Re: [clr-dev] Use of genericode

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

At 2008-01-10 14:56 +0100, Linda van den Brink wrote:
>---------- Forwarded message ----------
>From: Linda van den Brink <linda@dynasol.nl>
>Date: Jan 10, 2008 9:14 AM
>Subject: Use of genericode
>To: clr-dev@lists.oasis-open.org
>
>
>Hi all,
>
>For a Dutch customer I'm working on several information models. These
>information models are developed in cooperation with a Dutch
>organization that manages geo-standards in the public sector. The
>customer, and the geo-standards organization, are looking for a good
>way to manage codelists. I've recommended that they look at
>genericode, and in turn they are now asking me to look at this.
>
>First of all, is Genericode stable enough to be considered for production use?

Genericode 1.0 was confirmed as a "Committee Specification" in a vote 
in December.  Due to the holiday period, the posting of the finalized 
document hasn't happened and should be done by next week.  The only 
change between Committee Specification and Committee Draft 3 is the 
title.  Everything else is the same.

>And what about the context/value association spec?

It is an early work in progress.

>I noticed that it
>was heavily trimmed down compared to the previous version, which was a
>description of not only the CVA format but also a methodology for
>validation that relied on Schematron. I am curious about the reasons
>for removing this dependency on Schematron.

The committee felt that the responsibility is solely to specify file 
formats and the constraints of applications to use those file 
formats.  It was deemed out of scope to specify a validation task 
using a CVA file ... only the file itself.  One comment expressed at 
the time was "what if one wishes to use CVA files for a validation 
task that doesn't involve Schematron, as in a database?".  Since CVA 
files have a wider scope than just validation (e.g. it could be used 
to define data input constraints), and validation is wider than just 
Schematron, the committee should not specify a Schematron-based 
validation for fear of users thinking they couldn't do other kinds of 
tasks or other kinds of validation.

It was not a unanimous decision, but nevertheless, the committee 
decided to stick to only the declarative nature of file formats and 
the information they represent.

>Or is the validation methodology going to be in a separate spec?

There are no plans to make it a separate specification.  It will be 
up to industry to utilize CVA files as it wishes.

>I've been experimenting a bit with the Schematron implementation of
>the CVA validation available at Cranesoftwrights. Is it an
>implementation of the latest version of the spec?

Yes, it is an implementation of CVA 0.2D1, which is the latest 
published review version and the subject of ongoing discussions:

   http://www.oasis-open.org/committees/document.php?document_id=26175

Some important feedback has been posted on the committee list, so the 
specification is undergoing a number of changes.

I'm testing the specification changes by maintaining an 
implementation of Schematron-based validation in conjunction with 
editing the changes.  At the time each draft is released, I will try 
to have an implementation of that draft immediately available on 
Crane's web site.

>Finally, a different sort of question. The geo-standards group I'm
>involved in has been using UML to express all their information
>models. They generate their schemata from UML. I expect them to ask me
>if Genericode code lists could also be expressed in UML, but I don't
>see an obvious way of doing this. Any thoughts on this?

I don't know UML, so perhaps someone else on the committee or in the 
community should comment.  But, if you think of genericode as 
representing a keyed table, then if UML can represent a keyed table 
the answer to "can the content found in genericode code lists also be 
expressed in UML?" would be "yes".  Since genericode specifies a file 
format, one can't express genericode (per se) in UML, but one should 
be able to represent the content found in genericode in UML.

Note that I said it is a keyed table.  There may be many columns in 
the table, but each row in the table must be uniquely distinguished 
by one or more column values in that row.  Additionally, one may have 
different combinations of one or more column values in each row where 
each combination uniquely identifies each row.  Otherwise, it is a 
very generic table (no pun intended).

I hope this helps.

. . . . . . . . . . . Ken

--
Comprehensive in-depth XSLT2/XSL-FO1.1 classes: Austin TX,Jan-2008
World-wide corporate, govt. & user group XML, XSL and UBL training
RSS feeds:     publicly-available developer resources and training
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/c/
Box 266, Kars, Ontario CANADA K0A-2E0    +1(613)489-0999 (F:-0995)
Male Cancer Awareness Nov'07  http://www.CraneSoftwrights.com/c/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal