| 02:30.57 | *** join/#asterisk-dev infobot (ibot@208.53.50.136) |
| 02:30.57 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Check out our blog! blogs.asterisk.org -=- Follow on Twitter at @AsteriskDev |
| 06:01.01 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:6cfb:5947:bfca:be35) |
| 06:22.56 | *** join/#asterisk-dev sh_smith (~sh_smith@cpe-76-174-26-91.socal.res.rr.com) |
| 09:03.11 | *** join/#asterisk-dev jkroon (~jkroon@165.16.203.126) |
| 11:11.23 | seanbright | file: roger roger |
| 11:11.46 | file | Ack ack |
| 11:27.13 | seanbright | also: f*ck solaris |
| 11:27.15 | seanbright | :D |
| 11:33.10 | jkroon | seanbright, sorry about https://gerrit.asterisk.org/#/c/asterisk/+/10013/ |
| 11:36.16 | seanbright | jkroon: hmm? |
| 11:36.26 | seanbright | i know that is the problem |
| 11:36.28 | seanbright | i've reverted |
| 11:36.35 | seanbright | well, posted the revert for review |
| 11:39.53 | seanbright | jkroon: can you test a patch for me? |
| 11:40.01 | seanbright | re: ASTERISK-28048 |
| 11:43.50 | jkroon | revert will obviously fix it, is there an alternative fix perhaps? |
| 11:44.00 | seanbright | yes |
| 11:44.47 | seanbright | jkroon: http://paste.ubuntu.com/p/VQWNGpZ8VV/ |
| 11:45.50 | seanbright | kind of a big hammer. but should work. |
| 11:49.20 | file | ha |
| 11:51.26 | jkroon | ag no, that's a really crappy hammer. basically do what's currently being done by default? |
| 11:57.21 | seanbright | yes, it's a penalty for running the non-bundled version |
| 11:57.43 | off-nominal | begrudgingly installs pjproject |
| 12:03.41 | jkroon | lol, it's adequate performance wise currently, so I'm sure it'll work there too, distro's generally don't like bundled stuff. |
| 12:04.10 | jkroon | ok, using gerrit, how can I perform testing for you on my side? |
| 12:04.49 | off-nominal | don't bother. i'm changing it. |
| 12:04.56 | off-nominal | trying with a smaller hammer. |
| 12:04.59 | jkroon | still need to know *how* |
| 12:05.41 | off-nominal | well, the error occurs on compile, so you would just apply the patch and try to compile |
| 12:06.01 | off-nominal | maybe enable TEST_FRAMEWORK and run the unit tests too |
| 12:06.05 | jkroon | ie, i'd like to be able to test for you whatever you post. is there no way for me to use git via the gerrit remote to test posted patches? |
| 12:06.28 | off-nominal | git review -d <rev> |
| 12:06.31 | off-nominal | err |
| 12:06.34 | off-nominal | git review -d <review id> |
| 12:06.54 | jkroon | sorry - very limited experience with gerrit. |
| 12:06.57 | off-nominal | same |
| 12:07.07 | off-nominal | jkroon: https://wiki.asterisk.org/wiki/display/AST/Gerrit+Usage |
| 12:07.33 | off-nominal | file: pjproject 2.8 is released |
| 12:07.34 | off-nominal | time to upgrade |
| 12:09.34 | jkroon | we're currently using 2.7.2 but would probably be good to test against 2.8 as well. |
| 12:10.17 | file | off-nominal: it's in queue already |
| 12:10.58 | off-nominal | wonders of they changed sock.h... |
| 12:13.12 | jkroon | diff and see :) |
| 12:13.39 | jkroon | possibly (hopefully) to fix what's tripping us right now anyway. |
| 12:14.09 | file | the Solaris patch was given to them, but they didn't accept it if I recall - could be wrong, though |
| 12:14.32 | jkroon | we handed them some ./configure fixes too, i did not appreciate their attitude. |
| 12:20.04 | off-nominal | getting an unrelated error about 'pjsip_dlg_create_uas' when compiling against the non-bundled version |
| 12:20.13 | off-nominal | but that is above my pay grade |
| 12:20.33 | file | that can happen if there's a conflict on the system, two PJSIP installs for example |
| 12:20.42 | off-nominal | ah, ok cool |
| 12:26.01 | off-nominal | i would like to point out that i edited some trolling comments about solaris out of my commit message |
| 12:26.05 | off-nominal | incredible restraint |
| 12:26.30 | file | :D |
| 12:26.55 | off-nominal | but it DID say "(to the 4 of you running Asterisk on Solaris - you're welcome)" |
| 12:27.04 | off-nominal | so we can just enjoy that here |
| 12:28.15 | off-nominal | jkroon: git review -d 10097 |
| 12:29.44 | *** join/#asterisk-dev bford (uid283514@gateway/web/irccloud.com/x-oqikytoytwwaegoy) |
| 12:29.44 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 12:31.09 | off-nominal | i do like pjproject, but the level of abstraction is out of hand in places |
| 12:32.26 | jkroon | haven't worked with pjproject, but from what I've seen it looks pretty good. |
| 12:32.57 | file | it originates from olden times and targets many things, including embedded |
| 12:33.00 | file | which is a different world |
| 12:33.35 | jkroon | however if you tell me that you have no interested in fixing ./configure bugs and that the current behaviour is good enough and you don't care about fixing --with-ssl vs --without-ssl then I get somewhat annoyed. |
| 12:33.46 | off-nominal | but you would think that posix would be the thinnest layer |
| 12:33.54 | off-nominal | typedef struct sockaddr_storage pj_sockaddr_storage |
| 12:34.12 | off-nominal | but whatever |
| 12:34.21 | file | jkroon: it's because fundamentally PJSIP was meant to be bundled and included in things, where you have tight control and can change things as needed |
| 12:34.30 | file | it's not truly made to be generic in its build process or usage |
| 12:34.53 | file | we added shared library support to appease the distros but for good real world usage you want a custom build of it |
| 12:35.55 | file | that's partially why we do bundled - because we build it in the ideal configuration for use as a server in Asterisk, and include fixes that haven't seen a release, it's just better for real world users |
| 12:36.14 | jkroon | well, the same folks that package asterisk for gentoo is also the guys that package pjproject, which is mostly a guys working at LINX, myself, and some other contributors here and there. |
| 12:37.16 | file | sure, but our user base is mostly Ubuntu and CentOS/Fedora |
| 12:37.25 | file | which targets a generic build that is what it is |
| 12:37.48 | jkroon | point being if there things we can do differently to make it easier we're open to suggestions and pointers. |
| 12:39.15 | file | I don't think there's really anything you can do differently, PJSIP just wasn't designed or implemented to be built in a generic fashion for everyone - to really be useful it has to be customized to what is using it |
| 12:39.22 | file | now whether our server config is "good" for clients I don't know |
| 12:45.04 | jkroon | review/sean_bright/ASTERISK-28049 - compile tests at least. |
| 12:50.41 | off-nominal | jkroon: no code actually uses those functions yet, so a compilation test is sufficient |
| 13:02.52 | jkroon | you need me to +1 all branches or is one fine? |
| 13:03.30 | off-nominal | well they will all need a +1 if they are to eventually be committed |
| 13:04.36 | jkroon | ok, my comment can be interpreted the wrong way, would really like to see a function like that (which doesn't break compile) merged. |
| 13:04.58 | file | anything which is up for review gets reviewed with the goal of being merged |
| 13:05.45 | jkroon | yea, bad that this has to be reverted really, so what i'm trying to say really is that I hope the revert isn't the last of it. |
| 13:09.36 | file | he abandoned the revert |
| 13:14.00 | off-nominal | jkroon: i abandoned the revert. i uploaded a new review which (hopefully) fixes the issue. |
| 13:14.37 | off-nominal | file: i removed all pjsip installs except for the one and am still getting compile errors. |
| 13:14.44 | off-nominal | so something else might be going on |
| 13:14.48 | file | off-nominal: interesting! |
| 13:14.57 | off-nominal | pj_mutex_unlock(tsx->mutex); |
| 13:15.02 | off-nominal | error about mutex not being there |
| 13:15.10 | off-nominal | and the _uas one from earlier |
| 13:15.13 | off-nominal | will try to figure it out |
| 13:16.03 | gtjoseph | off-nominal: is ./configure reporting the correct status that matches what the install really is? |
| 13:17.02 | off-nominal | gtjoseph: please rephrase that in the form of something i can answer |
