Re: [Geopriv] questions/comments regarding draft-schulzrinne-geopriv-relo

From: Henning Schulzrinne ^lt;hgs@cs.columbia.edu>
Date: Sat Aug 12 2006 - 09:12:49 EDT

> In other words, you believe an HTTP Subscribe header could have broader
> use than RELO?

Yes, although I don't necessarily want to make this a major part of the
discussion. It's clearly purely optional functionality that doesn't harm
innocent bystanders.

>
> I'm not sure about this. We probably just need an opaque string to hold
> any arbitrary key. Either the key is found or it is not. The shape and
> smell of the key are unimportant. Perhaps there could be an optional
> "client-id" attribute. If present, it is used. If not, use the source
> IP address... with the obvious warning that if the contents of the
> client-id attribute is an IP address then this is a possible spoofing
> attempt.
>

The revised version of the draft defines an id attribute that describes
the format of the opaque string. Something like

<get-location id="CDP">cepsr7-1/17</get-location>

I think that's roughly what you're suggesting, if I understood correctly.

Henning

_______________________________________________
Geopriv mailing list
Geopriv@ietf.org
https://www1.ietf.org/mailman/listinfo/geopriv
Received on Sat, 12 Aug 2006 09:12:49 -0400

This archive was generated by hypermail 2.1.8 : Sat Aug 12 2006 - 09:40:04 EDT