Re: [xml-dev] UBL 2, W3C Schema extensibility, and multiple

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

At 2006-05-27 16:29 +0100, Fraser Goffin wrote:
>it is interesting to see how some of the rules for how a data standard
>must be used are enforced explicitly in the technical artefacts
>available to implementers, whereas others form part of specification
>or description of usage rules (NDR). Do you (UBL) always prefer the
>former if it is possible ?

I cannot speak for the committee, but I have my own opinions, largely 
centred around "where the rubber hits the road".

I believe the NDRs should be developed with a clear understanding of 
available real-world mechanisms so as to not inadvertently prescribe 
something that cannot be expressed or proscribe something that could 
be expressed that would meet the business requirements.

>I face some similar challenging questions from colleagues about
>whether it is important to be able to demonstrate compliance to a
>[public] standard through validation checks, or whether it is more
>important to be able to consume messages and do business regardless of
>whether standards are being complied with or not ?

A fine balance ... but one that I personally believe comes down to 
integrity and responsibility.  If the sender of the document 
undertakes the responsibility of describing their business 
requirement accurately, then the receiver should have no problems 
testing the integrity of the transmission.  The corollary is, I 
believe, that the receiver upon finding an issue of validation cannot 
assume anything about the integrity.  It is a line of first 
defence.  There may be a number of reasons why the integrity has been 
compromised, assuming the reason to be innocuous and continuing to 
accept the information seems foolhardy.

All that to say "yes, validation is critically important".

>This question just keeps coming back. Shall we accept messages which
>contain everything WE might need to process the requested business
>(i.e. we only check that it meets OUR requirements which may mean that
>it does not always explicitly meet a [public] standard), and ignore
>everything else (i.e. not even bother this stuff to check whether it
>is compliant to any standard).

Two questions implicit in your last parenthesized comment there, so 
two answers:

(1) - if the stuff you receive is not what you expect (it is invalid 
to the schema) there is an issue of integrity and why trust anything 
if something is wrong (what if the information you are working with 
is also wrong even if it is structure correctly?) ... schema 
validation isn't absolution for the content in the structure, but it 
is one thing that can be tested

(2) - if the stuff you receive includes what you not are interested 
it (but it is valid according to the schema), I think you should 
accept what you are interested in if what you are interested has been 
satisfied ... this is an issue that others looking at UBL are 
considering for their own situations.

Question (2) is going to come more to the fore as people subset 
vocabularies and mix vocabularies.  There is a premise I can make 
when using a subset:  if I have agreed to use the Small Business 
Subset, and the SBS portion of a UBL document meets the constraints 
for the SBS portion, then it is an acceptable SBS message.  There is 
a premise I can make when mixing vocabularies (and using something 
like NVDL for validation):  that the integrity of a 
namespace-qualified portion of an instance can be valid in isolation 
without impacting the integrity of the whole.

I hope this helps, Fraser ... BTW, I'm not convinced all these 
conversations should be cross-posted to both XML and UBL ... though I 
recognize there may be interested parties in both camps, those of us 
who subscribe to both are getting a lot of copies of the messages.

. . . . . . . . . . . Ken

--
Registration open for XSLT/XSL-FO training: Wash.,DC 2006-06-12/16
Also for XSL-FO/XSLT training:    Minneapolis, MN 2006-07-31/08-04
Also for XML/XSLT/XSL-FO/UBL training: Varo,Denmark 06-09-25/10-06
World-wide corporate, govt. & user group UBL, XSL, & XML training.
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/u/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Cancer Awareness Aug'05  http://www.CraneSoftwrights.com/u/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal