--- Log opened Mon Jan 18 00:00:42 2010
20100118 00:00:44< fendrin> Crab_: copy your trunk directory on your hard disk and make svn switch in the new copy. This is quite fast and does the trick.
20100118 00:00:58< Crab_> fendrin: I use git :)
20100118 00:01:51< fendrin> Crab_: Would like to use git as well but I am scared.
20100118 00:02:02< Crab_> fendrin: so, it'll be quick, but firstly it needs to fetch some revisions from that branch, because I synced only trunk. it's already on 40722, through :)
20100118 00:02:42< Ivanovic> Crab_: any update regarding your visa?
20100118 00:02:55< fendrin> Crab_: Do you need to run any server for git?
20100118 00:03:02< Crab_> Ivanovic: no, nothing yet.
20100118 00:03:06< Ivanovic> mkay
20100118 00:03:18< Ivanovic> just keep in mind that less than 2 weeks are left
20100118 00:03:29< Crab_> Ivanovic: yes, I know.
20100118 00:03:47< Crab_> 'a bit less than 3'
20100118 00:04:04< Ivanovic> right
20100118 00:04:09< Crab_> fendrin: no, just fetch a fairly recent git snapshot of the repository
20100118 00:05:12< fendrin> Crab_: How would that be done?
20100118 00:05:49< Crab_> Someone (Soliton, afair), does them from time to time, they can be downloaded.
20100118 00:06:28< Crab_> fendrin: http://wesnoth.org/files/hidden/git/wesnoth-git.tar.bz2 is the one from jan-11
20100118 00:06:42< Crab_> it's 1.6G
20100118 00:07:03< Ivanovic> Crab_: you know, there is a reason for the *hidden* in the filename
20100118 00:07:12< Ivanovic> the reason is that it is not meant to appear in public logs...
20100118 00:07:42< Crab_> Ivanovic: ok, sorry :) then I think that the ministry of truth will need to take action :)
20100118 00:07:52< Ivanovic> you could just have given him the link in a query...
20100118 00:08:45< fendrin> security through obscurity
20100118 00:08:50< stikonas> t
20100118 00:09:07< Ivanovic> fendrin: the matter is not security but bandwidth
20100118 00:09:41< Ivanovic> if there are too many downloading this tarball it will let our available bandwidth go through the roof...
20100118 00:09:44< stikonas> wasn't that git snapshot meant to be used. what other reason can be to put it there?
20100118 00:10:07< Ivanovic> stikonas: it is meant to be used by those who need it, not by everyone in the ne
20100118 00:10:10< Ivanovic> t
20100118 00:10:52< fendrin> Crab_: It's nicely on the way. Half an hour.
20100118 00:11:57 * stikonas is wondering what part of people here use git-svn import instead of svn repository
20100118 00:12:11< Crab_> fendrin: do you have a stack trace of that segfault ?
20100118 00:12:29< fendrin> Crab_: Yes, I have. Do you want me to pastebin it?
20100118 00:12:33< Crab_> yes
20100118 00:13:42< fendrin> Crab_: No I don't have one. It's inside of eclipse and I can't find a way to get it out there.
20100118 00:17:45< Crab_> fendrin: you can show a screenshot :) ok, compiling :)
20100118 00:19:24< fendrin> Crab_: I know from the debugger that the argument that is given to the constructor of terrain_filter, a vconfig object called cfg, only contains null pointers.
20100118 00:21:17-!- wesbot changed the topic of #wesnoth-dev to: released 1.7.12, announcing "soon" | string/feature freeze active! | 57 bugs, 245 feature requests, 7 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100118 00:26:21< CIA-53> fendrin * r40737 /branches/fendrin_pathfind/data/scenario-test.cfg: Renamed the subtags of tunnel.
20100118 00:26:45< fendrin> Crab_: ^ wesnoth -t will make it crash.
20100118 00:27:05< Crab_> fendrin: ok, thanks. as soon as the compile finishes, I'll check it out
20100118 00:27:08< fendrin> Crab_: How do you benchmark the ai? I would like to know if the new pathfinder makes the ai slower.
20100118 00:28:04< Crab_> fendrin: see /utils/ai_test
20100118 00:28:44< Crab_> in short: setup postgresql db and a webserver, modify config file, launch python script
20100118 00:29:18< Crab_> I can do a 'setup postgresql db and a webserver' part for you, and PM you a config file with username/password entered
20100118 00:29:43< Crab_> or, you can do it locally. then, you'll be able to launch ai-vs-ai batch testing and see the results at a url.
20100118 00:29:59< Crab_> one of the stats that is collected, is 'total time per turn'.
20100118 00:30:36< Crab_> if you change the pathfinding without changing the ai, this stat will quite accurately tell the impact of your changes.
20100118 00:30:37< fendrin> average time of pathfinding would be more usefull :-)
20100118 00:31:38< fendrin> Crab_: The newest change includes the execution of a suf and two slf every time the pathfinder is called.
20100118 00:31:54< Crab_> then you can either hack this in the script, or use alternate way to benchmark your changes - e.g. a profiler, such as google performance tools
20100118 00:32:12< Crab_> it can be used to collect only 'pathfinding' part of data
20100118 00:32:48< Crab_> and generate a graph like ftp://ftp.terraninfo.net/wesnoth/ai_2009may16_r35649_bs600.jpg
20100118 00:33:10< Crab_> or, there's a third quick&dirty way 
20100118 00:34:21< Crab_> take advantage of the fact that pathfinding is a noticable part of the ai turn and just measure 'turn time' on something like NR:Showdown with lots of units.
20100118 00:35:50< Crab_> --log-debug=ai/manager will show you total time for ai side turn, measured by using SDL_GetTicks(); at end and at start of turn.
20100118 00:37:29< fendrin> Crab_: My hope is that doing this filter once per call to the pathfinder is not costly. I don't want to mess with caching things.
20100118 00:50:43< AI0867> fendrin:
20100118 00:50:45< AI0867> +               teleport_pair locations;
20100118 00:50:45< AI0867> +
20100118 00:50:45< AI0867> +               it->get_teleport_pair(locations, u);
20100118 00:50:45< AI0867> +
20100118 00:50:47< AI0867> +               sources_.push_back(locations.first);
20100118 00:50:50< AI0867> +               destinations_.push_back(locations.second);
20100118 00:50:52< AI0867> there's something wrong here
20100118 00:51:05< AI0867> teleport_map constructor
20100118 00:51:18< AI0867> the teleport_pair is not initialized with anything, and then used
20100118 00:54:18< fendrin> AI0867: teleport_pair locations(); doesn't seem to work. Is it teleport_pair locations = teleport_pair(); ?
20100118 00:55:01< AI0867> it's initialized, but there's nothing in there, so why are you pushing it back?
20100118 00:55:23< AI0867> I think what you mean is: teleport_pair locations = it->get_teleport_pair(locations, u);
20100118 00:55:26< AI0867> or something like that
20100118 00:55:46< loonycyborg> Maybe get_teleport_pair is supposed to fill it..
20100118 00:55:47< AI0867> you're pushing default-constructed pairs in a loop
20100118 00:55:59< fendrin> AI0867: The get_teleport_pair fills the teleport_pair that is the first argument.
20100118 00:56:12< AI0867> ah
20100118 00:56:19< fendrin> AI0867: The method shouldn't be named "get".
20100118 00:56:34< fendrin> That leads to confusion.
20100118 00:57:03< loonycyborg> Why not return this pair by value?
20100118 00:57:55< AI0867>         source_filter_.get_locations(loc_pair.first);
20100118 00:57:55< AI0867>         target_filter_.get_locations(loc_pair.second);
20100118 00:58:00< AI0867> probably because of that
20100118 00:58:02< fendrin> loonycyborg: Because the slf system returns by reference arguments. It's so easy to pass that behaviour through. And it will be fast.
20100118 01:01:16< loonycyborg> fendrin: Anyway, I've built and got bt. Debugging suggests that your crash is caused by zero being passed as cfg arg to terrain_filter's constructor.
20100118 01:01:51< fendrin> loonycyborg: Right, that is what I have found as well. But why is it happenend?
20100118 01:01:59< fendrin> ing
20100118 01:02:09-!- SonIcco [n=SonIcco@pD9512216.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20100118 01:02:37< loonycyborg> cfg.get_children returns something invalid AFAICT.
20100118 01:03:01< loonycyborg> In teleport_group's constructor that is.
20100118 01:04:17-!- Skystriker [n=croseliu@pool-71-250-26-120.nwrknj.east.verizon.net] has quit [Read error: 60 (Operation timed out)]
20100118 01:08:05-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["ChatZilla 0.9.86 [Firefox 3.5.7/20091221164558]"]
20100118 01:10:50< Crab_> fendrin: works without crashes for me. what I'm doing wrong ?
20100118 01:11:26< fendrin> Crab_: Update to the latest version and start wesnoth with -t
20100118 01:12:06< Crab_> teleport worked ok
20100118 01:12:40< Crab_> ah, maybe it's not the latest version...
20100118 01:13:56< Crab_> yes, I'll need to recompile again (it hadn't picked r40736 on the first try)
20100118 01:15:28< fendrin> Crab_: Some older revisions do work fine, but they weren't prepared for persistence. And did not support filtering.
20100118 01:15:36< Crab_> ok
20100118 01:16:00< Crab_> I guess that the config goes out of scope after event handler returns..
20100118 01:16:34< Crab_> but, the segfault is in the teleport_group constructor, yes ?
20100118 01:17:05< fendrin> Right.
20100118 01:26:09-!- Tallken [n=f2f93bf5@93.102.35.9.rev.optimus.pt] has quit ["Leaving"]
20100118 01:29:24< Crab_> fendrin: (still compiling). in  teleport_group::teleport_group(vconfig cfg) ,  you make a copy of the original vconfig that you've got in the  WML_HANDLER_FUNCTION(tunnel...   is that copying necessary ?
20100118 01:30:07< fendrin> Crab_: I guess not. It is an attempt to fix the bug somehow.
20100118 01:31:53-!- Tallken [n=f2f93bf5@93.102.73.35.rev.optimus.pt] has joined #wesnoth-dev
20100118 01:37:51< loonycyborg> fendrin: It's definitely caused by one of cfg.get_children returning an empty container.
20100118 01:38:43< fendrin> loonycyborg: Hmmm, so it's only wrong wml code?
20100118 01:39:16< loonycyborg> Perhaps.
20100118 01:39:33< fendrin> loonycyborg: If the container is empty, what does first() return?
20100118 01:40:37< fendrin> s/first()/front()
20100118 01:41:31< loonycyborg> It's undefined behavior.
20100118 01:50:42-!- Chusslove [n=Chusslov@brsg-d9bef703.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20100118 01:52:12-!- Blueblaze [n=nick@99.158.47.180] has quit [Read error: 110 (Connection timed out)]
20100118 01:52:24< Crab_> fendrin: sorry, some problems while compiling. if the bug still survives till next evening, i'll look into it.
20100118 01:52:47< fendrin> Crab_: Thank you.
20100118 01:52:54-!- Crab_ [n=Crab_@wesnoth/developer/crab] has quit ["Leaving."]
20100118 01:57:16-!- Chusslove [n=Chusslov@brsg-d9bee921.pool.mediaWays.net] has joined #wesnoth-dev
20100118 02:01:04-!- fmunoz [n=chatzill@201-0-203-222.dial-up.telesp.net.br] has quit [Read error: 110 (Connection timed out)]
20100118 02:01:39-!- Skystriker [n=croseliu@pool-71-250-26-120.nwrknj.east.verizon.net] has joined #wesnoth-dev
20100118 02:11:01-!- Tallken [n=f2f93bf5@93.102.73.35.rev.optimus.pt] has quit ["Leaving"]
20100118 02:12:37-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20100118 02:22:59-!- Blueblaze [n=nick@adsl-99-158-47-180.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100118 02:25:59-!- esr [n=chatzill@static-71-162-243-5.phlapa.fios.verizon.net] has quit [Remote closed the connection]
20100118 02:28:09-!- umcreg [n=umcreg@190.22.100.179] has joined #wesnoth-dev
20100118 02:28:25-!- happygrue [n=George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20100118 02:29:46-!- umcreg [n=umcreg@190.22.100.179] has quit [Client Quit]
20100118 02:30:50-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20100118 02:31:07-!- Netsplit leguin.freenode.net <-> irc.freenode.net quits: loonybot, knotwork, Blueblaze, lukjad007, deekay, Zarel
20100118 02:35:46-!- Netsplit over, joins: lukjad007
20100118 02:44:57< AI0867> fendrin: what was the issue again?
20100118 02:45:09< AI0867> also, the silver mage's village-teleport isn't working
20100118 02:46:32< AI0867> wesnoth -t doesn't crash here
20100118 02:47:32-!- Blueblaze [n=nick@adsl-99-158-47-180.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100118 02:53:18-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20100118 02:53:18-!- deekay [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20100118 02:53:18-!- knotwork [n=markm@142.177.178.19] has joined #wesnoth-dev
20100118 03:05:52-!- deekay [n=dk@wesnoth/developer/dragonking] has quit []
20100118 03:10:39-!- shadowm_laptop [n=ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20100118 03:14:27-!- esr [n=chatzill@static-71-162-243-5.phlapa.fios.verizon.net] has joined #wesnoth-dev
20100118 03:15:22< fendrin> AI0867: wesnoth -t doesn't crash?
20100118 03:16:00-!- shadowm_laptop [n=ignacio@wesnoth/developer/shadowmaster] has quit ["foo quit message"]
20100118 03:17:16< fendrin> AI0867: Let me check...
20100118 03:19:41< CIA-53> ai0867 * r40738 /branches/fendrin_pathfind/src/ (game_events.cpp pathfind.cpp): Prevent crashes by incomplete [tunnel] tags
20100118 03:19:43< AI0867> I can get it to crash by renaming some of the tags in [tunnel]
20100118 03:19:46< AI0867> but I just fixed that
20100118 03:20:05< fendrin> AI0867: Thank you. 
20100118 03:20:19< AI0867> more importantly, the teleport ability no longer works
20100118 03:20:39< fendrin> AI0867: The village teleport is deactivaded since zookeeper wrote me a implementation in wml. 
20100118 03:22:05< fendrin> AI0867: http://wesnoth.pastebin.com/m6bf43191
20100118 03:27:31< AI0867> the WML doesn't work yet, as $unit isn't set yet
20100118 03:28:21< Skystriker> Are OOS errors on Dark Forecast on dev a known bug?
20100118 03:29:40< fendrin> AI0867: Ah yes. I would like to set $unit only for the location filters. 
20100118 03:31:05< AI0867> additionally, it'd require delayed substitution, as $unit.side is now evaluated at prestart, not when the filter is checked
20100118 03:31:15< fendrin> right
20100118 03:36:28< AI0867> and even when setting owner_side=1, it still doesn't work...
20100118 03:38:02-!- jee [n=jee@117.204.0.146] has joined #wesnoth-dev
20100118 03:50:30< fendrin> AI0867: Strange, the slf that contain only x and y pairs work well, the heavier ones fail.
20100118 03:55:13< Skystriker> I wish you could change your color ingame :P
20100118 03:56:53< Skystriker> If there is OOS errors, is it more likely the game version, or the addon?
20100118 03:57:05< Skystriker> Both dark forecast and Aetheryn's maps are giving me and my friends OOS errors
20100118 04:05:48-!- Ivanovic_ [n=ivanovic@dtmd-4db2cbb8.pool.mediaWays.net] has joined #wesnoth-dev
20100118 04:21:56-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 113 (No route to host)]
20100118 04:58:45< AI0867> fendrin: the issue is the [filter][not][/not][/filter] in the [target]
20100118 05:00:30< AI0867> removing it doesn't make the silver mage magically teleport into the unit in the target village, it stops as if its destination is blocked
20100118 05:01:07< AI0867> but pathfinding ignores this blockage, giving a wrong impression of where the unit can move
20100118 05:02:58< fendrin> AI0867: That was my thought as well. But removing the filter didn't work for me. 
20100118 05:05:52< AI0867> http://wesnoth.pastebin.com/d30cfe81f
20100118 05:05:57< AI0867> that's what my tunnel looks like
20100118 05:06:07< AI0867> I dumped it in with the other tunnels
20100118 05:06:33-!- noy [n=Noy@wesnoth/developer/noy] has quit [Read error: 104 (Connection reset by peer)]
20100118 05:08:58< fendrin> AI0867: I pasted it into the unit_type tag.
20100118 05:09:08-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100118 05:11:44-!- jee [n=jee@117.204.0.146] has left #wesnoth-dev []
20100118 05:12:03< fendrin> AI0867: It is working. 
20100118 05:20:19-!- dtiger [n=dtiger@dynamic-vpdn-93-125-68-75.telecom.by] has joined #wesnoth-dev
20100118 05:22:31-!- VM___ [n=emanuel@80.73.182.163] has quit [Client Quit]
20100118 05:26:07< AI0867> without the [target] [filter]?
20100118 05:26:15< fendrin> AI0867: yes
20100118 05:26:54< fendrin> AI0867: This is a wml coding problem. The filter won't work in other contexts neither.
20100118 06:10:22-!- Blueblaze is now known as Jaredu
20100118 06:10:27-!- Jaredu is now known as Blueblaze
20100118 06:11:09-!- dtiger [n=dtiger@dynamic-vpdn-93-125-68-75.telecom.by] has quit [Remote closed the connection]
20100118 06:19:08-!- Skystriker [n=croseliu@pool-71-250-26-120.nwrknj.east.verizon.net] has quit [Read error: 110 (Connection timed out)]
20100118 06:21:16-!- wesbot changed the topic of #wesnoth-dev to: released 1.7.12, announcing "soon" | string/feature freeze active! | 58 bugs, 245 feature requests, 7 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100118 06:35:22-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20100118 07:05:45< CIA-53> jetryl * r40739 /trunk/data/core/images/items/ (altar-evil.png altar.png key.png): Added new key image, and updated the altar images.
20100118 07:42:01-!- Blueblaze [n=nick@adsl-99-158-47-180.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20100118 07:50:37-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20100118 08:01:18-!- Blueblaze [n=nick@adsl-99-158-47-180.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100118 08:15:22-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20100118 08:16:53-!- fendrin [n=fabi@wesnoth/developer/fendrin] has quit [Remote closed the connection]
20100118 08:21:27-!- fendrin [n=fabi@88-134-75-97-dynip.superkabel.de] has joined #wesnoth-dev
20100118 09:27:14-!- Blueblaze [n=nick@adsl-99-158-47-180.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20100118 09:33:20-!- deekay [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20100118 10:03:15-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has quit [Read error: 60 (Operation timed out)]
20100118 10:05:40-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has joined #wesnoth-dev
20100118 10:10:50-!- tracktest [n=dtiger@mm-85-206-84-93.dynamic.pppoe.mgts.by] has joined #wesnoth-dev
20100118 10:11:00-!- tracktest [n=dtiger@mm-85-206-84-93.dynamic.pppoe.mgts.by] has quit ["Konversation terminated!"]
20100118 10:37:57-!- noy [n=Noy@wesnoth/developer/noy] has quit [Read error: 60 (Operation timed out)]
20100118 10:38:13-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100118 10:56:22-!- Ivanovic_ is now known as Ivanovic
20100118 10:56:47< Ivanovic> moin
20100118 11:08:12-!- loonybot [n=loonybot@ppp79-139-137-149.pppoe.spdop.ru] has joined #wesnoth-dev
20100118 11:09:02-!- loonycyborg [n=sergey@ppp79-139-137-149.pppoe.spdop.ru] has joined #wesnoth-dev
20100118 11:12:41-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20100118 11:48:48-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has joined #wesnoth-dev
20100118 12:41:11-!- mordante [n=mordante@87.215.201.26] has joined #wesnoth-dev
20100118 12:41:13< mordante> servus
20100118 12:42:12< mordante> Ivanovic, just tested here with SDL-1.2.14-3 and both Wesnoth 1.6.5 and trunk break on fvwm
20100118 12:42:22< Ivanovic> mordante: okay, good to know
20100118 12:42:29< mordante> I only get mouse up events and no mouse down events
20100118 12:42:38< Ivanovic> mordante: looks like at least 1.7.x works here on kde using this libsdl version
20100118 12:42:47< Ivanovic> though i get *millions* or "missed event" messages
20100118 12:43:32< Ivanovic> 20100118 12:43:24 warning gui/event: distributor mouse button left [wml_message_left]: SDL left button up. The mouse button is already up, we missed an event.
20100118 12:43:35< mordante> what happens when you start wesnoth with --new-widgets and press the tip of the day buttons also a lot of these messages?
20100118 12:44:04< Smar> I managed to hang up whole wesnoth by missed event... by pressing theme changing button while at menu with 1.7
20100118 12:44:14< mordante> (not sure whether I wired in the effect they should have, but can't test here at the moment)
20100118 12:44:26< Smar> otherwise atleast menu worked. 
20100118 12:44:27< Ivanovic> interesting, no such messages *at all* in the main menu
20100118 12:44:37< Ivanovic> the messages only start once in a game (eg two brothers)
20100118 12:44:42< mordante> Smar, that sounds like some other bug since that uses the old gui engine
20100118 12:44:46< Ivanovic> there i eg get it for every click in a dialog
20100118 12:45:03< Ivanovic> that is: no matter of using the old or new widgets, i get zero messages in the menu itself
20100118 12:45:07< Smar> mordante: I supposed so... I’ll wait for RC and look if I’m enough unlazy to file a bug
20100118 12:45:11< mordante> Ivanovic, you started with --new-widgets ?
20100118 12:45:17< Ivanovic> yes
20100118 12:45:30< Ivanovic> once without param, no messages in the main screen and things working
20100118 12:45:35< mordante> Smar, best file a bug now and not wait until we think we're almost ready for 1.8
20100118 12:45:41< Ivanovic> once with param, no messages in the main screen and things working
20100118 12:45:48< Smar> but I’m lazy :(
20100118 12:45:56< Ivanovic> though always *many* messages ingame, messages like the one mentioned above
20100118 12:46:18< Smar> 01.52.13 < Smar> 20100118 01:52:40 warning gui/event: Can't execute timer since it no longer exists.
20100118 12:46:21< Smar> 01.52.13 < Smar> 20100118 01:53:45 warning gui/event: distributor mouse button left [lobby_main]: SDL left button up. The mouse button is already up, we missed an event.
20100118 12:46:50< Smar> can’t do anything more atleast in 6 hours. 
20100118 12:47:21< mordante> Ivanovic, odd will test that when I'm home those message can occur every now and then but it shouldn't happen too often
20100118 12:47:34< mordante> they're mostly harmless ;-)
20100118 12:47:34< Ivanovic> mordante: i get them for every click over here
20100118 12:47:52< Ivanovic> every click that does something to a dialog (as in "next dialog please")
20100118 12:48:09< Ivanovic> using kde, wesnoth windowed and libsdl 1.2.14
20100118 12:48:19< Ivanovic> kde 4.3.4 that is
20100118 12:48:31< mordante> I'll test at home with 1.2.13 and fvwm see whether I can reproduce it
20100118 12:49:16< mordante> Smar, that first message looks somewhat suspicious the second harmless
20100118 12:49:39< mordante> Smar, but still best file a bug report so I won't forget about it
20100118 12:50:29< Smar> I’ll test it a bit more when I’m at home
20100118 12:50:36< mordante> thanks
20100118 12:50:54< mordante> teaser, sunset darkens the screen except for the redrawn areas, so it's a debug tool
20100118 13:31:59-!- dtiger [n=dtiger@mm-85-206-84-93.dynamic.pppoe.mgts.by] has joined #wesnoth-dev
20100118 13:32:12-!- dtiger [n=dtiger@mm-85-206-84-93.dynamic.pppoe.mgts.by] has quit [Client Quit]
20100118 13:56:09-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has quit [No route to host]
20100118 13:57:55< Ivanovic> mordante: so you will submit a debian bugreport now?
20100118 13:58:17< Ivanovic> that is: linking it somehow to the one in our tracker marked upstream and also linking in our tracker to the debian report
20100118 13:58:27< Ivanovic> (so that we have some "central reference" for the problem)
20100118 14:27:35< mordante> Ivanovic, yes I'll do that this evening
20100118 14:27:46< Ivanovic> okay
20100118 15:37:59-!- rosso [n=rosso@dslb-088-070-049-153.pools.arcor-ip.net] has quit [Remote closed the connection]
20100118 16:00:19-!- crimson_p [n=irchon@64.201.60.211] has joined #wesnoth-dev
20100118 16:01:01< crimson_p> just popping in to say, I got the email yesterday, but then completely forgot about it!
20100118 16:01:39< crimson_p> sorry. I'll upload a binary a bit later today though, once I actually get up
20100118 16:02:09-!- crimson_p [n=irchon@64.201.60.211] has quit [Remote closed the connection]
20100118 16:05:59< Ivanovic> mkay
20100118 16:06:23< Ivanovic> please ping me once it is online so that i can announce then
20100118 16:06:59< loonycyborg> Ivanovic: Too late. He popped out :P
20100118 16:07:06< Ivanovic> argh, right
20100118 16:08:17-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20100118 16:10:17< Ivanovic> crab_: IIRC you were the last one working on unit stuff, please have a look at this one: https://gna.org/bugs/index.php?15146
20100118 16:10:45< Ivanovic> and yeah, the backtrace does not help at all since it lacks debug symbols, but maybe the assert helps
20100118 16:10:59-!- EdB [n=edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has joined #wesnoth-dev
20100118 16:30:21-!- SonIcco [n=SonIcco@pD9510E43.dip0.t-ipconnect.de] has joined #wesnoth-dev
20100118 16:38:26< mordante> Ivanovic, Crab_ seems the reporter used a self-compiled binary, maybe ask to compile with debug symbols
20100118 16:46:06-!- EdB [n=edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has quit [Remote closed the connection]
20100118 16:47:43< Ivanovic> okay, the new po4a version (that should have the fix in for the generation of our manuals) is out and in gentoo
20100118 16:48:04< Ivanovic> lets see what happens when trying to generate the manual using it (and how many fuzzy/untranslated strings this will generate)
20100118 16:57:49< CIA-53> ivanovic * r40740 /trunk/doc/manual/manual.fi.html: regenerated doc files (not using po4a 0.38, first working version since 0.34!)
20100118 17:03:15< Ivanovic> okay, lovely
20100118 17:03:26< Ivanovic> of *course* there are many changes, it is just that the autotools recipe sucks
20100118 17:03:52< Ivanovic> manually removing the html and xml files before running "make update-po4a" results in *many* changed html files as well as changed po files
20100118 17:09:36< CIA-53> ivanovic * r40741 /trunk/ (60 files in 2 dirs): 
20100118 17:09:38< CIA-53> argh, autotools update recipe is far from perfect, it seems...
20100118 17:09:38< CIA-53> the result of the update is that *many* strings changed, though the benefit is that now some additional things should work nicely (eg alternate entries for images when the images are not found/loaded)
20100118 17:09:38< CIA-53> due to this change 26 fuzzy and 12 untranslated strings were added to po/wesnoth-manual/*.po, most of those should be "done in some mins" strings, so not too much of an issue (will notify people on the i18n ml)
20100118 17:09:39< CIA-53> due to the changes the following manuals were not complete enough anymore for generation: Danish, Galician and Traditional Chinese
20100118 17:11:45-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20100118 17:16:02-!- mordante [n=mordante@87.215.201.26] has quit ["Leaving"]
20100118 17:23:53-!- MikeJB [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20100118 17:24:04-!- MikeJB [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Client Quit]
20100118 17:25:18-!- MikeJB [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20100118 17:30:17< Ivanovic> esr: got a problem with utbs and pofix.py
20100118 17:30:34< Ivanovic> esr: could you make pofix ignore all lines starting with #
20100118 17:30:35< Ivanovic> ?
20100118 17:30:53< crimson_penguin> Ivanovic: I'm just building now
20100118 17:30:54< Ivanovic> in the utbs file there is this "string comment" included: #. [message]: speaker=$dwarf_name
20100118 17:31:08< Ivanovic> the problem is that we also have a replacement rule for $dwarf_name
20100118 17:31:10< MikeJB> Ivanovic: Any (extremely rough) estimates on the release of 1.8? RL has gotten in the way of the Latin translation a bit lately, but I do have translations I haven't been able to review yet I'd like to send in before 1.8's release.
20100118 17:31:16 * crimson_penguin can't believe he used part of the Wesnoth XCode project as reference yesterday (or tried to), yet completely forgot about the release
20100118 17:31:28< Ivanovic> MikeJB: cf wesnoth-i18n mailling list
20100118 17:32:57< MikeJB> Ivanovic: Thanks. Just checked the website. Apparently my email hasn't received the latest mail yet.
20100118 17:33:29< Ivanovic> i sent it 15mins ago, looks like gna is a little slow again with sending out mails...
20100118 17:36:00< crimson_penguin> Ivanovic: when did you wanna announce?
20100118 17:36:08< Ivanovic> crimson_penguin: when the mac binary is online
20100118 17:36:10< Ivanovic> ;)
20100118 17:36:28< crimson_penguin> I might be able to use our faster internet to get it up faster
20100118 17:37:52< CIA-53> ivanovic * r40742 /trunk/ (4 files in 3 dirs): updated German translation
20100118 17:43:08< crimson_penguin> Wesnoth has a lot of files!
20100118 17:43:21< crimson_penguin> I'm 2/3 done compiling
20100118 17:46:22< Ivanovic> :)
20100118 17:47:55< CIA-53> ivanovic * r40743 /trunk/src/gui/auxiliary/widget_definition/: set an svn:ignore entry
20100118 17:51:40-!- Zarel_ [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20100118 17:56:28< crimson_penguin> building disk image...
20100118 17:58:45-!- MikeJB [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Read error: 110 (Connection timed out)]
20100118 18:02:10< zookeeper> wouldn't a better way to auto-prune off-topic be to auto-lock the threads instead of deleting them?
20100118 18:03:27< esr> Ivanovic: Yes, I can do that.
20100118 18:03:33< Ivanovic> esr: would be great!
20100118 18:03:40< Ivanovic> zookeeper: no idea if this is possible
20100118 18:07:14-!- Sapient [n=patrickp@wesnoth/developer/sapient] has joined #wesnoth-dev
20100118 18:10:42-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has joined #wesnoth-dev
20100118 18:11:38< crimson_penguin> all ready to upload, now I just have to get it to my brother's computer somehow
20100118 18:21:16-!- wesbot changed the topic of #wesnoth-dev to: released 1.7.12, announcing "soon" | string/feature freeze active! | 59 bugs, 244 feature requests, 7 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100118 18:26:10-!- GNUtoo [n=GNUtoo@host82-133-dynamic.54-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20100118 18:26:34< GNUtoo> hi,on the freerunner _initial.cfg if I put 0 0 as resolution it works better
20100118 18:26:49< GNUtoo> at least it shows the multiplayer gui
20100118 18:27:13< GNUtoo> * in _initial.cfg
20100118 18:31:50-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20100118 18:32:38-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has quit [Read error: 60 (Operation timed out)]
20100118 18:41:37-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has joined #wesnoth-dev
20100118 18:44:49< crimson_penguin> Ivanovic: it's uploading, but I'm not sure if it's really gonna go faster - right now it's only going at 13KB/s, whereas the usual internet goes at 20, but then before it was going at 50-70, so we'll see
20100118 18:45:01< crimson_penguin> if I check on it in a bit and it's still going slow I'll switch to the other internet
20100118 18:55:20-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100118 19:00:17-!- Noyga [n=noyga@wesnoth/developer/noyga] has joined #wesnoth-dev
20100118 19:02:05-!- EdB [n=edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has joined #wesnoth-dev
20100118 19:13:15-!- Crab_ [n=Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20100118 19:14:20-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100118 19:14:39-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Client Quit]
20100118 19:14:53-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100118 19:15:01-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Client Quit]
20100118 19:15:15-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100118 19:15:25-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Client Quit]
20100118 19:15:37-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100118 19:15:45-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Client Quit]
20100118 19:16:01-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100118 19:16:42-!- stikonas [n=and@bcm-131-111-247-5.girton.cam.ac.uk] has joined #wesnoth-dev
20100118 19:19:53< zookeeper> wtf? http://www.wesnoth.org/forum/viewtopic.php?f=11&t=28571
20100118 19:20:31< zookeeper> i think i'll just...delete that. it's not worth locking.
20100118 19:20:42< Blarumyrran> Very Zen
20100118 19:25:40-!- rosso [n=rosso@dslb-088-070-049-153.pools.arcor-ip.net] has joined #wesnoth-dev
20100118 19:32:00< boucman> hey all
20100118 19:32:13< boucman> Crab_: another one http://www.wesnoth.org/forum/viewtopic.php?p=404471#p404471
20100118 19:32:27< Crab_> hi, boucman
20100118 19:33:06< Ivanovic> Crab_: left a message for you in the logs regarding a bug report
20100118 19:34:00< Crab_> Ivanovic: ok
20100118 19:35:29 * Crab_ wonders about the feasibility of developing a lua ai in a branch..
20100118 19:36:33-!- rosso [n=rosso@dslb-088-070-049-153.pools.arcor-ip.net] has quit [Remote closed the connection]
20100118 19:37:58< boucman> hmm
20100118 19:38:17< boucman> well, you'd need to discuss it with silene, but that's a worthy thing to do for 1.9
20100118 19:38:57< Crab_> boucman: well, it's definitely in plans for 1.9 , since I expect that it'll be easier to use than formula_ai
20100118 19:39:20< boucman> Crab_: how are we wrt 1.8 on the AI front ?
20100118 19:39:21< Crab_> boucman: but, I was thinking about proposing this idea to the poster in the post you've mentioned :)
20100118 19:39:28< boucman> did we switch to the new one yet ?
20100118 19:39:35< boucman> hehe
20100118 19:40:14< Crab_> boucman: since making a 'beta' of lua-based candidate actions / stages is almost an easycoding task now :)
20100118 19:40:29< boucman> hehe
20100118 19:40:47< Crab_> but, 'how to do it cleanly' will need discussions with silene.
20100118 19:40:55< boucman> he seemed to want to do c++, but i think that it's a bad idea to encourage c++ AIs at this point
20100118 19:41:07< boucman> since it's too easy to "break" the AI frontiere in C++
20100118 19:41:19< Crab_> well, 'make it possible to write ai in lua' is a c++ coding task :)
20100118 19:41:46< boucman> Crab_: i meant "writing an AI in c++"
20100118 19:42:01< Crab_> ok, yes, there are other interesting pure c++ things that are needed, too :)
20100118 19:42:23< Crab_> such as 'intelligent recruiting with multiple leaders'
20100118 19:42:35< boucman> Crab_: hmm
20100118 19:42:51< boucman> i don't see why that one would be c++
20100118 19:43:24< Crab_> boucman: why not ? it can be done in c++, basing on existing implementation, and it's interesting thing to do.
20100118 19:43:26< Crab_> about, 'how are we wrt 1.8 on the AI front ?', will answer in a few minutes, somewhat busy atm
20100118 19:43:35< boucman> k
20100118 19:44:14< boucman> well, do we want to encourage C++ AIs ? i thought we wanted to slowly move to more logic in WML/FAI/LUA to decrease the entry barrier...
20100118 19:45:10< Crab_> well, my plan was to push the 'core' of the ai towards being 'language-independent'
20100118 19:45:29< boucman> ok
20100118 19:45:40< Crab_> so, it works with 'things written in any language hidden behind a c++ interface'
20100118 19:45:52< Crab_> 'any language' can be c++, too, there's nothing wrong with that :)
20100118 19:46:30< Crab_> as we can mix-in parts written in different languages
20100118 19:46:36< boucman> ok, so I misunderstood your original statement, I thought you meant "it has to be done in c++"
20100118 19:47:15< Crab_> it all depends on the availability of interested coders
20100118 19:47:27< boucman> true
20100118 19:47:32< Crab_> if we've got someone who's willing to code good things in c++, then let it be c++ :)
20100118 19:47:51< Crab_> and that's why you're right, 'decreasing the entry barrier' is very important.
20100118 19:47:59< Crab_> I hope that lua will help there.
20100118 19:48:25< Ivanovic> to me it sounds good if the ai is modular, so that you can easily stick together parts
20100118 19:48:33< Ivanovic> no matter if they are c++, lua or fai
20100118 19:48:47< Crab_> Ivanovic: yes, it is modular :)
20100118 19:49:03< Crab_> Ivanovic: and you can stick together parts written in different languages
20100118 19:49:20< Crab_> c++ and fai, for now :)
20100118 19:49:22< Ivanovic> if that is easily possible, why not also write stuff in c++, as long as someone is willing to do so
20100118 19:49:32< Crab_> yes, exactly
20100118 19:49:39< Ivanovic> and yeah, making the api accessible to lua sounds perfectly fine, too
20100118 19:50:24-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20100118 19:50:29< silene> hi
20100118 19:50:34< Ivanovic> hi silene
20100118 19:50:36< Crab_> hi silene
20100118 19:50:58< silene> (i was just ringed about the lua discussion)
20100118 19:51:03< Ivanovic> :)
20100118 19:51:37< Crab_> silene: yes, we're hoping that lua will be a better way to write ai components than formula_ai is :)
20100118 19:51:40< silene> i don't envision any difficulty, the harder thing will probably to split the lua code into several files, since it will be going to grow way to big
20100118 19:52:14< silene> the c++ interface to lua code, i mean
20100118 19:53:13< Crab_> silene: I can write a small description  of the things that are needed to be accessible to lua code from 'ai' side , for you to comment and say what will be the best way to handle them. is this ok?
20100118 19:53:30< silene> anyway, i'm going offline again; my lecture for tomorrow morning is far from finished and i probably won't be home before midnight, argl...
20100118 19:53:43< silene> Crab_: yes, please do so
20100118 19:54:00< Crab_> ok, I'll try to find time for that today and leave a link in the logs.
20100118 19:55:11-!- silene [n=plouf@wesnoth/developer/silene] has quit ["back to trying to explain smt solvers"]
20100118 19:58:45-!- AnMaster_ [n=AnMaster@unaffiliated/anmaster] has joined #wesnoth-dev
20100118 20:03:41-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has quit [Success]
20100118 20:04:06-!- AnMaster_ [n=AnMaster@unaffiliated/anmaster] has quit [Read error: 60 (Operation timed out)]
20100118 20:05:56-!- mjs-de [n=mjs-de@wh.uni-dortmund.de] has joined #wesnoth-dev
20100118 20:12:48< crimson_penguin> Ivanovic: last I checked, it was still going at only 15 or so :-/
20100118 20:12:54< crimson_penguin> maybe I should just switch it over to this internet... but then it'll slow down my net ;P
20100118 20:12:55< Ivanovic> :(
20100118 20:14:07-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20100118 20:14:25< mordante> servus
20100118 20:16:09-!- AnMaster [n=AnMaster@unaffiliated/anmaster] has joined #wesnoth-dev
20100118 20:17:16< boucman> thespaceinvader: what happens when a drake fights over a chasm, will it land ?
20100118 20:18:05< thespaceinvader> it'll have to
20100118 20:18:18< thespaceinvader> that's a won't fix graphical bug at present
20100118 20:19:01< mordante> GNUtoo, which resolution did you change in _initial.cfg? (I assume TINY)
20100118 20:19:19< boucman> thespaceinvader: i expected so... that's fine
20100118 20:19:26< thespaceinvader> no worries
20100118 20:19:47< boucman> hopefully people will provide art and we'll just have to add it in
20100118 20:19:51< boucman> one can dream :)
20100118 20:20:13< thespaceinvader> well yeah
20100118 20:20:31< thespaceinvader> but it's such a niche that it is simply not worth the time to be honest
20100118 20:20:40< thespaceinvader> i'd love for there to be in-flight attacks, but... 
20100118 20:20:43< thespaceinvader> SO much work
20100118 20:21:02< boucman> indeed
20100118 20:21:13< boucman> that's why we had them landing in water in the first place
20100118 20:27:03-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20100118 20:32:35-!- Skystriker [n=croseliu@pool-71-250-26-120.nwrknj.east.verizon.net] has joined #wesnoth-dev
20100118 20:34:01-!- Sapient [n=patrickp@wesnoth/developer/sapient] has left #wesnoth-dev []
20100118 20:42:33< GNUtoo> mordante, I don't remember but I put 0x0 as resolution
20100118 20:42:50< GNUtoo> and after the OS update wesnoth is much faster
20100118 20:43:00< GNUtoo> but there are some OOM problems
20100118 20:43:21< GNUtoo> even with -DLOW_MEM or something like that
20100118 20:44:55< mordante> GNUtoo, can you post your _initial.cfg?
20100118 20:45:08< fendrin> Who knows about the usage of the SUF? What is the location argument for? The unit knows it's location on the map.
20100118 20:45:10< GNUtoo> ok 
20100118 20:45:26< mordante> maybe we really should send Wesnoth to weight watchers ;-)
20100118 20:46:09< boucman> fendrin: let me have look i've used it in the animation engine
20100118 20:46:35< GNUtoo> http://pastebin.com/d393cd4ae
20100118 20:47:25< fendrin> boucman:   bool matches_filter(const vconfig& cfg,const map_location& loc,bool use_flat_tod=false) const; in unit.hpp. It's the loc argument I am worrying about.
20100118 20:47:25< boucman> fendrin: basically, it's a call to unit::matches_filter
20100118 20:47:33< boucman> lemme check
20100118 20:47:34< GNUtoo> window_width = 0
20100118 20:47:38< GNUtoo> window_height = 0
20100118 20:47:42< GNUtoo> that's what I changed
20100118 20:48:11< mordante> the ones below GUI_TINY__RESOLUTION?
20100118 20:48:15< fendrin> boucman: I guess it's an relict from times where the unit didn't know about it's own position on the map.
20100118 20:48:42< boucman> fendrin: probably, but it might also be used as "would the unit match if it were on a particular location
20100118 20:48:51< mordante> ah I see, missed the pastebin line
20100118 20:49:00< boucman> since that can't be made a default argument
20100118 20:52:00< GNUtoo> mordante, yes
20100118 20:52:37< GNUtoo> mordante, in wesnoth 1.4 it worked out of the box without config file changes (nor rant but info)
20100118 20:54:07< mordante> did you also try with 480 x 640 (or what your resolution is) to 0x0 makes it the default 
20100118 20:54:57< mordante> true this new gui has been in development (might be some parts in 1.4 were done) so I'm not surprised it worked out of the box at that time
20100118 20:55:05< GNUtoo> mordante, the problem was thqat it worked only in 640x480 and not 480x640
20100118 20:55:29< mordante> this gui still needs more testing on tiny guis
20100118 20:55:33< GNUtoo> ok
20100118 20:55:41< GNUtoo> mordante, are you going to fosdem?
20100118 20:55:55< GNUtoo> I've the following screens:
20100118 20:56:00< GNUtoo> 480x640
20100118 20:56:04< GNUtoo> 480x320
20100118 20:56:06< mordante> yes that's why I asked to swap those numbers for height and width
20100118 20:56:15< GNUtoo> and eeepc but everybody already tested on eeepc
20100118 20:56:44< mordante> I'll be at the fosdem 
20100118 20:57:23< mordante> so if you take your system with you I would like to do some gui tuning
20100118 20:57:47< mordante> looking at tiny gui at my screen gives me headaches
20100118 20:59:01< GNUtoo> lol ok
20100118 21:01:13-!- Blueblaze [n=nick@adsl-99-158-47-180.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100118 21:03:02-!- EdB [n=edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has quit [Remote closed the connection]
20100118 21:04:42< GNUtoo> I bet I won't see pandoras at the fosdem
20100118 21:04:54< Ivanovic> correct
20100118 21:04:58< mordante> :-(
20100118 21:05:04< Ivanovic> though: you won't see them at fosdem 2010
20100118 21:05:10< Ivanovic> maybe at fosdem 2011
20100118 21:05:11< Ivanovic> ;)
20100118 21:05:30 * mordante remembers pandora and fosdem 2009...
20100118 21:05:34< Ivanovic> :)
20100118 21:05:52< Ivanovic> mordante: the classical "ready in 2 month" changed to "ready in 6 weeks" already
20100118 21:05:58< Ivanovic> so, hey, it improved
20100118 21:05:59< Ivanovic> ;)
20100118 21:07:14< mordante> now that duke nukem forever has been scrapped I assume /. needs something to replace it... :-P
20100118 21:08:23< loonycyborg> Reminds me of http://www.xkcd.com/678/
20100118 21:08:57< mordante> indeed
20100118 21:10:30< GNUtoo> lol ok
20100118 21:13:13< stikonas> these news are also interesting: http://forums.wesnoth.org/viewtopic.php?f=6&t=18607 It seems that somebody wants to port Wesnoth to Android
20100118 21:13:23< mordante> GNUtoo, can you test this patch http://paste.debian.net/57040/ (want to make sure the preprocessor stuff is parsed properly)
20100118 21:15:10< GNUtoo> stikonas, they should take wesnoth and turn it in a big JNI
20100118 21:15:54< stikonas> GNUtoo: isn't it possible to run native code on android?
20100118 21:16:13< GNUtoo> scummvm did it this way
20100118 21:16:50< GNUtoo> I don't know for non-rooted android tough,I heard that it was possible on newer version tough
20100118 21:17:32< boucman> there is a C api, but you have no access to phone functions afaik (doesn't matter for wesnoth, I guess)
20100118 21:18:25< mordante> it's also possible to use non-java code now?
20100118 21:19:01< boucman> it is, but it might be limited... I hope network is there, everything else should be fine (some SDL apps have already been ported)
20100118 21:19:10< boucman> and yes, I studied the question at some point :P
20100118 21:19:18< Ivanovic> ram, ram, ram
20100118 21:19:26< Ivanovic> that is the limiting factor with android phones
20100118 21:19:30< GNUtoo> yes ram is the issue
20100118 21:19:41-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has quit [Read error: 104 (Connection reset by peer)]
20100118 21:19:44< stikonas> Ivanovic: 512 MB sould be enough :)
20100118 21:19:45< Ivanovic> most are at such low ram levels, it is insane! (at least if you plan to run wesnoth)
20100118 21:19:51< GNUtoo> anyway I ran wesnoth on my htcdream but the htcdream was running under GNU/Linux
20100118 21:20:09< Ivanovic> stikonas: even 256MB are perfectly fine
20100118 21:20:14< GNUtoo> depends
20100118 21:20:17< GNUtoo> there is shared mem
20100118 21:20:19< Ivanovic> but stuff below requires quite some work
20100118 21:20:22< GNUtoo> between CPU and DSP
20100118 21:20:42< Ivanovic> GNUtoo: does not matter too much for wesnoth, unless you run several other apps, too
20100118 21:20:45< GNUtoo> freerunner: 128M htcdream:110M
20100118 21:21:06< GNUtoo> what does not mather? the shared ram?
20100118 21:21:11< Ivanovic> you don't need the dsp for wesnoth, since ogg most likely ain't supported in some advanced way
20100118 21:21:36< GNUtoo> what I mean is that the phone *needs* the shared mem to work
20100118 21:21:44< Ivanovic> that is "supported in any way beside 'gimme sound'"
20100118 21:22:01< GNUtoo> else all what is connected to the non-main cpu won't work
20100118 21:22:14< GNUtoo> I'm not even shure you can boot the phone without that
20100118 21:22:42< Crab_> boucman: returning to "how are we wrt 1.8 on the AI front ?"...
20100118 21:22:43< GNUtoo> so: 192M->110M under the htcdream
20100118 21:22:49< boucman> sure
20100118 21:23:02< GNUtoo> at least that is my understanding
20100118 21:23:39< Crab_> boucman: c++ side is mostly ok (needs a set of half-ready patches to be applied, and it can be frozen except for bugfixing)
20100118 21:23:54< boucman> k
20100118 21:24:20< Crab_> boucman: fai has got problems - it is difficult to use, and needs a library of useful snippets to make various things easier and faster
20100118 21:25:23< Crab_> the problem with fai is that it is functional language, and it's not easy to write code step-by-step, e.g. "calculate this, store, calculate that, store, combine and calculate again, loop a while, pick best, do"
20100118 21:25:31< boucman> Crab_: hmm, no info is passed through the network, so the library could be added (cautiously) in 1.8.x however, no wml syntax change in 1.8.x so i would focus on c++ side if I were you
20100118 21:25:52< Crab_> that library should be at least partly c++
20100118 21:26:32< boucman> hmm
20100118 21:26:44< boucman> in that case it's not 1.8.x compat
20100118 21:27:06< boucman> since it would most likely introduce stuff that can't be used on all versions
20100118 21:27:24< Crab_> well, I think that I'll expand c++ formula function table before 1.8 starts.
20100118 21:27:36< Crab_> i need pretty simple things there
20100118 21:27:49< Crab_> e.g. there's a large number of pathfinding questions that are easy to answer in c++
20100118 21:29:03< GNUtoo> mordante, your patch seem to work
20100118 21:29:12< boucman> Crab_: how much time do you need ?
20100118 21:29:15< GNUtoo> at least the multiplayer GUI worked
20100118 21:29:38< GNUtoo> now loading wesnoth and pray for not having OOM that I didn't have before(other days,wesnoth 1.4)
20100118 21:30:57< Crab_> boucman: for c++ side, one good evening to settle things right and commit all the stuff that needs committing.   for expanding formula_ai, - several days to ensure that the resulting set of functions is easy enough to code simple things simply
20100118 21:32:05< mordante> GNUtoo, ok then I think I've some ideas how to fix your problem properly
20100118 21:32:17< GNUtoo> ok thanks a lot
20100118 21:32:31-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has joined #wesnoth-dev
20100118 21:32:39< boucman> sounds good, 
20100118 21:32:46< GNUtoo> mordante, Ivanovic do you want videos of wesnoth under freerunner and htcdream?
20100118 21:33:04< mordante> it will even reduce the memory footprint (I don't expect a huge reduction)
20100118 21:33:20< mordante> GNUtoo, I'd like to see them
20100118 21:33:33< Ivanovic> GNUtoo: sure!
20100118 21:33:33-!- YogiHH [n=chatzill@d051178.adsl.hansenet.de] has joined #wesnoth-dev
20100118 21:33:41< GNUtoo> ok I'll do them
20100118 21:33:43< YogiHH> hello
20100118 21:33:50< Ivanovic> GNUtoo: i'd recommend uploading them to youtube and linking them in the release forum
20100118 21:34:09< Crab_> hi, YogiHH
20100118 21:34:14< GNUtoo> ok
20100118 21:34:26< fendrin> YogiHH: Hello
20100118 21:35:06< boucman> once 1.8 is out, we should really try to recontact the iPhone guy  and if he's MIA, scavange what can be from his repository
20100118 21:35:26< Crab_> better grab the repository today :)
20100118 21:35:34< boucman> Crab_: ???
20100118 21:35:38< Ivanovic> boucman: he himself is not MIA
20100118 21:35:50< boucman> it's at SF, it shouldn't disapear...
20100118 21:35:51< Ivanovic> boucman: today he posted that he just submitted "episode 3" to apple
20100118 21:35:56< Crab_> boucman: ah, ok.
20100118 21:35:56< Ivanovic> no, it is *NOT* at sf.net
20100118 21:36:08< boucman> oh, right, it was originally IIRC
20100118 21:36:12< Ivanovic> at sf.net he tried to host it, but it was not working out
20100118 21:36:23< Ivanovic> atm it is here: http://wesnoth.repositoryhosting.com/trac/wesnoth_wesnoth/
20100118 21:36:32< Ivanovic> yes, linked from the "iphone" link on the frontpage
20100118 21:36:37< boucman> well, i'll do a CO of his stuff just to make sure
20100118 21:37:48< mordante> I also have a CO
20100118 21:38:24< mordante> maybe I better get a git checkout
20100118 21:38:34< fendrin> Has he ever been asked to join #wesnoth-dev ?
20100118 21:39:18< boucman> not that I know of
20100118 21:39:25< Ivanovic> fendrin: i at least wrote him something regarding merging stuff back to mainline and what his point of view is
20100118 21:39:37< Ivanovic> but he never reacted after my pm
20100118 21:39:42< boucman> so did I, he never answered either
20100118 21:39:59< boucman> my guess is that he's not an "open source" guy if you see what I men
20100118 21:40:00< boucman> mean
20100118 21:40:19< Ivanovic> jo
20100118 21:40:48< fendrin> We should take care soon about his code. As time is passing by it wíll get more difficult to merge his branch back in.
20100118 21:41:08< boucman> fendrin: it's 1.6 based, so it won't get worse with time...
20100118 21:41:18< boucman> it's based on the latest 1.6 official release
20100118 21:41:58< fendrin> Porting that to trunk will get more difficult from time to time. Or do you want to delete trunk and continue from 1.6+iphone?
20100118 21:42:22< Ivanovic> and it is *not* based on the latest version of 1.6, but some older 1.6 release
20100118 21:42:51< boucman> fendrin: well, 1.6 and trunk are so different at this point that i'm not sure it'll make much of a difference anyway
20100118 21:43:19< boucman> we don't want the iPhone specific stuff anyway, we're more interested on the handling of the tiny screen and user input
20100118 21:43:31< boucman> (at least i'm not able to maintain the iphone part)
20100118 21:43:35< fendrin> boucman: And his memory vaste improvements.
20100118 21:43:41< Ivanovic> tiny screen, lower mem usage, opengl support
20100118 21:43:46< boucman> yes, that too :)
20100118 21:43:50< boucman> opengl ?
20100118 21:43:56< Ivanovic> yeah, he relies on opengl
20100118 21:44:15< boucman> hmm, not sure we want to make that mandatory, our last try at opengl wasn't exactly a success
20100118 21:44:54< fendrin> boucman: silvertree?
20100118 21:45:18< stikonas> fendrin: no, there is wesnoth_opengl branch in svn repository
20100118 21:45:24< boucman> no, I was thinking of the wesnoth-gl branch we had a long time ago... (1.2 or 1.4 IIRC)
20100118 21:45:29< boucman> never was released
20100118 21:45:39< fendrin> Why did it fail?
20100118 21:45:41< Ivanovic> boucman: that was by now ages ago
20100118 21:45:50< boucman> two problems :
20100118 21:46:01< Ivanovic> boucman: drivers for opengl improved a lot over time and opengl support *might* improve things for the current mobile devices
20100118 21:46:16< Ivanovic> those already have well working 3d chips which would help with the load
20100118 21:46:16< boucman> 1) optimizing drawings is completely different if you have a graphic hardware and when you don't
20100118 21:46:33< boucman> which meant rewriting most of the drawing code
20100118 21:46:45< Ivanovic> and open source opengl support for ati/intel cards is looking good, too
20100118 21:46:52< boucman> (animation side could be ported more easily now, but i don't know about the terrain)
20100118 21:46:54< Ivanovic> he *has* already done so for iphone
20100118 21:46:55< Ivanovic> ;)
20100118 21:47:26< boucman> 2) we had problems getting opengl support on all platforms at the time, but as Ivanovic pointed out, the situation might have improved
20100118 21:47:40< boucman> at the time I think windows was (suprisingly) the main problem
20100118 21:47:41< mordante> guess my gui wouldn't be too hard to port to OGL
20100118 21:47:53< Ivanovic> IIRC all the "current mobile platforms" have opengl
20100118 21:48:00< Ivanovic> plus mac, linux and windows
20100118 21:48:03< boucman> mordante: yes, the terrain drawing code is the part that might cause the most problems
20100118 21:48:04< mordante> no at the time Linux with open source drivers was the big problem
20100118 21:48:16< Ivanovic> that is: even opensolaris has opengl simply due to also using mesa
20100118 21:48:22< boucman> mordante: i'll trust you on that, it was a long time ago
20100118 21:48:22< mordante> boucman, not sure it will, it's already drawn stack based
20100118 21:48:46< boucman> then 1.9 might be a good time to start an new opengl branch :P
20100118 21:49:06< Ivanovic> boucman: i'd say "fosdem" sound good for having a look at the stuff kyle did
20100118 21:49:23< mordante> boucman, is it a joke or somewhat serious?
20100118 21:49:25< boucman> yes, that might be a subject
20100118 21:49:26< GNUtoo> opengl? if it's not mandatory it should be ok for me but I'm just an user
20100118 21:49:37< boucman> mordante: somewhat serious
20100118 21:49:48< mordante> ok
20100118 21:49:49< Ivanovic> GNUtoo: could you find out how it looks regarding opengl es support for your platforms?
20100118 21:49:50< GNUtoo> but if it's mandatory it could be a problem because we do not have free 3d drivers on phones
20100118 21:50:05< Ivanovic> GNUtoo: uhm, they don't have to be free, just working
20100118 21:50:06< Ivanovic> ;)
20100118 21:50:12< GNUtoo> on my platform I've only software 
20100118 21:50:19< GNUtoo> no hardware opengl
20100118 21:50:25< GNUtoo> because no free drivers
20100118 21:50:34< GNUtoo> for instance:
20100118 21:50:42< Ivanovic> and rendering opengl in software should not produce much more load than our current system using sdl in software
20100118 21:50:46< GNUtoo> openmoko: the 3d card is so bad that software is faster
20100118 21:51:02< boucman> GNUtoo: we'll probably make it optional for 1.9, and if all platform support it during the 1.9.x we might make it mandatory for 1.10
20100118 21:51:13< Ivanovic> unless we use particle effects and all those joys
20100118 21:51:18< boucman> in other word, if it's a problem for your platform, it probably won't become mandatory
20100118 21:51:26< GNUtoo> htcdream: under android there is a non-free opengl-es driver so not avaliable under GNU/Linux
20100118 21:51:30< mordante> Ivanovic, it does get a very low fps rate for frogatto with my current drivers
20100118 21:51:40< Ivanovic> mordante: what are you using?
20100118 21:51:49< Ivanovic> card+driver that is
20100118 21:52:07< GNUtoo> also most of the previous phones/devices do not have opengl
20100118 21:52:32< GNUtoo> + the only opengl driver for GNU/Linux is half unusable:
20100118 21:52:39< GNUtoo> http://arstechnica.com/open-source/news/2009/12/poulsbo-mess-casts-a-shadow-on-intels-moblin-project.ars
20100118 21:52:51< GNUtoo> powervr = poulsbo = gma500
20100118 21:53:00< mordante> Ivanovic, need to check but I run Debian stable so I expect that the drivers in testing are much faster
20100118 21:53:01< fendrin> Ivanovic: Are you sure that the iphone port does anything usefull with the opengl thing? I imagine it's just used to render in a single rectangle that is the screen.
20100118 21:53:25< boucman> yeah, i have gma500 on my eee, and it's a nuisance :(
20100118 21:53:35< boucman> (i didn't even bother installing the 3d stuff)
20100118 21:53:42< GNUtoo> indeed that card is in most of the curent devices:
20100118 21:53:45< GNUtoo> *pandora box
20100118 21:53:48< GNUtoo> *touchbook
20100118 21:53:52< GNUtoo> *beagle board
20100118 21:54:02< Ivanovic> mordante: what matters, when using the open source drivers, is the kernel (for the drm part) as well as mesa
20100118 21:54:03< fendrin> What do we want to make better with opengl? We have 2d images.
20100118 21:54:03< boucman> hopefully it will improve with time...
20100118 21:54:03< GNUtoo> *gumstix
20100118 21:54:05< GNUtoo> etc...
20100118 21:54:24< Ivanovic> GNUtoo: if things work with the pandora, which is *very* likely, then other systems should be fine, too
20100118 21:54:42< boucman> fendrin: we use a lot of CPU to handle invalidation and redrawing, ensuring that the correct part of the screen are redrawn
20100118 21:54:45< GNUtoo> I don't know...If I understand well they are stuck with an old kenrel
20100118 21:54:50< boucman> with opengl, we don't have to handle that
20100118 21:54:57< GNUtoo> ok
20100118 21:55:10< Ivanovic> GNUtoo: regarding what i heard from djwillis they are not in fact stuck with an old kernel
20100118 21:55:11< boucman> fendrin: were you the one asking about :sunset, btw ?
20100118 21:55:26< Ivanovic> they will ship .29 for the first release because it was tested over the last yeah
20100118 21:55:27< fendrin> boucman: no
20100118 21:55:29< boucman> k
20100118 21:55:36< GNUtoo> ok
20100118 21:55:37< Ivanovic> but plan on switching to .33 in the first "real" update
20100118 21:55:59< mordante> Ivanovic, radeon driver on 2.6.26-2-amd64 and Radeon HD 3200 (integrated on the motherboard)
20100118 21:56:08< Ivanovic> mordante: ah, okay
20100118 21:56:19< mordante> boucman, teaser was
20100118 21:56:19< Ivanovic> you doN't have any hardware accell!
20100118 21:56:44< GNUtoo> indeed there is a lot of problems with 3d
20100118 21:56:45< Ivanovic> you should update to a .32+ kernel to get 3d accell, .30 to get 2D accell (exa)
20100118 21:56:58< Ivanovic> and you need mesa 7.7+, too
20100118 21:57:06-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Excess Flood]
20100118 21:57:16< Ivanovic> beside this those onboard chips are slow as hell...
20100118 21:58:00< mordante> Ivanovic, I know that I only have software rendering, but you said software was as fast/slow as sdl ;-)
20100118 21:58:07< Ivanovic> boucman: personally i'd see opengl as option for the 1.9/1.10 series and maybe the default afterwords if it works
20100118 21:58:27< GNUtoo> will software rendering be fast enough?
20100118 21:58:37< mordante> I'm aware the on board chips aren't fast, but I don't really need it
20100118 21:58:44-!- crimson_penguin [n=ben@64.201.60.211] has joined #wesnoth-dev
20100118 21:58:58< mordante> if one day I need it I'll buy a separate card ;-)
20100118 21:58:59< boucman> Ivanovic: i more or less agree, it's just that if backward compatibility is a pain and all platform support it well in 1.9, we might make it mandatory 1.10
20100118 21:59:00< Ivanovic> mordante: your chip ain't even able get a plain video in a larger resolution done without any 2d accell
20100118 21:59:07< boucman> again, that is assuming all ports are fine with that
20100118 21:59:08< Ivanovic> so it does not say *anything* at all
20100118 21:59:09< Ivanovic> ;)
20100118 21:59:26-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Excess Flood]
20100118 22:00:03< Ivanovic> boucman: since 1.9 is meant to be started rather soon i'd guess that there will still be platforms with opengl handicapped drivers
20100118 22:00:13< boucman> most likely
20100118 22:00:59< Ivanovic> at least debian, but also *buntu (due to the lovely "we only give you new versions every half year"-policy)
20100118 22:01:15< mordante> Ivanovic, I had not problems with videos thusfar, not even with my old system (also without accel)
20100118 22:01:30< Ivanovic> mordante: the problem is resizing videos
20100118 22:01:35< Ivanovic> not normal decoding, that is harmless
20100118 22:02:02< Ivanovic> but without 2d accell, it is basically impossible to resize a video window to something above 1400x900 (that was the result of a test on my box)
20100118 22:02:44< mordante> I looked at upscaled versions without problems as well (max screen size 1280x1024)
20100118 22:03:02-!- crimson_penguin [n=ben@64.201.60.211] has joined #wesnoth-dev
20100118 22:03:18< mordante> and I know there are options if it gets too slow for me :-)
20100118 22:07:30< Ivanovic> okay, your resolution is just too low to make a real difference
20100118 22:07:46< mordante> indeed
20100118 22:08:14< Ivanovic> your cpu should be a little stronger than mine and video is done via plain X11 in shadowfb, there 1280x1024 should still be possible
20100118 22:08:22< Ivanovic> though one of your cores should be at some 90%
20100118 22:17:43-!- mjs-de [n=mjs-de@wh.uni-dortmund.de] has quit [Remote closed the connection]
20100118 22:19:46-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has quit [Read error: 104 (Connection reset by peer)]
20100118 22:30:17-!- Blarumyrran [n=Blarumyr@81-20-159-197.levira.ee] has joined #wesnoth-dev
20100118 22:30:56< fendrin> Ivanovic: The iphone ports uses unitmaps in png format for the animations. Would you like that to be ported?
20100118 22:32:52< CIA-53> crab * r40744 /trunk/ (9 files in 6 dirs): lua ai support engine, interface code and ai integration
20100118 22:33:29< boucman> fendrin: what are these ?
20100118 22:33:51< crimson_penguin> Ivanovic: ooonly and hour to go
20100118 22:34:16< fendrin> boucman: They are png that contains the animations for a unit.
20100118 22:34:38< Ivanovic> fendrin: you mean "unit sheets"?
20100118 22:34:53< boucman> fendrin: i just completed my checkout, i'll look at them
20100118 22:34:57< Ivanovic> as in: all frames stored in one larger image and some wml selects which areas to use"?
20100118 22:35:02< boucman> (at the moment i don't really see the difference with our frames
20100118 22:35:21< fendrin> Ivanovic: yes, like that.
20100118 22:35:45< Ivanovic> fendrin: so basically what frogatto already does and what jetryl once was again but now does support
20100118 22:35:48< fendrin> boucman: http://wesnoth.repositoryhosting.com/svn_public/wesnoth_wesnoth/trunk/res/data/core/images/unitmaps/
20100118 22:36:48< mordante> Smar, thanks for your bug report, I just confirmed the bug
20100118 22:37:00< boucman> fendrin: spritesheets were discussed and had been rejected earlier
20100118 22:37:22< fendrin> boucman: That is why I asked. 
20100118 22:37:38< mordante> boucman, jetrel changed his opinion and likes them now
20100118 22:37:44< Ivanovic> boucman: when they were discussed long ago they were rejected because jetryl was against it
20100118 22:38:03< zookeeper> IIRC jetrel would like spritesheets
20100118 22:38:15< Ivanovic> now he wants them
20100118 22:38:29< boucman> the main reason was jetrel found them hard to manipulate...
20100118 22:38:41< fendrin> They are doable with the wesnoth engine afaik. shadowmaster implemented the subloading of a image via wml.
20100118 22:38:44< boucman> jmm
20100118 22:39:10< mordante> IIRC added upon Jetrel's request
20100118 22:39:15< boucman> fendrin: i personally don't mind either way...
20100118 22:39:33< mordante> I think we should prefer what our artist prefer
20100118 22:39:54< mordante> artists*
20100118 22:40:16< boucman> afaik the main selling point of spritesheets was a better imge compression
20100118 22:40:47< Ivanovic> jupp
20100118 22:40:51< mordante> jup
20100118 22:42:25< boucman> fendrin: ok, i don't really mind spritesheets per-se, but we have to make sure it's implemented intelligently in the engine
20100118 22:43:04< boucman> looking at what the iphone port attained memory-wise i'm not too worried, but looking at the directory structure, i'm not sure how the unitmaps<->unit images mapping works
20100118 22:43:11< boucman> it doesn't seem to be specified in WML
20100118 22:43:31< Ivanovic> probably hardcoded
20100118 22:43:36< boucman> :(
20100118 22:43:36-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Excess Flood]
20100118 22:43:45< fendrin> The screen resolution is hardcoded as well.
20100118 22:44:16-!- crimson_penguin [n=ben@64.201.60.211] has joined #wesnoth-dev
20100118 22:44:32-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Excess Flood]
20100118 22:44:33< boucman> Ivanovic: seeing how each unit has a particular set of animations and frames, i'd be suprised if it really was hardcoded... you never know, but i would be suprised if there wern't a plain-text mapping somewhere
20100118 22:45:22-!- crimson_penguin [n=ben@64.201.60.211] has joined #wesnoth-dev
20100118 22:45:38-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Excess Flood]
20100118 22:46:04< Ivanovic> uhm, those "maps" are basically broken over here when just downloading them
20100118 22:46:17< boucman> ???
20100118 22:46:35< Ivanovic> i downloaded them and only the first "row" is shown completely
20100118 22:46:59< Ivanovic> that is: the files map.elves.fighter.png and map.drakes.fire.png
20100118 22:47:19< Ivanovic> and only the first line of images is shown plus the top pixels of the 2nd
20100118 22:47:30< boucman> oh
20100118 22:47:41< boucman> I had that problem too, but the svn checkout ones were ok
20100118 22:48:07< Ivanovic> interesting
20100118 22:49:28< Crab_> hmm, I get some linking errors in lobby_main... http://wesnoth.pastebin.com/m562b7b97
20100118 22:49:53-!- crimson_penguin [n=ben@64.201.60.211] has joined #wesnoth-dev
20100118 22:49:59< mordante> Crab_, which build system?
20100118 22:50:04< Crab_> scons
20100118 22:50:12< fendrin> I can't find a textfile. So I guess the position of the subimage is somehow derived from the filename.
20100118 22:52:21< boucman> hmm
20100118 22:52:23< loonycyborg> Crab_: Maybe you could try rm -rf build
20100118 22:52:34< Crab_> loonycyborg: will try
20100118 22:52:38< Crab_> thanks
20100118 22:52:40< boucman> possible, but what would be the size of the "square" around each sprite
20100118 22:53:47< boucman> looking at WML, images are named units/elves-wood/enchantress-magic-2.png
20100118 22:54:51< fendrin> boucman: yes, that is why i assumed that this filepath is somehow converted into a location in the unitmap.
20100118 22:54:52< Ivanovic> boucman: maybe some hardcoded mapping with "first attack" -> row1, "2nd attack" -> row2, "defense (miss)" -> row3, "defense (hit)" -> row4
20100118 22:55:02< boucman> and diffing with latest 1.6 svn, there are no difference in the WML file wrt animations, so the info is most likely elsewhere
20100118 22:55:33< fendrin> boucman: UnitTextureAtlas.cpp
20100118 22:56:23< boucman> :(
20100118 22:56:27< boucman> yeah that's it...
20100118 22:56:45< boucman> ok, fendrin, to answer your original question, no I don't want that in mainline
20100118 22:56:47< boucman> :P
20100118 22:57:09< fendrin> boucman: It's just impressive.
20100118 22:57:19< Ivanovic> what we would want in mainline might be "configurable as now, just pointing at some place in the image"
20100118 22:57:28< boucman> fendrin: when you think of it, not that much...
20100118 22:57:43< boucman> it's not much more work to write it as a cpp include as it is as a WML file
20100118 22:57:53< boucman> it's pure lazyness from my point of view
20100118 22:58:13< mordante> I'm off night
20100118 22:58:16< boucman> night
20100118 22:58:23-!- mordante [n=mordante@wesnoth/developer/mordante] has quit ["Leaving"]
20100118 23:00:42< fendrin> boucman: Well, the file is a nightmare. If the rest of the iphone port matches it I wouldn't try to port. Everything I have seen so far was hardcoded in some way.
20100118 23:01:03< boucman> hmm
20100118 23:01:10< boucman> that's a bit of a disapointment
20100118 23:01:23< boucman> did you have a look at the opengl stuff yet ?
20100118 23:01:45< fendrin> boucman: there is some in video.cpp
20100118 23:01:50-!- MikeJB [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #Wesnoth-dev
20100118 23:02:13< fendrin> There the resolution is hardcoded. The file is a hack.
20100118 23:04:11 * boucman is really disapointed :/
20100118 23:04:27< fendrin> boucman: About what code detail in particular?
20100118 23:04:43< boucman> what do you mean ?
20100118 23:06:16< fendrin> boucman: The opengl implementation can still be usefull. I haven't discovered where the black magic takes place now.
20100118 23:06:22-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100118 23:06:41< boucman> fendrin: from what you described so far, I don't think it's a smart implementation
20100118 23:06:46< loonycyborg> fendrin: Are all gl calls in that one file?
20100118 23:07:12< boucman> it might be that ogl is the only way to access video on iphone, but it's used as a drawing board with all our video routines doing the real work
20100118 23:07:58< fendrin> loonycyborg: For some reason the checkout from the iphone svn repository has all original wesnoth files be set ececutable and the new ones not. This is an easy way to see new files in the port.
20100118 23:08:47< fendrin> boucman: Right, that is what I have imagined some 20 minutes before. It is only used to render in a rectangle that is the screen.
20100118 23:09:12< fendrin> boucman: We would not gain anything if that is the case.
20100118 23:09:46< Ivanovic> yeah, we most likely have to really check the port for things that we maybe can use
20100118 23:10:14< boucman> the mem improvements would be nice if they could be used in 1.9
20100118 23:10:18< Ivanovic> that is look for the ways used to reduce the memory footprint (ways beside deactivating the ingame help when in a game)
20100118 23:11:01< Ivanovic> yeah, some of the improvements (many? all? no idea) are done by deactivating stuff
20100118 23:11:20< Ivanovic> no idea how much might be done via crude hacks making the code unreadable
20100118 23:11:51< loonycyborg> fendrin: I wasn't really wondering about new files. I wondered how exactly it uses gl, i.e. whether gl is restricted to one place or they're used everywhere.
20100118 23:12:46< fendrin> loonycyborg: A grep for RenderQueue.h gives not that much hits. I guess that is the interface to the whole opengl thing.
20100118 23:15:51< crimson_penguin> Ivanovic: 22 minutes :P
20100118 23:15:57< Ivanovic> hui
20100118 23:16:08< crimson_penguin> ugh, why's it going at 13?!
20100118 23:27:02-!- boucman [n=rosen@wesnoth/developer/boucman] has quit ["Leaving."]
20100118 23:29:19< CIA-53> caslav_ilic * r40745 /trunk/ (5 files in 2 dirs): Updated localized images for Russian.
20100118 23:33:01-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20100118 23:33:11-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20100118 23:44:47-!- asdasd_ [i=4cc8cb2b@gateway/web/freenode/x-regyvgzssofxrxaa] has joined #wesnoth-dev
20100118 23:44:51-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20100118 23:45:47-!- asdasd_ [i=4cc8cb2b@gateway/web/freenode/x-regyvgzssofxrxaa] has quit [Client Quit]
20100118 23:48:27-!- GNUtoo [n=GNUtoo@host82-133-dynamic.54-79-r.retail.telecomitalia.it] has quit ["Leaving"]
20100118 23:52:58 * fendrin wonders if anyone around ever tried the iphone port.
20100118 23:53:43< CIA-53> caslav_ilic * r40746 /trunk/po/wesnoth-manual/ (sr.po sr@latin.po): Updated translations for Serbian.
20100118 23:53:55< crimson_penguin> oh right, it's done; better go default it
20100118 23:54:25< CIA-53> caslav_ilic * r40747 /trunk/ (changelog players_changelog): Updated translations for Serbian (changelogs updated).
20100118 23:58:53< Ivanovic> crimson_penguin: so upload is compelted?
20100118 23:59:10< crimson_penguin> Ivanovic: yup, I'm just setting it as the default now
20100118 23:59:41< Ivanovic> can you also default the sources to anything beside windows and mac?
20100118 23:59:58< crimson_penguin> ok
--- Log closed Tue Jan 19 00:00:30 2010