Re: [sipcore] draft on rich call data and Call-Info

[Paul Kyzivat <[e-mail]>]

On 7/31/20 1:36 PM, Peterson, Jon wrote:
> 
> Hello,
> 
> Chris Wendt and have a draft that has split off from our work toward delivering rich call data (RCD) in the STIR WG. This is for carrying data elements like a caller name, a jCard, call reason, and potentially related elements that will be rendered to the called party to help them decide if they should pick up the phone. We ultimately decided that, instead of doing a one-off for conveying RCD within the PASSporT object, we should instead focus on how that information is best relayed in baseline SIP, and keep the STIR part of the problem confined to signing over those fields as they appear in SIP requests.
> 
> So this draft is the core SIP part:
> 
> https://urldefense.com/v3/__https://tools.ietf.org/html/draft-wendt-sipcore-callinfo-rcd-03__;!!N14HnBHF!qN-QsYPq0MvWk7reTKb984Mry91INiEDfIoc4rQhKA_mMt9ysIG4-Yz2itA0$
> 
> We'd be interested in the thoughts of the SIPCORE WG about the fundamental approach here, and ultimately if there's interest in adopting this as a WG item here.

This seems to be a good start.

I question the decision to add the "jCard" purpose. The existing "card" 
purpose is not bound to any particular representation. It makes perfect 
sense to use a purpose of "card" with content type of 
application/vcard+json.

Similarly, I think it need do discuss the need for the "call-reason" 
parameter. As the draft notes, the contents of the Subject header could 
be used for this purpose. I would like to some some further in 
investigation of why it is unsuitable before giving up on it.

There is a lot of overlap between jCard properties mentioned in section 
6 and other sip header fields that are capable of carrying similar 
information. At the least I think there should be some discussion of how 
these do/don't interact and how a caller should choose which to use.

	Thanks,
	Paul

	Thanks,
	Paul