Re: [emergency-msg] EM TC Notifications, Methods and Messaging 09-16-03 Meeting Minutes

[Art Botterell <acb@incident.com>]

At 12:35 PM -0400 9/20/03, R. Allen Wyke wrote:
>If they can't resolve a URI, how do they plan on validating the message
>against a schema?

Just speculating... but I imagine they could validate against a local 
cached schema... or even use a non-validating parser and handle any 
resulting data errors downstream in the application.  We aren't 
always talking about what we normally think of as "computers" here... 
many CAP consumers will be embedded "thin" devices.

Anyway, any consumer on a one-way link will have a problem with 
retrieving updates.  My understanding is that that's precisely why 
the warning-systems industry is interested in a standard... because 
once they start deploying firmware in commercial quantities it won't 
be feasible for them to chase a moving-target schema.

Certainly we're addressing a broader user community here than in some 
other OASIS projects.  Still, supporting use of CAP over one-way 
channels has been one of our explicit requirements since the 
beginning.  What I'm hearing here is feedback from users who care 
about that particular use case and aren't persuaded that we've met 
that requirement.

- Art