On Thu, Sep 11, 2008 at 8:25 PM, Paul Kyzivat <[EMAIL PROTECTED]> wrote:

>
>
> Jagan Mohan wrote:
>
>> Hi,
>>
>>   If the codec change requested in a Re-INVITE needs to be rejected by
>> UAS,
>> is it mandatory to send a 488 response to the Re-INVITE and continue
>> operating with the codec negotiated in the INVITE?
>>
>
> Would be helpful if you showed the specifics of the SDP in the initial and
> subsequent o/a.


Say, g.711-u law is the only codec negotiated in the initial INVITE
transaction and g.729 is the only codec received in the re-INVITE.
I have only one media line in all offer/answers.

Yes, RFC 3264 clearly tells that atleast one codec received in the offer,
should be sent in the answer if the offer needs to be accepted.

I just wanted to confirm that sending a 2xx response with g.711-u law codec
in the SDP is wrong when only g.729 codec is received in the re-INVITE
offer.

Thanks for the detailed info.

Thanks,
Jagan



>
> If the new offer has multiple codecs, and you object to one of them, then
> of course you may answer omitting that one. If there is only one codec
> offered in the reinvite, and you don't like it, then you could:
>
> - reject the reinvite with a 488, continuing with your old codec
> - successfully answer, accepting the codec, but in an inoperable way,
>  such as by specifying a=inactive or c=0.0.0.0
> - answer successfully, but reject the media line, by setting the port
>  to zero
>
> Of those, I would think that the 488 would normally be the most useful
> alternative.
>
>    From section 14 of RFC 3261, looks like it's not mandatory to send a 488
>> response and more a implementation specific problem. In that case, is it
>> ok
>> to send a 2xx response to the Re-INVITE with the codec negotiated in the
>> initial INVITE?
>>
>
> If your answer accepts the stream, then it must mention at least one of the
> codecs from the offer.
>
> You should look at RFC 3264 if you haven't already.
>
>        Thanks,
>        Paul
>
>     Or is this case addressed in any other standard?
>>
>> Thanks,
>> Jagan
>> _______________________________________________
>> Sip-implementors mailing list
>> [email protected]
>> https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
>>
>>
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to