hello yate - Rev 1846 not resolved problem one a one call by "one a one call" you mean the individual message not ack-ed from external Python issue, right? i suggest to capture with ethereal/wireshark the tcp port, look at the timing of exchanged text lines there should be pairs of %>message and % i mean %%>message and %% hello kill the mensagges is correct, both in time and in quantity messages are not lost and have no lag so there is little time between %%> and %%< and long time between these groups? it is as if it were an emergency gate, which open single-sided the message from the first call is unable to return to yate again, at TCP level is ok but in Yate is not? can you pastebin somewhere the output of tcpdump so i can look at the times? Pending enter a message in a second call to reopen the door, there the message of the first call may happen because the door became open yes kill i have pastebin minute * _monica_ has quit IRC ("Leaving") kill no problem four you i capture with ethereal/wireshark ? well, you can't put that in a pastebin ;-) * vircuser has joined #yate I pass messages to a file and paste in pastebin hi * killall-9 has quit IRC ("changing computer") * killall-9 has joined #yate * l-fy has quit IRC (Nick collision from services.) * vircuser is now known as l-fy done kill talk to my partner who has a yate in his notebbok for sniff and I said that there was no problem of packages or times, not lost in your notebook runs a old version yate and is not the problem the problem appeared with a new revision does this happen all the time or only on 1st message? 1st mensage = 1st call --> not connect 2st mensage = 2st call --> connect two calls at the same time call 1 and 2 connect at the same time 3st mensage = 3st call --> not connect 4st mensage = 4st call --> connect again two calls (3 call and 4 call) second mennsage re-open door and connect last calls can you please capture the communication between yate and python? i tried to reproduce with php and failed to ok kill, no problem now I think update python to the new release 2.4.5 (March 11, 2008) older version running 2.4.4 mmm Maybe this new release python causing the problem i think my partner has in his notebook a yate running OK one a one call as it arrives at the office will see that version of python and revision of yate is assigned running, and test re python upgrade to the new release and make a call i bet that is the new version of python if you can wait kill, I think it will be the best thing, at least to establish that the problem is guess that also want to solve the problem for the new release of python