Minutes: XDI TC Telecon Thursday 1-2:30PM PT 2010-10-07

[Drummond Reed <drummond.reed@xdi.org>]

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

Date:  Thursday, 07 October 2010 USA
Time:  1:00PM - 2:30PM Pacific Time (21:00-22:30 UTC) - SEE NOTE ABOVE

ATTENDING

Giovanni Bartolomeo
Michael Schwartz 
Bill Barnhill 
Joseph Boyle 
Drummond Reed 
Joe Johnston 
Markus Sabadello


THE GOTOMEETING SCREENSHARE URL IS:

 https://www2.gotomeeting.com/join/969244355

THE IDEARPAD LINK IS:

      http://xdi.idearpad.org/9

No password necessary.

Please try to preface each of your comments with your name so the transcription into the minutes is easier.
(Note: the link to last week is http://xdi.idearpad.org/6 - password was "turtle".)


1) GENERAL UPDATE & ANNOUNCEMENTS

Drummond welcomed Michael Schwartz as an official TC member. He also explained that he has now officially stepped down as Executive Director of the Information Card Foundation so he can focus exclusively on the Personal Data Ecosystem space. He encourages all TC members to take a look at http://personaldataecosystem.org. 

Markus also mentioned that the Personal Data Store project has been renamed to Project Nori, at http://projectnori.org. 

Giovanni mentioned that he is looking into working with a publisher on XDI-related projects, including a book. TC members were very enthusiastic about such a project, but were concerned about the amount of bandwidth they would have to contribute, especially during the period where we are trying to put out our 1.0 specs.


2) NEED TO PUSH FORWARD WITH 1.0 SPECIFICATION DRAFTS

There is an increasing need for 1.0 Working Drafts.

Drummond characterized the process in three stages: foundation, framing, and finish. Each will take at least a month, so we are looking at at least a three month process.

At the same time, we want to have multiple interoperable implementations being built and tested. That could slow down the speed at which we are able to move as well.

Bill made the suggestion that make the Serialization specification our top priority, because that's the leading gate to standardization.

Giovanni felt that the serialization format must be compatible with the semantics we develop or the two could get out of sync.

Bill wanted to note that he is currently working on a paper combining Event-model-F, a full formal model of events, with an implementation of a mereology (model of part-whole relations), with some basic concepts from DOLCE-Lite+DnS and SKOS.  This paper will combine these as an XDI dictionary sub-set with applied examples from the Activity Streams use cases.

A link (not a source) for Bill's paper, is http://plato.stanford.edu/entries/mereology/. 
One of the main resources Bill is using is www.inf.unibz.it/~artale/papers/appl-onto-07.pdf.
Another is http://www.uni-koblenz-landau.de/koblenz/fb4/AGStaab/eventmodel


3) JSON SERIALIZATION 

Drummond noted that have not as a group had a chance to review the final JSON serialization proposal resulting from the work at Whistler. 

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

Since this is now very stable, and forms the basis for other work on the "new  view" of the graph and the metagraph predicates, we should make sure it addresses the list of issues Bill had with  earlier proposals (at the bottom of the page).

Bill said he has reviewed the proposal and added comment at the bottom, but overall he is in favor o proceeding with it.

Drummond asked if there were any other open issues with it.

Giovanni suggested we review his Sept. 20 email,
http://lists.oasis-open.org/archives/xdi/201009/msg00030.html, 
especially point #2.

We discussed it and in the end due to lack of time we agreed to discuss via email this key point:
> Now, if "=drummond/+friend/$ is equivalent to =drummond+friend",  
> and $/=drummond/$ is equivalent, at least in my understanding, to  
> =drummond, why do we need to use these triples?

Giovanni asked whether there is a strong need that every XDI address must be expressed in the triple format (an XRI containing three segments). Drummond said no (at least to our knowledge at the time being).


4) UML DIAGRAMS

Last month Drummond posted a document summarizing some insights coming out of the XDI Retreat in mid-August:

  http://www.oasis-open.org/committees/download.php/39382/new-view-of-xdi-graph-model-v1.pdf

This subsequently led to comparison of UML diagrams for the "new  view" and "old view", as posted on the following wiki page:

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

We briefly reviewed these so TC members would be aware of the two "views" of the graph they represent.
 

5) SYMBOLIC METAGRAPH PREDICATES PROPOSAL

Drummond posted a proposal for using symbolic metagraph  predicates instead of $ words based on the insights coming out of the JSON serialization format discussions and the  "new view" of the graph.

  http://wiki.oasis-open.org/xdi/SymbolicMetagraphPredicates
  
#ACTION: Drummond to send an email and post a wiki page on the complete matrix that we would use to define the semantics of the metagraph symbols used as subjects, predicates, and objects.

Bill's question: If we have a graph containing the statement expressed by the XRI =Drummond/+friend/=Markus, and we are coloring the statements in the graph +Red or +Black, how then do we state an XRI such that the semantics implied are that the statement made by =Drummond/+friend/=Markus is '+Black'?

Drummond asks: who is making this assertion that this statement is "+black"?

Bill: For my question it is someone other than =Drummond, lets say =Bill

Drummond answers: =bill/+black/(=drummond/+friend/=markus)

Giovanni: I do not like this form, it is very far from conventional way to interpret an RDF triple and ambiguous (I read this as: =bill has an attribute +black which refers to the sentence (=drummond/+friend/=markus))

Bill: I can see that, but have to say I like it as it matches my thoughts on who the assertion authority is for a statement.

Giovanni: why not simply =bill/+assert/(=drummond/+friend/=markus)/+color/+black)

Bill: Hmm, that could work..but seems verbose? Not sure. Then you are saying that the statement is the assertion authority, I think.

Giovanni: I prefer a more verbose sentence, but semantically correct! :-) this is because otherwise we that some sentences are ambigously interpreted

Bill: Could we say 
=bill/+black/(=drummond/+friend/=markus)
entails...
=bill/+assert/(=drummond/+friend/=markus)/+color/+black)

Giovanni: problem with =bill/+black/(=drummond/+friend/=markus) is that it sounds ambiguous (see what I wrote before..)

Bill: True, but to me it isn't ambiguous. It say's Bill says (=drummond/+friend/=markus)  is +black, whatever Bill has defined +black to be.

Giovanni: ? can we go on via email maybe after the conf. call? This is an important point.

Bill: So how about if it was in the graph =Drummond, for comparison?

Drummond's answer: option a) just substitute =drummond for =bill, or b) see below.

Keep going drummond, just getting an idea down:
That works well because then you can state what statements are entailed through inferencing, I think via:
=Drummond/+inferred/(inferred XRI #1)
=Drummond/+inferred/(inferred XRI #2)
& so on. Correct?

Drummond: yes.

6) OTHER TOPICS
If we have time (this is our ongoing issues list):

* Link contracts examples.
* Equivalence semantics: close on whether  we need an additional $word that is the equivalent of Higgins Personal  Data Model (PDM) semantics of h:correlation, which is not as strong as  $is.

      http://lists.oasis-open.org/archives/xdi/201006/msg00036.html

* Cool URIs: continue previous discussion about the use of standard RDF URIs in XDI:

  http://lists.oasis-open.org/archives/xdi/201006/msg00023.html


7) NEXT CALL

The next call is next week at the regular time.