Re: [codelist] A formal comment received by the CLRTC

["Anthony B. Coates (Miley Watts)" <abcoates@mileywatts.com>]

On Tue, 31 Oct 2006 09:55:09 -0000, Paul Spencer  
<paul.spencer@boynings.co.uk> wrote:

> This comment of David's intrigued me:
>
> "d) Have separate optional sections for logical parts of the content - so
> the processing codes themselves are not mixed up with human readable
> descriptions, documentation and such.  Provide means to make these
> externally referenced to encourage use of registry services instead of
> a single "kitchen sink" XML approach."
>
> Ignoring the first bit as not being relevant to us, what is required to  
> get the best use out of a registry, say the ebXML regrep? Should we be  
> looking at a granularity of access below the complete list? Perhaps we  
> have missed a requirement here, but I do not have enough experience of  
> using registry services to be sure.

I think what David is talking about is being able to replace certain  
blocks of the XML with a reference to an external URL or Web service  
(possibly a parameterised reference) that is queried dynamically to get  
that piece of content.

I don't see a "real life" requirement for this for the code list XML  
itself.  That is to say, I don't think we have enough experience (I don't)  
to say that people actually would want to apply this technique (which is  
typically used for human-readable items like Web pages) to the code list  
XML itself.  I also can't say that we need a mechanism on top of what will  
be possible with derived code lists (version 2.0 requirements) and  
XInclude.

After all, we might find that people use the code list XML only to  
populate databases, not as part of complex Web services.

Cheers, Tony.
-- 
Anthony B. Coates
Senior Partner
Miley Watts LLP
Experts In Data
+44 (79) 0543 9026
Data standards participant: ISO 20022 (ISO 15022 XML), ISO 19312,  
UN/CEFACT TMG, MDDL, FpML, UBL.