| 01:16.55 | *** join/#asterisk-dev pchero (~pchero@211.178.226.108) |
| 01:16.55 | *** mode/#asterisk-dev [+o pchero] by ChanServ |
| 03:42.32 | *** join/#asterisk-dev electronic_eel_ (~quassel@213.240.182.47) |
| 04:14.58 | *** join/#asterisk-dev csavinovich (sid296765@gateway/web/irccloud.com/x-rsprbfpzeyhegijl) |
| 04:14.58 | *** mode/#asterisk-dev [+o csavinovich] by weber.freenode.net |
| 04:46.31 | *** join/#asterisk-dev tsearle (tsearlenim@gateway/shell/matrix.org/x-krybbypwiysedqtp) |
| 04:46.31 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
| 04:47.15 | *** join/#asterisk-dev DanJenkins[m] (dannimblea@gateway/shell/matrix.org/x-nqziaipjscpxawxf) |
| 04:47.15 | *** mode/#asterisk-dev [+o DanJenkins[m]] by ChanServ |
| 05:27.58 | *** join/#asterisk-dev coreyfarrell (~coreyfarr@rrcs-66-57-14-122.midsouth.biz.rr.com) |
| 05:27.58 | *** mode/#asterisk-dev [+o coreyfarrell] by ChanServ |
| 05:29.21 | *** join/#asterisk-dev spatel (~spatel@pool-96-237-230-175.bstnma.fios.verizon.net) |
| 06:14.22 | *** join/#asterisk-dev DanJenkins[m] (dannimblea@gateway/shell/matrix.org/x-kjbybhjymohzbbkw) |
| 06:14.22 | *** mode/#asterisk-dev [+o DanJenkins[m]] by ChanServ |
| 06:16.38 | *** join/#asterisk-dev tsearle (tsearlenim@gateway/shell/matrix.org/x-egpruifcthpbhgkn) |
| 06:16.38 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
| 07:02.05 | *** join/#asterisk-dev electronic_eel (~quassel@213.240.182.47) |
| 07:30.45 | *** join/#asterisk-dev elguero (~miguel323@c-73-238-205-3.hsd1.nh.comcast.net) |
| 07:30.46 | *** mode/#asterisk-dev [+o elguero] by ChanServ |
| 08:52.19 | *** join/#asterisk-dev elguero (~miguel323@c-73-238-205-3.hsd1.nh.comcast.net) |
| 08:52.19 | *** mode/#asterisk-dev [+o elguero] by ChanServ |
| 09:01.07 | *** join/#asterisk-dev spatel (~spatel@pool-96-237-230-175.bstnma.fios.verizon.net) |
| 10:49.15 | *** join/#asterisk-dev spatel (~spatel@pool-96-237-230-175.bstnma.fios.verizon.net) |
| 11:12.36 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01cb0583d31d00dd6c39e4e3d4e964.ipv6.abo.wanadoo.fr) |
| 13:07.19 | *** join/#asterisk-dev electronic_eel_ (~quassel@213.240.182.141) |
| 13:09.18 | *** join/#asterisk-dev electronic_eel (~quassel@213.240.182.141) |
| 13:28.36 | *** join/#asterisk-dev spatel (~spatel@pool-96-237-230-175.bstnma.fios.verizon.net) |
| 13:59.37 | *** join/#asterisk-dev spatel (~spatel@pool-96-237-230-175.bstnma.fios.verizon.net) |
| 14:01.46 | *** join/#asterisk-dev spatel (~spatel@pool-96-237-230-175.bstnma.fios.verizon.net) |
| 15:09.09 | coreyfarrell | not a huge surprise: https://lists.gnu.org/archive/html/autoconf/2021-01/msg00049.html |
| 15:16.11 | seanbright | tldr? |
| 15:17.33 | gtjoseph | cmake :) |
| 15:18.53 | gtjoseph | almost every current open source project i contribute to is moving or has moved to cmake. |
| 15:35.45 | *** join/#asterisk-dev bford (uid283514@gateway/web/irccloud.com/x-xygxotffpmsxawxz) |
| 15:35.45 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 15:44.56 | *** join/#asterisk-dev kharwell (uid358942@gateway/web/irccloud.com/x-egosqmnmciqftmvm) |
| 15:44.56 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 15:54.07 | *** join/#asterisk-dev sduthil (~sduthil@198-48-220-170.cpe.pppoe.ca) |
| 16:23.25 | coreyfarrell | seanbright: autoconf project appears to be faltering |
| 16:26.16 | coreyfarrell | gtjoseph: how would you deal with menuselect in any other build system? |
| 16:27.21 | file | menuselect isn't really buddy buddy with autoconf, is it? |
| 16:28.29 | coreyfarrell | no but the asterisk makefiles read the output of menuselect and use this to determine what gets built. |
| 16:29.40 | seanbright | i would assume that you would have to blow up the world to move away from autoconf |
| 16:30.11 | file | oh you're referring to using the information of menuselect to control the build process |
| 16:30.37 | coreyfarrell | menuselect requires info from the 'configure' stage, but then provides information which controls build flags and even what gets built. |
| 16:31.47 | coreyfarrell | I don't know cmake enough to say if you can inject "run menuselect" between detecting dependencies and generating build targets |
| 16:34.35 | seanbright | george knows |
| 16:39.09 | *** join/#asterisk-dev csavinovich (sid296765@gateway/web/irccloud.com/x-krwmijtxcjxrjiem) |
| 16:39.09 | *** mode/#asterisk-dev [+o csavinovich] by ChanServ |
| 16:39.31 | *** join/#asterisk-dev tuxd00d (sid401305@gateway/web/irccloud.com/x-nyuatocuelaandgm) |
| 16:39.39 | *** join/#asterisk-dev X-Rob (X-Rob@gateway/web/irccloud.com/x-rihayisatkanyxwo) |
| 16:42.09 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-jrqakgqjigohbdoi) |
| 16:42.09 | *** mode/#asterisk-dev [+o jameswf] by ChanServ |
| 16:43.54 | *** join/#asterisk-dev bford (uid283514@gateway/web/irccloud.com/x-tgrmevqjxefesfxm) |
| 16:43.54 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 16:45.14 | gtjoseph | heh |
| 16:45.28 | gtjoseph | cmake is pretty flexible and even has a GUI to go with it. |
| 16:45.29 | *** join/#asterisk-dev kharwell (uid358942@gateway/web/irccloud.com/x-hzpnfmhqcbdebcuz) |
| 16:45.29 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 16:45.59 | gtjoseph | but cmake's goal is exactly like autoconfs... it generates makefiles |
| 16:47.20 | gtjoseph | so i think no changes to menuselect would be needed as long as we put the same includes of menuselect.makeopts. |
| 16:47.28 | gtjoseph | into the cmake genreated makefiles |
| 16:49.16 | gtjoseph | one of the things i like about cmake is that the source tree remains pristine. all build artifacts go into a ./build directory. |
| 16:50.35 | *** join/#asterisk-dev DanFromUK (sid21651@gateway/web/irccloud.com/x-hgrztbuzawrklxln) |
| 16:51.06 | coreyfarrell | yeah that's nice. I'm not the biggest fan of cmake syntax but it's no worse than autoconf/make. |
| 16:54.31 | *** join/#asterisk-dev file (sid178970@asterisk/developer-and-muffin-lover/file) |
| 16:54.31 | *** mode/#asterisk-dev [+o file] by ChanServ |
| 16:58.19 | *** join/#asterisk-dev DanFromUK (sid21651@gateway/web/irccloud.com/x-rnoctsnvcuivyiej) |
| 17:01.18 | *** join/#asterisk-dev sahmed (4672ec3f@cpe-70-114-236-63.austin.res.rr.com) |
| 17:01.39 | gtjoseph | cmake isn't the only alternative of course. |
| 17:01.55 | gtjoseph | it's just the one i see most often for C/C++ projects |
| 17:19.37 | *** join/#asterisk-dev DanFromUK (sid21651@gateway/web/irccloud.com/x-ybeoamcvpwhggohh) |
| 17:20.54 | *** join/#asterisk-dev pchero (~pchero@211.178.226.108) |
| 17:20.54 | *** mode/#asterisk-dev [+o pchero] by ChanServ |
| 18:00.44 | *** join/#asterisk-dev X-Rob (X-Rob@gateway/web/irccloud.com/x-lymergyxnnxarirw) |
| 18:18.51 | *** join/#asterisk-dev X-Rob (X-Rob@gateway/web/irccloud.com/x-rcwehirycktqhxtu) |
| 18:23.15 | *** join/#asterisk-dev electronic_eel (~quassel@213.240.182.143) |
| 18:31.51 | *** join/#asterisk-dev X-Rob (X-Rob@gateway/web/irccloud.com/x-mydhvjhvxqyhkawh) |
| 18:56.49 | *** join/#asterisk-dev tzafrir (~tzafrir@lemon.tzafrir.org.il) |
| 18:56.49 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 19:15.13 | *** join/#asterisk-dev X-Rob (X-Rob@gateway/web/irccloud.com/x-mlhqelxpgtsvlvhe) |
| 19:59.26 | seanbright | is cmake one of those convention-over-configuration tools? |
| 20:01.36 | gtjoseph | elaborate? |
| 20:05.11 | seanbright | like if your code layout is such you don't have to do a lot of cmake configuration |
| 20:05.16 | seanbright | a la maven in java land |
| 20:06.21 | seanbright | for a simple maven project you just need a small pom.xml which has mostly metadata and then a java file at src/main/java/Foo.java |
| 20:06.26 | seanbright | and maven will just do the right thing |
| 20:06.32 | gtjoseph | Unfortunately, there's no standard way to declare what a C file's dependencies are, like "import" statements in java. |
| 20:07.02 | gtjoseph | so you still have to create cmake configuration files to declare what the dependencies are. |
| 20:07.08 | seanbright | well... import statements don't really declare dependencies, but i get your point |
| 20:08.03 | gtjoseph | yeah true. i guess it's the maven xml configs. |
| 20:08.49 | gtjoseph | still at least that's a "standard". cmake does have a bunch of pre-built rules for many of the common libraries. |
| 20:09.31 | seanbright | maven makes the common case easy but allows you to customize things |
| 20:09.36 | seanbright | was just wondering if cmake is similar |
| 20:09.45 | seanbright | just did a simple hello world with cmake and it's cool |
| 20:10.35 | gtjoseph | https://pastebin.com/3yvMq7x4 |
| 20:10.35 | seanbright | sean@gerrit:~/cm$ ./Tutorial |
| 20:10.36 | seanbright | Hello, World |
| 20:10.43 | seanbright | BAM... i did that with cmake |
| 20:10.47 | gtjoseph | :) |
| 20:11.00 | seanbright | fancy |
| 20:11.15 | seanbright | so yeah, let's convert to cmake |
| 20:11.23 | seanbright | let's discuss in october |
| 20:11.34 | seanbright | or maybe we can roll out own build system |
| 20:11.38 | seanbright | basterisk |
| 20:11.49 | gtjoseph | nothing's stopping you from create the cmake build system alongside autoconf :) |
| 20:12.15 | seanbright | i am going to do it once i am done all of these other things |
| 20:12.22 | gtjoseph | uh huh |
| 20:12.37 | seanbright | unfurls a 48 foot long todo list |
| 20:42.07 | jsmith | seanbright: If we're going to consider new build systems, can I suggest Meson instead of CMake? |
| 20:42.21 | jsmith | seanbright: It's a lot more modern, and a lot less fragile. |
| 20:44.11 | gtjoseph | jsmith: I'm sure he'll add Meson to the end of his 48 foot ling to-do list :) |
| 20:44.25 | jsmith | That'd be swell :-) |
| 20:49.54 | seanbright | 49 foot |
| 20:50.11 | seanbright | who things build systems are awesome?! ME SON! |
| 20:50.40 | seanbright | meson... get it? |
| 20:50.52 | coreyfarrell | jsmith: I've checked out meson, much nicer syntax. I'm not sure meson can support configure -> menuselect -> build stages? |
| 20:55.44 | gtjoseph | if we're serious about this, we cna always change menuselect. |
| 20:58.30 | seanbright | i guess i never understood why configure flags weren't sufficient (albeit not as user friendly) |
| 20:58.37 | gtjoseph | we could have menuselect generate a meson_options.txt file |
| 20:59.20 | gtjoseph | i think coreyfarrell is thinking more of enabling or disabling modules, etc. |
| 20:59.29 | gtjoseph | which is a lot easier from cmenuselect |
| 21:01.29 | gtjoseph | oh, interesting... there's an autotools to meson converter. |
| 21:01.55 | seanbright | i think we'd be better starting from scratch |
| 21:02.05 | seanbright | we all know that complete rewrites aren't an antipattern |
| 21:02.15 | gtjoseph | :) |
| 21:02.32 | gtjoseph | would be nice to run just to see what it comes up with. |
| 21:02.36 | seanbright | true |
| 21:02.48 | seanbright | our stuff is not idiomatic autotools though |
| 21:03.09 | seanbright | but can't hurt to try |
| 21:03.26 | gtjoseph | it's not idiomatic, just idiotical. |
| 21:03.33 | seanbright | confirmed |
| 21:03.38 | *** join/#asterisk-dev sahmed (4672ec3f@cpe-70-114-236-63.austin.res.rr.com) |
| 21:10.55 | coreyfarrell | I agree with seanbright about just having configure flags (including for modules). |
| 21:11.28 | gtjoseph | for pachage that's fine as they use menuselect from the connand line anyway |
| 21:11.35 | gtjoseph | packagers |
| 21:12.04 | gtjoseph | for end users it would be a big pain especially with inter-package dependencies |
| 21:12.41 | coreyfarrell | you mean inter-module dependencies? |
| 21:12.41 | gtjoseph | now personally, i think we should have use the kernel's menuconfig instead of writing our own menuselect. |
| 21:12.52 | gtjoseph | yeah sorrt...inter-module |
| 21:13.26 | seanbright | any licensing issues with that? |
| 21:14.16 | gtjoseph | with what? the kernel menuconfig? |
| 21:14.20 | seanbright | yes |
| 21:14.41 | seanbright | i think i misread when you said |
| 21:14.51 | seanbright | i thought you were advocating for using it |
| 21:14.57 | seanbright | but it seems you were just opining that we didn't |
| 21:15.32 | gtjoseph | yeah, i was just opinioning. |
| 21:15.44 | seanbright | so idiotical |
| 21:16.16 | gtjoseph | we have menuselect which reads the sources files and extracts the dependencies and module status so no need to change that. |
| 21:16.45 | gtjoseph | i just grumble every time i go to build a kernel and type "make menuselect" |
| 21:17.03 | gtjoseph | then come back to asterisk and do "make cmenuconfig" |
| 21:17.07 | gtjoseph | :) |
| 21:17.14 | seanbright | i thought we had support for menuconfig |
| 21:17.18 | seanbright | like as an alias |
| 21:17.27 | seanbright | if not, i approve you adding it |
| 21:17.31 | seanbright | +2 |
| 21:17.36 | gtjoseph | Ha! |
| 21:18.29 | seanbright | so today it's: |
| 21:18.45 | seanbright | determine what we can build -> select what we want to build -> build |
| 21:18.56 | gtjoseph | yep |
| 21:19.00 | seanbright | in a way it's a little silly |
| 21:19.14 | seanbright | so if i want to add a new module res_foo that relies on libfoo |
| 21:19.30 | seanbright | i have to embed the metadata in res_foo.c which will get picked up by make menuselect-tree |
| 21:19.50 | seanbright | but i also have to modify configure.ac to actually find that dependency |
| 21:20.01 | gtjoseph | yep |
| 21:20.24 | seanbright | ideally both would be generated automatically |
| 21:20.25 | file | plays space invaders in cmenuselect |
| 21:20.32 | seanbright | it's a chicken & egg thing |
| 21:20.52 | seanbright | i'm expecting a PM from file in a minute that says "yeah we're not changing the build system" |
| 21:21.17 | file | if you want to entertain the thought and think about it and propose something go for it |
| 21:21.59 | seanbright | does the choice of build system have any impact on packagers? |
| 21:22.17 | file | however at the moment I will allocate resources to moving the testsuite to python 3, myself :P |
| 21:22.28 | seanbright | s/2/3/ |
