Re: [ubl-dev] Hybrid approach to local vs. global

[Chin Chee-Kai <cheekai@SoftML.Net>]

Thank you, Jon, for extending the input channels to UBL-Dev on this very
important, interesting and at times controversial topic.

I've posted some time back regarding UBL's "global everything" approach
being a hindrance to wider adoption (or something to that effect.  As it has
been sometime back and I couldn't search it in the UBL-dev archive, the
precise wordings I used might differ).  In that thread, I suggested to have
only global types and local elements and argued the basis for that suggestion.

I'm sure most arguments would be familiar with readers on the list.  But some
illustration would be useful.  Most of what I'd like to say is found in
http://www.itsc.org.sg/synthesis/2006/3_XML.pdf

in which the primary ideas are to suggest:
(1) the Use Venetian blind without any global elements (all global types,
     all local elements)
(2) the problem with global elements during validation -
      introduction of additional unexpected "valid" data elements
      because of imports and inclusion of other UBL sub-modules.
      This, therefore, poses a potential security issue there as well.

Perhaps this might invite a lot of other countering debate, but I have not
seen viable alternatives to solve.  I welcome hearing opinions from other
camps.





Best Regards,
Chin Chee-Kai
ITSC

SoftML
Tel: +65-6820-2979
Fax: +65-6820-2979
Email: cheekai@SoftML.Net
http://SoftML.Net/



At 02:45 PM 2007-02-28 -0800, jon.bosak@sun.com wrote:
>Hello ubl-dev,
>
>In collaborating with UN/CEFACT, the OASIS UBL TC is tasked with
>contributing input to the work of UN/CEFACT ATG2 in developing a
>revised set of XML naming and design rules (NDRs) for future
>UN/CEFACT deliverables.  Part of this input consists of comments
>on the "hybrid" solution to the problem of global vs. local
>scoping that was adopted for the current UN/CEFACT NDRs.  The UBL
>TC is working to have its analysis ready by the end of its regular
>meeting 14 March and would be interested in hearing comments from
>the UBL developer community on this question prior to that time.
>The rationale for the "hybrid" approach can be found in the
>minutes of the ATG2 meeting held in Kongsberg June 2005.  Those
>minutes are available from the UBL TC site at
>
> 
>http://www.oasis-open.org/committees/download.php/22178/20050627-30-ATG2-minutes-Kongsberg.zip
>
>The relevant files are 20050627-30ATG2_Minutes.doc (Section 5) and
>20050627-30ATG2_Minutes_Annex3_HybridModelRuleChanges.ppt (all).
>
>We invite interested ubl-dev members to discuss the merits of the
>"hybrid" approach on this list and let us know what you think.
>Since we have a deadline of 14 March to prepare our input to the
>upcoming UN/CEFACT Forum meeting in Dublin, comments received
>after that date will probably not be useful.  Please note that the
>UBL TC knows no more about the rationale for the "hybrid" approach
>than what is provided in the documents referenced above; what you
>see is what we've got.
>
>Jon
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: ubl-dev-unsubscribe@lists.oasis-open.org
>For additional commands, e-mail: ubl-dev-help@lists.oasis-open.org