OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

ubl-lcsc message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]

Subject: Re: [ubl-lcsc] Latest update from Mike for Tuesday's call(s)

In case i miss the analysis team call, i want to make a few comment on 
Mike's remaining issues...

> Issues
> 1) I disagree strongly with the notion to call 'lines' as 'items' as I 
> feel this leads to potentially great confusion with the actual item 
> (i.e. article) information itself. Yes, the use of 'line' has a 
> document connotation but we are actually designing replacements for 
> 'documents', however much we think about data, process modeling!
> Please note that I took the opportunity of EAN being here to ask their 
> view. Like me they have similar concerns about "Xxxx Line" being 
> document-centric, but feel that "Xxxx Item" has even worse and 
> dangerous connotations!

I accept the comment about  'xxx line' being preferable to 'xxx item'. 
 However, i then come back to the point about Order Item - to be 
consistent this must be called Order Line.

> 2) I think the many-to-many situations are likely to either be 
> resolved when we assemble the components into specific 
> messages/documents. I also intend to do a quick review of the 
> relationship cardinality, as it is easy to forget that one is 
> modelling within a bounded sphere of our scope and NOT modelling the 
> entire world!

I too have been pondering the question of many-to-many associations and 
am coming to agree with this view.  relational theory on which our 
normalisation models are based does recognise these type of 
associations.  it is when implementers try to build database system (and 
other structures) that they must resolve these - becasue the technology 
does not support a way of navigating them.  As mike says, we resolve 
them when we assemble our documents.  we decide which is the 'one' and 
which is the 'many' by the direction we assembly them.
So i think we have a theoretical and practical argument for allowing 
many-to-many associations in the normalised model.
My suggestion is that we proceed with this approach and review as part 
of the next revision.

Probert, Sue wrote:

>Just before we left Washington Mike asked me to post to you all the attached
>update but unfortunately I could not get online again until arriving home.
> <<Model&Messages_2002-12-13.zip>> 
>Sue Probert
>Senior Director, Document Standards
>Commerce One
>Mobile: +44 7798 846652
>Tel: +44 1425 275117 or +44 1753 483000
>email: sue.probert@commerceone.com

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

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]

Powered by eList eXpress LLC