| 21:22.32 | seanbright | what's so hard about that? |
| 21:22.54 | file | it's just... magic! |
| 21:23.12 | seanbright | are there any tools to assist with that migration? |
| 21:23.23 | file | that's the easy part |
| 21:23.29 | gtjoseph | both cmake and meson are well supported so other than having to change their packaging scripts, it shouldn't be difficult. |
| 21:23.29 | seanbright | what's the hard part? |
| 21:23.41 | file | logs onto VPN |
| 21:23.56 | coreyfarrell | gtjoseph: third-party? |
| 21:24.01 | seanbright | RIP |
| 21:24.27 | seanbright | what if we just rewrote everything in c++ |
| 21:24.34 | seanbright | err, rust |
| 21:24.34 | gtjoseph | it'd just be a having the new system run a custom script that ruuns make for those. |
| 21:24.36 | file | gah dang it Confluence keyboard shortcuts |
| 21:25.15 | file | seanbright: so there is the python 2 to 3 thingy which helps, but there are still manual things that need adjustment and then functional things |
| 21:25.20 | file | https://usercontent.irccloud-cdn.com/file/GvTOYI3y/image.png |
| 21:26.10 | file | (more stuff could be found during the actual act of moving to python3) |
| 21:26.47 | seanbright | those are only in your internal bug tracker? |
| 21:27.13 | seanbright | what are the SWP #s? i have my methods... |
| 21:27.34 | file | SWP-10932 is the epic :P |
| 21:28.58 | file | I can make ASTERISK issues for things if you want to jump in |
| 21:29.39 | seanbright | nah |
| 21:29.47 | seanbright | i have 49 feet of todo list |
| 21:29.59 | seanbright | coreyfarrell loves converting python 2 to 3 |
| 21:30.33 | file | the automation conversion gets you fairly far, so lots of stuff does work |
| 21:30.40 | file | automation? automatic? |
| 21:31.36 | seanbright | i really don't know a lot about it. i know that print went from keyword to function which was the end of the world. |
| 21:32.12 | coreyfarrell | I did a bunch of bulk work on the testsuite python compatibility. up to someone else to take over. I nominate seanbright. |
| 21:34.21 | file | as sacrifice? |
| 21:35.11 | file | off topic question - but is Google Messages for web less... reliable... lately for anyone else? |
| 21:37.17 | gtjoseph | been working fine for me |
| 21:41.19 | seanbright | my day has been spent reading kamailio code |
| 21:41.35 | file | oh, that's a different mindset |
| 21:41.39 | seanbright | it is |
| 21:42.18 | seanbright | the documentation is not great |
| 21:42.27 | seanbright | but it's OSS, that is how the cookie crumbles |
| 21:42.49 | seanbright | so i got the new version of firefox and was reading about "super cookies" |
| 21:42.57 | seanbright | speaking of cookies |
| 21:43.06 | seanbright | and it's crazy the lengths trackers will go to |
| 21:43.18 | file | oh, anything exciting in the new version? |
| 21:44.48 | seanbright | more security |
| 21:45.00 | seanbright | i dunno, it's a dead man walking isn't it? |
| 21:45.25 | seanbright | i hate chrome but firefox has like... a 1 share |
| 21:45.50 | file | I would say it has a challenging future |
| 21:46.34 | file | I do wish Chrome, and others, would adopt container tabs though |
| 23:33.39 | *** join/#asterisk-dev DanJenkins[m] (dannimblea@gateway/shell/matrix.org/x-vqllhdwqrwlgarsx) |
| 23:33.39 | *** mode/#asterisk-dev [+o DanJenkins[m]] by ChanServ |
| 23:44.36 | *** join/#asterisk-dev tsearle (tsearlenim@gateway/shell/matrix.org/x-dakmfwikqaaduobt) |
| 23:44.36 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
