search for: ast_softhangup

...herefore, I've added the following hideous hack - er, code - to voicemail2.c. It starts right after the call to play_and_record() in leave_voicemail(). if (res != '#' && chan != NULL && !strncmp(chan->name, "Zap", 3)) { /* Hang up the Zap channel only */ ast_softhangup(chan, AST_SOFTHANGUP_EXPLICIT); } Obviously, it hangs up the channel after the voicemail has been recorded, if the # key wasn't pressed, if the channel still exists, and if it's a Zap channel. I couldn't see a way to do this with AGI. Question: is this safe? I used a soft hangup bec...

Thanks to ww and his patch on bug #104, I have successfully implemented Asterisk behind NAT without using STUN or anything crazy. It's quite straight forward. Until this gets tested enough and put into CVS, you will have to patch your chan_sip.c file to do this. I'm sure within the next few days this will get put merged into CVS if no one finds any problems. I tried this on chan_sip.c

I'm running the a very recent CVS version of asterisk on an RH9 machine. My problem is that my x100p takes about 10 seconds to detect a hangup. After that it takes about 10 more seconds for the the zaptel device to release the line. Here's an example of my console report: == Parsing '/var/spool/asterisk/voicemail/default/101/INBOX/msg0000.txt': == Parsing

...masquerade': /usr/src/asterisk/channel.c:2112: undefined reference to `pthread_kill' channel.o: In function `tonepair_generator': /usr/src/asterisk/channel.c:2418: undefined reference to `sin' /usr/src/asterisk/channel.c:2418: undefined reference to `sin' channel.o: In function `ast_softhangup': /usr/src/asterisk/channel.c:587: undefined reference to `pthread_kill' channel.o: In function `ast_softhangup_nolock': /usr/src/asterisk/channel.c:587: undefined reference to `pthread_kill' pbx.o: In function `ast_async_goto': /usr/src/asterisk/pbx.c:1873: undefined reference...