RE: [emergency] FW: Suggestion: Add incidentKeyword tocontentObject

From
Ellis, David <>
Date
2006-01-30T14:35:23+00:00
ID
Thread
RE: [emergency] FW: Suggestion: Add incidentKeyword tocontentObject
Mark

 

I am working on creating a National level
Alerting system.  The more “non” Distribution elements we put
in the EDXL Distribution Element the more likely a organization might try to
use them for routing and message could be misrouted or rejected.

 

Additionally, I am trying to use set
theory to validate distribution aggregation/dis-aggregation.  Adding
addition element which some organization uses for classifying messages causes
the number of validation matches to increase significantly and creates
significant delays delivering life saving time-sensitive warning messages.

 

David E. Ellis

Information Management Architect

(505) 844-6697

 

From: Mark Carlson -
Conneva, Inc. [mailto:] 

Sent: Saturday, January 28, 2006
3:18 PM

To: Ellis, David;


Subject: Re: [emergency] FW: Suggestion:
Add incidentKeyword to contentObject

 

David,

I am not able to post messages to the group so will just reply to you and
Patti.  

My proposal was to add an incidentKeyword with valueListURN/valueList pair that
would be used to classify the incident, not to use the valueListURN/valueList
pair to identify the incident.  If it makes sense to have incidentID and
incidentDescription, then adding a way to classify the incident also makes
sense and adds value.  I am working on a resourcing application in which
different business logic / routing rules will apply based on incident
type.  Sure, I can design a contentObject payload that contains
incidentID, desription and keyword information, but as the ID and description
are already present in the DE, that is counter-intuitive.  I'm not convinced
incident fields belong in the DE, but if they do, then a way to categorize the
incident seems necessary. 

Regards,

Mark Carlson

On 1/28/06, Ellis,
David <>
wrote: 

 

As you can see, the green IncidentID and incidentDescription
have not functionality in either explicit of Pub/Sub distribution.  In
fact, if we could predict incidents in advance and add them to a
valueListURN/value pair for determining distribution during incidents, the TC
should go to Las Vegas and win enough money to retire