Draft:  draft-ietf-sipping-service-identification-01.txt
Reviewer: Spencer Dawkins [spencer@mcsr-labs.org]
Review Date: 3/31/2008
Review Deadline: 4/11/2008
Status: RAI review 

Summary: Ready with comments.

Comments:
----------

Based on comments on the -00, the following comments remain:

- one asking for a reference to "Push-to-Talk".

Jonathan suggested just telling people to Google "Push-to-Talk", which could be the right answer but seems strange, given the existence of informative references sections in RFCs. My counterproposal was to ask the OMA Liaison to suggest a reference, since I don't know what a stable reference would be, either. Obviously, this is a nit.

- one raising a concern that was similar to a concern Jonathan raised in the draft, but I thought it was different and worth mentioning - the draft talks at some length about the effect that using opaque service identifiers has on innovation BETWEEN carriers, but ISTM there's a similar effect on innovation even WITHIN a single carrier.

If you define two service using Service Identifiers that are only used within your network, it's not at all clear how to combine them for a customer who wants to use both on the same call, and it's not clear how to extend a single service, if signaling isn't sufficient to describe services. 
I don't think this is worth more than two sentences, but if the message of this draft is "using opaque service identifers breaks things for everyone", it seemed appropriate to say "using opaque service identifers even breaks things for YOU, within YOUR network".