RE: Discussion Topics Re-Revisited

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

Terry Allen wrote:

>per ISO 11179, every data element has unique identifier.  By
>administrative necessity, every registered item will have to have
>a unique identifier.


Terry - agreed!  I have been standing on a soap (the non microsoft kind) box
and shouting this out for a while.  For an XML query to work into a
Repository like mechanism,  it must return one and one only results.  If you
allow cowboys to start randomly naming their data elements with non-unique
identifiers, the entire model won't funtion.

There are two issues that have come out of my rantings:

FIRST ISSUE:
1) What those unique identifiers will be?

My thoughts here are that they should be semantically meaningless for two
reasons -

  a) becuase they are being used as keys by applications (within ebXML) for
queries and applications don;t care about the semantic meanign of a word,
they care about syntax only.  Humans who query a repository to find data
element can search on more meaningful points to locate the semantic
equivalent that suits them.

  b) The political upheavel it could potentially cause.  Some have suggested
that a first come first serve basis should apply.  Well,  if I go to a
repository authority first and name VIN number data element
"Fords_Rules!!!", I think that general motors may get upset.  This is pretty
drastic but I have seen [nameless] standards people red in the face shouting
becuase they want their way or the highway.

SECOND ISSUE:
2) Are empty elements in an XML document instance required to reference a
data element unique identifier so they may be retrieved from a repository?

I think yes.  IT is important to know what the node is, even when it is
blank.  This would allow for less chance of error at the receiving and in a
document choreography.

My $0.02 CAD worth

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