| 00:04.22 | *** join/#asterisk-dev Deeewayne (~dwayne@c-71-207-214-190.hsd1.al.comcast.net) |
| 00:04.22 | *** mode/#asterisk-dev [+o Deeewayne] by ChanServ |
| 00:19.29 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 00:19.29 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 00:37.27 | *** join/#asterisk-dev lmsteffan (~laurent@reef.ac-noumea.nc) |
| 00:37.31 | *** join/#asterisk-dev lmsteffan_ (~lmsteffan@reef.ac-noumea.nc) |
| 01:12.09 | *** join/#asterisk-dev moy (~moy@CPE001cdfec4cee-CM00080dab8485.cpe.net.cable.rogers.com) |
| 02:00.11 | *** join/#asterisk-dev jeffg (~jeffg@pdpc/supporter/professional/jeffg) |
| 02:43.36 | *** join/#asterisk-dev elguero (~miguel323@ns1.nashuacs.com) |
| 02:45.13 | *** join/#asterisk-dev mog (~mog@c-68-62-169-225.hsd1.al.comcast.net) |
| 02:45.13 | *** mode/#asterisk-dev [+o mog] by ChanServ |
| 03:39.30 | *** join/#asterisk-dev alecdavis (~alec@202.36.76.21) |
| 05:54.24 | *** join/#asterisk-dev mog (~mog@c-68-62-169-225.hsd1.al.comcast.net) |
| 05:54.24 | *** mode/#asterisk-dev [+o mog] by ChanServ |
| 07:13.38 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 07:13.38 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 07:36.08 | wedhorn | mvanbaak: how's that skinny _transmit stuff coming along |
| 08:22.40 | *** join/#asterisk-dev elguero (~miguel323@ns1.nashuacs.com) |
| 08:39.39 | *** join/#asterisk-dev tamiel (~tamiel@213.30.183.226) |
| 09:02.36 | *** join/#asterisk-dev felipe_ (~felipe@my.nada.kth.se) |
| 11:25.35 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 11:25.35 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 11:26.59 | snuff-home | mornin kp |
| 11:50.00 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 11:50.00 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 12:27.32 | *** join/#asterisk-dev caio1982 (~caio1982@CAcert-br/caio1982) |
| 12:45.48 | *** join/#asterisk-dev eliel (~eliels@201.234.94.226) |
| 12:45.48 | *** mode/#asterisk-dev [+o eliel] by ChanServ |
| 13:26.41 | *** join/#asterisk-dev elguero (~miguel323@ns1.nashuacs.com) |
| 13:38.42 | *** join/#asterisk-dev bklang_ (~bklang@tesla.alkaloid.net) |
| 13:53.03 | *** join/#asterisk-dev moy (~chatzilla@74.12.131.89) |
| 14:07.37 | *** join/#asterisk-dev pabelanger (~pabelange@yoda.kanatek.com) |
| 14:07.48 | pabelanger | morning |
| 14:08.06 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 14:08.06 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 14:48.01 | *** join/#asterisk-dev eliel (~eliels@201.234.94.226) |
| 14:48.01 | *** mode/#asterisk-dev [+o eliel] by ChanServ |
| 14:50.46 | eliel | hello all |
| 14:52.46 | leifmadsen | howdy! |
| 14:55.21 | *** join/#asterisk-dev jpeeler (~jpeeler@asterisk/digium-software-dev/jpeeler) |
| 14:55.21 | *** mode/#asterisk-dev [+o jpeeler] by ChanServ |
| 14:55.56 | eliel | leifmadsen: hey! |
| 14:57.05 | leifmadsen | how goes today? |
| 14:59.24 | eliel | great! and you? |
| 15:00.11 | eliel | how about the xml docs? do you have something done? |
| 15:02.12 | leifmadsen | something done? |
| 15:02.21 | leifmadsen | going good -- I think I might get a coffee soon and have a snack |
| 15:02.26 | leifmadsen | (I didn't get a coffee yesterday :)) |
| 15:02.59 | pabelanger | Roll up the rim to win! |
| 15:04.42 | leifmadsen | lol |
| 15:04.46 | leifmadsen | I don't drink that crap anymore |
| 15:04.49 | leifmadsen | I drink real coffee |
| 15:05.23 | eliel | leifmadsen: a site with all the apps/funcs/agis documentation? |
| 15:06.05 | leifmadsen | eliel: http://www.asterisk.org/docs |
| 15:06.24 | leifmadsen | eliel: (it's out of date unfortunately, but maybe some day we'll get the web master to update it) |
| 15:07.15 | eliel | leifmadsen: nice work! |
| 15:07.28 | russellb | leifmadsen: or we'll set up our own site ... |
| 15:07.58 | eliel | like asteriskdocs.org (??) |
| 15:08.10 | leifmadsen | that could be a good site for something like that :) |
| 15:08.52 | russellb | leifmadsen: well I'm also thinking something to replace asterisk.org/docs |
| 15:08.57 | russellb | maybe we just link over there, whatever |
| 15:09.26 | russellb | since we have no access to update that, and it doesn't seem to be getting any attention, that's the only thing we can do I guess |
| 15:09.30 | leifmadsen | russellb: ya, jsmith owns that domain, so we can likely re-use it for whatever we want |
| 15:09.34 | *** join/#asterisk-dev zerohalo (~zerohalo@173-13-92-17-NewEngland.hfc.comcastbusiness.net) |
| 15:09.58 | leifmadsen | russellb: ya, if we could get the existing code (jsmith might have that?) which converts the XML docs to HTML |
| 15:10.08 | leifmadsen | then we could probably build a simple site |
| 15:10.21 | leifmadsen | or reference.asterisk.org |
| 15:10.28 | leifmadsen | (on a VM that we can actually control) |
| 15:10.33 | russellb | another thing I have been thinking about, since I'm on an atlassian tools kick, is wondering if we could use confluence |
| 15:10.38 | russellb | it's a wiki |
| 15:10.43 | leifmadsen | hmmm |
| 15:10.54 | russellb | need to look into its APIs a bit more ... |
| 15:10.58 | leifmadsen | I've always been hesitant to build a wiki and getting into the voip-info.org problem |
| 15:11.04 | russellb | well, access controlled |
| 15:11.10 | leifmadsen | I'd be in favour of that |
| 15:11.33 | russellb | it would be nice because it would also be a place to store development process, policies, etc |
| 15:11.41 | russellb | stuff that is spread between doxygen and asterisk.org/developers |
| 15:11.42 | leifmadsen | especially if the APIs are cool enough to allow updates remotely |
| 15:11.48 | russellb | right |
| 15:11.51 | leifmadsen | yes I agree! |
| 15:11.54 | russellb | all of their other tools have awesome APIs |
| 15:12.11 | leifmadsen | well, we could talk about that today too if you want |
| 15:12.15 | leifmadsen | :) |
| 15:12.33 | russellb | indeed. |
| 15:14.33 | russellb | leifmadsen: http://confluence.atlassian.com/display/CONFDEV/Remote+API+Specification |
| 15:14.37 | russellb | looks pretty flexible. |
| 15:15.04 | russellb | that is confluence itself, and they have it set up so any user can add comments at the bottom of pages |
| 15:15.42 | leifmadsen | russellb: from what I just saw on the APIs (Comments) might even be able to build something like: http://labs.oreilly.com/ofps.html |
| 15:15.54 | leifmadsen | although that's a bit more advanced than what we're talking about here |
| 15:16.03 | russellb | yeah, that's fancy stuff, though |
| 15:16.23 | russellb | anyway, I'm just thinking that if we deem this important, we may just need to do it our own way, heh. |
| 15:16.34 | leifmadsen | ya, unfortuantely |
| 15:16.40 | leifmadsen | but yes, documentation I would consider important |
| 15:17.02 | russellb | :-) |
| 15:17.10 | file | raises eyebrow |
| 15:17.21 | eliel | russellb: another project for gsoc could be create a webservice (inside the asterisk http server) to return the data api result in json, xml, etc... |
| 15:17.34 | leifmadsen | russellb: did you watch Two and a Half Men last night? There was a note at the end that said the smiley face died on March 14, 2010 |
| 15:17.46 | russellb | yes, there are a lot of cool things that could be done with the data API, but we need a lot of good data providers first :-) |
| 15:17.56 | russellb | leifmadsen: nope. |
| 15:18.42 | eliel | russellb: agree |
| 15:20.27 | *** join/#asterisk-dev kpfleming (~kpfleming@asterisk/digium-director-of-software-tech/kpfleming) |
| 15:20.27 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 15:28.05 | *** join/#asterisk-dev marv (~timr@router.asteriasgi.com) |
| 15:31.31 | *** join/#asterisk-dev CunningPike (~CunningPi@204.239.8.157) |
| 15:32.44 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 15:32.53 | *** join/#asterisk-dev ctooley (~ctooley@c-98-193-232-154.hsd1.tn.comcast.net) |
| 15:32.53 | *** mode/#asterisk-dev [+o ctooley] by ChanServ |
| 15:33.09 | *** join/#asterisk-dev Netgeeks (~chris@173.11.68.155) |
| 15:43.08 | leifmadsen | jmls: hey! |
| 15:43.12 | leifmadsen | M14889 |
| 15:43.15 | MuffinMan | [closed] [Asterisk] Applications/app_voicemail/IMAP 0014889: Thread-specific vm_state tracking issue if a voicemail is left immediately after a restart. reported by jaroth https://issues.asterisk.org/view.php?id=14889 |
| 15:43.15 | leifmadsen | errr |
| 15:43.17 | leifmadsen | M14899 |
| 15:43.19 | MuffinMan | [feedback] [Asterisk] Core/NewFeature 0014899: [patch] Add environment variables to #exec for config files, add new AST_BUILD dialplan function reported by jmls https://issues.asterisk.org/view.php?id=14899 |
| 15:48.05 | *** join/#asterisk-dev ebroad (~EB@72.11.213.195) |
| 15:57.43 | *** join/#asterisk-dev Netgeeks (~chris@173.11.68.155) |
| 17:04.54 | seanbright | la da dee dee |
| 17:05.56 | russellb | orly |
| 17:06.10 | seanbright | for realz |
| 17:12.21 | eliel | russellb: thanks for the review, do you want me to move the test cases to the test api before merging? |
| 17:12.35 | russellb | please :-) |
| 17:12.40 | eliel | russellb: ok |
| 17:24.12 | *** join/#asterisk-dev sruffell (~sruffell@asterisk/the-kernel-guy/sruffell) |
| 17:24.12 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 17:24.30 | *** join/#asterisk-dev TheDavidFactor (~chatzilla@nc-71-0-16-133.dhcp.embarqhsd.net) |
| 17:27.26 | eliel | russellb: i don't know why but automerge is not doing its work with the gsoc branch |
| 17:27.51 | russellb | i will look |
| 17:27.56 | *** join/#asterisk-dev zoid_99 (~christoph@router.asteriasgi.com) |
| 17:28.24 | *** join/#asterisk-dev Slugs_ (Slugs_@c-76-97-205-31.hsd1.ga.comcast.net) |
| 17:28.47 | russellb | the automerge property isn't set |
| 17:29.29 | eliel | debian:~/dev/asterisk/gsoc# svn pg automerge . |
| 17:29.30 | russellb | should be fixed |
| 17:29.34 | eliel | * |
| 17:29.47 | russellb | it wasn't set when I checked it out ... |
| 17:30.09 | russellb | team/group/data_api_gsoc2009 right? |
| 17:30.26 | eliel | yes!, thanks |
| 17:36.35 | *** join/#asterisk-dev codefreeze-lap (~murf@mail.parsetree.com) |
| 17:36.35 | *** mode/#asterisk-dev [+o codefreeze-lap] by ChanServ |
| 17:45.45 | tzafrir | again, is it just me, or test/Makefile has MODULE_PREFIX=test (instead of tests)? |
| 17:45.48 | tzafrir | fix it? |
| 17:46.21 | kpfleming | no, it's fine the way it is |
| 17:46.26 | kpfleming | why do you think it is wrong? |
| 17:46.31 | *** join/#asterisk-dev anonymouz666 (~anonymouz@187-28-37-118.poolip.RJO.embratel.net.br) |
| 17:46.53 | zoid_99 | I'm trying to work with an ASTOBJ specifically the sip_registry object. I need to be able to look one up using ASTOBJ_CONTAINER_FIND but I can't seem to be able to set the reg->name value on creation. I set it using ast_copy_string(reg->name,user1.userpart,sizeof(reg->name)); but doesn'tseem todo anything. |
| 17:46.59 | tzafrir | tests don't get excluded if they are listed as excluded in makeopts |
| 17:47.06 | tzafrir | err... |
| 17:47.12 | tzafrir | in menuselect.makeopts |
| 17:47.21 | seanbright | zoid_99: what version of asterisk is this? |
| 17:47.24 | kpfleming | are you talking about MODULE_PREFIX or MENUSELECT_CATEGORY? |
| 17:47.27 | zoid_99 | 1.6.2.x |
| 17:47.35 | zoid_99 | and 1.4.20 |
| 17:47.48 | zoid_99 | er.. 1.4.30 |
| 17:47.50 | kpfleming | 1.6.2 does not use ASTOBJ in chan_sip |
| 17:47.55 | kpfleming | it uses ao2 |
| 17:47.57 | tzafrir | MODULE_PREFIX |
| 17:49.12 | tzafrir | hmm... right. |
| 17:49.36 | kpfleming | MENUSELECT_CATEGORY should be TESTS, so it matches the directory name |
| 17:49.39 | kpfleming | MODULE_PREFIX is not related |
| 17:49.56 | zoid_99 | it seems to still use it for registry object? ASTOBJ_INIT(reg); |
| 17:50.14 | kpfleming | in 1.6.2? that's surprising, but OK |
| 17:50.57 | zoid_99 | yeah.. 1.6.2.6 |
| 17:53.19 | tzafrir | bah. this is the only non "xml" category whose name is not the name of the directory |
| 17:53.57 | kpfleming | right... so feel free to fix it |
| 18:17.17 | *** join/#asterisk-dev atis_work (~atis_work@193.238.212.171) |
| 18:36.02 | tzafrir | include/asterisk/tests.c has a #ifdef TEST_FRAMEWORK |
| 18:36.10 | tzafrir | Where is TEST_FRAMEWORK defined? |
| 18:36.28 | kpfleming | it's a compiler flag from one of the menuselect files |
| 18:37.34 | tzafrir | tests/test_substitution.c fails: test_substitution.c:260: error: implicit declaration of function âast_cli_generatorâ |
| 18:37.56 | tzafrir | (And I actually don't use dummy-select in that tree. Before I did) |
| 18:39.08 | tzafrir | With NOISY_BUILD I don't see that define added in the command-line |
| 18:39.25 | russellb | it's in include/asterisk/buildopts.h probably |
| 18:56.41 | tzafrir | ok. it was an issue with the category change, then |
| 18:57.43 | tzafrir | those files were not yet properly disabled |
| 19:15.26 | *** join/#asterisk-dev Deeewayne (~dwayne@75.76.254.162) |
| 19:15.26 | *** mode/#asterisk-dev [+o Deeewayne] by ChanServ |
| 19:16.39 | *** join/#asterisk-dev atis_work (~atis_work@193.238.212.171) |
| 19:28.14 | *** join/#asterisk-dev pabelanger (~pabelange@yoda.kanatek.com) |
| 19:43.15 | bklang_ | Corydon76-dig: Hey Tilghman, I've found another place (this time in AMI) where it's segfaulting on Solaris because a null is passed to sprintf(). Do you want me to add information to the ticket you used to fix AGI or do you prefer a new ticket? |
| 19:44.24 | kpfleming | quote: "What audio codec is better, G.711a or G.711u ???" |
| 19:44.36 | russellb | kpfleming: yes |
| 19:44.45 | Qwell | "Depends on whether or not you are a silly European." |
| 19:44.54 | russellb | lol!!! |
| 19:44.57 | russellb | amazing. |
| 19:45.16 | kpfleming | personally, i find a-Law to be much warmer and more inviting than u-Law |
| 19:45.20 | russellb | bklang_: a ticket would be ideal |
| 19:45.26 | bklang_ | russellb: a new ticket? |
| 19:45.32 | russellb | yes |
| 19:45.35 | bklang_ | Ok, thanks |
| 19:45.39 | russellb | np |
| 19:45.52 | russellb | Corydon76-dig is out today |
| 19:55.15 | tzafrir | Is it possible to create reviews from a anon svn working copy? |
| 19:55.56 | tzafrir | (I currently still can't use origsvn on my Squeeze laptop) |
| 19:55.56 | kpfleming | by uploading a patch, yes |
| 19:56.14 | tzafrir | How should the patch be formatted? |
| 19:56.30 | kpfleming | normal unified diff |
| 19:58.31 | russellb | you can run post-review from an svn.digium.com checkout |
| 20:06.43 | tzafrir | hmm... 'svn switch' from trunk to 1.6.0 was not such a grand idea |
| 20:06.43 | kpfleming | nope |
| 20:06.43 | kpfleming | bigggggg difference |
| 20:09.28 | tzafrir | That is: made the working copy a non-working copy |
| 20:10.26 | *** join/#asterisk-dev elguero (~miguel323@ns1.nashuacs.com) |
| 20:10.48 | tzafrir | 'make distclean' before that make it work, though |
| 20:17.07 | *** join/#asterisk-dev atis_work (~atis_work@193.238.212.171) |
| 20:28.32 | moy | question about main/translate.c ... isn't a break missing around line 751 in 1.4, is finding any translator providing the same src/dst, but then sets t = NULL and continues, which seems to be causing a segfault in my box |
| 20:33.04 | Qwell | moy: sounds right |
| 20:34.11 | Qwell | I can't imagine there being any reason for a translator being in the list multiple times. |
| 20:35.00 | moy | not the same translator, different modules, same src/dst codecs |
| 20:36.49 | Corydon76-dig | yawns |
| 20:51.09 | *** join/#asterisk-dev sivadnz (~alec@210.48.127.227) |
| 21:08.53 | seanbright | hmm |
| 21:09.04 | seanbright | breaking ABI in 1.4? |
| 21:09.11 | seanbright | that seems like a horrible horrible horrible idea |
| 21:09.13 | seanbright | that is all. |
| 21:09.59 | kpfleming | it's not a great idea, but it's better than causing ABI to change based on compilation-time choices |
| 21:10.49 | seanbright | why now? |
| 21:11.10 | kpfleming | to stop the damage |
| 21:11.42 | kpfleming | and it doesn't affect all external modules, and it also doesn't solve the problem for all external modules |
| 21:11.45 | kpfleming | but it helps |
| 21:11.50 | seanbright | what issue # does this address? |
| 21:11.59 | seanbright | i'm guessing JIRA something |
| 21:12.01 | kpfleming | yes |
| 21:12.04 | seanbright | ahhh |
| 21:12.23 | kpfleming | it was brought to our attention via some incompatibilities with our binary modules, but it applies to other modules as well |
| 21:12.31 | seanbright | well that explains all changes that go against the release policy :-) |
| 21:13.05 | kpfleming | it also affects all version where asterisk-addons is not part of the asterisk tree... right now changing those compilation time flags requires rebuilding -addons as well, or risking very weird behavior |
| 21:13.19 | Qwell | the modules simply won't load |
| 21:13.29 | kpfleming | no, i mean before the patch |
| 21:13.33 | Qwell | ahh, right |
| 21:14.05 | kpfleming | after the patch, the behavior is much more deterministic. |
| 21:14.22 | Qwell | Right. You'll just get symbol resolution errors |
| 21:14.22 | kpfleming | and it provides us a model for how to do it right in the future, and never cause this problem again |
| 21:14.46 | seanbright | well, at least we made note of it in the CHANGES file |
| 21:15.20 | Qwell | well, CHANGES is the wrong place. but you're right |
| 21:15.24 | kpfleming | it will also be prominently included in the release announcement |
| 21:15.35 | seanbright | Q(well) |
| 21:15.53 | seanbright | awesome |
| 21:16.19 | Qwell | hrm. where should I be putting this? |
| 21:16.38 | kpfleming | UPGRADE.txt |
| 21:16.39 | Qwell | both CHANGES and UPGRADE.txt are for upgrading major versions |
| 21:16.43 | seanbright | you should just post it directly to nerdvittles.com |
| 21:16.52 | seanbright | it will be there 15 seconds after the next release |
| 21:16.57 | kpfleming | no, UPGRADE.txt also has sections for upgrades between point releases |
| 21:17.02 | Qwell | kpfleming: ahh, okay |
| 21:17.03 | kpfleming | or can, it has in the past |
| 21:21.32 | seanbright | oh well |
| 21:22.16 | *** join/#asterisk-dev akshay07b (~a464aa04@gateway/web/freenode/x-qtmihjgptscbhpbi) |
| 21:24.22 | akshay07b | what all things need to be taken in consideration before editing astobj2 |
| 21:27.38 | seanbright | here's my thing |
| 21:28.00 | seanbright | we aren't supposed to break ABI in release branches |
| 21:28.07 | seanbright | we do anyway when there is: |
| 21:28.09 | seanbright | a) good reason |
| 21:28.15 | seanbright | b) a paying customer needs it |
| 21:28.20 | Qwell | There is good reason |
| 21:28.22 | seanbright | so here is what i suggest |
| 21:28.31 | seanbright | we change the policy to include those 2 things |
| 21:28.37 | russellb | Qwell: I actually thought this was going to be possible without breaking ABI |
| 21:28.38 | seanbright | and i'll stop trolling on it |
| 21:28.41 | russellb | I guess I should have reviewed the patch |
| 21:28.49 | Qwell | russellb: no, by definition the ABI has to change |
| 21:29.07 | russellb | no, if the ABI changes to include both signatures, it doesn't break |
| 21:29.11 | russellb | it changes by addition, not by change |
| 21:29.26 | Qwell | it can't include both. they were different and conflicting. That was the problem. |
| 21:29.33 | kpfleming | right |
| 21:29.38 | kpfleming | the same function name had two signatures |
| 21:29.56 | kpfleming | in some cases, at least |
| 21:30.06 | russellb | ok, then I'm not sure I agree that it should be in 1.4 (or any release), either |
| 21:31.20 | kpfleming | Qwell: do you have any written notes that demonstrate what functions would appear with which names depending on the compilation flags? |
| 21:31.33 | kpfleming | it would be useful to see visually what that situation was prior to this patch |
| 21:31.36 | russellb | what's a function that can't work and requires a change rather than an addition? I'd like to go look to make sure I understand |
| 21:32.03 | Qwell | I'll double-check, but I believe ast_mutex_*() |
| 21:32.43 | russellb | k, let me know, i have a meeting in a few. can we temporarily revert it until we complete this discussion? |
| 21:33.05 | Qwell | for ao2_*lock(), that function would exist without DEBUG_THREADS, and with DEBUG_THREADS it would instead by __ao2_*lock() |
| 21:33.40 | kpfleming | ok, that can be solved by providing both functions if needed |
| 21:33.51 | russellb | k, so for that, if the ABI always exposes both, we're good |
| 21:37.11 | kpfleming | Qwell: for everybody's sake, it would be good to make a table (spreadsheet, whatever) listing each affected function on its own row, and then have columns that show how it would appear given each potential combination of compilation options that affect it |
| 21:37.36 | akshay07b | what all things need to be taken in consideration before editing astobj2 |
| 21:37.49 | Qwell | It's more than what would appear - it's how things get initialized and such as well. |
| 21:38.01 | kpfleming | well, let's start with the function names |
| 21:38.34 | Qwell | post-processed? |
| 21:38.52 | seanbright | Qwell: sorry for making more work for you |
| 21:39.13 | kpfleming | yes, what the actual symbol names would be, not macro names |
| 21:39.19 | *** join/#asterisk-dev xpot-mobile (~xpot@66.60.101.91) |
| 21:44.46 | *** join/#asterisk-dev akshay07b (~a464aa04@gateway/web/freenode/x-qyoftaiofqyopkph) |
| 21:49.59 | akshay07b | I had a query regarding the extension of astobj2 |
| 21:50.22 | kpfleming | ask a specific question, not a question about a question |
| 21:52.18 | Kobaz | i have a question about the question for questioning questions that result in questions |
| 21:52.27 | kpfleming | shush |
| 21:52.42 | Kobaz | heh |
| 21:53.14 | russellb | guys, he is a student interested in working with us for the Google Summer of Code program |
| 21:53.40 | russellb | given that, let's be gentle :-) |
| 21:53.50 | kpfleming | trying to be helpful, at least |
| 21:54.03 | russellb | looks at Kobaz |
| 21:54.35 | Kobaz | looks back |
| 21:56.44 | Qwell | kpfleming: clarification needed. How it would look *in Asterisk*? Because, they were mostly all static inline... |
| 21:57.12 | kpfleming | if they were static inline, then there couldn't be any symbol resolution errors, or mismatched signatures |
| 21:57.28 | kpfleming | or even just 'static', never mind the 'inline' part |
| 21:58.01 | Qwell | then none of the mutex ones are going to matter |
| 21:58.42 | Qwell | it's just ao2_{try,un,}lock and ast_channel_{try,un,}lock |
| 21:59.09 | kpfleming | then you may have done more work than was necessary :-( |
| 21:59.25 | russellb | well, it cleans up lock.h immensely |
| 21:59.30 | russellb | which is still awesome for trunk, at least |
| 21:59.35 | kpfleming | yes |
| 22:00.04 | Qwell | kpfleming: well, no. since if something initializes a lock and something elses uses that lock...things will be...err...screwed. |
| 22:00.36 | kpfleming | if the something elses were compiled with different flags then |
| 22:00.42 | Qwell | right |
| 22:00.49 | russellb | when would that occur, though? |
| 22:00.52 | russellb | is that possible? |
| 22:00.57 | kpfleming | yes |
| 22:00.59 | russellb | channel.h... |
| 22:01.01 | kpfleming | chan_ooh323 |
| 22:01.15 | russellb | damn you ast_channel |
| 22:01.19 | kpfleming | yes |
| 22:01.53 | russellb | so yeah, the mutex stuff is required for embedded locks in public API shared data structures |
| 22:01.57 | russellb | ast_channel being the most blatant example |
| 22:02.02 | russellb | (though no longer true in trunk, at least) |
| 22:02.17 | russellb | that may actually be the only example |
| 22:02.51 | russellb | negative, there are others. |
| 22:05.42 | russellb | I think we should go with this rework in trunk only. Even before this change, modules will refuse to load if they were not compiled with the same build flags. |
| 22:06.09 | kpfleming | well... some modules. not all of them, unfortunately, but that's a digium commercial product problem. |
| 22:06.38 | russellb | ah. Yeah, we should make those modules have the buildsum or whatever it is, so that they only load if the compiler flags we expect were used. |
| 22:06.40 | kpfleming | if we can fix ao2_lock() and ao2_unlock() to provide both signatures regardless of compile-time options, that would solve the immediate issues and not break ABI |
| 22:06.55 | russellb | Okay, I'm perfectly fine with that, then, for all releases |
| 22:07.14 | russellb | and sorry I didn't weigh in pre-commit ... |
| 22:08.10 | Qwell | https://reviewboard.asterisk.org/r/508/diff/#index_header |
| 22:08.38 | Qwell | main/astobj2.c, line 143. How do you rectify that when ast_mutex_lock() is static inline? |
| 22:09.31 | russellb | that should be okay. astobj2 locks are private to main/astobj2.c |
| 22:09.38 | kpfleming | correct |
| 22:09.40 | russellb | no other file will ever touch them |
| 22:09.44 | kpfleming | nobody has access to them except via the API calls |
| 22:19.14 | sivadnz | How do we fix the broken URL in sounds/Makefile that are now broken, for asterisk tag releases 1.4.0 - 1.4.10, they refer to ftp.digium.com, not downloads.asterisk.org? |
| 22:19.57 | Qwell | sivadnz: Those releases are old and not useful. They won't be changed. |
| 22:22.14 | sivadnz | they help testing regressions. |
| 22:22.26 | russellb | besides, you can just run menuselect and tell it not to download sounds |
| 22:22.31 | russellb | sounds are bundled with the tarballs |
| 22:22.35 | russellb | if you want more, you can always download manually |
| 22:22.57 | kpfleming | put ftp.digium.com into /etc/hosts |
| 22:23.22 | russellb | ooh |
| 22:23.30 | russellb | kpfleming: you and your ideas... |
| 22:23.34 | kpfleming | that *might* work |
| 22:23.42 | kpfleming | since we have it on a dedicated IP now |
| 22:23.43 | russellb | i think downloads is the default virtual host |
| 22:23.49 | russellb | and that |
| 22:23.50 | kpfleming | when it was name-based virtual hosting it would have failed |
| 22:23.57 | sivadnz | look what I started :) |
| 22:24.13 | russellb | resolved! |
| 22:24.21 | sivadnz | how? |
| 22:24.25 | russellb | ^^^^^^^ |
| 22:24.35 | jmls | leifmadsen: really late pong ... just got home |
| 22:24.39 | russellb | 1) disable sounds downloads in menuselect. or 2) use /etc/hosts |
| 22:25.06 | sivadnz | or change the Makefile. |
| 22:25.18 | kpfleming | any of the above |
| 22:25.22 | russellb | they are existing releases |
| 22:25.26 | russellb | you can't edit the past |
| 22:25.30 | kpfleming | i don't know why ftp.digium.com is still in DNS, to be honest |
| 22:25.53 | russellb | we could have them CNAM it to downloads or whatever |
| 22:27.29 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 22:29.19 | sivadnz | I used 1.4.0 to check how many looped local channels I could get, 70 odd - no segfault though. trunk segfaults, 1.4 branch segfaults, 1.6.1 branch segfaults. |
