--- Log opened Sun May 29 00:00:28 2011
20110529 00:06:46< zaroth> boucman: well, when talking about help not being called from main menu, do you mean removing the button?
20110529 00:09:23< boucman> zaroth: if it simplifies a lot of code, or avoid parsing a whole lot of WML, why not
20110529 00:09:41< boucman> it's not an aim, but if it helps it's an idea worth investigating
20110529 00:09:56< boucman> ok, time to sleep, see you all tomorow
20110529 00:10:12-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Quit: Leaving.]
20110529 00:30:56-!- gabba [~gabba@wesnoth/developer/gabba] has left #wesnoth-dev []
20110529 00:37:20-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Quit: Valkier]
20110529 00:42:42-!- mjs-de [~mjs-de@d119223.adsl.hansenet.de] has quit [Remote host closed the connection]
20110529 00:47:21-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has joined #wesnoth-dev
20110529 01:01:58-!- noy [~Noy@wesnoth/developer/noy] has quit [Ping timeout: 246 seconds]
20110529 01:09:08-!- anonymissimus [~chatzilla@HSI-KBW-095-208-006-134.hsi5.kabel-badenwuerttemberg.de] has quit [Quit: ChatZilla 0.9.86.1 [Firefox 3.6.2/20100316074819]]
20110529 01:14:37-!- deekay [~dk@wesnoth/developer/dragonking] has quit [Quit: deekay]
20110529 01:19:49-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has joined #wesnoth-dev
20110529 01:38:58-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20110529 01:58:34-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Quit: Valkier]
20110529 02:01:31-!- zookeeper [~l@wesnoth/developer/zookeeper] has quit []
20110529 02:13:19-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has quit [Quit: Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz]
20110529 02:13:32-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection]
20110529 02:18:49-!- fendrin [~fabi@88-134-17-152-dynip.superkabel.de] has quit [Remote host closed the connection]
20110529 02:22:00-!- dipseydoodle [~mwkich@70.15.235.113.res-cmts.ovr.ptd.net] has quit [Quit: dipseydoodle]
20110529 02:26:59-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20110529 02:32:45-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20110529 02:42:22-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has left #wesnoth-dev []
20110529 02:57:36-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20110529 03:04:50-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has joined #wesnoth-dev
20110529 03:11:40-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has quit [Ping timeout: 260 seconds]
20110529 03:17:46-!- un214 [~un214@adsl-75-45-20-64.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20110529 03:18:41-!- Gambit is now known as ZoidberGambit
20110529 03:23:22-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has joined #wesnoth-dev
20110529 03:28:57-!- Upth [~ogmar@c-76-105-37-245.hsd1.ca.comcast.net] has joined #wesnoth-dev
20110529 03:28:57-!- Upth is now known as Upthorn
20110529 03:49:12-!- clanehin [~quassel@cpe-174-109-037-217.nc.res.rr.com] has quit [Remote host closed the connection]
20110529 04:22:03-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20110529 04:24:01-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has quit [Ping timeout: 240 seconds]
20110529 04:34:06-!- atomicbomb [~quassel@61.94.49.232] has joined #wesnoth-dev
20110529 04:44:55-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Ping timeout: 246 seconds]
20110529 04:49:45-!- ZoidberGambit is now known as Gambit
20110529 04:54:21-!- Ivanovic_ [~ivanovic@dtmd-4db2c25d.pool.mediaWays.net] has joined #wesnoth-dev
20110529 04:55:44-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has left #wesnoth-dev []
20110529 04:57:45-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 248 seconds]
20110529 04:58:17-!- Ivanovic_ is now known as Ivanovic
20110529 05:09:04-!- fendrin [~fabi@88-134-17-152-dynip.superkabel.de] has joined #wesnoth-dev
20110529 05:13:29-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has quit [Ping timeout: 252 seconds]
20110529 05:34:55-!- beetlenaut [~dan@174.32.63.204] has quit [Remote host closed the connection]
20110529 05:36:55-!- beetlenaut [~dan@174.32.63.204] has joined #wesnoth-dev
20110529 05:37:30-!- un214 [~un214@adsl-75-45-20-64.dsl.scrm01.sbcglobal.net] has quit [Remote host closed the connection]
20110529 05:47:06-!- beetlenaut [~dan@174.32.63.204] has quit [Remote host closed the connection]
20110529 05:52:36-!- shadowm_laptop is now known as Shadowmbit
20110529 05:53:28-!- Gambit is now known as Gambomaster
20110529 05:53:34-!- Gambomaster [~Gambit@wesnoth/developer/grickit] has quit [Disconnected by services]
20110529 05:53:53-!- Gambit [~Gambit@wesnoth/developer/grickit] has joined #wesnoth-dev
20110529 05:55:06-!- Gambit is now known as Shadowanchor
20110529 06:00:06-!- beetlenaut [~dan@174.32.63.204] has joined #wesnoth-dev
20110529 06:00:15-!- beetlenaut [~dan@174.32.63.204] has quit [Remote host closed the connection]
20110529 06:00:31-!- Shadowmbit is now known as Gambomaster
20110529 06:14:28-!- Gambomaster [ignacio@wesnoth/developer/shadowmaster] has quit [Remote host closed the connection]
20110529 06:15:57-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20110529 06:16:04-!- Shadowanchor is now known as Gambit
20110529 06:25:32-!- Gambit [~Gambit@wesnoth/developer/grickit] has quit [Quit: Twas in the darkest depths of Mordor, I met a girl so fair. But Gollum and the evil one crept up and slipped away with her.]
20110529 07:20:27-!- hopman- [~chris@c-71-202-164-232.hsd1.ca.comcast.net] has quit [Ping timeout: 255 seconds]
20110529 07:20:32-!- hopman- [~chris@c-71-202-164-232.hsd1.ca.comcast.net] has joined #wesnoth-dev
20110529 08:04:04-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20110529 08:54:01-!- Upthorn [~ogmar@c-76-105-37-245.hsd1.ca.comcast.net] has quit [Ping timeout: 248 seconds]
20110529 09:10:32-!- zaroth is now known as zar0th
20110529 09:10:41-!- zar0th is now known as zaroth
20110529 09:16:14-!- zookeeper [~l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20110529 09:36:40-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 260 seconds]
20110529 09:39:18-!- negusnyul [~negusnyul@h47-207.pool95-168.dyn.tolna.net] has joined #wesnoth-dev
20110529 09:46:34-!- MeccaGod [~majs@h146n3fls33o279.telia.com] has joined #wesnoth-dev
20110529 09:52:01-!- GeorgeSebastian [~georgeSeb@unaffiliated/georgesebastian] has joined #wesnoth-dev
20110529 09:56:09-!- MeccaGod [~majs@h146n3fls33o279.telia.com] has quit [Ping timeout: 260 seconds]
20110529 10:00:16-!- MeccaGod [majs@h146n3fls33o279.telia.com] has joined #wesnoth-dev
20110529 10:07:29-!- Crab_ [~Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20110529 10:20:08-!- Ivanovic [~ivanovic@dtmd-4db2c25d.pool.mediaWays.net] has quit [Changing host]
20110529 10:20:08-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20110529 10:21:16< Ivanovic> moin
20110529 10:30:53-!- Blueblaze [~Blueblaze@adsl-76-202-22-76.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20110529 10:47:26-!- MeccaGod [majs@h146n3fls33o279.telia.com] has quit []
20110529 10:58:21< zaroth> Crab_: did you look into questycaptcha?
20110529 10:59:46-!- stikonas [~and@bcm-131-111-216-209.girton.cam.ac.uk] has joined #wesnoth-dev
20110529 10:59:46-!- stikonas [~and@bcm-131-111-216-209.girton.cam.ac.uk] has quit [Changing host]
20110529 10:59:46-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20110529 11:05:17< Crab_> zaroth: no, not yet. only taken a look at the docs.
20110529 11:12:19-!- negusnyul [~negusnyul@h47-207.pool95-168.dyn.tolna.net] has quit [Remote host closed the connection]
20110529 11:30:03-!- loonybot [~loonybot@46.138.69.145] has joined #wesnoth-dev
20110529 11:30:03-!- loonybot [~loonybot@46.138.69.145] has quit [Changing host]
20110529 11:30:03-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-dev
20110529 11:30:27-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20110529 11:47:37-!- atomicbomb [~quassel@61.94.49.232] has quit [Quit: take a bath and dinner]
20110529 12:00:28-!- Octalot [~noct@host86-147-146-20.range86-147.btcentralplus.com] has quit [Remote host closed the connection]
20110529 12:07:55-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20110529 12:15:58< zaroth> boucman: are you here?
20110529 12:17:28< CIA-94> zaroth * r49697 /trunk/doc/man/wesnoth.6: Added information about --new-syntax to manpage
20110529 12:17:43< CIA-94> zaroth * r49698 /trunk/src/ (5 files): 
20110529 12:17:43< CIA-94> Created game_controller_new, a new game controller class using
20110529 12:17:44< CIA-94> game_controller_abstract interface.
20110529 12:17:46< CIA-94> zaroth * r49699 /trunk/src/ (9 files): 
20110529 12:17:46< CIA-94> added game_controller_new, which takes over game_controller
20110529 12:17:46< CIA-94> responsibilities in case of --new-syntax
20110529 12:18:21-!- atomicbomb [~quassel@61.94.49.188] has joined #wesnoth-dev
20110529 12:19:07< zaroth> Crab_: do you have a second? I'd like to move the init_video() from the game_controller to game_controller_abstract
20110529 12:19:21< Crab_> zaroth: yes, I've got some time now
20110529 12:19:35< zaroth> since I don't intend changing anything about the resources loading stuff such as this
20110529 12:20:00< zaroth> but it depends on quite a few class members, which are currently exclusive to game_controller
20110529 12:20:09< zaroth> and I wonder what approach should I take
20110529 12:20:11-!- mjs-de [~mjs-de@d119197.adsl.hansenet.de] has joined #wesnoth-dev
20110529 12:20:56< zaroth> also I don't very much like the fact that constructor of game_controller is currently a commandline parser, part II
20110529 12:22:14< zaroth> so far I moved to game_controller_abstract only the absolute minimum to get wesnoth to compile
20110529 12:22:21< zaroth> now I'd like to have the main menu running
20110529 12:22:28< zaroth> and I need video for this ;-)
20110529 12:24:59< Crab_> If I look a the list of members of game_controller, I see that there's quite different stuff in there
20110529 12:25:35< zaroth> I wonder what would you suggest, moving some of the game_controller options up to the game_controller_abstract and part of the commandline parsing as well?
20110529 12:26:23< zaroth> or make the new game_controller completely commandline switches agnostic (this would be easiest for now, but would take a lot of work to get the switches working later...)
20110529 12:26:40< Crab_> but there's a list of managers which is scoped to wesnoth's main menu, and which is not related to the way *how* we do the game loop, but more related to the list of service operations we can do
20110529 12:27:12< Crab_> firstly, about the command-line parsing
20110529 12:27:36< Crab_> I suggest making a separate class for command-line options.
20110529 12:28:11< zaroth> isn't current game_config:: namespace suited for that?
20110529 12:28:12< Crab_> there will be 1 instance of it created, with responsibility of 'translate from --options to values stored in public members'.
20110529 12:28:24< zaroth> (I'd be all for the name change, though)
20110529 12:28:30< Crab_> no, game_config:: has meaning in game
20110529 12:29:00< Crab_> and I am thinking of a helper struct just to leave the low-level command-line parsing stuff behind.
20110529 12:29:48< Crab_> then, at the start of wesnoth, we would be able to parse the command-line options, get a 'command_line_arguments' object and use it hereafter.
20110529 12:30:06< zaroth> there are at least some things in game_config:: that don't have much meaning in game
20110529 12:30:37< zaroth> revision, path, wesnoth_program_dir are only a few such examples
20110529 12:30:49< zaroth> or my new_syntax, for that matter
20110529 12:31:28< Crab_> so, instead of         '   for(arg_ = 1; arg_ != argc_; ++arg_) {  ...     if(val == "--fps") {     preferences::set_show_fps(true);  '  we'll have ' preferences::set_show_fps(command_line_options.show_fps); '
20110529 12:32:33< Crab_> zaroth: by 'in game' I mean that for stuff in game_config, it is used in the code, not just stored.
20110529 12:33:20< Crab_> zaroth: and for the new options class, I just want to do the translation between 'not-yet-parsed-command-line-arguments' and a neat c++ class with public members of correct types.
20110529 12:33:40< Crab_> zaroth: basically, 'from string array to c++ data structure' step.
20110529 12:34:28-!- Crendgrim [~crend@77-22-173-219-dynip.superkabel.de] has joined #wesnoth-dev
20110529 12:35:15< Crab_> then the game_controller would be free of low-level code for that
20110529 12:35:54< zaroth> hmm, I'm not yet sure of how different commandline switches could be parsed and various traps that sure are hiding in there
20110529 12:36:19< zaroth> but since currently commandline parsing is spread only between three functions, it shouldn't be that hard
20110529 12:36:41< Crab_> you can leave it for later, of course, but if you want to clean it up, it's relatively simple.
20110529 12:36:51-!- atomicbomb [~quassel@61.94.49.188] has quit [Quit: Go home]
20110529 12:36:57< zaroth> what do you think about that: http://www.boost.org/doc/libs/1_46_1/doc/html/program_options.html ?
20110529 12:37:39< Crab_> basically, by using a helper class to store the parsed stuff, we avoid the need to know the references to the object which we need to modify, when we parse.
20110529 12:37:42< zaroth> I recently used that library for a program of my own to parse the commandline and store things, it was quite a pleasant experience
20110529 12:38:16< Crab_> yes, using a library to deal with the step is a good idea.
20110529 12:38:34< Crab_> but, to use it, we need that cleanup to be done in the process 
20110529 12:39:11< zaroth> yes, I just thought that the commandline parser class could be written easier with use of this library
20110529 12:39:27< Crab_> it will give you more than just a parser, which is a good thing
20110529 12:39:45< Crab_> returning to game_controller...
20110529 12:40:38-!- MeccaGod [~majs@h146n3fls33o279.telia.com] has joined #wesnoth-dev
20110529 12:43:19< Crab_> after the options are dealt with, the constructor of play_controller would be much neater
20110529 12:43:45< zaroth> you meant game_ or play_ ?
20110529 12:43:52< Crab_> sorry, game_
20110529 12:45:06-!- dacovale [~niklas@h55eb1ca0.selukra.dyn.perspektivbredband.net] has joined #wesnoth-dev
20110529 12:45:10< Crab_> then, there are two kinds of stuff remaining - stuff related to initialization of the game (video, languages, etc), and stuff related to 'what to do next' logic.
20110529 12:45:25< zaroth> yes, but I'm still not sure how to initialize the video, should I add a few parameters to the init_video(), which would be the no_gui and other relevant things from command_line_options?
20110529 12:46:05< Crab_> zaroth: abstract_game_controller would have a const ref to command_line_parameters object.
20110529 12:46:34< Crab_> zaroth: so, since you went with abstract_.. route, then you'd just keep everything related to initialization of the game (video, languages, etc) in abstract_game_controller
20110529 12:46:56< dacovale> wesbot: seen JonJon
20110529 12:46:57< wesbot> dacovale: The person with the nick JonJon last spoke 5d 19h ago.  5d 19h ago they left with the message: Quit: Page closed
20110529 12:46:58< Crab_> and leave the non-abstract implementations to deal with the logic.
20110529 12:47:08< zaroth> yeah, init_* except init_config() ;-)
20110529 12:47:45< Crab_> zaroth: part of init_config as well.
20110529 12:48:00< Crab_> zaroth: note that the game config contains both stuff related to wesnoth's core and to scenarios/campaigns/etc
20110529 12:48:12< Crab_> zaroth: i.e. the gui definitions, music lists, etc, are in WML, too.
20110529 12:48:24< Crab_> zaroth: and that's not that good, actually.
20110529 12:48:25< zaroth> well, I wanted to change that
20110529 12:48:57< Crab_> zaroth: so, you would still need a partial 'init_config' in abstract class to load the stuff required for the interface
20110529 12:49:00< zaroth> and use two hardwired/ files for the bootstrap of game_controller_new config and game_instance config during the transition period
20110529 12:49:12< Crab_> zaroth: it will be unaffected by defines, btw.
20110529 12:49:47< zaroth> but I'm not sure if the single cache won't foil my tries... Cache currently is a singleton, right?
20110529 12:50:27< Crab_> zaroth: used via game_config::config_cache::instance(), yes
20110529 12:50:40< Crab_> zaroth: but in your approach you'll need more than one, I think
20110529 12:50:58< Crendgrim> zaroth, Crab_: The captcha change didn't help much. There are again 5 new spams today
20110529 12:51:07< Crab_> Crendgrim: wasn't done yet.
20110529 12:51:12< Crendgrim> oh, okay
20110529 12:51:16< Crendgrim> sorry ;)
20110529 12:52:04< Crab_> zaroth: since, most likely, we'll want to separately use the (1) stuff-required-to-start-up-wesnoth, (2) stuff-required-to-populate-the-game-lists, stuff-for-current-game
20110529 12:52:05< zaroth> Crendgrim: yeah, you can logout and go to http://wiki.wesnoth.org/index.php?title=Special:UserLogin&type=signup to check if it's changed yet ;-)
20110529 12:52:41< zaroth> Crab_: yes, that's what I wanted the hardwired/main_config.cfg and hardwired/game_instance.cfg for
20110529 12:52:47< Crendgrim> zaroth: I didn't check yet... or I could do that by using another browser ;)
20110529 12:53:02< zaroth> not to disturb the current data/ directory structure too much
20110529 12:53:23< zaroth> but there is one breach in this
20110529 12:53:37< Crab_> which one ?
20110529 12:53:39< zaroth> help button from the main menu needs (2)
20110529 12:53:55< zaroth> and that's the only case of such a breach 
20110529 12:54:18< zaroth> but I guess we could show a help without units/traits and such, just the general stuff
20110529 12:54:19< Crab_> why it is a only case ? 'play campaign' has to get a list of campaigns somehow )
20110529 12:54:54< zaroth> well, but play campaign creates a game_instance, which is the primary user of (1), so I don't consider it a breach ;-)
20110529 12:55:11< Crab_> can't we create a game instance for help, as well ?
20110529 12:55:28< Crab_> since help which shows a list of units is dependent on specific era/set of addons/etc
20110529 12:56:02< Crab_> so, we need  a set-of-configs-just-like-in-a-particular-game to be loaded.
20110529 12:57:08< zaroth> that would result in a bit of loading time each time the help button is pressed or require to create hold this game_instance within gameloop, but yes, can be done
20110529 12:57:56< Crab_> yes, but now this time is always spent.
20110529 12:58:05< zaroth> (each time the help button is pressed from main menu, that is)
20110529 12:58:10< Crab_> and after your change, it would be only spent if someone actually presses the button
20110529 12:58:16< Crab_> only 'the first time' if it's cached
20110529 12:58:32< Crab_> so, it would be better then now.
20110529 12:59:50< zaroth> hmm... I'm also not sure if I would like game.cpp to create game_instance, since it's not how the encapsulation was supposed to be (it was supposed to be gameloop - > game_controller -> game_instance)...
20110529 12:59:56< zaroth> but it's a minor issue
20110529 13:01:04< Crab_> well, the code which controls help should be of similar interface to the code which controls the 'game'
20110529 13:01:35< Crab_> since they both will do the same thing - ensure that the correct configs are loaded, and show some gui/gui2 dialogs
20110529 13:02:02< Crab_> as of now, for MP, this is a connect to server & lobby dialogs, for sp, this is 'select campaign' dialog, etc
20110529 13:02:08< Crab_> for help, this would be the help dialog.
20110529 13:02:11< Crab_> what's that different ?
20110529 13:02:37< Crab_> we just don't have a way for launching the game from help ) but, conceptually, its more-or-less the same
20110529 13:03:02< zaroth> yes, it can be seen that way
20110529 13:05:35< Crab_> note that we want to structure our new game controller in such a way which would allow us to go from MP to SP, and, technically, it should allow to go from SP to help (or lobby to any other dialog), and continue with the game afterwards
20110529 13:07:59< Crab_> so, in the end, it's possible that both 'go to MP', 'help', and 'play campaign' would be handled by same code.
20110529 13:10:02< Crab_> this would allow to add new types of in-campaign transitions, as well.
20110529 13:10:17< zaroth> a moment of reflection: I thought I would be working with mp campaigns, and here I am, writing a commandline parser class
20110529 13:10:27< zaroth> ;-)
20110529 13:10:48< Crab_> zaroth: if it takes too much time, just move it aside with a simple refactoring, without trying to change the way it works.
20110529 13:11:26< Crab_> zaroth: actually, you can just add that in-between class, move the code which parses the command line to single place outside the game_controller, and modify the game_controller to use that in-between class.
20110529 13:11:39< Crab_> so, you'd be able to reuse the old command-line parsing logic
20110529 13:11:48< Crab_> albeit reading code is more hard than writing it...
20110529 13:12:56< zaroth> well, I have to both read and write it anyway, since I have to understand how it's being parsed to use proper storing methods
20110529 13:13:38< Crab_> yes, but in many cases it's easier to rewrite-from-scratch instead of making the (good) old code work.
20110529 13:13:52< Crab_> but the code is good, it's just it feels different...
20110529 13:15:52-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20110529 13:23:14< zaroth> by the way, in case a commandline option is parsed directly to game_config::, should my class do it instead of exposing it publicly in its interface?
20110529 13:23:33< zaroth> oh well, I guess it'll need to be judged on a per-case basis
20110529 13:23:39< Crab_> I'd just expose it publicly.
20110529 13:24:21< Crab_> but, if you hate duplication, assign to game_config:: . just be consistent.
20110529 13:26:25< Crab_> i.e., make it a mandatory thing for things in game_config::
20110529 13:29:58-!- Crab_ [~Crab_@wesnoth/developer/crab] has quit [Quit: Leaving.]
20110529 13:33:55< zaroth> I guess I'll pass on it, then ;-)
20110529 13:34:08< zaroth> consistency wins with duplication
20110529 13:51:19-!- Gallaecio [~quassel@88.103.82.122] has quit [Remote host closed the connection]
20110529 13:53:43-!- Gallaecio [~quassel@88.103.82.122] has joined #wesnoth-dev
20110529 14:03:08-!- Gambit [~Gambit@wesnoth/developer/grickit] has joined #wesnoth-dev
20110529 14:16:09-!- Gallaecio [~quassel@88.103.82.122] has quit [Read error: Connection reset by peer]
20110529 14:16:25-!- negusnyul [~negusnyul@dsl4E5CCA87.pool.t-online.hu] has joined #wesnoth-dev
20110529 14:18:19-!- Gallaecio [~quassel@88.103.82.122] has joined #wesnoth-dev
20110529 14:29:00-!- Gallaecio [~quassel@88.103.82.122] has quit [Remote host closed the connection]
20110529 14:31:13-!- Gallaecio [~quassel@88.103.82.122] has joined #wesnoth-dev
20110529 15:00:13-!- GeorgeSebastian [~georgeSeb@unaffiliated/georgesebastian] has quit [Ping timeout: 240 seconds]
20110529 15:08:44-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has quit [Quit: I ATE'NT DEAD]
20110529 15:08:51-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection]
20110529 15:22:46-!- GeorgeSebastian [~georgeSeb@117.231.121.198] has joined #wesnoth-dev
20110529 15:22:46-!- GeorgeSebastian [~georgeSeb@117.231.121.198] has quit [Changing host]
20110529 15:22:46-!- GeorgeSebastian [~georgeSeb@unaffiliated/georgesebastian] has joined #wesnoth-dev
20110529 15:26:01-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-dev
20110529 15:28:15-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20110529 15:33:42-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20110529 16:03:07-!- Crendgrim [~crend@77-22-173-219-dynip.superkabel.de] has quit [Quit: Konversation terminated!]
20110529 16:12:43-!- Crendgrim [~crend@77-22-173-219-dynip.superkabel.de] has joined #wesnoth-dev
20110529 16:37:10< fendrin> zaroth: I have filled a feature request some time ago. It is now part of your gsoc project, so I assign it to you so it won't get forgotten.
20110529 16:37:34< zaroth> fendrin: sure
20110529 16:38:59< fendrin> zaroth: It's no extra work, just "difficult levels for MP Campaigns", I just hate ancient Feature Requests to clutter the interface.
20110529 17:21:33-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has joined #wesnoth-dev
20110529 17:25:22< Nephro> fendrin, hi, do you have the access to assign patches/bugs/frs to people on gna?
20110529 17:26:17< fendrin> Nephro: Sure, every developer has. Or do I misunderstand your question?
20110529 17:26:59< Nephro> fendrin, http://gna.org/patch/?2674 can you assign this to Crab, please? He wanted to see it, but must've missed in the logs...
20110529 17:27:45< fendrin> Nephro: No problem. Nephro I would liked to assign a bug to you but could not find "nephro" in the list. Is your gna name different?
20110529 17:28:01< Nephro> Submitted by: 	Dmitry K. <nephro>
20110529 17:28:13< Nephro> (on gna)
20110529 17:29:56< fendrin> Nephro: It seems you are missing some group membership to be a recipient of bugs.
20110529 17:30:15< fendrin> Nephro: I have assigned 2674 to crab.
20110529 17:31:17< Nephro> I think I haven't yet received svn access. Probably because my task was pretty big and had to be significantly refactored. All the process was delayed by my exam session
20110529 17:31:45< Nephro> But that should be solved, once Crab sees(and hopefully approves) 2674
20110529 17:32:03< Nephro> fendrin, for now, you can just give me a link to the bug, I'll check it out
20110529 17:32:09< fendrin> Nephro: No problem. Crab will also take care of the bug report.
20110529 17:32:29< fendrin> Nephro: https://gna.org/bugs/?13119
20110529 17:32:53< Nephro> thanks
20110529 17:33:30< fendrin> You are welcome.
20110529 17:38:16-!- Cookiee [~quassel@unaffiliated/cookiee] has joined #wesnoth-dev
20110529 17:38:43-!- Cookiee [~quassel@unaffiliated/cookiee] has left #wesnoth-dev []
20110529 18:03:19-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Quit: Valkier]
20110529 18:03:54-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20110529 18:18:22-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has joined #wesnoth-dev
20110529 18:28:25-!- GeorgeSebastian [~georgeSeb@unaffiliated/georgesebastian] has quit [Ping timeout: 248 seconds]
20110529 18:37:43-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Quit: Valkier]
20110529 18:48:24-!- anonymissimus [~chatzilla@HSI-KBW-095-208-006-134.hsi5.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20110529 18:51:35< anonymissimus> fendrin: did you succeed in making emacs lua mode indent correctly ? found this: http://stackoverflow.com/questions/4643206/how-to-configure-indentation-in-emacs-lua-mode
20110529 18:52:59-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has joined #wesnoth-dev
20110529 18:53:28< fendrin> anonymissimus: No, not yet. I will have a look, thank you.
20110529 18:55:49-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has left #wesnoth-dev []
20110529 18:59:12< CIA-94> fendrin * r49700 /trunk/src/ (unit.cpp unit.hpp): Renamed the [recall_filter] tag to [filter_recall].
20110529 19:04:36-!- mordante [~mordante@roadie.xs4all.nl] has joined #wesnoth-dev
20110529 19:04:39-!- mordante [~mordante@roadie.xs4all.nl] has quit [Changing host]
20110529 19:04:39-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20110529 19:04:58< mordante> servus
20110529 19:06:18-!- Crend [~crend@77-22-173-219-dynip.superkabel.de] has joined #wesnoth-dev
20110529 19:06:49-!- Crendgrim [~crend@77-22-173-219-dynip.superkabel.de] has quit [Ping timeout: 248 seconds]
20110529 19:07:35-!- Crend is now known as Crendgrim
20110529 19:07:55< mordante> zaroth, regarding boost program options grzywacz once looked at it, but ran into problems, not sure what however
20110529 19:08:49< zaroth> mordante: I also found a few problems, but I think I will solve them changing the syntax slightly in few cases
20110529 19:09:10< zaroth> making it more explicit
20110529 19:09:26< mordante> zaroth, ok, maybe if grzywacz lurks around you can ask him about what he ran intp
20110529 19:09:29< mordante> into*
20110529 19:10:09< zaroth> the quality of commandline parsing should be overall better anyway, since it does all the typechecking automatically etc.
20110529 19:10:43< mordante> I agree, I used boost programming options for some other projects
20110529 19:13:34< zaroth> mordante: did you use multitoken() in it?
20110529 19:14:15< zaroth> because we have a few commands that use something like this syntax in wesnoth currently
20110529 19:14:41< zaroth> e.g. --campaign
20110529 19:14:42< mordante> doesn't sound familiar, so I think not
20110529 19:15:12< zaroth> and I can find the modificator to make it multitoken here: http://www.boost.org/doc/libs/1_46_1/doc/html/boost/program_options/typed_value.html
20110529 19:15:25< zaroth> but the min_tokens and max_tokens appear to be getters
20110529 19:15:33< zaroth> where are the setters? couldn't find them in boost docs
20110529 19:16:38< zaroth> or maybe it's yet another obscure C++ idiom somewhere I'm not familiar yet with...
20110529 19:19:12-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20110529 19:19:22< mordante> no that doesn't sound familiar
20110529 19:19:31-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has left #wesnoth-dev []
20110529 19:20:43-!- Upth [~ogmar@c-76-105-37-245.hsd1.ca.comcast.net] has joined #wesnoth-dev
20110529 19:20:43-!- Upth is now known as Upthorn
20110529 19:31:15< CIA-94> zaroth * r49701 /trunk/ (6 files in 2 dirs): 
20110529 19:31:15< CIA-94> Created new command_options object, which will be responsible for
20110529 19:31:15< CIA-94> commandline parsing. Introduces Boost.Program_options dependency.
20110529 19:32:08-!- wesbot changed the topic of #wesnoth-dev to: 218 bugs, 315 feature requests, 28 patches | logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://pastebin.com | http://imagebin.org
20110529 19:32:33< zaroth> 315 feature requests!
20110529 19:33:18< zaroth> and these are just the open ones, right?
20110529 19:36:10< mordante> I'm off bye
20110529 19:36:17-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20110529 19:40:01-!- Blueblaze [~Blueblaze@adsl-76-202-22-76.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20110529 19:58:24-!- ancestral [~ancestral@67-6-61-230.mpls.qwest.net] has joined #wesnoth-dev
20110529 20:01:06-!- ancestral [~ancestral@67-6-61-230.mpls.qwest.net] has quit [Client Quit]
20110529 20:02:21-!- ancestral [~ancestral@67-6-61-230.mpls.qwest.net] has joined #wesnoth-dev
20110529 20:06:26-!- Netsplit *.net <-> *.split quits: iwaim___
20110529 20:11:13-!- Netsplit over, joins: iwaim___
20110529 20:14:28-!- Crab_ [~Crab_@62.80.190.34] has joined #wesnoth-dev
20110529 20:14:35-!- Crab_ [~Crab_@62.80.190.34] has quit [Changing host]
20110529 20:14:35-!- Crab_ [~Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20110529 20:16:19-!- MeccaGod [~majs@h146n3fls33o279.telia.com] has quit [Ping timeout: 240 seconds]
20110529 20:16:35-!- Upthorn [~ogmar@c-76-105-37-245.hsd1.ca.comcast.net] has quit [Read error: Connection reset by peer]
20110529 20:16:57< Nephro> Hello, Crab_ ! I've posted the patch on GNA...
20110529 20:17:05< Crab_> hi, I've seen it
20110529 20:17:22< Crab_> I will post the review today, in a few hours
20110529 20:17:23< Nephro> Is it any good?
20110529 20:17:46< Nephro> oh, okay.
20110529 20:18:01< Nephro> I'll wait then.
20110529 20:18:30< Crab_> ok. you can take the time to study formula ai, to see what functions it exposes, note that most of them are already exposed via LuaWML
20110529 20:18:47-!- vultraz [~chatzilla@124.109.10.222] has joined #wesnoth-dev
20110529 20:19:50< Nephro> Crab_, you mean this list: http://wiki.wesnoth.org/FormulaAI_Functions?
20110529 20:20:00< Crab_> yes, or the source code in src/ai/formula/
20110529 20:20:35< Crab_> AI_specific_functions from there
20110529 20:20:54-!- MeccaGod [majs@h146n3fls33o279.telia.com] has joined #wesnoth-dev
20110529 20:27:53< anonymissimus> fendrin: your recall filter rename misses at least the changelog
20110529 20:28:01< anonymissimus> >:
20110529 20:28:10< anonymissimus> >:(
20110529 20:33:33< CIA-94> fendrin * r49702 /trunk/changelog: [recruit_filter] --> [filter_recruit]
20110529 20:39:14-!- Crendgrim [~crend@77-22-173-219-dynip.superkabel.de] has quit [Remote host closed the connection]
20110529 20:40:16< CIA-94> espreon * r49703 /trunk/src/ (5 files): Set svn:eol-style and svn:keywords.
20110529 20:54:41< zaroth> ah, sweet.
20110529 20:55:22< zaroth> it appears that adding find_package() in CMakelists or checkBoost in sconstruct is not enough
20110529 20:55:36< zaroth> what do I do to actually link with boost.program_options?
20110529 20:58:20< zaroth> I solved it in CMake, still looking where to add the library in scons
20110529 21:00:04< loonycyborg> zaroth: You don't need to do anything else in scons.
20110529 21:00:34< loonycyborg> CheckBoost happens to be enough.
20110529 21:00:45< zaroth> loonycyborg: how do you know that?
20110529 21:01:01< loonycyborg> Because it works for me.
20110529 21:01:16< zaroth> by testing my revision? I'm not using program_options there yet
20110529 21:01:22< loonycyborg> And because I made that CheckBoost originally.
20110529 21:01:34< zaroth> I just started using it in local revision, that's when it broke
20110529 21:01:49< zaroth> ah, alright then, I believe you ;-)
20110529 21:01:55< loonycyborg> It passes -lboost_program_options.
20110529 21:02:13-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Quit: Valkier]
20110529 21:04:35< Crab_> zaroth: note that when you add a new dependency, it's a good thing to shoot an email to dev-ml, just in case.
20110529 21:05:10< zaroth> Crab_: okay
20110529 21:05:48< zaroth> did you have a look, btw? It's just a header now in the svn, but I tried to keep it as consistent as possible
20110529 21:08:25< Crab_> well, so far I wanted to take a look at one of the new bugs, but wesnoth refused to compile because of program-options :)
20110529 21:08:57< Crab_> so I have to install a few libs for it to work, or hack it out / switch to earlier wesnoth version.
20110529 21:12:44-!- Upthorn [~ogmar@c-76-105-37-245.hsd1.ca.comcast.net] has joined #wesnoth-dev
20110529 21:12:44-!- Upthorn is now known as Upth
20110529 21:15:53< Crab_> zaroth: thanks
20110529 21:16:45< anonymissimus> zaroth: also, if this breaks compilation on win it's nice to know as much as possible ;)
20110529 21:17:11< zaroth> well, it's not like we're using a completely new library...
20110529 21:17:31< zaroth> it's just another component of boost ;-)
20110529 21:17:58< anonymissimus> thats good; hopefully it's in one of the already used libs/dlls
20110529 21:18:20< loonycyborg> Nope.
20110529 21:18:20< anonymissimus> if it fails it's usually the linking of boost
20110529 21:19:05< loonycyborg> I'll have to build another boost library on next release.
20110529 21:19:32< anonymissimus> _oh_ _nos_
20110529 21:20:34< loonycyborg> zaroth: But I really suggest you only add dependencies to buildsystem once they're *really* required.
20110529 21:21:01< loonycyborg> If you want to commit your wip code then do that in a branch, or in local git-svn
20110529 21:21:46< Crab_> loonycyborg: the code in question is simple enough to convert, I guess it won't take more than 24hours before the new lib will be required. of course the lack of advance warning is not cool.
20110529 21:22:05< zaroth> well, I don't see myself adding new deps anytime soon
20110529 21:22:16< zaroth> except this one
20110529 21:22:25-!- Upth [~ogmar@c-76-105-37-245.hsd1.ca.comcast.net] has quit [Quit: Some message that is totally family friendly and work safe.]
20110529 21:29:08-!- Blueblaze [~Blueblaze@adsl-76-202-22-76.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20110529 21:29:42< anonymissimus> if all Linux only-s would make a windows built before their commits life was easier :P
20110529 21:32:22-!- anonymissimus [~chatzilla@HSI-KBW-095-208-006-134.hsi5.kabel-badenwuerttemberg.de] has quit [Quit: ChatZilla 0.9.86.1 [Firefox 3.6.13/20101206122310]]
20110529 21:35:39-!- anonymissimus [~chatzilla@HSI-KBW-095-208-006-134.hsi5.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20110529 21:53:32-!- Octalot [~noct@host86-147-146-20.range86-147.btcentralplus.com] has joined #wesnoth-dev
20110529 21:56:11-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has joined #wesnoth-dev
20110529 22:01:47-!- MeccaGod [majs@h146n3fls33o279.telia.com] has quit []
20110529 22:05:19-!- negusnyul [~negusnyul@dsl4E5CCA87.pool.t-online.hu] has quit [Remote host closed the connection]
20110529 22:07:34-!- erl [~erl@vetinte.spelaroll.se] has quit [Ping timeout: 260 seconds]
20110529 22:29:20< anonymissimus> yep, MSVC fails to link
20110529 22:30:07< anonymissimus> on Linux doing apt-get install libboost-all-dev solves this I guess
20110529 22:32:04< zaroth> well, I didn't have to do even that, since in gentoo every library is -dev
20110529 22:32:20< zaroth> but I certainly ought to try and build wesnoth on MSVC one day
20110529 22:32:28< Espreon> Good luck...
20110529 22:32:34< zaroth> this could be an enlightening experience
20110529 22:37:09< loonycyborg> Not only that, the entire boost is a single ebuild.
20110529 22:37:34< loonycyborg> Totally unlike debian with its libboost-program-options etc
20110529 22:39:36< zaroth> by the way, we use foo_ for members
20110529 22:39:49< zaroth> do we have something like that for local variables?
20110529 22:40:00< loonycyborg> afaik no.
20110529 22:40:36< zaroth> that is, I have a local variable and I'd like to explicitly name it as such to avoid name collision with public members, which due to our policy don't have any _ or sth like that
20110529 22:40:38< loonycyborg> _ is affixed to private members only, but it's not like you'll have any other :P
20110529 22:41:06< Crab_> fendrin: ok, I've found the reasons of the crash that  LoW is having.
20110529 22:41:40< Crab_> Nephro: (this will bite the lua code, as well)  ^
20110529 22:41:42< zaroth> loonycyborg: erm... http://pastebin.com/034Re1gD
20110529 22:41:54< fendrin> Crab_: What is it?
20110529 22:41:55< Crab_> fendrin: formula ai has a attribute 'recruits_of_side', a function which returns an array of recruits for the side
20110529 22:41:55< zaroth> private: 5 public: a lot
20110529 22:42:01< zaroth> but I guess it's an exception ;-)
20110529 22:42:09< Crab_> fendrin: and it doesn't take leader-specific recruits into consideration.
20110529 22:42:29< Crab_> fendrin: later, one of DKs undocumented functions for calculation of map ownership receives that as a parameter and does 'divide by 0'
20110529 22:43:47< fendrin> Crab_: So it should also divide by zero if there is a side without recall list.
20110529 22:43:55< Crab_> fendrin: it ignores recall.
20110529 22:44:13< Crab_> fendrin: that DKs function is only used with recruit list
20110529 22:44:20< fendrin> Ah yes.
20110529 22:44:20< Crab_> fendrin: one of the fixes is to make that 'recruits_of_side' return all leader-specific recruits, as well
20110529 22:44:37< Crab_> fendrin: another fix is to change LoW2 back from DKs recruitment to default one.
20110529 22:44:47< loonycyborg> zaroth: Honestly I'm not even sure that having them all in a single class is a good idea.
20110529 22:45:13< Crab_> fendrin: opinion ?
20110529 22:45:56< zaroth> loonycyborg: well, that's what Crab_ and I brainstormed today, any other ideas to avoid having commandline parsing spread in 3 places in code?
20110529 22:46:14< fendrin> Crab_: Is it a matter of opinions?
20110529 22:46:20< loonycyborg> Other ideas would be to have global variables in appropriate places and have the po code set them,
20110529 22:46:20-!- Blueblaze [~Blueblaze@adsl-76-202-22-76.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20110529 22:47:00< loonycyborg> or just expose the variables_map globally.
20110529 22:47:08< Crab_> loonycyborg: the idea was exactly to get 1 'global' object with vars and have the po code set them.
20110529 22:47:24< Crab_> thus splitting low-level vs high-level parsing of option list
20110529 22:48:18< Crab_> fendrin: yes, I just wanted to know an opinion
20110529 22:48:38< fendrin> Crab_: Changing LoW to not use it will not fix the bug. There are several things to do. 
20110529 22:49:01< zaroth> loonycyborg: hm, I thought about commandline_options doing that (setting stuff in different places), but in several cases it's a bit tricky to do
20110529 22:49:02< Crab_> fendrin: well, if the idea is to deprecate formula_ai (keeping the formula language internally) in the long run...
20110529 22:49:56< fendrin> Crab_: hmmm, I don't know much about your ai plans to have an option on that.
20110529 22:49:57< zaroth> and I decided that simply outsourcing the current approach is easy enough to change later and solving my problem that it was blocking, so - good enough
20110529 22:51:34< zaroth> besides, thanks to boost::optional access to commandline_options members is pretty uniform
20110529 22:52:37< Crab_> for now, I will fix the 'get recruits of side' to return all potential recruits. this will break some other assumptions and lead to other issues, but they will be less serious
20110529 22:52:49< loonycyborg> It seems like a repetition to me, you both declare a class member and define a po option. It would be more tolerable if they were in whatever places they actually affect.
20110529 22:54:35< Crab_> loonycyborg: maybe then add a way for users to 'register' command-line option receivers ? i.e., if some code needs a command-line option, it'll add it via a small wrapper object, like it's done with log domains.
20110529 22:54:46< anonymissimus> zaroth: ok well, thx to timotei's download tip I've apparently solved MSVC linking
20110529 22:55:03< Crab_> fendrin: the problem is, that now 'what units side can recruit now?' and 'what side can recruit potentially?' are two different questions.
20110529 22:55:27< zaroth> loonycyborg: well, but thanks to doing it that way I can handle some parsing stuff, such as resolution AxB within the command_options class
20110529 22:55:28< fendrin> Crab_: There is another related issue. You use a "minimum_recruit_cost" integer in your ai code. This value does not respect the leader recruit list.
20110529 22:55:28< loonycyborg> Anyway, local variables only shadow class member, perhaps there's a way to still access them. It's definitely possible to access shadowed global variables using ::
20110529 22:55:49< Crab_> fendrin: yes, that's true, too.
20110529 22:55:57< anonymissimus> interestingly, coblocks does still link fine; where does it get the new lib from I wonder
20110529 22:56:09< loonycyborg> Crab_: I like that approach.
20110529 22:56:17< Crab_> anonymissimus: maybe it doesn't compile the new file at all ?
20110529 22:56:19< zaroth> and expose it as std::pair<int,int> - that actually is the kind of parsing job which should be done in commandline_options, not in game_controller 
20110529 22:56:39< anonymissimus> codeBLocks ? checking...
20110529 22:57:10< Crab_> fendrin: maybe add functions for 'potential minimum recruit cost' ,'potential average recruit cost', 'potential recruit list' ?
20110529 22:58:01< zaroth> and I wouldn't be able to do that with just having option returning a string via program_options::variable_map
20110529 22:59:07< loonycyborg> zaroth: Another silly suggestion: make them global variables instead of class members :P
20110529 23:00:05< Crab_> zaroth: so, imagine that you have code which needs a '--fps' int command-line parameter,  and you just add a 'command_line_parameter<int> fps_' to your class or 'static command_line_parameter<int> fps' to the top of the cpp file
20110529 23:00:33< Crab_> zaroth: and, internally, it will create the option if it is not yet created, and, when options are parsed, it will store the value and allow the code to quickly get it.
20110529 23:00:41< Crab_> zaroth: like it's with log domains.
20110529 23:00:48< anonymissimus> Crab_: no, I added all the new files
20110529 23:01:11< Crab_> anonymissimus: ok, then it's indeed funny if it links.
20110529 23:01:43< anonymissimus> perhaps cygwin is responsible or some ming install
20110529 23:02:15< loonycyborg> anonymissimus: What do you see in the build log?
20110529 23:02:26< anonymissimus> nothing
20110529 23:02:52< anonymissimus> nothing unusual, that is
20110529 23:03:13-!- samyazza [5b96dd24@gateway/web/freenode/ip.91.150.221.36] has joined #wesnoth-dev
20110529 23:03:30< zaroth> Crab_: augh... can I get my approach working first?
20110529 23:04:23< zaroth> because, you know, it's not that simple, I'm passing some function pointers to validator functions, which are private members of commandline_options
20110529 23:04:41< loonycyborg> anonymissimus: Oh, about program_options. It isn't actually needed by the sourcecode yet.
20110529 23:05:05< zaroth> like for example this --resolution AxB
20110529 23:05:22< loonycyborg> It's only buildsystems that got updated.
20110529 23:05:53< zaroth> loonycyborg: the header is needed, just linking isn't
20110529 23:06:15< Crab_> zaroth: simple enough,  command_line_parameter<resolution> resolution :)) or command_line_parameter<std::string> resolution(pointer_to_validator_in_the_constructor)
20110529 23:06:16< zaroth> and with this approach of registering the validator functions would be spread all over the place
20110529 23:06:37< Crab_> zaroth: no, if you go with ' command_line_parameter<resolution> resolution ' route
20110529 23:06:46< Crab_> zaroth: but, you're welcome to do it your way;
20110529 23:06:51< Crab_> zaroth: just suggesting alternatives
20110529 23:07:19< zaroth> I certainly want to learn how the logging system is currently done one day
20110529 23:07:47-!- erl [~erl@vetinte.spelaroll.se] has joined #wesnoth-dev
20110529 23:07:52< zaroth> but for now, I just want to get this commandline issue out of the way as quick as possible with a reasonably clean approach
20110529 23:08:04< Crab_> yes, that's a good idea.
20110529 23:08:34< Crab_> as a side effect, you'll be more confident in the input data to the game_controller routines, since your code will take part in parsing it
20110529 23:10:05-!- Vorpal [~AnMaster@unaffiliated/anmaster] has quit [Ping timeout: 244 seconds]
20110529 23:10:29< anonymissimus> loonycyborg: that doesnt make any differebce for compiling, does it ?
20110529 23:11:14< loonycyborg> anonymissimus: As zaroth only headers are needed and generally they're available.
20110529 23:11:24< loonycyborg> *As zaroth siad
20110529 23:11:28< loonycyborg> *said
20110529 23:11:46-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20110529 23:11:53< anonymissimus> does that explain why MSVC built didn't complain about missing dll ?
20110529 23:12:07< loonycyborg> Yes.
20110529 23:12:10< anonymissimus> usually that comes during execution
20110529 23:12:45< anonymissimus> aha, hm but shouldnt the mingw linker need a .a file upon linking ?
20110529 23:13:13< loonycyborg> Only if your sourcecode actually references symbols from that .a file.
20110529 23:13:27< zaroth> anonymissimus: yes, .dlls are needed for linking (actually using the library), headers are needed for compiling (which can so happen that includes a header but uses nothing from inside at all, as in this case)
20110529 23:14:43< anonymissimus> well, MSVC needed the lib anyway
20110529 23:16:55< anonymissimus> where can I get .a files for boost in case it'll break ?
20110529 23:17:39-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has quit [Quit: Valkier]
20110529 23:18:42-!- Valkier [~Valkier@c-174-55-104-2.hsd1.pa.comcast.net] has joined #wesnoth-dev
20110529 23:22:23< zaroth> dammit! a wild obscure linking error appears
20110529 23:22:27< loonycyborg> Make them yourself.
20110529 23:22:47 * zaroth goes to bisecting
20110529 23:22:52< zaroth> s/to//
20110529 23:22:57< loonycyborg> Where did you get those that you're currently using?
20110529 23:25:01< loonycyborg> anonymissimus: ^
20110529 23:25:21-!- isaac [~isaac@heal.cauterized.net] has joined #wesnoth-dev
20110529 23:25:21-!- isaac [~isaac@heal.cauterized.net] has quit [Changing host]
20110529 23:25:21-!- isaac [~isaac@debian/developer/isaac] has joined #wesnoth-dev
20110529 23:26:01< zaroth> anyone can make sense of this? http://pastebin.com/8Tm3ikjp I didn't touch neither hotkeys.cpp nor image_modifications.cpp ...
20110529 23:27:01< zaroth> and I commented out all the possible offending new code and the issue stays
20110529 23:27:08< zaroth> so it's probably caused by something I removed
20110529 23:27:35< anonymissimus> loonycyborg: it was a download from the wesnoth forum
20110529 23:28:12< Crab_> zaroth: util.o (from util.cpp) should contain that lexical_cast_default
20110529 23:28:22< anonymissimus> from the CodeBlocks compilation guide on windows and following
20110529 23:28:32< zaroth> Crab_: I didn't touch util.cpp either
20110529 23:28:46< zaroth> I just changed commandline_options.cpp and game_controller.cpp
20110529 23:29:03< loonycyborg> anonymissimus: Either convince the person who make them to build program_options too
20110529 23:29:10< loonycyborg> or build it yourself.
20110529 23:29:17< anonymissimus> it's a supposedly "old" build system which I excpect to break any moment :P
20110529 23:29:30< Crab_> zaroth: have you changed the buildsystem-related files ?
20110529 23:29:55< anonymissimus> loonycyborg: that means building boost with the same compiler ?
20110529 23:30:04< loonycyborg> Yes.
20110529 23:30:19< loonycyborg> ./bootstrap.sh; ./bjam --with-program-options
20110529 23:30:20< zaroth> Crab_: git status shows only game.cpp, game_controller.cpp and commandline_options.?pp
20110529 23:30:23< anonymissimus> _oh_ _no_
20110529 23:30:27< loonycyborg> Should work in cygwin
20110529 23:30:54< anonymissimus> aha hm ok noted thx
20110529 23:31:17< loonycyborg> It was somewhat more complex for me with plain cmd.exe and bootstrap.bat
20110529 23:31:29< Crab_> zaroth: if you git-stash everything, would it link ?
20110529 23:31:38< Crab_> zaroth: if no, then the error appeared earlier
20110529 23:32:06< zaroth> Crab_: yeah, it sure will... I never commit noncompiling stuff
20110529 23:32:17< zaroth> I will probably generate patch and add lines step by step
20110529 23:32:46< Crab_> zaroth: can you show the full diff ?
20110529 23:32:47< loonycyborg> Actually that was --with-program_options
20110529 23:34:43< zaroth> Crab_: http://pastebin.com/5rN0EJ2N
20110529 23:35:04< zaroth> I removed the #include "display.hpp" which got there by accident and the problem stays
20110529 23:36:15< zaroth> I totally should've made smaller commits
20110529 23:36:40< Crab_> zaroth: game.cpp no longer includes any lexical_cast_default<int> calls ?
20110529 23:37:40< zaroth> Crab_: it still has one call
20110529 23:50:58-!- mjs-de [~mjs-de@d119197.adsl.hansenet.de] has quit [Remote host closed the connection]
20110529 23:52:20< CIA-94> anonymissimus * r49704 /trunk/src/game_controller_abstract.cpp: fixed no newline at end of file
20110529 23:52:42< CIA-94> anonymissimus * r49705 /trunk/projectfiles/VC9/wesnoth.vcproj: 
20110529 23:52:42< CIA-94> update MSVC project file
20110529 23:52:42< CIA-94> new files needed:
20110529 23:52:42< CIA-94> libboost_program_options-vc90-mt-gd-1_42.lib
20110529 23:52:42< CIA-94> boost_program_options-vc90-mt-gd-1_42.dll
20110529 23:52:42< CIA-94> can be get from http://www.boostpro.com/download/
20110529 23:56:41-!- zookeeper [~l@wesnoth/developer/zookeeper] has quit []
20110529 23:56:56< Crab_> Nephro: ok, I've taken a look at the patch
20110529 23:57:27< Crab_> Nephro: it is mostly ok. a lot of (small-and-not) things that I don't like, but, mostly, it is the correct approach
20110529 23:58:01< Crab_> Nephro: two ways to go: 1) commit it and then fix the bugs 2) work on it without committing, fixing stuff
--- Log closed Mon May 30 00:00:37 2011