Minutes: XDI TC Telecon Thursday 2011-07-07

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

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

Date:  Thursday, 07 July 2011 USA

Time:  1:00PM - 2:30PM Pacific Time (20:00-21:30 UTC)

ATTENDING

Joseph Boyle 

Giovanni Bartolomeo

Drummond Reed 

Bill Barnhill

Mike Schwartz

THE IDEARPAD LINK FOR TODAY IS:

     http://xdi.idearpad.org/35

1) OPENXDI PROJECT UPDATE

Mike reported that they have implemented a good portion of link contracts. Mike explained that he is very interested in $if statements for a particular use case: evaluating a condition for a link contract. The use case that requires this is parental approval of link contracts for their children to do data sharing with other children. Mike called this "declarative security".

We also discussed $not. Mike explained that they are implementing $not statements in link contracts but the one unexpected situation was that $do$not/$get is harder to implement because it requires deleting subgraphs from the $get graph.

Bill said he has been using $where clauses, based on a proposal he sent to the list in May, to define conditions in link contracts.

# BILL to send a summary of his $where proposal to the list (or post to the wiki) so Mike and others can evaluate it.

2) MAPPING TO RDF CONCEPTS

We reviewed Giovanni's email about OWL RDF statements and mapping them to XDI:

  http://lists.oasis-open.org/archives/xdi/201107/msg00004.html

In this email Giovanni proposes that we can provide bridges between XDI and RDF -- specifically that in the XDI 1.0 specifications, we put in a normative mapping to OWL and RDF concepts.

Giovanni put it this way: "For many features we have introduced into XDI, there are ways to describe them in traditional RDF, OWL, SKOS, etc. So if we decide to have them in the XDI 1.0 standard, we should provide a normative section containing their mapping into existing ontologies."

Bill  then explained his issue with the difference between ObjectProperty and  DatatypeProperty in RDF. The two concepts are different, and require different processing in RDF, which Bill says has caused unnecessary overhead. In Bill's view, the ability to express literals as  (data:,xxx) references avolds the need to have to make this distinction.

# ACTION ITEM: Everyone send an email to the list with their thoughts about the use of the ! literal syntax.

Drummond explained that he's begun thinking about expressing RDF graphs in the XDI graph model, i.e., both visual depictions and as XDI statements. He would like to include several examples in the next version of the XDI Graph Patterns document to show what these look like.

Giovanni then suggested that we look at how Linked Data starts with a fundamental unit of data: subject/predicate/object triples. He feels XDI has a more powerful formulation because it allows you to go inside these triple statements, to express concepts like context. RDF tried to do the same thing with reification but it has not been successful. 

Giovanni pointed out that Tim Berners-Lee has talked about how the RDF model could be extended to have sub-units inside RDF statements. This is what Giovanni feels is our biggest achievement with XDI. He feels this is highly relevant to the subject of provenance -- an issue currently being widely discussed in Linked Data.

Giovanni summarized it by saying XDI provides a simpler and more natural approach to the subject of reification and provenance tracking.

Drummond suggested that there are options for defining normative mappings from RDF/OWL ontology concepts (terms) to XDI:

    * The first would be to just use the XRI cross-reference format of the actual RDF or OWL URI for a term.

    * The second would be to do #1 and in addition define a shorter, more readable synonym (an XDI $word) to mean the same thing.

    * The third would be only define the XDI $word, and in the XDI specification specify the normative mapping from this $word to the RDF/OWL term, but not actually define or use an XRI cross-reference to that term.

Giovanni prefers the third option because not only can we define synonyms for the RDF/OWL term, but we can also define other uses of the term. This approach is consistent with defining new concepts that do not exist natively in RDF. He used the analogy of adding a word in Italian that does not have a direct synonym in English, but the meaning of which can be explained in English.

Drummond thought that was a good analogy and in fact one which we should use in the introduction to the specs.

We then discussed a specific mechanism proposed by Giovanni whereby the XRI 3.0 specification concept of "base URI" would be applied to specify the Named Graph URI needed to express the provenance of Linked Data. Drummond agreed this should work well, and encouraged Giovanni to write that up that as a specific component of the RDF/OWL mappings we will do in the XDI 1.0 specifications.

2) EQUIVALENCE, INVERSE, AND $SAME PROPOSAL

See Drummond's email to the list and subsequent discussion.

  http://lists.oasis-open.org/archives/xdi/201107/msg00007.html

We were only able to begin to touch on this topic, with Drummond explaining the need for canonical synonyms when traversing the XDI graph. We will continue the discussion next week.

3) NEXT CALL

The next call is next week at the regular time.

------------

ONGOING ISSUES LIST

Each of these is a candidate for the agenda for future calls.

* DO WE NEED SEPARATE METAGRAPH WORDS FOR EQUIVALENCE AND INVERSION? (added 2011-06-30 - Giovanni)

This is an open issue because does not have a direct corallary in RDF.

* SYNONYM HANDLING (added 2011-06-30 - Giovanni)

This remains an open issue because it raises challenges with compatibility with RDF.

* TRANSACTIONAL INTEGRITY FOR XDI (added 2011-03-24)

Since  versioning, as one example, involves multiple transactions that must be  commited as a group, we will need to address transactional integrity.  Specifically, we need to define how this will be handled at the protocol  level, vs. the implementation level.

* PROPOSED CONSTRUCTS/OPERATORS FOR XDI

Discuss the following wiki page originally posted by Giovanni:

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

* DICTIONARY STRUCTURE

Mike would like an example of the PDX dictionary as soon as we can do it.

*   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