| 00:28.12 | *** join/#asterisk-dev Netgeeks (~chris@173.11.68.156) |
| 00:28.12 | *** join/#asterisk-dev file (~file@asterisk/developer-and-muffin-lover/file) |
| 00:28.12 | *** join/#asterisk-dev JunK-Y (~junky@pdpc/supporter/active/junk-y) |
| 00:28.12 | *** join/#asterisk-dev WIMPy (~wimpy@e183095026.adsl.alicedsl.de) |
| 00:28.12 | *** join/#asterisk-dev Qwell (~north@pdpc/sponsor/digium/Qwell) |
| 00:28.15 | *** mode/#asterisk-dev [+oo file Qwell] by pratchett.freenode.net |
| 02:52.16 | *** join/#asterisk-dev abesamthomas (~abesamtho@61.11.125.45) |
| 04:05.55 | *** join/#asterisk-dev mjordan (~mjordan@user-69-1-6-49.knology.net) |
| 04:05.55 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 04:06.26 | *** part/#asterisk-dev mjordan (~mjordan@user-69-1-6-49.knology.net) |
| 06:03.09 | *** join/#asterisk-dev oron_laptop (~oron_lapt@line25-214.adsl.actcom.co.il) |
| 07:21.43 | *** join/#asterisk-dev schmidts (~schmidts@Lo20-rt01.lm33.sil.at) |
| 07:21.45 | schmidts | good morning |
| 07:35.34 | *** join/#asterisk-dev Freeaqingme (~Freeaqing@91.214.168.110) |
| 08:00.44 | *** join/#asterisk-dev lftsy (~lftsy@194.38.160.113) |
| 08:32.24 | *** join/#asterisk-dev oron_laptop (~oron_lapt@local.xorcom.com) |
| 09:06.26 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-155-170.cablep.bezeqint.net) |
| 09:06.26 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 09:27.58 | *** join/#asterisk-dev pnlarsson (~niklas@fw1.gml.g.icnet.infracom.se) |
| 09:53.50 | *** join/#asterisk-dev frawd (~francois@221.red-80-28-139.adsl.static.ccgg.telefonica.net) |
| 10:19.46 | *** join/#asterisk-dev hehol (~hehol@buero-gw.dortmund.loca.net) |
| 11:16.06 | *** join/#asterisk-dev jsmith (~jsmith@fedora/jsmith) |
| 11:16.06 | *** mode/#asterisk-dev [+o jsmith] by ChanServ |
| 11:38.43 | *** join/#asterisk-dev oron_laptop (~oron_lapt@local.xorcom.com) |
| 12:24.53 | *** join/#asterisk-dev w9sh (~chatzilla@64.238.96.125) |
| 12:35.46 | *** join/#asterisk-dev mjordan (~mjordan@user-69-1-6-49.knology.net) |
| 12:35.46 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 12:40.01 | *** join/#asterisk-dev moy (~moy@216.172.42.74) |
| 12:40.29 | *** join/#asterisk-dev kchehab (~kchehab@77.42.241.66) |
| 12:43.00 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 12:43.01 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 13:23.49 | *** join/#asterisk-dev ccesario (~ccesario@187.17.166.162) |
| 13:37.19 | schmidts | is it normal that the last entry of a linked list is set to NULL while the first is still a legal value? |
| 13:37.26 | schmidts | $2 = {first = 0x6574617473766564, last = 0x0} |
| 13:38.51 | *** join/#asterisk-dev manitou_ (~manitou@93.170.30.243) |
| 13:39.44 | kpfleming | schmidts: i wouldn't think so |
| 13:40.16 | schmidts | ok :( |
| 13:40.59 | schmidts | i am trying to build a linklist of allready allocated and unused buckets so its not necessary to alloc one each time we link an object into an ao2 container |
| 13:46.04 | kpfleming | it would probably be easier and just as effective to change the bucket structure to have a internal array that can hold 8 (or 16) entries at once |
| 13:51.57 | schmidts | but this would only work for normal hashed container and not for these with a hash value of 1 |
| 13:53.16 | kpfleming | richard mudgett is working on a different type of container implementation that would be useful for those cases though |
| 13:53.46 | schmidts | ok |
| 14:05.16 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 14:05.16 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 14:06.00 | *** join/#asterisk-dev serafie (~erin@nat/digium/x-wopkikdovjxbpfsq) |
| 14:19.20 | *** join/#asterisk-dev mjordan (~mjordan@nat/digium/x-bgrqjbxvxmhixeqb) |
