RE: Just what is Interoperability?

[John McClure <hypergrove@olympus.net>]

In the DCN, semantic equivalence may be indicated directly by an XML element
in an instance document, and it may be indicated in the RDF Schema-based
dictionary pointed at by XML elements in the instance document. If in the
dictionary, and not in the instance document, then the dictionary's value
can be assumed as the default by choice of one's software or of a software's
user.

In the DCN, values for datatypes such as date are wrapped by an XML element
representing a first-class resource, called <Entry>. Below, an <identifier>
element is used to provide the raw semantic mapping I think you want. The
'range' attribute on the <instant> element is used to indicate a range of
values applicable to the instant, in this case you can infer that the date
applicable to the ship-date is up-to and including 05/01/01.

<Entry xsi:type='Time' name='RequestedShipDate'>
   <identifier identifier='unedifact:de2005'/>
   <instant range='/2001-05-01'/>
</Entry>

A second, perhaps more explicit and less error-prone, representation is
equally possible, one that states the role(s) being fulfilled by the
time-entry. Note that a 'range' attribute is unnecessary on the <instant>
element, just use the attribute for a scalar value, ie. 'value'.

<Entry xsi:type='Time' name='RequestedShipDate'>
   <identifier identifier='unedifact:de2005'/>
   <instant value='2001-05-01'/>
   <list xsi:type='roles' name='entryRoles'>
     <Role xsi:type='EntryRole' name='Deadline'/>
   </list>
</Entry>

To indicate a default mapping within the DCN Dictionary is as trivial as it
is in the instance document. Below is the actual code for this entry in the
dictionary (v110). I have now inserted the same <identifier> element used
above, albeit qualified by the dcn namespace prefix.

<rdfs:Class rdf:ID="RequestedShipDate">
	<rdfs:subClassOf rdf:resource="#Time"/>
	<rdfs:subClassOf rdf:resource="#Entry"/>
	<rdfs:label xml:lang="en">requested ship date</rdfs:label>
	<rdfs:comment xml:lang="en">
	Date the buyer requested the product be shipped.
	</rdfs:comment>
	<dc:Date>2001-01-25</dc:Date>
	<dc:Rights>Source: PIP Business Dictionary</dc:Rights>
	<dcn:identifier identifier='unedifact:de2005'/>
</rdfs:Class>

> -----Original Message-----
> From: William J. Kammerer [mailto:wkammerer@foresightcorp.com]
> Sent: Friday, April 06, 2001 10:13 AM
> To: ebXML Core
> Subject: Just what is Interoperability?
>
>
> Since in some circles it appears a fait accompli xCBL will be chosen as
> the basis of ebXML's core components, I thought it might be nice to
> check it out at http://www.xcbl.org/.
>
> I talked earlier today about the EDIFACT ORDERS message and how the DTM
> segment is used in it. xCBL contains an analogous Order document, which
> in turn contains an OrderHeader. In that header are the
> RequestedShipByDate, RequestedDeliverByDate, PromiseDate, ValidityDates,
> and CancelByDate elements - all with descriptive names - augmented by
> the ListOfDateCoded element which can contain any number of DateCoded
> elements which are other, non-specified, dates qualified by a code from
> a code list dwarfing the already complex EDIFACT D.E. 2005.  I reckon
> the explicitly named date elements are derived from commonly used date
> qualifiers in EDIFACT D.E. 2005 (or the analog in ASC X12 D.E. 374
> Date/Time Qualifier).
>
> Arofan Gregory had earlier said to me: "I don't understand what your
> interest is in ebXML. If you want to use EDIFACT - by all means, use
> EDIFACT. But when you take away the ability to determine semantic
> equivalence across syntaxes, you abandon the goal of interoperability
> that was ebXML's charter at the core component level."
>
> Can Arofan or someone please describe how "to determine semantic
> equivalence across [the xCBL and UN/EDIFACT] syntaxes?"  I didn't see
> any annotations in the DTDs or schemas which could be used to equate,
> say, RequestedShipByDate with the date in a DTM segment qualified by a
> code like "10" (Shipment date/time, requested).  Are there any formal
> models for these classes with the relevant tie-backs?  If this
> information wasn't recorded, then however can we determine semantic
> equivalence?
>
> If such information is lost or was never maintained, then I'm not too
> concerned.  I have no use for auto-converting xCBL schemas or documents
> into EDIFACT MIGs or messages, respectively.
>
> I understand "interoperability" to be the standardized exchange of
> business information between two organizations, regardless how disparate
> their internal processes.  Interoperability has nothing to do with the
> sterile exercise of converting EDIFACT to xCBL or vice versa.
>
> William J. Kammerer
> FORESIGHT Corp.
> 4950 Blazer Pkwy.
> Dublin, OH USA 43017-3305
> +1 614 791-1600
>
> Visit FORESIGHT Corp. at http://www.foresightcorp.com/
> "accelerating time-to-trade"
>
>
>
> ------------------------------------------------------------------
> To unsubscribe from this elist send a message with the single word
> "unsubscribe" in the body to: ebxml-core-request@lists.ebxml.org
>