| 00:19.59 | dan_j | I was really looking for a one line description of each function. Many arent documented and hard to understand if its the first time looking at the source code. |
| 00:22.53 | coreyfarrell | dan_j: unfortunately if it's not documented in the code, it's not documented :( |
| 00:23.45 | coreyfarrell | most documentation is in the headers, some is in the sources |
| 00:45.05 | dan_j | Ok. I guess im trying to run before i can walk. Will keep at it. Thanks. |
| 01:12.14 | file | it also depends on the edge and use |
| 01:12.16 | file | er age |
| 05:20.35 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 09:09.42 | *** join/#asterisk-dev pchero (~pchero@109.70.54.56) |
| 10:31.52 | dan_j | Whats the purpose of autoservice ? Specifically, ast_autoservice_chan_hangup_peer , ast_autoservice_start and ast_autoservice_stop ? |
| 10:34.46 | dan_j | from what i can figure out, thats destroying old channels that are no longer in use, but trying to figure out what start and stop do. |
| 10:41.49 | dan_j | I'm looking at this bug as a learning exercise. https://issues.asterisk.org/jira/browse/ASTERISK-25691 |
| 10:43.04 | dan_j | In app_dial.c do_privacy function, I can see it doing an ast_autoservice_start and ast_autoservice_stop and then an ast_autoservice_chan_hangup_peer. But in order source code I've looked at, it only has the hangup function. Just wondering how they relate. |
| 10:44.58 | dan_j | order = other |
| 11:13.04 | *** join/#asterisk-dev pchero (~pchero@109.70.54.56) |
| 13:09.30 | *** join/#asterisk-dev hehol (~Adium@ip-178-203-97-214.hsi10.unitymediagroup.de) |
| 13:44.12 | file | dan_j, app_dial is one of the largest complicated things due to its feature set and age - autoservice on the other hand isn't that bad, it hands a channel off to another thread to have it be serviced for them while the current thread does something on another channel |
| 13:44.34 | file | ast_autoservice_chan_hangup_peer does that whilst another channel is hung up |
| 13:45.14 | file | in this case because of hangup handlers, as they execute dialplan logic they may take a long time to run |
| 13:48.46 | dan_j | Autoservice only deals with hangups, or other things too? |
| 13:50.03 | file | it defers hangups, it's purpose is to read frames from a channel so the queue doesn't get backed up |
| 13:50.32 | file | generally a frame for every 20ms of audio, so gotta keep reading it in (even if being discarded) |
| 13:52.25 | file | ast_autoservice_chan_hangup_peer is a helper function for a common pattern that had to occur as a result of the hangup handler functionality going in |
| 13:58.31 | dan_j | Ok. Ill try to look at the autoservice code again to try to understand, now that youve given me some background |
| 14:00.26 | dan_j | Oh. Also, whats the purpose of ao2 objects? A storage system? |
| 14:01.16 | file | it's a reference counting and container system for objects |
| 14:34.19 | dan_j | ok. thanks for the help. |
| 16:06.47 | *** join/#asterisk-dev lnxslck (~lnxslck@84.90.160.135) |
| 16:09.37 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 16:17.52 | *** join/#asterisk-dev gtjoseph (~gtj0@unaffiliated/gtj) |
| 17:05.58 | *** join/#asterisk-dev lnxslck (~lnxslck@av3-84-90-160-135.netvisao.pt) |
| 17:19.14 | *** join/#asterisk-dev craigify (~craigify@ip184-186-18-123.no.no.cox.net) |
| 17:42.33 | file | neat, I have enough bandwidth to load gerrit |
| 18:44.54 | *** join/#asterisk-dev pchero (~pchero@109.70.54.56) |
| 19:32.58 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 21:45.01 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-179-40-172.cust.bezeqint.net) |
| 21:45.01 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 22:00.23 | coreyfarrell | does anyone know how to add an 'int' to a hash? We have ast_str_hash / ast_str_hash_add, I'm looking for ast_int_hash_add. |
| 22:03.49 | file | define add... |
| 22:07.02 | coreyfarrell | when you already have a hash and want to resume calculating the hash by adding more data |
| 22:07.55 | coreyfarrell | the goal is to produce a hash for an ao2_container using multiple fields (strings and at least one int). |
| 22:08.02 | file | there is no hashing algorithm in the tree based around that, you'd be making your own |
| 22:08.27 | file | provided it produces the same value given the same set of data and is reasonably distributed, then it's fine |
| 22:12.55 | coreyfarrell | ok and for the purpose of ao2_container I don't need to worry about byte ordering, the hash never needs to be stable across different platforms? |
| 22:13.08 | file | it does not |
| 22:19.18 | coreyfarrell | so I'm thinking I should just create functions similar to ast_str_hash / ast_str_hash_add that will operate on (void* buffer, size_t len), use len to control the loop instead of checking for '\0' |
| 22:47.49 | *** join/#asterisk-dev opticron (~opticron@75.76.201.152) |
| 22:47.49 | *** mode/#asterisk-dev [+o opticron] by ChanServ |
| 22:57.44 | *** join/#asterisk-dev craigify (~craigify@ip184-186-18-123.no.no.cox.net) |
| 22:57.44 | *** join/#asterisk-dev gtjoseph (~gtj0@unaffiliated/gtj) |
| 22:57.44 | *** join/#asterisk-dev moy (sid47040@gateway/web/irccloud.com/x-tolhknkmiernjwop) |
| 22:57.44 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-opnjwborhfkhogqp) |
| 22:57.44 | *** join/#asterisk-dev drmessano (sid104353@pdpc/supporter/active/drmessano) |
| 22:57.44 | *** join/#asterisk-dev jrrose (jon@nat/digium/x-atzgjfajtnrtykad) |
| 23:17.49 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 23:25.03 | *** join/#asterisk-dev snuff-work (~blatus@203.87.61.22) |
| 23:25.03 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |