| 00:21.23 | *** join/#asterisk-dev pabelanger (~pabelange@asterisk/contributor-and-bug-marshal/pabelanger) |
| 00:21.23 | *** mode/#asterisk-dev [+o pabelanger] by ChanServ |
| 00:24.36 | *** part/#asterisk-dev alecdavis (~alecdavis@202.36.76.21) |
| 00:25.52 | *** join/#asterisk-dev italorossi (~italoross@187.61.168.117) |
| 00:32.34 | *** join/#asterisk-dev suneye (~atcmmi@116.7.102.64) |
| 02:27.52 | *** part/#asterisk-dev w9sh (~sph@50-79-224-193-static.hfc.comcastbusiness.net) |
| 03:28.32 | *** join/#asterisk-dev moy (~moy@UNVLON55-1176057127.sdsl.bell.ca) |
| 03:29.47 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 03:31.10 | *** join/#asterisk-dev zerohalo (~zerohalo@74.61.196.236) |
| 03:43.26 | *** join/#asterisk-dev serafie (~erin@user-24-214-173-250.knology.net) |
| 03:58.49 | *** join/#asterisk-dev jrose_atDigium (~jrose_atD@nat/digium/x-bqwdrfegkfabsxnw) |
| 05:49.39 | *** join/#asterisk-dev zerohalo (~zerohalo@74.61.196.236) |
| 06:27.44 | *** join/#asterisk-dev oej (~Adium@2001:16d8:cc57:1000::42:1006) |
| 06:27.45 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 06:44.07 | *** join/#asterisk-dev bulkorok (~chatzilla@gw1.pinguin.ag) |
| 06:49.04 | *** join/#asterisk-dev oej (~Adium@2001:16d8:cc57:1000::42:1006) |
| 06:49.04 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 08:29.07 | *** join/#asterisk-dev oej (~Adium@2001:16d8:cc57:1000::42:1006) |
| 08:29.08 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 09:36.18 | *** join/#asterisk-dev italoros_ (~italoross@187.61.168.117) |
| 09:42.52 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 09:42.52 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 10:04.23 | *** join/#asterisk-dev tonyclewis (uid6025@gateway/web/irccloud.com/x-lpwjmciofhegjfcj) |
| 10:44.42 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 10:44.42 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 10:46.55 | *** join/#asterisk-dev oej (~Adium@h87-96-134-129.dynamic.se.alltele.net) |
| 10:46.56 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 11:02.40 | *** join/#asterisk-dev Greenlight (~email@cpc1-dund9-0-0-cust142.16-4.cable.virginmedia.com) |
| 11:13.34 | *** join/#asterisk-dev w9sh (~sph@50-79-224-193-static.hfc.comcastbusiness.net) |
| 12:15.38 | *** join/#asterisk-dev seanbright (~sean@asterisk/contributor-and-bug-marshal/seanbright) |
| 12:15.38 | *** mode/#asterisk-dev [+o seanbright] by ChanServ |
| 12:56.13 | *** join/#asterisk-dev serafie (~erin@nat/digium/x-qrnjaavwprpcpuyi) |
| 12:59.33 | *** join/#asterisk-dev serafie1 (~erin@user-24-214-173-250.knology.net) |
| 13:31.11 | *** join/#asterisk-dev GameGamer43 (uid5533@gateway/web/irccloud.com/x-jjdkwcxlvqyamrzy) |
| 13:51.17 | *** join/#asterisk-dev malcolmd (~malcolmd@pdpc/sponsor/digium/malcolmd) |
| 13:51.17 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 13:53.53 | *** join/#asterisk-dev putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 13:53.53 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 13:55.23 | *** join/#asterisk-dev mjordan (~mjordan@nat/digium/x-jqufhjycpzacowos) |
| 13:55.23 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 13:56.01 | *** join/#asterisk-dev sruffell (~sruffell@asterisk/the-kernel-guy/sruffell) |
| 13:56.01 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 13:57.22 | file | g'day all |
| 13:59.41 | snuff-home | morning... |
| 14:00.36 | file | owww I hit my shoulder on a pointy part of my laptop |
| 14:01.05 | snuff-home | is trying to think how'd u'd manage that |
| 14:01.29 | file | it involves being on a bus and my legs not being on the floor |
| 14:01.41 | snuff-home | ah |
| 14:02.38 | file | snuff-home, I remember reading my scrollback awhile ago and seeing you talking about stasis and not getting it... I think... what didn't you get? |
| 14:04.10 | snuff-home | so is there 'grunt work' for stasis that can be done.. similar to cli/manager cmds for pjsip/gulp/whatever we calling it |
| 14:07.22 | file | that I fail to have an answer to really |
| 14:07.30 | file | jrose_atDigium, poke |
| 14:08.10 | snuff-home | also will re-test my obsd issue with the extra fix again.. been rather busy with real work .. geoservers and SLD (xml) |
| 14:09.38 | opticron | the biggest loss that I can think of in moving from the event system to stasis was the cache dumping CLI commands that the event system had built-in |
| 14:11.06 | opticron | snuff-home, one other thing that comes to mind is refactoring things like "core show channels" that currently access channels directly |
| 14:11.37 | opticron | ideally, those commands would read from the channel state cache |
| 14:11.49 | opticron | or bridge state cache or whatever else |
| 14:12.47 | mjordan | I'll second opticron's point. We really need to have the channel/bridge inspection commands be hitting stasis cache and not the live objects |
| 14:13.02 | mjordan | that would be a *fantastic* patch, and should address both CLI and AMI commands |
| 14:13.10 | file | oh heck yes |
| 14:13.11 | opticron | :D |
| 14:13.55 | mjordan | the great thing about that is it eliminates two major bugs that Asterisk has always suffered from: (1) deadlocks caused by CLI/AMI commands and their threads obtaining the channel container; (2) performance impacts while a CLI/AMI command holds the channel container, blocking all channel drivers or other modules from accessing channels |
| 14:14.15 | sruffell | "We're here live in asterisk-dev, folks, where excitement is building on a potential patch snuff-home could work on…." |
| 14:14.24 | mjordan | I should ammend that: Asterisk hasn't always suffered from that. But it's happened a lot. |
| 14:14.36 | mjordan | sruffell :-) |
| 14:15.05 | mjordan | that patch will get written at some point, but it'd be a great help (and a good intro to "how to get things out of Stasis-caches") |
| 14:15.16 | snuff-home | heh.. i'm not the most reliable at getting stuff done.. but i can always try.. |
| 14:15.19 | sruffell | That does actually sound pretty cool to me and I'm not even up on the internals. I just know I have seen slow CLI commands in the past... |
| 14:16.29 | mjordan | The Stasis cache maintains a snapshot of the most relevant channel/bridge information (and other things as well) - so consumers of channel/bridge state can hit the cache as opposed to impacting objects that should only be manipulated by threads that need real-time performance assurances |
| 14:17.15 | snuff-home | mm so we have a pointer to the statis-cache stuff? |
| 14:17.30 | mjordan | snuff-home: one thing you'll find is that certain commands - like 'core show channel foo' - may display a bit more information than you can get out of the cache. For now, my preference would be to just remove the extra data shown, and deal with the small loss of information |
| 14:17.39 | mjordan | snuff-home: sure, give me just a second |
| 14:17.41 | snuff-home | k |
| 14:18.32 | snuff-home | i hope to publish over w/end to reviewboard at least the 'cli' versions of show endpoint(s)/transport(s)/aors |
| 14:18.36 | mjordan | Overview of Stasis (which doesn't really talk about the cache, but still provides a good idea of what is going on): https://wiki.asterisk.org/wiki/display/AST/Stasis+Message+Bus |
| 14:18.43 | mjordan | snuff-home: :-) |
| 14:20.04 | opticron | snuff-home, for actually pulling information out of the cache, stasis_cache_get and stasis_cache_dump in stasis.h will be important |
| 14:20.07 | mjordan | Example of getting channel state from the cache: ast_channel_snapshot_get_latest in stasis_channels.c |
| 14:20.21 | mjordan | and what opticron said :-) |
| 14:22.17 | opticron | oh...that's going to be a significant change in functionality |
| 14:22.24 | mjordan | opticron: leedm777: looking at the stasis cache API, we may eventually want the ability to query the cache with a provided callback function. For example, 'core show channels foo' is going to provide the channel name, not the channel unique ID. It'd be easiest to have an API call that will do an ao2_callback and use a provided callback function |
| 14:22.27 | opticron | since the cache keys on uniqueids |
| 14:22.50 | mjordan | we can't change that. People are used to channel names; channel unique IDs won't work for CLI commands. |
| 14:22.54 | opticron | yeah |
| 14:23.01 | opticron | that's what I figured |
| 14:23.37 | mjordan | if we allowed someone to query the cache with a passed in callback function - compatible with ao2_callback - it'd be easy to allow querying on channel names |
| 14:23.42 | opticron | mjordan, I tried to add a callback approach for dumping when I added it, but that got shot down :) |
| 14:24.00 | mjordan | well, we can discuss it again when leedm777 gets in. We have a use case now :-) |
| 14:24.01 | sruffell | Just out of curiosity, Is the CLI still explicitly excluded from the definition of a published API? I.e. nobody should be scripting against it since it's free to change at any time, even on a stable branch? |
| 14:24.05 | opticron | so I returned a container which you can run callbacks on as well |
| 14:24.26 | opticron | sruffell, correct |
| 14:24.35 | opticron | you really shouldn't be scripting against the CLI |
| 14:24.39 | sruffell | ok, just checking. |
| 14:24.42 | mjordan | sruffell: that's my opinion, but less because it's free to change. It's purpose was to provide human readable output, not script output. If you need script output, imo, you should be using the appropriate facilities for that |
| 14:24.55 | file | indeed |
| 14:24.55 | sruffell | nods |
| 14:24.59 | mjordan | opticron: I don't like returning a container |
| 14:25.07 | file | tharrr be dragons down there |
| 14:25.20 | opticron | and now, scripting will work really well with Stasis-HTTP |
| 14:25.21 | mjordan | opticron: exposing the container means you've broken your abstraction layer completely |
| 14:25.35 | opticron | mjordan, it's not exposing the container directly, it generates a new one |
| 14:25.40 | opticron | also, I agree |
| 14:25.50 | mjordan | returning an iterator is usually preferable |
| 14:26.07 | opticron | ao2_iterators make me cringe |
| 14:26.14 | mjordan | why? |
| 14:26.24 | opticron | because they don't follow the model of other ao2 objects |
| 14:26.30 | mjordan | how so? |
| 14:26.34 | opticron | they themselves are not ao2 objects |
| 14:26.44 | mjordan | meh? |
| 14:27.07 | mjordan | it makes it easy to declare an iterator on the stack, which is usually where you want it |
| 14:27.15 | sruffell | opticron: Think it's worth changing the DAHDI tests to use trunk / Stasis at this point? |
| 14:27.30 | *** join/#asterisk-dev kharwell (~kharwell@nat/digium/x-aahulyiliyrmxvul) |
| 14:27.42 | mjordan | allocating iterators out on the heap when they're scoped to live within a single stack frame 99.999% of the time seems a bit wasteful |
| 14:28.07 | opticron | sruffell, probably, it'll mean less breakage in the future and you may have to do a partial rewrite for 12 anyway |
| 14:28.58 | sruffell | kk |
| 14:29.16 | opticron | a significant chunk of AMI is changing |
| 14:40.53 | *** join/#asterisk-dev moy (~moy@UNVLON55-1176057127.sdsl.bell.ca) |
| 14:40.59 | *** part/#asterisk-dev russellb (~russellb@redhat/russellb) |
| 14:45.39 | *** join/#asterisk-dev moy (~moy@UNVLON55-1176057127.sdsl.bell.ca) |
| 14:48.12 | *** join/#asterisk-dev moy (~moy@UNVLON55-1176057127.sdsl.bell.ca) |
| 14:59.30 | file | The answer is Canada. |
| 15:08.55 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 15:14.38 | file | jrose_atDigium, moo |
| 15:33.54 | Qwell | mjordan: Since nothing is actually using ast_manager_publish_message(), I think I'm going to make it ast_manager_publish_event(), which adds the type and class to the args, and lets you send a blob directly. |
| 15:34.33 | Qwell | I was using it in my queues stuffs, but don't need it there. It was super handy. int ast_manager_publish_event(const char *type, int class_type, struct ast_json *obj); |
| 15:41.22 | mjordan | works for me |
| 15:42.25 | file | mjordan, I think security in gulp will end up being a core part... and not a separate module |
| 15:42.38 | mjordan | works for me |
| 15:42.53 | mjordan | just make it extensible :-) |
| 15:43.00 | file | yeah |
| 15:43.05 | file | using XML! |
| 15:44.57 | Qwell | eXtensible to the eXtreme! |
| 15:56.35 | *** join/#asterisk-dev WIMPy (~wimpy@e183095026.adsl.alicedsl.de) |
| 16:39.02 | *** join/#asterisk-dev serafie (~erin@user-24-214-173-250.knology.net) |
| 16:57.02 | Qwell | mjordan: https://reviewboard.asterisk.org/r/2577/ fwiw |
| 18:11.50 | putnopvut | Reviewboard is being a bit strange right now. Trying to post my features config changes, and this happens: http://fpaste.org/15568/99374871/ |
| 18:12.25 | file | O.o |
| 18:12.31 | putnopvut | That was my reaction, too! |
| 18:13.01 | file | schnitzel sandwich or meatball sandwich... |
| 18:13.03 | file | decisions, decisions |
| 18:17.18 | *** join/#asterisk-dev angler (~angler@pdpc/sponsor/digium/angler) |
| 18:19.14 | *** join/#asterisk-dev leedm777 (~leedm777@nat/digium/x-swgfnfmwxkwfzffn) |
| 18:30.03 | leifmadsen | file: always meatball |
| 18:30.24 | file | leifmadsen, I went for schnitzel |
| 18:31.52 | leifmadsen | then you've made a huge mistake |
| 18:32.30 | file | I don't make mistakes. I make informed decisions. |
| 18:37.05 | opticron | file made the correct decision |
| 18:43.03 | *** join/#asterisk-dev acidfu (~nib@unaffiliated/acidmen) |
| 18:53.58 | *** join/#asterisk-dev pabelanger (~pabelange@asterisk/contributor-and-bug-marshal/pabelanger) |
| 18:53.58 | *** mode/#asterisk-dev [+o pabelanger] by ChanServ |
| 20:47.39 | *** join/#asterisk-dev plUmbro (~gesnaud@cdf35-1-78-230-252-134.fbx.proxad.net) |
| 20:53.33 | *** join/#asterisk-dev oron_laptop (~oron_lapt@line25-214.adsl.actcom.co.il) |
| 20:53.40 | *** part/#asterisk-dev oron_laptop (~oron_lapt@line25-214.adsl.actcom.co.il) |
| 21:10.32 | *** join/#asterisk-dev elguero (~miguel323@2001:470:1f06:12c4::2) |
| 21:49.47 | *** part/#asterisk-dev mjordan (~mjordan@nat/digium/x-jqufhjycpzacowos) |
| 22:23.41 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-28-58.cablep.bezeqint.net) |
| 22:23.42 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 22:25.54 | *** join/#asterisk-dev pabelanger (~pabelange@asterisk/contributor-and-bug-marshal/pabelanger) |
| 22:25.54 | *** mode/#asterisk-dev [+o pabelanger] by ChanServ |
| 22:37.11 | *** join/#asterisk-dev tm1000 (tm1000@2600:3c01::f03c:91ff:fe70:fce6) |
| 22:42.44 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-28-58.cablep.bezeqint.net) |
| 22:42.44 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 23:32.59 | *** join/#asterisk-dev snuff-work (~snuffy@110.141.36.196) |
| 23:33.00 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 23:37.13 | snuff-work | evening ppls.. |
| 23:37.36 | snuff-work | the BSD fix.. almost worked small copy n paste error |