--- Log opened Sun Apr 12 00:00:05 2015
--- Day changed Sun Apr 12 2015
20150412 00:00:05< gfgtdf> iceiceice: even with that commit mesage i dont really know why it does these cheks
20150412 00:00:26< iceiceice> gfgtdf: it sounds like... hes trying to hack it so that the whiteboard fake units are visible
20150412 00:00:42< iceiceice> even when they move to a currently fogged spot
20150412 00:02:23< gfgtdf> iceiceice: i dont think it makes sense to plan unit moving to a fogges spot
20150412 00:02:41< gfgtdf> iceiceice: esp since one doesnt knwo whether that hex might be wall or similar.
20150412 00:03:17< iceiceice> well
20150412 00:03:24< iceiceice> if its only fogged and not shrouded it makes sense
20150412 00:03:45< gfgtdf> iceiceice: ah ok
20150412 00:04:38< iceiceice> it doesnt amke sense the code though
20150412 00:04:47< gfgtdf> iceiceice: but in the current implementation it also shows for shrouded hexes i think
20150412 00:04:50< iceiceice> the code should just say "if i'm a ghost unit then i am visible"
20150412 00:04:59< iceiceice> and that should handle all of the whiteboard cases
20150412 00:05:38< gfgtdf> iceiceice: yes but afaik there is no easy way to check whether a unit is ghost.
20150412 00:06:53< gfgtdf> iceiceice: also you shouldnt see teh enmie ghost units
20150412 00:07:05< iceiceice> the enemy ghost units aren't transmitted over network though
20150412 00:07:09< iceiceice> ohh hmm
20150412 00:07:12< iceiceice> well i think 
20150412 00:07:48< iceiceice> i still think the current code looks very dirty and your code looks much cleaner
20150412 00:07:53< iceiceice> i dont think the whiteboard should need that stuff
20150412 00:08:00< iceiceice> if it does we can change it more easily i think
20150412 00:08:09< iceiceice> b/c the display object knows which units are ghosts i think
20150412 00:08:26< iceiceice> we could just add a "am i a ghost" function probably anyways
20150412 00:09:59-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150412 00:11:21-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has joined #wesnoth-dev
20150412 00:16:55-!- [Relic] [~Relic]@2602:306:33a3:6d30:c12a:16f1:3738:e91] has joined #wesnoth-dev
20150412 00:26:07< vultraz> iceiceice: this is off topic to the issue, but do you think we should remove shroud_data=? I don't think anyone uses it
20150412 00:26:38< iceiceice> in scenario?
20150412 00:26:42< iceiceice> i think its needed
20150412 00:26:49< vultraz> no, side
20150412 00:26:54< iceiceice> so,
20150412 00:27:00< iceiceice> when you reload a save game it has to reconstruct the shroud and fog data
20150412 00:27:10< iceiceice> you cant do it just by looking at where the units are
20150412 00:27:16< iceiceice> because wml ight have used place_shroud
20150412 00:27:22< iceiceice> so it has to be written down by the game when you save
20150412 00:27:44< iceiceice> so the feature always has to be there i think
20150412 00:29:58< vultraz> still, dunno why the player gets to mess with it
20150412 00:30:00< vultraz> http://wiki.wesnoth.org/BuildingScenariosShroudData
20150412 00:55:24-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20150412 01:00:57-!- prkc [~prkc@54020B6B.dsl.pool.telekom.hu] has quit [Remote host closed the connection]
20150412 01:03:19-!- gfgtdf [~chatzilla@f054141061.adsl.alicedsl.de] has quit [Read error: Connection reset by peer]
20150412 01:08:35-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20150412 01:11:53-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150412 01:15:39< iceiceice> vultraz: i mean because why not
20150412 01:15:55< iceiceice> ok, having a detailed page like that is pretty silly 
20150412 01:16:06< iceiceice> because thats not something someone would normally want to do
20150412 01:16:14< iceiceice> but theres no point to just disable features for no gain
20150412 01:16:28< iceiceice> its not like its less stuff for us to maintain
20150412 01:29:40-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20150412 01:32:07< shadowm> ./episode3/scenarios/08C_Breakdown.cfg:1499:            shroud_data=$shroud_store
20150412 01:32:13< shadowm> ./episode3/scenarios/08A_Interim.cfg:1753:            shroud_data=$floor1_shroudmap
20150412 01:32:18< shadowm> ./episode1/scenarios/12_The_Queen.cfg:374:                    shroud_data=$rtw_end_gamestate.shroudmap
20150412 01:32:23 * shadowm kicks vultraz.
20150412 01:32:37-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150412 01:34:09-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150412 01:34:40< iceiceice> i see we've moved on from squeaky hammer to full-on beat downs :O
20150412 01:37:43-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20150412 01:39:27-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Quit: tomreyn]
20150412 02:03:14< iceiceice> hmmm i'm getting a wierd preprocessor error:
20150412 02:03:15< iceiceice> The following add-on had errors and could not be loaded:
20150412 02:03:16< iceiceice>     /home/chris/.local/share/wesnoth/1.13/data/add-ons/Silly_AI_Example/_main.cfg
20150412 02:03:16< iceiceice> Please report this to the author or maintainer of this add-on.
20150412 02:03:16< iceiceice> Details:
20150412 02:03:16< iceiceice>     Undefined macro in #ifver/#ifnver first argument: 'WESNOTH_VERSION'
20150412 02:03:17< iceiceice>     at ~add-ons/Silly_AI_Example/_main.cfg:26
20150412 02:03:39< iceiceice> shouldn't WESNOTH_VERSION always be defined?
20150412 02:04:59< iceiceice> i am working around it with this:
20150412 02:04:59< iceiceice> #ifndef WESNOTH_VERSION
20150412 02:04:59< iceiceice> #define WESNOTH_VERSION
20150412 02:05:00< iceiceice> 1.13.0
20150412 02:05:00< iceiceice> #enddef
20150412 02:05:00< iceiceice> #endif
20150412 02:07:54< shadowm> Just F5.
20150412 02:08:45< shadowm> It's a bug with the config cache code that causes that to happen for add-ons that don't need to be recached and previously failed a loading step.
20150412 02:09:49< irker523> wesnoth: Chris Beck wesnoth:master e28a6bfcc6eb / src/serialization/preprocessor.cpp: give a better error message in preprocessor when #ifver fails http://git.io/vvkav
20150412 02:09:51< irker523> wesnoth: Chris Beck wesnoth:master a5d3c64d10da / / (4 files in 3 dirs): move "modify_ai" to the game lua kernel http://git.io/vvkaf
20150412 02:10:28-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20150412 02:19:43-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150412 02:34:51-!- ancientcc [~ancientcc@27.115.116.212] has joined #wesnoth-dev
20150412 02:39:43-!- ancientcc [~ancientcc@27.115.116.212] has quit [Remote host closed the connection]
20150412 02:46:46-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 02:50:04-!- Necrosporus [~Necrospor@unaffiliated/necrosporus] has joined #wesnoth-dev
20150412 02:51:56-!- Necrosporus_ [~Necrospor@unaffiliated/necrosporus] has quit [Ping timeout: 240 seconds]
20150412 02:52:23-!- ancientcc [~ancientcc@211.162.34.205] has quit [Quit: Leaving]
20150412 02:53:34< irker523> wesnoth: Ignacio R. Morelle wesnoth:master fe64aa9fdac4 / CMakeLists.txt src/CMakeLists.txt: cmake: Remove ENABLE_BOOST_FILESYSTEM option http://git.io/vvk61
20150412 02:56:22< irker523> wesnoth: Chris Beck wesnoth:master f3a639fc6284 / / (4 files in 3 dirs): move [print] to game_lua_kernel http://git.io/vvkie
20150412 02:56:24< irker523> wesnoth: Chris Beck wesnoth:master 1bd5cfb7568e / src/scripting/lua_kernel_base.cpp: fixup lua dofile to use a protected call http://git.io/vvkiv
20150412 02:56:26< irker523> wesnoth: Chris Beck wesnoth:master 8f6d3a1a6dfc / CMakeLists.txt src/CMakeLists.txt: Merge branch 'master' of https://github.com/wesnoth/wesnoth http://git.io/vvkif
20150412 02:59:34< vultraz> uhhhhh
20150412 02:59:48< vultraz> ok
20150412 02:59:49< vultraz> what is this..
20150412 03:02:32-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 03:03:01< vultraz> so this happens in 1.12 https://www.dropbox.com/s/xjdjrvf9hp3uqtm/1.12mouseover%20undraw.png?dl=0
20150412 03:03:55< vultraz> the text undrawing, not the error, that's me
20150412 03:04:37< vultraz> mousing over the text while that window is open undraws it. not a thing in master. bleh
20150412 03:04:43< iceiceice> vultraz: i think something is screwd upwith the wesnoth version macro...
