[ebxml-dev] RE: CC Specification Review

["Probert, Sue" <Sue.Probert@commerceone.com>]

Title: RE: CC Specification Review

I think it is important that we don't become too pedantic re syntax neutral. Instead my suggestion would be that we concentrate on remembering the very good reasons why the 'syntax-neutral' CC and BP approach was adopted by ebXML in the first place.

These were borne from the genuine desire of those who have been through earlier semantic specific standardisation exercises in these areas, to ensure that we don't have to 'do it all again' for future new syntaxes too. This is to minimise 'wheel reinvention' and also to promote future increased interoperability across different syntax implementations.

Hence the vision/strategy was agreed to be 'Let's define our business processes and core component definitions in such a way that a) global adoption is promoted (including multi-lingual aspects, b) any size/type organisation can relate implementations to them and c) semantic ambiguity is removed to the maximum possible'.

regards

Sue

Sue Probert
Senior Director, Document Standards
Commerce One
Tel: +44 1332 342080
email: sue.probert@commerceone.com

-----Original Message-----
From: John Larmouth [mailto:j.larmouth@salford.ac.uk]
Sent: 30 April 2002 13:39
To: Philip Goatly
Cc: CRAWFORD, Mark; ebxml-dev@lists.ebxml.org; eBTWG List
Subject: Re: CC Specification Review

Philip Goatly wrote:

> I don't believe there is anything that is syntax neutral
 
Hm.  I have some sympathy with that.  We have no notation for expressing
pure semantics that is devoid of syntax overtones.  If you use ordinary
English, you can get a lot of syntax neutrality, but you are imprecise,
and cannot be processed by machines.  There is a difficult balance
between formal specification (allowing machine processing of the
specification) and syntax neutrality.

But that does not mean we should not try!

One alternative to syntax neutrality is to specify a sufficiently large
number of disparate syntaxes (with equal weight for each) so that
neutrality is achieved by sheer weight of numbers!  That is why it seems
important to many that representations of core components should be
developed in XSD,Relax,ASN.1,EDIFACT Graphical Syntax, and ABNF - to
mention but a few!

John L

--
   Prof John Larmouth
   Larmouth T&PDS Ltd
   (Training and Protocol Development Services)
   1 Blueberry Road                    
   Bowdon                               j.larmouth@salford.ac.uk
   Cheshire WA14 3LS                    Tel: +44 161 928 1605
   England                              Fax: +44 161 928 8069

----------------------------------------------------------------
To subscribe or unsubscribe from this elist use the subscription
manager: <http://lists.ebtwg.org/ob/adm.pl>