--- Log opened Mon Apr 06 00:00:05 2015
--- Day changed Mon Apr 06 2015
20150406 00:00:05< mattsc> Did you change that?  Should the project file come with a setting?
20150406 00:00:08< ancestral> Changed, cleaned, building
20150406 00:00:28< ancestral> I did change it…
20150406 00:00:57-!- Kexoth [~kex@78.157.29.160] has joined #wesnoth-dev
20150406 00:01:00< ancestral> from 10.8. 1.12 built okay with 10.6 as target
20150406 00:01:06< ancestral> Let’s see
20150406 00:01:10< mattsc> Yes, I am using 10.8 for ‘Os X Development Target'
20150406 00:01:22< ancestral> “Deployment target”?
20150406 00:01:34< mattsc> Umm, yes. of course.  :P
20150406 00:02:03< mattsc> I’d suggest that you try to run things without changing anything first and see if that works.
20150406 00:02:08< ancestral> 160/564…
20150406 00:02:13< mattsc> Do you have the github app installed?
20150406 00:02:18< ancestral> Fair enough, guilty as charge
20150406 00:02:25< ancestral> I do have the GitHub app
20150406 00:02:35< ancestral> I’ve just been cloning from command line so far
20150406 00:03:03< mattsc> well, it might not be the problem, when thingsdon’t work, it’s just easier to eliminate potential problems (even if they are unlikely)
20150406 00:03:36< mattsc> well, you can do it from the CL just as well - just reset the Xcode project file to what it was and try again (if the current attempt does not work, I mean)
20150406 00:03:45< gfgtdf> ancestral: looks like here somone had teh same problem: https://github.com/Homebrew/homebrew/issues/31386
20150406 00:05:16< ancestral> Maybe you’re right, maybe it has to do with Pango
20150406 00:05:37< ancestral> I could try building the pango libs with homebrew
20150406 00:05:44< shadowm> iceiceice: Boost.locale is still optional on master or is the INSTALL file out of date in that regard as well? (I'm fixing it up right now.)
20150406 00:05:59< mattsc> ancestral: and just to add my usual disclaimer: I really don’t understand all these things very well.  I’ve gotten to where I am mostly by trial and error.  So it’s quite possible that something is not set up right.
20150406 00:06:17< mattsc> It’s one of the reasons why I’d really like somebody else to take over.  :)
20150406 00:06:26< ancestral> Understood
20150406 00:06:28< gfgtdf> shadowm: it should be optional on non-windows.
20150406 00:06:34< ancestral> Well, FWIW, 1.12 is building okay for me
20150406 00:06:53< mattsc> I didn’t build the libraries for 1.12 though.  crimson_penguin did.
20150406 00:07:05< ancestral> So I think I can do the next 1.12 release
20150406 00:07:08< ancestral> Ah
20150406 00:07:33< shadowm> gfgtdf: Hm, okay... is it *recommended* to have it, though? I assume yes.
20150406 00:07:40< mattsc> ancestral: for 1.12, have you built a dmg yet?
20150406 00:07:49< mattsc> *have you test-built a dmg ...
20150406 00:07:58< gfgtdf> shadowm: boost::local is needed unless you use libintl
20150406 00:08:05< ancestral> mattsc: Failed the same way with the 10.8 deployment target
20150406 00:08:09< gfgtdf> this is gettext vs gettext_boost file
20150406 00:08:14< ancestral> mattsc: That’s the next step, make the dmg
20150406 00:08:27< shadowm> gfgtdf: Yes, so which is the recommended alternative?
20150406 00:08:33< mattsc> ancestral: okay, sounds good (on 1.12)
20150406 00:08:47< gfgtdf> shadowm: boost locale.
20150406 00:08:56< ancestral> Let me give that a go now
20150406 00:09:01< mattsc> on 1.13, not sure what to say next; it does sound like it might be a problem with one of the libraries.
20150406 00:09:16< mattsc> ancestral: I have to take off again shortly
20150406 00:09:24< iceiceice> shadowm: the INSTALL file correctly states that boost locale is optional, and libintl is an alternative
20150406 00:09:27< ancestral> Sure, I’ll PM you if I run into trouble
20150406 00:09:28< mattsc> did you still want me to send you the executable?
20150406 00:09:39< ancestral> If it’s not any trouble, sure
20150406 00:10:18< mattsc> let me see if I can get my Dropbox account to work.  I only use it once every 6 months or so...
20150406 00:12:35< shadowm> iceiceice, gfgtdf I'd like to remove the paragraph about disabling Boost.filesystem and Boost.locale support since 1) it's partly outdated (the cmake option is supposed to be removed now, isn't it?); 2) it seems too confusing for the casual reader; 3) it describes an *unsupported* configuration.
20150406 00:12:42< mattsc> ancestral: Seems to be working, but it claims that it’s going to take 15 min.  I’ll send you a PM with the link once it’s done.
20150406 00:12:47< mattsc> Bye for now.
20150406 00:12:53< ancestral> Bye
20150406 00:12:59< ancestral> Thanks
20150406 00:13:06< iceiceice> shadowm: yeah it was way too confusing in the first place,
20150406 00:13:09< shadowm> As I understand it, this information is only relevant for our OS X packagers and Ivanovic (OpenPandora packager).
20150406 00:13:13< mattsc> No, on the contrary, thank you.
20150406 00:13:25< shadowm> Everyone else should either get the supported configuration working or ask us for help directly, I think.
20150406 00:13:26< iceiceice> or anyone else who wants to build for those platforms
20150406 00:13:37< gfgtdf> are there problems with bfs on osx ?
20150406 00:13:40-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: Not here much atm. Will check logs for my nick and forum PMs.]
20150406 00:13:56< iceiceice> gfgtdf: i think for OS X 10.5 it is harder to find
20150406 00:14:13< iceiceice> i dont know that its not available but mattsc gave up eventually
20150406 00:14:19< iceiceice> thats what i remember anyways
20150406 00:14:31< iceiceice> i think we agreed only to support 10.6 from now on
20150406 00:14:58< iceiceice> or, only to target 10.6+ anyways
20150406 00:15:57< irker921> wesnoth: Ignacio R. Morelle wesnoth:master 53dbd1fe0c5a / INSTALL: Recommend Boost.locale, mention history is usually part of libreadline http://git.io/ve8Uf
20150406 00:18:53-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20150406 00:25:28< ancestral> Yeah
20150406 00:25:48< ancestral> FWIW, Macs that can’t run 10.6 are over 10 years old
20150406 00:26:11< ancestral> And the newest 32-bit Macs are 8-9 years old
20150406 00:29:41< shadowm> iceiceice, gfgtdf: What's the Boost multiindex library? I can't find it in my distribution's repositories or even http://www.boost.org/doc/libs/ .
20150406 00:30:00< shadowm> Oh never mind, it's right in front of me with an added hyphen.
20150406 00:30:32< gfgtdf> shadowm: i think we dont use boost multiindex directly
20150406 00:32:21< gfgtdf> shadowm: im wrng
20150406 00:57:24< irker921> wesnoth: Ignacio R. Morelle wesnoth:master 4c7b7b5a151c / SConstruct: scons: 'test' is also built by the 'all' target http://git.io/ve8tp
20150406 00:58:19< shadowm>  IIRC SkyOne is stuck with an old OS X version.
20150406 00:59:03< irker921> wesnoth: Ignacio R. Morelle wesnoth:1.12 eb2a30baccc6 / SConstruct: scons: 'test' is also built by the 'all' target http://git.io/ve8qU
20150406 01:09:18< shadowm> I'm rewriting INSTALL.
20150406 01:09:22-!- trewe [~trewe@2001:8a0:d108:a601:626c:66ff:fe92:9b7c] has quit [Quit: quit]
20150406 01:09:55< shadowm> Purging the autotools migration documentation that nobody needs anymore, amongst other things.
20150406 01:15:19-!- gfgtdf_ [~chatzilla@x50abb1d6.dyn.telefonica.de] has joined #wesnoth-dev
20150406 01:18:25-!- gfgtdf [~chatzilla@f054143211.adsl.alicedsl.de] has quit [Ping timeout: 255 seconds]
20150406 01:18:29-!- gfgtdf_ is now known as gfgtdf
20150406 01:18:50-!- anoob [~anoob@189-30-87-247.jvece702.dsl.brasiltelecom.net.br] has quit [Ping timeout: 272 seconds]
20150406 01:23:00-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20150406 01:23:14-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20150406 01:26:06-!- gfgtdf [~chatzilla@x50abb1d6.dyn.telefonica.de] has quit [Quit: ChatZilla 0.9.91.1 [Firefox 37.0.1/20150402191859]]
20150406 01:32:49-!- prkc [~prkc@4E5CD609.dsl.pool.telekom.hu] has quit [Ping timeout: 255 seconds]
20150406 02:03:00-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150406 02:19:38-!- new_one [~new_one@50.15.126.83] has quit [Remote host closed the connection]
20150406 02:49:09-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20150406 02:51:04-!- Necrosporus_ [~Necrospor@unaffiliated/necrosporus] has quit [Ping timeout: 252 seconds]
20150406 03:22:58-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has quit [Read error: Connection reset by peer]
20150406 03:25:49-!- c74d [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20150406 03:59:13-!- irker921 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20150406 04:53:03-!- Kexoth [~kex@78.157.29.160] has quit [Remote host closed the connection]
20150406 05:22:03-!- Kwandulin [~Miranda@p5B0096AC.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 07:00:15< Ivanovic> shadowm: i have no clue if fribidi still works on dev for me
20150406 07:00:18-!- [Relic] [~Relic]@2602:306:33a3:6d30:29ae:a05a:1958:432c] has quit [Quit: Leaving]
20150406 07:00:41< Ivanovic> shadowm: i think once i heard the feedback from the hebrew translation that fribidi does not get all the requirements right anyway i started not to care too much anymore
20150406 07:00:59< Ivanovic> (for hebrew the GUI2 work is required to correctly link the letters)
20150406 07:08:36< shadowm> Ivanovic: Well, cmake can't find fribidi but scons can, and both work fine on 1.12. To me that seems like a bug that needs to be fixed before releasing 1.13.0.
20150406 07:08:57< Ivanovic> sounds about right
20150406 07:12:23< Ivanovic> Could not find FriBiDi. Disabling FriBiDi support.
20150406 07:16:35< Ivanovic> shadowm: do you also get a warning in game_lua_kernel.cpp ?
20150406 07:16:56< Ivanovic> http://pastebin.com/geY27bxw
20150406 07:32:53< vultraz> I don't
20150406 07:33:31< shadowm> Ivanovic: No. What compiler and version is that?
20150406 07:39:17< shadowm> clang 3.4 here, but I just tested and found no problems with gcc 4.9 either.
20150406 07:39:58< shadowm> Ivanovic: From a quick glance at the code, I suspect your Boost version might be more relevant in this case. 1.55.0 here.
20150406 07:40:17< Ivanovic> gcc
20150406 07:40:40< Ivanovic> boost-1.56.0
20150406 07:40:45< Ivanovic> gcc-4.9.2
20150406 07:41:03-!- iwaim__ [~iwaim@2001:2c0:40e:2002:0:4:14:80] has quit [Remote host closed the connection]
20150406 07:41:33< shadowm> Well, 4.9.2 here too. So it must be something about Boost 1.56.0.
20150406 07:41:37< shadowm> iceiceice: ^
20150406 07:41:58< Ivanovic> shadowm: the code was committed by gfgtdf
20150406 07:42:07< shadowm> gfgtdf: ^
20150406 07:43:20< shadowm> Whoever understands what's going on there, really.
20150406 07:43:53-!- iwaim [~iwaim@2001:2c0:40e:2002:0:4:14:80] has joined #wesnoth-dev
20150406 07:45:56< shadowm> The underlying implementation of boost::tuple is an eyesore to look at and I don't even have the conflicting version.
20150406 07:53:41< shadowm> Sucks that it's giving address offsets and all the types involved are non-trivial and I really can't find my way through Boost's implementation detail jungle.
20150406 07:58:08< Ivanovic> anyway, I am off for the next hours, have fun
20150406 08:33:45-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20150406 08:53:29-!- kex [~kex@78.157.29.160] has joined #wesnoth-dev
20150406 08:58:02-!- kex [~kex@78.157.29.160] has quit [Ping timeout: 244 seconds]
20150406 09:16:10-!- mjs-de [~mjs-de@p508C873D.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 09:34:53-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 256 seconds]
20150406 09:40:18-!- Appleman1234 [~Appleman1@210.164.9.13] has joined #wesnoth-dev
20150406 09:58:38-!- cib0 [~cib@p5DC743F0.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 10:04:37-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20150406 10:12:25-!- gfgtdf [~chatzilla@x50abb1d6.dyn.telefonica.de] has joined #wesnoth-dev
20150406 10:12:36< gfgtdf> shadowm: that warning is caused by boost optional
20150406 10:13:06< gfgtdf> shadowm: see here: http://stackoverflow.com/questions/21755206/how-to-get-around-gcc-void-b-4-may-be-used-uninitialized-in-this-funct
20150406 10:16:21-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 248 seconds]
20150406 10:22:39-!- cib0 [~cib@p5DC743F0.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20150406 10:29:41-!- Appleman1234 [~Appleman1@210.164.9.13] has quit [Ping timeout: 248 seconds]
20150406 10:31:13-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150406 11:05:52-!- kex [~kex@78.157.29.160] has joined #wesnoth-dev
20150406 11:14:23-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20150406 11:16:53-!- enchilado [enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20150406 11:25:27-!- Kwandulin [~Miranda@p5B0096AC.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20150406 11:44:57-!- prkc [~prkc@4E5CD609.dsl.pool.telekom.hu] has joined #wesnoth-dev
20150406 12:04:04-!- kex [~kex@78.157.29.160] has quit [Remote host closed the connection]
20150406 12:15:28-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer]
20150406 12:16:03-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20150406 12:26:30-!- Gallaecio [~quassel@84.120.113.26.dyn.user.ono.com] has joined #wesnoth-dev
20150406 12:27:01-!- gfgtdf [~chatzilla@x50abb1d6.dyn.telefonica.de] has quit [Ping timeout: 248 seconds]
20150406 12:27:31< Gallaecio> Please, whoever has commit rights: https://paste.kde.org/pjsjlbefe
20150406 12:27:48< Gallaecio> It’s a quoting issue I noticed while translating master.
20150406 12:28:30< Gallaecio> (yes, I am translating master, I am a masochist :))
20150406 12:35:35-!- Kwandulin [~Miranda@p5B0096AC.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 12:54:29-!- mjs-de [~mjs-de@p508C873D.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20150406 13:14:02-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 256 seconds]
20150406 13:26:09-!- cib0 [~cib@p5DC743F0.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 13:26:48-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150406 13:53:20-!- Netsplit *.net <-> *.split quits: Gambit, stikonas, vultraz, Espreon, higgins, Necrosporus, cib0, iwaim, Frainz, noy,  (+9 more, use /NETSPLIT to show all of them)
20150406 13:56:35-!- prkc [~prkc@4E5CD609.dsl.pool.telekom.hu] has joined #wesnoth-dev
20150406 13:56:35-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20150406 13:56:35-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150406 13:56:35-!- yann [~dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20150406 13:56:35-!- heirecka [~heirecka@exherbo/developer/heirecka] has joined #wesnoth-dev
20150406 13:57:50-!- Gambit [~derek@wesnoth/developer/grickit] has joined #wesnoth-dev
20150406 13:57:57-!- iwaim [~iwaim@2001:2c0:40e:2002:0:4:14:80] has joined #wesnoth-dev
20150406 13:57:57-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20150406 13:57:57-!- higgins [~higgins@105.ip-167-114-152.net] has joined #wesnoth-dev
20150406 13:58:16-!- esr [~esr@wesnoth/developer/esr] has joined #wesnoth-dev
20150406 13:58:16-!- Espreon [~espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20150406 13:58:16-!- Samual [~Samual@xonotic/core-team/Samual] has joined #wesnoth-dev
20150406 13:58:16-!- EliDupree [~quassel@66-189-34-122.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20150406 13:58:16-!- Frainz [~Frainz@mmisc.de] has joined #wesnoth-dev
20150406 13:58:16-!- elias [~allefant@allefant.com] has joined #wesnoth-dev
20150406 13:58:16-!- DDR [~david@ec2.happyspork.com] has joined #wesnoth-dev
20150406 13:58:42-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20150406 13:58:42-!- timotei_ [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20150406 14:00:26-!- Netsplit *.net <-> *.split quits: stikonas, iwaim, higgins
20150406 14:00:48-!- Netsplit over, joins: iwaim
20150406 14:01:10-!- Netsplit over, joins: stikonas, higgins
20150406 14:31:35-!- iceiceice [~chris@ext-74.ias.edu] has joined #wesnoth-dev
20150406 14:31:35-!- iceiceice [~chris@ext-74.ias.edu] has quit [Changing host]
20150406 14:31:35-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20150406 14:31:57< iceiceice> shadowm, Ivanovic: please don't page me in on any compilation problems specific to cmake.
20150406 14:32:08< iceiceice> cmake should be using the exact same compilation flags as scons.
20150406 14:32:17< iceiceice> if that ever does not happen, that is a bug in the cmake script.
20150406 14:32:41< iceiceice> please fix all such bugs so that you do not introduce bugs into the release builds.
