| 05:47.44 | *** join/#asterisk-dev _Vile (~vile@freeswitch/developer/vile) |
| 06:51.09 | *** join/#asterisk-dev jkroon (~jkroon@165.16.203.57) |
| 07:03.52 | *** join/#asterisk-dev tsearle (~tsearle@37.19.10.33) |
| 07:03.52 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
| 07:13.46 | *** join/#asterisk-dev _Vile (~vile@freeswitch/developer/vile) |
| 07:23.26 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:b749:768d:ed77:7c9a) |
| 07:23.29 | *** join/#asterisk-dev pchero_work (~pchero@87.213.240.121) |
| 09:05.08 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 10:53.10 | *** join/#asterisk-dev jkroon (~jkroon@41.114.9.2) |
| 11:32.58 | *** join/#asterisk-dev jkroon (~jkroon@41.114.57.104) |
| 12:22.29 | *** join/#asterisk-dev jkroon (~jkroon@41.114.56.247) |
| 12:34.31 | *** join/#asterisk-dev pc-m (~pcm@modemcable094.94-70-69.static.videotron.ca) |
| 13:37.25 | *** join/#asterisk-dev CELYA_1 (~Thunderbi@lstlambert-657-1-90-73.w80-13.abo.wanadoo.fr) |
| 14:13.14 | *** join/#asterisk-dev _Vile (~vile@freeswitch/developer/vile) |
| 14:34.55 | *** join/#asterisk-dev cresl1n (uid299068@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 14:34.55 | *** mode/#asterisk-dev [+o cresl1n] by ChanServ |
| 15:16.15 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-iuvyrcdeupzfswfb) |
| 15:16.15 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 15:29.52 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-nvxlouuqfnmbtuub) |
| 15:29.52 | *** mode/#asterisk-dev [+o rmudgett] by ChanServ |
| 15:32.04 | *** join/#asterisk-dev bford (uid283514@gateway/web/irccloud.com/x-anqaqqtrscgegqsf) |
| 15:32.04 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 16:06.23 | *** join/#asterisk-dev _Vile (~vile@freeswitch/developer/vile) |
| 16:44.16 | gtjoseph | coreyfarrell: re "module reload sounds http"... i just tested that and I get No such module 'sounds' |
| 16:44.16 | gtjoseph | Module 'http' reloaded successfully. |
| 16:47.01 | coreyfarrell | oh I misread the code, saw the break and associated it with the loop instead of the switch. so please remove the .reload callback from 16 and master, no need for special note in CHANGES. |
| 16:47.23 | coreyfarrell | gtjoseph: misread the cli.c code that is. |
| 16:47.51 | gtjoseph | so what do you want me to remove? |
| 16:48.31 | coreyfarrell | https://gerrit.asterisk.org/#/c/asterisk/+/10924/1/main/sounds.c@341 and line 340, then delete the reload_module function. |
| 16:49.01 | coreyfarrell | that `.reload = reload_module` is what registers 'module reload sounds' in 16 and master. |
| 16:49.07 | gtjoseph | oh i gotcha. i forgot that 16+ was different. |
| 16:50.52 | coreyfarrell | so really the comments I left on 13 are invalid.. but the findings on master re: ao2_cleanup(variants) and if (failed_index) do apply to 13 so the -1 stands. |
| 16:53.44 | *** join/#asterisk-dev jkroon (~jkroon@165.16.204.40) |
| 17:37.03 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:b749:768d:ed77:7c9a) |
| 17:42.21 | coreyfarrell | is the cmp_fn for ao2_container_alloc_list optional for a container that will only be used by an ao2_iterator? |
| 17:49.15 | coreyfarrell | gtjoseph: I just gave +1 on your media_index changes.. made a comment on 16 and master that you can either act on or ignore. |
| 17:57.01 | rmudgett | coreyfarrell: The cmp_fn is only used by ao2_find() |
| 19:46.06 | seanbright | does this appear correct? |
| 19:46.07 | seanbright | [res_pjsip] |
| 19:46.07 | seanbright | endpoint = config,pjsip.conf,criteria=type=endpoint |
| 19:46.07 | seanbright | endpoint/cache = memory_cache,object_lifetime_maximum=3000 |
| 19:46.08 | seanbright | endpoint = realtime,ps_endpoints |
| 19:46.40 | seanbright | because it keeps going to realtime for the same endpoint, which is confusing to me |