| 14:19.20 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 14:52.39 | *** join/#asterisk-dev putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:52.39 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:53.44 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 14:53.44 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 15:00.24 | pabelanger | russellb: email to -users sent |
| 15:02.37 | pabelanger | Another day, another code review request. https://reviewboard.asterisk.org/r/1767/ |
| 15:03.26 | pabelanger | I'm not sure I've done the correct thing here. Since the original code did not care about the return results of write() and read(), I've updated the code to at least return error codes |
| 15:06.57 | russellb | pabelanger: ack |
| 15:08.11 | opticron | where are chan_usbradio and app_rpt actually maintained now if not in the main asterisk tree? |
| 15:08.26 | opticron | i guess I should hit google with that... |
| 15:11.04 | russellb | would be good to ask the guy that was talking about that in here last week ... if i could remember the nick |
| 15:14.24 | russellb | http://svn.ohnosec.org/viewvc/projects/allstar/?root=svn |
| 15:14.51 | opticron | oh wow...they stopped on 1.4 |
| 15:15.13 | opticron | I saw that but didn't think it was up to date |
| 15:15.21 | russellb | it has commits as of yesterday |
| 15:15.27 | opticron | yeah |
| 15:15.32 | pabelanger | <--this guy |
| 15:15.38 | pabelanger | and no, I don't want to maintain it |
| 15:15.48 | russellb | no, in that repo i just linked to |
| 15:15.53 | pabelanger | O I C |
| 15:16.04 | russellb | that's where it is maintained now |
| 15:16.15 | pabelanger | ya, then lets remove it |
| 15:19.07 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 15:19.07 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 15:19.10 | leifmadsen | wish we could remove it from existing branches |
| 15:19.22 | leifmadsen | it should at least be disabled by default, if not already |
| 15:22.47 | pabelanger | leifmadsen: as long as it compiles with --enable-dev-mode, I'm fine with it in 1.8, 10. Which it does now |
| 15:25.06 | leifmadsen | pabelanger: coolio |
| 15:41.30 | kpfleming | putnopvut: INV_TERMINATED is really the state for a completed (2xx response) INVITE transaction? that sounds terrible. |
| 15:42.28 | kpfleming | but once again, yay for SIP, with three different places to send the ACK depending on the response |
| 15:42.39 | Juggie | wanders in |
| 15:47.49 | seanbright | well look what the cat dragged in |
| 15:47.54 | seanbright | :P |
| 15:50.03 | putnopvut | kpfleming: yeah, when a 2xx is received, that's considered the end of the INVITE transaction. The ACK is considered separate, and that's why it uses the learned route set. On a non-2xx the ACK is considered part of the transaction and is sent to the same destination as the INVITE. |
| 15:50.50 | kpfleming | ok. i was more confused by the apparent misnaming of INV_COMPLETED and INV_TERMINATED. receiving a 2xx response doesn't seem like 'terminating' to me. |
| 15:53.17 | *** join/#asterisk-dev sruffell (~Adium@asterisk/the-kernel-guy/sruffell) |
| 15:53.17 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 15:53.50 | putnopvut | Lovely, I made bamboo unhappy |
| 15:54.10 | Juggie | lol... |
| 15:54.17 | Juggie | hi seanbright. |
| 15:54.17 | seanbright | howdy |
| 15:54.42 | putnopvut | Well, I made bamboo unhappy with my commit apparently, but it's not in code I changed. It's ooh323 code that it's complaining about. WEIRD |
| 15:54.49 | Juggie | I don't always idle :) |
| 15:54.57 | Qwell | putnopvut: same thing happened to terry |
| 15:57.22 | Juggie | seanbright, did digium rope you in yet? :) |
| 15:58.48 | seanbright | Juggie: no sir |
| 15:59.01 | seanbright | i just hang out here so i can have them fix my problems for me |
| 15:59.10 | Juggie | hah |
| 15:59.30 | Juggie | i've been playing w/ sphinx + cdrrecords for the last week |
| 15:59.50 | Juggie | the goal was instant lookup based on originating number from a table of 100+Million CDR Records. |
| 16:00.33 | Juggie | (sphinx search) |
| 16:03.47 | Juggie | it worked out well though I am still not sure if I would keep it around or invest more time into getting the same performance out of SQL w/ a proper index. |
| 16:05.32 | seanbright | that sounds really boring |
| 16:05.34 | seanbright | what else is going on? |
| 16:05.36 | seanbright | heh |
| 16:05.40 | Juggie | lol. |
| 16:05.54 | Juggie | i write php code all day and manage a team of web developers so not a lot. |
| 16:06.12 | kpfleming | so the third circle of hell, then |
| 16:06.31 | Juggie | kinda |
| 16:06.44 | seanbright | you're in good company |
| 16:06.52 | Juggie | there is some sembelence of order as what we do is write web apps to support telephony infrastructure |
| 16:06.56 | russellb | how do you write code all day and manage a team of devs |
| 16:06.57 | Qwell | Juggie: still at revenue canada? |
| 16:06.58 | seanbright | i've been going through pcaps all week |
| 16:07.02 | russellb | sounds like you don't actually write php code all day |
| 16:07.03 | russellb | :-p |
| 16:07.09 | Juggie | Qwell, we are called CRA now but yes. |
| 16:07.19 | Juggie | russellb, not all day :P |
| 16:07.24 | Juggie | 1hour, close enough. |
| 16:07.26 | Juggie | :) |
| 16:07.29 | Qwell | Juggie: I finally just got used to revenue coming before Canada, and now you change it on me? sheesh. |
| 16:07.44 | Juggie | Qwell, its been CRA for 5+ years. |
| 16:07.50 | file | Canada Revenue Agency. |
| 16:07.55 | Qwell | I'm a bit behind. |
| 16:08.00 | Juggie | we have 30+ * boxes now. |
| 16:08.02 | Juggie | whee |
| 16:08.25 | russellb | the plural of box is boxen |
| 16:08.34 | Juggie | the sad part is, they are on 1.4 |
| 16:08.58 | russellb | exposed to the internet? |
| 16:09.00 | Juggie | no. |
| 16:09.10 | Juggie | i dont manage them. |
| 16:09.15 | Juggie | not my area. |
| 16:09.16 | russellb | but if you did ... |
| 16:09.19 | russellb | they would be? |
| 16:09.29 | russellb | also, hi Juggie ! |
| 16:09.30 | Juggie | if i did they would be upgraded long ago |
| 16:09.34 | Juggie | hi russellb |
| 16:09.37 | Juggie | hows the new gig? |
| 16:09.42 | russellb | good. |
| 16:09.58 | russellb | i sit around and look pretty in a red fedora all day |
| 16:10.04 | pabelanger | putnopvut: no, you are good |
| 16:10.27 | Juggie | russellb, who does redhat hate more, centos or oracle (for also forking rhel) :) |
| 16:10.45 | russellb | heh, oracle probably |
| 16:10.50 | Juggie | :P |
| 16:10.53 | Qwell | Juggie: As far as I understand it, CentOS is quite welcome. |
| 16:10.54 | kpfleming | they are unbreakable! |
| 16:11.45 | Juggie | but I think 30+ boxen with 2-4PRI's each |
| 16:11.47 | Juggie | i dont recall. |
| 16:12.01 | Juggie | mostly 4 I believe. |
| 16:13.06 | Juggie | we have some small call centers on them as well, though one of our dev's wrote his own distributed agent queing engine. |
| 16:13.23 | russellb | seems to be common once people want to scale queues ... |
| 16:13.38 | seanbright | we did the same here |
| 16:13.41 | Juggie | ya for sure. |
| 16:13.50 | Juggie | if i had the time |
| 16:13.53 | Juggie | (which i don't) |
| 16:14.03 | Juggie | i would look to blend asterisk + queues + redis |
| 16:14.20 | russellb | why redis? |
| 16:14.28 | Juggie | as it has all the internal structs you wuold need |
| 16:14.44 | Juggie | lists w/ priority, etc... |
| 16:15.02 | Juggie | also its clustered and async backed to disk |
