OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

xdi message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Minutes: XDI TC Telecon Thursday 1-2PM PT 2009-04-30

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

Date:  Thursday, 30 April 2009 USA
Time:  1:00PM - 2:00PM Pacific Time (20:00-21:00 UTC)


Bill Barnhill
Markus Sabadello
Giovanni Bartolomeo
Drummond Reed



Our main topic was discussing the most recent email from Giovanni and
Drummond's response:


The following notes are taken from the chat room transcript.

Drummond began with a question: according to Giovanni's proposal, does the
XDI RDF statement =drummond/+cat imply the XDI RDF subject =drummond+cat?

Giovanni felt it does.

Drummond gave the examples of XDI RDF statements identifying his two cats:
=drummond/+cat/=drummond*hobbes, and =drummond/+cat/=drummond*stormy

There was consensus that =drummond+cat identifies the set of all members of
=drummond/+cat. (See also the notes further below.)

We then discussed identifying individual members of this set. Giovanni
asserted that =drummond/+cat/*hobbes  infers =drummond+cat*hobbes. Bill
asked if =drummond/+cat/*hobbes infers =drummond/+cat/=drummond*hobbes? To
put it another way, if *hobbes is relative is the base XRI to which it is
relative automatically the subject of the statement or the document XRI, or
are they the same? I.e., is every document the statements describing a
single subject, or contexts within that subject?

Drummond clarified that every XDI RDF subject is relative to its context;
every XDI RDF predicate is relative to its subject, and every XDI RDF object
is relative to its predicate.

Giovanni said that, rather than =drummond/+cat/*hobbes, he prefers the form
in which all the XRIs are absolute, i.e.,

Drummond next asked if, by Giovanni's proposal, +x/+y/+z  infers that the
set exists +x+y+z, and also that it is a member of the set +x+y. In other
words, does 
+x/+y/+z infers BOTH +x+y and +x+y+z?

And if so, what is inferred by +x+y+z+k?

This kicked off a long discussion of how XDI RDF statements identify
set/member relationships. The concensus was that every XDI RDF subject can
represent both a set and a member of another set.

Bill summed it up this way, "So our consensus at this point can be described
in the following way:  XDI RDF statements +S that consist of only XDI RDF
subjects represent the set of relationships going out from +S; the form +S+P
is the set of relationships of type +P; where relationship is a 2-tuple of
(type, object}. He then asked if you query for +S+P, what you should get?

For example, if you do a query on =drummond, you will get back the graph
root on =drummond, but if you do a query on =drummond+email, do you get back
a graph rooted on =drummond/+email?

Bill's answer is that the XDI graph rooted in +S that is filtered so that
there are no arcs out of +S that are not +P

At that point Drummond asked a key question: are we saying that
=drummond+email identifies the same set as =drummond/+email?

Giovanni said no, if you query =drummond+email, you get all statements with
subject =drummond+email. Drummond agreed.

Drummond then clarified his question: is that graph defined to be the same
as the graph of all objects that satisfy =drummond/+email? Or, as Bill put
it more specifically, is =drummond+email equivalent to (=drummond/+email)
(note the cross-reference encapsulation of the latter which makes it a
single XDI RDF subject).

This was a particularly deep insight, because it shows that if the pattern
is nested more deeply than two XRI subsegments, then the recursion is
relative to each set in order of precedence (left to right). In other words,
the XDI RDF subject +x+y+z would be equivalent to ((+x/+y)/+z), and the
subject +x+y+z+k would be equivalent to (((+x/+y)/+z)/+k).


The next call will be at the standard time next Thursday, May 7.

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]