20150412 03:04:44< vultraz> Will file a bug later
20150412 03:05:04< iceiceice> Message "shadows of deception requries 1.11.11 or later" ...
20150412 03:05:11< iceiceice> if thats 1.12 why are you seeing that
20150412 03:05:12< vultraz> I said, ignore that :P
20150412 03:05:36< vultraz> I deliberately changed my version failsafe in my addon to cause that to pop up
20150412 03:06:33-!- ancientcc [~ancientcc@112.65.211.213] has quit [Remote host closed the connection]
20150412 03:09:54< vultraz> if v(ver, '>=', '1.12.2') then
20150412 03:09:55< vultraz> 	do_bug( _ "Shadows of Deception requires Wesnoth 1.11.11 or later.", false)
20150412 03:09:57< vultraz> end
20150412 03:09:58< vultraz> my own code
20150412 03:10:03< vultraz> I just didn't change the message
20150412 03:10:22< vultraz> iceiceice: the bug is that a  gui2 dialog there causes undrawing of the text when you mouse over
20150412 03:10:44< iceiceice> of the tooltip?
20150412 03:10:52< vultraz> the storytext
20150412 03:11:05< iceiceice> vultraz: i think all of those bugs are intractable probably :/
20150412 03:11:16< iceiceice> unless we port story text to gui2
20150412 03:11:31< vultraz> It doesn't happen in master
20150412 03:11:39< vultraz> also confirmed to happen with the screenshot dialog in the editopr
20150412 03:11:58< iceiceice> i don't have confidence that i could figure out why it doesnt or doesnt happen in any case in any reasonable amount of time
20150412 03:12:12< iceiceice> (where reasonable is proportional to how much i care about this bug)
20150412 03:12:47< iceiceice> maybe someone else has more expertise
20150412 03:12:55-!- ancientcc [~ancientcc@123.155.153.55] has joined #wesnoth-dev
20150412 03:15:09-!- ancientcc [~ancientcc@123.155.153.55] has quit [Client Quit]
20150412 03:17:06< shadowm> 00:04:45 <iceiceice> vultraz: i think something is screwd upwith the wesnoth version macro...
20150412 03:17:13< shadowm> I already explained this above.
20150412 03:17:40< vultraz> shadowm: he was referring to my screenshto
20150412 03:17:45< shadowm> Or is it some other heretofore unmentioned thing that I might want to know as the author of the #ifdef feature?
20150412 03:17:58< iceiceice> shadowm: i couldn't get it to work using f5
20150412 03:18:10< shadowm> Why would the "wesnoth version macro" have anything to do with GUI drawing routines being crap? :p
20150412 03:18:29< vultraz> A better question is who fixed it in master!
20150412 03:18:33< shadowm> iceiceice: Purge cache + reload then.
20150412 03:22:30< shadowm> If you could submit a bug report and assign it to me (I have more pressing things to do atm) I'll dig up later my IRC rant where I think I managed to get fairly close to the root cause of this.
20150412 03:22:40< shadowm> From like 2 years ago.
20150412 03:24:00-!- ancientcc [~ancientcc@27.115.116.212] has joined #wesnoth-dev
20150412 03:25:55-!- ancientcc [~ancientcc@27.115.116.212] has quit [Client Quit]
20150412 03:26:57-!- ancientcc [~ancientcc@123.157.71.143] has joined #wesnoth-dev
20150412 03:27:50-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Quit: Leaving]
20150412 03:41:29-!- new_one [~new_one@128.42.74.222] has quit [Remote host closed the connection]
20150412 03:49:41-!- Kwandulin [~Miranda@p5B00971C.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150412 03:54:16-!- ancientcc [~ancientcc@123.157.71.143] has quit [Quit: Leaving]
20150412 03:56:46-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 03:57:27-!- ancientcc [~ancientcc@211.162.34.205] has quit [Remote host closed the connection]
20150412 04:26:20-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has quit [Ping timeout: 276 seconds]
20150412 04:37:29-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20150412 04:58:45-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 04:59:13-!- kex [~kex@78.157.29.160] has quit [Remote host closed the connection]
20150412 05:01:06-!- ancientcc [~ancientcc@112.65.211.213] has quit [Client Quit]
20150412 05:02:15-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 05:03:00-!- ancientcc [~ancientcc@112.65.211.213] has quit [Remote host closed the connection]
20150412 05:05:13< shadowm> gfgtdf: I'm backporting commit 5a55d7ee012f93045e697eff5f2ea33041254345 to 1.12.
20150412 05:13:09-!- kex [~kex@78.157.29.160] has joined #wesnoth-dev
20150412 05:15:35< irker523> wesnoth: gfgtdf wesnoth:1.12 b79a88d05cb0 / src/filesystem_boost.cpp: fixup 20e9706c11f511b5ae48c3ee5666c13e3d3a78c7 http://git.io/vvIv7
20150412 05:15:37< irker523> wesnoth: Ignacio R. Morelle wesnoth:1.12 18557d342e65 / src/filesystem_boost.cpp: fs: Demote a few BFS-specific error log statements to info level http://git.io/vvIv5
20150412 05:17:01< irker523> wesnoth: Ignacio R. Morelle wesnoth:master 542b3a290492 / src/filesystem_boost.cpp: fs: Demote a few BFS-specific error log statements to info level http://git.io/vvIvj
20150412 05:30:31-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has joined #wesnoth-dev
20150412 05:32:02-!- [Relic] [~Relic]@2602:306:33a3:6d30:c12a:16f1:3738:e91] has quit [Quit: Leaving]
20150412 05:49:13-!- c74d3 [~c74d3a4eb@2002:4404:712c:0:76de:2bff:fed4:2766] has joined #wesnoth-dev
20150412 05:53:57-!- Netsplit *.net <-> *.split quits: c74d
20150412 06:11:13-!- kex [~kex@78.157.29.160] has quit [Remote host closed the connection]
20150412 06:11:52< shadowm> ancestral: Did you ever test the window minimize patch?
20150412 06:12:26< ancestral> shadowm: No, I was focused gettinng 1.12 out the door.
20150412 06:13:19< ancestral> I have the day off Sunday. I am now focused on building 1.13. I can test your path on Sunday.
20150412 06:13:26< ancestral> *patch
20150412 06:15:01< shadowm> I may not release 1.13.0 today, btw.
20150412 06:17:40< ancestral> Then I can build with master
20150412 06:48:59-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20150412 06:58:38-!- cib0 [~cib@p5DC75B4D.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150412 07:20:34-!- cib0 [~cib@p5DC75B4D.dip0.t-ipconnect.de] has quit [Quit: Leaving]
20150412 07:22:55-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20150412 07:39:39-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20150412 07:54:21-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 07:54:35-!- ancientcc [~ancientcc@112.65.211.213] has quit [Remote host closed the connection]
20150412 07:56:00-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 07:56:37-!- ancientcc [~ancientcc@112.65.211.213] has quit [Remote host closed the connection]
20150412 07:57:47< irker523> wesnoth: Ignacio R. Morelle website:master d87170d64a92 / start/1.12/ (63 files in 2 dirs): pofix pass for the 1.12.2 release http://git.io/vvIBr
20150412 07:57:50< irker523> wesnoth: Ignacio R. Morelle wesnoth:master 3ce8f05219f6 / utils/pofix.py: pofix: Add download path string fixes for the 1.12.2 release http://git.io/vvIBo
20150412 08:41:33-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 08:42:39-!- ancientcc [~ancientcc@112.65.211.213] has quit [Remote host closed the connection]
20150412 08:42:59-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 08:43:24-!- ancientcc [~ancientcc@112.65.211.213] has quit [Remote host closed the connection]
20150412 08:44:38-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 08:45:12-!- ancientcc [~ancientcc@112.65.211.213] has quit [Remote host closed the connection]
20150412 08:45:45< shadowm> Ivanovic: Isn't it too drastic to remove the Pandora option entirely? Removing its weird dependency exceptions should allow for it to work again once you or whoever gets the dependencies sorted out.
20150412 08:46:10< Ivanovic> what it initially was supposed to do is only to allow the resolution of 800x480
20150412 08:46:18< Ivanovic> no idea what others added in there to piggyback on it
20150412 08:46:39< Ivanovic> plus: i don't understand the other file system "workarounds" like using the old path when deactivating nls
20150412 08:47:14< Ivanovic> e.g. the gentoo builds have a flag where nls can be deactivated (for those who don't use translations), yet they might want to use the "more recent" filesystem implementation anyway
20150412 08:47:27< Ivanovic> so all the piggybacking in there is, uhm, extremely strange IMO
20150412 08:47:33-!- ancientcc [~ancientcc@27.115.116.212] has joined #wesnoth-dev
20150412 08:48:14-!- ancientcc [~ancientcc@27.115.116.212] has quit [Client Quit]
20150412 08:48:14< shadowm> I think ENABLE_NLS only switches on/off compiling translations, not the underlying implementation in the game binaries.
