RE: [xri] Potential breakthrough

["Barnhill, William" <barnhill_william@bah.com>]

Title: RE: [xri] Potential breakthrough

Marty, you said "It must be entirely _IMPOSSIBLE_ for Ootao to assign the identifier
Boeing uses for a mountain to a radio transmitter. It must be
_IMPOSSIBLE_ for Ootao to assign ANY identifier in the @boeing
namespace."

I agree, but isn't the identifier @ootao@boeing*123 in the @ootao namespace, not the @boeing namespace?

To put it another way using the owl vs mountain example: Let's say ootao tracks owl nesting places, and that an Owl resides on the mountain @boeing*123. Then @ootao@boeing*123 might be used to identify 'the owls living on' (@ootao's information context) 'the mountain' (information derived from the meaning of 'mountain', as ideally @boeing*123 would have a link to +mountain) 'identified by @boeing*123'.

Granted that @ootao*(@boeing*123) is a more precise way of putting it, but it seems logical that @ootao could, and would for readability, just use @ootao@boeing*123.

One problem I could see is who holds the data? To me @ootao*(@boeing*123) implies that @ootao is the data authority, wheras @ootao@boeing*123 implies that either @ootao or @boeing is the data authority.

Do you see the above use case as something that should be doable, or something that shouldn't be doable?

Also, are you saying that @ootao@boeing*123 should always imply that @boeing is the data authority?

My take is that would mean that if @ootao wants to make any statements about @boeing*xxx where xxx is an identifier in the @boeing namespace, @ootao would need to use @ootao*(@boeing*xxx) form, right?

Bill

-----Original Message-----
From: Schleiff, Marty [mailto:marty.schleiff@boeing.com]
Sent: Thu 5/3/2007 1:24 PM
To: Drummond Reed; xri@lists.oasis-open.org
Subject: RE: [xri] Potential breakthrough

Drummond,

Three messages down in this thread you said that both of the following
should be expressed as "@ootao@boeing*reputation".

        1.1) How do you propose to represent "@boeing*reputation" in the
context of "@ootao"?

       

        1.2) How do you propose to represent "*reputation" in the
context of  "@ootao@boeing"?

I can't agree with this, so two messages down I told the Owls &
Mountains story to illustrate the problem with your view.

One message down you responded that:

        2.1) A mountain /\/\/\/\ named "@boeing*123" in the context of
"@ootao" should be "@ootao@boeing*123".
        2.2) An owl oVo named "123" in the context of Ootao's
Boeing-built transmitter should be "@ootao*(@boeing)*123". 

I agree completely with this.

Unless you misunderstood the original questions (1.1 and 1.2), your
answers to 1.x seem inconsistent with your suggestions for 2.x. I'll
assume I didn't ask the original questions clearly enough.

Here's something you said one message down that you're not stating
clearly enough (resulting in an incorrect statement):

        "While it's entirely _possible_ for Ootao to decide it wants to
assign the identifier Boeing uses for a mountain (@boeing*123) to an
radio transmitter".

It must be entirely _IMPOSSIBLE_ for Ootao to assign the identifier
Boeing uses for a mountain to a radio transmitter. It must be
_IMPOSSIBLE_ for Ootao to assign ANY identifier in the @boeing
namespace.