| 16:15.07 | seanbright | sexy |
| 16:15.15 | Juggie | its memcached + sexy |
| 16:15.26 | seanbright | i'm using mongodb + redis + haskell |
| 16:15.31 | seanbright | well... they are installed |
| 16:15.37 | seanbright | actually that's not even true |
| 16:15.42 | seanbright | i've just heard of them |
| 16:15.48 | russellb | lolz. |
| 16:16.05 | Juggie | but it has push and pop does transactions, can be atomic, etc. |
| 16:16.06 | seanbright | i'm old school RDBMS |
| 16:16.19 | seanbright | even when it doesn't make sense and is completely inefficient |
| 16:16.25 | seanbright | because that's how i roll |
| 16:16.33 | Juggie | i did some reading on mango |
| 16:16.37 | Juggie | it tempted me |
| 16:16.40 | Juggie | but i returned to sql |
| 16:17.09 | russellb | well i wrote my distributed queue system with erlang + hadoop + zeromq + ms access + visual basic |
| 16:17.18 | file | russellb, sounds awesome |
| 16:17.28 | russellb | it is |
| 16:17.30 | seanbright | i actually want to use mongodb for our scripting data |
| 16:17.38 | seanbright | right now we just php serialize all of them into a database column |
| 16:17.43 | Juggie | mangodb is very JS like for querying |
| 16:17.45 | seanbright | which is god awful when you want to query on them |
| 16:17.54 | seanbright | but if they were in mongodb it would be trivial |
| 16:18.05 | Juggie | postgres just added JSON support i think |
| 16:18.08 | Juggie | not sure if mysql has it |
| 16:18.19 | seanbright | nice |
| 16:18.36 | Juggie | but you should convert to json over serialize |
| 16:18.41 | Juggie | its one in the same but json is more portable |
| 16:18.50 | seanbright | true story |
| 16:19.05 | seanbright | either way i can't query on it while it's in the database column |
| 16:19.15 | Juggie | i believe w/ postgres you can now |
| 16:19.18 | Juggie | as it understands json |
| 16:19.19 | seanbright | well then i am sold |
| 16:19.23 | Juggie | and has functions to work w/ it |
| 16:19.56 | Juggie | http://arstechnica.com/business/news/2012/01/native-json-features-submitted-for-postgresql-92.ars |
| 16:22.08 | Juggie | mango looks nice for that though for sure being able to have abstract records is a plus |
| 16:22.21 | seanbright | indeed. |
| 16:22.26 | seanbright | but... asterisk is awesome |
| 16:22.56 | seanbright | we obviously needs to be able to store cdrs in mongodb |
| 16:23.05 | Juggie | cdr_mango |
| 16:23.12 | Juggie | :) |
| 16:24.11 | Juggie | hm |
| 16:24.39 | Juggie | just got an email from dell for 'Dell Voice' which lets you make and receive calls for free on android or iphone and provides a free new canadian number? |
| 16:24.40 | Juggie | odd. |
| 16:26.17 | file | it's fongo |
| 16:26.35 | Juggie | never heard of them |
| 16:39.51 | *** join/#asterisk-dev otherwiseguy (~otherwise@rrcs-173-197-0-67.west.biz.rr.com) |
| 16:54.49 | Juggie | the site barely works |
| 17:39.10 | *** part/#asterisk-dev mjordan (~mjordan@nat/digium/x-bgrqjbxvxmhixeqb) |
| 17:44.44 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 18:00.35 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 18:00.35 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 18:25.15 | acidfoo | what means "IE" in AST_EVENT_IE_MAILBOX |
| 18:26.10 | seanbright | information element |
| 18:27.15 | acidfoo | thank you ;) |
| 18:29.32 | seanbright | you're welcome |
| 18:36.35 | Kobaz | do de do |
| 18:36.53 | Kobaz | so i think i found a bug in mixmonitor but i can't reproduce it |
| 18:37.04 | Kobaz | well it's reproducible but not always |
| 18:37.10 | Kobaz | audio is saved to disk out of order |
| 18:47.05 | *** join/#asterisk-dev pnlarsson (~niklas@fw1.gml.g.icnet.infracom.se) |
| 18:56.50 | *** join/#asterisk-dev zerohalo (~zerohalo@74-94-141-204-ma.hfc.comcastbusiness.net) |
| 19:01.47 | wdoekes2 | file: re: no valid transports: so I can just commit that in 1.8? |
| 19:36.47 | acidfoo | is there a function to query a voicemail to get informations like the number of message in the box, etc |
| 19:36.54 | acidfoo | something else than the event subsystem |
| 19:37.15 | acidfoo | well, I want to poke it when I need |
| 19:37.24 | acidfoo | thanks in advance |
| 19:39.39 | acidfoo | in the asterisk 1.8 serie ;) |
| 19:39.45 | jsmith | system function, or dialplan function? |
| 19:39.52 | acidfoo | system |
| 19:40.16 | jsmith | I'm sure there is... but I don't know it off the top of my head -- but looking at the minivm code should make it easy to find |
| 19:40.38 | acidfoo | ok, ill dig into minivm thanks |
| 20:13.58 | acidfoo | well, I don't find it :| |
| 20:23.13 | *** join/#asterisk-dev alecdavis (~alec@202.36.76.21) |
| 20:30.02 | Kobaz | so who knows mixmonitor well? |
| 20:32.52 | jsmith | Kobaz: (and is willing to publicly admit it?!?) |
| 20:33.05 | jsmith | ducks and hides |
| 20:33.15 | seanbright | mixmonitor is 95% audiohooks |
| 20:36.00 | *** join/#asterisk-dev tamiel (~tamiel@85-171-171-179.rev.numericable.fr) |
| 20:43.54 | Kobaz | nice |
| 20:44.04 | Kobaz | so what would cause mixmonitor to write audio out of sequence? |
| 20:52.32 | seanbright | Kobaz: receiving it out of sequence? |
| 20:52.46 | seanbright | audiohooks don't do any jitterbuffering |
| 20:52.57 | seanbright | so if frames are out of order that is how they will be written |
| 20:53.59 | leifmadsen | neato thought |
| 20:54.06 | leifmadsen | JITTERBUFFER() function? |
| 20:56.28 | seanbright | i think enabling it on the endpoints would be sufficient |
| 20:56.42 | seanbright | but i'm not 100% on that |
| 20:58.04 | leifmadsen | well you could more easily test if that may be the case by adding a jitterbuffer in front of the mixmonitor, wouldn't it? |
| 20:58.17 | leifmadsen | s/wouldn't it/couldn't you/ |
| 20:59.38 | jsmith | Thank goodness for local channels and the /j flag :-) |
| 21:00.33 | russellb | jsmith: Asterisk 10 has a much cleaner way of doing it, luckily |
| 21:00.43 | jsmith | russellb: Oh? I'd love to learn more :-) |
| 21:00.44 | russellb | use a function to enable it on a channel |
| 21:01.39 | leifmadsen | pointed out the function already :) |
| 21:01.59 | russellb | https://wiki.asterisk.org/wiki/display/AST/Asterisk+10+Function_JITTERBUFFER |
| 21:07.19 | *** join/#asterisk-dev goddva (~glarsen@77.40.154.242) |
| 21:11.42 | jsmith | Whee! |
| 21:11.45 | jsmith | That's awesome :-) |
| 21:13.23 | Kobaz | mm |
| 21:13.25 | Kobaz | so anyway |
| 21:13.27 | Kobaz | here's an example |
| 21:13.33 | Kobaz | http://www.kobaz.net/misc/1329956043.4232070.WAV |
| 21:13.44 | Kobaz | in the real call, a talks, b talks, a talks |
| 21:13.49 | Kobaz | in the recording, b is talking over a |
| 21:13.58 | russellb | silence suppression? |
| 21:14.07 | Kobaz | on the endpoint? |
| 21:14.11 | *** join/#asterisk-dev otherwiseguy (~otherwise@rrcs-173-197-0-67.west.biz.rr.com) |
| 21:14.20 | russellb | yeah, where it's only sending audio when talking |
| 21:14.26 | russellb | can't remember of mixmonitor handles that properly |
| 21:14.39 | Kobaz | i have this |
| 21:14.40 | russellb | or monitor |
| 21:14.41 | Kobaz | [general] |
| 21:14.41 | Kobaz | transmit_silence => yes |
| 21:14.41 | russellb | or whatever |
| 21:14.49 | Kobaz | one endpoint is a dahdi device |
| 21:14.52 | Kobaz | the other endpoint is a polycom |
| 21:15.17 | Kobaz | i think this may have started happening when i upgraded dahdi |
| 21:22.51 | seanbright | yeah, we have similiar recordings here |
| 21:22.55 | seanbright | no dahdi device involved |
| 21:28.21 | Kobaz | ah okay |
| 21:28.25 | Kobaz | using dahdi for timing? |
| 21:28.30 | seanbright | yeah |
| 21:28.33 | seanbright | and trascoding |
| 21:28.36 | Kobaz | what version? |
| 21:28.39 | seanbright | 2.6.0 |
| 21:28.42 | Kobaz | ah |
| 21:28.50 | Kobaz | notice any problems with using an older one? |
| 21:29.03 | Kobaz | like did you start getting weird recordings after an upgrade |
| 21:29.09 | Kobaz | or it's just been happening |
| 21:29.22 | seanbright | nothing concrete |
| 21:29.31 | seanbright | we're just starting to notice this particular problem lately |
| 21:29.43 | Kobaz | yeah me too |
| 21:29.57 | Kobaz | it started happening a few months ago but i thought it went away |
| 21:29.57 | seanbright | but we've moved from 1.6.0 to 1.6.2 to 1.8.8 in a fairly quick timeframe |
| 21:30.04 | seanbright | hard to say when things went south |
| 21:30.05 | Kobaz | yeah i'm on 1.8.5 |
| 21:30.21 | seanbright | it doesn't happen on all calls, just some |
| 21:30.25 | Kobaz | yeah |
| 21:30.30 | Kobaz | what kernel version? |
| 21:30.52 | seanbright | Linux my-server 2.6.32-36-server #79-Ubuntu SMP Tue Nov 8 22:44:38 UTC 2011 x86_64 GNU/Linux |
| 21:30.56 | Kobaz | mm |
| 21:31.02 | Kobaz | mines much older, so that's probably not it |
| 21:31.03 | seanbright | if you're able to replicate it easily, a packet capture would be useful |
| 21:31.26 | Kobaz | Linux alinksA 2.6.27.38 #1 SMP PREEMPT Mon Oct 26 17:02:17 EDT 2009 i686 GNU/Linux |
| 21:31.29 | seanbright | actually that problable wouldn't help |
| 21:31.52 | seanbright | we have that without using mixmonitor |
| 21:32.03 | seanbright | we just use straight monitor and mix later |
| 21:32.13 | Kobaz | oh |
| 21:32.17 | seanbright | one leg just seems to get less data |
| 21:32.24 | Kobaz | i had problems with monitor |
| 21:32.32 | Kobaz | because i run cron jobs that suck up lots of io |
| 21:32.47 | Qwell | Kobaz: ramdisk |
| 21:32.48 | Kobaz | so calls would drop audio because monitor is inline in the audio bridging |
| 21:32.57 | seanbright | we don't want to waste transcoders (g729->whatever) |
| 21:33.08 | seanbright | so we have a process that processes the recordings in batches after the call ends |
| 21:33.15 | Kobaz | using mixmonitor fixed it because it's in a thread |
| 21:33.19 | Kobaz | yeah me too |
| 21:33.24 | Kobaz | i encode to mp3 when it's all done |
| 21:33.28 | seanbright | yeah, same. |
| 21:33.44 | Kobaz | every 5 minutes i do a new mp3 batch |
| 21:33.46 | seanbright | well, mixmonitor isn't really in a thread |
| 21:33.55 | seanbright | well, let me rephrase |
| 21:34.03 | seanbright | audiohooks are "inline" in the audio bridging |
| 21:34.42 | seanbright | and each frame is transcoded and put into a buffer |
| 21:34.43 | Kobaz | i remember looking at mixmonitor and it spawns a thread somewhere |
| 21:34.53 | seanbright | you're right, there is a thread |
| 21:34.57 | Kobaz | yeah it buffers the audio and then the writer is a thread |
| 21:35.11 | seanbright | and that thread wakes up and mixes the audio and dumps it back out to disk |
| 21:35.28 | Kobaz | so as long as both legs are dumping their packets |
| 21:35.31 | Kobaz | silence or not |
| 21:35.51 | Kobaz | stuff should be in order |
| 21:36.08 | seanbright | do you transcode? |
| 21:36.17 | Kobaz | nope |
| 21:36.26 | Kobaz | actually |
| 21:36.26 | Kobaz | well |
| 21:36.28 | Kobaz | yeah |
| 21:36.33 | seanbright | ulaw from dahdi -> ulaw on polycom? |
| 21:36.34 | Kobaz | record from ulaw to WAV |
| 21:36.50 | Kobaz | call is dahdi <=> polycom |
| 21:36.51 | Kobaz | yeah |
| 21:36.58 | seanbright | but it's ulaw the whole way? |
| 21:37.04 | seanbright | and why WAV? |
| 21:37.13 | Kobaz | no idea |
| 21:37.18 | Kobaz | that's just what i picked forever ago |