| 22:30.04 | russellb | it seems to me that it should work until you run out of RAM |
| 22:30.09 | sivadnz | and stumbled over the broken URL, easily fixed, but thought I'd ask, for next time. |
| 22:30.26 | russellb | stack consumption in a given thread shouldn't grow in that scenario ... |
| 22:30.42 | russellb | i mean, obviously something is wrong, i just can't think of why that would kill it |
| 22:31.25 | kpfleming | did 1.4.0 have res_random_segfault? |
| 22:31.25 | Kobaz | but every dial to a local channel is more calls on the stack |
| 22:31.54 | kpfleming | no, in the heap |
| 22:32.11 | russellb | stacks in different threads, though |
| 22:32.34 | russellb | an yeah, lots of heap |
| 22:32.36 | Kobaz | oh yeah |
| 22:32.57 | russellb | kpfleming: res_random_segfault() was not approved for inclusion :-( |
| 22:33.03 | russellb | s/()// |
| 22:33.10 | kpfleming | damn |
| 22:33.33 | Kobaz | s/a/b/g |
| 22:33.44 | *** part/#asterisk-dev sivadnz (~alec@210.48.127.227) |
| 22:34.10 | russellb | thinks |
| 22:34.14 | russellb | wow, alec was up quite early |
| 22:35.46 | Kobaz | almost time to go swim |
| 22:50.00 | *** join/#asterisk-dev cobolfoo (~cobolfoo@bas7-quebec14-1177838444.dsl.bell.ca) |
| 22:50.57 | cobolfoo | I am using DAHDI to monitor a analog line (TDM) which I dont control, I want to spy on DTMF pressed, there is a way to do this? Do I have to implement my own goertzel DTMF detection based on linear audio? |
| 22:52.00 | russellb | O.O |
| 22:52.41 | kpfleming | fails at wanting to help someone spy |
| 22:52.53 | russellb | kpfleming: same here. |
| 22:54.00 | jmls | runs for cover |
| 22:54.04 | jmls | night all |
| 22:55.12 | seanbright | i don't know enough about dahdi to help |
| 22:55.25 | seanbright | does it do dtmf detection? |
| 22:55.30 | russellb | yeahhhhhhh... i don't either |
| 22:55.41 | alecdavis | don't need dahdi. callerid boxes from ebay work. |
| 22:55.51 | seanbright | russellb: the difference is that i am telling the truth :-) |
| 22:56.01 | russellb | touche |
| 22:56.35 | russellb | alecdavis: if the DTMF he was wanting to spy on was DTMF encoded CallerID, I guess :-) |
| 22:56.44 | alecdavis | russelb: timezones help making look as though I'm up early. |
| 22:56.51 | russellb | alecdavis: :-) |
| 22:56.56 | alecdavis | yes |
| 22:57.03 | Qwell | alecdavis: so what you're saying is that you're actually just up *really* late? :D |
| 22:58.38 | alecdavis | of course, If you meant I went to bed late. |
| 22:59.49 | alecdavis | ^^^^ sivadnz is alecdavis, mirc picked my alternative, and I missed it. |
| 23:01.29 | cobolfoo | caller ID use FSK not DTMF here |
| 23:01.39 | alecdavis | sean and russell: dsp.c? |
| 23:01.51 | cobolfoo | btw I use the term spy to let you understand I am not initiating the phone call |
| 23:02.44 | seanbright | sure, you could start a dsp on the channel, but that's assuming you're in asterisk |
| 23:03.40 | russellb | i refuse to participate in this conversation, heh |
| 23:04.01 | tzafrir | cobolfoo, dahdi_monitor? |
| 23:04.51 | cobolfoo | dahdi_monitor only record audio data, I already do that by doing a open("/dev/dahdi/channel") |
| 23:05.11 | cobolfoo | I can extract DTMF from the audio data but I have the feeling that dahdi can already do that |
| 23:05.40 | tzafrir | dahdi (kernel) won't do that |
| 23:08.32 | cobolfoo | In user mode, there is events like DAHDI_EVENT_DTMFUP but as far as I know it only get called when you are the one who control the line with dahdi |
| 23:12.58 | cobolfoo | Aslo, with wctdm devices, you can retrieve 109 registers, someone know which registers do what? :) |
| 23:22.25 | tzafrir | cobolfoo, is it FXS or FXO? |
| 23:22.39 | cobolfoo | FXS |
| 23:23.09 | cobolfoo | FXS line connected between CARRIER and a PHONE |
| 23:23.17 | leifmadsen | jmls: ack-pong? |