Re: WG Review: Call Control UUI for SIP (cuss)

Hi,

let's not start discussing (again) whether this information in carried
in a header field or in a body part. Let's discuss the requirement below
instead, which seems to be what needs to be clarified in the charter
(the details on how to encode this will be largely determined by the
requirements the solution should meet).

>  5. SIP elements may need to apply policy about passing and screening
>   the information.

I agree the charter needs to be clearer on what types of policies are
expected and what is needed to apply them. Discussing about the
granularity of those policies would be useful as well. The charter
should clarify the questions below:

Would a proxy reject a whole request because it carries some type of
information (without the proxy knowing the exact contents of the
information)?... or would the proxy remove the information and proxy the
remainder of the request?... or would the proxy need access to the
contents of the information in order to decide what to do? Also, will
intermediaries applying policies be proxies or some type of a B2BUA?

Thanks,

Gonzalo

On 01/07/2010 5:31 AM, Cullen Jennings wrote:
> 
> On Jun 29, 2010, at 3:25 AM, Elwell, John wrote:
> 
>> Cullen,
>>
>> Whilst neither agreeing nor disagreeing with the charter, I did not find anything in the charter that said the information had to be in the SIP header rather than in the body. On what basis do you make that deduction?
>>
>> John
> 
> When I read 
>  5. SIP elements may need to apply policy about passing and screening
>   the information.
> 
> And the discussion about it's not just UA. I reached that perhaps flawed conclusion that proxies needed to be able to change the information when "screening" and thus it needed to be in a header. Note I would have far less of an issue with an opaque container for proprietary information if it was in a body instead of a header and had the types of constraints that SIP-T has.  
> 
> _______________________________________________
> Ietf mailing list
> Ietf@ietf.org
> https://www.ietf.org/mailman/listinfo/ietf
> 