -->
			=example
				$add
					/
						{some-stuff}
				$del
					/
						{some-other-stuff}
				$get
					/
						{some-more-stuff}
________________________


We discussed if a $do can contain an another $do? Yes, they should be able
to be recursive, although the return rules need to be very clear.

We discussed conditional operations, stored procedures, and whether a client
should be able to request that a server evaluate something and execute a
subgraph. John made the point that stored procedures without conditionals
would be fairly limited. However it is a slippery slope RE how far we can go
in that direction.


3) IIW PREP

We discussed XDI-related sessions for IIW. The consensus was that the agenda
is likely to be very crowded, but that we should schedule one session for a
review of the XDI4J utilities by Markus.

# MARKUS to offer an XDI4J utilities session.


4) NEXT MEETING/STARTING SPEC DRAFTING

There will be no meeting next week due to Internet Identity Workshop and the
XRI TC F2F. Drummond suggested that the following week we should begin
assigning spec editing tasks for the first three specs we identified we
wanted to write first (see http://wiki.oasis-open.org/xdi/XdiOneSpecs).