20150406 14:34:46< iceiceice> also please fix the bug where cmake release builds are -O0 by default
20150406 14:35:29< iceiceice> which causes anyone who follows instructions like this site: http://linuxg.net/how-to-install-wesnoth-1-12-0-on-ubuntu-14-10-ubuntu-14-04-and-derivatives/
20150406 14:35:35< iceiceice> to get a very bad impression of the game
20150406 14:44:36-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20150406 15:01:58-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20150406 15:15:16-!- [Relic] [~Relic]@2602:306:33a3:6d30:8ce9:ee44:3997:e303] has joined #wesnoth-dev
20150406 15:17:17-!- prkc [~prkc@4E5CD609.dsl.pool.telekom.hu] has quit [Remote host closed the connection]
20150406 15:45:02-!- Kwandulin [~Miranda@p5B0096AC.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20150406 15:55:39-!- gfgtdf [~chatzilla@x50abb1d6.dyn.telefonica.de] has joined #wesnoth-dev
20150406 16:04:33-!- kex [~kex@78.157.29.160] has joined #wesnoth-dev
20150406 16:09:08-!- kex [~kex@78.157.29.160] has quit [Ping timeout: 245 seconds]
20150406 16:14:31-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 250 seconds]
20150406 16:28:16-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150406 16:34:53-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has joined #wesnoth-dev
20150406 16:42:00-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 256 seconds]
20150406 16:50:01-!- mjs-de [~mjs-de@p508C873D.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 16:54:21-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150406 17:03:22-!- iceiceice [~chris@50.245.222.235] has joined #wesnoth-dev
20150406 17:03:37-!- iceiceice [~chris@50.245.222.235] has quit [Changing host]
20150406 17:03:37-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20150406 17:12:57-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20150406 17:56:53-!- kex [~kex@92.53.39.44] has joined #wesnoth-dev
20150406 18:34:28-!- irker182 [~irker@uruz.ai0867.net] has joined #wesnoth-dev
20150406 18:34:28< irker182> wesnoth: Chris Beck wesnoth:master 17b72bce4d88 / RELEASE_NOTES: Update RELEASE_NOTES http://git.io/veEvN
20150406 18:36:37-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 248 seconds]
20150406 18:51:09-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150406 18:53:56< irker182> wesnoth: Charles Dang wesnoth:master 03ca1a62a3f0 / changelog players_changelog: Updated changelogs http://git.io/veEYZ
20150406 19:00:08< gfgtdf> vultraz: you whether it is pssible to use a treeview  gui2 widget from lua?
20150406 19:00:49< vultraz> I don't know
20150406 19:00:52< vultraz> Probably not
20150406 19:01:12< vultraz> it's not in any of the lua gui2 functions
20150406 19:12:27< gfgtdf> vultraz: hm ok
20150406 19:12:47< iceiceice> gfgtdf:
20150406 19:12:53< gfgtdf> iceiceice: ?
20150406 19:12:55< iceiceice> i think maybe we should copy paste this line to the line where you have assertion?
20150406 19:12:55< iceiceice> https://github.com/wesnoth/wesnoth/commit/e243fdb0e408f12a807ea92b0555723bdfd5bb65#diff-6042ddb155eaeb75d4794d483a34429eL276
20150406 19:15:11< gfgtdf> iceiceice: y that what fabi said too
20150406 19:15:19< gfgtdf> iceiceice: http://www.wesnoth.org/irclogs/2015/02/%23wesnoth-dev.2015-02-03.log at 15:20
20150406 19:16:27< iceiceice> i'm just going to fix it if we are going to release soon
20150406 19:16:40< gfgtdf> iceiceice: ok
20150406 19:16:57< iceiceice> maybe some fail fast thing is better but i dont think it can do much harm anyways
20150406 19:17:26< iceiceice> if we are planning to release soon
20150406 19:17:58< gfgtdf> iceiceice: what you mean by fail fast ?
20150406 19:18:33< iceiceice> idk, i dont acutally know the theme code,
20150406 19:18:39< iceiceice> maybe the actions are never supposed to be missing
20150406 19:19:10< iceiceice> or it would be better to build the theme before prestart but mark it as "hidden" or something
20150406 19:21:14< iceiceice> usually i dont like to do "if null, ignore" its just sometimes not a very good idea :p
20150406 19:21:22< iceiceice> but here i think its right
20150406 19:23:44-!- Gallaecio [~quassel@84.120.113.26.dyn.user.ono.com] has quit [Remote host closed the connection]
20150406 19:24:20< irker182> wesnoth: Chris Beck wesnoth:1.12 23f788d16f65 / src/hotkey/command_executor.cpp: fix a segfault, reported #23423 http://git.io/veE0B
20150406 19:25:25< gfgtdf> iceiceice: i think teh bug report number is wrong
20150406 19:25:28< irker182> wesnoth: Chris Beck wesnoth:1.12 00c85e0031dc / changelog: update changelog http://git.io/veE0b
20150406 19:25:37< iceiceice> oh you're right
20150406 19:25:40< iceiceice> shoot
20150406 19:25:53< iceiceice> :///
20150406 19:26:01< iceiceice> should i force push it
20150406 19:26:07< iceiceice> or not worth it
20150406 19:26:47< iceiceice> i'm going to force push to fix it, its only 1.12 branch
20150406 19:26:50< iceiceice> not like people are pulling that constantly
20150406 19:27:06< gfgtdf> iceiceice: hm
20150406 19:27:24< gfgtdf> iceiceice: forcepushing will oynl change teha commti right ?
20150406 19:27:29< iceiceice> yes
20150406 19:28:25< irker182> wesnoth: Chris Beck wesnoth:1.12 42ce10d135d6 / src/hotkey/command_executor.cpp: fix a segfault, reported #23243 http://git.io/veEuW
20150406 19:28:27< irker182> wesnoth: Chris Beck wesnoth:1.12 39aac03cf6aa / changelog: update changelog http://git.io/veEu8
20150406 19:37:57< irker182> wesnoth: Chris Beck wesnoth:master 50dacfcdbf3a / src/hotkey/command_executor.cpp: fix a segfault, reported #23243 http://git.io/veEVO
20150406 19:37:59< irker182> wesnoth: Chris Beck wesnoth:master 5e978c42fcf4 / changelog: update changelog http://git.io/veEV3
20150406 19:42:21-!- travis-ci [~travis-ci@ec2-54-80-166-53.compute-1.amazonaws.com] has joined #wesnoth-dev
20150406 19:42:22< travis-ci> wesnoth/wesnoth#6191 (1.12 - 00c85e0 : Chris Beck): The build has errored.
