Hi, >>> - a PRACK. But a PRACK can only be sent in response to a reliable >>> provisional. The assumption here is that the answer has not been sent >>> in a reliable provisional yet. So the PRACK would only be an option >>> if a reliable provisional *without* SDP was sent after sending an >>> answer in an unreliable provisional. This is a very weird case. >> >> Is this even allowed? I would think that once an answer has been sent in >> an unreliable provisional response, the identical answer MUST be >> included in (at least) the first reliable provisional response if one is >> sent. Allowing any sort of reliable response that does not contain an >> identical copy of the answer that was sent in an unreliable response >> seems like a recipe for disaster. > >Its hard to imagine why the UAS would do such an odd thing. But AFAIK it >is not *forbidden* from sending a reliable provisional without the answer. > >I don't see it as a recipe for disaster if the UAC is carefully >constructed. But based on the range of behavior in the wild, I think >this might have a high probability of working unpredictably. I agree. This is a problem I have seen over the years, and it's not only related to SDP: implementations make assumptions about the presence of information elements, even if the presence/non-presence is not required, and even if the implementation doesn't need the information for anything. Now, we can always blame implementors for not reading the specs carefully enough, but the fact that we are working on the offer/answer clarification draft, and the re-INVITE clarification draft, shows that SDP offer/answer for SIP is a mess - and the CapNeg vulcano hasn't even erupted yet :) Regards. Christer _______________________________________________ Sipping mailing list https://www.ietf.org/mailman/listinfo/sipping This list is for NEW development of the application of SIP Use sip-implementors@xxxxxxxxxxxxxxx for questions on current sip Use sip@xxxxxxxx for new developments of core SIP
