| 00:04.37 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 00:22.44 | *** part/#asterisk-dev mjordan (~matt@nat/digium/x-rvztklxtrwsoccsh) |
| 00:34.43 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 00:40.18 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.5) |
| 01:35.28 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 01:40.00 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.173) |
| 02:05.26 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 02:35.19 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 02:40.40 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.183) |
| 03:35.21 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 03:40.23 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.212) |
| 03:47.25 | wedhorn | snuff-work: ping |
| 03:48.16 | snuff-work | wedhorn: pong |
| 03:48.41 | wedhorn | you still got that ccm box around? |
| 03:48.47 | snuff-work | yes |
| 03:49.00 | wedhorn | hehe, wrong answer.... |
| 03:49.29 | wedhorn | I thing I've found a bit of an ugly bug, and no idea why |
| 03:49.46 | snuff-work | dont envisage getting rid of it anytime soon completely |
| 03:50.17 | wedhorn | If you have a call on a v17 7961, and another call comes in, you lose the ability to press the number buttons |
| 03:50.52 | wedhorn | ie, go to voicemail, then have an inbound call, and you can no longer enter numbers into the voicemail connection |
| 03:51.26 | snuff-work | seems odd.. the other call is ringing in background right? |
| 03:51.29 | wedhorn | I assume ccm doesn't have this issue |
| 03:52.04 | wedhorn | doesn't matter, if it's still ringing, the digits are directed at that sub, if the call hangs up, the device no longer sends the digits at all |
| 03:52.44 | snuff-work | i see what u mean |
| 03:53.13 | snuff-work | even after the call goes away.. still can't hit digits on phone connected to vm |
| 03:55.06 | snuff-work | test |
| 04:00.19 | *** join/#asterisk-dev snuff-work (~snuffy@210.8.167.13) |
| 04:00.19 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 04:00.38 | wedhorn | must smell |
| 04:00.56 | snuff-work | i see what u mean.. after that ringing call no can has dtmf |
| 04:00.56 | wedhorn | everytime I talk to someone today, they up and disappear ... |
| 04:01.45 | wedhorn | yep, but no idea why. Assume ccm must send something to control what sends the digits |
| 04:02.06 | snuff-work | my pidgin was going flakey.. had to restart it as wasn't sure i was even actually getting msgs |
| 04:02.10 | wedhorn | really need a pcap of two calls to the same device at the same time |
| 04:02.35 | wedhorn | wasn't having a go at you, just having a rough day ;) |
| 04:03.00 | snuff-work | didn't take it like that.. no just stating the fact =) |
| 04:04.20 | wedhorn | anyway, if something is sent for each call, it could be something when the device is registered (anyway, a v6 7960 doesn't have this issue) |
| 04:04.57 | snuff-work | mm.. well i may be able to get a test done soonish probably tomorrow |
| 04:06.36 | wedhorn | okay, cool. I've most of a patch to transfer a call before answering (even works while on another call without putting the active call on hold). Seems to work apart from the issue above |
| 04:08.39 | snuff-work | nice |
| 04:35.26 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 04:39.57 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.123) |
| 05:35.16 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 05:40.40 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.198) |
| 06:24.11 | *** join/#asterisk-dev oej (~Adium@h87-96-134-129.dynamic.se.alltele.net) |
| 06:24.11 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 06:35.19 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 06:40.17 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.220) |
| 06:47.41 | *** join/#asterisk-dev oej (~Adium@h87-96-134-129.dynamic.se.alltele.net) |
| 06:47.41 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 07:02.58 | snuff-work | small question.. how are you meant to reload pjsip.conf ? |
| 07:09.01 | snuff-work | minor bug with the 'converter' from sip -> pjsip.. seems like it cares alot about the 'order' things are in the file |
| 07:12.40 | snuff-work | so if ur 'identify' is above 'endpoint' = fail |
| 07:35.24 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 07:39.58 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.148) |
| 08:03.33 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.148) |
| 08:08.16 | *** join/#asterisk-dev bulkorok (~Benjamin@85.183.61.47) |
| 08:20.28 | *** join/#asterisk-dev oej (~Adium@h87-96-134-129.dynamic.se.alltele.net) |
| 08:20.28 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 08:35.15 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 08:35.51 | skrusty | morning |
| 08:50.13 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.4) |
| 09:12.25 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.4) |
| 09:27.57 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 09:27.57 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 09:35.18 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 09:38.50 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.125) |
| 10:00.05 | *** join/#asterisk-dev danjenkins (~dan@cpc65687-folk2-2-0-cust207.1-2.cable.virginmedia.com) |
| 10:06.43 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.125) |
| 10:15.17 | *** join/#asterisk-dev vlad_starkov (~vlad_star@94.228.255.31) |
| 10:16.02 | *** join/#asterisk-dev jmls (~julian@77.107.171.82) |
| 10:16.11 | jmls | good morrow |
| 10:35.18 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 10:38.31 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.139) |
| 11:35.22 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 12:15.22 | *** join/#asterisk-dev pc-m (~pcm@modemcable094.94-70-69.static.videotron.ca) |
| 12:22.49 | *** join/#asterisk-dev roramirez (~rodrigo@pc-34-179-160-190.cm.vtr.net) |
| 12:35.18 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 12:58.50 | *** join/#asterisk-dev drmessano (~nonya@pdpc/supporter/active/drmessano) |
| 13:02.35 | *** join/#asterisk-dev drmessano (~nonya@pdpc/supporter/active/drmessano) |
| 13:33.17 | *** join/#asterisk-dev drmessano (~nonya@pdpc/supporter/active/drmessano) |
| 13:35.19 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 13:35.39 | *** join/#asterisk-dev vlad_starkov (~vlad_star@212.34.52.66) |
| 13:42.38 | *** join/#asterisk-dev drmessano (~nonya@pdpc/supporter/active/drmessano) |
| 13:44.11 | file | whoops |
| 13:44.28 | *** join/#asterisk-dev drmessano (~nonya@pdpc/supporter/active/drmessano) |
| 13:46.16 | *** join/#asterisk-dev mjordan (~matt@nat/digium/x-qygmpuvkgyelofys) |
| 13:46.16 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 13:57.03 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 14:08.00 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 14:20.26 | *** join/#asterisk-dev gtjoseph (~gtjoseph@unaffiliated/gtj) |
| 14:30.57 | *** join/#asterisk-dev chazzam (~chazz@donutokyo.info) |
| 14:37.53 | *** join/#asterisk-dev sruffell (~sruffell@asterisk/the-kernel-guy/sruffell) |
| 14:37.53 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 14:40.24 | *** join/#asterisk-dev ipengineer (~zconkle@static-71-252-134-63.dllstx.fios.verizon.net) |
| 15:10.53 | *** join/#asterisk-dev kharwell (~kharwell@nat/digium/x-ozfurvlqxbohthzz) |
| 15:11.36 | *** join/#asterisk-dev zigg (~matt@unaffiliated/zigg) |
| 15:11.39 | *** part/#asterisk-dev zigg (~matt@unaffiliated/zigg) |
| 15:13.26 | *** join/#asterisk-dev putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 15:13.26 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 15:15.04 | *** join/#asterisk-dev ipengineer_ (~zconkle@static-71-252-134-63.dllstx.fios.verizon.net) |
| 15:29.22 | gtjoseph | putnopvut or file: res_pjsip opens sorcery twice, once in config_system, then again in pjsip_configuration (which is the one used by all other pjsip modules). Any reason for 2? Seems to work fine if they all use the forst one opened by config_system. |
| 15:32.28 | file | points at putnopvut |
| 15:33.28 | gtjoseph | i'm thinking further on ASTERISK-22537 |
| 15:34.37 | file | nods |
| 15:38.31 | file | I think he had a reason... but I can't remember what it was |
| 15:39.01 | opticron | system options are not reloadable |
| 15:39.16 | opticron | could that be why? |
| 15:39.40 | gtjoseph | neither is transport but they share the second instance |
| 15:42.54 | file | ah I know why it's done there specifically... because it has to be done before initializing PJSIP... but may not have thought about just keeping the sorcery instance around |
| 15:43.06 | putnopvut | Yeah, I'm thinking back on the sorcery instance used for system options and trying to remember exactly what the deal was with that. My initial recollection was just as opticron said, that system options should not be reloadable. At the time of creation, I don't know that sorcery allowed for specific options to be marked as "not reloadable' |
| 15:44.12 | mjordan | The config framework lets you do stuff like that, but I'm not sure sorcery exposes that level of detail. |
| 15:44.37 | mjordan | it doesn't really mark things as 'not reloadable', but you can carry over run-time data and ignore changes in configuration if you want to |
| 15:45.38 | gtjoseph | Hmmm. transport uses the same instance from pjsip_configuration it's not reloadable (you get a warning when you reload res_pjsip) |
| 15:47.03 | gtjoseph | ast_sorcery_object_register_no_reload |
| 15:47.49 | gtjoseph | maybe just change config_system to use ast_sorcery_object_register_no_reload instead of ast_sorcery_object_register |
| 15:48.52 | putnopvut | Yeah, I don't know if that existed at the time that the system config options were being created. I have no qualms about changing to use that instead of a separate sorcery instance. |
| 15:49.21 | file | ditto |
| 15:50.03 | gtjoseph | gotcha. it'll help things out if a module only gets 1 sorcery instance. |
| 16:27.53 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 16:38.58 | *** join/#asterisk-dev vlad_starkov (~vlad_star@212.34.52.66) |
| 16:42.51 | mjordan | putnopvut: file: opticron: kharwell: anyone else - can you guys look at the CHANGES/UPGRADE file in 12 and double check that we aren't missing anything else? |
| 16:43.41 | file | I shall do so whilst eating lunch |
| 16:45.58 | putnopvut | mjordan: sure thing |
| 16:48.06 | kharwell | looks |
| 16:48.43 | gtjoseph | Any chance of a commit of https://reviewboard.asterisk.org/r/3104 (PJSIP CLI Part 2) this week? |
| 16:48.47 | putnopvut | mjordan: security events logged over AMI is a new thing for 12.1, I believe. |
| 16:49.18 | *** join/#asterisk-dev otherwiseguy (~otherwise@2605:a601:43e:3501:3e97:eff:fe14:ef66) |
| 16:54.41 | file | hum |
| 16:54.49 | putnopvut | hums |
| 16:55.43 | *** join/#asterisk-dev Pullphinger (~Pullphing@c-24-13-69-42.hsd1.il.comcast.net) |
| 16:56.05 | file | mjordan, the HTTP server now supports chunked Transfer-Encoding |
| 16:56.36 | mjordan | file: thanks |
| 16:57.21 | *** join/#asterisk-dev drmessano (~nonya@pdpc/supporter/active/drmessano) |
| 16:57.43 | mjordan | did we put the security AMI events in manager itself? |
| 16:58.04 | putnopvut | mjordan: the code got put into main/security_events.c |
| 16:58.23 | mjordan | ah. And AMI just implicitly raises them. |
| 16:58.29 | putnopvut | Yeah. |
| 16:58.48 | putnopvut | But the idea that there's a "security" read permission now is new, along with the events you'll get. |
| 17:00.09 | mjordan | hm. |
| 17:00.13 | mjordan | We didn't document the events. |
| 17:00.37 | putnopvut | crud |
| 17:01.33 | *** part/#asterisk-dev bulkorok (~Benjamin@85.183.61.47) |
| 17:01.38 | mjordan | opticron: jrose_atDigium: kharwell: if you guys can, look back through your changes and make sure everything is documented on the wiki, in CHANGES, and UPGRADE.txt. We really can't make an RC without having everything well documented. |
| 17:02.04 | jrose_atDigium | Sure thing |
| 17:05.34 | mjordan | I'll work on a patch for documenting the security events |
| 17:12.04 | opticron | ok, on it |
| 17:14.12 | kharwell | mjordan: only thing that is not on the wiki are the realtime db changes, but there is currently no place keeping track of realtime table structures on the wiki that I can see |
| 17:15.20 | mjordan | jrose_atDigium: is there any reason why we opted to not add the optional IEs in the AMI security events? |
| 17:15.42 | mjordan | kharwell: yeah, that's okay. We don't have an automated mechanism to turn SQL into wiki |
| 17:21.00 | putnopvut | Speaking of realtime stuff, I have an alembic question. If I'm understanding how alembic works, it will simply run whatever your upgrade or downgrade scripts tell it to. Since we've bundled all of our realtime tables into a single upgrade path, it means that if you're using alembic, there's no way to only create a subset of database tables supported by Asterisk. Is that correct? |
| 17:21.23 | mjordan | pretty much, with the exception of voicemail. |
| 17:21.26 | mjordan | :-\ |
| 17:21.37 | putnopvut | So if I only wanted, say, sippeers, then using alembic means I'll get all the realtime tables created and have to delete them myself if I don't want them. |
| 17:22.03 | mjordan | I'm not sure if there's a way to say 'only upgrade this table' |
| 17:22.15 | putnopvut | I think we'd have had to construct our upgrade paths to work that way. |
| 17:22.34 | putnopvut | So we'd have separate upgrade paths for each DB table. |
| 17:23.09 | putnopvut | Then we could provide a script of our own that could be used to wrap multiple calls to alembic to upgrade/downgrade whichever table people care about. |
| 17:23.36 | pabelanger | then you'd need to run multiple alembic version tables in your database |
| 17:23.47 | putnopvut | Yep. |
| 17:23.57 | pabelanger | you might as well have different databases then |
| 17:24.01 | pabelanger | 1 for each table |
| 17:24.07 | mjordan | I wouldn't go that far |
| 17:24.29 | mjordan | The assumption we're making is that if you run realtime, you want to have the schema for all tables, since you might run realtime for any of the modules |
| 17:24.37 | pabelanger | I don't understand why you'd want to rev up a specific table only |
| 17:24.40 | mjordan | it's not a terrible assumption, but it's also highly unlikely to always be true |
| 17:24.44 | *** join/#asterisk-dev newtonr (~newtonr@nat/digium/x-nkugdopdnimsoxmm) |
| 17:24.45 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 17:24.55 | mjordan | it's less about rev'ing a particular table, and more about now creating a bunch of tables for stuff you won't use |
| 17:25.00 | mjordan | s/now/not |
| 17:25.03 | pabelanger | Oh, right |
| 17:25.11 | pabelanger | I see what you are saying |
| 17:26.31 | putnopvut | Asterisk is an interesting case where people don't necessarily need the entire set of tables provided by realtime. They may only wish to have sippeers but have static queue members, or some other such thing. It's not a case where the Asterisk software relies on a certain database schema in order to operate. |
| 17:26.54 | putnopvut | That was probably a terrible way of wording what I was trying to say. |
| 17:27.09 | putnopvut | But I've gone on record before as saying that I hate typing conversations in real-time :) |
| 17:28.08 | mjordan | nice to see we have some security events that can never be raised... |
| 17:28.22 | pabelanger | Ya, your point is valid. The only way to do it right would be per module alembic scripts, tracking separate versioning for it. Seems doable, but could get complicated fast |
| 17:31.00 | leifmadsen | it does almost seem easier to manage though by providing all the things, then letting the operator create another alembic script or something to delete the unneeded tables |
| 17:31.10 | leifmadsen | rather than taking on all-the-iterations |
| 17:33.40 | mjordan | jrose_atDigium: did you get the question previously? |
| 17:37.07 | opticron | I only found the one item that wasn't listed |
| 17:37.29 | mjordan | opticron: which was? |
| 17:38.05 | gtjoseph | file: did you get bitten by a hashtable when you were a kid or something. :) |
| 17:38.32 | mjordan | HASH ALL THE THINGS |
| 17:38.37 | putnopvut | gtjoseph: he's been locked in the land of performance recently, and there are certain things that just raise red flags for him now :) |
| 17:39.02 | file | not exactly red flags |
| 17:39.09 | putnopvut | orange flags? |
| 17:39.19 | file | hashtabs are from pre-ao2 times and I don't want to encourage their use |
| 17:39.31 | putnopvut | oh, hashtab in particular |
| 17:39.39 | putnopvut | I was thinking of astobj2 hash tables :) |
| 17:40.06 | file | hash tables have their use. |
| 17:40.17 | gtjoseph | asterisk needs some @deprecated annotation mechanism. |
| 17:41.14 | mjordan | well |
| 17:41.19 | mjordan | we actually have that |
| 17:41.20 | mjordan | :-D |
| 17:41.28 | mjordan | do you mean on C APIs, or configuration values? |
| 17:41.35 | mjordan | or entire modules? |
| 17:41.42 | gtjoseph | C APIs |
| 17:42.08 | mjordan | http://www.stack.nl/~dimitri/doxygen/manual/commands.html#cmddeprecated |
| 17:43.34 | file | groans as he gets to codec_* |
| 17:43.42 | putnopvut | There's also this: http://gcc.gnu.org/onlinedocs/gcc-3.2.3/gcc/Type-Attributes.html |
| 17:43.59 | opticron | mjordan, the change in subscription behavior for device states, already added to UPGRADE.txt |
| 17:44.09 | putnopvut | There's a deprecated attribute you can set on stuff. But I think that's more for if you provide a library to be used externally. |
| 17:44.47 | mjordan | yeah |
| 17:44.55 | mjordan | it will throw a warning if it gets used |
| 17:45.05 | mjordan | we should probably put that on some of our channel.h functions :-) |
| 17:45.22 | gtjoseph | I need something that causes eclipse to flash in red letters if you try to use a deprecated api. :) |
| 17:45.54 | file | Eclipse. How... nice. |
| 17:45.57 | mjordan | hey now |
| 17:46.09 | mjordan | some of us used it for quite a long time, before we got tired of the CDT puking all over chan_sip |
| 17:46.29 | gtjoseph | I guess I can go back to 'ed' |
| 17:47.01 | coreyf1513 | I still use eclipse from time to time for call graphs |
| 17:48.27 | mjordan | Won't lie: I like the Subversion plugin for browsing history |
| 17:48.39 | mjordan | and I do miss "Find all references" |
| 17:50.20 | gtjoseph | I used to do "find all references" by running my thumb over the top edges of a punch card stack. |
| 17:51.15 | file | takes a brief siesta and reads translate.h some |
| 17:51.58 | file | mjordan, you should rewrite this over a weekend |
| 17:52.20 | file | preferably before I get to translate.c |
| 17:52.52 | mjordan | gtjoseph: my favorite punch card joke - http://www.criticalcommons.org/Members/ccManager/clips/the-simpsons-on-immigration-and-technological/view |
| 17:52.59 | mjordan | file: meh |
| 17:53.12 | file | harsh |
| 17:53.39 | mjordan | there's a lot of dragons in there |
| 17:56.36 | gtjoseph | LOL |
| 18:09.20 | *** join/#asterisk-dev danjenkins (~dan@cpc65687-folk2-2-0-cust207.1-2.cable.virginmedia.com) |
| 18:14.48 | *** join/#asterisk-dev danjenkins (~dan@cpc65687-folk2-2-0-cust207.1-2.cable.virginmedia.com) |
| 18:23.07 | mjordan | jrose_atDigium: your commit in r407419 isn't going to make much sense to end users |
| 18:24.18 | mjordan | jrose_atDigium: a Bridge object isn't a thing that they can manipulate/be aware of. ARI has a bridges resource that has operations on it, but 'creator' isn't a query parameter in those operations. The events may have that field, but in that case, the affected events should be noted in the change |
| 18:24.31 | mjordan | jrose_atDigium: and that change also affected all of the Bridge related AMI events |
| 18:24.46 | jrose_atDigium | uwS8Rp8 |
| 18:24.55 | mjordan | I hope that's not a password |
| 18:24.55 | jrose_atDigium | damnit. |
| 18:24.59 | mjordan | :-) |
| 18:25.09 | mjordan | ducks and hides |
| 18:28.03 | jrose_atDigium | It's the bridge snapshot JSON representations that get the fields. That's what I meant by object. I'll adjust it to be more specific... and add AMI, but it affects all events containing bridge snapshots, so I'll have to figure out what all of those are -_- |
| 18:29.09 | tzafrir | http://xkcd.com/936/ - if you accidentally paste "correct horse battery staple" to the channel, people may not suspect what it is |
| 18:29.40 | mjordan | jrose_atDigium: welcome to the wonderful world of correctly documenting features |