20150412 08:48:33-!- ancientcc [~ancientcc@27.115.116.212] has joined #wesnoth-dev
20150412 08:48:40< Ivanovic> then i understood some info by iceiceice wrong
20150412 08:49:04-!- ancientcc [~ancientcc@27.115.116.212] has quit [Client Quit]
20150412 08:49:49< shadowm> Okay, here's what ENABLE_PANDORA does at the CMake level (on master) besides passing -DPANDORA to the C++ compiler:
20150412 08:50:16-!- ancientcc [~ancientcc@27.115.116.212] has joined #wesnoth-dev
20150412 08:50:21< shadowm> 1) Disable vorbisfile dependency (otherwise mandatory)
20150412 08:50:52< shadowm> 2) Disable Boost.filesystem + Boost.locale dependencies (otherwise mandatory on master, with the exception that libintl may be used instead of Boost.locale)
20150412 08:51:38< shadowm> Though it seems the CMake logic goes more along the lines of "Boost.locale is always mandatory unless ENABLE_PANDORA=ON" in this case, which is not what I was told.
20150412 08:52:16-!- ancientcc [~ancientcc@27.115.116.212] has quit [Client Quit]
20150412 08:52:34< shadowm> 3) Require SDL_mixer >= 1.2.0 instead of SDL_mixer >= 1.2.12 (this works on 1.12, but on master there's code that will break at compile time unless SDL_mixer is >= 1.2.12)
20150412 08:53:03< shadowm> And that's it.
20150412 08:54:22< shadowm> The -DPANDORA changes are somewhat non-trivial, primarily compiling different code for one of our image scaling algorithms.
20150412 08:55:02< shadowm> I believe this is because you and mordante found performance issues with our default algorithm when run on the Pandora's processor.
20150412 08:55:40-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 08:56:07-!- ancientcc [~ancientcc@211.162.34.205] has quit [Remote host closed the connection]
20150412 08:56:45< shadowm> The other -DPANDORA changes are just excluding code that depends on the aforementioned vorbisfile dependency, and selecting the Pandora theme (optimized for 800x480) as the default Pandora theme instead of Default.
20150412 09:00:55< shadowm> SCons has no canned config support for Pandora-optimized builds at all. That said, it does have an option that specifically forces using libintl instead of Boost.locale for builds, which is probably why I was told that Boost.locale is not a mandatory dependency. :\
20150412 09:01:29< shadowm> I guess nobody has bothered to add such an option to CMake yet.
20150412 09:02:32< shadowm> Or actually, I just inadvertently *removed* the option since ENABLE_BOOST_FILESYSTEM doubled for both Boost.filesystem and Boost.gettext.
20150412 09:02:37< shadowm> /o\
20150412 09:03:14< shadowm> So I guess I have to bring it back in some form to get Boost.gettext's optionality back with CMake. *headdesk*
20150412 09:05:10< iceiceice> shadowm: here was the point of diveregence with regards to boost filesystem and gettext
20150412 09:05:11< iceiceice> https://github.com/wesnoth/wesnoth/pull/347
20150412 09:05:49< iceiceice> err
20150412 09:05:52< iceiceice> the conversation not the commits
20150412 09:06:04< iceiceice> this was the point of divergence
20150412 09:06:05< iceiceice> https://github.com/wesnoth/wesnoth/pull/351
20150412 09:06:08< shadowm> 06:00:55 <shadowm> SCons has no canned config support for Pandora-optimized builds at all. That said, it does have an option that specifically
20150412 09:06:09< iceiceice> which i sent an email about
20150412 09:06:22< shadowm> Sorry, I didn't meant to paste that.
20150412 09:06:58< shadowm> OKAY SO WHAT I NEED TO DO IS
20150412 09:07:29< shadowm> 1) (Re)add an option to CMake to make Boost.gettext optional in favor of libintl, complain loudly if running on Windows with this option enabled.
20150412 09:07:38< shadowm> 2) Pandora???
20150412 09:07:59< irker523> wesnoth: Nils Kneuper wesnoth:master 7513477224e2 / CMakeLists.txt src/CMakeLists.txt src/sound_music_track.cpp: remove most OpenPandora specific items http://git.io/vvIPJ
20150412 09:08:42< shadowm> 0) Get me a Boost.gettext-free test environment.
20150412 09:08:51-!- ancientcc [~ancientcc@123.157.71.143] has joined #wesnoth-dev
20150412 09:09:08< iceiceice> i thought boost random is also not available on pandora?
20150412 09:09:14< iceiceice> (atm?)
20150412 09:09:19< iceiceice> maybe i misremember thogu
20150412 09:09:28< shadowm> I guess this is an auspicious time to make me a Debian jessie chroot since I'll probably stick to testing next.
20150412 09:09:35< Ivanovic> i removed anything pandora specific besides the stuff required to have it work nicely under 800x480
20150412 09:09:38< iceiceice> regardless someone will have to gather a bunch of libs
20150412 09:09:57< Ivanovic> the question is: in which cases would we *NOT* want to use the boost.gettext stuff?
20150412 09:10:32< Ivanovic> meaning: i remember that we required those two paths for the 1.12 series due to issues with the pandora version
20150412 09:10:38< Ivanovic> now we can get rid of the two, right?
20150412 09:10:57< iceiceice> relevant: http://forums.wesnoth.org/viewtopic.php?f=10&t=41196
20150412 09:11:32-!- ancientcc [~ancientcc@123.157.71.143] has quit [Read error: Connection reset by peer]
20150412 09:12:51< Ivanovic> honestly, i don't understand our filesystem and lib setup these days
20150412 09:12:59< shadowm> IIRC the only remaining issue with Boost.gettext was having to bump the Apple OS X version requirements?
20150412 09:13:04< Ivanovic> the question is for which situations is which version best?
20150412 09:13:28< Ivanovic> is it "windows only" where you really need boost.gettext or do other versions also benefit from it?
20150412 09:13:37< Ivanovic> what is the stuff about the boost filesystem thing?
20150412 09:13:52< Ivanovic> are there versions that benefit from turning it off or can this be on all over the board now?
20150412 09:14:05< Ivanovic> (remember: forget thinking about the pandora version, this is no longer relevant for master!)
20150412 09:14:58< iceiceice> so gfgtdf is not here, and he knows the most about both of them
20150412 09:15:24< iceiceice> his opinion, at the time of that forum post, was that the old filesystem code is a maintanence burden,
20150412 09:15:31< iceiceice> he didnt think that the libintl code was much of a burden
20150412 09:15:33< Ivanovic> if there are sane reasons to have flags, i vote for having specific flags that handle only one thing
20150412 09:16:14< Ivanovic> no more "roll all things into the pandora flag"
20150412 09:16:28< iceiceice> yes so on master we more or less have that now
20150412 09:16:34< iceiceice> in scons
20150412 09:16:39< Ivanovic> yes, i think i heard some comments like "If you want to have the old and functional stuff, compile with the pandora flag"
20150412 09:16:50< Ivanovic> iceiceice: assume that i have no clue what *IS* in there
20150412 09:17:39< iceiceice> on master right now, you must use boost filesystem
20150412 09:17:47< iceiceice> you may select libintl or boost locale
20150412 09:17:54< iceiceice> if you are on windows, it complains if you dont pick locale
20150412 09:18:02< iceiceice> otherwise all of this stuff has been solved i believe
20150412 09:18:09< iceiceice> we could nuke that option also i guess,
20150412 09:18:12-!- Appleman1234 [~Appleman1@CPE-121-208-6-89.bqzl1.cha.bigpond.net.au] has quit [Ping timeout: 276 seconds]
20150412 09:18:24< iceiceice> but since gfgtdf spoke in favor of keeping the old code i dont really see a reason
20150412 09:18:29< iceiceice> i dont think its confusing anymore
20150412 09:18:55< iceiceice> it was pretty confusing before techtonik made that pr and then i made this commit, https://github.com/wesnoth/wesnoth/pull/351
20150412 09:19:07< iceiceice> because at that time boost file system => boost locale
20150412 09:19:31< iceiceice> for no really good reason except that the only people who didnt want file system also happened to not want locale
20150412 09:19:55< iceiceice> but that kind of stuff is gone now, i think all the options are now exactly what they say on the tin
20150412 09:20:28< Ivanovic> so let me summarize what I understand:
20150412 09:21:04< Ivanovic> 1) there is no more option in cmake to use the old filesystem code, in scons it can still be selected but should probably be removed since it might be too much of a hazzle to maintain the functionality
20150412 09:21:36< iceiceice> in scons the old fs code cannot be selected
20150412 09:21:48< iceiceice> that pr (351) was what eliminated it
20150412 09:22:05< Ivanovic> 2) in cmake there is currently no way to use the libintl codepath instead of boost.gettext while this is possible in scons, here some systems (e.g. android) require the old implementation to work and there are some issues with the OSX build, too, if boost.gettext is used
20150412 09:22:06< iceiceice> that pr was not ported to cmake, until possibly shadowm did it just now
20150412 09:22:22< Ivanovic> okay, so '1' is already fixed
20150412 09:22:34< iceiceice> i think (2) is not solved right now
20150412 09:22:41< iceiceice> PR 351 basically did (1) and (2) in one shot on scons
20150412 09:22:47< Ivanovic> and '2' means adding in a special flag for deselecting boost.gettext and switching to libintl instead
20150412 09:22:51< iceiceice> and i didnt do either on cmake
20150412 09:22:57< shadowm> Yes, I'll have to look into doing (2) as I previously said above.
20150412 09:23:01-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 09:23:04< iceiceice> ok,  think we are all on the same page
20150412 09:23:20< Ivanovic> great, this solves a lot of confusion on my side
20150412 09:23:26< Ivanovic> thanks for your patience with me
20150412 09:23:41< shadowm> (Incidentally, my opinion re OS X and master? Bump version requirements if necessary, I don't care if it breaks PPC or an OS release from 2007. It's 2015, and if you need to replace your hardware to get a newer version, well, you bought an Apple product, you should've known this would happen. :p I'm overprotective of our requirements on 1.12 solely because it's a stable branch and breaking that kind of promise in stable without an ...
20150412 09:23:47< shadowm> ... irrefutable excuse is poor project management.)
20150412 09:24:13< irker523> wesnoth: Charles Dang wesnoth:master 3378e0696682 / data/campaigns/Northern_Rebirth/ (8 files in 2 dirs): NR: readded uses of SUPPORTER_DEATH_HANDLER http://git.io/vvIMq
20150412 09:24:27< Ivanovic> shadowm: also sounds like a decent solution to me
20150412 09:26:45-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20150412 09:29:16-!- mjs-de [~mjs-de@f049196090.adsl.alicedsl.de] has joined #wesnoth-dev
20150412 09:29:40-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150412 09:30:17-!- noy [~Noy@wesnoth/developer/noy] has quit [Client Quit]
20150412 09:32:48< shadowm> zookeeper: I like loyalists.ogg. :(
20150412 09:34:45< zookeeper> shadowm, we can keep it just for you
20150412 09:35:00< iceiceice> i also like loyalist.ogg 
20150412 09:35:12< iceiceice> tell him to make a drake anthem, we dont have one i think?
20150412 09:35:29< iceiceice> i guess he could make a khalifate sound track
20150412 09:35:33< shadowm> We had a really loud synth track for them back in 1.0.
20150412 09:35:38< iceiceice> really?
20150412 09:35:48< shadowm> At least, it was assigned to the drakes faction in MP games.
20150412 09:36:16< shadowm> Maybe that's the actual reason people used to hate them so much.
20150412 09:37:58< vultraz> bumba showed me some old synth tracks from the old NR
20150412 09:38:03< vultraz> if that's what it sounded like
20150412 09:38:06< vultraz> I can see why :P
20150412 09:38:19< zookeeper> yeah the old drake track was quite terrible
20150412 09:38:23< shadowm> It was objectively worse.
20150412 09:38:57< shadowm> And yes, I did play NR back when it had the synth music loops and the peculiar prose.
20150412 09:39:10-!- ancientcc [~ancientcc@140.207.223.191] has quit [Ping timeout: 264 seconds]
20150412 09:39:23< zookeeper> oh the "Come on guyz!" prose?
20150412 09:39:35< shadowm> That, yes.
20150412 09:40:06-!- ancientcc [~ancientcc@123.157.71.143] has joined #wesnoth-dev
20150412 09:40:29< zookeeper> well, i've still never played it past the second scenario
20150412 09:41:43< iceiceice> y me either :)
20150412 09:41:59< iceiceice> i think i cheated and looked at a walkthrough for the second scenario
20150412 09:41:59-!- rayblade53 [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20150412 09:42:05< iceiceice> but then i didnt continue with the campaign
20150412 09:42:10< iceiceice> after i saw what the walkthrough entailed :p
20150412 09:42:16-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Ping timeout: 240 seconds]
20150412 09:42:17< shadowm> It gets better.
20150412 09:42:17-!- rayblade53 is now known as vultraz
20150412 09:43:02< shadowm> And by better, I mean 30-min turns on the last scenario. (Okay, but seriously, the AI was significantly less efficient back in the day.)
20150412 09:45:42< vultraz> iceiceice: perhaps you would consider playing it again now that I'm working on it
20150412 09:46:07< zookeeper> iceiceice, and did you know that originally, each MP faction had its own music track, and the music would change on every side turn depending on which faction's turn it was? :P
20150412 09:46:17< iceiceice> wow thats brilliant :p
20150412 09:46:29< iceiceice> yeah i guess i would try again
20150412 09:46:34< iceiceice> i didnt try to play a wesnoth campaign in a while
20150412 09:47:03< shadowm> I misremembered. The drake track wasn't the loud one: https://raw.githubusercontent.com/wesnoth/wesnoth/1.0/music/wesnoth-6.ogg (drake track)
20150412 09:47:13< iceiceice> i think i won my last ladder match after a streak of like 10 losses... so i was planning to retire on that :p
20150412 09:47:34< shadowm> Here's the loud one: https://raw.githubusercontent.com/wesnoth/wesnoth/1.0/music/wesnoth-7.ogg (Knalgan Alliance track)
20150412 09:47:50< zookeeper> oh, that loud one
20150412 09:47:57< zookeeper> with the horrible clipping
20150412 09:48:06< iceiceice> haha
20150412 09:49:35< shadowm> Most of these tracks would turn out slightly slower and with lower pitch in-game because they were recorded at 48 KHz but we configure SDL_mixer (which can't do resampling reliably/at all?) to do the more standard 44.1 KHz instead.
20150412 09:50:01 * zookeeper blinks
20150412 09:50:02< zookeeper> really?
20150412 09:50:17< shadowm> Yes.
20150412 09:50:20< zookeeper> awesome
20150412 09:50:58< shadowm> So wesnoth-6.ogg was slightly more bearable in-game than it sounds on a proper player.
20150412 09:51:21< shadowm> That had no effect on the awful clipping of -7 and -8, of course.
20150412 09:51:42-!- ancientcc [~ancientcc@123.157.71.143] has quit [Ping timeout: 256 seconds]
20150412 09:55:05-!- Ravana_ [SZ_Bot@unaffiliated/ravana/x-2327071] has quit [Disconnected by services]
20150412 09:55:05-!- SZ_Bot [SZ_Bot@27-83-235-80.dyn.estpak.ee] has joined #wesnoth-dev
20150412 09:55:14-!- SZ_Bot is now known as Ravana_
20150412 09:55:15-!- Ravana_ [SZ_Bot@27-83-235-80.dyn.estpak.ee] has quit [Changing host]
20150412 09:55:15-!- Ravana_ [SZ_Bot@unaffiliated/ravana/x-2327071] has joined #wesnoth-dev
20150412 09:57:21< irker523> wesnoth: Nils Kneuper wesnoth:master 2d1e0804aa83 / CMakeLists.txt changelog: enforce cmake default build type (bug #23445) http://git.io/vvI5p
20150412 09:57:56< irker523> wesnoth: Nils Kneuper wesnoth:1.12 caf812dd56e0 / CMakeLists.txt changelog: enforce cmake default build type (bug #23445) http://git.io/vvIdf
20150412 10:03:48-!- Appleman1234 [~Appleman1@CPE-121-208-6-89.bqzl1.cha.bigpond.net.au] has joined #wesnoth-dev
20150412 10:04:56-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 10:09:08-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 245 seconds]
20150412 10:11:32< Ivanovic> shadowm: i just had a look at the FindFriBiDi.cmake file and i don't understand why it does not find cmake over here...
20150412 10:12:10-!- kex [~kex@78.157.29.160] has joined #wesnoth-dev
20150412 10:12:32< Ivanovic> since there is a define for fribidi_charset_to_unicode in fribidi-char-sets.h
20150412 10:14:12< Ivanovic> using fribidi-0.19.6 over here
20150412 10:14:49< Ivanovic> anyway, why are we still supporting the ancient fribidi?
20150412 10:15:00< Ivanovic> that is: could we significantly ease the overall checking of the file?
20150412 10:15:45-!- Kwandulin [~Miranda@p5B00971C.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20150412 10:16:36< iceiceice> Ivanovic: thanks for fixing bug 23445
20150412 10:17:20-!- kex [~kex@78.157.29.160] has quit [Ping timeout: 276 seconds]
20150412 10:18:22< Ivanovic> shadowm: is there a reason why you check for CHECK_SYMBOL_EXISTS and not CHECK_FUNCTION_EXISTS ?
20150412 10:18:30< Ivanovic> since the latter does find fribidi over here
20150412 10:24:09-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 256 seconds]
20150412 10:34:08-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20150412 10:37:16< Ivanovic> shadowm: okay, things seem to be very off
20150412 10:37:25< Ivanovic> shadowm: where does this compile test in the file come from?
20150412 10:37:32< Ivanovic> and how is it actually supposed to work?
20150412 10:39:12-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150412 10:43:35< Ivanovic> that is: if i manually add all the compile requirements i can make it compile, sadly it will not work with the "autosetup" in this cmake line
20150412 10:47:29-!- vultraz [~chatzilla@124.109.10.167] has quit [Read error: Connection reset by peer]
20150412 10:48:28-!- vultraz [~chatzilla@124.109.10.167] has joined #wesnoth-dev
20150412 10:48:36-!- ancientcc_ [~ancientcc@123.155.153.55] has joined #wesnoth-dev
20150412 10:48:37-!- vultraz [~chatzilla@124.109.10.167] has quit [Changing host]
20150412 10:48:37-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20150412 10:51:03-!- ancientcc [~ancientcc@140.207.223.191] has quit [Ping timeout: 250 seconds]
20150412 10:52:43< Ivanovic> that is i am currently looking at removing the support for fribidi <0.10.5
20150412 10:52:56-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has quit [Read error: Connection reset by peer]
20150412 10:53:26-!- vultraz [~chatzilla@wesnoth/developer/vultraz] has joined #wesnoth-dev
20150412 10:55:25< Ivanovic> okay, if removing all the other stuff so that i can actually build again using fribidi it gives me a warning that some of the functions we use are deprecated in fribidi 0.19.6
20150412 10:55:35< Ivanovic> /src/wesnoth-master/src/font.cpp: In member function 'void font::{anonymous}::text_surface::bidi_cvt()':
20150412 10:55:36< Ivanovic> /src/wesnoth-master/src/font.cpp:631:2: error: 'FriBidiLevel fribidi_log2vis(const FriBidiChar*, FriBidiStrIndex, FriBidiParType*, FriBidiChar*, FriBidiStrIndex*, FriBidiStrIndex*, FriBidiLevel*)' is deprecated (declared at /usr/include/fribidi/fribidi-deprecated.h:198) [-Werror=deprecated-declarations]
20150412 10:55:38< Ivanovic>   fribidi_log2vis(bidi_logical, n, &base_dir, bidi_visual, NULL, NULL, NULL);
20150412 10:57:15< Ivanovic> so my position on fribidi right now is: screw fribidi!
20150412 10:57:16< Ivanovic> ;)
20150412 11:02:50< Ivanovic> shadowm: for me, so that the compile test of the fribidi mini program actually works, i need to set the following things on commandline and i guess cmake is not setting them all, so the check is bulls*** anyway...
20150412 11:02:53< Ivanovic> gcc fribidi_test.c -isystem /usr/include/fribidi -isystem /usr/include/glib-2.0 -isystem /usr/lib64/glib-2.0/include
20150412 11:03:23< Ivanovic> plus i need to first switch to using CHECK_FUNCTION_EXISTS instead of CHECK_SYMBOL_EXISTS
20150412 11:08:42< Ivanovic> shadowm: in fact scons was changed to only allow fribidi >=0.10.9, so some of the check code can probably be altered
20150412 11:09:16< Ivanovic> yet I don't know the differences between the version but I guess the old fribidi support can be removed
20150412 11:10:42< irker523> wesnoth: Charles Dang wesnoth:master c82438854601 / changelog players_changelog: Fixed a bunch of typos http://git.io/vvLtM
20150412 11:12:10-!- Kwandulin [~Miranda@p5B00971C.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150412 11:22:50-!- markus_ [~mjs-de@f048110082.adsl.alicedsl.de] has joined #wesnoth-dev
20150412 11:22:59-!- mjs-de [~mjs-de@f049196090.adsl.alicedsl.de] has quit [Ping timeout: 276 seconds]
20150412 11:24:10< irker523> wesnoth: Nils Kneuper wesnoth:master 05d6e7dd9bd0 / CMakeLists.txt RELEASE_NOTES changelog cmake/FindFriBiDi.cmake src/font.cpp: remove option to use ancient FriBiDi http://git.io/vvLOU
20150412 11:39:23-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has joined #wesnoth-dev
20150412 11:48:49< vultraz> "The boost filesystem & locale dep may be disabled, if compiling with scons, by passing it an argument." <- didn't we get rid of this?
20150412 11:49:53-!- ancientcc_ [~ancientcc@123.155.153.55] has quit [Quit: Leaving]
20150412 11:50:26-!- anoob [~anoob@187-7-51-153.jvece702.dsl.brasiltelecom.net.br] has joined #wesnoth-dev
20150412 11:50:36-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 11:56:51< irker523> wesnoth: Nils Kneuper wesnoth:master df09cda92e08 / cmake/FindFriBiDi.cmake: format cleanup for FriBiDi finding in cmake http://git.io/vvLWF
20150412 11:59:00-!- ancientcc [~ancientcc@140.207.223.191] has quit [Remote host closed the connection]
20150412 11:59:43< irker523> wesnoth: Charles Dang wesnoth:master 640922c5aa4c / RELEASE_NOTES: Formatting and typo fixup http://git.io/vvLlw
20150412 12:03:43-!- new_one [~new_one@128.42.95.49] has joined #wesnoth-dev
20150412 12:06:08-!- bumbadadabum [~bumbadada@d155109.upc-d.chello.nl] has joined #wesnoth-dev
20150412 12:06:26< bumbadadabum> Hope I'm not late to the party
20150412 12:07:02< irker523> wesnoth: Bär Halberkamp wesnoth:master 98c9bc4e5ae3 / changelog players_changelog: updated changelog http://git.io/vvL46
20150412 12:10:28-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 12:10:35-!- ancientcc [~ancientcc@211.162.34.205] has quit [Client Quit]
20150412 12:11:14-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 12:12:05-!- ancientcc [~ancientcc@211.162.34.205] has quit [Client Quit]
20150412 12:26:07-!- kingdom_ [~kingdom@140.207.223.191] has joined #wesnoth-dev
20150412 12:26:28-!- Kwandulin [~Miranda@p5B00971C.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20150412 12:27:13-!- kingdom_ [~kingdom@140.207.223.191] has quit [Remote host closed the connection]
20150412 12:29:20-!- kingdom_ [~kingdom@140.207.223.191] has joined #wesnoth-dev
20150412 12:30:01-!- kingdom_ [~kingdom@140.207.223.191] has quit [Read error: Connection reset by peer]
20150412 12:30:27-!- kingdom_ [~kingdom@140.207.223.191] has joined #wesnoth-dev
20150412 12:33:00-!- kingdom_ [~kingdom@140.207.223.191] has quit [Remote host closed the connection]
20150412 12:41:08-!- markus_ is now known as mjs-de
20150412 12:46:21-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 12:47:01-!- ancientcc [~ancientcc@211.162.34.205] has quit [Client Quit]
20150412 12:47:34-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 12:47:52-!- ancientcc [~ancientcc@211.162.34.205] has quit [Remote host closed the connection]
20150412 12:54:14-!- ancientcc [~ancientcc@211.162.34.205] has joined #wesnoth-dev
20150412 12:55:27-!- ancientcc [~ancientcc@211.162.34.205] has quit [Read error: Connection reset by peer]
20150412 13:07:39-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 13:08:53-!- ancientcc [~ancientcc@140.207.223.191] has quit [Client Quit]
20150412 13:10:01-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 13:12:37< irker523> wesnoth: Nils Kneuper wesnoth:master 76c6cc1de3f0 / CMakeLists.txt changelog cmake/FindFriBiDi.cmake src/CMakeLists.txt: switch to better way finding FriBiDi in cmake http://git.io/vvLXb
20150412 13:16:49-!- Kwandulin [~Miranda@p5B00971C.dip0.t-ipconnect.de] has joined #wesnoth-dev
20150412 13:17:34< Ivanovic> shadowm: okay, for 1.13.x fribidi should now be detected and working for cmake
20150412 13:17:46< Ivanovic> shadowm: the only question is if this should be backported to 1.12
20150412 13:18:15< Ivanovic> i'll leave this one to you, downside is that it adds a dependency to a more recent fribidi (before any version would have been accepted)
20150412 13:18:16-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Ping timeout: 240 seconds]
20150412 13:25:02< Ivanovic> shadowm: you mentioned you redid the INSTALL file, i just checked and there is no mentioning of FriBiDi in there, is this intentional?
20150412 13:27:34< Ivanovic> bumbadadabum: btw i asked for your cloak in #freenode, but it looks as if the freenode staff is not around today...
20150412 13:27:51< bumbadadabum> ah
20150412 13:27:52< Ivanovic> at least they were not around at 10am and now more than 5h it seems as if still nobody is around
20150412 13:28:18< bumbadadabum> maybe freenode staff alsp gets sunday off :p
20150412 13:28:24< bumbadadabum> *also
20150412 13:47:23-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20150412 13:50:49-!- trewe [~trewe@2001:8a0:d11d:e301:626c:66ff:fe92:9b7c] has joined #wesnoth-dev
20150412 14:29:19-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Quit: Konversation terminated!]
20150412 14:30:31-!- ancientcc [~ancientcc@140.207.223.191] has quit [Quit: Leaving]
20150412 14:32:47-!- ancientcc [~ancientcc@112.65.211.213] has joined #wesnoth-dev
20150412 14:38:15-!- ancientcc [~ancientcc@112.65.211.213] has quit [Quit: Leaving]
20150412 14:38:18-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev
20150412 14:38:56-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 14:39:08-!- ancientcc [~ancientcc@140.207.223.191] has quit [Read error: Connection reset by peer]
20150412 14:39:30-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 14:49:05-!- gfgtdf [~chatzilla@f054161041.adsl.alicedsl.de] has joined #wesnoth-dev
20150412 14:49:19-!- ancientcc [~ancientcc@140.207.223.191] has quit [Quit: Leaving]
20150412 14:50:38-!- ancientcc [~ancientcc@140.207.223.191] has joined #wesnoth-dev
20150412 14:50:56< gfgtdf> iceiceice: whats the reason for moving functions to lua like in https://github.com/wesnoth/wesnoth/commit/f3a639fc628412b97b1d2e31afd72f5395767e77 ?
20150412 14:51:41< iceiceice> so that they use the lua_kernel pointers and not resources::
20150412 14:52:02< gfgtdf> iceiceice: why do you not want to use resources ?
20150412 14:52:14< iceiceice> because if you copy the game state, you need to use different poitners
20150412 14:53:27< gfgtdf> iceiceice: why? i'd think that even with multiple game states there is only one currently running game state and resources would then point to that one.
20150412 14:54:32< iceiceice> suppose that you want to allow a simulation based ai
20150412 14:54:46< iceiceice> you are going to want to be able to have not just like one or two game states,
20150412 14:54:51< iceiceice> maybe a stack of them
20150412 14:55:12< iceiceice> you could make some kidn of "resources context manager" class that like switches all the resources poitners out
20150412 14:55:21< iceiceice> but its really hideous
20150412 14:55:27< iceiceice> and probably will cause segfaults
20150412 14:55:43< iceiceice> its conceptually much simpler for the implementation of wml to use the lua kernel pointers
20150412 14:56:17< iceiceice> those resources pointers to things like "unit_map" and "team" shouldnt' really exist probably
20150412 14:56:25< iceiceice> to "game_config_manager", ok
20150412 15:01:13< gfgtdf> iceiceice: do you know how bug the gamestate structure is?
20150412 15:01:36< gfgtdf> iceiceice: liek compared the to game config i mean.
20150412 15:01:53< iceiceice> i didnt measure it, no
20150412 15:02:03< gfgtdf> iceiceice: hm ok
20150412 15:02:57< gfgtdf> iceiceice: but in case of teh commit above the only resrouces tghat are used are resources::screen
20150412 15:03:55< iceiceice> yeah
20150412 15:04:07< iceiceice> i would like to delete the whole src/game_events/action_wml.cpp file though
20150412 15:04:14< iceiceice> and move all that to game_lua_kernel
20150412 15:05:43< gfgtdf> iceiceice: you bnow what teh difference is between game_display::get_singleton() and resrouces::screen ?
20150412 15:06:56< iceiceice> i think its pretty minimal
20150412 15:07:05< iceiceice> resources::screen is managed by play_controller object
20150412 15:07:21< iceiceice> so its only different i guess during the play_controller construction
20150412 15:07:35< gfgtdf> iceiceice: i actualy think the WML_HANDLER_FUNCTION macro is more easy to use than the lua way with static luaL_Reg const callbacks[] 
20150412 15:08:05< iceiceice> it's not really object oriented though
20150412 15:08:13< iceiceice> it defines these crappy static functions
20150412 15:08:19< iceiceice> that do not have anything in scope
20150412 15:08:43< iceiceice> i dont think the lua way is much different
20150412 15:08:46< gfgtdf> iceiceice: hm yes, but i dnt have to do the declaration/definition stuff.
20150412 15:09:05< iceiceice> i find the wml handler function confusing sometimes
20150412 15:09:13< iceiceice> because it doesnt tell me what types the arguments are
20150412 15:09:15< iceiceice> and i always have to remember
20150412 15:09:56< iceiceice> do you think there is some way to define member functions using a macro like that?
20150412 15:10:33< gfgtdf> iceiceice: hm no i dont think taht possible
20150412 15:10:49< gfgtdf> iceiceice: teh problme is taht c++ doenst allo you to initilize objects inisde a class
20150412 15:10:57< gfgtdf> iceiceice: you have to define them outsiside
20150412 15:11:17< iceiceice> i guess the alternative would be like, pass them a context object where they can get pointers from or something
20150412 15:11:53< iceiceice> idk i think i would rather to debug the simpler version where they are member functions
20150412 15:12:29< gfgtdf> iceiceice: hm y thats another point
20150412 15:12:58< gfgtdf> iceiceice: i dont remember actualyl when i had to debug those functions last time
20150412 15:13:07< iceiceice> idk if you see a way thats better than what i've been doing i would like to know
20150412 15:13:11< gfgtdf> iceiceice: so i cannot say whether that is more complicated
20150412 15:15:58-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has joined #wesnoth-dev
20150412 15:17:44< gfgtdf> iceiceice: i currently dont and i think it s limitation of c++, but im still unpahhy with it.
20150412 15:18:27< iceiceice> hmm
20150412 15:19:02< iceiceice> so another thing i thought was to use a secondary class that is only in the .cpp file
20150412 15:19:34< iceiceice> all those lua callbacks should be private anyways
20150412 15:20:03< iceiceice> so the .cpp file could have some structure like
20150412 15:20:38< iceiceice> class handlers { WML_HANDLER_FUNCTION( ... ) ... };
20150412 15:21:11< iceiceice> maybe the registration can be done using like boost pp cat macros
20150412 15:21:45< iceiceice> i think it could be done but idk if its worth it
20150412 15:22:29< gfgtdf> iceiceice: no that thing what isad warlier is teh you cannot use WML_HANDLER_FUNCTION  inside a class
20150412 15:22:54< iceiceice> you dont need to initialize anything in the class
20150412 15:23:00-!- Appleman1234 [~Appleman1@CPE-121-208-6-89.bqzl1.cha.bigpond.net.au] has quit [Ping timeout: 272 seconds]
20150412 15:23:00< iceiceice> it would be a different macro
20150412 15:23:18< iceiceice> in the class it would just define and declare the function
20150412 15:23:20< gfgtdf> iceiceice: WML_HANDLER_FUNCTION indint see how it coudl be done different
20150412 15:23:35< gfgtdf> iceiceice: and when youd it be added to lua ?
20150412 15:23:50< iceiceice> i think it would work kind of like the make enum thing
20150412 15:24:13< iceiceice> you would provide macros to define some tuple ( name, body ) ( name, body ) ...
