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

 


Help: OASIS Mailing Lists Help | MarkMail Help

xri message

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


Subject: RE: [xdi] Global Cross-Reference Proposal


Hi,

it seems that this topic underlines a slightly different approach between the two TCs. I've been not very involved in XRI TC, as my focus is more on semantics - so please pardon me if I'm entering this "inter-TC" thread from a pure XDI perspective.
I guess that the different viewpoints are due to the fact that XRI TC focus is more on syntax, whereas XDI, in order to fulfill its goals, needs also a very powerful semantic model. In particular, the capability to "scare quote" a statement, that is, to find a way to express beliefs of entities about other entities is a fundamental one in intelligent systems (see for example belief in BDI model in software agents, http://en.wikipedia.org/wiki/BDI_software_agent#Beliefs).
Personally, I think, this feature should be supported by XDI and the proposed syntax is one of the most straightforward we can have. I also understand this has a cost from a pure XRI perspective, as Les says, aesthetics, and probably the need to explain to XRI users the difference between the two options, but, making a balance between pros and cons, I think it worths to be supported.

Thanks,
Giovanni

At 20.34 10/11/2008, Chasen, Les wrote:
I see this as two ways to say the same thing.  In both cases =Drummond is a cross reference to @xpinion.    In Drummond’s proposal I guess one of them refers to =Drummond in scary quotes and the other doesn’t.  Do we really need that?
 
The original reason (and AFAICT the only reason) for this syntax was aesthetics.  I agree it is easier to read and type the more compact syntax of @xpinion=Drummond.  However, inventing a distinction between the two cross references creates confusion. 
 
From: Barnhill, William [USA] [ mailto:barnhill_william@bah.com]
Sent: Monday, November 10, 2008 2:14 PM
To: Chasen, Les; Giovanni Bartolomeo; Drummond Reed
Cc: OASIS - XDI TC; OASIS XRI TC
Subject: RE: [xdi] Global Cross-Reference Proposal
 
So here's a use case that's near and dear to me:
 
Let's say someone has a site that is an XRI/XDI enabled version of Epinions.com, @xpinions. @xpinions wants to make an assertion about =Drummond that they control, not that =Drummond controls. Let's say that Drummond is communitive =>
@xpionions*(=Drummond)/$is/(+communitive)   [Btw, <- correct? shorter way of stating?]
 
Let's say they also want to reference/pt to a statement that =Drummond says he's open =>
@xpinions=Drummond/$is/(+open)
 
Does the above make sense and show need for both? Also is syntax correct for v3, and is there a more concise way to express it?
 
Thanks,
=Bill.Barnhill
 
 

From: Chasen, Les [ mailto:les.chasen@neustar.biz]
Sent: Monday, November 10, 2008 2:03 PM
To: Giovanni Bartolomeo; Drummond Reed
Cc: OASIS - XDI TC; Barnhill, William [USA]; OASIS XRI TC
Subject: RE: [xdi] Global Cross-Reference Proposal
I don't see the importance here.  You are introducing a semantic "scare quote" difference between the current "local" cross references and the proposed "global" cross references.  I find that distinction so very nuanced that it is confusing.  I don't understand why we want to introduce this into XRI.
 
 
From: Giovanni Bartolomeo [ mailto:giovanni.bartolomeo@uniroma2.it]
Sent: Monday, November 10, 2008 1:01 PM
To: Drummond Reed
Cc: 'OASIS - XDI TC'; Barnhill, William [USA]; Chasen, Les; 'OASIS XRI TC'
Subject: RE: [xdi] Global Cross-Reference Proposal
 
Thanks Drummond, it clarifies indeed! Wow, that's a very powerful mechanism!
And should also answer Bill's question:

"If the following is now a reference to the global =Drummond:
      @cordance=drummond              <== global scope – NOT ALLOWED IN XRI 2.0
then how would @cordance make a statement 'about' =Drummond?"

is my understanding correct?

Thanks!
Giovanni

At 18.13 10/11/2008, Drummond Reed wrote:
Giovanni,

I forgot to cc the XDI TC list on my reply to Les this morning. I'm
including it below just to clarify the question you asked. Note that your
point about the XDI statement being reflected only applies to XRIs in an XDI
context (which is of course what the XDI TC is focused on). In XRI 3.0 the
semantics are more universal, i.e., what does an XRI cross-reference mean in
_any_ context.

Best,

=Drummond

********** DRUMMOND'S EARLIER REPLY TO LES ***************

Les, you're right that in both @cordance=drummond and @cordance*(=drummond),
you can infer a semantic relationship between the subsegments (@cordance and
=drummond) and their global XRIs.

However, in the latter case, I think the only interpretation you can infer
is that the entity being referred to by *(=drummond) is being referred to IN
A DIFFERENT CONTEXT than =drummond.

