RE: [emergency] GJXDM vs EDXL Distribution isses

["Gary Poindexter" <gpoindexter@wi.rr.com>]

I think Rex is being extremely generous when he says "the GJXDM ... appears
to start from a model and then builds the base or core set of functionality
types and then puts those into a hierarchy that allows for an
Object-Oriented descent of inheritance of properties from those core types."
"Appears" is the keyword. Having worked with this beast from it's inception,
I believe the theory was much better than the implementation and appearance
is deceiving.

I believe he is on target with "...developing a best practice to sort out
elements from the XSDs and identify when to use one over the other for
specific purposes and issue guidelines for using namespace prefixes
correctly..."

The gjxdm tried to be all things to all people and thus is full of
ambiguity. Creating a best practice based on real-world
experience/implementations to eliminate the ambiguity and tighten up
structures will be helpful to this TC as well as others. If the gjxdm shoe
fits, there is no reason not to wear it. Being precise and truly
hierarchical should trump following the gjxdm.

My $.02,
gary

-----Original Message-----
From: Rex Brooks [mailto:rexb@starbourne.com]
Sent: Wednesday, December 29, 2004 1:33 PM
To: Ham, Gary A; emergency@lists.oasis-open.org
Subject: Re: [emergency] GJXDM vs EDXL Distribution isses


Hi Everyone,

On a personal note, the GJXDM practice is more like the way I
normally work, in that it appears to start from a model and then
builds the base or core set of functionality types and then puts
those into a hierarchy that allows for an Object-Oriented descent of
inheritance of properties from those core types. As with all things,
there are pros and cons to hierarchical v. flat or horitzonatlly
organized classes developed out of or into schemas. I think this
needs to have some careful thought, especially from the Justice side
of the picture since all it takes to invalidate a more or less
monolithic system is just a few critical exceptions, which require
the whole edifice to be reevaluated at the least and rebuilt at the
worst.

We are far enough along now with both DON and DOJ schools that it is
imperative that the overall guidelines for governmental practice in
the US needs to be harmonized, no matter what it takes, and it also
needs to be harmonized with the international community as well, or
else our standards work is going to create more problems than it
solves in pretty short order.

Do I have a solution? Nope, not yet. However, one thing that should
be brought up at that meeting Jan. 7 is developing a best practice to
sort out elements from the XSDs and identify when to use one over the
other for specific purposes and issue guidelines for using namespace
prefixes correctly so that none of the existing standards need be
invalidated at this point. Then we need to find a way to get the
standards-writers, like us, to agree on some basic core types,
exactly like what the Core Data Types Focus Group appears to be aimed
at doing--except that we in this TC need to take that out to the
larger international community, i.e. W3C and ISO and IEEE, etc.

I believe the impetus is there, and some liaisons are already in
place, but it aint gonna be easy.

Cheers!
Rex


At 9:34 AM -0500 12/29/04, Ham, Gary A wrote:
>The GJXDM subschema generator is up again today so I ran another
>difference against it
>
>EDXL Distribution has an element "eventType" that is type as an
>enumeration.
>
>GJXDM has an Type "EventType" wich is a data type for multiple kinds of
>events and has the subelements of EventDate, EventDescriptionText,
>EventName, EventStatus, Event Time, and Event Type Text
>
>These are fundamentally different.  GJXDM uses Type in its naming
>conventions almost exclusively for organizing a data type and not for
>categorization.
>
>In general the GJXDM uses "code" for ennumerated value sets of all kinds
>and CategoryCode  or TypeCode for classification such as we have in
>EDXL.  This is true even if the code instances are complete words.
>
>To be GJXDM compliant we would probably have to change the "eventType"
>to something more akin to "EmergencyEventTypeCode"
>
>There is probably more of this to come. Your thoughts..............?
>
>
>R/s
>
>
>Gary A. Ham
>Senior Research Scientist
>Battelle Memorial Institute
>540-288-5611 (office)
>703-869-6241 (cell)
>"You would be surprised what you can accomplish when you do not care who
>gets the credit." - Harry S. Truman
>
>To unsubscribe from this mailing list (and be removed from the
>roster of the OASIS TC), go to
>http://www.oasis-open.org/apps/org/workgroup/emergency/members/leave_workgr
oup.php.


--
Rex Brooks
President, CEO
Starbourne Communications Design
GeoAddress: 1361-A Addison, Berkeley, CA, 94702 USA, Earth
W3Address: http://www.starbourne.com
Email: rexb@starbourne.com
Tel: 510-849-2309

To unsubscribe from this mailing list (and be removed from the roster of the
OASIS TC), go to
http://www.oasis-open.org/apps/org/workgroup/emergency/members/leave_workgro
up.php.