20150412 15:24:27< iceiceice> then you have one macro that expands to give the member function declarations and definitions
20150412 15:24:30< iceiceice> and one that makes the reg stuff
20150412 15:26:53-!- [Relic] [~Relic]@2602:306:33a3:6d30:3d10:c61b:58b:6a26] has joined #wesnoth-dev
20150412 15:27:46< gfgtdf> iceiceice: hm ok i see
20150412 15:28:51< gfgtdf> iceiceice: hm no i dont think thats worth it
20150412 15:29:25< iceiceice> maybe its a little simpler if you use a registry table like it has now
20150412 15:29:35< iceiceice> and the reg stuff in the ctor refers to that
20150412 15:29:43< iceiceice> so its only one collection of macros
20150412 15:29:46< iceiceice> but its still complicated
20150412 15:30:10< iceiceice> oh isee what you mean, i guess thats probably not allowed
20150412 15:30:24< iceiceice> because of the initializing in class thing
20150412 15:39:24< gfgtdf> iceiceice: you know what this ise: https://github.com/wesnoth/wesnoth/blob/1.12/src/play_controller.cpp#L821
20150412 15:41:03< iceiceice> no i dont
20150412 15:48:22< gfgtdf> iceiceice: you have an opinion about https://gna.org/bugs/?22974 ? I think we sure need some type of modification that cannot be disabled, but i wonder whether we should not instead of adding memeber hidden= make it a new tag like resource]
20150412 15:53:30< iceiceice> yeah i think there should be some kind of [resource] tag
20150412 15:53:47< iceiceice> basically a macro that does not expand until scenario is loaded
20150412 15:54:26< iceiceice> idk about avoiding to write {MY_COOL_EVENTS} in every scenario
20150412 15:54:37< iceiceice> he could just use a macro to define his scenarios :p
20150412 15:54:59< gfgtdf> iceiceice: still it will bloat teh wml config if you have teh same evet in every scenario
20150412 15:55:55< iceiceice> if its like [resource] id = MY_COOL_EVENTS [content] [event]...[/event] [event]...[/event] [/content] [/resource]
20150412 15:56:28< iceiceice> and then you put like [load_resource] id = MY_COOL_EVENTS [/load_resource] in each scenario
20150412 15:56:39< iceiceice> and it doesnt get expanded until scenario is loaded
20150412 15:56:43< iceiceice> then i guess there's no bloat
20150412 15:56:57< iceiceice> b/c its only in the game config once
20150412 15:57:55< gfgtdf> iceiceice: do you mean leik having lado_resource liek a action wml tag ?
20150412 15:58:24< iceiceice> i guess
20150412 15:58:43< iceiceice> it coudl be a little differnet maybe
20150412 15:58:54< iceiceice> like maybe before preload events,
20150412 15:59:05< iceiceice> we crawl the entire scenario config and expand all the load_resource
20150412 15:59:07< iceiceice> but not before that
20150412 15:59:59< iceiceice> even at the end of mp connect maybe
20150412 16:00:07< gfgtdf> iceiceice: hm but tehn id do it in teh same code that laods teh [modification] wml into the scenario
20150412 16:00:09< gfgtdf> then*
20150412 16:00:25< iceiceice> yeah maybe
20150412 16:01:55< gfgtdf> iceiceice: hm ok that sounds quite trivial to implement.
20150412 16:02:45< iceiceice> y
20150412 16:02:51< iceiceice> i didnt try to do it before because
20150412 16:02:55< iceiceice> theres a lot of ways to do it
20150412 16:03:00< iceiceice> dont want to have like 6 ways to do it :p
20150412 16:03:12< iceiceice> but there should be a good alternative to macro inclusion
20150412 16:03:21< iceiceice> maybe this is the good way
20150412 16:03:53< iceiceice> i guess that they already can use wesnoth.do_file sort of to do this
20150412 16:04:16< gfgtdf> iceiceice: that drawback of this compared to teh solution suggested in eth bugreport is that with this resources cannot define [option]s
20150412 16:04:19-!- ancientcc [~ancientcc@140.207.223.191] has quit [Remote host closed the connection]
20150412 16:04:39< gfgtdf> iceiceice: do_fiel can only execute lua not wml
20150412 16:06:31< iceiceice> i think if you combined it with [insert_tag] you could get dynamically loaded wml
20150412 16:06:41< iceiceice> but ... i dont think anyone does that
20150412 16:07:40< gfgtdf> iceiceice: insert tg dont work there
20150412 16:07:48< gfgtdf> iceiceice: ah ok
20150412 16:07:59< gfgtdf> iceiceice: no, still
20150412 16:08:11< gfgtdf> iceiceice: youÄ'd need a way to load varaibles
20150412 16:08:48< iceiceice> i didnt really think it through
20150412 16:08:59< iceiceice> i dont usually use insert_tag hehe
20150412 16:12:58-!- irker523 [~irker@uruz.ai0867.net] has quit [Quit: transmission timeout]
20150412 16:13:51-!- kingdom_ [~kingdom@140.207.223.191] has joined #wesnoth-dev
20150412 16:16:40-!- kingdom_ [~kingdom@140.207.223.191] has quit [Remote host closed the connection]
20150412 16:22:54-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20150412 16:23:48-!- horrowin1 [~Icedove@2a02:810a:8b40:85c:21b:fcff:fee3:c3ff] has joined #wesnoth-dev
20150412 16:24:10-!- kingdom_ [~kingdom@211.162.34.205] has joined #wesnoth-dev
20150412 16:27:07-!- kingdom_ [~kingdom@211.162.34.205] has quit [Remote host closed the connection]
20150412 16:27:08< Ravana_> I would like to know if it is specifically forbidden to make self-modifying addons
20150412 16:31:41-!- Kwandulin [~Miranda@p5B00971C.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer]
20150412 16:33:01< gfgtdf> Ravana_: what you mean by self modifying ?
20150412 16:33:47< Ravana_> addon than can change its event source during gameplay
20150412 16:40:08< gfgtdf> Ravana_: h not that i am  aware of but why woudl you wna to do that ?
20150412 16:40:32< Ravana_> mostly to see if I can
20150412 16:43:00< Ravana_> way to keep track of which version of addon different people have without having to learn lua would also help
20150412 16:43:43< gfgtdf> iceiceice: i mkae a simple imeplemtatiogn of [resource]: https://github.com/wesnoth/wesnoth/pull/401
20150412 16:44:09< gfgtdf> iceiceice: but im not very convinced yet, esp ebcasue it has not flat ""require_resource" to know wheth erevery user needs to have it installed
20150412 16:44:21< gfgtdf> flag*
20150412 16:47:33< iceiceice> gfgtdf: i think maybe [resource] should be an SP feature also?
20150412 16:47:44< gfgtdf> iceiceice: it is
20150412 16:47:55< iceiceice> ok i misunderstood then
20150412 16:48:10< gfgtdf> iceiceice:  is for mp and for sp
20150412 16:48:24< gfgtdf> iceiceice: afaik [modification]s are too
20150412 16:50:18< iceiceice> so i think it is different from modification
20150412 16:50:20< iceiceice> in that the user cannot attach or detach resource
20150412 16:50:29< gfgtdf> iceiceice: y
20150412 16:50:43< iceiceice> so i did add a feature "require_scenario"
20150412 16:50:58< iceiceice> i didnt test it very well,
20150412 16:51:02< gfgtdf> iceiceice: what does it do ?
20150412 16:51:15< iceiceice> its similar to putting require_campaign in a campaign
20150412 16:51:38< iceiceice> well, i guess its a little different
20150412 16:51:43< iceiceice> the idea is that you should have the content on your machine already
20150412 16:51:46< gfgtdf> iceiceice: for [multiplayer] ?
20150412 16:51:56< iceiceice> yes
20150412 16:52:27< iceiceice> https://github.com/wesnoth/wesnoth/commit/7a4f025fabcb53ef816b1b1ba1a1eaa7fc0fef07
20150412 16:52:55< gfgtdf> iceiceice: is there currently a way for have require addon X with version > Y  without uning any of thse addosn content explicitly ?
20150412 16:53:21< gfgtdf> iceiceice: like an addon that adds unit but without using its era.
20150412 16:53:48< iceiceice> i dont think so
20150412 16:53:55< gfgtdf> iceiceice: for exampel if want all reople to have some lua pack version > 5 installed.
20150412 16:54:07< iceiceice> so
20150412 16:54:26< iceiceice> i have to look at the code again to be sure, maybe theres a trick
20150412 16:54:32< iceiceice> but right now i didnt intend that you could do that
20150412 16:54:50< iceiceice> fwiw i think its usually better to clone the content to your add-on
20150412 16:55:00< iceiceice> because people dont expect that anyone else will be depending on their content and they change it at random
20150412 16:55:13< iceiceice> in my experience :p
20150412 16:55:20< gfgtdf> iceiceice: hm ok
20150412 16:55:52< iceiceice> there is a way to make another add-on a dependency 
20150412 16:55:54< iceiceice> in the pbl file
20150412 16:56:04< iceiceice> but thats all i think
20150412 16:57:39< gfgtdf> iceiceice: ok
20150412 17:18:59-!- travis-ci [~travis-ci@ec2-54-147-218-195.compute-1.amazonaws.com] has joined #wesnoth-dev
20150412 17:18:59< travis-ci> gfgtdf/wesnoth-old#462 (wml_resources - d3652e1 : gfgtdf): The build passed.
20150412 17:18:59< travis-ci> Build details : http://travis-ci.org/gfgtdf/wesnoth-old/builds/58171024
20150412 17:18:59-!- travis-ci [~travis-ci@ec2-54-147-218-195.compute-1.amazonaws.com] has left #wesnoth-dev []
20150412 18:35:29-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20150412 18:47:36-!- Haudegen [~quassel@85.124.51.57] has quit [Ping timeout: 276 seconds]
20150412 19:03:25-!- Haudegen [~quassel@85.124.51.57] has joined #wesnoth-dev
20150412 19:21:34-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has quit [Quit: i go nstuf kthxbai]
20150412 20:18:21< shadowm> Ivanovic: No, I need to double-check INSTALL.
20150412 20:20:32-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150412 20:21:38< shadowm> Huh.
20150412 20:21:47< Ivanovic> attack of the clones?
20150412 20:21:48< shadowm> fribidi isn't mentioned in 1.12's INSTALL either.
20150412 20:22:53< Ivanovic> hehe, that is a nice one...
20150412 20:23:06< Ivanovic> so fix the build recipe or not?
20150412 20:23:24< Ivanovic> no idea on how many systems fribidi will actually work if the binaries were created with cmake
20150412 20:23:54< shadowm> It compiles with fribidi support on 1.12 for me, but I don't know if it actually works.
20150412 20:23:58< Ivanovic> (and i am surprised that there seems to be nobody on the scons side getting warnings with fribidi functions we used being marked deprecated)
20150412 20:24:15< Ivanovic> shadowm: just start the game, select hebrew and go into the ingame help
20150412 20:24:17< shadowm> What fribidi version do you have?
20150412 20:24:34< Ivanovic> if the items on the left side that are translated appear on the right, then it works
20150412 20:24:38< Ivanovic> 0.19.6
20150412 20:24:50< shadowm> Same as mine and I don't get warnings.
20150412 20:25:32< Ivanovic> with strict compilation turned on before i did my changes to the recipe?
20150412 20:25:50< shadowm> I always build with strict on.
20150412 20:26:05< Ivanovic> as mentioned, check how it displays the hebrew translation in the ingame help
20150412 20:26:11< shadowm> 1.12 help with hebrew selected: http://i.imgur.com/agy08Uz.png
20150412 20:26:30< Ivanovic> okay, seems to be working
20150412 20:26:46-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Ping timeout: 250 seconds]
20150412 20:27:01< Ivanovic> hmm, building with gcc?
20150412 20:27:10< shadowm> No, that's clang 3.4.
20150412 20:27:11< Ivanovic> maybe clang does not support the deprecation warning?
20150412 20:27:28< shadowm> I've built with gcc on a few occasions and didn't get any warnings either.
20150412 20:27:40< Ivanovic> do you have a file like this? /usr/include/fribidi/fribidi-deprecated.h
20150412 20:27:54< shadowm> Yes.
20150412 20:28:27< Ivanovic>      FRIBIDI_ENTRY FriBidiLevel fribidi_log2vis (
20150412 20:28:28< shadowm> (I also built the 1.12.1 and 1.12.2 tag test builds with gcc.)
20150412 20:28:44< Ivanovic> in our code we are using the function fribidi_log2vis
20150412 20:28:53< Ivanovic>      FRIBIDI_GNUC_WARN_UNUSED FRIBIDI_GNUC_DEPRECATED;
20150412 20:29:11< shadowm> I see that there.
20150412 20:30:15< Ivanovic> this does look to me as if the behavior on my system with throwing the warning is correct
20150412 20:30:26< Ivanovic> no idea why you are not seeing it though
20150412 20:30:49< Ivanovic> gcc-Version 4.9.2 (Gentoo 4.9.2 p1.2, pie-0.6.2) 
20150412 20:37:14< shadowm> g++ -o build/release/font.o -c -std=c++98 -pthread -W -Wall -O2 -D_GNU_SOURCE=1 -D_REENTRANT -DHAVE_LIBDBUS -DHAVE_CONFIG_H -D_X11 -DHAVE_FRIBIDI -Ibuild -Isrc -I/usr/include -I/usr/include/SDL -I/usr/include/pango-1.0 -I/usr/include/cairo -I/usr/include/glib-2.0 -I/usr/lib/x86_64-linux-gnu/glib-2.0/include -I/usr/include/pixman-1 -I/usr/include/freetype2 -I/usr/include/libpng12 -I/usr/include/dbus-1.0 ...
20150412 20:37:20< shadowm> ... -I/usr/lib/x86_64-linux-gnu/dbus-1.0/include -I/usr/include/fribidi src/font.cpp
20150412 20:37:26< shadowm> No warnings. :\
20150412 20:37:46< shadowm> 4.9shadowm@nanacore:~/src/wesnoth-release-clean/1.12.2% g++ --version
20150412 20:37:50< shadowm> g++ (Debian 4.9.2-10) 4.9.2
20150412 20:50:08< Ivanovic> http://pastebin.com/Wq5Pek2f
20150412 20:53:38-!- prkc [~prkc@catv-89-134-173-244.catv.broadband.hu] has joined #wesnoth-dev
20150412 20:54:15-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has quit [Ping timeout: 250 seconds]
20150412 20:56:15-!- shadowm_desktop [ignacio@wesnoth/developer/shadowm] has joined #wesnoth-dev
20150412 21:10:32< Ivanovic> shadowm: this required a tiny patch to cmakelists.txt: http://pastebin.com/W3fVZdwW
20150412 21:12:20< Ivanovic> and this cmake command in case it matters: http://pastebin.com/y8kfUG2N
20150412 21:17:44-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20150412 21:26:36-!- iceiceice [~chris@wesnoth/developer/iceiceice] has quit [Ping timeout: 252 seconds]
20150412 21:40:02-!- iceiceice [~chris@wesnoth/developer/iceiceice] has joined #wesnoth-dev
20150412 22:12:57-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection]
20150412 22:35:05-!- Appleman1234 [~Appleman1@CPE-121-208-6-89.bqzl1.cha.bigpond.net.au] has joined #wesnoth-dev
20150412 22:35:55-!- horrowin1 [~Icedove@2a02:810a:8b40:85c:21b:fcff:fee3:c3ff] has quit [Quit: horrowin1]
20150412 22:38:30< bumbadadabum> function wml_actions.modify_ai(cfg)
20150412 22:38:31< bumbadadabum> 	wesnoth.modify_ai_wml(cfg)
20150412 22:38:31< bumbadadabum> end
20150412 22:38:42< bumbadadabum> this seems to be a nil and causes my micro ais to crash
20150412 22:39:37< bumbadadabum> has anyone here recently changed that
20150412 22:39:55< bumbadadabum> or should I just bug mattsc about it?
20150412 22:40:46< iceiceice> damn
20150412 22:40:50< iceiceice> i changed it like a few hours ago
20150412 22:41:10< bumbadadabum> did you?
20150412 22:41:11< iceiceice> you should probably bug me about it
20150412 22:41:26< iceiceice> this was the commit:
20150412 22:41:27< iceiceice> https://github.com/wesnoth/wesnoth/commit/a5d3c64d10da5f34341e09256a14483dd6cfe9e1
20150412 22:41:57< bumbadadabum> I should recompile then
20150412 22:42:07< bumbadadabum> I might have done that maybe half an hour before the commit
20150412 22:42:33< iceiceice> man i dont see the bug
20150412 22:42:53< bumbadadabum> it's probably me
20150412 22:43:07< bumbadadabum> (it's 99% likely me)
20150412 22:43:09< bumbadadabum> (as usual)
20150412 22:43:11< iceiceice> oh i see :)
20150412 22:43:59< bumbadadabum> I was just so excited to finally have a use for the no brainer attack micro ai
20150412 22:44:58< bumbadadabum> since the AI doesn't seem to want to ignore my leader otherwise
20150412 22:45:22< bumbadadabum> either that or the ai keeps attacking it because it's the weakest unit on the side
20150412 22:54:32-!- un214 [~un214@2602:306:cdde:61a9:56a0:50ff:fe57:101d] has joined #wesnoth-dev
20150412 23:19:18-!- anoob [~anoob@187-7-51-153.jvece702.dsl.brasiltelecom.net.br] has quit [Quit: Leaving]
20150412 23:34:33-!- ancestral [~ancestral@174-20-187-184.mpls.qwest.net] has joined #wesnoth-dev
20150412 23:40:53-!- tomreyn [~tomreyn@megaglest/team/tomreyn] has quit [Remote host closed the connection]
20150412 23:59:23-!- mjs-de [~mjs-de@f048110082.adsl.alicedsl.de] has quit [Remote host closed the connection]
--- Log closed Mon Apr 13 00:00:27 2015