| 21:37.19 | seanbright | and you mean wav49? |
| 21:37.21 | Kobaz | i need to switch to something better |
| 21:37.23 | seanbright | gsm encoded wav? |
| 21:37.33 | Kobaz | whatever format MixMonitor encoded when you do file.WAV |
| 21:37.36 | jsmith | Interesting... what's up with sending out SIP notifies with 0 as the port number on contact? It does a lovely number on U-verse routers :-/ |
| 21:37.37 | seanbright | yeah |
| 21:37.38 | Kobaz | yeah gsm |
| 21:37.41 | seanbright | stop doing that |
| 21:37.43 | jsmith | (Send a SIP notify, watch the router reboot) |
| 21:37.44 | seanbright | go to slinear |
| 21:37.45 | Kobaz | hah |
| 21:37.48 | Kobaz | k |
| 21:37.54 | Qwell | jsmith: seriously? |
| 21:37.58 | jsmith | Qwell: Seriously :-( |
| 21:38.04 | Kobaz | does sox do slinear? |
| 21:38.06 | Qwell | that's a serious security flaw |
| 21:38.07 | seanbright | it does |
| 21:38.09 | Kobaz | ah nice |
| 21:38.12 | jsmith | Qwell: One of my clients has at least six different customers complaining about it |
| 21:38.12 | seanbright | it's called 'raw' |
| 21:38.13 | Kobaz | okay so i can wav it and then mp3 it |
| 21:38.15 | Kobaz | ah |
| 21:38.28 | Kobaz | okay i'll try that |
| 21:38.30 | Kobaz | gotta go to dinner |
| 21:38.30 | jsmith | Qwell: Took us a little while to figure out the cause... |
| 21:38.49 | Qwell | jsmith: I don't know the answer to your question, unfortunately. I wonder if an incoming unauthed packet would do it... |
| 21:38.58 | jsmith | Qwell: Apparently the SIP ALGs on the firewall can't handle dealing with port 0 |
| 21:39.12 | seanbright | Kobaz: that won't help your audio problem though |
| 21:39.43 | Kobaz | oh |
| 21:39.50 | Kobaz | i did notice the recordings sound pretty crappy |
| 21:39.53 | Kobaz | but hmm |
| 21:39.54 | Kobaz | okay |
| 21:39.55 | jsmith | Qwell: I'll send you over a ticket number in a little while, if you care to stare at the carnage :-p |
| 21:40.01 | Kobaz | we need to work on this later |
| 21:40.06 | Kobaz | this is totally killing me |
| 21:40.07 | Qwell | u-verse ticket? I'd love to see that. |
| 21:40.10 | Kobaz | and probably you too |
| 21:40.41 | seanbright | Kobaz: it doesn't happen that frequently, but it's a bit of a pain, yes. |
| 21:40.47 | seanbright | i would like to get it resolved |
| 21:41.26 | Kobaz | anyways |
| 21:41.29 | Kobaz | busting out for birthday dinner |
| 21:42.37 | jsmith | Qwell: https://issues.asterisk.org/jira/browse/ASTERISK-19430 |
| 21:42.47 | Qwell | aww, I was hoping for a u-verse ticket. :( |
| 21:43.06 | Qwell | oh. That must be why Tony is trying to talk to me on jabber. |
| 21:43.10 | jsmith | Qwell: Yeah, right... I'm not that crazy. |
| 21:43.14 | jsmith | Qwell: Probably so :-p |
| 21:43.37 | Qwell | wait |
| 21:43.51 | Qwell | I was right? It actually is an unauthed incoming packet that does it? |
| 21:43.53 | Qwell | That is incredible. |
| 21:44.28 | file | that's... weird |
| 21:45.49 | Qwell | Do we maybe want to make this private? |
| 21:45.55 | Qwell | I think that might be wise... |
| 21:48.56 | Qwell | jsmith: Hopefully you can still see that issue |
| 21:49.08 | jsmith | Qwell: Sure, if you can add me to it |
| 21:49.38 | jsmith | Qwell: Nope, I can't see it any more :-( |
| 21:49.46 | jsmith | Qwell: Can you add me as a watcher? |
| 21:49.56 | Qwell | jsmith: I'll ping espiceland. I don't think watcher will do it (though I'll certainly try) |
| 21:50.10 | jsmith | Qwell: Ok, thanks... |
| 21:54.40 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 21:54.40 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 22:16.26 | *** join/#asterisk-dev mjordan (~mjordan@nat/digium/x-xabrwmvgvalbpaji) |
| 22:16.26 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |