Re: [ubl-comment] UBL comments on ebXML Core Components TechnicalSpecification v1.8

["Eve L. Maler" <eve.maler@sun.com>]

Ron,

You're right that codes get mapped to meanings, and that identifiers 
convey uniqueness.  But I think the proposal as stated is defensible. 
For example, it makes the point that sometimes codes can "play the role" 
of an identifier (that is, sometimes you have a piece of information 
that both can be mapped to a meaning and, when used on a particular 
object, indicates uniqueness of that object).  So making them be a 
mutually exclusive choice is unhelpful.

The suggestion in the proposal is to allow identifiers to be 
*represented* in a variety of ways (at the RT level, e.g. as Codes or 
Names), while allowing the identifier-ness to be captured slightly 
higher up (at the property level).  So nothing is being lost.

	Eve

Schuldt, Ron L wrote:
> UBL Team,
> 
> I concur with most of the comments contained in the document prepared by the
> UBL team. However, the subject of CODE versus IDENTIFIER is not being
> portrayed properly.
> 
> In the example for Country Code, AU represents Australia and the processor
> would need to refer to a look-up table to convert from "AU" to its meaning -
> namely "Australia"
> 
> In comparison, an "Employee Identifier" could be something like 
> 123-45-6789 and the processor does not need to refer to a look-up table but
> simply takes the value captured between the two tags e.g., 
> <EmployeeIdentifier>123-45-6789</EmployeeIdentifier>. Except for validating
> that the string has the right characteristics, the processor does not
> necessarily need to refer to a look-up table. Typically an identifier is a
> "key" that is used to join two or more tables. Idntifiers are necessary keys
> for topics such as Part Identifier, Person Identifier (since names cannot be
> considered unique), Enterprise Identifier (typically assigned by a
> registration authority such as DUNS), Engineering Drawing Document
> Identifier, etc. In general, identifiers are used when when the population
> of the set is continually growing and some activity or system is continually
> adding new identifiers.
> 
> To simplify the difference, a CODE requires a processor to refer to a
> look-up table to convert to the actual instance whereas an IDENTIFIER does
> not require the processor to refer to a look-up table but rather captures
> the instance value contained between the start tag and the end tag.
> 
> Therefore, I strongly recommend that Proposal 9 be deleted or at least
> revised to instead request further clarification of the differences. If you
> concur with my examples above, perhaps they could frame a proposed
> clarification.
> 
> Ronald L. Schuldt
> Senior Staff Systems Architect
> Lockheed Martin Enterprise Information Systems
> 11757 W. Ken Caryl Ave. #F521 MP DC5694
> Littleton, CO 80127
> 303-977-1414
> ron.l.schuldt@lmco.com
-- 
Eve Maler                                    +1 781 442 3190
Sun Microsystems XML Technology Center   eve.maler @ sun.com