| 18:30.46 | mjordan | tzafrir: +1 |
| 18:37.42 | jrose_atDigium | diving emu rapid waffle - bonus points for being all trochees |
| 18:38.23 | file | nobody mind my replicating media formats branches... |
| 18:38.35 | mjordan | file: I stopped awhile back |
| 18:39.13 | file | I'm doing something crazy and breaking my work up into multiple reviews as things progress, with them being merged into a staging branch |
| 18:40.07 | file | CRAZY I SAY, CRAZY |
| 18:40.53 | mjordan | yay! We might be able to review it and not go mad |
| 18:41.00 | file | that's my plan |
| 18:54.57 | kharwell | I have a sorcery observer registered on pjsip endpoints. endpoints are configured in realtime. I add a endpoint to the table. When does the observer created callback get called? |
| 18:57.30 | putnopvut | kharwell: probably whenever sorcery actually creates the endpoint structure, which I guess would be on the next lookup of that endpoint. (note, I have not checked the code for this) |
| 18:58.08 | kharwell | yeah that's what I figured and what I am looking at in the code, but my callback doesn't seem to get called |
| 18:59.11 | kharwell | it is there though as I can list it with pjsip show endpoints |
| 19:00.11 | putnopvut | Is a different observer callback being called instead? |
| 19:01.10 | putnopvut | Like maybe loaded()? |
| 19:01.13 | kharwell | what do you mean? like it is calling updated instead of created? none are currently being called |
| 19:01.23 | putnopvut | Yeah, that's what I was meaning. |
| 19:01.28 | file | created and updated are only called if created and updated within Asterisk |
| 19:02.54 | kharwell | so there is no way to be notified if an object is created/updated in a database? |
| 19:03.21 | file | nope, because doing so would either require 1. Constantly querying the database looking for changes or 2. A notification mechanism from the database itself |
| 19:04.30 | kharwell | hmm yeah I guess if you waited to notify until it was created in asterisk it wouldn't be in "realtime" necessarily |
| 19:17.23 | *** join/#asterisk-dev danjenkins (~dan@cpc65687-folk2-2-0-cust207.1-2.cable.virginmedia.com) |
| 19:18.22 | *** join/#asterisk-dev snuff-work (~snuffy@210.8.167.13) |
| 19:18.22 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 19:46.09 | *** join/#asterisk-dev ipengineer (~zconkle@static-71-252-134-64.dllstx.fios.verizon.net) |
| 20:05.08 | kharwell | file: just curious - when I set up pjsip endpoints for realtime I still had one configured in pjsip.conf, but that didn't seem to get loaded. Is it meant to be one or the other? |
| 20:05.14 | *** join/#asterisk-dev oej (~Adium@h87-96-134-129.dynamic.se.alltele.net) |
| 20:05.14 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 20:05.44 | file | you can have multiple but configuration overrules the default built-in so you need to add it in order of preference to the sorcery.conf file |
| 20:06.30 | file | config,pjsip.conf,criteria=type=endpoint |
| 20:06.41 | kharwell | okay so I didn't have the sip.conf even specified in sorcery.conf so it just used the realtime one |
| 20:06.52 | file | (yes, that needs to be documented) |
| 20:07.13 | file | kharwell, correct |
| 20:07.39 | kharwell | but it auto defaults to pjsip.conf for other non realtime types |
| 20:08.03 | kharwell | for instance I had aors in pjsip.conf, but endpoints in realtime and it seem to pick that stuff up |
| 20:08.08 | file | correct |
| 20:08.18 | kharwell | cool thanks! |
| 20:08.24 | file | when you define an object type you can also specify a default configuration |
| 20:08.49 | file | the PJSIP ones are the deployment model that makes the most sense (almost everything in configuration files, with contacts being persisted to the astdb) |
| 20:09.25 | *** join/#asterisk-dev ipengineer (~zconkle@static-71-252-134-64.dllstx.fios.verizon.net) |
| 20:11.23 | kharwell | ah yes I think I have noticed that in various places (pjsip_configuration for instance) |
| 20:11.33 | file | yup |
| 20:11.59 | kharwell | pretty cool so if I wanted to write out to my own back end then I just need to create a wizard for it and configure in sorcery.conf? |