| 19:48.55 | file | cache before endpoint |
| 19:48.59 | file | order matters |
| 19:49.17 | file | like, right up at the top... |
| 19:49.27 | file | otherwise it should be okay... I think |
| 19:49.57 | seanbright | ok |
| 19:56.32 | seanbright | i'm confused |
| 19:56.45 | seanbright | sorcery_memory_cache_retrieve_id |
| 19:56.54 | seanbright | it doesn't appear to actually put anything in cache? |
| 19:57.38 | file | sorcery is responsible for doing that |
| 19:59.12 | gtjoseph | seanbright: look in sorcery.c ast_sorcery_create, update, et al. |
| 19:59.46 | seanbright | yeah, i'm not using create or anything |
| 19:59.51 | seanbright | just retrieve_by_id |
| 20:00.21 | file | retrieve_by_id in sorcery.c has functionality that after lookup it will then call into the caches to create it |
| 20:00.44 | seanbright | yeah, i'll try to figure out why that is completely broken on my system |
| 20:05.16 | seanbright | meh, i give up |
| 20:05.22 | seanbright | something is not correct with sorcery caching |
| 20:05.28 | seanbright | you're welcome |
| 20:05.31 | gtjoseph | did you read the wiki page on caching? |
| 20:05.31 | seanbright | :P |
| 20:05.54 | gtjoseph | was that a "no"? |
| 20:06.05 | seanbright | no, it was a continutation of my previous statement |
| 20:06.13 | seanbright | if you scroll up 20 lines, you will see the config i used |
| 20:06.20 | seanbright | which is almost verbatim from the wiki |
| 20:06.45 | seanbright | potentially dumb question - does [res_pjsip] apply to res_pjsip_*? |
| 20:06.55 | gtjoseph | nope |
| 20:07.01 | seanbright | well there ya go |
| 20:07.11 | gtjoseph | it applyies to the object res_pjsip exposes |
| 20:07.48 | seanbright | that sentence does not make sense |
| 20:07.52 | seanbright | at least not to me |
| 20:08.01 | seanbright | how about this: forget it, i just won't cache |
| 20:08.17 | gtjoseph | so if you want to cache "identify" you need to put that in a section named res_pjsip_endpoint_identifier_ip |
| 20:08.30 | gtjoseph | same for outbound registration |
| 20:08.42 | gtjoseph | the section needs to be res_pjsip_outbound_registration |
| 20:08.58 | seanbright | then reading the wiki would be a complete waste of time |
| 20:09.08 | seanbright | because it doesn't show that |
| 20:09.11 | seanbright | everything is [res_pjsip] |
| 20:09.27 | seanbright | but i appreciate the condescending question :D |
| 20:12.00 | seanbright | so if something inside res_pjsip.so calls retrieve_by_id("endpoint", "foo"); |
| 20:12.13 | gtjoseph | it just so happens that the wiki page examples all use res_pjsip base objects. |
| 20:12.18 | seanbright | and something else inside res_pjsip_something_else.so calls retrieve_by_id("endpoint", "foo"); |
| 20:12.21 | seanbright | the cache will not work? |
| 20:12.32 | gtjoseph | it should |
| 20:12.34 | seanbright | well, not that it won't work |
| 20:12.41 | seanbright | just that it won't share the objects between the two? |
| 20:13.04 | gtjoseph | the key is that the res_pjsip modules get the sorcery instance from res_pjsip |
| 20:13.16 | gtjoseph | ast_sip_get_sorcery() or something |
| 20:13.29 | gtjoseph | so they all work off the same instance. |
| 20:13.52 | gtjoseph | sorcery.conf controls objects modules expose, not use. |
| 20:14.34 | seanbright | ok, so i need to add a section to sorcery.conf for each module that starts with res_pjsip, correct? |
| 20:15.03 | gtjoseph | if it exposes object, yes. |
| 20:15.17 | gtjoseph | not for objects it uses. |
| 20:16.20 | gtjoseph | the sample sorcery.conf has better examples. |
| 20:16.26 | seanbright | not really |
| 20:16.31 | seanbright | unless i am looking at the wrong sample |
| 20:17.35 | gtjoseph | https://pastebin.com/xYtcYfci |
| 20:17.39 | gtjoseph | No? |
| 20:18.15 | gtjoseph | doesn't show registration but it does show identify, outbound-publish |
