Well, we haven't exactly codified who is responsible for applying patches. Usually it just goes to whoever knows the code fairly well, and is around. However, IETF is happening next week, so a lot of people are in the air right now (both figuratively and literally). DUM is something that I have just started wading into, and I am uneasy about applying patches without feedback from those who wrote most of that code.
Scott, have you looked at this?
As for when the next release is, the answer is soon (I had originally intended to designate 1.1-RC2 as the official release this evening, but since a couple of bugs have been discovered in the last few days, I'll have to wait for the fixes and cut RC3, probably sometime early next week.)
Best regards, Byron Campen
Byron I do not think that a call to handler->onFailure() is necessary, the aim is "in fine" to do the unPublish. Just another question, who is responsible of merging this patch into the reSIProcate project ? Any idea for the next release date ? Best Regards Fabrice ROUILLIER Good find. Now, would it be necessary to call handler->onFailure() in this case? Is getting a 412 considered a "failure" for an unPUBLISH? (As far as intent goes, it seems not to me)
Best regards, Byron Campen Dear reSIProcate team, I find a bug in the implementation of the "ClientPublication" class when handling response to a 412 message received from server. You previously remove the "SIP-if-match" tag and republish the document. This SHALL NOT be done if the 412 response is received when trying to end the publication (Expires header set to 0) In that case nothing more have to be done ! In fact publishing again with expires set to 0 and without a Sip-if-match will raise a 412 again and again !!!
void ClientPublication::dispatch(const SipMessage& msg) { ... if (code == 412) { // Receive a 412 while ending a publication, nothing more to do in this case. if(mPublish->header(h_Expires).value() != 0 ) { InfoLog(<< "SIPIfMatch failed -- republish"); mPublish->remove(h_SIPIfMatch); update(mDocument); return; } else { delete this; return; } } else if (code == 423) // interval too short ... } Hope this will be corrected in next candidate release Best Regards Fabrice ROUILLIER _______________________________________________ resiprocate-devel mailing list
|