Re: [Sip] Terminology (was RE: Fwd:I-DACTION:draft-rosenberg-sip-target-uri-delivery-01.txt

["Christer Holmberg" <christer.holmberg at ericsson.com>]

Hi,

 

I think we should use those definitions as base. That way I think we would be able to progress faster.

 

Because, I think we really need to agree on the definitions FIRST. There is no idea on discussing detailed wording in 4244bis if we don't have a common understanding of what we mean.

 

Regards,

 

Christer

 

---

From: Hans Erik van Elburg [mailto:ietf.hanserik at gmail.com]
Sent: 13. maaliskuuta 2009 12:19
To: Dean Willis
Cc: Christer Holmberg; Mary Barnes; sip at ietf.org; Elwell, John
Subject: Re: [Sip] Terminology (was RE: Fwd:I-DACTION:draft-rosenberg-sip-target-uri-delivery-01.txt

Hi Dean,

Yes this was exactly how we used/defined the terms at the time we where discussing the ua-loose-route vs Target-header solutions.

I still think that this way they are most intuitive and closest to how you would use the terms in natural language.

/Hans Erik van Elburg

On Thu, Mar 12, 2009 at 8:27 PM, Dean Willis <dean.willis at softarmor.com> wrote:

On Mar 12, 2009, at 1:18 PM, Christer Holmberg wrote:

A couple of question:

1. Does the 4244bis definition of "retarget" cover the freephone case, when the R-URI is replaced with the AoR of B - not with the contact of B - no matter whether it's done based on location service, configuration or whatever?

By the way, I believe that the way I use the words, this is a "reroute" and not a "retarget". In the freephone case, I would still expect the destination to have an awareness that it is the target of freephone calls and have credentials for the freephone number such that it could appropriately authenticate its responses if we had a means to do so.

2. Related to the first question, when you say "...and thus chaning the target of the request", what is the defintion of "target"?

My definition is that a "retarget" changes the expected identity of the expected responder, introducing the possibility of an "unanticipated respondent" scenario.

This is, AFAIK, very different from the 4244 terminology, which I've always held to be not particularly useful.

--
Dean