Fwd: Re: [codelist] Interesting article about extending enumeration lists in XML Schema

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

(retry due to SMTP problems)

Date: Wed, 24 Sep 2008 21:23:49 -0700
To: codelist@lists.oasis-open.org
From: "G. Ken Holman" <gkholman@CraneSoftwrights.com>
Subject: Re: [codelist] Interesting article about extending 
enumeration lists in XML Schema

At 2008-09-25 04:05 +0000, kumar.sydney@gmail.com wrote:
>Came across this interesting article.

Yes, I also saw reference to this and it was in today's Robin Cover summary.

>Rules out genericode based approach in the study.
>
>http://www.ibm.com/developerworks/library/x-extenum/

Sure he does ... but from what I read his proposal doesn't address 
everything that genericode does for deployments such as UBL because 
of his focus.

The only genericode lists that are "extended" in UBL are those 
without definition ... that is to say they are a restriction of an 
infinite set of all possible values.

I see using genericode with context/value association allowing 
different users to have different restrictions on lists without 
having different schemas.

Furthermore genericode with context/value association allows 
different restrictions on the same enumeration used in different 
contexts of the one document.  I don't believe this can be addressed 
in W3C Schema when using global types.

While, yes, you could put value-level meta data and list-level meta 
data in appinfo constructs in a W3C Schema enumeration, genericode 
allows one to selectively use the available keys without having to 
change the file in any way.

So I don't think his "ruling out" of genericode really affects people 
considering solutions that need the flexibility not offered by W3C Schema.

In fact, I think that since the article is *about* extending lists in 
W3C Schema (based on the title) it makes sense that he cannot talk 
about genericode.  His observation about genericode that reads 
"Adding new technology requirements can be too costly or 
time-consuming in some settings." is, I think too subjective ... if a 
user's problem is solved by genericode and not by W3C Schema then 
either they change their requirements or they accept that they need 
as much software as is needed to solve their problem.  There will be 
many implementations of genericode and context/value association ... 
in many contexts such as document data entry or document validation.

But did you notice the last bullet of the conclusion?

   "If you want to keep enumerated values out of the parser, consider 
the Genericode approach"

Sounds to me like he is affirming genericode's existence and role 
when users are prepared to do without W3C schema enumerations to get 
the advantages of genericode and context/value association.

. . . . . . . . . . Ken


--
Upcoming XSLT/XSL-FO hands-on courses:      Wellington, NZ 2009-01
Training tools: Comprehensive interactive XSLT/XPath 1.0/2.0 video
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/o/
Male Cancer Awareness Nov'07  http://www.CraneSoftwrights.com/o/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal