Re: [Fwd: Re: [ubl-dev] UBL Adoption Group?]

["Stephen Green" <stephengreenubl@gmail.com>]

Thanks Roberto

I'll responded to you offlist but here are the salient points
(not wanting to make * all * the details public unnecessarily)

* those are good insights worth keeping and developing I think

* the history meant we had plenty of reason to cover the core
identification and having started to do that there came very high
level requests to start a group

* the way to do it is being well defined in test assertion guideline
efforts (see http://www.w3.org/QA/2006/03/minutes_of_qa_ig_f2f_at_the_w3.html )
and CEFACT already has such an effort starting too for the CEFACT
equivalent which is still in progress

* I've had interest from outside Europe which suggests Europe output
might need tayloring to other countries'/continents' needs

* already Japan gave good input into identifying the 'core'

* SystML and Invinet found good agreement on the 'core' in independant
efforts - SystML's for global use and Invinet's for Spain's use so there
is potential to use outside Spain too

So in summary, the technology, content and politics all seem to have
come together now making it quite feasible to take further steps as
necessary

All the best and thanks for patiently discussing this with me publicly

Steve



On 25/05/07, roberto@javest.com <roberto@javest.com> wrote:
> forgot to include ubl-dev...
>
> ---------------------------- Original Message ----------------------------
> Subject: Re: [ubl-dev] UBL Adoption Group?
> From:    roberto@javest.com
> Date:    Fri, May 25, 2007 10:47 am
> To:      "Stephen Green" <stephen.green@bristol.gov.uk>
> --------------------------------------------------------------------------
>
> Hi Stephen,
>
> I see, you are right UBL is really more than EDI, this is because of XML
> family of standards used (e.g. an XSLT definition is a piece of a
> controller)
>
> The logic to accept or skip an incoming message could be located into many
> places:
>
> 1) Expressed using XML (like schematron or as a result of a pipeline of
> validations).
> 2) hard coded client-side using Javascript
> 3) hard coded server-side using JSP/Servlet/PHP/ASP
>
> Whatever a software house will choose the "right" way or not, I think an
> interoperability test could be as simple as providing a set of UBL
> instances and XSDs that sufficiently cover the main set of customization
> use cases (subsetting, profiling, derivation, extensions) for the purpose
> of a "public" interop test (available as WS or ebXML service).
>
> Then if a software, driver, engine, ... will be able to pass all published
> interop tests, it could be elegible to show an "UBL Conformant" Logo on
> their software as quality symbol.
>
> I put some fantasy here, but I think is not a so bad idea, what you think
> about ?
>
> Regards,
>
> Roberto Cisternino
>
> > Good morning Roberto
> >
> > Interesting point. What, then, about standards or test assertions
> related to the 'controller'? I guess this gets to the point in an
> interesting way: UBL was mostly about standardising the 'model'
> > if you think of things as MVC. But was it really? UBL is by name:
> >
> > 1. universal - it is not just for Europe
> > 2. for business - it has to provide what is needed to conduct business
> (universally)
> > 3. a language - it has to make sense equally not just to the sender
> ('speaker') but to the receiver too
> > (and they are likely to be on different continents - why else would they
> be interested in UBL)
> >
> > For this to all be so it has to include more than just model.
> >
> > Sure, it includes a model and provides a model to a document
> >
> > But the document it the focus (for conducting business B2B) and not the
> application which uses it. So as a document it is more than data and
> structure,
> > it is logic and function too. These aspects have to work and they have
> to work
> > universally.
> >
> > Interesting discussion. To me this what I'm trying in SystML to further
> and follow
> > through but a key part of that (as other early adopting * customers * of
> UBL
> > tended to agree) is identification of a 'core', due to the size and
> complexity which
> > has been for some a real barrier. Once that core is identified, and what
> I'm getting
> > from yourself, Roberto, is that the core is not just model but
> controller too (and
> > Ken would add 'view too' from the perspective of writing human interface
> stylesheets),
> > then there are further steps to take, which is what I get from Tim's
> comments.
> > In SystML I'd like to ensure that the core is identified without losing
> the 'universal'
> > aspect of UBL - not allwoing it to be just a core for a 'european
> business language'
> > which would negate its true distinctive value. So catering for
> S.American, N.American,
> > Middle Eastern, Carribean, South East Asian, ...  global requirements is
> what UBL set
> > out to do and I think the time has come to make moves to fulfill that
> goal; perhaps
> > applying relevant aspects of lessons learned in Europe and Asia, yes.
> >
> > A way other standards-related efforts have well-proven for this purpose
> is through
> > globally scoped profiling efforts - first identify a core then provide
> profiles for more
> > specific things (in this case localisation and vertical customisation).
> The SBS was an
> > initial effort for UBL 1.0. Maybe it will do the core identification for
> 2.0. I'm just thinking
> > about there being an umbrella now for this like was originally strongly
> suggested for
> > the SBS - some said create another TC for it or Adoption TC. That was
> too expensive
> > then. Plus now there is the option of a Member Section which seems most
> relevant.
> > I remember our CEO even asked for us to do it back then. Maybe now is
> the time.
> >
> > Many thanks for weedling this out of me.
> >
> > All the best
> >
> > Steve
> >
> >
> > ------------------------------------------------------------
> > Stephen Green
> >
> > Senior IT Officer
> > Bristol City Council
> > Room G34, Romney House
> > Romney Avenue
> > Bristol  BS99 3HB
> > Tel: 0117 922 3794
> > Fax: 0117 922 4877
> > Email: stephen_green@bristol.gov.uk
> >
> >
> >
> >>>> <roberto@javest.com> 25/05/07 06:50 >>>
> > ...
> >> conflicting concepts of minor aspects of the various subsets. There could,
> >> for instance, be some subsets which disallow non-subset data and others
> which
> >> tolerate it with some qualifications (like the 'must understand' versus
> 'can ignore' principles). These approaches haven't been aligned so
> there are chances for conflicting systems which could probably be
> avoided with a
> >> seeking of a common understanding and formalised approach.
> >
> > Mmm... as a developer I think such tolerance is not the task of UBL in a
> MVC paradigm, as the "Controller" should take care of this.
> > The reason is an XML based system MUST NOT ignore validations expecially
> the 1st structural/lexical step (XSD) otherwise we return back to HTML.
> >
> > A tolerant system should use a filter befor validate (as proposed by KH)
> or just block the incoming instance and let the user to design a
> specific mapping to import it (an EDI practise).
> >
> > ______________________________________________________________________
> Please note the new simpler name for our website:
> > http://www.bristol.gov.uk
> >
> > Our email addresses have also changed - visit
> > http://www.bristol.gov.uk/bigchange for further details.
> >
> > Sign-up for our email bulletin giving news, have-your-say  and event
> information at: http://www.bristol.gov.uk/newsdirect
> >
> >
> >
> >
> > --------------------------------------------------------------------- To
> unsubscribe, e-mail: ubl-dev-unsubscribe@lists.oasis-open.org For
> additional commands, e-mail: ubl-dev-help@lists.oasis-open.org
> >
> >
>
>
> Roberto Cisternino
>
>
>
> Roberto Cisternino
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: ubl-dev-unsubscribe@lists.oasis-open.org
> For additional commands, e-mail: ubl-dev-help@lists.oasis-open.org
>
>


-- 
Stephen Green

Partner
SystML, http://www.systml.co.uk
Tel: +44 (0) 117 9541606

http://www.biblegateway.com/passage/?search=matthew+22:37 .. and voice