--- Log opened Fri Dec 31 00:00:41 2010
--- Day changed Fri Dec 31 2010
20101231 00:00:41-!- vcap [~vcap@AReims-551-1-101-51.w86-207.abo.wanadoo.fr] has joined #wesnoth-dev
20101231 00:01:58-!- mjs-de [~mjs-de@p3EE22F55.dip.t-dialin.net] has quit [Ping timeout: 260 seconds]
20101231 00:02:26-!- mjs-de [~mjs-de@p3EE271AE.dip.t-dialin.net] has joined #wesnoth-dev
20101231 00:05:32-!- vcap [~vcap@AReims-551-1-101-51.w86-207.abo.wanadoo.fr] has quit [Read error: Operation timed out]
20101231 00:06:54-!- GNUtoo|laptop [~gnutoo@host7-149-dynamic.51-79-r.retail.telecomitalia.it] has quit [Quit: Program received signal SIGSEGV, Segmentation fault.]
20101231 00:08:05-!- Upth [ogmar@adsl-75-26-166-84.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20101231 00:08:05-!- Upth is now known as Upthorn
20101231 00:13:12-!- mjs-de [~mjs-de@p3EE271AE.dip.t-dialin.net] has quit [Remote host closed the connection]
20101231 00:24:25-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Ping timeout: 276 seconds]
20101231 00:49:38-!- vcap [~vcap@AReims-551-1-36-161.w86-215.abo.wanadoo.fr] has joined #wesnoth-dev
20101231 00:51:52-!- Johannes13 [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 276 seconds]
20101231 01:03:17-!- Grimling [~floris_ki@91.181.56.191] has quit [Ping timeout: 276 seconds]
20101231 01:04:37-!- Grimling [~floris_ki@91.181.26.59] has joined #wesnoth-dev
20101231 01:17:06-!- eoc [~eoc@pD956153A.dip0.t-ipconnect.de] has quit [Quit: Leaving]
20101231 01:17:11-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20101231 01:24:58-!- fstltna [~fstltna@184.235.220.222] has joined #wesnoth-dev
20101231 01:45:03-!- Cookiee [~quassel@unaffiliated/cookiee] has joined #wesnoth-dev
20101231 01:50:53-!- Cookiee [~quassel@unaffiliated/cookiee] has quit [Ping timeout: 272 seconds]
20101231 01:54:49-!- Cookiee [~quassel@unaffiliated/cookiee] has joined #wesnoth-dev
20101231 01:57:19-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has quit [Quit: Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz]
20101231 01:58:25-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection]
20101231 01:59:01-!- Cookiee is now known as Cookie
20101231 01:59:31-!- Cookie is now known as Guest45872
20101231 01:59:55-!- ancestral [~ancestral@mobile-166-137-140-158.mycingular.net] has joined #wesnoth-dev
20101231 02:01:08-!- Guest45872 [~quassel@unaffiliated/cookiee] has quit [Quit: OH. LOOK ITS A COOKIE THERE, RIGHT THERE *runs away*]
20101231 02:01:31-!- Cookiee [~quassel@unaffiliated/cookiee] has joined #wesnoth-dev
20101231 02:15:58-!- ancestral [~ancestral@mobile-166-137-140-158.mycingular.net] has quit [Quit: Smell ya later!]
20101231 02:17:02-!- ancestral [~ancestral@mobile-166-137-140-158.mycingular.net] has joined #wesnoth-dev
20101231 02:40:10-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20101231 02:58:15< neb1> wesbot seen Crab
20101231 02:58:15< wesbot> neb1: Sorry, I don't know of Crab.
20101231 02:58:19< neb1> wesbot seen Crab_
20101231 02:58:19< wesbot> neb1: The person with the nick Crab_ last spoke 22h 23m ago.  22h 23m ago they left with the message: Quit: Leaving.
20101231 03:06:14-!- ancestral [~ancestral@mobile-166-137-140-158.mycingular.net] has quit [Quit: Smell ya later!]
20101231 03:08:22-!- Grimling [~floris_ki@91.181.26.59] has left #wesnoth-dev []
20101231 03:10:41-!- fendrin [~fabi@77-20-109-69-dynip.superkabel.de] has joined #wesnoth-dev
20101231 03:10:41-!- fendrin [~fabi@77-20-109-69-dynip.superkabel.de] has quit [Changing host]
20101231 03:10:41-!- fendrin [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20101231 03:13:15-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20101231 03:30:35-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Read error: Connection reset by peer]
20101231 03:31:16-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20101231 03:31:19< neb1> Crab_: I followed your advice and answered with a full answer ALL of the questions about the changes I would make in wesnoth... so I think I finished my work for the GCi task http://www.google-melange.com/gci/task/show/google/gci2010/wesnoth/t129047209781
20101231 03:31:27< neb1> Crab_: http://www.google-melange.com/gci/task/show/google/gci2010/wesnoth/t129047209781
20101231 03:31:51< neb1> my bad, Crab_: http://wiki.wesnoth.org/GCI/Playtesting/Heir_to_the_Throne/Challenging   
20101231 03:39:40-!- Gambit [~Gambit@wesnoth/developer/grickit] has quit [Quit: trying something]
20101231 03:40:10-!- Gambit [~Gambit@wesnoth/developer/grickit] has joined #wesnoth-dev
20101231 04:04:17-!- Ivanovic_ [~ivanovic@dtmd-4d0bf0bb.pool.mediaWays.net] has joined #wesnoth-dev
20101231 04:04:17-!- Ivanovic_ [~ivanovic@dtmd-4d0bf0bb.pool.mediaWays.net] has quit [Changing host]
20101231 04:04:17-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20101231 04:07:47-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 240 seconds]
20101231 04:08:11-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 265 seconds]
20101231 04:08:18-!- Ivanovic_ is now known as Ivanovic
20101231 04:09:16-!- Blueblaze [~Blueblaze@99.148.246.110] has quit [Quit: Blueblaze]
20101231 04:12:17-!- }neb1{ [~laurion@f79-118-217-246.pitesti.rdsnet.ro] has joined #wesnoth-dev
20101231 04:13:38-!- neb1 [kenzy@79.118.216.169] has quit [Ping timeout: 240 seconds]
20101231 04:32:50-!- neb1 [neb1@f79-118-217-124.pitesti.rdsnet.ro] has joined #wesnoth-dev
20101231 04:33:48-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20101231 04:34:00-!- PetePorty [~Pete@pc-159-54-120-200.cm.vtr.net] has joined #wesnoth-dev
20101231 04:35:27-!- }neb1{ [~laurion@f79-118-217-246.pitesti.rdsnet.ro] has quit [Ping timeout: 246 seconds]
20101231 04:56:57-!- neb1 [neb1@f79-118-217-124.pitesti.rdsnet.ro] has quit []
20101231 04:57:03-!- neb1 [neb1@f79-118-217-124.pitesti.rdsnet.ro] has joined #wesnoth-dev
20101231 05:07:14-!- neb1 [neb1@f79-118-217-124.pitesti.rdsnet.ro] has quit []
20101231 05:21:52-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20101231 05:49:38-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20101231 05:49:38-!- Gambit [~Gambit@wesnoth/developer/grickit] has quit [Read error: Connection reset by peer]
20101231 06:28:51-!- PeterPorty [~Pete@pc-159-54-120-200.cm.vtr.net] has joined #wesnoth-dev
20101231 06:29:57-!- PeterPorty is now known as PeteyPorty
20101231 06:39:11-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20101231 06:56:06-!- shadowm_laptop is now known as evilshadowmaster
20101231 07:45:16-!- zookeeper [~l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20101231 07:45:49-!- evilshadowmaster [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 240 seconds]
20101231 07:47:54-!- PeteyPorty [~Pete@pc-159-54-120-200.cm.vtr.net] has quit [Quit: Saliendo]
20101231 07:48:01-!- evilshadowmaster [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20101231 07:48:04-!- PetePorty [~Pete@pc-159-54-120-200.cm.vtr.net] has quit [Read error: Connection reset by peer]
20101231 07:59:58-!- evilshadowmaster [ignacio@wesnoth/developer/shadowmaster] has quit [Disconnected by services]
20101231 08:00:05-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20101231 08:06:09< CIA-84> zookeeper * r48135 /trunk/data/campaigns/Heir_To_The_Throne/scenarios/04_The_Bay_of_Pearls.cfg: When Delfador appears to speak, use his stored self so the player doesn't think his XP has vanished.
20101231 08:31:56-!- Blueblaze [~Blueblaze@adsl-99-148-246-110.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20101231 08:33:08-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Quit: ...]
20101231 08:34:53-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has quit [Ping timeout: 264 seconds]
20101231 08:35:18-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 260 seconds]
20101231 08:39:39-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20101231 09:16:22-!- Upth [ogmar@adsl-75-26-166-84.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20101231 09:20:00-!- Upthorn [ogmar@adsl-75-26-166-84.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 246 seconds]
20101231 09:20:54-!- Upth [ogmar@adsl-75-26-166-84.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 265 seconds]
20101231 09:20:56-!- Upthorn [ogmar@adsl-75-26-166-84.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20101231 09:35:03-!- timotei [~timotei@89.43.196.13] has joined #wesnoth-dev
20101231 09:35:03-!- timotei [~timotei@89.43.196.13] has quit [Changing host]
20101231 09:35:03-!- timotei [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20101231 09:35:26-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20101231 09:51:16-!- eoc [~eoc@pD9561687.dip0.t-ipconnect.de] has joined #wesnoth-dev
20101231 10:10:49-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20101231 10:16:33< timotei> Happy new year guys
20101231 10:18:09< Cookiee> new years eve :D
20101231 10:19:37-!- timotei [~timotei@wesnoth/developer/timotei] has quit [Quit: Leaving]
20101231 10:25:54< Ivanovic> moin
20101231 10:28:33-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has joined #wesnoth-dev
20101231 10:29:48< Appleman1234> Cookiee, Happy New Years Eve ..
20101231 10:45:53-!- mordante [~mordante@roadie.xs4all.nl] has joined #wesnoth-dev
20101231 10:45:53-!- mordante [~mordante@roadie.xs4all.nl] has quit [Changing host]
20101231 10:45:53-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20101231 10:46:05< mordante> servus
20101231 10:48:27< Espreon> mordante: Ic grete þe.
20101231 10:48:34< mordante> hi Espreon 
20101231 10:52:51< Ivanovic> AI0867: your task requires review: http://socghop.appspot.com/gci/task/show/google/gci2010/wesnoth/t129345906203
20101231 10:53:45-!- GNUtoo|laptop [~gnutoo@host7-149-dynamic.51-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20101231 10:54:41< Cookiee> Appleman1234: you too! :)
20101231 10:56:11-!- MeccaGod [~majs@h14n4fls31o279.telia.com] has joined #wesnoth-dev
20101231 10:58:39< mordante> AI0867, which campaign does lightfigher want to add? I ran it the last time on the 28th, running it now again
20101231 10:58:56-!- loonybot [~loonybot@94.29.61.177] has joined #wesnoth-dev
20101231 10:58:56-!- loonybot [~loonybot@94.29.61.177] has quit [Changing host]
20101231 10:58:56-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-dev
20101231 10:59:12< Ivanovic> mordante: how is work with po4a stuff going?
20101231 10:59:54-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20101231 10:59:54< Espreon> mordante: Probably The Settlers of Light.
20101231 11:01:06< Espreon> mordante: Hmmm, well, TSoL is for 1.9, and there's no 1.9 branch in Wescamp's SVN repository... so... yeah.
20101231 11:02:33-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has quit [Remote host closed the connection]
20101231 11:04:52< mordante> Ivanovic, ask me later today, want to work on it today
20101231 11:04:59< Ivanovic> mordante: cool!
20101231 11:08:21-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has joined #wesnoth-dev
20101231 11:08:50-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has quit [Remote host closed the connection]
20101231 11:11:13-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has joined #wesnoth-dev
20101231 11:11:50-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has quit [Remote host closed the connection]
20101231 11:12:23-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has joined #wesnoth-dev
20101231 11:13:43-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has quit [Remote host closed the connection]
20101231 11:16:14-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has joined #wesnoth-dev
20101231 11:19:51-!- grzywacz [~grzywacz@89-78-180-180.dynamic.chello.pl] has joined #wesnoth-dev
20101231 11:19:51-!- grzywacz [~grzywacz@89-78-180-180.dynamic.chello.pl] has quit [Changing host]
20101231 11:19:51-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20101231 11:21:17< AI0867> mordante: I don't actually know, he was just complaining about it not showing up, while following the instructions on PblWML
20101231 11:21:47< AI0867> Ivanovic: I know
20101231 11:22:06< AI0867> I've gotten gmail to stop spam-shelving melange's mails
20101231 11:30:44< mordante> AI0867, ok just talking with Espreon about it and it seems to be a 1.9 campaign which aren't uploaded yet
20101231 11:41:54-!- mjs-de [~mjs-de@p3EE271AE.dip.t-dialin.net] has joined #wesnoth-dev
20101231 11:42:17< mordante> AI0867, Espreon will look at uploading 1.9 later first want to look at the po4a stuff
20101231 11:42:48< Espreon> OK.
20101231 11:45:55-!- Grimling [~Grimling@91.181.246.135] has joined #wesnoth-dev
20101231 11:47:40< CIA-84> ivanovic * r48136 /branches/1.8/po/ (7 files in 7 dirs): updated Korean translation
20101231 11:48:15< CIA-84> ivanovic * r48137 /trunk/ (13 files in 11 dirs): updated Japanese and Korean translation
20101231 11:57:55-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has joined #wesnoth-dev
20101231 12:00:38-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has quit [Client Quit]
20101231 13:15:05-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has joined #wesnoth-dev
20101231 13:17:59-!- Crendgrim [~crend@p54883BCB.dip.t-dialin.net] has joined #wesnoth-dev
20101231 13:26:26-!- Grimling [~Grimling@91.181.246.135] has quit [Quit: Grimling]
20101231 13:28:16< mordante> afk
20101231 13:29:21-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has quit [Quit: Ex-Chat]
20101231 13:37:25< Ivanovic> mordante: the po4a update stuff looks good at a first look
20101231 13:37:33< Ivanovic> here is the output on my end: http://pastebin.com/XWayBEkC
20101231 13:37:57< Ivanovic> one slight thing to change: when discarding it always displays "Discard the translation of ../../doc/manual/manual.en.xml"
20101231 13:38:10< Ivanovic> and when merging against the po file it states "Update pot file."
20101231 13:39:33< Ivanovic> the generated po files as well as the pot file look okay, too
20101231 13:39:41< Ivanovic> (though the header differs in the pot file!)
20101231 13:39:47< Ivanovic> -# Copyright (C) YEAR Wesnoth Development Team
20101231 13:39:49< Ivanovic> +# Copyright (C) YEAR Free Software Foundation, Inc.
20101231 13:41:02< Ivanovic> and the html files appear to be fine, too
20101231 13:49:42-!- Gambit [~Gambit@wesnoth/developer/grickit] has joined #wesnoth-dev
20101231 13:54:40-!- Guest68841 [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20101231 13:55:11-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has quit [Remote host closed the connection]
20101231 13:55:35-!- Guest68841 is now known as janebot
20101231 14:25:49-!- Grimling [~floris_ki@91.181.246.135] has joined #wesnoth-dev
20101231 14:30:08-!- Elvish_Pillager [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20101231 14:34:09< mordante> thanks for testing Ivanovic 
20101231 14:34:51< mordante> the "Discarded" message is created by po4a-translate and I agree silly
20101231 14:35:14< mordante> that's why I added more messages to the output to see which translation "fails"
20101231 14:35:43< mordante> the copyright issue is easy to fix, but I need to upgrade my toolchain 
20101231 14:35:50< mordante> there is an commented out fix for it
20101231 14:37:20< Ivanovic> okay
20101231 14:40:03< mordante> how badly do you want the stuff working?
20101231 14:40:26< mordante> I can either make it work first 'single-threaded' or directly 'multi-threaded'
20101231 14:40:50< Ivanovic> mordante: "something working" right now is a good start
20101231 14:41:03< Ivanovic> so just commit what you got and extend it later so that it is multithreaded
20101231 14:41:42< mordante> ok, will do some minor polishing and get the manpages up then
20101231 14:42:06< mordante> once that works I'll commit, since that will leave me without autotools to verify
20101231 14:45:50< Ivanovic> sounds like a plan to me
20101231 15:09:08< CIA-84> ivanovic * r48138 /trunk/po/wesnoth-sotbe/fi.po: updated Finnish translation
20101231 15:25:13-!- Grimling [~floris_ki@91.181.246.135] has left #wesnoth-dev []
20101231 15:25:21-!- Grimling [~floris_ki@91.181.246.135] has joined #wesnoth-dev
20101231 15:28:28-!- Crab_ [~Crab_@62.80.190.34] has joined #wesnoth-dev
20101231 15:28:32-!- Crab_ [~Crab_@62.80.190.34] has quit [Changing host]
20101231 15:28:33-!- Crab_ [~Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20101231 15:31:52< Grimling> Crab_: hi
20101231 15:33:04-!- PetePorty [~Pete@pc-8-253-120-200.cm.vtr.net] has joined #wesnoth-dev
20101231 15:40:38< Crab_> hello
20101231 15:41:22< Grimling> Crab_: a few days ago I've sent my diff-files for the bugs as a form pm
20101231 15:41:27< Grimling> forum*
20101231 15:41:40< Grimling> just to know, have you already looked at them?
20101231 15:42:02< Crab_> yes, I've looked at them a bit
20101231 15:42:08< Ivanovic> Grimling: you can see this yourself
20101231 15:42:17< Ivanovic> Grimling: in the pm dialog in the forums you got two kinds
20101231 15:42:26< Ivanovic> one is outbox, the other sent (or something along those lines)
20101231 15:42:40< Crab_> Ivanovic: no, 'looked at the PM' and 'looked at the attachment' is somewhat different :)
20101231 15:42:55< Grimling> :p
20101231 15:42:59< Ivanovic> it is in the first kind when just sent but not read, after the recipent read (aka "saw") the message it is automagically in the other
20101231 15:43:18< Ivanovic> Crab_: pah
20101231 15:43:36< Grimling> Ivanovic: didn't know, thanks, but i meant the files :p
20101231 15:43:37< Crab_> Grimling: but I haven't got the time yet to test them, will do that later today or tomorrow.
20101231 15:43:43-!- Cookiee [~quassel@unaffiliated/cookiee] has quit [Remote host closed the connection]
20101231 15:44:41< Grimling> Crab_: ok, take your time, i'm now working on a translation task for wesnoth-help
20101231 15:50:22< Crab_> ok. definitely till the end of tomorrow. the code looks good (there are a few things to improve), but I need to compile,test, commit if all's mostly ok
20101231 15:51:06< Grimling> and can you make the task for it?
20101231 15:51:44< Crab_> yes, I will, but since you're working on a different task atm, what's the point ?
20101231 15:52:21< Crab_> as you won't be able to claim it right now, right ?
20101231 15:52:43-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20101231 15:52:45< Grimling> yes, that's right, but when he's ready, i'm just waiting for the review of the current task, so i hope it won't take long
20101231 15:52:49< Crab_> ok
20101231 16:19:22< fstltna> Hi! Where do I find the script that converts .mp3 files to .ogg files?
20101231 16:21:20< Ivanovic> "the script"?
20101231 16:21:25< Ivanovic> there is no such single script that we use
20101231 16:21:40< fstltna> where it does the conversion at build-time
20101231 16:21:51< Ivanovic> usually our composers directly output ogg files when writing the music
20101231 16:21:55< Ivanovic> there is no such script
20101231 16:22:03< Ivanovic> wesnoth does not have any mp3 stuff anywhere
20101231 16:22:11< fstltna> Something is converting .mp3 files to .ogg files in the build
20101231 16:22:17< Ivanovic> no, it is not
20101231 16:22:22< fstltna> ok, it must be in the iPhone version
20101231 16:22:26< shadowmaster> fstltna: are you talking of the iWhatever port?
20101231 16:22:28< Ivanovic> why do you think something like this would be done?
20101231 16:22:54< fstltna> To compress the mp4 files to ogg to save space
20101231 16:22:55< fstltna> The iphone version definately does this
20101231 16:23:04< fstltna> mp3 i meant
20101231 16:23:20< shadowmaster> please realize that the iWhatever ports differ a *lot* in architecture from the mainline distribution
20101231 16:23:28< Ivanovic> fstltna: the iphone version can't convert anything from mp3 to ogg
20101231 16:23:28< fstltna> I am trying to find where it does it to make sure it is working properly/how to judge if it is worth it...
20101231 16:23:42< Ivanovic> fstltna: since there is *NO* mp3 stuff in any of the wesnoth repositories
20101231 16:23:55< Ivanovic> where should the iphone port have mp3 files from?
20101231 16:24:13< Ivanovic> that is: they could use mp3 files somehow, but then it would be converting from ogg (what we have) to mp3
20101231 16:24:41< fstltna> it does it in the mp3->ogg direction, not the other way around
20101231 16:24:58< fstltna> Not all the files are mp3, mainly just music files
20101231 16:24:59< Ivanovic> fstltna: where is the source you base this statement on?
20101231 16:25:51< fstltna> It is in the iPhone build
20101231 16:25:57< Ivanovic> url?
20101231 16:26:20< fstltna> Some of the files are ogg all the time, other get converted from mp3 to ogg at package build-time
20101231 16:26:36< Ivanovic> url to the stuff you mean??
20101231 16:26:40< shadowmaster> fstltna: maybe the sound files?
20101231 16:26:52< fstltna> The sound (not music) files are ogg
20101231 16:27:00< shadowmaster> ISTR something about the .wav files not being quite .wav files anymore in that fork
20101231 16:27:16< fstltna> No wav files...
20101231 16:27:41< Ivanovic> http://wesnoth.repositoryhosting.com/trac/wesnoth_wesnoth/browser/trunk/res/data/core/music
20101231 16:27:43< Ivanovic> interesting
20101231 16:27:52< Ivanovic> looks like kyle has really changed *everything*
20101231 16:28:00< Ivanovic> there are only mp3 files in this folder, no ogg files at all
20101231 16:28:08< fstltna> yes he has :)
20101231 16:28:13< fstltna> right
20101231 16:28:33< fstltna> Like I said it does the conversion at build time and converts the mp3 to ogg
20101231 16:28:39< Ivanovic> fstltna: for anything regarding his build (and why this might be a sane approach for him) you have to ask kyle
20101231 16:28:52< Ivanovic> best do so via the forums in the iphone specific threads that already exists
20101231 16:28:53< fstltna> I will, but he is slow to respond
20101231 16:29:02< fstltna> ok thanks
20101231 16:29:05< Ivanovic> we got no idea about this at all
20101231 16:30:10< fstltna> i will dig into it further on my own...
20101231 16:30:35< Ivanovic> fstltna: if i were you i'd just make a checkout of the repository and grep for mp3
20101231 16:31:10< fstltna> The main reason I am asking is because my fork has extensive muysic tracks, and I want to make sure the best tradeoff of size vs. qualioty is made...
20101231 16:31:25< Ivanovic> then just use the geniune ogg files
20101231 16:31:30< fstltna> I have a checkout of it, and a SVN for my fork or his work
20101231 16:31:50< fstltna> There are no gweniune ogg files for what I am doing, the music has been replaced
20101231 16:31:55< Ivanovic> they are likely to sound a whole lot better than anything you get by converting from ogg to mp3 (as kyle must have done at one time) and then back to ogg again
20101231 16:32:03< fstltna> right
20101231 16:32:16< fstltna> There are not from ogg, they are from mp3 native
20101231 16:32:23< Ivanovic> there are genuine ogg fiels
20101231 16:32:32< Ivanovic> just download them from the "mainline" repository!
20101231 16:32:34< Ivanovic> http://svn.gna.org/viewcvs/wesnoth/trunk/data/core/music/
20101231 16:32:35< fstltna> The whole music files have been replaced
20101231 16:32:55< Ivanovic> and the names are the same beside that the "changed" ones end in mp3
20101231 16:33:04< fstltna> I'm working on a steampunk total conversion for the iOS devices
20101231 16:33:23< fstltna> right
20101231 16:33:41< Ivanovic> ahh, you want to use your *own* music that you got as mp3 files
20101231 16:33:49< fstltna> exactly! LOL
20101231 16:34:11< Ivanovic> talk to kyle, he should know what is done how and where
20101231 16:34:32< fstltna> Some of the source files are 1mb big - I want to make sure that they are downgraded to 128kbs 22khz files
20101231 16:34:57< Ivanovic> search the internet for "lame"
20101231 16:34:58< fstltna> I will ask him if I can't figure it out. As I said he is kind of slow at responding sometimes
20101231 16:35:14< Ivanovic> with this program you should be able to easily reencode stuff the way you want it
20101231 16:35:43< fstltna> he is probobly using the Apple-supplied audioconversion command-line too
20101231 16:35:47< Ivanovic> (using a shell and some for magic it should be no problem to convert all mp3 files in a given folder to a "low quality version"
20101231 16:36:03< fstltna> right
20101231 16:38:14< fstltna> i will dig into a bit more and ask him if I can't figure it out... Thanks for the pointers!
