RE: How to handle Validation failures was RE: [Sipping-emergency] proposed charter, new wg on emergency

I did not mean to imply that the group, or IETF agreed on validation.  I was
really only addressing Nate.  Sorry for the confusion.

I can get any number of professionals in the emergency call space to talk to
you about validation.  Nate is a good resource.  It's a fundamental
requirement when you have location.

It's separable in the sense that you could have two entirely separate
mechanisms to determine routing and to validate a location.  However, I
observe that both involve the same sequence of actions: you take a location,
and look it up in a database to determine where to route the call.  If you
don't find the location in the database, then it's not a valid location.
You can do the lookup any time, and use any mechanism to query, but you need
to use a lookup to determine boundaries, and that lookup can, by
implication, validate the address.  That makes the job easy, which is good,
but the real issue is that it's a requirement, a pressing requirement,
that won't wait.

It's not backend if you want the USER to be able to determine if the
location is valid.  I care, and I think it needs to be a first class
requirement.  Whether it's backend or not, it needs to be standardized
because the parts that deliver the address and the parts that have the
database are independent of each other and need to be able to work together.

The expertise is not an issue.  The validation problem is simple.  Any
complications are really in the nature of the representation of the
information and are in any event shared with the routing.

Brian

> -----Original Message-----
> From: Peterson, Jon [mailto:jon.peterson@neustar.biz]
> Sent: Monday, September 27, 2004 7:42 PM
> To: 'Brian Rosen'; 'Marc Linsner'; sipping-emergency@ietf.org
> Subject: RE: How to handle Validation failures was RE: [Sipping-emergency]
> proposed charter, new wg on emergency
> 
> 
> > I changed the subject, as we are off the subject of a new charter,
> > as we agree validation should be in that charter.
> >
> 
> Brian,
> 
> I'm not sure that everyone on this distribution agrees that validation
> should be in the scope of this proposed working group. In fact, until I
> read
> your mail, I had gotten the sense that several voices (including James,
> Marc
> and Nate) were questioning this idea.
> 
> I think validation as such is probably outside of the core expertise of
> the
> IETF, whereas routing is more likely to be a place where we can do some
> solid work. The expertise of the IETF does not lie in subjects like
> whether
> Conrad St is in Butler County or Pine Township, nor in recommending how
> you
> might figure something like that out. Moreover, this seems to be a back-
> end
> application issue, not a protocol issue, as I think a number of people
> have
> already pointed out.
> 
> I think that validation is a separable operation from routing, even if it
> is
> a prerequisite in some operational models. An umbrella specification that
> described everything you needed to do in order to provide emergency
> services
> might include validation, but that doesn't mean that the routing protocol
> should change in any way because validation was or was not performed. But
> I
> don't think the IETF has the breadth to tackle that whole umbrella.
> 
> Jon Peterson
> NeuStar, Inc.
> 

_______________________________________________
Sipping-emergency mailing list
Sipping-emergency@ietf.org
https://www1.ietf.org/mailman/listinfo/sipping-emergency