| 00:18.30 | *** join/#asterisk-dev droemel (~droemel@p4FCACC55.dip.t-dialin.net) |
| 00:29.39 | *** join/#asterisk-dev Qwell (~north@pdpc/sponsor/digium/Qwell) |
| 00:29.50 | *** mode/#asterisk-dev [+o Qwell] by ChanServ |
| 01:51.22 | Corydon76-home | mvanbaak: weird. You do know that res_curl is a prereq for res_config_curl, right? |
| 01:55.41 | mvanbaak | uhhuh |
| 01:57.00 | mvanbaak | in the end I ran; rm *curl* |
| 01:57.09 | mvanbaak | in the asterisk modules dir |
| 03:00.36 | *** join/#asterisk-dev seraphie (~erin@75.76.38.159) |
| 05:33.53 | *** join/#asterisk-dev irroot (~irroot@197.173.201.193) |
| 06:17.03 | *** join/#asterisk-dev irroot (~irroot@197.170.39.99) |
| 08:20.15 | *** join/#asterisk-dev hehol (~Adium@2a01:198:71d:0:21f:d0ff:fea1:568e) |
| 09:38.48 | *** join/#asterisk-dev Fritz09 (~Adium@pop1-2535.catv.wtnet.de) |
| 10:25.19 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 10:25.19 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 10:42.52 | mvanbaak | meh, I found out how to get asterisk in a 'Stopped' state |
| 10:42.54 | mvanbaak | unkillable |
| 10:43.04 | mvanbaak | asterisk -rx 'core restart now' |
| 10:44.24 | mvanbaak | 50523 ?? Ts 2:46.58 asterisk -U asterisk -G asterisk |
| 10:44.46 | mvanbaak | now I have to reboot my machine :( |
| 10:50.54 | *** join/#asterisk-dev mvanbaak (~michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 10:50.54 | *** mode/#asterisk-dev [+o mvanbaak] by ChanServ |
| 11:03.55 | irroot | mvanbaak odd is it when you only do it -rx that it barfs ?? |
| 12:23.34 | mvanbaak | irroot: no, if I do 'asterisk -r' and then type 'core restart now' it happens as well |
| 12:33.01 | irroot | that sux |
| 12:33.34 | irroot | im at home and dont want to break things at office ill check it out and try it what ver you run |
| 12:48.09 | mvanbaak | trunk |
| 12:48.13 | mvanbaak | on FreeBSD 8.2 |
| 12:48.26 | mvanbaak | Asterisk SVN-trunk-r334236 built by mvanbaak @ ifrid.vanbaak.info on a amd64 running FreeBSD on 2011-09-01 21:51:04 UTC |
| 13:19.29 | irroot | i dont got fleabsd only penguins |
| 13:32.04 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-84-111-20-179.red.bezeqint.net) |
| 13:32.04 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 14:29.39 | Corydon76-home | mvanbaak: Are you using the dahdi-freebsd kernel drivers? A lockup like that should be impossible from userspace alone and generally indicates a kernel (or kernel driver) bug |
| 15:02.15 | *** join/#asterisk-dev acidfu (~nib@206-248-162-54.dsl.teksavvy.com) |
| 15:23.20 | *** join/#asterisk-dev chazzam (~chazz@173-24-236-90.client.mchsi.com) |
| 15:40.21 | otherwiseguy | Corydon76-home: With the realtime API as it is and the whole "sometimes we don't pass a key, we pass a key space operator" thing, it makes it hard(er) to properly escape column names (in fact, I'd like to put them in backquotes, but don't really want to go to the trouble of splitting the string on space, etc.). It really is a horrible abuse of the realtime API that has cropped up in realtime usage. "mailbox LIKE", '%", etc. As if LIKE is supposed to |
| 15:40.21 | otherwiseguy | valid across all realtime backends. It seems like it would be much better to have the API take alternating (key, op, value) varargs, then define the ops as RT_EQ/LT/GT/LTE/GTE/LIKE, etc. and let the backends define equivalents for themselves. |
| 15:41.27 | otherwiseguy | instead of silly string parsing games |
| 15:42.39 | otherwiseguy | You'd still be stuck with having to define a wildcard value that would need to be parsed for (defaulting to '%' most likely), but it would make things better |
| 15:48.39 | mvanbaak | Corydon76-home: no, no dahdi stuff here |
| 16:39.48 | otherwiseguy | hmm, actually backticks around field names seems to be a mysql thing, should be double quoted for column and table names in sqlite3. But, the point is still the same. The realtime API could use some work to make it more generic. |
| 17:18.47 | pabelanger | Is chan_iax2 different then chan_sip, where it can bind to multiple IP interfaces, but chan_sip cannot? |
| 17:22.53 | pabelanger | Also, looks like svn.asterisk.org is lagging behind origsvn.digium.com |
| 17:24.02 | irroot | pabelanger agreed on the SVN thing been cursing SVN "svn ps XXX -F ...." adds cr/lf and breaks auto merge *facepalm* |
| 17:24.41 | irroot | ps that problem with round robin logged and redid the whole rr code in there |
| 17:50.25 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-155-146.cablep.bezeqint.net) |
| 17:50.25 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 19:09.32 | *** join/#asterisk-dev Gugge (~gugge@46.30.120.1) |
| 19:24.11 | *** join/#asterisk-dev Corydon76-home (blue@c-69-137-80-31.hsd1.tn.comcast.net) |
| 19:24.11 | *** mode/#asterisk-dev [+o Corydon76-home] by ChanServ |
| 19:25.41 | *** join/#asterisk-dev Corydon76-home (mauve@c-69-137-80-31.hsd1.tn.comcast.net) |
| 19:25.41 | *** mode/#asterisk-dev [+o Corydon76-home] by ChanServ |
| 19:26.06 | *** join/#asterisk-dev otherwiseguy (~terry@69.76.217.40) |
| 19:27.43 | *** join/#asterisk-dev Corydon76-home (four@c-69-137-80-31.hsd1.tn.comcast.net) |
| 19:27.43 | *** mode/#asterisk-dev [+o Corydon76-home] by ChanServ |
| 19:28.01 | otherwiseguy | Corydon76-home: howdy |
| 20:42.49 | *** join/#asterisk-dev Corydon76-dig (gold@c-69-137-80-31.hsd1.tn.comcast.net) |
| 20:42.49 | *** mode/#asterisk-dev [+o Corydon76-dig] by ChanServ |
| 21:24.17 | *** join/#asterisk-dev bbryant1 (~brett@c-174-56-132-225.hsd1.sc.comcast.net) |
| 21:24.17 | *** mode/#asterisk-dev [+o bbryant1] by ChanServ |
| 21:29.12 | pabelanger | Wow, I just looked at how the channel drivers do the logic for bindaddr, and each channel drivers does it differently :( |
| 21:29.48 | Corydon76-home | finka: ENOCONTEXT |
| 21:43.02 | Corydon76-home | pabelanger: yes, it appears that chan_iax can work with multiple sockets, although it appears that the socket chosen is based upon a peer setting, not based upon from which address it received a packet |
| 21:43.24 | Corydon76-home | Or on which socket, that is |
| 21:47.46 | pabelanger | Corydon76-home: Right, comparing chan_iax to other channel drivers, it actually binds at the point of reading the bindaddr setting, while other channel drivers do it after the file has been parsed. I'l working on a patch to try something now, but I was surprised to see chan_iax allowing multiple bindaddr. |
| 21:48.24 | pabelanger | Some generic testing, with multiple bindaddr does not look good. I think asterisk is getting confused on which socket to send the reply |
| 21:48.39 | Corydon76-home | So in order to use a second or third bindaddr, you need to set a peer's sourceaddress setting to one of those bound IPs |
| 21:49.25 | pabelanger | Well, for my purpose, I'm not looking to use a 2nd or 3rd bindaddr but simple override the previous defined one |
| 21:49.39 | pabelanger | like all other channel drivers |
| 21:50.06 | Corydon76-home | You could effectively have an internal IAX peer with no password, and external IPs would not be able to falsify that connection, because it would arrive on the wrong bound address |
| 21:50.38 | Corydon76-home | You'll break existing behavior if you change the IAX2 driver to do that |
| 21:50.51 | pabelanger | I know :( |
| 21:51.01 | Corydon76-home | A better modification would be to modify the other drivers to allow binding to multiple addresses |
| 21:51.22 | pabelanger | nods |
| 21:51.38 | Corydon76-home | Preferably in the expected way, whereby a connection coming in on one IP address will be forever bound to that IP. |
| 21:52.46 | pabelanger | For now, I'm trying to see if binding multiple address for chan_iax2 is supported. Then we can decided if the configuration file is a bug or not. |
| 21:52.58 | pabelanger | If it can, then it is the only channel driver to do so |
| 21:53.05 | Corydon76-home | pabelanger: It is supported, in that peculiar way. |
| 21:53.43 | Corydon76-home | Not the expected way, but one in which peers are restricted to a particular bound address |
| 21:54.04 | Corydon76-home | Search for "sourceaddress" in the IAX2 code to see |
| 21:56.47 | pabelanger | Damn, this is going to be a problem for me and the testsuite. All other configuration files with bindaddr, can be overrided (sp?) at a later point in the same section. Now I have to figure out something specific for chan_iax |
| 21:57.06 | Corydon76-home | overridden |
| 21:57.33 | pabelanger | heh, fail |
| 21:57.53 | pabelanger | oh well, would be nice is all configuration files worked the same |
| 21:57.54 | pabelanger | <PROTECTED> |
| 22:01.24 | Corydon76-home | Well, you can. Just specify a sourceaddress for each IAX2 peer, and you're homefree. |
| 22:01.46 | Corydon76-home | where the sourceaddress matches your last bindaddr in the file. |
| 22:03.59 | pabelanger | Well, the fact our iax.conf.sample files does not reference sourceaddress is not good :( |
| 22:04.04 | pabelanger | let me see what it does |
| 22:05.32 | pabelanger | Ya, that looks to be at the peer level, I'm looking for something at the global. EG: starting 2 instances of asterisk on the same box. Wanting 1 to bind to 127.0.0.1 and the second to 127.0.0.2 |
| 22:08.44 | Corydon76-home | Ah. Yeah, you'll need a completely different iax.conf for that |
| 22:09.28 | Corydon76-home | What's the reason for doing all the #include magic? |
| 22:11.45 | Corydon76-home | pabelanger: then again, if something is already bound to 0.0.0.0, port 4569, I think another process trying to do the same thing will simply fail on that initial bind |
| 22:12.12 | Corydon76-home | and then the second bound address becomes default for that second process |
| 22:13.01 | Corydon76-home | Not perfectly optimal, but it might work |
| 22:21.20 | *** join/#asterisk-dev otherwiseguy (~terry@69.76.217.40) |
| 22:25.22 | Corydon76-home | pokes otherwiseguy |
| 22:46.32 | *** join/#asterisk-dev BigMike (~Miranda@p57AD91F5.dip0.t-ipconnect.de) |
| 23:13.08 | *** join/#asterisk-dev Fritz09 (~Adium@pop1-2535.catv.wtnet.de) |
| 23:19.06 | *** join/#asterisk-dev seraphie (~erin@75.76.38.159) |