| 00:00.02 | gtjoseph | no problem |
| 00:00.23 | gtjoseph | i think you hold the record for fastest time to testsuite running |
| 00:00.42 | dan_j | i followed the wiki yesterday on getting it set up |
| 00:00.57 | gtjoseph | :) |
| 00:01.03 | dan_j | whoever wrote the wiki did a good job it seems |
| 00:02.15 | dan_j | gtjoseph: btw, i saw a comment to your test for the rtp fix |
| 00:02.29 | gtjoseph | ? |
| 00:03.29 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-pjhnpuhxaforxpaj) |
| 00:03.40 | gtjoseph | are you referring to the 127.0.0.x stuff? |
| 00:04.24 | dan_j | ah, never mind. didnt see there was a subsequent comment |
| 00:05.41 | dan_j | night all |
| 00:06.02 | gtjoseph | good night! |
| 00:44.45 | *** join/#asterisk-dev newtonr (RustyNewto@nat/digium/x-vyypypkzvgngbxvg) |
| 00:44.45 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 01:31.30 | *** join/#asterisk-dev gtjoseph (~gtj0@unaffiliated/gtj) |
| 01:39.37 | *** join/#asterisk-dev snuff-work (~blatus@203.87.61.22) |
| 01:39.38 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 01:43.32 | file | rmudgett, well... that's better |
| 01:44.02 | rmudgett | The new CLI command? |
| 01:44.29 | file | yes |
| 01:45.12 | rmudgett | Corey came up with the new approach |
| 01:45.42 | gtjoseph | does the new approach allow us to inspect the info from code? |
| 01:45.53 | file | gtjoseph, ha ha |
| 01:45.58 | file | gtjoseph, you're funny |
| 01:46.31 | gtjoseph | wasn't trying to be, :) |
| 01:48.13 | rmudgett | file: Looks like when I disable the pjsip pool cache the memory doesn't balloon under the stress test. Only one run so far. |
| 01:48.34 | file | rmudgett, did you try dropping the max capacity on the caching pool too? |
| 01:48.41 | file | our value may be too big |
| 01:49.00 | rmudgett | Where is that parameter. I didn't see one. |
| 01:50.11 | coreyfarrell | rmudgett: the only thing I was just thinking is do we want to use the "info" keyword for this CLI command? I mean since "info" is a thing in SIP, maybe "buildopts" or similar? |
| 01:50.28 | file | rmudgett, it's in res_pjsip.c when it creates the caching pool |
| 01:50.31 | file | it's hardcoded to 1024 * 1024 right now |
| 01:50.46 | coreyfarrell | who will ever need more than 1mb of memory? |
| 01:50.55 | rmudgett | coreyfarrell: That was something I was wanting. Buildops seems a better word than info. |
| 01:51.06 | file | coreyfarrell, what that actually means I'm not sure |
| 01:54.08 | coreyfarrell | file: hmm apparently it was 640k that bill gates was rumored to say nobody will ever need more memory |
| 01:54.30 | gtjoseph | ok you guys so seriously... why not do a pj_dump_config once at startup and regex the result into a container? |
| 01:55.29 | coreyfarrell | gtjoseph: that would require keeping that memory always allocated |
| 01:55.41 | gtjoseph | so |
| 01:56.24 | gtjoseph | just thinking ahead |
| 01:56.27 | coreyfarrell | this is meant to be a rarely used troubleshooting tool |
| 01:56.37 | gtjoseph | i do still need PJ_MAX_HOSTNAME |
| 01:57.19 | rmudgett | You have to change that in the pj/config_site.h file |
| 01:57.31 | gtjoseph | i don't want to change it |
| 01:57.57 | gtjoseph | <PROTECTED> |
| 01:58.09 | rmudgett | Include site/user specific configuration to control PJLIB features. |
| 01:58.15 | rmudgett | YOU MUST CREATE THIS FILE YOURSELF!! |
| 01:58.26 | coreyfarrell | gtjoseph: personally I think pjproject needs API calls to retrieve build options.. hacking them out of a dump to logs seems like a bad idea |
| 01:58.29 | gtjoseph | no, the builder of pjproject must include it |
| 01:58.42 | gtjoseph | it defaults to 128 |
| 01:58.47 | gtjoseph | but THEY can change iot |
| 01:58.56 | gtjoseph | coreyfarrell: no kidding |
| 01:59.04 | gtjoseph | they also need to format their codebase |
| 01:59.16 | gtjoseph | instead of left justifying everything |
| 01:59.53 | rmudgett | Use tabs of 8 with space indentation of 4 |
| 01:59.59 | rmudgett | Looks much better then |
| 02:00.29 | gtjoseph | rmudgett: do you get what i'm after? |
| 02:05.04 | rmudgett | PJ_MAX_HOSTNAME is available as a define when you include pjlib.h |
| 02:05.27 | rmudgett | Though you have to define non-default values in pj/config_site.h |
| 02:05.27 | gtjoseph | at compile time |
| 02:05.34 | gtjoseph | not at run time |
| 02:06.24 | rmudgett | pjsip does have a lot of things defined at compile time. Such as system timers. |
| 02:06.36 | gtjoseph | yep, but this one can cause asterisk to crash |
| 02:06.38 | rmudgett | The default is 3k timers |
| 02:07.02 | rmudgett | So can running out of timers |
| 03:04.03 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 03:31.59 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 03:54.01 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 04:13.25 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 04:36.31 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 06:12.02 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-179-40-172.cust.bezeqint.net) |
| 06:12.03 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 06:19.38 | *** join/#asterisk-dev goddva (~goddva@77.40.154.242) |
| 06:54.19 | *** join/#asterisk-dev Juggie (~Juggie@unaffiliated/juggie) |
| 06:58.54 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 08:00.44 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 08:10.10 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 08:14.41 | *** join/#asterisk-dev lumasepa (~gestoip@193.145.124.30) |
| 08:21.55 | *** join/#asterisk-dev Demon_VoIP (~demon@109.60.135.186) |
| 08:22.22 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 08:22.22 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 08:53.06 | *** join/#asterisk-dev Tim_Toady (~fuzzy@snf-33276.vm.okeanos.grnet.gr) |
| 09:07.14 | lumasepa | Hi, I'm developing a module for asterisk that implements some functions to use redis from the dialplan and I'm having some issues, if someone can help me, I'm using the hiredis library and in hiredis there are something called redisContext that holds the info of the connection with redis, when you use a multithread pattern each thread should use different redisContext, as I'm developing a dialplan function the code runs in a channel thread and that |
| 09:35.04 | tzafrir | lumasepa, and did you have any questions? |
| 09:37.19 | lumasepa | Yes, how the asterisk resources work ? |
| 09:38.12 | lumasepa | and what do you think is the better way to do that (only one connection to the redis server) |
| 12:56.00 | stefan27 | hi lumasepa, I have no knowledge of what you're working on, I just wanted to tell you that your first message ended with "...in a channel thread and that" -not a complete sentence? |
| 13:17.21 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 13:34.29 | file | eyes seanbright |
| 13:34.34 | file | seanbright, you don't have to keep abandoning |
| 13:34.39 | seanbright | yes i do |
| 13:34.44 | seanbright | because i am done |
| 13:34.54 | seanbright | fix it, don't fix it, i don't really care |
| 13:34.57 | seanbright | heh |
| 13:35.01 | file | huh, does 11 not have ast_module_reload? |
| 13:36.20 | seanbright | i don't know what the build log says, so i'll have to take your word for it |
| 13:36.52 | file | it says |
| 13:36.53 | file | 2016-01-13 07:32:22 res_musiconhold.c: In function 'handle_cli_moh_reload': |
| 13:36.53 | file | 2016-01-13 07:32:22 res_musiconhold.c:1840: error: implicit declaration of function 'ast_module_reload' |
| 13:37.15 | seanbright | weird |
| 13:37.33 | file | in 11 it's in include/asterisk/_private.h |
| 13:37.40 | seanbright | indeed |
| 13:37.41 | file | it got moved to module.h in 13 |
| 13:37.46 | seanbright | gotcha |
| 13:38.30 | seanbright | i'm tapping out either way |
| 13:38.46 | file | I'm intrigued that Jenkins won't let you access it, it's using openid just like gerrit does and against the same database |
| 13:38.54 | seanbright | yeah, i'm able to get in now |
| 13:39.00 | seanbright | don't know what was happening yesterday |
| 13:39.21 | file | ack then I won't investigate further |
| 13:42.55 | seanbright | um |
| 13:43.14 | file | there, fixed |
| 13:43.18 | seanbright | ok, great |
| 13:43.29 | seanbright | how do i give you ownership of the reviews? |
| 13:43.36 | seanbright | i don't want to do it anymore |
| 13:43.47 | file | anyone can update reviews |
| 13:43.51 | seanbright | ok great |
| 13:44.17 | file | it's still your commit though, so in the grand scheme of things you still get direct credit |
| 13:44.38 | seanbright | well, that is my primary concern |
| 13:47.45 | file | let it snow, let it snow, let it snow |
| 13:47.56 | seanbright | how do i update the commit message on the cherry picked reviews? |
| 13:48.15 | file | you can click the Edit Message button on them |
| 13:48.20 | file | in gerrit |
| 13:48.27 | seanbright | simple enough |
| 13:49.29 | *** join/#asterisk-dev leedm777 (leedm777@nat/digium/x-pvanseoqzlmkxghs) |
| 14:08.22 | stefan27 | when asterisk pbx encounters ExecIf(x?a:b) and a=App1(${VAR1}) does asterisk really need to expand the variables in a like ${VAR1} in the event x is evaluted to false, which it does? |
| 14:09.52 | stefan27 | it's a minor problem for me to rewrite these dialplan calls, but i had an expression in a which only "compiles" when x is true |
| 14:11.20 | dan_j | When running testsuite, is it possible to examine the AMI output? |
| 14:12.27 | file | stefan27, I... have no idea ^_^ |
| 14:13.05 | file | dan_j, the testsuite creates log files in the 'logs' directory which should include AMI |
| 14:15.45 | WIMPy | stefan27: I'm pretty sure, variable evaluation heppens long before any appliction is executed. |
| 14:40.57 | WIMPy | stefan27: Yes. Fun thing: Try to use DB_DELETE for different entries. It looks correct, but both will be gone. |
| 14:42.14 | seanbright | yikes |
| 14:42.19 | seanbright | yeah, that's "fun" |
| 14:42.47 | WIMPy | Good opportunities for really obscure dialplans. |
| 14:50.35 | dan_j | Any idea why this test is just flying through before it gets the AOR_CONTACT_ADDED? |
| 14:50.36 | dan_j | http://pastebin.com/aV7DamPu |
| 14:50.41 | dan_j | http://pastebin.com/DGvCmNMk |
| 14:57.41 | lumasepa | stefan27, the continuation is : "means that every time that the functions are called it have to connect to the redis server send the command use the reply and release the connection, so my question is, there is some way to make only one connection to the redis server ? for example make a thread that handles the redisContext and all the request to redis are sent by this thread, or maybe making a resource but I don't know how the resources work, and |
| 14:58.44 | lumasepa | it seem that the text was truncated |
| 15:02.32 | lumasepa | the full question : |
| 15:02.35 | lumasepa | Hi, I'm developing a module for asterisk that implements some functions to use redis from the dialplan and I'm having some issues. |
| 15:02.53 | lumasepa | <PROTECTED> |
| 15:03.01 | lumasepa | As I'm developing a dialplan function the code runs in a channel thread and that means that every time that the functions are called it have to connect to the redis server send the command use the reply and release the connection. |
| 15:03.07 | lumasepa | So my question is, there is some way to make only one connection to the redis server ? for example make a thread that handles the redisContext and all the request to redis are sent by this thread, or maybe making a resource but I don't know how the resources work, and if they are executed in a separated thread |
| 15:04.41 | lumasepa | I'm gona go out now, if someone answers I will see it tomorrow |
| 15:05.57 | jsmith | lumasepa: You might look at the code in res_odbc |
| 15:06.32 | jsmith | lumasepa: It maintains connections to different ODBC-compliant databases, and you might find something there that helps you |
| 15:10.43 | file | lumasepa, you can either do a connection per thread using thread local storage, you can have a pool of connections which are pulled and used and then returned to the pool, or you can have 1 connection that is protected - the threading is up to you |
| 15:22.30 | dan_j | file: is a test supposed to wait to receive the ami events, or do i need to force it to wait before exiting? |
| 15:23.13 | file | I do not remember the details of that. |
| 15:23.53 | gtjoseph | dan_j: it should wait |
| 15:27.29 | dan_j | Yeh, I see it now. How can I make a pause between the register and the unregister? |
| 15:27.54 | dan_j | It's registering and unregistering before my 'Action: Getvar' is executed. |
| 15:28.05 | gtjoseph | <pause/> |
| 15:28.22 | gtjoseph | or <pause milliseconds=nnn/> |
| 15:28.54 | dan_j | safer to put a time on it? |
| 15:29.19 | gtjoseph | i think the default is 1 second |
| 15:29.23 | gtjoseph | so it should be safe |
| 15:34.46 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-bawggorgvdcapmti) |
| 15:40.15 | dan_j | How do I check for a response? I'm using pluggable_modules.EventActionModule at the moment. |
| 15:41.08 | gtjoseph | it's a pair of request/response. givce me a sec to look |
| 15:44.07 | dan_j | At the moment, I used ami-actions to send the request. |
| 15:44.21 | gtjoseph | oh, yeah... |
| 15:44.25 | dan_j | Looked through the sample tests but couldnt see anything obvious |
| 15:44.28 | gtjoseph | then use an ami-events: |
| 15:44.35 | gtjoseph | to capture events |
| 15:44.44 | dan_j | It doesnt seem to like it |
| 15:44.46 | dan_j | 1sec |
| 15:45.05 | gtjoseph | look at qualify/max_initial_qualify_time |
| 15:46.12 | dan_j | I dont see it requesting any information over ami in that test? |
| 15:46.29 | gtjoseph | yeah, i used a dummy ami-start: to kick off the process |
| 15:46.42 | gtjoseph | but the ami-event: should work |
| 15:46.58 | dan_j | Here's my test so far, http://pastebin.com/2hN6iJUJ , It's not happy with the Response section |
| 15:47.10 | dan_j | It says that i dont have an Event there which is required. |
| 15:48.07 | gtjoseph | yeah, you HAVE to start with an action |
| 15:48.16 | gtjoseph | or a placeholder like ami-start: |
| 15:48.51 | dan_j | really? If i remove the bottom section, it works fine. |
| 15:49.01 | gtjoseph | i know |
| 15:49.02 | dan_j | just doesnt check the response |
| 15:49.07 | gtjoseph | it's an odd module |
| 15:49.49 | dan_j | OK. I've added ami-start |
| 15:49.57 | dan_j | But getting this 'No event specified to match on. Aborting test' |
| 15:50.05 | gtjoseph | and you need a stop-test: at the end |
| 15:50.31 | gtjoseph | hmmm |
| 15:51.31 | gtjoseph | past the entire yaml and let me try it |
| 15:51.42 | gtjoseph | paste |
| 15:53.45 | dan_j | This is what I've got so far. |
| 15:53.46 | dan_j | http://pastebin.com/AcTYUbh2 |
| 15:54.10 | dan_j | just seen a typo |
| 15:54.29 | gtjoseph | and add the stop-test: at the end |
| 15:54.50 | gtjoseph | also paste the error |
| 15:55.51 | dan_j | Updated version |
| 15:55.52 | dan_j | http://pastebin.com/M44yQKVF |
| 15:55.58 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-zcrjlvmhmedfyend) |
| 15:56.16 | gtjoseph | ok, how about the error |
| 15:57.05 | dan_j | Sorry, just copying and pasting it out of my terminal window |
| 15:57.06 | dan_j | http://pastebin.com/7a0VJ9Qc |
| 15:57.43 | dan_j | Also, if I remove the response section, i get this |
| 15:57.49 | dan_j | Exception: Pluggable component 'stop-test' not recognized |
| 15:58.50 | dan_j | Ah stop_test |
| 15:59.08 | gtjoseph | oops my fault |
| 16:00.01 | dan_j | np |
| 16:01.51 | dan_j | Where are the modules stored? |
| 16:02.45 | gtjoseph | testsuite/lib/python/asterisk |
| 16:05.33 | gtjoseph | i hate pasting yaml |
| 16:06.27 | dan_j | Looks like that module only allows for Events and Actions following events |
| 16:07.24 | dan_j | But not responses |
| 16:07.25 | gtjoseph | give me a few minutes. i KNOW i have this working elsewhere |
| 16:08.02 | gtjoseph | maybe it's in combination with an EventModule |
| 16:12.29 | dan_j | Maybe I can use the originations module. It looks like it can trigger an originate based on an event such as the registration of 'alice' |
| 16:12.51 | gtjoseph | actually, use ListDialPlan |
| 16:13.08 | gtjoseph | look at... |
| 16:13.18 | dan_j | Is that an AMI action? |
| 16:13.19 | gtjoseph | config_wizard/hints |
| 16:13.47 | gtjoseph | actually it's ShowDialPlan |
| 16:13.53 | gtjoseph | and the events are ListDialPlan |
| 16:14.02 | gtjoseph | it's exactly what you want |
| 16:16.08 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 16:17.15 | dan_j | Yep |
| 16:17.17 | dan_j | Nice one |
| 16:17.48 | gtjoseph | i knew i did something like what you needed but i couldn't remember where. |
| 16:18.17 | file | gtjoseph, I have that happen often! |
| 16:18.41 | gtjoseph | yeah, but you have an excuse.... you have WAY more stuff here than I do. |
| 16:28.26 | dan_j | wow. it worked! fantastic! |
| 16:28.38 | dan_j | Now i need to test for the unregister and removal of the extension |
| 16:28.59 | dan_j | How do I test for the lack of something in the dialplan? count: 0 ? |
| 16:29.20 | gtjoseph | iirc: yes |
| 16:30.24 | dan_j | I hope eventually I wont have to bother you any more! One day! |
| 16:32.38 | dan_j | Am i correct that no full logs are saved if the test is successful? |
| 16:33.32 | dan_j | And that I run cleanup-test-remnants.sh before committing? |
| 16:37.19 | gtjoseph | i forget about the logs but cleanup is always a good idea even though the commit will ignore leftover stuff |
| 16:39.19 | dan_j | Does ShowDialplan depend on anything? |
| 16:40.02 | gtjoseph | Nope |
| 16:45.19 | dan_j | How do I restart the commit again? a .swp file has made it in. |
| 16:49.08 | dan_j | nm. i got it |
| 16:49.14 | dan_j | from irc history |
| 16:59.56 | gtjoseph | dan_j: you got a .swp file in the commit |
| 17:00.03 | dan_j | removed |
| 17:00.28 | gtjoseph | ah |
| 17:00.39 | dan_j | but i'm missing something i think |
| 17:00.41 | dan_j | 1sec |
| 17:00.54 | gtjoseph | the update to tests.yaml inm the parent |
| 17:01.01 | dan_j | yeh |
| 17:01.38 | dan_j | it's 'dir' ? |
| 17:01.44 | dan_j | not 'test' |
| 17:01.48 | gtjoseph | no test |
| 17:02.02 | gtjoseph | test for a single test, dir for a directory of tests |
| 17:02.39 | dan_j | I had to put a slightly longer pause in. |
| 17:02.46 | dan_j | 1 second didnt do it. |
| 17:02.55 | dan_j | just put 5 seconds, should i try to reduce it? |
| 17:03.15 | gtjoseph | i think that's fine |
| 17:03.52 | file | when it goes in we'll find out fast how it behaves on slowwwwwwww systems |
| 17:04.18 | dan_j | Ok. I'm checking the test and it's failing. Will commit the missing change in a moment |
| 17:05.25 | dan_j | Hmm. sometimes it fails and sometimes it works. It's not removing the extension quickly enough once 'alice' unregisters. |
| 17:05.50 | dan_j | Anyway to put a delay between the ami-event and the subsequent action? |
| 17:06.13 | gtjoseph | hmmm |
| 17:06.43 | gtjoseph | are you testing for the CONTACT_REMOVED event? |
| 17:07.03 | dan_j | Yep |
| 17:07.09 | gtjoseph | yeah, just saw |
| 17:07.57 | gtjoseph | so the final ListDialplan is still showing a count of 1 |
| 17:08.30 | dan_j | sometimes yes, sometimes no |
| 17:08.32 | dan_j | <PROTECTED> |
| 17:08.39 | gtjoseph | when you were sniffing the AMI were there any other events you could wait for? |
| 17:08.49 | gtjoseph | that came after the REMOVED |
| 17:09.14 | dan_j | Maybe. I closed that terminal window. I'll run another test to see. |
| 17:14.36 | dan_j | Ok, changed it to something else. Seems ok now. |
| 17:19.56 | dan_j | RE: https://issues.asterisk.org/jira/browse/ASTERISK-25686 is what the user's done considered a close of the issue, or does asterisk need fixing? |
| 17:20.57 | gtjoseph | looks like alembic needs fixing |
| 17:21.15 | gtjoseph | maybe |
| 17:21.22 | gtjoseph | need to look at that script |
| 17:22.01 | gtjoseph | are you volunteering to take that issue now that you know how to mess with alembic? |
| 17:22.18 | dan_j | Yeh, looks like a simple enough issue to take a look at. |
| 17:22.26 | gtjoseph | go for it |
| 17:23.36 | dan_j | Just to check, I'm going to see how chan_sip and realtime sip stores quality internally, will then look at pjsip to see why it's a double and if its just an issue with the db |
| 17:25.58 | dan_j | Do I need to cherrypick the regcontext feature? |
| 17:34.20 | gtjoseph | cherrypick: yes |
| 17:34.36 | gtjoseph | don't look at chan_sip however |
| 17:35.07 | gtjoseph | look at res_pjsip/location.c |
| 17:40.15 | dan_j | ok. |
| 17:41.24 | dan_j | How does cherrypick manage to port the changes? Simply because the files are similar? |
| 17:41.51 | gtjoseph | yep. |
| 17:42.13 | gtjoseph | well, iut's a little more complicated but essentially |
| 17:42.25 | rmudgett | Look at git cherry-pick man page |
| 17:44.17 | dan_j | I just used the web interface to do it |
| 17:44.29 | gtjoseph | that's the best way |
| 17:44.47 | dan_j | Doesn't include the issue number though. |
| 17:45.25 | dan_j | unless i've done it wrong |
| 17:45.35 | gtjoseph | no, that's 12 nit |
| 17:45.38 | gtjoseph | 1 nit |
| 17:45.46 | gtjoseph | you have to add the topic |
| 17:46.57 | dan_j | Ok. Got it. |
| 18:03.57 | dan_j | pjsip stores qualify frequency as an integer, and qualify timeout as a double. But does that mean that asterisk is wrong, or the schema? Is it necessary to have qualify timeout being a double? |
| 18:05.33 | dan_j | I suppose it is because frequency is now in seconds in pjsip, and i assume timeout is the same and needs to be higher resolution to include ms |
| 18:06.54 | gtjoseph | i wrote that and i'm never wrong. :) |
| 18:06.55 | dan_j | gtjoseph: to start with a new issue, do i just issue 'git checkout -b ASTERISK-XXXX' or does that end up adding to my previous work? |
| 18:07.26 | gtjoseph | as long as XXXX is new you're OK |
| 18:10.07 | *** join/#asterisk-dev GameGamer43 (sid5533@gateway/web/irccloud.com/x-ihkjkudeeyfvahoy) |
| 18:14.34 | *** join/#asterisk-dev tm1000 (sid6728@gateway/web/irccloud.com/x-etiazhuziubxzwvq) |
| 18:14.40 | *** join/#asterisk-dev drmessano (sid104353@pdpc/supporter/active/drmessano) |
| 18:17.23 | *** join/#asterisk-dev dan_j (sid21651@gateway/web/irccloud.com/x-owbpidantmycmoin) |
| 18:17.57 | *** join/#asterisk-dev Synthase_ (uid63346@gateway/web/irccloud.com/x-vhzhexdmpcgckovz) |
| 18:19.13 | gtjoseph | dan_j: you back? |
| 18:19.41 | dan_j | Yep. My irc cloud client seems to have been blacklisted on an irc server |
| 18:19.54 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-opnjwborhfkhogqp) |
| 18:20.01 | dan_j | Just sorting out the kids. Will look at the script shortly |
| 18:20.03 | *** join/#asterisk-dev X-Rob (sid14615@gateway/web/irccloud.com/x-amexveozuktdqzvk) |
| 18:20.20 | dan_j | Wow. Didnt realise how many people use irccloud |
| 18:27.10 | coreyfarrell | dan_j: when you run 'git checkout -b ASTERISK-XXXX' it bases that new branch on whatever branch was previously checked out. 'git checkout 13 -b ASTERISK-XXXX' will force it to base the new branch on 13. |
| 18:29.39 | dan_j | Ok. Will redo it. |
| 18:29.43 | *** join/#asterisk-dev moy (sid47040@gateway/web/irccloud.com/x-tolhknkmiernjwop) |
| 18:30.14 | coreyfarrell | dan_j: well if you were on 13 (or whatever base you want) to start then you are fine |
| 18:30.18 | gtjoseph | so now i need an automatic fact-checker as well as an automatic spell checker on my posts. :) |
| 18:32.22 | dan_j | coreyfarrell: Yes. I was on 13 but I was working on a different issue, so now the new issue will be based on the fix of the old issue. |
| 18:33.52 | gtjoseph | dan_j: for the new alembic issue, you want to be directly off of 13. |
| 18:34.14 | gtjoseph | wait, maybe not |
| 18:34.44 | gtjoseph | since your first issue modified alembic you DO need to be based off your earlier issue |
| 18:35.02 | dan_j | But it modified a different script? |
| 18:35.20 | coreyfarrell | dan_j: also about something I saw earlier, the .swp file getting into gerrit is a good example of why I don't use 'git commit -a'. I always use 'git status', then 'git add folder/file' before 'git commit' without the -a. |
| 18:35.28 | gtjoseph | yeah but the version hashs will be messed up so the chain will be nbroken |
| 18:36.05 | gtjoseph | git will be happy but alembic won't |
| 18:36.26 | dan_j | coreyfarrell: noted. I also altered the exclude file to remove swp |
| 18:36.44 | dan_j | Ok. Ill check out my issue and then checkout again for the new issue. |
| 18:37.16 | gtjoseph | the first thing i do when on a new repo is to run my devel tools to see if it generates any unexpected files then add them to .git/info.excludes |
| 18:38.32 | coreyfarrell | gtjoseph: maybe we should add '.*.swp' to /.gitignore of both repos? |
| 18:38.44 | gtjoseph | i was debating that |
| 18:38.54 | gtjoseph | i suppose it's a common enough extension |
| 18:39.03 | gtjoseph | so it wouldn't hurt |
| 18:39.35 | gtjoseph | dan_j: can do that as a followup. :) |
