--- Log opened Tue May 25 00:00:31 2010
20100525 00:07:33< AI0867> I've removed vconfig::vconfig() on my machine, replacing all its uses, but I'm not sure that's the right solution...
20100525 00:07:51-!- Crab_ [~Crab_@wesnoth/developer/crab] has quit [Quit: Leaving.]
20100525 00:07:55< AI0867> the problem is that there are essentially two reasons to construct a vconfig with no arguments
20100525 00:08:13< AI0867> one is to create an empty vconfig, signaling an empty underlying config
20100525 00:08:53< AI0867> the other is because you don't want to actually construct one yet, and pass it as a non-const reference to be assigned to (operator=) later
20100525 00:09:40< AI0867> the former is rare, and can be done by calling vconfig(config(), true)
20100525 00:10:07< AI0867> the problem is that if the second is the default, it will keep getting used wrong...
20100525 00:11:37< AI0867> named constructors?
20100525 00:16:14-!- Blueblaze [~nick@adsl-99-182-52-253.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100525 00:17:52-!- _jbx_ [~jbailey@12.190.80.225] has quit [Quit: It was a new day yesterday, but it's an old day now.]
20100525 00:58:14-!- mjs-de [~mjs-de@vpw.wh.uni-dortmund.de] has quit [Remote host closed the connection]
20100525 00:59:25< CIA-86> eleazar * r42856 /trunk/data/core/ (terrain-graphics.cfg terrain.cfg): fixing a glitch that caused dry grass to display as leaf-litter instead. Also changing the default base of most villages that was 'Gg' to 'Gs' for better appearance with a variety of terrains.
20100525 01:01:19-!- Sirp__ [~97c1dc1b@wesnoth/developer/dave] has left #wesnoth-dev []
20100525 01:04:27< Elvish_Pillager> Does anyone here know whether [message] with side_for= not including the player whose turn it is works in MP yet?
20100525 01:07:32-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100525 01:13:34-!- phlaem [~a@e178081079.adsl.alicedsl.de] has quit [Quit: Leaving]
20100525 01:20:49-!- Netsplit *.net <-> *.split quits: billynux
20100525 01:20:54< AI0867> no idea
20100525 01:32:36-!- thespaceinvader_ [~chatzilla@cpc3-whit4-2-0-cust871.know.cable.virginmedia.com] has joined #wesnoth-dev
20100525 01:32:46-!- thespaceinvader_ [~chatzilla@cpc3-whit4-2-0-cust871.know.cable.virginmedia.com] has quit [Client Quit]
20100525 01:33:53-!- Bob_The_Mighty [~chatzilla@cpc4-brig15-0-0-cust904.3-3.cable.virginmedia.com] has joined #wesnoth-dev
20100525 01:36:15-!- thespaceinvader [~chatzilla@wesnoth/artist/thespaceinvader] has quit [Ping timeout: 265 seconds]
20100525 01:56:53-!- Skystriker [~croselius@pool-70-111-210-233.nwrk.east.verizon.net] has quit [Ping timeout: 240 seconds]
20100525 01:57:40-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has quit [Quit: Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz]
20100525 01:58:15-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection]
20100525 02:01:46-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has quit [Quit: Leaving]
20100525 02:06:53< CIA-86> eleazar * r42857 /trunk/data/campaigns/Heir_To_The_Throne/maps/ (25 files): cosmetic changes to most HttT maps to use new and improved terrains.
20100525 02:16:33< CIA-86> eleazar * r42858 /trunk/ (3 files in 3 dirs): created impassable snowy mountains.
20100525 02:18:04< shadowmaster> about time
20100525 02:18:44< shadowmaster> now I can use snowy mountains at will in my maps
20100525 02:28:38-!- Gambit is now known as Grickit
20100525 02:28:59-!- Grickit is now known as Gambit
20100525 02:42:01-!- Blueblaze [~nick@adsl-99-182-52-253.dsl.hstntx.sbcglobal.net] has quit [Ping timeout: 264 seconds]
20100525 02:47:41-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100525 02:55:11< CIA-86> ai0867 * r42859 /trunk/src/ (6 files in 2 dirs): 
20100525 02:55:11< CIA-86> Split vconfig no-arg constructor into unconstructed_vconfig() and
20100525 02:55:11< CIA-86> empty_vconfig(), the latter of which is backed by an empty config object,
20100525 02:55:11< CIA-86> rather than NULL.
20100525 02:55:11< CIA-86> This should eliminate segfaults due to accidental dereferencing of
20100525 02:55:12< CIA-86> the nullpointer in default-constructed vconfigs
20100525 03:05:35-!- Blueblaze [~irchon@166.133.232.179] has joined #wesnoth-dev
20100525 03:07:01-!- Blueblaze [~irchon@166.133.232.179] has quit [Remote host closed the connection]
20100525 03:07:05< CIA-86> espreon * r42860 /trunk/data/core/images/terrain/cloud-snow-tile.png: 
20100525 03:07:05< CIA-86> Ran wesnoth-optipng:
20100525 03:07:05< CIA-86> Overall statistics (only for files with a smaller recompressed size):
20100525 03:07:05< CIA-86> Original size: 15 KiB on 1 files
20100525 03:07:05< CIA-86> Optimized size: 12 KiB
20100525 03:07:06< CIA-86> Total saving: 2 KiB = 16% decrease
20100525 03:07:27-!- Blueblaze [~irchon@166.133.232.179] has joined #wesnoth-dev
20100525 03:09:44< CIA-86> eleazar * r42861 /trunk/data/campaigns/ (17 files in 2 dirs): cosmetic changes to UtBS and tSG maps in use new terrains.
20100525 03:12:39-!- nmmachado [~chatzilla@189.38.200.116] has joined #wesnoth-dev
20100525 03:28:46-!- k23z__ [k23z__@unaffiliated/k23z--/x-2536701] has joined #wesnoth-dev
20100525 03:32:26< nmmachado> Hi, Someone is a translator of the Portuguese language?
20100525 03:39:47-!- nmmachado [~chatzilla@189.38.200.116] has quit [Remote host closed the connection]
20100525 03:41:41-!- Blueblaze [~irchon@166.133.232.179] has quit [Quit: Blueblaze]
20100525 03:42:11-!- ancestral [~ancestral@mobile-166-137-142-227.mycingular.net] has joined #wesnoth-dev
20100525 03:46:25-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100525 04:09:52-!- ancestral [~ancestral@mobile-166-137-142-227.mycingular.net] has quit [Quit: Colloquy for iPhone - http://colloquy.mobi]
20100525 04:10:59-!- ancestral [~ancestral@97-116-112-18.mpls.qwest.net] has joined #wesnoth-dev
20100525 04:14:23-!- Blueblaze [~nick@adsl-99-182-52-253.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100525 04:15:52-!- un214 [~quassel@adsl-75-45-4-19.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20100525 04:16:32< un214> for some reason I decided to redo that work done a long time ago to have a multiplayer gender preference
20100525 04:21:33-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20100525 04:21:44-!- Bob_The_Mighty [~chatzilla@cpc4-brig15-0-0-cust904.3-3.cable.virginmedia.com] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539]]
20100525 04:21:56-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20100525 04:51:57< CIA-86> eleazar * r42862 /trunk/data/core/images/terrain/ (163 files in 2 dirs): moving some frozen terrain to a subfolder.
20100525 04:52:41< CIA-86> eleazar * r42863 /trunk/data/core/ (terrain-graphics.cfg terrain.cfg): moving some frozen terrain to a subfolder. part 2 - complete.
20100525 04:55:12-!- un214 [~quassel@adsl-75-45-4-19.dsl.scrm01.sbcglobal.net] has quit [Remote host closed the connection]
20100525 04:57:14-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20100525 05:00:29-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 240 seconds]
20100525 05:01:19-!- Ivanovic_ is now known as Ivanovic
20100525 05:07:29< CIA-86> eleazar * r42864 /trunk/data/core/ (271 files in 3 dirs): moved the unwalkable terrains into a subfolder.
20100525 05:10:34-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Ping timeout: 264 seconds]
20100525 05:12:35-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20100525 05:17:16< CIA-86> eleazar * r42865 /trunk/data/core/ (25 files in 4 dirs): moving various editor tiles to appropriate subfolders.
20100525 05:17:33-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Ping timeout: 276 seconds]
20100525 05:18:52-!- crimson_penguin [~ben@64.201.60.211] has joined #wesnoth-dev
20100525 05:18:58-!- crimson_penguin [~ben@64.201.60.211] has quit [Changing host]
20100525 05:18:58-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20100525 05:21:55-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100525 05:24:17-!- ancestral [~ancestral@97-116-112-18.mpls.qwest.net] has quit [Remote host closed the connection]
20100525 05:24:40-!- ancestral [~ancestral@97-116-112-18.mpls.qwest.net] has joined #wesnoth-dev
20100525 05:24:52-!- cib0 [~Miranda@p5DC432EE.dip.t-dialin.net] has quit [Ping timeout: 265 seconds]
20100525 05:25:01< CIA-86> eleazar * r42866 /trunk/data/core/ (24 files in 3 dirs): moved desert plants into subfolder
20100525 05:52:52-!- Christheturtle [~yeti-of-d@5e0efac7.bb.sky.com] has quit []
20100525 06:00:45-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20100525 06:06:55-!- silene [~plouf@bau91-1-82-239-244-109.fbx.proxad.net] has joined #wesnoth-dev
20100525 06:07:47-!- silene [~plouf@bau91-1-82-239-244-109.fbx.proxad.net] has quit [Changing host]
20100525 06:07:47-!- silene [~plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20100525 06:19:32-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100525 06:21:03-!- Gambit [~Gambit@pa-67-234-73-7.dhcp.embarqhsd.net] has quit [Quit: Leaving.]
20100525 06:43:46-!- timotei [~timo@wesnoth/developer/timotei] has joined #wesnoth-dev
20100525 06:58:02-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100525 07:25:06-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20100525 08:11:49-!- zookeeper [~l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20100525 08:23:13-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Quit: ...]
20100525 08:25:41-!- silene [~plouf@wesnoth/developer/silene] has quit [Quit: Leaving.]
20100525 08:28:20-!- Upthorn [ogmar@adsl-75-26-196-149.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 240 seconds]
20100525 08:28:23-!- Upthorn [ogmar@adsl-75-26-196-149.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20100525 09:00:59-!- ancestral [~ancestral@97-116-112-18.mpls.qwest.net] has quit [Quit: And that’s the end of THAT chapter.]
20100525 09:37:22< Ivanovic> moin
20100525 09:37:35< timotei> hello
20100525 09:40:22-!- euschn [~eugen@wesnoth/developer/euschn] has joined #wesnoth-dev
20100525 09:47:01-!- timotei [~timo@wesnoth/developer/timotei] has quit [Remote host closed the connection]
20100525 09:51:37-!- timotei [~timo@wesnoth/developer/timotei] has joined #wesnoth-dev
20100525 09:53:45< timotei> If I make a custom campaign, and put it in my home folder/wesnoth/data , should I put it in add-ons dir, or just in data?
20100525 09:55:21< timotei> fendrin: ^
20100525 09:55:53< fendrin> timotei: It is only searched for in add-ons dir
20100525 09:55:53< zookeeper> campaigns and other add-ons should always go in the add-ons dir
20100525 09:56:05< timotei> ok
20100525 09:56:06< timotei> thanks
20100525 10:18:37-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100525 10:27:01-!- k23z__ [k23z__@unaffiliated/k23z--/x-2536701] has quit [Ping timeout: 264 seconds]
20100525 10:54:16-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100525 10:56:53-!- timotei [~timo@wesnoth/developer/timotei] has quit [Read error: Operation timed out]
20100525 11:14:54-!- stikonas [~and@bcm-131-111-247-104.girton.cam.ac.uk] has joined #wesnoth-dev
20100525 11:14:54-!- stikonas [~and@bcm-131-111-247-104.girton.cam.ac.uk] has quit [Changing host]
20100525 11:14:54-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100525 11:15:57-!- loonybot [~loonybot@79.139.137.200] has joined #wesnoth-dev
20100525 11:15:57-!- loonybot [~loonybot@79.139.137.200] has quit [Changing host]
20100525 11:15:57-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-dev
20100525 11:16:32-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20100525 11:17:28-!- k23z__ [k23z__@unaffiliated/k23z--/x-2536701] has joined #wesnoth-dev
20100525 11:22:28-!- mjs-de [~mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20100525 11:27:53-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has quit [Quit: Hi! I'm a quit message virus vaccine. If you see a quit message virus, don't replace your quit message with it!]
20100525 11:29:16-!- thespaceinvader [~chatzilla@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20100525 11:38:53-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100525 11:44:18-!- Blueblaze [~nick@adsl-99-182-52-253.dsl.hstntx.sbcglobal.net] has quit [Ping timeout: 276 seconds]
20100525 11:54:17-!- zhoubug [~river@218.249.160.100] has joined #wesnoth-dev
20100525 11:54:25-!- zhoubug [~river@218.249.160.100] has left #wesnoth-dev []
20100525 12:03:18-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has joined #wesnoth-dev
20100525 12:06:38-!- k23z__ [k23z__@unaffiliated/k23z--/x-2536701] has quit [Ping timeout: 272 seconds]
20100525 12:30:24< Upthorn> hrm.
20100525 12:30:52< Upthorn> According to the spec I currently have up for persistence wml, I am going to need to modify event handlers to specify whether their scope is campaign or era
20100525 12:31:18< Upthorn> in order to allow for per-addon default namespace settings
20100525 12:31:42-!- Crab_ [~Crab@wesnoth/developer/crab] has joined #wesnoth-dev
20100525 12:32:22< Crab_> Upthorn: hi
20100525 12:32:36< Upthorn> hi crab
20100525 12:33:03< Upthorn> did you appear because of what I just said?
20100525 12:34:32< Upthorn> Is it a good idea or a bad idea to track whether each event handler comes from a campaign or an era?
20100525 12:35:46< Upthorn> I am not certain that it is a good idea, but I do not see what harm this would do, aside from increasing the size of the event_handler struct by one enum
20100525 12:36:10< Crab_> Upthorn: yes, you have summoned me :)
20100525 12:37:21< Crab_> Upthorn: I think that we can trick the preprocessor into handling the default namespace for us
20100525 12:38:18< Upthorn> That is the other solution I have been tossing around. I guess that is better because it is a one-time processing impact, rather than a continuous memory impact.
20100525 12:39:42< Crab_> Upthorn: http://wesnoth.pastebin.com/4zL3J6Lc
20100525 12:40:27< Crab_> so, during the time preprocessor includes all files in {scenarios}, DEFAULT_NAMESPACE expands to my_addon
20100525 12:41:49< Upthorn> that works to a certain extent, but what about [get_global_variable] from_global="my_variable_name" to_local="foo" [/get_global_variable]?
20100525 12:42:17< Crab_> well, that's uncorrect syntax, in that case :)
20100525 12:42:19< Upthorn> under the current spec, this should work so long as a default namespace is somehow defined
20100525 12:44:40< Upthorn> more to the point, [get_global_variable] namespace="^" from_global="my_variable_name" to_local="foo" [/get_global_variable], where global namespace is defined as "my_world.addon"
20100525 12:45:06< Upthorn> under the spec I wrote, this should get the global variable from "my_world"
20100525 12:45:50< Crab_> yes, I've read it.  ^{DEFAULT_NAMESPACE} will work for that, too
20100525 12:47:01< Upthorn> Would it be acceptable to have the preprocessor add a namespace={DEFAULT_NAMESPACE} to global variable events that lack the namespace attribute?
20100525 12:47:50< Crab_> it's not how the preprocessor works.. it has to replace something by something
20100525 12:48:10< Upthorn> it couldn't replace the entire [get_global_variable] tag?
20100525 12:49:17< Crab_> well, it can expand a macro like {GET_GLOBAL_VARIABLE_FROM_DEFAULT_NAMESPACE parameter1 parameter2}
20100525 12:49:25< Crab_> to the [get_global_variable] tag
20100525 12:49:47< Crab_> but it cannot post-process tags, that's handled later.
20100525 12:50:12< Crab_> note that, with lua, text can be reinterpreted as wml at runtime
20100525 12:51:03< Crab_> so, it's quite hard to reliably add 'origin' information to wml tags / wml events
20100525 12:51:34< Upthorn> Do you think post-processing tags is a good solution, or it is better just to provide perhaps the best solution is to provide a {GET_GLOBAL (namespace) from_global to_local (side)} macro
20100525 12:51:56< Upthorn> err
20100525 12:52:06< Upthorn> somehow two sentences collided there.
20100525 12:52:21< Crab_> post-processing tags is a buggy approach, because there are places where it's quite hard to postprocess tags
20100525 12:52:37< Crab_> for example, see http://wiki.wesnoth.org/LuaWML:Events#wesnoth.fire
20100525 12:52:43< Upthorn> Do you think post-processing tags is a good solution, or it is better just to provide a {GET_GLOBAL (namespace) from_global to_local (side)} macro (along with {SET_...} and {CLEAR_...})
20100525 12:54:11< Upthorn> I see, post-processing Lua in the way I describe would be very difficult.
20100525 12:54:31< Upthorn> Lua-generated WML, that is.
20100525 12:54:36< Crab_> yes
20100525 12:55:07< Upthorn> so I should go with a macro solution for now at least
20100525 12:55:08< Crab_> I don't like to force the 'use the big macroses' approach, too, I think that 'provide an example of how default namespace can be implemented using a small preprocessor macro' is better.
20100525 12:56:11< Upthorn> when you say "small preprocessor macro", do you mean that the macro name is small, or the macro function is small?
20100525 12:56:39< Crab_> 'only a small part of [get_global_variable] tag is a macro' http://wesnoth.pastebin.com/4zL3J6Lc
20100525 12:57:51< Upthorn> because SET_GLOBAL ({SET_GLOBAL (namespace) from_local to_global (side)}) has a small name, but a large function, and namespace={DEFAULT_NAMESPACE} has a large name but a small function.
20100525 12:58:35< Crab_> there's another potential solution, as well , using WML variables instead of preprocessor macroses
20100525 12:58:41< Upthorn> my tendency is to prefer solutions which requiresthe fewest repeated keystrokes of developers.
20100525 12:59:22< Crab_> e.g., WML variables like '$era_default_namespace' and '$scenario_default_namespace'
20100525 12:59:32< Crab_> what do you think about it ?
20100525 13:00:10< Crab_> it can be implemented at user-level (set_variable at prestart) or at engine level.
20100525 13:00:48< Upthorn> -- I would prefer to provide {SET_GLOBAL params}, because that way UMC developers don't have to type "namespace={DEFAULT_NAMESPACE}" out when they want the default namespace
20100525 13:01:18< Crab_> ( something like http://wesnoth.pastebin.com/DnW3rqmW )
20100525 13:01:58< Crab_> note that nothing prevents the  {SET_GLOBAL } macroses to be coded in every and each case
20100525 13:02:41< Upthorn> I guess I am making unnecessary work for myself.
20100525 13:03:11< Upthorn> I can add {SET_GLOBAL } etc macros as an optional task if there is time
20100525 13:03:25< Crab_> yes, event the users can code them for themselves, if they want
20100525 13:03:33< Crab_> as users can easily define their own macroses.
20100525 13:03:38< Upthorn> and otherwise the users can code them for themselves and there is no real problem
20100525 13:04:03< Crab_> the things that engine supports, those are important
20100525 13:04:20< Upthorn> yes. I understand now.
20100525 13:04:45< Crab_> you can help the users by suggesting a way to do that, yes
20100525 13:04:53< Upthorn> I need to learn to recognize my tendencies to over-engineer.
20100525 13:05:59< Crab_> for example, documentation like 'if you define the scenario_default_namespace attribute in [campaign], then the engine will automatically set a WML variable $scenario_default_namespace for you',
20100525 13:06:18< Crab_> or documentation like "the best way to have a default namespace for your campaign is to use a macro like that : "
20100525 13:06:40 * Crab_ often has those tendencies, too
20100525 13:12:38< Upthorn> yeah. I had some heated conversations with scummvm devs about whether or not it was necessary to pre-define pixel format-descriptor constructors for all common pixel formats
20100525 13:14:53< Upthorn> when all that was really needed was setA setR setG setB...
20100525 13:20:56< Crab_> happens sometimes. but, it's best to get basic things working before getting fancy.
20100525 13:21:33< Upthorn> yes. I just sometimes have too wide a definition of "basic" :)
20100525 13:22:44< Crab_> :)
20100525 13:34:44< Upthorn> Also, I apologize that I've been unable to do any significant work on the first day of coding, I had to have an emergency extraction of one of my wisdom teeth.
20100525 13:41:37-!- Ken_Oh [~briang@static-71-178-174-220.washdc.fios.verizon.net] has joined #wesnoth-dev
20100525 13:50:06-!- thespaceinvader [~chatzilla@wesnoth/artist/thespaceinvader] has quit [Ping timeout: 245 seconds]
20100525 13:53:01-!- esr [~chatzilla@static-71-162-243-5.phlapa.fios.verizon.net] has quit [Ping timeout: 245 seconds]
20100525 13:53:41< Crab_> Upthorn: you have done some work in advance while coding your SP prototype, so it's ok.
20100525 13:55:32< Crab_> Upthorn: but I would like to note that your timeline in the wiki is not yet updated to account for all the mandatory deliverables (we've talked about that)
20100525 13:55:58< Crab_> Upthorn: it's better to do that, to make sure that the timeline can be used to monitor the progress.
20100525 13:56:37< Crab_> as we need to 'reserve' time for all of those things in  http://wiki.wesnoth.org/index.php?title=SummerOfCodeProposal_JodyNorthup#Mandatory
20100525 13:56:41< Upthorn> I understand. I will be sure to get that done tomorrow/today.
20100525 13:56:43< Upthorn> yeah
20100525 13:57:25< Upthorn> which is, after I sleep but hopefully before next 0700UTC
20100525 14:00:13-!- Johannes13 [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20100525 14:00:28< Crab_> Upthorn: ok. btw, during what time do you prefer to be online ?
20100525 14:03:07< Upthorn> Hrm. Well I am about to go to bed (it is 0501 local time) so a 2-3 hours earlier or ~8 hours later than this time would be good
20100525 14:04:52< Upthorn> Ideally I'd like to be getting to bed at more like 300 local, so the best times for me would be between about 1900 UTC and about 1000 UTC
20100525 14:07:17< Upthorn> but I don't have any commitments aside from a gaming group all-day wednesdays (which will mostly be thursdays in your timezone) and possibly an AD&D group on parts of fridays (which will be entirely saturdays, your time)
20100525 14:07:58< Upthorn> Crab_: so if the suggested hours are not good, let me know ones which would work better.
20100525 14:10:41< Crab_> ok, so, something in 2:00 UTC - 5:00 UTC is ok ?
20100525 14:11:57< Crab_> and yes, AD&D is a good thing :)
20100525 14:11:58< Crab_> ( I am a DM for 10+ years, btw )
20100525 14:31:13< shadowmaster> kdevelop 4 is really better for working with wesnoth
20100525 14:31:26< shadowmaster> problem: it probably (haven't checked) consumes a lot of RAM.
20100525 14:31:58< shadowmaster> since I can already run out of RAM when compiling wesnoth with -j 1 when KDE is running...
20100525 14:36:58< Upthorn> yeah, 2:00 - 5:00 UTC works fine
20100525 14:37:18< Upthorn> I assume you mean morning, and not 14:00 - 17:00
20100525 14:37:19< Crab_> ok, thanks
20100525 14:37:29< Crab_> yes, that's morning
20100525 14:37:45< Crab_> (that is early morning for me)
20100525 14:47:00-!- thespaceinvader [~chatzilla@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20100525 15:02:00-!- esr [~chatzilla@static-71-162-243-5.phlapa.fios.verizon.net] has joined #wesnoth-dev
20100525 15:14:17-!- faryshta [~faryshta@189.146.40.14] has joined #wesnoth-dev
20100525 15:15:18-!- FAAB [~huajie@219.142.239.173] has joined #wesnoth-dev
20100525 15:16:52-!- cib0 [~Miranda@p5DD356CC.dip.t-dialin.net] has joined #wesnoth-dev
20100525 15:17:20< CIA-86> eleazar * r42867 /trunk/data/core/images/terrain/flat/ (42 files): actually adding the graphics for the dry grass i thought i had added weeks ago. Sorry about that.
20100525 15:18:10< FAAB> hi Crab_, I filed a new version of patch https://gna.org/patch/?1666 . Let me know if there are more changes to do in it.
20100525 15:18:48< Crab_> FAAB: ok
20100525 15:19:49-!- FAAB [~huajie@219.142.239.173] has quit [Client Quit]
20100525 15:22:56< Crab_> FAAB: the patch looks ok (haven't tried it yet), there's several minor issues which I'll fix when committing, if a test shows it works ok.
20100525 15:24:26< Crab_> FAAB: I expect that I'll test it in +15 h, will commit if it's ok.
20100525 15:32:22-!- Gambit [~Gambit@pa-67-234-73-7.dhcp.embarqhsd.net] has joined #wesnoth-dev
20100525 15:35:47-!- Gambit [~Gambit@pa-67-234-73-7.dhcp.embarqhsd.net] has quit [Client Quit]
20100525 15:37:14-!- faryshta [~faryshta@189.146.40.14] has quit [Quit: Saliendo]
20100525 16:03:18-!- Rhonda [~rhonda@wesnoth/developer/rhonda] has quit [Quit: rebooting]
20100525 16:03:32-!- Gallaecio [~Gallaecio@232.158.60.213.dynamic.mundo-r.com] has joined #wesnoth-dev
20100525 16:06:31-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20100525 16:09:25-!- _jbx_ [~jbailey@12.190.80.225] has joined #wesnoth-dev
20100525 16:09:32< CIA-86> ivanovic * r42868 /trunk/po/ (26 files in 26 dirs): updated Japanese translation
20100525 16:09:33< CIA-86> ivanovic * r42869 /branches/1.8/po/ (wesnoth-low/ja.po wesnoth-sof/ja.po wesnoth-sotbe/ja.po): updated Japanese translation
20100525 16:14:20-!- Rhonda [~rhonda@anguilla.debian.or.at] has joined #wesnoth-dev
20100525 16:22:05< Ivanovic> AI0867: has r42830 fixed bug #16020 ? https://gna.org/bugs/index.php?16020
20100525 16:32:44< CIA-86> ivanovic * r42870 /branches/1.8/ (changelog players_changelog): add (many!) missing entries in the changelogs, change the format to be as uniform, as possible...
20100525 16:32:49< Ivanovic> AI0867: if it does, please mark the report accordingly
20100525 16:33:14< Ivanovic> AI0867: if it does not, please update the 1.8 changelog, since i listed your fix from r42830 in there together with this bug report
20100525 16:34:12-!- dtiger [~dtiger@dynamic-vpdn-93-125-64-133.telecom.by] has joined #wesnoth-dev
20100525 16:38:51-!- euschn [~eugen@wesnoth/developer/euschn] has quit [Quit: Leaving.]
20100525 16:41:03-!- wesbot [~wesbot@wesnoth/bot/wesbot] has quit [Remote host closed the connection]
20100525 16:47:53-!- wesbot [~wesbot@wesnoth/bot/wesbot] has joined #wesnoth-dev
20100525 16:54:58-!- Gambit [~Gambit@pa-67-234-73-7.dhcp.embarqhsd.net] has joined #wesnoth-dev
20100525 17:02:11-!- k23z__ [k23z__@unaffiliated/k23z--/x-2536701] has joined #wesnoth-dev
20100525 17:32:00-!- Sapient [~patrickp@wesnoth/developer/sapient] has joined #wesnoth-dev
20100525 17:33:12-!- Sapient [~patrickp@wesnoth/developer/sapient] has left #wesnoth-dev []
20100525 17:38:46-!- cib0 [~Miranda@p5DD356CC.dip.t-dialin.net] has quit [Read error: Connection reset by peer]
20100525 17:50:04-!- FAAB [~huajie@219.142.239.173] has joined #wesnoth-dev
20100525 17:57:09-!- Johannes13_ [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20100525 18:00:29-!- Johannes13 [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 240 seconds]
20100525 18:04:27-!- Christheturtle [~yeti-of-d@5e0efac7.bb.sky.com] has joined #wesnoth-dev
20100525 18:06:49-!- Upthorn [ogmar@adsl-75-26-196-149.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 276 seconds]
20100525 18:06:53-!- Upthorn [ogmar@adsl-75-26-201-1.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20100525 18:07:00-!- Upth [~ogmar@adsl-75-26-196-149.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 272 seconds]
20100525 18:07:24-!- Upth [~ogmar@adsl-75-26-201-1.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20100525 18:08:46-!- Crab_ [~Crab@wesnoth/developer/crab] has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
20100525 18:37:06-!- FAAB [~huajie@219.142.239.173] has quit [Quit: Leaving.]
20100525 18:38:37-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100525 18:40:21-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100525 18:54:24-!- silene [~plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20100525 18:54:38< silene> hi
20100525 19:48:21-!- Blueblaze [~nick@adsl-99-182-52-253.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100525 19:54:10-!- timotei [~timo@wesnoth/developer/timotei] has joined #wesnoth-dev
20100525 19:56:55-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20100525 19:57:07< mordante> servus
20100525 19:59:40< timotei> hello mordante
20100525 19:59:46< mordante> hi timotei 
20100525 20:01:36-!- silene [~plouf@wesnoth/developer/silene] has quit [Quit: Leaving.]
20100525 20:06:14-!- Crab_ [~Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20100525 20:08:38-!- Christheturtle [~yeti-of-d@5e0efac7.bb.sky.com] has quit [Disconnected by services]
20100525 20:09:30-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100525 20:09:45-!- Christheturtle1 [~Christhet@5e0efac7.bb.sky.com] has joined #wesnoth-dev
20100525 20:16:40-!- k23z__ [k23z__@unaffiliated/k23z--/x-2536701] has quit [Quit: Leaving]
20100525 20:17:42-!- Christheturtle1 [~Christhet@5e0efac7.bb.sky.com] has left #wesnoth-dev []
20100525 20:19:04-!- Christheturtle [~Christhet@5e0efac7.bb.sky.com] has joined #wesnoth-dev
20100525 20:19:36-!- Bob_The_Mighty [~chatzilla@cpc4-brig15-0-0-cust904.3-3.cable.virginmedia.com] has joined #wesnoth-dev
20100525 20:35:23-!- Crab_ [~Crab_@wesnoth/developer/crab] has left #wesnoth-dev []
20100525 20:36:15-!- gabba [~gabba@wesnoth/developer/gabba] has joined #wesnoth-dev
20100525 20:36:35< gabba> bonjour
20100525 20:37:25< mordante> servus gabba 
20100525 20:37:41-!- silene [~plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20100525 20:40:28< gabba> hi mordante
20100525 20:41:06< timotei> fendrin: hey 
20100525 20:41:16< fendrin> hello timotei
20100525 20:41:51< timotei> fendrin: I'm starting to work with ant for the plugin.  how should I proceed with the depenedencies? I have 2 .jar-s
20100525 20:42:05< timotei> should I add them on svn? or require them from global Eclipse ClassPath
20100525 20:42:27< timotei> Crab or the others told me it's not good to add the dependencies in svn
20100525 20:42:54< silene> timotei: please no .jar in svn; and if you could remove the ones already there, it would be even better
20100525 20:43:25< timotei> silene: I don't think there are currently any jars afair
20100525 20:43:59< silene> timotei: there are 7 in utils/java
20100525 20:44:15< timotei> silene: :-S let me check
20100525 20:44:48< timotei> silene: that are a "must"
20100525 20:44:53< timotei> silene: that's the update site...
20100525 20:45:14< fendrin> timotei: silene is very right here, no libs in our svn that are external upstream
20100525 20:45:20< silene> so what? they are compiled from thing, right? so you should put the source in the svn, not the compiled files
20100525 20:45:35< timotei> silene: some versions of the already compiled plugin have to be hosted somewhere, so the users can install it in eclipse
20100525 20:46:27< silene> timotei: the svn is not a hosting place
20100525 20:46:46< timotei> silene: well, you're right
20100525 20:46:49< fendrin> timotei: write a script that builds them and puts them into the plugin install folder on the webserver. Only the source belongs into svn
20100525 21:07:22-!- phlaem [~a@e178088107.adsl.alicedsl.de] has joined #wesnoth-dev
20100525 21:14:48< timotei> wiki.wesnoth.org down?
20100525 21:15:54< Espreon> Not for me.
20100525 21:16:44< Christheturtle> Not that I know of.
20100525 21:19:36-!- elias [~elias@allegro/developer/allefant] has joined #wesnoth-dev
20100525 21:21:13-!- billynux [~billy@wesnoth/developer/billynux] has joined #wesnoth-dev
20100525 21:21:31< billynux> hi all / mordante 
20100525 21:23:54-!- Bob_The_Mighty [~chatzilla@cpc4-brig15-0-0-cust904.3-3.cable.virginmedia.com] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539]]
20100525 21:30:54< mordante> hi billynux 
20100525 21:31:03< mordante> billynux, did you read the log of yesterday?
20100525 21:31:19< billynux> nope, I'll get to it
20100525 21:31:47< mordante> it was only a one line reply to your question
20100525 21:33:09< timotei> Ivanovic: ping
20100525 21:33:42< billynux> yes, saw it now
20100525 21:33:43-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20100525 21:33:47< mordante> timotei, don't expect a pong soon, he's afk
20100525 21:33:58< timotei> mordante: np with that
20100525 21:34:07< mordante> billynux, ok
20100525 21:34:24< billynux> so, I fixed that double free today, and the fact that a second call to free would leak
20100525 21:34:42< billynux> mordante, but.. I do agree that an auto_ptr would be better
20100525 21:35:53-!- fendrin [~fabi@wesnoth/developer/fendrin] has quit [Remote host closed the connection]
20100525 21:37:32< mordante> good
20100525 21:38:10-!- fendrin [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20100525 21:41:56< silene> billynux: whenever possible, don't use auto_ptr, use move_ptr (aka unique_ptr), it is less buggy
20100525 21:42:53< mordante> silene, is unique_ptr already available somewhere?
20100525 21:43:09< billynux> mordante: :@ I'm back to that double free error (using auto_ptr). Check http://code.google.com/p/async-net-api/source/detail?r=34 vs. http://pastebin.com/bQHx8yDv
20100525 21:43:21-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has quit [Read error: Connection reset by peer]
20100525 21:43:22< billynux> ok silene, I don't know move_ptr ...
20100525 21:43:27< billynux> checking it out now
20100525 21:43:30< silene> mordante: yes, in gcc; but anyway, move_ptr and scoped_ptr have been available for years in boost
20100525 21:43:33< mordante> I thought it wasn't available yet, but I agree unique_ptr is better as auto_ptr
20100525 21:44:16-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has joined #wesnoth-dev
20100525 21:45:40< timotei> mordante: I have currently done something, and git is on "(no branch)". How do I switch back to HEAD?
20100525 21:45:43< timotei> git checkout HEAD?
20100525 21:45:56< silene> git checkout master
20100525 21:45:58< timotei> of course, maintaining my commits
20100525 21:46:22< billynux> silene, there is a disclaimer in the first page I find RE move_ptr that says it isn't boost, unique_ptr is though
20100525 21:46:25< timotei> ok, it works. thanks
20100525 21:46:25< mordante> the version in gcc should be an extension so probably not be available in all compilers, better use the boost version
20100525 21:46:28< timotei> thanks silene
20100525 21:47:04< mordante> IIRC the name in C++0x is unique_ptr
20100525 21:47:13< silene> billynux: in boost, it is called static_move_ptr and scoped_ptr (depending on which version you need)
20100525 21:48:56< mordante> ah never noticed that one in boost, didn't expect it in the container_ptr library but in the smart_ptr libarary
20100525 21:49:02< mordante> library*
20100525 21:52:06< billynux> but boost doesn't seem to provide much doc. for static_move_ptr
20100525 21:54:55< CIA-86> eleazar * r42871 /trunk/data/core/images/terrain/ (23 files in 2 dirs): putting hills in a subfolder, and adding a new dry hills terrain. Step 1.
20100525 21:57:47< CIA-86> timotei * r42872 /trunk/utils/java/eclipse_plugin/changelog: eclipse plugin: added a changelog to track plugin's changes
20100525 21:57:55< CIA-86> timotei * r42873 /trunk/utils/java/eclipse_plugin/.settings/ (org.eclipse.core.runtime.prefs org.eclipse.jdt.core.prefs): eclipse plugin: this don't need to be included
20100525 21:58:00< CIA-86> timotei * r42874 /trunk/utils/java/eclipse_plugin/ (5 files in 4 dirs): eclipse plugin: add support for and building for campaign projects
20100525 21:58:26< mordante> I'm off night
20100525 21:58:43< billynux> mordante, did you see that pastebin?
20100525 21:58:48< billynux> vs. the diff?
20100525 21:58:57< mordante> billynux, which one?
20100525 21:59:08< billynux> billynux> mordante: :@ I'm back to that double free error (using auto_ptr). Check http://code.google.com/p/async-net-api/source/detail?r=34 vs. http://pastebin.com/bQHx8yDv
20100525 21:59:41< billynux> Today I removed a double free error, but it came back with the use of auto_ptr
20100525 21:59:55< billynux> I'm gonna see if unique_ptr may fix this
20100525 22:00:04< mordante> no missed that one, will have a look
20100525 22:00:15< CIA-86> eleazar * r42875 /trunk/data/core/ (19 files in 2 dirs): adding new dry hills terrain. Step 2. Done -- for now, transistions could be made smarter.
20100525 22:01:15< mordante> billynux, how does your dtor look?
20100525 22:01:19-!- Ken_Oh [~briang@static-71-178-174-220.washdc.fios.verizon.net] has quit [Read error: Connection reset by peer]
20100525 22:02:05< billynux> mordante, http://code.google.com/p/async-net-api/source/browse/trunk/asio_server.cpp?spec=svn34&r=34#31
20100525 22:02:39< billynux> ... memory management is giving me headaches :(
20100525 22:03:49< mordante> I don't see a double free in the pastebin, but the usage of std::auto_ptr is wrong, acceptor_ should be the std::auto_ptr
20100525 22:04:09< mordante> std::auto_ptr deletes the pointer when it goes out of scope
20100525 22:04:55< mordante> also make sure you don't manually delete the pointer owned by std::auto_ptr
20100525 22:08:01< billynux> mordante, yes, acceptor_ is an auto_ptr (in the pastebin, where I get the double free, which I don't see either)
20100525 22:09:13< mordante> then why assign the new pointer first to a temporary std::auto_ptr?
20100525 22:10:43< billynux> I can't construct it in the constructor
20100525 22:11:05< mordante> I mean line 12 in the pastebin
20100525 22:11:09< billynux> mordante, doing acceptor_ = std::auto_ptr<tcp::acceptor>(new tcp::acceptor( ... doesn't help either
20100525 22:11:34< billynux> yes, I know
20100525 22:13:02< mordante> know what?
20100525 22:13:11< billynux> that you mean pastebin:12
20100525 22:13:20< mordante> oh ok
20100525 22:13:45< billynux> not using the temp. doesn't change it either -> acceptor_ = std::auto_ptr<tcp::acceptor>(new ...
20100525 22:16:20< mordante> you shouldn't assignment, but reset the pointer of acceptor_
20100525 22:16:20< silene> don't create an auto_ptr, just assign the raw ptr to acceptor_ (that being said, it doesn't explain your issue)
20100525 22:18:07< billynux> mordante / silene: no, I'm guessing my issue could be explained by io_services's destructor manually deleting my acceptor object, but it doesn't make sense
20100525 22:19:35< silene> billynux: i don't know it  that's your issue, bue don't forget that, as soon as an auto_ptr is copied to a temporary (whether willingly or not), the object it points to is destroyed
20100525 22:19:55< silene> if* but*
20100525 22:20:01-!- YogiHH [YogiHH@d009076.adsl.hansenet.de] has joined #wesnoth-dev
20100525 22:21:13< YogiHH> hello
20100525 22:21:16< billynux> ok, but tcp::acceptor* new_acceptor(new...  /*   and */ acceptor_.reset(new_acceptor); doesn't work either
20100525 22:21:21< mordante> hi YogiHH 
20100525 22:21:47< mordante> reset wants a raw pointer
20100525 22:22:05< billynux> he gets it
20100525 22:22:29-!- dtiger [~dtiger@dynamic-vpdn-93-125-64-133.telecom.by] has quit [Ping timeout: 240 seconds]
20100525 22:22:51< silene> billynux: use valgrind, it will tell you where the object is destroyed the first time
20100525 22:23:02< billynux> whats wrong with http://code.google.com/p/async-net-api/source/diff?spec=svn34&r=34&format=side&path=/trunk/asio_server.cpp ? -> I'm guessing there is a memleak in the destructor, but a delete there issues a double free ... now. How/why was it freed?
20100525 22:23:18< billynux> silene, I'm trying to, but it isn't helping much
20100525 22:23:28< YogiHH> mordante: Is there a simple way to background-colour the headers of a gui2 table?
20100525 22:23:33< billynux> (most likely I'm not reading its output well)
20100525 22:24:01< silene> billynux: it should give you a backtrace pointing to the first delete
20100525 22:24:44< silene> more precisely, two backtraces, the first one pointing to the second delete, and the second one pointing to the first delete
20100525 22:25:13< billynux> silene, what is the option for that? --leak-check=yes ?
20100525 22:25:35< mordante> YogiHH, define a new label with a background colour and use that one
20100525 22:25:52< silene> billynux: no option, that's the default behavior; what's its output?
20100525 22:26:01< billynux> 1' -> pastebin
20100525 22:26:13< YogiHH> mordante: i tried that, could it be that background colours don't work for labels?
20100525 22:26:41< mordante> YogiHH, can you paste your code?
20100525 22:27:17< YogiHH> mordante: yes, will take a moment
20100525 22:28:04< zookeeper> hi, YogiHH
20100525 22:28:14< billynux> silene, http://pastebin.com/VULrn1FQ
20100525 22:29:38-!- mjs-de [~mjs-de@vpw.wh.uni-dortmund.de] has quit [Remote host closed the connection]
20100525 22:29:40< silene> billynux: it means that your program didn't have any issue, not even a double free
20100525 22:29:45< zookeeper> YogiHH, have you looked into http://gna.org/bugs/?15950 yet?
20100525 22:30:04< YogiHH> zookeeper: nope, not yet
20100525 22:30:31< billynux> I guessed that, but how come just running it I get that double free error... it is very confusing... also.. I got that different runs mean different outputs/errors, which I attribute to threading
20100525 22:30:52< mordante> do you copy the server to the thread?
20100525 22:31:02< mordante> server object*
20100525 22:31:09< silene> billynux: unfortunately, it means that it's not a double free but rather a memory corruption, and you are unlucky because it depends on the actual memory layout, hence not reproducible in valgrind
20100525 22:31:11< billynux> let me see
20100525 22:31:19< Christheturtle> Does anyone here happen to be fluent in Python? I'm working on porting Wesnoth to Android.. And I might need assistance in the nearish future.
20100525 22:31:52< silene> Christheturtle: wesnoth doesn't use python; are you trying to port wesnoth to python?
20100525 22:32:17< Christheturtle> Yes, and then into Android. It's quite a bit task.
20100525 22:32:42< Greywhind> i know Python, but i don't know Wesnoth terribly well yet
20100525 22:32:46< billynux> mordante, well the thread is running the io object... which will call the handlers of the server (asio_server) object
20100525 22:33:07< silene> if you don't know python well enough, it will be just impossible for you to convert wesnoth to this language, sorry
20100525 22:33:23< Christheturtle> I'm fluent in Python.
20100525 22:33:26< Christheturtle> brb
20100525 22:33:38< billynux> (I think you confused chris with grey silene)
20100525 22:34:16< billynux> mordante, silene :  ... thinking if the thread may outlive the asio_server object and call a handler after it was destroyed ... (but that would give a segfault?)
20100525 22:35:09< silene> billynux: ? i'm just stating the obvious; converting 205k lines of C++ to python is just a daunting task
20100525 22:35:16< billynux> and... the asio_server destructor stops the io object... which should kill the thread... mmm... maybe do a t.join() there?
20100525 22:35:28 * billynux agrees
20100525 22:35:45< Christheturtle> (back for a moment) I know it is, its a mammoth task. Thats why I will need other Python programmers soon. But as I said, brb.
20100525 22:36:11< billynux> I *love* dynamic languages, but I don't find them very suitable for game dev. (yet)
20100525 22:36:38< mordante> I like them for small things and especially when it involves regexes
20100525 22:37:14< Christheturtle> (back for a moment again, i'm really going afk now) I'll most likely be using some Java too.
20100525 22:37:34< mordante> billynux, maybe leave the pointer dangling for now and add a FIXME
20100525 22:38:03< mordante> when you have test code working in the Wesnoth repro without extra dependencies I'll take a look
20100525 22:38:07-!- Christheturtle is now known as Christheturtle|A
20100525 22:38:13< billynux> well, rev. 34 "works" (but I believe there is a memleak in the destructor, if I put a delete there, it issues that double free err.)
20100525 22:38:34< billynux> ok... I don't want to spend too much time on it now
20100525 22:38:35-!- Christheturtle|A is now known as Christt|AFK
20100525 22:38:42< mordante> (or maybe look at the code later see what goes wrong, really about to sleep)
20100525 22:38:44< billynux> but... it bugs me! a lot
20100525 22:38:51< billynux> ok mordante, no prob.
20100525 22:39:07< billynux> I'd be damn
20100525 22:39:17< mordante> I agree, but maybe disable the delete with and add the comment and the solution might pop up later
20100525 22:39:21< billynux> with io_thread_.join() in the dest. everything seems ok
20100525 22:39:39< mordante> dest ?= destructor?
20100525 22:39:43< billynux> yes
20100525 22:39:45< mordante> ok
20100525 22:39:57< silene> billynux: then try replacing auto_ptr with scoped_ptr, just to be sure you are not leaking the pointer somewhere
20100525 22:40:06< billynux> so... I guess I was right, the thread outlived the server, corrupting memory
20100525 22:40:33< mordante> yes make sure threads are done before pulling their resources away ;-)
20100525 22:40:36< silene> forget what i just said, i'm not synchrnous
20100525 22:40:37< billynux> silene, but is not a scoped_ptr what I want (assuming semantics of scopeness)
20100525 22:40:53< billynux> :)
20100525 22:41:05< silene> billynux: it is, you want the pointer to be stuck to the class and not go anywhere else
20100525 22:41:22< billynux> I just wrongly assumed the thread would die graciously 
20100525 22:41:58< billynux> ah, different kind of scope then
20100525 22:42:04< mordante> no threads don't die graciously and if they do on Linux, start to bet what they do on Windows ;-)
20100525 22:42:11< billynux> scoped_lock uses { } scope
20100525 22:42:12< silene> but anyway, if the join works, it really was just a memory corruption not related to the pointer itself
20100525 22:42:24< zookeeper> YogiHH, do you think you'll have time before 1.8.2?
20100525 22:42:27< mordante> you need to signal them to stop and use join to make sure they terminate properly
20100525 22:42:35< billynux> silene,  so it appears
20100525 22:43:07< YogiHH> zookeeper: when is that planned?
20100525 22:43:15-!- Bocom_ [~Bocom@c-b7cfe255.013-31-6b736412.cust.bredbandsbolaget.se] has quit [Read error: Connection reset by peer]
20100525 22:43:29-!- Bocom_ [~Bocom@c-b7cfe255.013-31-6b736412.cust.bredbandsbolaget.se] has joined #wesnoth-dev
20100525 22:43:34< billynux> thanks mordante, in this scape, i stop the io object (which boost's doc. ensures me that makes the run method return ASAP, thus finishing the thread)
20100525 22:44:01< CIA-86> elias * r42876 /trunk/data/tools/wesnoth/wmlparser2.py: 
20100525 22:44:01< CIA-86> new python wmlparser:
20100525 22:44:01< CIA-86> - simplified the high level API
20100525 22:44:01< CIA-86> - added support for escaping " with "" within "quoted" strings
20100525 22:44:01< CIA-86> - added support for a,b,c=1,2,3 style assignments
20100525 22:44:02< CIA-86> - added support for <<lua code quoting>>
20100525 22:44:02< CIA-86> - added unit tests for all of the above
20100525 22:44:34< billynux> http://code.google.com/p/async-net-api/source/detail?r=35
20100525 22:44:51< billynux> s/scape/case/
20100525 22:45:09< mordante> ASAP doesn't mean the `main thread' does the delete before, in multi-threading you need to synchronize if you want to make assumptions
20100525 22:45:22< billynux> true
20100525 22:45:57< YogiHH> mordante: http://wesnoth.pastebin.com/bsh0yYMS
20100525 22:46:14< silene> YogiHH: see topic, in a few days
20100525 22:46:36< billynux> valgrind's output looks very good now, except: still reachable: 8 bytes in 1 blocks
20100525 22:46:36< YogiHH> ah, thanks
20100525 22:47:00< YogiHH> zookeeper: I see what i can do, but time is a bit tight atm.
20100525 22:47:10< mordante> billynux, use --show-reachable=yes to see what leaks
20100525 22:47:14< billynux> k
20100525 22:47:24< timotei> fendrin: do the campaigns need to have a certain folder name?
20100525 22:47:34< timotei> fendrin: or any "suggestive" enough name?
20100525 22:47:46< fendrin> no, there is no restriction or enforcement
20100525 22:47:52-!- wesbot changed the topic of #wesnoth-dev to: 1.8.2 planned on 31st May, stringfreeze for branches/1.8! | 113 bugs, 280 feature requests, 18 patches | logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100525 22:47:58< timotei> ok
20100525 22:48:46< mordante> YogiHH, you need to define a label which draws a filled rectangle on the background and the use definition="myfilledlabel" instead of default
20100525 22:49:17< YogiHH> mordante: oh, ok
20100525 22:50:52< zookeeper> YogiHH, all right. that's just pretty much the only serious campaign bug i can see...and it's rather mysterious so i'd expect it to break a lot of other things too, yet apparently it hasn't.
20100525 22:51:03< zookeeper> maybe i should just try to reproduce it myself...
20100525 22:51:10-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100525 22:51:24< CIA-86> eleazar * r42877 /trunk/data/core/terrain-graphics.cfg: fixed water transition glitch with dry hills.
20100525 22:53:01< mordante> really off now
20100525 22:53:15< billynux> see ya'
20100525 22:53:39-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20100525 22:56:36-!- Christt|AFK is now known as Christheturtle
20100525 22:56:38-!- silene [~plouf@wesnoth/developer/silene] has quit [Quit: Leaving.]
20100525 22:56:49< billynux> I'm off too
20100525 22:56:52-!- billynux [~billy@wesnoth/developer/billynux] has quit [Quit: Leaving]
20100525 22:56:57< boucman> hey all
20100525 22:57:04< Christheturtle> Ohai thar.
20100525 22:57:26-!- timotei [~timo@wesnoth/developer/timotei] has quit [Remote host closed the connection]
20100525 22:57:29< boucman> gabba: so, what's up ?
20100525 22:57:37< gabba> hey boucman
20100525 22:58:19< gabba> I'm finalizing the arrow framework design in my uml editor (BOUML)
20100525 22:58:56< boucman> ok, could you show me your current WIP so I can comment ?
20100525 22:59:01< gabba> I simplified it a lot
20100525 22:59:47< boucman> sounds good
20100525 22:59:56< gabba> Yes, I could either 1- send/push somewhere the bouml files, 2- take a screenshot (less informative I guess), 3- show you a preliminary code generation 
20100525 23:00:24< gabba> Getting the code generation to work wasn't as easy as I thought :-/
20100525 23:00:24-!- silene [~plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20100525 23:00:45< boucman> i don't have bouml (though it's probably an apt-get away) so screenshot is probably the simplest for tonight
20100525 23:01:26< gabba> ok, screenshot upcoming
20100525 23:01:38< boucman> I don't like code generation... it usually don't produce clean stuff unless the underlying UML is "perfect" so if you want to go that way, I would advise to generate once and never go back to the modeler afterward
20100525 23:01:51< boucman> UML is good for planning and design....
20100525 23:03:50< zookeeper> uh, i was just about to fill a bug report but then i searched for an existing one and found https://gna.org/bugs/?9158
20100525 23:04:04-!- Christheturtle is now known as Christt|AFK
20100525 23:04:08 * zookeeper considers that an annoying bug
20100525 23:04:39< zookeeper> i was trying to sift through the chat in a savefile and found it was completely cluttered with lobby messages
20100525 23:05:21< gabba> boucman: http://imagebin.org/98301
20100525 23:05:35< zookeeper> it's a BoL save on turn 608 with a size of 5mb and i bet at least 1-2mb of that is just lobby messages :P
20100525 23:06:35< gabba> boucman: yes, I'm just gonna use it to save copy-paste work, but once the code is integrated to wesnoth I don't think the UML tool will be useful
20100525 23:06:53< boucman> ok, sounds fine
20100525 23:07:25< boucman> gabba: first question while I read
20100525 23:07:28< gabba> boucman: right now I'm figuring out all the method parameter types
20100525 23:08:04< boucman> set_visible() what do you intend to use it for ? is it related to hoteseat play and only displaying for a given team ?
20100525 23:08:24< boucman> and don't (try to figure out the parameters) it's simpler to add them as the need comes
20100525 23:08:38< gabba> boucman: no, it's much more low-level, like when you show/hide a widget
20100525 23:09:36< boucman> ok... most things displayed don't have that idea of being set as invisible, my advice would be to remove the method a-priori, and add it back if you reach a point where you need it
20100525 23:09:40< gabba> boucman: the arrow starts hidden so you can set path and other parameters without any visual mess
20100525 23:10:14< boucman> the useal way to do that is simply not to cal display::draw, since the game is monothreaded
20100525 23:10:27< CIA-86> silene * r42878 /trunk/ (changelog src/actions.cpp src/unit.cpp src/unit.hpp): 
20100525 23:10:27< CIA-86> Reverted r42557 (patch #1668) as it just works around the real bug.
20100525 23:10:27< CIA-86> Fixed units being hidden anew after reloading. (Fix for bug #15889.)
20100525 23:10:27< CIA-86> The patch changes the "hidden" status to "uncovered", which has the opposite semantic and hopefully a less misleading name.
20100525 23:10:27< CIA-86> No backward compatibility is needed, since the saved status is currently useless and not worth restoring.
20100525 23:10:38< gabba> boucman: mmm, right
20100525 23:12:16< gabba> boucman: if you really want I can remove it, but all it will do is cause the arrow to return an empty list when asked for its images
20100525 23:12:28< boucman> (detail: display::getImagesFor and display::arrow_changed are probably private)
20100525 23:13:13< gabba> getImagesFor: ok
20100525 23:13:16< boucman> gabba: yes, i'd rather have you remove it, dead code is costly, and it's trivial to add back if the need arise
20100525 23:13:42< gabba> ok, no problem
20100525 23:14:07< gabba> display::arrow_changed is called by the arrows when they change, so needs to be public
20100525 23:15:05< boucman> oh, yes
20100525 23:16:32< boucman> my guess is that display::getImagesFor  will be called from a single place : display::draw_arrows which itself will be simply to factorize the arrow code out of display::draw
20100525 23:16:44< CIA-86> eleazar * r42879 /trunk/ (29 files in 7 dirs): more campaign maps updated to use new terrains.
20100525 23:17:03< boucman> and display::getImagesFor will be a simple foreach loop, so not worth doing a function for...
20100525 23:19:19< gabba> Yes, I'll try to do what makes sense there
20100525 23:20:10< boucman> ok good. I wanted to make sure you had the spirit not to stay "chained" to your UML design once hit with reality
20100525 23:20:24< CIA-86> silene * r42880 /branches/1.8/src/ (actions.cpp unit.cpp unit.hpp): 
20100525 23:20:24< CIA-86> Fixed units being hidden anew after reloading. (Fix for bug #15889.)
20100525 23:20:24< CIA-86> The patch changes the "hidden" status to "uncovered", which has the opposite semantic and hopefully a less misleading name.
20100525 23:20:24< CIA-86> No backward compatibility is needed, since the saved status is currently useless and not worth restoring.
20100525 23:20:24< CIA-86> Backported from trunk r42878.
20100525 23:21:11< boucman> and a remark : you went with the "display is in charge of drawing arrows" and not "display calls arrows that draw themselves" that fine.
20100525 23:21:32< gabba> boucman: it's always a temptation :P, but I'm not afraid to get my hands dirty in the code either
20100525 23:22:34< zookeeper> silene, probably still worth it to mention that status renaming in the changelog
20100525 23:22:37< gabba> yes, "display is in charge of drawing arrows" is mostly an efficiency question, remember, the whole data structure debate
20100525 23:22:38 * boucman is in charge of refactoring a 10GLOC project where 60% is useless autogenerated layers from UML tools
20100525 23:22:54< boucman> yup...
20100525 23:23:02< zookeeper> actually a wmllint rule to convert it wouldn't hurt either, since probably someone's trying to use it for something even if it doesn't work right
20100525 23:23:41< gabba> urgh, your refactoring project sounds painful... what do you mean exactly by "autogenerated layers" though?
20100525 23:24:21-!- Johannes13_ [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 265 seconds]
20100525 23:25:09< boucman> it's an Ada project based on UML
20100525 23:25:36< boucman> however, the OO concepts of Ada are very different from the C++ classes (which is the only thing UML was really designed for)
20100525 23:26:17< boucman> in particular a tagged type (i.e with inheritence and dispatching) is not related to encapsulation, unlike classes where public/private/protected are the main encapsulation tools
20100525 23:26:29-!- silene [~plouf@wesnoth/developer/silene] has quit [Read error: Operation timed out]
20100525 23:26:32< boucman> in Ada this is done via private parts in packages
20100525 23:26:44< boucman> the problem is the code generator is totally braindead
20100525 23:26:59< boucman> so it generates some complicated structures to emulate C++ classes
20100525 23:27:37< boucman> that are then used to implement the UML classes, to implement features that Ada has in the first place, but in an inefficient and unreadable way
20100525 23:28:13< gabba> Using a code generator for a very large project sounds like trouble to me, even in C++, so I can't imagine the result with a not-so-adapted language
20100525 23:28:15< boucman> in particular everything is done through multiple layers of dispatching calls (equivalent to virtual methods) so all the info that could be solved/optimized at compile time is lost
20100525 23:28:39< boucman> anyway, enought ranting
20100525 23:31:55< gabba> one design choice I did was to force the client code to create arrows through display::createNewArrow(); but said client code is still in charge of deleting the arrow
20100525 23:32:21< gabba> the arrow destructor will notify display to erase its images
20100525 23:32:40< boucman> that makes sense since the client is the only one to know the meaning of the arrow
20100525 23:35:27< boucman> you feel ready to start coding now ? or do you think there is more design needed (I personally think you're good to go, further design considerations will come while coding)
20100525 23:37:09< gabba> I still have a couple things to adjust before generating and starting on the method bodies
20100525 23:37:55< gabba> but in any case I'm already coding : when you write stuff such as std::map<map_location, std::list<std::pair<surface, arrow const*>>>, you're past the design phase :)
20100525 23:39:32-!- ilor_ [~ilor@wesnoth/developer/ilor] has joined #wesnoth-dev
20100525 23:39:32< boucman> hehe, indeed
20100525 23:40:22-!- YogiHH [YogiHH@d009076.adsl.hansenet.de] has left #wesnoth-dev []
20100525 23:42:01-!- ilor [~ilor@wesnoth/developer/ilor] has quit [Ping timeout: 264 seconds]
20100525 23:43:27< CIA-86> boucman * r42881 /trunk/data/core/terrain-graphics/ (compat.cfg util.cfg): move some macros from util.cfg to compat.cfg since we have better way to do that now. They are unused in mainline
20100525 23:45:09< gabba> boucman: btw, in general can we add "using std::map" and stuff like that to .cpp files? I know "using namespace std" is a bad idea, but the former could cut down on clutter
20100525 23:45:55< gabba> the other question would be, can I add some "using" statements to display.cpp, and if so, do I do anything to other code that uses the prefixes all over the place
20100525 23:46:33< boucman> gabba: i'm not a C++ expert, try to catch silene on the subject, but it sounds like a good idea
20100525 23:46:57< boucman> make sure to do it in a separate commit though, to simplify debugging if need be
20100525 23:47:03< gabba> sure
20100525 23:48:26< Ivanovic> timotei: please supply your pings with some content, or i can't help when finally back again...
20100525 23:51:09-!- elias [~elias@allegro/developer/allefant] has quit [Quit: Leaving]
20100525 23:52:31-!- zookeeper [~l@wesnoth/developer/zookeeper] has quit []
--- Log closed Wed May 26 00:00:37 2010