[ubl-lcsc] [Analysis Team] Re: Updated Comments List

[Tim McGrath <tmcgrath@portcomm.com.au>]

Once again you have done a great job in summarizing the issues.  I have 
changed the distribution of this to the LC list (with a prefix to denote 
its an Analysis discussion).  This way we keep a record of the 
discussion and allow others to follow our progress.  It also saves 
having to keep separate mailing groups.

Whilst I agree with your synopsis of the pragmatic approach, i think we 
should not lose track of the bigger picture.

[normalized/UML models (without context)]  --->  [hierarchical/message 
models (with context)]

In practical terms this is a feedback loop, as we develop our message 
models we discover things that should be in the 'bricks' of our 
normalized model.  However, ultimately it will be the normalized 
model/UML model that is the definitive source of our library.

Also, it is not just the cardinalties that may change between the 
normalized and hierarchical model, we can change the direction of the 
associations, choose to explicitly define some that must occur, etc... 
 What we cannot do is extend the cardinalities (e.g. if we have a 0..1 
we cannot decide in some contexts this may be 0..n.)

This is why we should be quite generic in our normalized model, it is 
the context (e.g. when i want an 'Order' message) that determines the 
way we assembly the bricks.  In a practical sense, as you correctly 
point out, the normalized 'bricks' are actually still in the context of 
our scope statement and are not really trying to capture the whole 
world.  But we do need a common model to use as a framework.

I have added my comments to your Disposition of Comments document.  If 
you plan a call today or tomorrow to finalize this please let me know 
and i will try to participate.

Michael Adcock wrote:

>Hi, all!
>
>I have brought together the various comments, and noted all those on
>which we have made a decision. 
>
>To those that have yet to be considered, I have added Stig's and Monica
>Martin's issues, giving a proposed response.
>
>All of these need clear and immediate resolution by Thursday 5th in
>order for us to meet the message assembly timeline. 
>
>There was some discussion about whether the model we are building is a
>complete model. The Normalised Model spreadsheet is a set of 
>potentially re-usable "bricks" and will ultimately contain all the
>bricks we need to build all the messages in our scope. The priority for
>the class diagramming, as I see it, is to target individual messages
>first as that is our most urgent task. In doing so, I do not see any
>danger of accidentally designing the same thing different ways, and the
>way I plan to do it with Rational Rose would prevent this.
>Significantly, the cardinalities (when I get them back in) may differ
>between messages, so it is important to keep separate message-oriented
>models. These can then be brought together into the overall class
>diagram at relative leisure. However I would caution that we are
>modelling within our scope, we are NOT modelling the world!
>
>I will park the idea of separate page sized part-models for the
>interim. However I think that we will need to consider these when we
>prepare printed documentation. 
>
>regards
>
>Mike Adcock
>Standards & Security Unit
>APACS - Association for Payment Clearing Services
>Mercury House, Triton Court
>14 Finsbury Square
>London EC2A 1LQ
>Tel: +44 (0) 20 7711 6318
>Fax: +44 (0) 20 7711 6299
>e-mail: michael.adcock@apacs.org.uk
>
>
>
>**********************************************************************
>The opinions expressed are those of the individual and not the company.
>  Internet communications are not secure and therefore APACS does not  
>     accept legal responsibility for the contents of this message.
>  If the reader of this message is not the intended recipient, or the
>employee responsible for delivering this communication to the intended
>recipient, you are hereby notified that any disclosure, distribution or
>   copying of this communication is strictly prohibited.  If you have
> received this communication in error, please notify us immediately by
>          telephone to arrange for its return.  Thank you.
>**********************************************************************
>  
>

-- 
regards
tim mcgrath
fremantle  western australia 6160
phone: +618 93352228  fax: +618 93352142

Attachment: Disposition of comment-1-tim.doc
Description: MS-Word document