| 18:39.54 | coreyfarrell | I think starting the line with '.' is important that way we only get files that are .filename.swap, not filename.swp |
| 18:40.09 | coreyfarrell | I meant .filename.swp |
| 18:40.41 | gtjoseph | makes sense |
| 18:40.55 | gtjoseph | **/.*.swp |
| 18:41.20 | gtjoseph | i'm off to buy a powerball ticket |
| 18:42.41 | *** join/#asterisk-dev Demon_VoIP (~demon@ip253.net222.n37.ru) |
| 18:43.23 | Demon_VoIP | gtjoseph, https://issues.asterisk.org/jira/browse/ASTERISK-25439 SegFault in the same lines today. |
| 18:43.56 | gtjoseph | ok, i'll look when i get back from lunch |
| 18:44.14 | Demon_VoIP | gtjoseph, Bon appetit |
| 19:34.37 | gtjoseph | Demon_VoIP: so no hang today, just the segfault so i'm going to concentrate on that and we'll handle the hangs separately. |
| 19:40.39 | Demon_VoIP | gtjoseph, yes. I happy not hangs today. And hope segfault will be fixed :) |
| 19:41.08 | gtjoseph | i'm going to give you a pjproject patch to see if we can trap the segfault. |
| 19:41.19 | gtjoseph | probably in an hour |
| 19:41.37 | Demon_VoIP | gtjoseph, ok. i'll recompile and check |
| 19:59.44 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 20:02.05 | dan_j | gtjoseph: just a case of changing sa.Integer to sa.Decimal ? |
| 20:02.32 | gtjoseph | look at one of the other scripts that changes an existing column. |
| 20:02.39 | gtjoseph | i've not done that myself |
| 20:03.05 | dan_j | Oh, i see. I have to make a new script to change the existing column to a decimal. Not just update the old script. |
| 20:03.24 | gtjoseph | correct |
| 20:03.53 | dan_j | ok |
| 20:09.03 | dan_j | since this is a fix, I leave downgrade blank? |
| 20:09.16 | dan_j | or actually, do the same for upgrades and downgrades? |
| 20:11.48 | gtjoseph | look at one of the scripts that changes a column size |
| 20:11.52 | gtjoseph | i'm not sure |
| 20:13.48 | coreyfarrell | dan_j: simple example @ https://github.com/asterisk/asterisk/blob/master/contrib/ast-db-manage/cdr/versions/54cde9847798_expand_accountcode_to_80.py |
| 20:14.30 | dan_j | Is there a quick way to undo a checkout in order to start again? |
| 20:15.17 | gtjoseph | git reset --hard HEAD to clean pending changes |
| 20:15.24 | gtjoseph | or gis teset --HARD <commit> |
| 20:15.32 | gtjoseph | git reset --hard <commit> |
| 20:15.39 | gtjoseph | to get back to a specific point |
| 20:16.30 | coreyfarrell | if you want to go back one commit you can use: git reset --hard HEAD~1 |
| 20:17.21 | coreyfarrell | one thing though 'git reset --hard' does not erase untracked (new) files. |
| 20:18.52 | *** join/#asterisk-dev superscrat (~asanders@173-21-89-217.client.mchsi.com) |
| 20:19.09 | dan_j | Ok. Confused. Just going to reclone and start fresh. I accidentally called the new script 'quality' instead of 'qualify' |
| 20:19.37 | gtjoseph | oh, don't do that |
| 20:19.51 | gtjoseph | just do a 'git checkout 13 -- <filename> |
| 20:20.04 | dan_j | I didnt commit yet |
| 20:20.18 | gtjoseph | just rename it then |
| 20:20.19 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-179-40-172.cust.bezeqint.net) |
| 20:20.19 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 20:20.54 | gtjoseph | oh, unless you have to redo the alembic |
| 20:21.11 | dan_j | oh. ok. didnt realise. hadnt started the code in it yet. |
| 20:21.45 | gtjoseph | you should NEVER have to reclone. |
| 20:21.58 | dan_j | just got to learn the ropes |
| 20:22.04 | gtjoseph | :) |
| 20:22.49 | file | only when you do absolutely naughty insane things with git may you need to reclone, and even then |
| 20:22.56 | coreyfarrell | dan_j: if you pastebin 'git status' we'd know where you are starting from |
| 20:23.26 | gtjoseph | you can also just delete the branch and start over... git checkout 13 ; git branch -D <badbranch> ; git checkout -b <goodbranch> 13 ; |
| 20:23.48 | gtjoseph | and a git branch |
| 20:24.52 | dan_j | ah, thats what i wanted |
| 20:24.54 | dan_j | thanks |
| 20:32.48 | gtjoseph | hey file is there an easy way to get just pjsip debug stuff without getting the rest of asterisk debug stuff? |
| 20:33.12 | file | you may be able to specify the filename option to core set debug (I don't use it, don't remember the format) and limit it to pjsip |
| 20:33.36 | gtjoseph | yeah, tried that |
| 20:33.39 | gtjoseph | oh well |
| 20:39.00 | coreyfarrell | gtjoseph: I think it's actually done by module, not filename. |
| 20:39.37 | coreyfarrell | gtjoseph: so 'core set debug 6 res_pjsip', not res_pjsip.c |
| 20:39.51 | gtjoseph | oh, no kidding. thanks. :) |
| 20:40.16 | gtjoseph | i always forget that tab completion lies for debug |
| 20:40.43 | coreyfarrell | I can't say if that actually works, but that's what the code looks like it should do |
| 20:42.00 | gtjoseph | it does work |
| 20:42.18 | gtjoseph | i get pjproject plis res_pjsip |
| 20:42.20 | gtjoseph | plus |
| 20:43.19 | coreyfarrell | gtjoseph: probably want to enable debug for res_pjsip_log_forwarder ? |
| 20:43.59 | gtjoseph | nope |
| 20:44.03 | gtjoseph | res_pjsip it is |
