RE: [xdi] Resource and Data

Will, I'm totally with you. I strongly
favor UML terms myself.

 

Is the noun you're talking about the
potential name for a container for subresources? If so, since subresources
represent association objects describing the association of their parent
resource with another resource, <association> would seem to be a good
choice. Bil Barnill has suggested <assoc> for brevity.

 

Loren has suggested calling it
<meta> since its all metadata describing the association.

 

Of course, per the earlier messages in
this thread, before we discuss what to call it, we need to decide if adding a
container element for subresources is the best approach. So far the discussions
I've been party to have all agreed that a container for subresources is not strictly
necessary – the parent Resource already serves as a container, although
of not just the subresources alone (since it also includes the Data branch). But
the subresources are all easily addressable as a group using XPath.

 

I'll try to summarize the arguments as I
understand them both for and against a container element for Subresources.

 

FOR

The argument for (which Bill Barnhill will
be posting a detailed version of) is that it provides an efficient way in XML
to specify the set of subresources that describe the association of a Resource
to other Resources.

 

Another argument for, which both Bill and
Loren have made, is that the subresource container element is actually a better
way to model the "XDI tuple" since it aggregates all subresources
into the "predicate" branch the same way the Data element aggregates
all of its contained resources (if any) into one branch. So the tuple really
becomes "Resource, Association, Data" with XRIs serving to identify a
Resource in all 3 legs.

 

AGAINST

The argument against is that the XML serialization
of the XDI metamodel itself already provides a container (the parent Resource)
and that adding another container just to turn the subresources into their own
branch only adds XML processing overhead.

 

Another argument against is that the XDI
tuple is already represented by Resource, XRI, Data, and that adding another
element only adds processing overhead that will add up in very large XDI data
stores (imagine XML databases containing billions of XDI documents.)

 

*****

Whew! This is going to be a tough one.

 

=Drummond 

 

From: Will St Germain
IV [mailto:] 

Sent: Wednesday, August 18, 2004 3:34 PM

To: Drummond Reed

Cc: 'Loren West';


Subject: RE: [xdi] Resource and
Data

 

Drummon, 

The
more we learn UML as a language to communicate, it seems reasonable to align to
the same nouns when we talk about moving data from one persistence to another?  (note,
as an observer I may be totally out of touch with the topic:) 

The
nouns that come to mind are: 

Instance 

Class 

Attribute 

DataType 

Association 

Reference:
 http://www.omg.org/technology/documents/formal/uml.htm 

Thanks,

Will St. Germain IV



Senior Transition and Information Architect,

303-924-1547

 

  
  
"Drummond Reed"
  <> 

  
08/18/2004
  03:59 PM 

  
  
  
   

    
    
To

    
    
    
"'Loren West'"
    <>, <>
    

    
   

   

    
    
cc

    
    
    
 

    
   

   

    
    
Subject

    
    
    
RE: [xdi] Resource and Data

    
   

  
  
 

  
   

    
    
 

    
    
    
 

    
   

  
  

  
 

Both
Loren and Bill have brought up the potential need for a container

element for subresources (Resources contained by a
parent Resource in the v5

metaschema). Both of them have good arguments for
it, and I've shared with

them at least one good argument against it.

The net net is that it's a wonderful, meaty topic
that we'll be diving into

on today's call.

=Drummond 

-----Original Message-----

From: Loren West [mailto:] 

Sent: Wednesday, August 18, 2004 2:46 PM

To: 

Subject: RE: [xdi] Resource and Data

Funny.  I read Gabe's request to mean the
difference between <data>

and it's sibling <resource> elements.
 Drummond read the request

as the difference between <data> and it's
parent <resource>.

Now I'm more convinced than ever that we need a
different name

for the <resource> elements that are
siblings to <data> (or

at least group them under a <meta> tag).
 They often times

get confused with the parent <resource>
element.

=Loren

-----Original Message-----

From: Drummond Reed
[mailto:] 

Sent: Wednesday, August 18, 2004 2:34 PM

To: 'Wachob, Gabe'; 

Subject: [xdi] Resource and Data

Gabe, it's a fine question. We've been working
with the Resource/XRI/Data

element names for long enough now that no one has
bothered to ask this basic

question.

I will give a very precise definition according to
RFC2396bis and XRI 1.0.

An XDI "resource" is an entity that can
be represented in an XDI document

and identified by an XRI. Period. It's abstract in
the extreme.

XDI "data", by contrast, is any
information that can be represented in an

XDI document that IS NOT identifiable by itself,
but must be placed in the

context of one or more XRIs in order to become
identifiable.

Example:

<data>43</data>

The value "43" in this XML element is
XPath-addressable (in the scope of a

specific XML document) but not XRI-addressable.

<resource><xri>=Mary.Jones/(+age)</xri><data>43</data></resource>

Now this data has become identifiable - in both a
local and global context -

with an XRI and thus now meets the RFC2396bis and
XRI 1.0 definition of a

resource.

Of course, data does not have to be just a literal
value. Complex data can

be a collection of resources. However, a
collection of resources (each of

which itself is identifiable) needs to be put in
the context of one or more

XRIs to be identifiable as a *collection*, so it
still meets the definition

of data.

The next step beyond this simplest definition is
to compare the XDI tuple

with the RDF tuple, and talk about Resource as
Subject, XRI as Predicate,

and Data as Object. That was addressed in the
powerpoint uploaded for the

last TC call

(http://www.oasis-open.org/committees/download.php/8454/How-XDI-Builds-On-RD

F-v1.ppt). 

With your knowledge of RDF, I would love your
feedback on that view of

Resource and Data.

=Drummond 

-----Original Message-----

From: Wachob, Gabe [mailto:] 

Sent: Wednesday, August 18, 2004 11:27 AM

To: ; 

Subject: RE: [xdi] Groups -
draft-xdi-metaschema-v5.xsd uploaded

Drummond-

             
  Can you explain in a sentence or two what the meaning of resource

vs. data is? This doesn't seem to be documented
anywhere - you may have

discussed this on the phone calls, but I don't see
explanation in the

minutes (not that I'd expect such content to be in
minutes). 

             
  Thanks

             
  -Gabe

__________________________________________________



Chief Systems Architect

Technology Strategies and Standards

Visa International 

Phone: +1.650.432.3696   Fax: +1.650.554.6817

> -----Original Message-----

> From: 
[mailto:]

> Sent: Tuesday, August 17, 2004 11:39 PM

> To: 

> Subject: [xdi] Groups -
draft-xdi-metaschema-v5.xsd uploaded

> 

> 

> The document draft-xdi-metaschema-v5.xsd has
been submitted 

> by Drummond Reed ()
to the OASIS 

> XRI Data Interchange (XDI) TC document
repository.

> 

> Document Description:

> v5 Proposed XDI Metaschema (only change is
Comment element eliminated)

> 

> Download Document:  

>
http://www.oasis-open.org/apps/org/workgroup/xdi/download.php/

> 8761/draft-xdi-metaschema-v5.xsd

> 

> View Document Details:

>
http://www.oasis-open.org/apps/org/workgroup/xdi/document.php?

> document_id=8761

> 

> 

> PLEASE NOTE:  If the above links do not
work for you, your 

> email application

> may be breaking the link into two pieces.
 You may be able to 

> copy and paste

> the entire link address into the address
field of your web browser.

> 

> 

>