RE: SubmissionPackage DTD

["Duane Nickull" <duane@xmlglobal.com>]

Thank you Len:

We (XML Global Research and ebXML Project teams) have identified several key
components that a repository single data element must have associated with
it based on needs from businesses and systems.  As the final depiction for
each item will be in XML syntax, I use the term "element" to describe a
component for each RDE.

For instance, when a single data element is standardised (ready for
consumption by whomever or whatever has access to the repository/registry)
it must have an element which contains a human readable string to describe
in human terms what the data element is.  This will help intuitively solve
the problems of recognitions of existing data elements to avoid potential
duplication.

It is also imperative to have a semantically neutral unique ID for each
item.  This key is the application search key.  It must be unique becuase
each query into a repository must result in one and only one query result.
Otherwise,  the system will not be automated.

It has been suggested that we use a unique ID which is semantically
meaningful to at least one person on the planet.  The logic there is while
no one can define a key which is universally acceptable but at least some
will be able to make use of it.  <IMHO> this is a *really* bad idea.  First
off, the political upheavel of EDI vs. other standards competing for
defining this semantically meaningful component will stall all future work.
Secondly,   the unique key is for a Machine to find only,  not a human.
Therefore,  the machine does not care about semantics on the query side,
only that it can find the item and the item is unique.  Once it has a
pointer to the item, the semantics for the calling app can be derived by
performing a "get" type request on the item and examining what it is
equivalent to.  I human shouldbe able to search on the semantic terms it
wants.

In addition to the unique key,  the data elements must also contain an has
of equivalency keys which must be searchable by the human interface (GUI) to
the repository.  This will allow repository objects to be found easier.  If
I decide to search for a BrandX_XML data element using a human interface
into the repositories API, I know what that is and can then determine if it
is equivalent to any other semantic data definition.

PERFORMANCE ISSUES:

It is largely unfeasable to have a hub and spoke type system for the
functionality needed for a globally architected system.  Therefor,  the
calling applications should be required to Cache the objects once they are
retrieved to alleviate further wasted bandwidth duplicating queries.  The
problem is that with a cache,  there must exist some mechanism to tell the
cache how often to refresh the object by querying the original object.
Thus,  the need for an "expires_on"  or "refresh_in" type of element
associated with each item submission may be appreciated by companies like us
who are almost finished architecting ebXML type applications.

As a driving force behind the ebXML Technical Architecture team,  I am going
to hold the architecture to the highest standards for efficient design.  I
have many issues with the current reg rep spec although I truly believe that
the work is heading in the right direction.

Duane Nickull
*******************************
President, Software Systems Architect
XML GLobal Technologies
http://www.xmlglobal.com



-----Original Message-----
From: owner-regrep@lists.oasis-open.org
[mailto:owner-regrep@lists.oasis-open.org]On Behalf Of Len Gallagher
Sent: Wednesday, May 17, 2000 4:01 PM
To: regrep@lists.oasis-open.org
Subject: RE: SubmissionPackage DTD



Duane,

I guess we're all in the same fix -- looking for data that satisfies a
given DTD.  We'll probably have to generate some.  If so, we'll keep you in
mind.

-- Len



At 06:21 PM 5/17/00 , you wrote:
>Len:
>
>This is Duane Nickull from XML GLobal.  We have built a registry and
>repository like system called goXML.com.  We are looking for some content
>which is similar to the ISO 11179 metadata elements.
>
>Do you have any data prepared which corresponds to the DTD's you posted to
>the OASIS Reg Rep list?  If you do,  are you willing to forward us some of
>this meta data.
>
>Duane Nickull
>
>-----Original Message-----
>From: owner-regrep@lists.oasis-open.org
>[mailto:owner-regrep@lists.oasis-open.org]On Behalf Of Len Gallagher
>Sent: Thursday, April 13, 2000 2:46 PM
>To: regrep@oasis-open.org
>Subject: SubmissionPackage DTD
>
>
>
>RegRep,
>
>I was on a roll today so attempted to define a SubmissionPackage that would
>allow a Submitting Organization the ability to request registration of
>multiple items in the same submission, to request withdrawal or replacement
>of previously registered items, or to register a "package" of items.
>
>A preliminary attempt is attached as a "txt" file.
>
>Sincerely,
>Len Gallagher
>

**************************************************************
Len Gallagher                             LGallagher@nist.gov
NIST                                      Work: 301-975-3251
Bldg 820  Room 562                        Home: 301-424-1928
Gaithersburg, MD 20899-8970 USA           Fax: 301-948-6213
**************************************************************