| 13:17.09 | off-nominal | define "correct status" |
| 13:17.43 | jkroon | off-nominal, so change to review, because 10097 still points at the revert even though it's been abandoned. |
| 13:18.02 | off-nominal | https://gerrit.asterisk.org/#/c/asterisk/+/10097/ |
| 13:18.06 | off-nominal | that is not the revert |
| 13:18.24 | jkroon | hahaha, so your new version is what I did as a temp fix to get the thing to compile yesterday. |
| 13:18.26 | off-nominal | this was the revert before i abandoned it: https://gerrit.asterisk.org/#/c/asterisk/+/10093/ |
| 13:18.55 | off-nominal | gtjoseph: in pjproject (vanilla), 'pjsip_transaction' has a 'mutex_b', not a 'mutex' |
| 13:19.10 | off-nominal | ahh, grp_lock |
| 13:19.12 | off-nominal | let me see |
| 13:19.13 | gtjoseph | when you run configure, it will report the features the version of pjproject it found has |
| 13:19.22 | off-nominal | right, sorry |
| 13:19.34 | gtjoseph | PJ_TRANSACTION_GRP_LOCK |
| 13:19.41 | gtjoseph | PJSIP_DLG_CREATE_UAS_AND_INC_LOCK |
| 13:19.42 | gtjoseph | etc |
| 13:22.07 | off-nominal | yeah, so it doesn't match with what is actually there |
| 13:23.12 | off-nominal | i have a theory |
| 13:23.32 | file | it could be bunnies? |
| 13:23.49 | jkroon | must've gotten confused here ... conference calls messed me up bad today. |
| 13:23.52 | off-nominal | i think i set PKG_CONFIG_PATH to the wrong value |
| 13:23.59 | off-nominal | hmm, maybe not |
| 13:24.24 | gtjoseph | is your install in the standard system paths? |
| 13:24.27 | off-nominal | no |
| 13:24.30 | off-nominal | /opt/pjproject |
| 13:24.31 | gtjoseph | ah |
| 13:24.45 | off-nominal | PKG_CONFIG_PATH=/opt/pjproject/lib/pkgconfig/ ./configure --enable-dev-mode --with-pjproject=/opt/pjproject/ |
| 13:24.52 | off-nominal | is that wrong? |
| 13:25.16 | off-nominal | i mean, it's not working, so clearly it is wrong. |
| 13:25.17 | gtjoseph | is /opt/pjproject the source tree or an install tree? |
| 13:25.47 | off-nominal | install |
| 13:26.12 | off-nominal | i untar'd it. copied asterisk's config_site.h and user.mak. make dep; make; make install |
| 13:26.23 | off-nominal | and used third-party's configure line as well |
| 13:26.42 | off-nominal | i will just install in the system path |
| 13:26.46 | off-nominal | as much as i hate doing that |
| 13:27.19 | gtjoseph | so when you configured pjproject, you used --prefix=/opt/pjproject --libdir=/opt/pjproject/lib, etc? |
| 13:27.37 | off-nominal | not libdir |
| 13:27.47 | gtjoseph | or did you use DESTDIR=/opt/pjproject when you did make install? |
| 13:28.01 | off-nominal | libdir isn't set in any of the third-party makefiles |
| 13:28.11 | off-nominal | i just set --prefix=/opt/pjproject |
| 13:28.17 | *** join/#asterisk-dev cresl1n (uid299068@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 13:28.17 | *** mode/#asterisk-dev [+o cresl1n] by ChanServ |
| 13:28.48 | gtjoseph | so it created /opt/pjproject/lib. /opt/pjprojectinclude, etc? |
| 13:28.52 | off-nominal | yes |
| 13:28.53 | gtjoseph | so it created /opt/pjproject/lib. /opt/pjproject/include, etc? |
| 13:28.56 | off-nominal | yes |
| 13:29.00 | off-nominal | yes. |
| 13:29.16 | gtjoseph | :). |
| 13:29.20 | off-nominal | :D |
| 13:29.24 | off-nominal | SHIT BROKE |
| 13:29.36 | gtjoseph | final question... did you scrub /usr/local |
| 13:29.47 | off-nominal | scrub: command not found |
| 13:29.56 | off-nominal | i did an updatedb, locate pjsip.h |
| 13:30.05 | off-nominal | the only place that shows up is /opt/project |
| 13:30.13 | off-nominal | pjproject* |
| 13:30.39 | off-nominal | ok, so i want to install in standard location - should i just drop the --prefix entirely? or set it to /usr or /usr/local? |
| 13:30.56 | gtjoseph | the default is /usr/local i believe |
| 13:31.02 | off-nominal | ok, i will give that a whirl |
| 13:31.06 | gtjoseph | do i always just --prefix=/usr |
| 13:31.10 | gtjoseph | BUT... |
| 13:31.20 | off-nominal | BUT SOFT |
| 13:31.22 | gtjoseph | are you on a redhat based distro with a /usr/lib64? |
| 13:31.29 | off-nominal | what light from yonder window breaks |
| 13:31.33 | off-nominal | i am not |
| 13:31.46 | off-nominal | do i look like an asshole to you? |
| 13:31.49 | off-nominal | (don't answer that) |
| 13:31.59 | gtjoseph | ok, then use --prefix=/usr then when you do a make install, specify DESTDIR=/opt/pjproject |
| 13:32.11 | gtjoseph | and make sure /opt/pjproject is empty before you do. |
| 13:32.13 | off-nominal | that's... weird |
| 13:32.17 | off-nominal | will try it |
| 13:33.34 | off-nominal | DESTDIR=/opt/pjproject make install |
| 13:33.35 | off-nominal | ? |
| 13:33.54 | gtjoseph | the DESTDIR can come before or after the make but yeah |
| 13:34.18 | off-nominal | ok |
| 13:34.50 | off-nominal | how should i run asterisk's configure now? |
| 13:34.59 | gtjoseph | DESTDIR is a fairly standard way of installing to alternate paths. |
| 13:35.10 | off-nominal | but then why set prefix? |
| 13:35.11 | gtjoseph | ---with-pjproject=/opt/.pjproject |
| 13:35.17 | gtjoseph | ---with-pjproject=/opt/pjproject |
| 13:35.30 | gtjoseph | otherwise it'll go into /opt/pjproject/usr/local |
| 13:35.59 | jkroon | ok, my votes are in tact. |
| 13:36.17 | off-nominal | i see |
| 13:36.18 | off-nominal | ok |
| 13:36.40 | *** join/#asterisk-dev CELYA_ (~Thunderbi@92.184.96.210) |
| 13:36.44 | jkroon | we install pjproject into /usr so it's "standard" (whatever that implies nowadays). |
| 13:36.53 | off-nominal | The PJPROJECT installation appears to be missing or broken. |
| 13:37.07 | off-nominal | do i need to set PKG_CONFIG_PATH |
| 13:37.08 | off-nominal | ? |
| 13:37.14 | gtjoseph | never had to use it |
| 13:37.34 | gtjoseph | try --with-pjproject=/opt/pjproject/usr |
| 13:37.59 | gtjoseph | and let me look at my older scripts to see if i have a working example. |
| 13:38.56 | off-nominal | this is precisely why everyone should used the bundled version |
| 13:39.10 | file | "it just works" |
| 13:39.12 | gtjoseph | :P |
| 13:39.16 | off-nominal | doing /opt/pjproject/usr didn't change anything |
| 13:41.25 | off-nominal | oh. might be ld path. |
| 13:41.58 | gtjoseph | look at config.log to see if it couldn't find the includes or the libs |
| 13:42.33 | off-nominal | No package 'libpjproject' found |
| 13:42.38 | off-nominal | which is a pkg-config error |
| 13:42.42 | gtjoseph | yeah ok |
| 13:43.06 | off-nominal | trying again after setting PKG_CONFIG_PATH |
| 13:43.10 | gtjoseph | yep |
| 13:44.24 | off-nominal | didn't fail, but didn't find the right options |
| 13:44.31 | off-nominal | i... give up |
| 13:44.33 | off-nominal | I GIVE UP |
| 13:44.51 | gtjoseph | is it a mix of yes and nos or are they all nos? |
| 13:44.57 | off-nominal | all nos |
| 13:45.26 | gtjoseph | yeah then it's still not finding the files. check the config.log before you give up. |
| 13:45.38 | off-nominal | it's ld path |
| 13:45.45 | off-nominal | ldconfig -p doesn't show anything |
| 13:45.51 | jkroon | if you look in /opt/pjproject, is there a usr folder, or include and lib? |
| 13:46.52 | gtjoseph | dumb question... did you --enable-shared with you built pjproject? |
| 13:46.57 | off-nominal | no |
| 13:47.01 | off-nominal | should i have done that? |
| 13:47.02 | off-nominal | heh |
| 13:47.17 | gtjoseph | then only the .a files were generated which non bundled won't use. L:) |
| 13:47.20 | off-nominal | ok, so i guess my first mistake was using the configure stuff from third-party/pjproject |
| 13:47.43 | gtjoseph | well, that's a good start but for a shared install, you need --enable-shared |
| 13:47.55 | gtjoseph | becaue pjproject by default only builds static libs |
| 13:48.06 | off-nominal | right |
| 13:48.16 | off-nominal | so how do we get the last 30 minutes back? |
| 13:48.31 | gtjoseph | we need the way back time machine |
| 13:48.53 | off-nominal | will work on it |
| 13:52.30 | off-nominal | checking for pjsip_tsx_layer_find_tsx2 in -lpjsip... yes |
| 13:52.32 | off-nominal | huzzah |
| 13:52.55 | gtjoseph | YAY! |
| 13:53.07 | off-nominal | so to recap... |
| 13:53.12 | off-nominal | if you do it right, it will work |
| 13:53.19 | gtjoseph | did you still need to set PKG_CONFIG_PATH? |
| 13:53.26 | off-nominal | i just did it, didn't try without |
| 13:53.33 | gtjoseph | give it a shot |
| 13:53.35 | off-nominal | but i don't know how it would find it otherwise |
| 13:53.38 | off-nominal | i will |
| 13:53.45 | off-nominal | but i want to run this stupid unit test first |
| 13:53.50 | gtjoseph | no worries |
| 13:54.07 | off-nominal | i went back to --prefix=/opt/pjproject |
| 13:54.09 | off-nominal | fwiw |
| 13:54.19 | gtjoseph | without the DESTDIR? |
| 13:54.22 | off-nominal | yes |
| 13:54.30 | gtjoseph | ok. |
| 13:54.59 | gtjoseph | the only real difference is that you can use DESTDIR to different places without having to reconfigure and recompile. |
| 13:57.46 | off-nominal | well i prefer not having the 'usr' dir in /opt/project |
| 13:57.50 | off-nominal | pjproject* |
| 13:57.58 | off-nominal | without PKG_CONFIG_PATH it doesn't work |
| 13:58.13 | off-nominal | because pkg-config doesn't know where to look for the libpjproject.pc file |
| 13:59.39 | gtjoseph | yep ok. thanks for checking. |
| 13:59.44 | seanbright | sure |
| 14:00.27 | seanbright | thanks for the hel |
| 14:00.29 | seanbright | p |
| 14:00.50 | gtjoseph | hell was probably more appropriate. |
| 14:02.09 | seanbright | nah, bro |
| 14:07.44 | *** join/#asterisk-dev sduthil (~sduthil@modemcable094.94-70-69.static.videotron.ca) |
| 14:09.23 | seanbright | i love working with marketing people |
| 14:09.40 | file | eyes seanbright |
| 14:09.45 | seanbright | they feel that the solution to every problem is "add another queue!" |
| 14:10.03 | file | isn't it? |
| 14:10.16 | gtjoseph | queues for everyone! |
| 14:12.26 | seanbright | i think at this point they have more queues than butts in seats |
| 14:12.46 | seanbright | and that isn't a joke |
| 14:13.45 | file | add more! MOAR! |
| 14:18.21 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-txpwenpoqrjmumax) |
| 14:18.21 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 14:26.03 | file | cresl1n: do you still have MOH qualms? |
| 14:26.52 | cresl1n | I need some more time this morning to look at it |
| 14:28.21 | file | loans cresl1n some time |
| 14:35.44 | coreyfarrell | file / gtjoseph: sorry I failed to fix the tag on https://gerrit.asterisk.org/10077 so it was missed with your jansson-destdir reviews. |
| 14:36.02 | gtjoseph | np |
| 14:36.31 | file | ponders |
| 14:36.35 | coreyfarrell | also crazy that gerrit doesn't have an option to stop it from renaming tags on cherry-pick. I don't understand any situation where that would be useful. |
| 14:36.44 | file | coreyfarrell: yeah. that. |
| 14:36.58 | file | I was just pondering how hard it would be to write a bot |
| 14:37.40 | file | I don't think that bad |
| 14:38.03 | file | the event should contain all the info, and it's just a comparison of the branch and end of topi |
| 14:38.04 | file | c |
| 14:39.50 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-emromkvqnwtooyml) |
| 14:40.36 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-kuyocsffdswihely) |
| 14:40.36 | *** mode/#asterisk-dev [+o rmudgett] by ChanServ |
| 14:41.46 | file | I may poke at that this weekend or tonight |
| 14:42.42 | file | if I were fancier I'd do it as a plugin, but Java is in my list of languages that I avoid |
| 14:42.46 | file | along with PHP and Perl |
| 14:44.36 | coreyfarrell | agree on all 3... wish I could avoid Perl. |
| 14:45.30 | gtjoseph | file: you may jut be able to drop something in gerrit-home/hooks like the merge script. |
| 14:45.40 | file | gtjoseph: ah good point |
| 14:51.16 | file | yeah, should be possible |
| 15:12.22 | coreyfarrell | for stasis functions which use ast_json_deep_copy, does that have any purpose beyond protecting against jansson <= 2.10 threading issues? I'm thinking since Asterisk 16+ requires a thread-safe version we can switch them to ast_json_ref, I'm not aware of any situations where we mutate json objects after initial creation. |
| 15:14.51 | file | gtjoseph: I don't think I can do it over SSH |
| 15:18.05 | gtjoseph | file: can you use wget or curl to run an rest command? |
| 15:18.37 | file | yes, it just requires different auth |
| 15:19.41 | file | and I don't have a service account to use for that purpose |
| 15:19.54 | gtjoseph | use the jenkins2 one. |
| 15:20.04 | file | can you log in as it? |
| 15:20.21 | gtjoseph | not from the ui but you can from rest |
| 15:20.44 | gtjoseph | that's how the test results summaries are posted. |
| 15:20.56 | file | ah |
| 15:21.33 | file | I'll dabble in that tonight then |
| 15:21.43 | file | do you have the credentials for the REST API somewhere? |
| 15:21.43 | gtjoseph | cool. |
| 15:21.56 | gtjoseph | yeah. i'll send them to you. |
| 15:22.00 | file | merci |
| 15:23.14 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:e0c3:32cb:7782:91fe) |
| 15:33.32 | file | it's theoretically done |
| 15:34.52 | file | hrm |
| 15:35.49 | file | oh |
| 15:36.01 | rmudgett | You might get it if you hum a few bars more. :) |
| 15:38.37 | file | well this is silly |
| 15:39.20 | rmudgett | humming bars? |
| 15:39.26 | file | I don't think I get the review identifier I can use with the REST API to update the topic |
| 15:39.50 | gtjoseph | figures |
| 15:41.29 | file | hum |
| 15:41.39 | gtjoseph | have you echo'd the enture comman dline passed in to see what you get? |
| 15:42.21 | file | not yet, about to do that annoyingly |
| 15:43.10 | file | oh that's silly |
| 15:43.15 | file | I got it, but it's not what you'd expect |
| 15:43.55 | gtjoseph | it never is |
| 15:49.52 | file | done. |
| 15:50.20 | file | it will now automatically update the topic |
| 15:50.58 | file | the value is "change_url" |
| 15:51.13 | file | which I guess means, this is the change identifier you use in URLs |
| 15:51.20 | file | instead of here is a URL for the change |
| 15:52.08 | gtjoseph | :) |
| 15:58.47 | file | that's my foray into bash scripting for this month |
| 15:58.47 | file | :D |
| 16:27.43 | coreyfarrell | gtjoseph: FYI I just reopened ASTERISK-27121, the recent stasis_cache patch causes a new leak (sorry I didn't get a chance to test it last week). I'm trying to investigate but my understanding of stasis_cache is pretty limited. |
| 16:28.10 | gtjoseph | i don't think it's a new leak but an old one |
| 16:28.39 | gtjoseph | are we talkign ref leak or memory leak? |
| 16:29.12 | coreyfarrell | ref leak, I've attached refs.txt to the ticket. I did verify that reverting the stasis_cache patch makes it go away. |
| 16:30.20 | gtjoseph | i think it's the reference to the stasis_message_type which is an ao2 object for some reason. |
| 16:30.50 | gtjoseph | anyway, i've got a patch in process that kills caching the stasis_subscription_change messages altogether. |
| 16:30.56 | coreyfarrell | I'm not 100% sure but I think the root leak is from stasis_cache.c:224 cache_entry_create. |
| 16:31.48 | gtjoseph | yeah it adds a ref to the message type |
| 16:32.10 | gtjoseph | anyway, i'll take a look. |
| 16:36.54 | coreyfarrell | oh I think I found the issue, cache_remove returns a reference which the caller has to clean. |
| 16:37.05 | gtjoseph | yep, just saw that. :) |
| 16:37.45 | gtjoseph | anyway, i'll deal with it. |
| 16:37.50 | coreyfarrell | cool thanks! |
| 16:37.57 | gtjoseph | no problem |
| 16:38.33 | file | I don't know how I feel about that |
| 16:38.46 | gtjoseph | about what? |
| 16:39.17 | file | It returning what you have to clean up, but shrug |
| 16:40.26 | gtjoseph | it's not assuming you want to get rid of the message that was attached to it |
| 16:40.33 | coreyfarrell | adding attribute_warn_unused_result would probably be nice. |
| 16:41.01 | *** join/#asterisk-dev jkroon (~jkroon@165.16.204.171) |
| 16:41.03 | gtjoseph | it's removing the cache entry and returning the message that was cached. |
| 17:26.36 | coreyfarrell | is 16.0.0 being released already? |
| 17:27.02 | coreyfarrell | asking because I saw things in asterisk-bugs mailing list being tagged to that release version.. |
| 17:27.33 | file | 16.0.0-rc2 is today |
| 17:28.05 | coreyfarrell | ah ok thanks |
| 17:28.22 | file | 16.0.0 is Astricon |
| 17:28.28 | file | probably, most likely |
| 17:29.28 | coreyfarrell | oh I guess I should have known another rc would have to be released since stuff was merged to 16.0. |
| 18:05.24 | *** join/#asterisk-dev opticron (~opticron@75.76.209.103) |
| 18:05.24 | *** mode/#asterisk-dev [+o opticron] by ChanServ |
| 18:05.42 | gtjoseph | has anyone ever set up voicemail using the imap backend? |
| 18:11.07 | seanbright | yes |
| 18:11.19 | seanbright | i mean... not me. but certainly someone has. |
| 18:12.14 | gtjoseph | thanks :) |
| 18:12.25 | seanbright | anytime! |
| 18:23.15 | *** join/#asterisk-dev fblackburn (~fblackbur@modemcable094.94-70-69.static.videotron.ca) |
| 18:35.01 | fblackburn | Hi, I play with ARI and I have made a small patch to allow to create bridge associated to an application. It is something that make sens for you/the project? The patch add `app` as query string to POST /bridges (as channel does) |
| 18:40.26 | gtjoseph | fblackburn: off hand i don't see why not. you should submit the patch for comments by following the instructions at https://wiki.asterisk.org/wiki/display/AST/Patch+Contribution+Process |
| 19:16.43 | fblackburn | Ok thanks, I follow `Asterisk Issue Guidelines` and for feature it tells me to discuss here 8) |
| 19:24.49 | *** join/#asterisk-dev leedm777 (textual@nat/digium/x-qqzcrcwgxoawastc) |
| 19:24.49 | *** mode/#asterisk-dev [+o leedm777] by ChanServ |
| 19:28.04 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01cb0583d31d0094e0956583ad8123.ipv6.abo.wanadoo.fr) |
| 19:55.23 | jameswf | this modou lo guy tries a different approach.... |
| 19:55.35 | jameswf | on the dev-list.... help me tax people |
| 20:38.12 | seanbright | jameswf: i'm calling you 'dear' from now on |
| 21:08.51 | rmudgett | kharwell: Why are we even bothering to cherry-pick https://gerrit.asterisk.org/#/c/testsuite/+/10102/ to 15.6? |
| 21:09.32 | kharwell | good question |
| 21:11.05 | kharwell | gtjoseph asked to have them (or other ones) cherry picked to those branches. I figured either way was fine, so just cherry picked away |
| 21:11.50 | rmudgett | It's not like the testsuite is going to be run on that branch much. |
| 21:11.54 | gtjoseph | if we do point releases in those branches we'll want the tests fixed in the corresponding testsuie branch |
| 21:13.49 | rmudgett | gtjoseph: Doesn't seem as if the testsuite is running on 15.6 |
| 21:22.03 | *** join/#asterisk-dev sahmed (~sahmed@host-212-68-252-215.dynamic.voo.be) |
| 21:22.22 | sahmed | Hello, Good Morning |
| 21:23.49 | sahmed | I need a code base help, is there any way to get peer rtp port information from ast_rtp_instance structure? |
| 21:25.16 | file | https://github.com/asterisk/asterisk/blob/master/include/asterisk/rtp_engine.h#L1145 |
| 21:27.04 | *** join/#asterisk-dev CELYA_ (~Thunderbi@LFbn-1-5359-81.w90-105.abo.wanadoo.fr) |
| 21:27.54 | sahmed | @file lets check this one :) |
| 22:04.50 | sahmed | @file this is exactly the same what I am looking for, Thanks a lot |
| 22:06.07 | sahmed | @file do you have any idea can we configure rtp port such way that peer channel port will be next event port. is there any configuration like that? |
| 22:06.25 | file | I don't understand the question |
| 22:07.02 | file | we don't control what ports the remote side chooses, only our own |
| 22:07.12 | sahmed | I mean if one leg took 10000 port for rtp the peer channel take 10002 |
| 22:07.17 | file | and for ours it randomly chooses from the configured range - there is no configuration to change that |
| 22:07.18 | *** join/#asterisk-dev tdonahue (~tdonahue@2600:8805:d500:fb00:1aba:515a:ee6f:cd17) |
| 22:08.17 | sahmed | Got it, thanks for quick answer :) |
| 22:14.59 | kharwell | automatic topic update. so nice |
| 22:16.21 | *** join/#asterisk-dev tdonahue (~tdonahue@2600:8805:d500:fb00:1aba:515a:ee6f:cd17) |
| 22:18.47 | file | kharwell: ooo yessssssss |
| 22:20.41 | *** join/#asterisk-dev tdonahue (~tdonahue@2600:8805:d500:fb00:1aba:515a:ee6f:cd17) |
| 22:27.54 | coreyfarrell | yeah thanks for dealing with that! |
| 22:28.26 | file | there is also a hook that can actually prevent pushes |
| 22:28.37 | file | I'm pondering whether there's anything actually useful we can do with it |
| 22:28.42 | coreyfarrell | what do you mean? |
| 22:29.36 | file | https://gerrit.googlesource.com/plugins/hooks/+/master/src/main/resources/Documentation/hooks.md those are the hooks |
| 22:30.22 | file | I don't thinkkk there is anything |
| 22:31.17 | *** join/#asterisk-dev tdonahue (~tdonahue@2600:8805:d500:fb00:1aba:515a:ee6f:cd17) |
| 22:32.29 | coreyfarrell | if you're talking about to block people from posting reviews with 'git review' then I don't think that'd be useful. only people with CLA can post to gerrit and if absolutely required you can always delete a review. |
| 22:33.11 | file | I was more thinking a method to just-in-case block security reviews that may have gone up to the wrong place |
| 22:33.11 | coreyfarrell | for most issues I think the current way is better - accept the review and let jenkins post -1 if an automated check fails |
| 22:33.30 | coreyfarrell | not sure how you would even catch that |
| 22:33.44 | coreyfarrell | like how you would identify something as a security review |
| 22:34.07 | file | try to retrieve it on JIRA and if it's forbidden? |
| 22:35.11 | *** join/#asterisk-dev tdonahue (~tdonahue@2600:8805:d500:fb00:1aba:515a:ee6f:cd17) |
| 22:35.18 | file | something to ponder, something to ponder |
| 22:36.22 | coreyfarrell | that would also force you to unlock the JIRA issue when it's time to publish the security release |
| 22:37.16 | file | we're supposed to do that anyway |
| 22:37.25 | coreyfarrell | oh I meant that as a good thing.. |
| 22:37.30 | file | yeah |
| 22:38.33 | *** join/#asterisk-dev tdonahue (~tdonahue@2600:8805:d500:fb00:1aba:515a:ee6f:cd17) |
| 22:43.22 | coreyfarrell | only other thing I'd say is the hook should have logic so it passes if JIRA is unreachable |
| 22:46.42 | *** join/#asterisk-dev tdonahue (~tdonahue@2600:8805:d500:fb00:1aba:515a:ee6f:cd17) |
| 23:01.27 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-txpwenpoqrjmumax) |
| 23:31.15 | leifmadsen | if someone wanted to add some sort of automated build stuff each time a new tag was created for Asterisk, what would be the recommended place to start? |
| 23:31.24 | leifmadsen | (not a per-commit build, just per build...) |
| 23:31.27 | leifmadsen | s/build/tag/ |
| 23:31.42 | leifmadsen | sighs |
| 23:31.51 | leifmadsen | s/per build/per tag/ |
| 23:38.59 | coreyfarrell | leifmadsen: once the tag is created I'd think it is too late because you can't change the tag once it's created. it seems like the extra 'release only' tests should run against each new commit to the X.X branches (like 15.6 and 16.0) |
| 23:41.44 | leifmadsen | coreyfarrell: I don't want to change anything post tag; I want to build a container |
| 23:43.01 | coreyfarrell | maybe I'm misunderstanding, you mean like to publish docker images on each release? |
| 23:43.13 | leifmadsen | yes, RPMs, containers, etc |
| 23:43.22 | leifmadsen | so the code should already be immutable |
| 23:43.26 | leifmadsen | I'm not looking at running a test |
| 23:43.41 | leifmadsen | hence the original request about adding automated build stuffz |
| 23:44.25 | leifmadsen | of course I can always execute against the github mirror and just watch for an event there, but it seems like something that should live closer to home |
| 23:44.42 | leifmadsen | my desire would be to leverage buildah to create OCI compatible Asterisk containers |
| 23:45.00 | leifmadsen | (which I already do with a build script here, that I'd be happy to donate to the project) |
| 23:45.18 | leifmadsen | I'm just sure enough things have changed that I'm not sure where that type of trigger would happen |
| 23:46.13 | leifmadsen | afk, sim racing league starts in 15 mins |
| 23:49.54 | *** join/#asterisk-dev snuff-work (~snuff-wor@210.9.148.102) |
| 23:49.54 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |