To bit a bit harsh/blunt, this could be interpreted as saying you are not
trying to produce a universal standard, just a framework that small
communities (of big companies) can use for their own purposes.
Seriously if UBL is to become universal, and certainly the EU have
indicated that they would like it to, such things have to be defined
to that many implementations can inter-operate without having to
find out first which flavor your trading partner has adopted.
The existence of additional agreements in order to use UBL could
be seen as an admission of failure.
I can quite see that the Codes fall into at least three categories
and that these need to be handled in different manners:-
1) Those derived from International lists, such as countries and
currencies. This also applies to things like units of measure which
are the subject of international standards. Deriving those from their
external source is the only way to go and this seems to be done.
2) Common usage within an industry. For example in the cereal seed
business they can be packed by weight or seed count. Seed counts
are not much use in the milk business, nor are things like thousand
grain weights. Likewise the airfreight and maritime logistics worlds
have well developed codes. Either there need to be a super set of all which
might be unwieldy, or they need to be clearly marked as to where they come
from - which you do well.
3) Structural or protocol codes. These are the ones that are
sadly lacking, and are absolutely vital for interoperability.
There may need to be multiple level codes, so a standardized one
and an explanatory one, but the standardized one is a basic
requirement for a usable general standard.
The Internet Email standards (or come to that the rest of the Internet) would
not have been widely adopted if such items had been left open. And in fact
the problem areas are exactly those that were loosely defined. This has given
rise to such things as at least two confirmation of delivery implementations,
and the lack of definition about the tie-up of the outer SMTP envelope and the
inner message envelope is the reason why most of the spam emails
get through.
(getting off my high horse)
David
On Thursday, 28 September 2017 10:14:05 BST G. Ken Holman wrote:
> Thank you, David, for this input.
>
> I have distilled your comments as follows and provided an initial
> response in an attempt to give you a timely answer. This response
> will be reviewed by the UBL Technical Committee for any additional comment.
>
> (1) - the representation of positive/negative responses
>
> It is intended that the <cac:DocumentResponse> ASBIE capture the
> representation of the response:
>
> http://docs.oasis-open.org/ubl/os-UBL-2.1/mod/summary/reports/UBL-AllDocumen
> ts-2.1.html#Table_ApplicationResponse.DocumentResponse
>
> The structure of the ABIE is outlined here:
>
> http://docs.oasis-open.org/ubl/os-UBL-2.1/mod/summary/reports/UBL-AllDocumen
> ts-2.1.html#Table_DocumentResponse.Details
>
> ... in which one finds the response ASBIE to this ABIE:
>
> http://docs.oasis-open.org/ubl/os-UBL-2.1/mod/summary/reports/UBL-AllDocumen
> ts-2.1.html#Table_Response.Details
>
> ... in which one can place the response code as defined by the users
> of the customization of UBL.
>
> The UN/CEFACT Core Component Type for codes defines optional
> BBIE-level metadata in attributes that can be used to identify the
> code list associated with the coded value.
>
> (2) - the lack of standardization of the use of any coded values in UBL
>
> The absence of any specification of particular code lists in the UBL
> conformance clauses is intentional. The committee has imposed no
> constraints of any kind on the values of codes (other than the
> lexical constraint of being a normalized string, per the UN/CEFACT
> Core Component Type).
>
> Accordingly, the conformance clauses state only conformance to schema
> constraints and additional document constraints not expressed in the schema:
>
> http://docs.oasis-open.org/ubl/os-UBL-2.1/UBL-2.1.html#S-CONFORMANCE
>
> Understanding the important role of code lists, UBL 2.1 includes a
> non-normative annex describing one possible method of a user
> community imposing coded value constraints on top of the
> conformance-related document structure constraints:
>
> http://docs.oasis-open.org/ubl/os-UBL-2.1/UBL-2.1.html#A-UBL-2.1-CODE-LISTS-
> AND-TWO-PHASE-VALIDATION
>
> In support of this candidate method, and to demonstrate its use, the
> committee has created genericode representations of popularly-used
> code lists users may find helpful in their use of UBL. The
> demonstration environment includes an XSLT stylesheet that
> incorporates the provided selection of illustrative code lists.
>
> Early in the development process the UBL committee determined that
> code lists are sufficiently varied and sufficiently fluid so as to be
> applied differently in all contexts of use of UBL worldwide and over
> time. Especially given how code lists can change from their
> publisher, and can change in which codes are used perhaps on a daily
> basis, baking in coded values into the UBL schemas would make the
> schema expressions extremely fragile.
>
> Whereas, by divorcing coded values from the conformance schemas, one
> can claim at all times schema structural conformance to the read-only
> published specification with full backward compatibility to future
> revisions of UBL. Our committee responsibility ends at the
> schemas. And to be consistent, *all* coded values are external to
> the schemas, even for concepts not likely to change (latitude and
> longitude directions for an example of this).
>
> This puts the onus on the user community to define for their
> particular community what the coded value constraints shall be and
> how they are to be validated.
>
> . . . . . . . Ken
>
> At 2017-09-28 12:32 +0100, David Goodenough wrote:
> >but the ApplicationResponse object strikes me as strangely disappointing.
> >
> >
> >
> >Looking at the flow diagrams, there are references to negative and positive
> >
> >responses, this is not explicitly represented in the ApplicationResponse or
> >
> >any of its sub-objects.
> >
> >
> >
> >There also does not seem to be a standard set of StatusReasonCodes or
> >
> >ResponseCodes. Surely there are a bunch of specific responses, many of
> >
> >which are shown in the flow diagrams, which could, or rather should be
> >
> >defined for interoperability. There might be supplimentary information
> >
> >which is implementation specific and might be used by humans to solve
> >
> >a problem, but the basic logic flow states should be defined in the
> >
> >standard.
> >
> >
> >
> >There is a DocumentStatusCode, but that is only used for transport
> >
> >related activities.
> >
> >
> >
> >Given that I am new to this lot, it might be that I have simply not found
> >
> >the code lists, but there is actually no reference to either
> >StatusReasonCode
> >
> >or ResponseCode in the:-
> >
> >
> >
> >Universal Business Language Version 2.2
> >
> >Committee Specification Draft 01 /
> >
> >Public Review Draft 01
> >
> >21 December 2016
> >
> >
> >
> >and no genericode definition for either that I can find, and no codes
> >
> >listed on the flow diagrams.
> >
> >David
> >
> >(Confused of Broadwell)
>
> --
> Contact info, blog, articles, etc. http://www.CraneSoftwrights.com/o/ |
> Check our site for free XML, XSLT, XSL-FO and UBL developer resources |
> Streaming hands-on XSLT/XPath 2 training class @ US$45 (5 hours free) |