| 20:45.21 | gtjoseph | Demon_VoIP: https://pastebin.com/u9zyStK0 patch to pjproject 2.5.4. apply with patch -p1. Also building pjproject with DEBUG would help. |
| 20:45.52 | gtjoseph | maybe we can catch the parameters to the dns calls |
| 20:50.42 | gtjoseph | Demon_VoIP: oops, that's pjproject 2.4.5 |
| 20:53.18 | Demon_VoIP | what's wrong? |
| 20:53.48 | gtjoseph | not sure yet but the patch adds some debugging message and tests |
| 20:54.21 | gtjoseph | also, make sure yo do a 'make clean |
| 20:54.41 | gtjoseph | and re-run configure |
| 20:55.04 | Demon_VoIP | and where the log will be? |
| 20:55.22 | gtjoseph | same as now |
| 20:57.31 | gtjoseph | hmmm. the log may be an issue,,, you might have to set 'core set debug 5 res_pjsip' |
| 20:59.45 | gtjoseph | Demon_VoIP: actually, i'm going to give you a new patch |
| 21:00.46 | Demon_VoIP | :) ok. i am waiting |
| 21:01.02 | gtjoseph | 5 minutes |
| 21:07.04 | gtjoseph | Demon_VoIP: new patch https://pastebin.com/MmS9VBvX |
| 21:08.18 | gtjoseph | so now, the debug message will show up as asterisk warnings without you having to 'core set debug' |
| 21:10.19 | gtjoseph | for the team: i didn't know this but if yo're messing with pjproject, you can do 'PJ_LOG(1, ...' and the messages will show as asterisk errors, and 'PJ_LOG(2, ...' will show as warnings |
| 21:37.31 | *** join/#asterisk-dev superscrat (~asanders@173-21-89-217.client.mchsi.com) |
| 21:51.09 | dan_j | gtjoseph: whats your thoughts on this bug? Should it do the same for both upgrades and downgrades? |
| 21:51.41 | dan_j | actually, i'll look at a script thats a fix and see what it does. |
| 21:51.47 | gtjoseph | i'm going to have to defer to file on that one |
| 21:51.56 | file | I don't understand the question |
| 21:52.15 | file | despite being wrong, a downgrade should revert it back to the way it was before the upgrade |
| 21:52.15 | gtjoseph | he's doing a fix where the type of an alembic column has changed |
| 21:52.23 | gtjoseph | there you go |
| 21:52.39 | dan_j | ok. noted |
| 22:10.18 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 22:45.48 | file | maybe I'll look at the issue I filed for making CI check alembic this weekend... |
| 22:46.11 | gtjoseph | that'd be nice |
| 22:47.04 | file | I should check the CentOS agents to see if they have it |
| 22:47.21 | file | the answer is no |
| 22:47.29 | file | superscrat, hi |
| 22:48.19 | gtjoseph | dan_j: i think there's an issue with 2008 |
| 22:48.44 | dan_j | cant cherrypick alembic? |
| 22:48.49 | file | goes to give superscrat more work |
| 22:49.04 | gtjoseph | only if the histories are the same |
| 22:49.18 | dan_j | in that case maybe my other fix has a similar problem. |
| 22:49.20 | gtjoseph | actually 2007 might have an issue as well |
| 22:49.21 | dan_j | the regcontext one |
| 22:49.30 | dan_j | oh dear |
| 22:49.30 | gtjoseph | no, that one's ok |
| 22:50.10 | gtjoseph | from the contrib/ast-db-manage directory, do 'alembic -c ./config.ini.sample branches' |
| 22:51.22 | gtjoseph | actually 2007 is OK |
| 22:52.00 | dan_j | I need to git review -d 2008 first? |
| 22:52.07 | gtjoseph | yep |
| 22:52.12 | dan_j | actually, i only have the 13 branch cloned |
| 22:52.42 | gtjoseph | doesn't matter |
| 22:52.57 | gtjoseph | when you do a clone, you get everything |
| 22:53.27 | gtjoseph | when you do a git review -d, the correct branch will be created for you |
| 22:53.41 | dan_j | ok. with 2007, i didnt get anything. with 2008, i get UserWarning: Revision 136885b81223 referenced from 136885b81223 -> 423f34ad36e2 (head), fix pjsip qualify timeout is not present |
| 22:54.57 | *** join/#asterisk-dev putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 22:54.57 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 22:55.25 | gtjoseph | i think what you need to do is do a clean git review -d 2003 to get your last master patch |
| 22:55.58 | gtjoseph | hmmm. hold on a sec and let me look |
| 22:56.38 | coreyfarrell | gtjoseph: that shouldn't be required. when you do 'git review -d' it should do a fetch from gerrit, pulling the actual commit (and all that lead to it) |
| 22:57.13 | gtjoseph | he hqas to re-run alembic revision though |
| 22:57.29 | gtjoseph | so i think he needs to start with a clean base again |
| 22:57.32 | coreyfarrell | oh sorry I'll be quiet |
| 22:57.50 | dan_j | Just checked, i have the script there with the following |
| 22:57.52 | dan_j | Revision ID: 423f34ad36e2 |
| 22:57.53 | dan_j | Revises: 136885b81223 |
| 22:58.11 | file | it's a cherry pick, so it doesn't have the change to add the regcontext stuff as a dependency |
| 22:58.13 | gtjoseph | yep but 13688 doesn't exist in master |
| 22:58.22 | gtjoseph | exactly |
| 22:58.49 | gtjoseph | so do a git review -d on the base which is 2003 in your case |
| 22:59.19 | gtjoseph | then re-run the alembic revision |
| 22:59.22 | dan_j | ok. i get it. |
| 22:59.42 | gtjoseph | and copy/paste your changes into the new file |
| 23:01.27 | dan_j | http://pastebin.com/iCD1tyPJ |
| 23:03.04 | gtjoseph | rats, 2004 i'm sorry |
