Re: [regrep] Implementing CCTS in Registry - further thoughts

[David RR Webber - XML ebusiness <Gnosis_@compuserve.com>]

Matt,

This is definately in line with my thinking.  However - the last
time we visited this - the assembly being part of the atomic
structure gave us lots of grief - in terms of making it 
difficult for people to understand (we figured it out - but
other people found the schema hard to fathom).

Therefore - this time around I'm seeing we can do the
generic atomic XML structure - with the split of details
between the logical "blue helmet" core component,
and then the related children being the physical 
entities underneath that - for the realization in various
dictionaries - OAGI, UCC/EAN, RosettaNet, EDI, et al.

Then the CAM mechanism handles the collections
and packages - BIEs and those assemblies.

I believe these 3 devices will be clear and succinct
for people to follow and build.

Cheers, DW.
==================================================
Message text written by Matthew MacKenzie
> 
I think the best approach would be to define a generic XML structure to 
house the component, with some sort of indication saying "this core 
component is  atomic", or "this core component is a compount component, 
made up of | derived from |etc ...".

Haven't spent time on CCTS, but we may even be able to layer a namespace 
onto RELAX-NG or W3C schema to serialize a component...

-matt
<