| 20:18.25 | seanbright | i don't need most of that nonsense |
| 20:18.33 | seanbright | just endpoint, auth, and aor |
| 20:18.52 | seanbright | and caching. preferably the kind that works. |
| 20:18.54 | gtjoseph | well then you only need a section for res_pjsip. |
| 20:19.00 | seanbright | but i'm over it. it doesn't work. |
| 20:19.18 | gtjoseph | try full_backend_cache |
| 20:19.31 | seanbright | no |
| 20:19.44 | gtjoseph | ok |
| 20:20.18 | seanbright | my endpoints are created/destroyed dynamically and i am using res_config_curl |
| 20:20.27 | seanbright | so i don't bother implementing the "list" realtime operations |
| 20:20.32 | seanbright | which i assume the full backend cache option does |
| 20:20.56 | gtjoseph | that's what i was getting at. |
| 20:21.21 | seanbright | what were you getting at? |
| 20:21.38 | seanbright | the memory cache requires the list ops? |
| 20:22.32 | gtjoseph | i don't remember. i didn't think so. |
| 20:32.46 | seanbright | i think i see the problem... maybe |
| 20:34.20 | seanbright | a mismatch between the id that is being retrieved and the one being cached |
| 20:35.03 | gtjoseph | something getting mangled? |
| 20:35.19 | seanbright | so the endpoint id is foo@bar.com |
| 20:35.27 | seanbright | the object getting put into cache has the id 'foo' |
| 20:35.29 | seanbright | or so it seems |
| 20:35.37 | seanbright | i am still looking in to it |
| 20:38.03 | seanbright | this may be surprising, but i think it is my fault |
| 20:38.16 | seanbright | i know - i find it unbelievable as well |
| 20:38.29 | gtjoseph | :) |
| 20:44.41 | seanbright | turns out i'm a huge asshole. |
| 20:44.48 | seanbright | who'd of thunk it?! |
| 20:44.53 | gtjoseph | NEWS FLASH!!! |
| 20:45.18 | gtjoseph | can't be any worse than the stupid stuff i've done |
| 20:45.20 | seanbright | asterisk was asking for foo@bar.com |
| 20:45.27 | seanbright | i was returning an object whose id was 'foo' |
| 20:45.55 | seanbright | so whole not technically a bug, maybe retrieve_by_id should check that the id requested is the same as the id returned |
| 20:46.02 | seanbright | but that is secondary |
| 20:46.42 | file | maybe as a developer assertion... I'm not sure I'd want that in the path... as it means something has broken a contract... |
| 20:48.51 | seanbright | the alternative would be that res_sorcery_realtime throws away the returned id and uses the requested on? |
| 20:48.55 | seanbright | meh |
| 20:49.02 | seanbright | i already hate that idea |
| 20:53.00 | seanbright | is there a way in dialplan to get the name of the transport that a pjsip request came in on? |
| 20:53.56 | file | don't think so |
| 20:54.01 | seanbright | k |
| 20:54.02 | file | https://wiki.asterisk.org/wiki/display/AST/Asterisk+13+Function_CHANNEL |
| 20:54.03 | gtjoseph | can't think of one offhand unless you have symmetric transport turned on. |
| 20:54.13 | gtjoseph | ah |
| 20:54.30 | seanbright | thanks |
| 21:26.41 | seanbright | gtjoseph: sorry for being next-level douchey |
| 21:27.18 | *** join/#asterisk-dev _Vile (~vile@freeswitch/developer/vile) |
| 21:51.15 | *** join/#asterisk-dev _Vile (~vile@freeswitch/developer/vile) |
| 22:36.47 | *** join/#asterisk-dev pchero_work (~pchero@dhcp-077-249-058-090.chello.nl) |
| 22:43.36 | *** join/#asterisk-dev snuff-work (~snuff-wor@202-161-112-134.tpgi.com.au) |
| 22:43.36 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 22:48.05 | *** join/#asterisk-dev Deeewayne (~dwayne@2605:a601:a2a0:73f:89b0:669b:cccb:b0be) |
| 22:48.05 | *** mode/#asterisk-dev [+o Deeewayne] by ChanServ |
| 23:07.39 | *** join/#asterisk-dev pchero_work (~pchero@dhcp-077-249-058-090.chello.nl) |
| 23:15.33 | *** join/#asterisk-dev pchero_work (~pchero@dhcp-077-249-058-090.chello.nl) |