comments Re: [ubl] First Draft of Chapter 1 Customization document

[stephen.green@systml.co.uk]

comments

*Excellent* and very welcome overall.  I've a few comments below:

3.1
"...A major advantage of UBL conformance is that the resulting  
instances can be processed by any off-the-shelf UBL application. This  
negates the need for bespoke software or modifications to UBL  
applications."

Not all off-the-shelf UBL applications can process all conformant instances.
Most off-the-shelf UBL applications will only be able to handle subsets and
if the application subset isn't a subset of the conformant customized subset
there may be a problem, I think. However, there is a likelihood that such
mismatches might be handled more gracefully and more affordably than with
compliant customizations, I think.

3.2
"...Schema-valid instances of a compatible customization are not  
necessarily schema-valid instances of UBL; however, the reverse may be  
true."

Need to state what is meant by 'the reverse'. There are various logical
answers to what 'the reverse' might mean, plus it might not be easy for
the reader to deduce.

3.2.2.1
"...or indeed a completely new data may be required."
I think that should be "...a completely new datatype ..."

3.2.4
"...
* Customizing a Data Type creates a new BBIE
* Customizing a BBIE creates a new ABIE
* A customized ABIE means creating a new ABIE
* Customizing an ASBIE creates a new ABIE
* Any new ABIE means a new document model "

is it also true that
a customized ABIE means creating new ASBIE(s)
?

General comment on 'compliance' (re semantics):
need to cover changes of meaning of any BIE - I suggest semantics have to be
preserved (to be compliant with UBL's design rule for its own development)
such that any definition change which amounts to a semantic change requires
a new BIE too. The ripple effect then applies to this kind of change too.
Also it warrants a new qualifier as well, in short it amounts to the same
kinds of changes as a change to the structure of a BIE or datatype. The same
applies to a change to the semantics of a datatype.


Best regards and much appreciation
-- 
Stephen D. Green

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

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



Quoting mavis.cournane@cognitran.com:

> please find a draft copy of Chapter 1 of the Customization document
> for discussion this afternoon in the plenary.
> Note that Sections 1-3 only are relevant for this, other sections will
> form subsequent chapters and are place holders only.
>
> Regards
> Mavis and Mike