RE: [ubl-ndrsc] Tiny UBL Extension Example

[Matthew Gertner <matthew.gertner@acepoint.cz>]

Bill,

Thanks for the comments. I knew my document must be full of errors since
I was tired when I wrote it. The "xsd:refinement" thing is a perfect
example.

Regarding the context drivers, I have no strong feelings about this. My
inclination would be to leave it more open as I propose in the paper,
but if you or anyone else is adament about one approach or the other and
wants to duke it out, I will be happy to stand on the sidelines and
watch. 

I appreciate the desirability of well-formed examples, but if I do this,
you will start asking for VALID documents as well, and we all know what
a slipperly slope that is. ;-) Seriously, I will fix this and the
"xsd:refinement" thing and post a new version.

BTW: I'm ccing the NDR list since I think your comments might be of
interest to others as well. Maybe someone will read this and put the
document in the NDR document repository on the portal (hint, hint).

Matt

> -----Original Message-----
> From: Burcham, Bill [mailto:Bill_Burcham@stercomm.com]
> Sent: Thursday, February 06, 2003 3:58 PM
> To: 'Matthew Gertner'
> Subject: RE: [ubl-ndrsc] Tiny UBL Extension Example
> 
> Quickie:
> 
> I don't think this tag is valid: "xsd:refinement" -- did you mean
> "xsd:restriction"?
> 
> As you already pointed out, examples are useful.  At a minimum the
paper
> needs an appendix with schemas and valid instances.
> 
> An opinion on the context annotations: I'd rather see the context
> categories
> be part of the vocabulary, e.g.:
> 
> <xsd:annotation>
> 	<xsd:documentation>
> 		<ubl:Context>
> 
>
<ubl:IndustryClassificationContext>Automotive</ubl:IndustryClassificatio
nC
> on
> text>
> 
> <ubl:GeopoliticalContext>France</ubl:GeopoliticalContext>
> 		</ubl:Context>
> 	</xsd:documentation>
> </xsd:annotation>
> 
> Reason I like that is that it represents more of a commitment to the
> categories.  We've been too whishy-washy on those for too long.  On
the
> other hand I can see the wisdom of leaving that list more open longer
> term.
> I think though, that could be addressed by versioning the context
category
> schema in the near term and waiting to see if a thousand categories
bloom
> later.  I think we should err on the side of specificity in this
> situation.
> 
> (also that snippet in your doc isn't well-formed -- missing closing
tags -
> -
> I realize it's a bit tedious but we gotta machine-validate code that
goes
> in
> our docs).