| 20:12.08 | file | yes |
| 20:12.48 | file | the idea of being able to programatically build a hierarchy of wizards has crossed my mind too... |
| 20:13.29 | file | hasn't bubbled to the surface though |
| 20:14.52 | kharwell | carbonates file |
| 20:15.26 | file | nobody has done it yet but that would potentially allow some clever stuff... which I will elaborate on in a future blog post |
| 20:21.21 | *** join/#asterisk-dev oej (~Adium@2.68.105.202.mobile.tre.se) |
| 20:21.21 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 20:28.30 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 20:50.27 | mjordan | btw: the 'how do I make it pull endpoints from realtime/static conf file' should probably be put up on the wiki |
| 21:02.53 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.192) |
| 21:10.09 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.192) |
| 21:18.07 | *** join/#asterisk-dev vlad_starkov (~vlad_star@91.206.59.136) |
| 21:40.02 | *** join/#asterisk-dev danjenkins (~dan@cpc65687-folk2-2-0-cust207.1-2.cable.virginmedia.com) |
| 22:00.42 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@64.235.216.2) |
| 22:02.25 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-28-58.cablep.bezeqint.net) |
| 22:02.25 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 22:18.10 | *** join/#asterisk-dev vlad_sta_ (~vlad_star@109.188.126.192) |
| 23:04.29 | *** join/#asterisk-dev roderickm (~roderickm@67.63.143.254) |
| 23:11.30 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.126.192) |
| 23:23.27 | seanbright | potentially dumb question to solidify my understanding of audiohooks |
| 23:24.17 | seanbright | there is a SIP channel A that is G729 and a SIP channel B that is also G729 |
| 23:24.32 | seanbright | a third channel, C, does a chanspy on channel A |
| 23:24.47 | seanbright | or a mixmonitor or something else that uses an audiohook |
| 23:25.07 | seanbright | then the audio will be transcoded to slinear for mixing |
| 23:25.22 | seanbright | if at the same time, another channel, D connects and also does a chanspy on channel A |
| 23:25.39 | file | the answer is once |
| 23:25.45 | seanbright | ok |
| 23:28.41 | seanbright | so the slinear factory is shared between hooks? |
| 23:29.15 | seanbright | or because the hooks are invoked serially they pick up work that was done before hand? |
| 23:30.12 | file | each hook doesn't invoke translation, the process at the start/finish does the translation |
| 23:30.57 | seanbright | so a follow up... |
| 23:31.20 | seanbright | is there a way to tell asterisk that you always want channels to be transcoded on the way in/out? |
| 23:31.25 | seanbright | even if they could be natively bridged? |
| 23:31.59 | file | not afaik |
| 23:32.00 | mjordan | hm. |
| 23:32.18 | seanbright | ok |
| 23:32.21 | mjordan | seanbright: what are you thinking of where that would be useful? |
| 23:32.25 | file | you maybe could do things to force the behavior |
| 23:32.42 | mjordan | if toss a hook on it, that would force it to not natively bridge (I think) |
| 23:32.55 | file | correct |
| 23:33.12 | seanbright | so VOLUME with a gain of 0? |
| 23:33.29 | file | that would do it |
| 23:33.32 | seanbright | we're having an issue in 11 when we spy on a channel that is involved in a bridge |
| 23:33.38 | mjordan | heh - func_jitterbuffer |
| 23:33.58 | seanbright | it causes the audio for the bridged parties to go to shit |
| 23:33.59 | mjordan | but VOLUME is better, yeah |
| 23:34.15 | file | interestin' |
| 23:34.52 | seanbright | and my assumption is that because audiohooks "get in the way" of frame handling, that the transcoding is somehow slowing everything down |
| 23:34.54 | mjordan | well, the audio probably shouldn't go to shit regardless |
| 23:35.23 | mjordan | seanbright: does it happen in a controlled environment? |
| 23:35.29 | seanbright | just in production |
| 23:35.34 | seanbright | :) |
| 23:35.35 | mjordan | drat |
| 23:35.51 | seanbright | there are a lot of moving pieces of replicate |
| 23:36.02 | seanbright | and we don't have spare transcoder cards sitting around |
| 23:36.46 | seanbright | i am going to test tomorrow morning when no one is dialing to see if it happens with no load on the server |
| 23:37.13 | seanbright | oddly, i'm hoping that it does manifest that way |
| 23:37.21 | seanbright | so i can't just shrug it off as a load issue |
| 23:38.30 | mjordan | that's kind of what I hoped, honestly. Load sucks. Asterisk forgetting a translation path - that's something that can be fixed |
| 23:38.55 | mjordan | seanbright: just curious, what version of 11? |
| 23:39.00 | seanbright | 11M |
| 23:39.03 | seanbright | :-) |
| 23:39.10 | mjordan | but about what revision? |
| 23:39.17 | mjordan | and did you get the performance improvement thing? |
| 23:39.18 | seanbright | HEADish |
| 23:39.29 | seanbright | which performance improvement thing? |
| 23:39.51 | seanbright | i'm using the git mirror that tzafrir has set up |
| 23:40.13 | seanbright | r407103 |
| 23:40.31 | seanbright | + some of my own patches. nothing related to audio at all though. |
| 23:40.54 | mjordan | r405431 |
| 23:41.20 | seanbright | we're using codec_opus from that github repo that someone set up, and when i have a single channel over WS sitting in MusicOnHold, codec_opus has a usage count of 6 |
| 23:41.29 | mjordan | It wouldn't be audio related, but if you're hitting a lot of CPU, that might alleviate some of it |
| 23:41.30 | *** join/#asterisk-dev otherwiseguy (~otherwise@2605:a601:43e:3501:2677:3ff:fe8f:2730) |
| 23:41.40 | mjordan | basically, Asterisk would always format every single verbose message. All the time. |
| 23:41.46 | seanbright | ah |
| 23:41.50 | seanbright | yeah, we're newer than that. |
| 23:41.56 | mjordan | k. Nuts. |
| 23:42.35 | seanbright | what would be awesome is if spy hooks were run out of band |
| 23:42.41 | seanbright | i think that would go a long way |
| 23:43.16 | seanbright | i am going to try the VOLUME/JITTERBUFFER thing and see if that improves things at all |
| 23:43.49 | seanbright | we'll use more transcoders overall, but if it helps with the quality problems it will be worth it |
| 23:44.12 | *** part/#asterisk-dev mjordan (~matt@nat/digium/x-qygmpuvkgyelofys) |
| 23:44.15 | seanbright | kbye |
| 23:47.33 | file | I'm here. I'm always here. |
| 23:51.21 | snuff-work | yes we know we shan't forget you ;) |
| 23:51.53 | file | May the odds be ever in your favour. |
| 23:52.21 | file | seanbright, your comment about pushing it out of band is true... keeping the media path fast and short is nifty |
| 23:52.51 | snuff-work | i noticed with the sip to pjsip converter.. that if your definitions aren't in the right order pjsip throws its toys away |
| 23:53.52 | snuff-work | also.. how do u even 'reload' your pjsip config on the fly.. or is it not possible or just forgot to add a cli cmd ? |
| 23:54.02 | file | module reload res_pjsip.so |
| 23:54.40 | snuff-work | in my pjsip.conf.. if i had 'identify' is above 'endpoint' = fail |
| 23:55.04 | file | weird |
| 23:55.24 | snuff-work | get msg like |
| 23:55.25 | snuff-work | [2014-02-05 17:38:09] WARNING[21523]: res_pjsip_endpoint_identifier_ip.c:144 ip_identify: Identify section 'syd-in' points to endpoint 'syd-in' but endpoint could not be looked up |
| 23:55.58 | file | that would mean the endpoint wasn't loaded... |
| 23:56.02 | file | did anything show up on the console? |
| 23:56.06 | file | when the module was loaded that is |
| 23:57.33 | snuff-work | mm.. seems so |
| 23:57.41 | snuff-work | [2014-02-05 17:37:54] ERROR[21503]: config_options.c:719 aco_process_var: Error parsing tos_audio=ef at line 97 of |
| 23:57.42 | snuff-work | [2014-02-05 17:37:54] ERROR[21503]: res_sorcery_config.c:275 sorcery_config_internal_load: Could not create an object of type 'endpoint' with id 'syd-in' from configuration file 'pjsip.conf' |
| 23:58.10 | gtjoseph | tos symbolics don't work yet |
| 23:58.36 | gtjoseph | it's on my list for this week. :) |
| 23:59.30 | snuff-work | but would be nicer if it just 'skipped' conf vars it didn't know rather than bailing on making it altogether ;( |
| 23:59.44 | file | it won't allow you to have a broken configuration |