XDI TC Minutes
Following are the minutes of the unofficial telecon of the XDI TC at:
Date: Friday, 29 November 2013 USA
Time: 09:00AM - 10:30AM Pacific Time (16:00-17:30 UTC)
ATTENDING
Markus Sabadello
Joseph Boyle
Drummond Reed
GUESTS
Andy Dale
NEWS & UPDATES
New XDI-Related Paper
Drummond provided an update about a new privacy-related paper to be launched at the Privacy by Design User Forum in Toronto on Dec 5 in which XDI semantic data interchange plays a key role. He will send a link to the paper as soon as it is published.
PRESENTATIONS/DISCUSSIONS
Progress on Working Drafts
Drummond reported that he had completed the first draft text for the Design Goals section and posted it to the wiki. Joseph is in the processing of incorporating it into the spec. Drummond then went over the proposed structure of the XDI Graph Model section.
https://wiki.oasis-open.org/xdi/XdiDesignGoals
Drummond and Joseph’s goal is to have the text and ABNF far enough so that Working Draft 01 can be published this coming week.
XDI Symbol Table
We noted the new table Dan Blum has developed:
https://wiki.oasis-open.org/xdi/XdiSymbolTable
ABNF Update
Joseph and Drummond have been continuing to work on the ABNF for XDI Core 1.0. As sections are finished they are be mirrored on the XDI TC wiki ABNF page:
https://wiki.oasis-open.org/xdi/XdiAbnf
Joseph has done fantastic work in structuring the ABNF into sections that make it much easier to explain and digest. Drummond plans to do a close review this weekend so the ABNF is ready to go for Working Draft 01.
Terminology
On the list we have been discussing two terminology questions:
“Global graph” vs. “outer graph”.
Alternatives to the term “root” that help emphazise the heterarchical nature of the XDI graph model.
In the discussion on the call, there was a consensus that the term “common graph” and “common root” worked better than “global graph” and “global root” because it would stress the peer-to-peer community nature of the XDI graph model. Drummond will update the spec to use this term.
There was also the sense the the term “root” is unavoidable when dealing with directed acyclic graphs, which is what XDI contextual arcs produce, so we should use it but constantly be educating XDI developers that independent XDI graphs themselves are not hierarchical but heterarchical.
Merging Graphs and Graph-Relative vs. Authority-Relative Statements
We continued discussion about whether certain statements in an XDI graph should be “graph-relative” or “root-relative”, e.g.:
<$public><$key>&/&/”...”
or “authority-relative”, e.g.:
=markus<$public><$key>&/&/”...”
For this discussion, we reviewed the following diagram that shows two peer roots in relationship to the common root.
Key points from the discussion:
The two peer root nodes below the common root contain only cached copies of other XDI graphs. Although we have only discussed caching in terms of the XDI discovery spec, in fact it can apply broadly to all forms of XDI data obtained via an XDI $get operation.
All other subgraphs under the common root fall into one of three categories:
Attributes of the graph for which the graph itself is authoritiative.
-
Authorities for which this graph is authoritative (as defined by peer root $ref statements, of which there is one example above).
Authorities for which this graph is not authoritative, but which this graph contains a copy of a portion of the authoritative subgraph.
Drummond suggested the following “bright line” definition to distinguish cached vs. copied data.
Cached data is data shared under an XDI $get permission. It is always stored under the authoritative peer root node subject to cache control parameters from the authority.
Copied data is data shared under an XDI $copy permission. It is always stored under the common root and is not subject to any cache control semantics because it is an authorized copy (and, if specified by the link contract, will be updated via push—synchronized—when it changes).
Drummond emphasized again how the “graph merge rule”—that any set of XDI statements can be copied exactly bit-for-bit from one XDI peer root graph to another without conflict—establishes a clean, strong foundation for all future interoperability of XDI.
DECISION POINTS FOR THIS CALL
None scheduled.
DECISION POINT QUEUE REVIEW
The decision queue stack is shown on the following three auto-generated Category pages:
https://wiki.oasis-open.org/xdi/CategoryLastCall
https://wiki.oasis-open.org/xdi/CategoryCurrent
https://wiki.oasis-open.org/xdi/CategoryHighPriority
See also this list of proposals that need to be developed:
https://wiki.oasis-open.org/xdi/XdiPendingIssues
NEXT CALL
The next call is next week at the regular time. Note that Drummond and Dan will not be able to attend due to travel returning from the Privacy by Design User Forum conference.