| 23:03.09 | gtjoseph | 2003 is the test |
| 23:04.08 | dan_j | nuts. and i also checked that it was 2003. |
| 23:05.16 | dan_j | how do i switch to master? |
| 23:05.57 | gtjoseph | git checkout master |
| 23:06.15 | dan_j | oh never mind. 2004 is already on master so i'm already on master now. correct? |
| 23:06.22 | gtjoseph | since it's a known branch, yo don't have to do the '-b' |
| 23:06.29 | gtjoseph | correct |
| 23:07.23 | gtjoseph | if yo want to double check, you can do a 'git diff master' to see the difference |
| 23:07.42 | gtjoseph | it should only be the changes from your 2004 review |
| 23:09.33 | dan_j | fatal: ambiguous argument 'master': unknown revision or path not in the working tree. |
| 23:10.02 | gtjoseph | ah, nevermind,. you don't have master locally yet. don't worry about it |
| 23:11.01 | *** join/#asterisk-dev DivideBy0 (~DivideBy0@unaffiliated/divideby0x0) |
| 23:11.20 | dan_j | the script from 13 is still there so it wont let me add one with the same na,e |
| 23:11.21 | dan_j | name |
| 23:11.38 | dan_j | git status doesnt say anything about disconnected files. |
| 23:11.39 | *** join/#asterisk-dev lnxslck (~lnxslck@av3-84-90-160-135.netvisao.pt) |
| 23:12.23 | gtjoseph | do a git remote -v for me |
| 23:12.38 | gtjoseph | and a git branch and put it in a pastebin |
| 23:12.59 | gtjoseph | and git status while you're at it. :) |
| 23:14.57 | dan_j | http://pastebin.com/zHWQ7N3D |
| 23:15.34 | gtjoseph | ok, do a git checkout master |
| 23:16.05 | dan_j | error: pathspec 'master' did not match any file(s) known to git |
| 23:16.06 | gtjoseph | it should create a local branch 'master' that tracks gerrit/master |
| 23:16.20 | gtjoseph | ok, do a git checkout gerrit/master |
| 23:16.20 | dan_j | i originally cloned using -b 13 |
| 23:16.34 | dan_j | sorry, remove the -b |
| 23:16.36 | gtjoseph | that should be ok |
| 23:16.49 | dan_j | You are in 'detached HEAD' state. |
| 23:16.54 | dan_j | so now review -d? |
| 23:17.19 | gtjoseph | sigh |
| 23:17.59 | gtjoseph | ok, do a git fetch gerrit |
| 23:18.55 | dan_j | oh dear. i'm really sorry to mess you around. |
| 23:19.02 | gtjoseph | then git checkout -b master |
| 23:19.03 | dan_j | I've been on the wrong terminal window |
| 23:19.09 | gtjoseph | HA! |
| 23:19.20 | dan_j | one sec. actually. not for that long |
| 23:20.50 | dan_j | Heres the pastebin from the correct server |
| 23:20.51 | dan_j | http://pastebin.com/wpSstLYY |
| 23:21.11 | gtjoseph | ok, that's fine |
| 23:21.26 | gtjoseph | do git checkout -b master gerrit/master |
| 23:21.40 | dan_j | ok |
| 23:21.46 | dan_j | switched to master branch |
| 23:21.50 | gtjoseph | perfectr |
| 23:22.06 | gtjoseph | now delete the 2 review/* branches |
| 23:23.43 | dan_j | The branch 'review/daniel_journo/ASTERISK-25686' is not fully merged. |
| 23:23.47 | dan_j | just delete anyway? |
| 23:23.54 | gtjoseph | yep, use -D |
| 23:24.05 | dan_j | ok, both gone |
| 23:24.33 | gtjoseph | cool. now go ahead and do the git review -d 2004 |
| 23:25.53 | dan_j | I've got another branch called 'ASTERISK-25686' should that be deleted too? |
| 23:26.06 | gtjoseph | yes actually |
| 23:26.15 | gtjoseph | because you're going to recreate it now |
| 23:29.27 | gtjoseph | when you're ready and assuming you're in the eview/daniel_journo/ASTERISK-25670 branch... |
| 23:29.49 | gtjoseph | just to a git checkout -b ASTERISK-25686 |
| 23:33.02 | dan_j | hmmm |
| 23:33.09 | dan_j | <PROTECTED> |
| 23:33.15 | dan_j | Only a single head supported so far... |
| 23:33.35 | gtjoseph | never seen that before |
| 23:33.44 | gtjoseph | did it create the file? |
| 23:34.47 | dan_j | cant see it, but i do have a pyc file which is based on the old script |
| 23:35.02 | gtjoseph | ignroe that |
| 23:35.37 | dan_j | Googe |
| 23:35.39 | dan_j | google |
| 23:35.40 | dan_j | This error means you have two revisions referencing the same down_revision. Just look at the last few revisions you've created and see which ones reference the same down_revision |
| 23:36.06 | gtjoseph | oh yeah, branches isn't clean |
| 23:36.41 | dan_j | quicker just to clone master? |
| 23:37.02 | gtjoseph | if you do you'll have an issue |
| 23:37.26 | gtjoseph | maybe better to wait until your other changes are merged though |
| 23:38.47 | gtjoseph | so i'd actually put 2008 in draft status (click the WIP button) until then |
| 23:39.26 | dan_j | ok. can do. |
| 23:39.47 | gtjoseph | then when 1984, 2003 and 2004 get merged yo can rebase 2008 directly on master |
| 23:40.05 | dan_j | so the merge will continue for 2007? |
| 23:40.30 | dan_j | or should i WIP that one too? |
| 23:42.14 | gtjoseph | that one's ok so i'd leave it |
| 23:42.50 | dan_j | ok. thanks for your help. i'll monitor and add it once the other stuffs merged |
| 23:52.24 | *** join/#asterisk-dev infobot (ibot@rikers.org) |
| 23:52.24 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Hell hath frozen over: Asterisk is moving to Git -=- Follow on Twitter at @AsteriskDev |