[Ecrit] urn:service:sos.country-specific

[Brian Rosen <[e-mail]>]

I am the expert reviewer for a request from 3GPP to add urn:service:sos.country-specific to the IANA registry.

Please reference RFC7163 and reference: 24.229 12.14.0 available at http://www.3gpp.org/ftp/Specs/archive/24_series/24.229/24229-ce0.zip <http://www.3gpp.org/ftp/Specs/archive/24_series/24.229/24229-ce0.zip> (cf subclause 7.11.1)

The text in 23.229, says, in part:
This sub-service type is used before an appropriate sub-service type is registered with IANA for a specific emergency service. 
The key word for me is “before”, meaning the general intention is to ultimately do  regular registration for an emergency service.  This is similar to a general experimental service, or an “X-“ service.  No registration is needed, no guarantee of interoperability.

The IETF has long history with these kinds of “escape” mechanisms, most of it bad.

In this particular case, I am very confused why it is thought necessary to have this kind of escape mechanism.  While new kinds of emergency services do appear from time to call, such as eCall, it’s rare; very rare.  It’s also definitely not country specific.  The registry is currently managed with expert review, and thus the bar to new services is low.  Even an experimental service that ultimately fails to gain traction would not have any negative consequences if the listing persisted in the registry.

When 7163 was being discussed, we never imagined that a generic escape mechanism was needed, because of these factors.  This request is clearly outside what we expected would occur, and it creates a very wide hole for non-interoperable implementations, especially where the same service exists in more than one country.

On the other hand, we don’t have criteria in 7163 that would allow me, as expert reviewer, to just say no.

My suggested course of action is to send a liaison to CT1 and request that they withdraw this request in favor of just using the registry as it is designed to work.  If 3GPP insists that they must have this feature, I’m not sure I have sufficient grounds to reject it, but I seek the advice of the work group on that.

Brian