The rationale is the same as the example I gave in natural language on the
wiki page, which is when quotes are used in English to explicitly
communicate that a particular word/phrase is being used in "quotes",
specifically what are called "scare quotes" (thanks to John Bradley for
educating me about that phrase). There's a good Wikipedia article on scare
quotes:

         http://en.wikipedia.org/wiki/Scare_quote

The example I used on the wiki page was:

        Boeing "international" employment contract signature date

In that phrase, the only thing an English reader can interpret about the
word "international" is that it is being used in a different context than
the normal use of the word. I think it's safe to say that in order to
understand what context applies (and therefore what the word really means in
that context), you have to look either at the immediate context of the full
sentence, or the full paragraph, or the full context in which it appears.

Which is exactly the meaning of "local context symbols" in XRI.

Therefore the following line up nicely:

        Boeing "international" employment contract signature date
         @boeing*(+international)+employment$contract$sig$d

...and...

        Boeing international employment contract signature date
         @boeing+international+employment$contract$sig$d

I don't think there's any more semantics that can be interpreted from the
difference. However I think you can see why the issue is so important to XRI
3.0 -- in XRI 2.0 the _only_ cross-reference that can be expressed in the
"scare quote" version, whereas in normal natural language (as in typical XDI
RDF), it's the global cross-reference version that's needed.

Hope this helps,

=Drummond


> -----Original Message-----
> From: Giovanni Bartolomeo [ mailto:giovanni.bartolomeo@uniroma2.it]
> Sent: Monday, November 10, 2008 9:04 AM
> To: Chasen, Les; Drummond Reed; OASIS XRI TC
> Cc: OASIS - XDI TC
> Subject: RE: [xdi] Global Cross-Reference Proposal
>
> Hello, I'm not sure to understand your last statement:
>
>  > In both cases the definition behind =Drummond comes from @cordance.
>
> Maybe I'm getting wrong, but in XDI (which means basically in this
> XRI3.0 proposal), @cordance=Drummond can be composed if and only if
> the statement @cordance/$has/=Drummond (aggregation) holds. According
> to my understanding, this should sound: =Drummond does not come from
> @cordance, rather, it is @cordance which is an aggregation of people
> like =Drummond.
> What do you think?
>
> Thanks,
> Giovanni
>
> At 17.30 07/11/2008, Chasen, Les wrote:
> >I still do not understand how @cordance=Drummond and
> >@cordance*(=Drummond) differ.  I understand you are saying that one is
> >global and the other is local but they are both cross references to
> >=Drummond.  In both cases the definition behind =Drummond comes from
> >@cordance.
> >
> >
> > > -----Original Message-----
> > > From: Drummond Reed [ mailto:drummond.reed@cordance.net]
> > > Sent: Friday, November 07, 2008 2:54 AM
> > > To: 'OASIS XRI TC'
> > > Cc: 'OASIS - XDI TC'
> > > Subject: [xdi] Global Cross-Reference Proposal
> > >
> > > I finally had time tonight to complete the writeup of the Global
> > > Cross-Reference proposal for XRI Syntax 3.0. This is based mostly on
> > > requirements from the XDI TC that go back to the dawn of the XDI RDF
> > > model.
> > > We have been discussing them on our last few weekly telecons in order
> > > to
> > > document them in detail before the XRI TC F2F next week.
> > >
> > > The full proposal is written up at:
> > >
> > >       http://wiki.oasis-open.org/xri/XriThree/GlobalCrossReferences
> > >
> > > Since this is the proposal with the biggest impact on XRI 3.0 syntax,
> >I
> > > urge
> > > all XRI TC members to read through it and comment on the list.
> > >
> > > Thanks,
> > >
> > > =Drummond
> > >
> > >
> > >
> > > ---------------------------------------------------------------------
> > > To unsubscribe from this mail list, you must leave the OASIS TC that
> > > generates this mail.  Follow this link to all your TCs in OASIS at:
> > > https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
> >
> >
> >---------------------------------------------------------------------
> >To unsubscribe from this mail list, you must leave the OASIS TC that
> >generates this mail.  Follow this link to all your TCs in OASIS at:
> > https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
> >
> >
> >No virus found in this incoming message.
> >Checked by AVG - http://www.avg.com
> >Version: 8.0.175 / Virus Database: 270.9.0/1773 - Release Date:
> >07/11/2008 9.08
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail.  Follow this link to all your TCs in OASIS at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php



No virus found in this incoming message.
Checked by AVG - http://www.avg.com
Version: 8.0.175 / Virus Database: 270.9.0/1773 - Release Date: 07/11/2008 9.08
 

Internal Virus Database is out of date.
Checked by AVG - http://www.avg.com
Version: 8.0.175 / Virus Database: 270.9.0/1773 - Release Date: 07/11/2008 9.08



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