Hi, >-----Original Message----- >From: [EMAIL PROTECTED] >[mailto:[EMAIL PROTECTED] On >Behalf Of ext Bernhard Suttner >Sent: Wednesday, June 04, 2008 5:54 PM >To: Pekka Pessi >Cc: [EMAIL PROTECTED] >Subject: Re: [Sofia-sip-devel] Ignoring Duplicate SDP > >Hi, > >thanks for your fast answer. > >> You could hack nua_session.c not to ignore SDP, but compare it with >> previous one, and if they differ, generate new offer (and ignore it) >> and then feed new sdp as an answer to it. > >ok. I have another "hack": > > else if (!session_get_description(sip, &sdp, &len)) > /* No SDP */; > //else if (cr->cr_answer_recv) { > /* Ignore spurious answers after completing O/A */ > //LOG3("ignoring duplicate"); > //sdp = NULL; > //} > else if (cr->cr_offer_sent) { > >(just commented the code). > >Do you think that this can break something? > >Will there by a _complete_ implemenation of such scenarios because I >think there are many use-cases like this? Just think about, that a PBX >has also a media server (same IP but different port) and will play a >cool ringback tone in Session Progress and later it want to connect the >user according to the 200 OK SDP.
And mixing the ringback tone into the stream that's ultimately going to carry voice is not an option? Remember, each network address change brings a potential NAT traversal process, with associated potential for breakage. So far I haven't seen any PBX changing the SDP between early and complete phases. Also, as Pekka has mentioned, it hits some really weird SOA-in-SIP scenarios; I'm not sure it's even allowed by the specs. Best regards, Mikhail ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://sourceforge.net/services/buy/index.php _______________________________________________ Sofia-sip-devel mailing list Sofia-sip-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sofia-sip-devel