20150406 19:42:22< travis-ci> Build details : http://travis-ci.org/wesnoth/wesnoth/builds/57379838
20150406 19:42:22-!- travis-ci [~travis-ci@ec2-54-80-166-53.compute-1.amazonaws.com] has left #wesnoth-dev []
20150406 19:53:00-!- travis-ci [~travis-ci@ec2-54-162-185-52.compute-1.amazonaws.com] has joined #wesnoth-dev
20150406 19:53:01< travis-ci> wesnoth/wesnoth#6190 (1.12 - 23f788d : Chris Beck): The build has errored.
20150406 19:53:01< travis-ci> Build details : http://travis-ci.org/wesnoth/wesnoth/builds/57379684
20150406 19:53:01-!- travis-ci [~travis-ci@ec2-54-162-185-52.compute-1.amazonaws.com] has left #wesnoth-dev []
20150406 19:54:58-!- new_one [~new_one@128.42.80.23] has joined #wesnoth-dev
20150406 20:00:31< nurupo> when in doubpt -- don't force push
20150406 20:00:52< nurupo> even if you are not in doubpt -- don't force push
20150406 20:01:02< iceiceice> i dont agree, i think it was fine there
20150406 20:01:42< iceiceice> if it was 1 year ago i wouldnt have done it, if it wasn't caught immediately i wouldnt have done it
20150406 20:02:15< nurupo> haven't read anything about "there", it's just my general rule :)
20150406 20:22:30-!- cib0 [~cib@p5DC743F0.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 20:28:54< shadowm> iceiceice: Did you actually force-push?
20150406 20:29:41< iceiceice> yes
20150406 20:29:43< shadowm> I thought I told you people to not do that on wesnoth/wesnoth ever.
20150406 20:29:48< shadowm> What the hell?
20150406 20:30:07< shadowm> I can understand a mistake like last time, but intentionally?
20150406 20:30:52< shadowm> For purely cosmetical reasons, no less.
20150406 20:32:26-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150406 20:32:31< iceiceice> did you get bad refs?
20150406 20:33:16< shadowm> I don't care whether anyone reports problems or not, the mentality of "it's okay if nobody notices/nothing bad happens by luck" is harmful to everyone.
20150406 20:33:28< iceiceice> i made a judgment
20150406 20:33:43< shadowm> And I am the repository administrator and I have decided your judgment is wrong.
20150406 20:33:45-!- new_one [~new_one@128.42.80.23] has quit [Quit: WeeChat 1.1.1]
20150406 20:33:54< iceiceice> you are welcome to your opinion
20150406 20:35:34< shadowm> Look, I just don't want this to happen ever again. Next time you find yourself wondering whether to force-push to wesnoth/wesnoth you will ask me and/or Ivanovic and/or AI0867 first and do as we decide. Is that clear?
20150406 20:39:13-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Read error: Connection reset by peer]
20150406 20:41:16-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 240 seconds]
20150406 20:42:21< iceiceice> sure
20150406 20:43:44-!- kex [~kex@92.53.39.44] has quit [Remote host closed the connection]
20150406 20:44:43-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150406 20:45:38< shadowm> And FYI baldras pulls all branches and tags from wesnoth/wesnoth periodically, and gettext.w.o is the primary automated consumer. How often it does this is beside the point.
20150406 20:46:32< iceiceice> that would be good information to share with developers
20150406 20:46:38< iceiceice> for instance on the wiki
20150406 20:46:42< shadowm> It's not relevant to developers.
20150406 20:46:45< iceiceice> it is relevant
20150406 20:47:00< iceiceice> if i'd known that it would have affected my decision to force push earlier
20150406 20:48:58< shadowm> Your decision shouldn't be affected by a factor that's considerably less relevant than the existence of an indefinite minus one number of users pulling from a public upstream repository at unspecified times.
20150406 20:50:11< iceiceice> well i dont have to explain to you why it would be
20150406 20:50:31< shadowm> Compared to that set of people, any issues arising from baldras specifically pulling before and after a force-push are trivially solved.
20150406 20:50:41< iceiceice> most programmers are intelligent people, they usually prefer to know why they shouldn't do things rather than that they shouldn't do things
20150406 20:50:53< iceiceice> i think i actually have a pretty good idea of how many people are pulling 1.12 branch and how often
20150406 20:51:17< shadowm> You think you do because you think all those people are on IRC.
20150406 20:52:15< iceiceice> how many times do you think 1.12 branch tip is pulled a day.
20150406 20:52:34< iceiceice> it's something we can ask github
20150406 20:52:59< shadowm> Statistics are less relevant than the actual possibilities here.
20150406 20:54:08< iceiceice> yes, i think that if someone got bad refs on 1.12 branch tip, they would almost surely come on irc and we could help them
20150406 20:54:13< shadowm> Also, the git push manpage says this: "In such a case, and only if you are certain that nobody in the meantime fetched your earlier commit A (and started building on top of it), you can run "git push --force" to overwrite it. In other words, "git push --force" is a method reserved for a case where you do mean to lose history."
20150406 20:54:17< iceiceice> if baldras got fouled, that's more seriously a pain in the ass
20150406 20:54:22< shadowm> And the point is you cannot be certain.
20150406 20:54:31< iceiceice> yes thats right, i made a judgment
20150406 20:54:46< shadowm> Why invite the possibility of giving someone trouble for a tiny cosmetic change?
20150406 20:55:01-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150406 20:55:04< iceiceice> i made a judgement about the relative possibilities of each and the costs of each, obviously
20150406 20:55:21-!- mjs-de [~mjs-de@p508C873D.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20150406 20:55:32< iceiceice> i dont know what we are arguing about
20150406 20:55:51-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 276 seconds]
20150406 20:55:51< iceiceice> are you seriously trying to argue that it was objectively wrong to force push there?
20150406 20:56:01< shadowm> Yes.
20150406 20:56:04< iceiceice> i can see that its against your policy
20150406 20:56:05< iceiceice> thats fine
20150406 20:57:04-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150406 20:58:24-!- new_one [~new_one@128.42.86.21] has joined #wesnoth-dev
20150406 21:05:40-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20150406 21:06:12-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20150406 21:07:55-!- Crendgrim_ [~crend@wesnoth/forum-moderator/crendgrim] has joined #wesnoth-dev
20150406 21:08:49-!- Crendgrim [~crend@wesnoth/forum-moderator/crendgrim] has quit [Ping timeout: 264 seconds]
20150406 21:10:52-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20150406 21:34:22-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20150406 21:37:18-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20150406 21:37:20-!- ancestral [~ancestral@63.92.240.233] has quit [Client Quit]
20150406 21:42:54-!- ancestral [~ancestral@63.92.240.233] has joined #wesnoth-dev
20150406 21:45:22-!- cib0 [~cib@p5DC743F0.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
20150406 21:48:03-!- kex [~kex@78.157.29.160] has joined #wesnoth-dev
20150406 21:55:48-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 245 seconds]
20150406 22:05:26< shadowm> CMake has no officially-sanctioned mechanism to select a default CMAKE_BUILD_TYPE, it seems.
20150406 22:06:47< shadowm> gfgtdf: What I get from the post on SO you linked earlier is that we should use -Wno-maybe-uninitialized?
20150406 22:07:49-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20150406 22:08:59< shadowm> This curious because the issue doesn't happen with cmake for me, so it's not a "cmake-specific problem" like iceiceice said. It's something about Ivanovic's Boost version, or maybe some patch to gcc 4.9.2 one of us has/doesn't have.
20150406 22:09:17< shadowm> And the only -W flag scons passes besides -W -Wall -Werror is -Wno-unised-local-typedefs.
20150406 22:09:33-!- new_one [~new_one@128.42.86.21] has quit [Quit: WeeChat 1.1.1]
20150406 22:10:23< shadowm> cmake does pass a few scons doesn't, though: -Wno-strict-aliasing -Wno-literal-suffix, -Wold-style-cast.
20150406 22:11:15< iceiceice> i think i added no-unused-local-typedefs to suppress errors in boost
20150406 22:11:19< iceiceice> so another thing is,
20150406 22:11:20< shadowm> The base set of -W -Wall -Werror -Wno-unused-local-typedefs is passed by both scons and cmake.
20150406 22:11:31< iceiceice> i realized that we aren't actually including libraries in the proper way perhaps
20150406 22:11:44-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150406 22:11:47< iceiceice> one of the gcc options is to include libraries using -isystem insteaad of -I
20150406 22:11:53< shadowm> I guess we should add -Wno-strict-aliasing -Wno-literal-suffix -Wold-style-cast to scons.
20150406 22:11:54< iceiceice> if you do that, then ti doesn't report warnings found in library code
20150406 22:12:08< iceiceice> i made a patch to do that in anura which was merged recently
20150406 22:12:11< shadowm> (Even though none of the warnings in question are emitted atm. I don't know..)
20150406 22:12:29< iceiceice> i think maybe we should just use -isystem for libs in both scons and cmake
20150406 22:12:36< iceiceice> but i dont actually know how to do that in either right now
20150406 22:12:51< iceiceice> it would probably fix a lot of these headaches if it just killed all the boost warnings
20150406 22:13:17< iceiceice> https://github.com/anura-engine/anura/commit/10f7e1c97733e5d192f314cecbc4a96e671fbf8b
20150406 22:15:09< shadowm> We use -isystem with cmake atm, but Boost isn't in the list of options, presumably because it's always included with paths relative to /usr/include.
20150406 22:15:25< shadowm> Surely you don't need to -isystem /usr/include?
20150406 22:16:12< iceiceice> i think you should probably
20150406 22:16:22< iceiceice> since we dont want to hear about warnings in /usr/include
20150406 22:16:28< iceiceice> it sounds like its hard to do this in scons
20150406 22:16:28< iceiceice> http://stackoverflow.com/questions/2441047/how-do-i-set-scons-system-include-path
20150406 22:16:35< iceiceice> but its not a problem right now i guess
20150406 22:17:01< iceiceice> i think everything that's not a header in our src/ directory should be -isystem probably
20150406 22:17:04< shadowm> Also, from Ivanovic's paste, gcc understands the error to be in our code rather than in a Boost header due to template instantiation.
20150406 22:17:19< shadowm> *warning treated as error
20150406 22:17:40< iceiceice> yeah i dont know how that interaction would work
20150406 22:17:51< iceiceice> it might suppress it, it might not i guess
20150406 22:17:58< gfgtdf> shadowm: i dont know more that that page, If this is a false positive i reccomend to disable the warning, at least for that code section
20150406 22:18:16< iceiceice> i think i am already suppressing maybe-uninitialized in the xbrz code
20150406 22:18:41< iceiceice> https://github.com/wesnoth/wesnoth/blob/master/src/xBRZ/xbrz.cpp#L49
20150406 22:23:08-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Read error: Connection reset by peer]
20150406 22:27:38-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150406 22:30:40-!- ancestral [~ancestral@63.92.240.233] has quit [Quit: i go nstuf kthxbai]
20150406 22:31:20< shadowm> Oh huh.
20150406 22:31:35< shadowm> This is weird. I got the same error as Ivanovic with scons this time but I didn't get it with cmake earlier.
20150406 22:32:24< shadowm> I was making a debug build with cmake then and now this is an optimized (-O3) build, so it must be a bad interaction with one of the -O3 optimizations.
20150406 22:32:52< shadowm> cmake uses -O3 by default for optimized (Release) builds too, scons uses -O2 by default but I always use -O3 instead.
20150406 22:34:39< iceiceice> from what i read in the cmake docs way back when,
20150406 22:34:52< iceiceice> cmake is somewhere between scons and make in terms of build rigor
20150406 22:35:18< iceiceice> it might be that you needed to clean build or something
20150406 22:35:50< shadowm> No.
20150406 22:36:16< shadowm> I'll try -O0 with scons to see what happens but I'm pretty sure now it's optimizations.
20150406 22:36:53< shadowm> Also, it *was* a pristine build dir.
20150406 22:37:21< shadowm> (And cmake uses make.)
20150406 22:42:40< shadowm> Yep, it's an optimization's fault.
20150406 22:46:55< shadowm> :|
20150406 22:47:33< shadowm> An optimization that's actually in -O2, not just -O3.
20150406 22:47:35< iceiceice> shadowm: "Cmake uses make" is besides the point, cmake also does additional checks beyond what make does
20150406 22:47:53< shadowm> What gcc version are you building with there?
20150406 22:48:02< iceiceice> that is what is referred to in the last sentence on wiki in cmake section
20150406 22:48:10< iceiceice> it doesnt do as much as scons though
20150406 22:48:13< iceiceice> its also faster though
20150406 22:48:16< iceiceice> ummm
20150406 22:48:19< iceiceice> i think im gcc 4.9.2
20150406 22:48:28< shadowm> Are you only building debug builds?
20150406 22:48:37< shadowm> Or do you have strict mode off or something?
20150406 22:48:43< iceiceice> i usually build -O0 or -O2 and i use strict mode i think
20150406 22:48:47< iceiceice> but i usually use clang
20150406 22:48:54< shadowm> Well, I get the error with -O2.
20150406 22:48:54< iceiceice> what exact build is failing for you?
20150406 22:49:03< iceiceice> 1.12 with gcc -O2?
20150406 22:49:08< shadowm> master.
20150406 22:49:19< shadowm> Ivanovic was trying to build master.
20150406 22:49:36< shadowm> I also get it with -O1.
20150406 22:49:59< iceiceice> i usually build master with release option and -O0
20150406 22:50:05< shadowm> Now the question is what optimization is causing this and whether the warning is a false positive or it's actually producing bad code.
20150406 22:50:07< iceiceice> i'm going to do that with gcc now
20150406 22:50:30< shadowm> -Os is unaffected.
20150406 22:51:08< iceiceice> do you have ivanovic' pastebin?
20150406 22:51:13< shadowm> Save yourself some time and just copy the command line and try to build src/scripting/game_lua_kernel.cpp alone.
20150406 22:53:15< shadowm> iceiceice: http://pastebin.com/PQrKywmm
20150406 22:53:27< shadowm> The last run at the bottom obviously succeeded.
20150406 22:53:52< shadowm> The -O1 run seems to provide more information than -O2 and -O3.
20150406 22:55:18< iceiceice> i think its just constructor (2) here: http://en.cppreference.com/w/cpp/container/vector/vector
20150406 22:55:37< iceiceice> if there's anything actually uninitialized there i think its a bug in the STL
20150406 22:55:46< iceiceice> so i would bet that its a false positive
20150406 22:56:36< iceiceice> oh thats the wrong line
20150406 22:57:03< iceiceice> i still dont think it should be uninitialized
20150406 22:57:43< iceiceice> my build completed i didnt get an error
20150406 22:57:56< iceiceice> also travis build gcc -O0 builds frequently
20150406 22:58:54< iceiceice> http://pastebin.com/cyAmz6mh
20150406 23:04:46< iceiceice> my boost version is 1.54 also
20150406 23:04:48< iceiceice> that might be relevant
20150406 23:04:58< iceiceice> i guess that you have boost 1.56 since you had this boost core stuff that i dont have
20150406 23:05:12< iceiceice> 1.56 is not available in mint yet
20150406 23:12:09-!- irker182 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20150406 23:13:14< shadowm> I said I didn't have Boost.core, I copied over the relevant file to my project instead. I have Boost 1.55.
20150406 23:13:19-!- kex [~kex@78.157.29.160] has quit [Ping timeout: 255 seconds]
20150406 23:13:37-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 264 seconds]
20150406 23:14:12< shadowm> iceiceice: You are using -O0, of course it'll not hit this warning.
20150406 23:14:28< iceiceice> oh
20150406 23:14:49< shadowm> There's a reason I advised running the compiler directly without scons or cmake.
20150406 23:14:51-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 252 seconds]
20150406 23:15:13< iceiceice> travis is running gcc -O2 builds
20150406 23:15:34< shadowm> It's that it takes only a few seconds regardless of your current options because you don't need to wait for a zillion other files to build with different options.
20150406 23:15:36< iceiceice> oh maybe not
20150406 23:15:44< iceiceice> i think gcc -O2 was too slow so i disabled it on travis
20150406 23:15:48< shadowm> You said travis is not using gcc 4.9.
20150406 23:15:54< iceiceice> yeah thats true also
20150406 23:18:21< iceiceice> ok i get maybe uninitialized warning
20150406 23:18:30-!- kex [~kex@89.205.97.191] has joined #wesnoth-dev
20150406 23:21:49< shadowm> Tuple members are 0-based, right?
20150406 23:21:58< shadowm> Yeah, they are.
20150406 23:23:44< shadowm> 1657 guarantees that the first board().teams().size() tuples in the vector are fully initialized.
20150406 23:23:59< shadowm> The vector's size remains constant afterwards AFAICT.
20150406 23:24:30< iceiceice> its not that line
20150406 23:24:32< iceiceice> gfgtdf was right
20150406 23:24:35< iceiceice> its the boost ::optional
20150406 23:24:54< shadowm> In the loop starting at 1691 result is a copy of any one of the tuples in the vector.
20150406 23:25:39< shadowm> The problematic lines copy the tuple's members.
20150406 23:25:53< shadowm> Which are guaranteed to be fully initialized.
20150406 23:26:56< iceiceice> so, i think we should consider to pass -Wno-maybe-uninitialized, because this is pretty lame
20150406 23:27:09< iceiceice> but if not i think we should do so in this file
20150406 23:27:23< shadowm> Is there a way to test that the code produced by the compiler here is correct anyway?
20150406 23:27:47< iceiceice> run coverity and see if it complains?
20150406 23:27:52< shadowm> Because warning or not, it's not unheard of for compiler optimizations to break things.
20150406 23:28:14< iceiceice> you could stick asserts in for all of the members
20150406 23:28:30< iceiceice> i guess its not easy to assert that something was initialized though actually
20150406 23:28:56< shadowm> I could use valgrind. How do I trigger this code?
20150406 23:29:13< iceiceice> use end_level
20150406 23:29:16< iceiceice> heh,
20150406 23:29:23< shadowm> Any settings will do?
20150406 23:29:24< iceiceice> you could probably just use valgrind with any of the unit tests
20150406 23:30:01< shadowm> You mean the WML unit tests, right?
20150406 23:30:04< iceiceice> yes
20150406 23:30:15< iceiceice> i think if the "carryover_percentage" fields are not present in the wml, then they will be supplied/
20150406 23:30:16< iceiceice> ?
20150406 23:31:02< iceiceice> i dont think those are present in the wml unit tests
20150406 23:31:41< iceiceice> if you want to change what the [end_level] tags look like in the unit tests, you can change this macro:
20150406 23:31:41< iceiceice> https://github.com/wesnoth/wesnoth/blob/master/data/test/macros/wml_unit_test_macros.cfg#L13
20150406 23:31:47< shadowm> I don't know what the apply_visitor() calls do.
20150406 23:31:59< shadowm> This code seems rather unusual in that regard.
20150406 23:33:11< iceiceice> i think its using boost::variant
20150406 23:33:20< shadowm> gfgtdf: I think you wrote this.
20150406 23:33:40< shadowm> What's the reason for using apply_visitor() instead of e.g. to_int() or such?
20150406 23:34:20< iceiceice> i think because he wants to directly get an optional
20150406 23:34:47< shadowm> But he uses to_bool() in the same block and that returns a plain bool.
20150406 23:34:53-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 246 seconds]
20150406 23:35:59< iceiceice> there's a code comment around 1619
20150406 23:36:18< gfgtdf> shadowm: i think that si becasue for "carryover_add" we allow to spcify a default in [scenario] but not for bonus
20150406 23:36:46< gfgtdf> iceiceice: that comment is just abotu teh imeplmentation of that struct
20150406 23:36:49-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150406 23:36:49< iceiceice> oh
20150406 23:37:05< iceiceice> y i see
20150406 23:37:10< gfgtdf> iceiceice: about why i use yes_no , true_false explicitly and not bool
20150406 23:37:26-!- cib0 [~cib@p5DC743F0.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150406 23:37:53< iceiceice> shadowm: i think it depends on exactly what variable you think might be uninitialized,
20150406 23:37:59< iceiceice> its not really clear in the warning imo
20150406 23:38:13< iceiceice> but i think if you are really worried you could use valgrind with wml unit tests
20150406 23:38:14< shadowm> Members 2 and 3 of the tuple.
20150406 23:38:24< shadowm> That's carryover_percentage and carryover_add.
20150406 23:39:08< shadowm> This code checks for the case where they are not provided, right?
20150406 23:39:52< shadowm> I see the boost::optional<>::get() calls and noticed the code says that calling that if the optional isn't initialized is UB.
20150406 23:40:06< shadowm> *the code comments in the Boost header
20150406 23:40:30-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20150406 23:40:34< iceiceice> i think thats boost::tuple<>::get()
20150406 23:40:46< shadowm> No, I'm in optional.hpp.
20150406 23:40:48< iceiceice> i dont see where we are calling boost::optional::get
20150406 23:41:13< shadowm> Ah, you are right, it's boost::tuple<>::get().
20150406 23:42:23< shadowm> team's methods take plain values but we call optional<>::get() though.
20150406 23:42:43< shadowm> On lines 1713 and 1716.
20150406 23:43:05< shadowm> I think that's actually a problem?
20150406 23:44:14< gfgtdf> shadowm: not we chach whther teh value is set in the check above
20150406 23:44:30< gfgtdf> s/not/no
20150406 23:44:42< gfgtdf> catch whether the*
20150406 23:45:16< iceiceice> yeah its fine
20150406 23:45:41< shadowm> Right, the implicit conversion is the initialized state.
20150406 23:48:26< shadowm> Hm, I got an invalid read warning from memcheck.
20150406 23:48:37< shadowm> It's not from this code yet, though, because Wesnoth isn't even finished initializing.
20150406 23:49:37< shadowm> Ugh, not invalid read, just use of uninit value. Still not sure if it's important.
20150406 23:50:33< shadowm> http://pastebin.com/PAcwHraJ -- it may come from library code but I'm not sure.
20150406 23:51:35< shadowm> A plain `lua wesnoth.wml_actions.endlevel {}` didn't trigger any warnings.
20150406 23:52:25< iceiceice> yeah i get that also
20150406 23:52:32< iceiceice> the uninit value
20150406 23:53:51< shadowm> No warnings with `lua wesnoth.wml_actions.endlevel { result = "victory", bonus = true, carryover_percentage = "1000", carryover_add = true }` either.
20150406 23:54:09< iceiceice> do you see any problems here?
20150406 23:54:09< iceiceice> https://github.com/wesnoth/wesnoth/blob/master/src/mt_rng.cpp#L98
20150406 23:56:06< shadowm> The only size 8 value there is sizeof(uint32_t) which is not a variable, so I don't think so.
20150406 23:56:46< shadowm> It's probably just library noise.
20150406 23:58:35< shadowm> So is it possible to silence the warnings in that section of game_lua_kernel.cpp specifically, or should we just -Wno-maybe-uninitialized everywhere?
20150406 23:58:46< iceiceice> the memcheck thing i'm actually worried about actually
20150406 23:59:19< iceiceice> shadowm: if you want i will disable Wno-maybe-unitialized in that file
20150406 23:59:32< iceiceice> i think we didnt encounter it anywhere besides there and xbrz
--- Log closed Tue Apr 07 00:00:26 2015