20101231 16:40:17-!- Grimling [~floris_ki@91.181.246.135] has left #wesnoth-dev []
20101231 16:40:56< Ivanovic> for i in *mp3; do lame -h --cbr -b 128 --resample 22.05 $i reencoded/$i; done
20101231 16:41:25< Ivanovic> something like this should reencode the mp3 file as cbr 128kbps with 22.05kHz
20101231 16:41:56< Ivanovic> (make sure that the folder "reencoded/" inside the current one exists
20101231 16:43:05< fstltna> thanks
20101231 16:43:07< fstltna> :)
20101231 16:48:20-!- eoc [~eoc@pD9561687.dip0.t-ipconnect.de] has quit [Quit: Leaving]
20101231 16:53:01< fstltna> I did a grep of "afconvert" and "lame" and didn't find anything. I suppose I will have to ask in the forum...
20101231 16:56:03-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has joined #wesnoth-dev
20101231 17:00:45< CIA-84> silene * r48139 /trunk/src/ (game_display.cpp reports.cpp reports.hpp): Removed redundant members reports::report_data::mouseover_hex and selected_hex.
20101231 17:00:51< CIA-84> silene * r48140 /trunk/src/ (game_display.cpp reports.cpp reports.hpp): Removed redundant member reports::report_data::level.
20101231 17:00:53< CIA-84> silene * r48141 /trunk/src/ (game_display.cpp reports.cpp reports.hpp): Removed redundant member reports::report_data::observers.
20101231 17:00:56< CIA-84> silene * r48142 /trunk/src/ (game_display.hpp playmp_controller.cpp playturn.cpp): Removed redundant member game_display::get_playing_team. Added game_display::playing_side.
20101231 17:01:03< CIA-84> silene * r48143 /trunk/src/ (game_display.cpp reports.cpp reports.hpp): Removed redundant member reports::report_data::active_side.
20101231 17:01:06< CIA-84> silene * r48144 /trunk/src/ (game_display.cpp reports.cpp reports.hpp): Removed redundant member reports::report_data::viewing_side.
20101231 17:01:06< CIA-84> silene * r48145 /trunk/src/ (5 files in 2 dirs): Removed class reports::report_data.
20101231 17:01:09< CIA-84> silene * r48146 /trunk/src/scripting/lua.cpp: Added Lua accessor for the unit displayed in the sidebar.
20101231 17:01:19< CIA-84> silene * r48147 /trunk/src/ (reports.cpp reports.hpp): Added parameter to circumvent dynamic report generators.
20101231 17:01:21< CIA-84> silene * r48148 /trunk/src/scripting/lua.cpp: 
20101231 17:01:21< CIA-84> Changed theme item handling to a wesnoth.theme_items table, so that it is similar to wesnoth.wml_actions.
20101231 17:01:21< CIA-84> The wesnoth.register_theme_item function is still used to cause report generators to show up in the table.
20101231 17:01:22< CIA-84> silene * r48149 /trunk/data/scenario-test.cfg: Implemented a whole new status as an example of modifying report generators.
20101231 17:27:18-!- PetePorty [~Pete@pc-8-253-120-200.cm.vtr.net] has quit [Read error: Connection reset by peer]
20101231 17:27:18-!- mjs-de [~mjs-de@p3EE271AE.dip.t-dialin.net] has quit [Ping timeout: 260 seconds]
20101231 17:27:58-!- mjs-de [~mjs-de@p3EE271AE.dip.t-dialin.net] has joined #wesnoth-dev
20101231 17:29:03-!- PetePorty [~Pete@pc-8-253-120-200.cm.vtr.net] has joined #wesnoth-dev
20101231 17:29:19-!- Johannes13 [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20101231 17:31:44-!- PetePorty [~Pete@pc-8-253-120-200.cm.vtr.net] has quit [Read error: No route to host]
20101231 17:33:43-!- PetePorty [~Pete@pc-8-253-120-200.cm.vtr.net] has joined #wesnoth-dev
20101231 17:34:35-!- vcap [~vcap@AReims-551-1-36-161.w86-215.abo.wanadoo.fr] has quit [Ping timeout: 240 seconds]
20101231 17:34:53-!- EdB [~edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has joined #wesnoth-dev
20101231 17:34:55-!- vcap [~vcap@AReims-551-1-82-242.w92-155.abo.wanadoo.fr] has joined #wesnoth-dev
20101231 17:45:54-!- neb1 [neb1@f79-118-255-149.pitesti.rdsnet.ro] has joined #wesnoth-dev
20101231 17:49:56-!- vcap_ [~vcap@AReims-551-1-143-146.w90-18.abo.wanadoo.fr] has joined #wesnoth-dev
20101231 17:52:57-!- vcap [~vcap@AReims-551-1-82-242.w92-155.abo.wanadoo.fr] has quit [Ping timeout: 255 seconds]
20101231 17:59:24-!- Crendgrim [~crend@p54883BCB.dip.t-dialin.net] has quit [Remote host closed the connection]
20101231 18:01:11-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20101231 18:07:49-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has left #wesnoth-dev ["Ex-Chat"]
20101231 18:12:14-!- vcap_ [~vcap@AReims-551-1-143-146.w90-18.abo.wanadoo.fr] has quit [Ping timeout: 260 seconds]
20101231 18:24:19-!- vcap [~vcap@AReims-551-1-112-237.w86-207.abo.wanadoo.fr] has joined #wesnoth-dev
20101231 18:28:51-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20101231 18:33:27-!- vcap [~vcap@AReims-551-1-112-237.w86-207.abo.wanadoo.fr] has quit [Ping timeout: 255 seconds]
20101231 18:40:24-!- vcap [~vcap@AReims-551-1-11-148.w90-58.abo.wanadoo.fr] has joined #wesnoth-dev
20101231 18:42:00-!- fstltna [~fstltna@184.235.220.222] has quit [Ping timeout: 255 seconds]
20101231 18:43:21-!- EdB [~edb@tss37-1-89-84-18-220.dsl.club-internet.fr] has quit [Remote host closed the connection]
20101231 18:45:23-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20101231 19:01:49-!- vcap [~vcap@AReims-551-1-11-148.w90-58.abo.wanadoo.fr] has quit [Ping timeout: 240 seconds]
20101231 19:11:08< mordante> I'm off have a happy new year
20101231 19:11:16-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has quit [Remote host closed the connection]
20101231 19:11:24-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20101231 19:22:31-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has joined #wesnoth-dev
20101231 19:36:55-!- Mussious [~kamil@dgd58.neoplus.adsl.tpnet.pl] has quit [Quit: Ex-Chat]
20101231 19:48:29-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: GO, GET TO THE CHOPPAH!!!]
20101231 19:48:54-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20101231 19:56:48-!- skyfaller [~skyfaller@wikipedia/Skyfaller] has joined #wesnoth-dev
20101231 19:58:10< skyfaller> Crab_: hi there, Espreon said to talk to you if I was curious about the Wesnoth AI? :)
20101231 19:58:28< Crab_> hello, yes, he's right
20101231 19:58:32< Crab_> how I can help ?
20101231 19:59:26< skyfaller> well, I'm pretty noobish, so I may ask lots of stupid questions and not be very helpful, but I feel like the AI doesn't take into consideration many important things while playing, so I've been reading up on it to try to see how it might be improved
20101231 20:00:14< skyfaller> I came across http://wiki.wesnoth.org/WhyWritingAWesnothAIIsHard and one thing that immediately occurred to me is that perhaps a better place to look for inspiration than chess AIs would be to look at poker bots
20101231 20:01:12< skyfaller> because I also play poker semi-regularly and I see a lot of overlap between the skills I use in poker and the skills I use in Wesnoth
20101231 20:02:15< Crab_> how do they work ?
20101231 20:02:31< skyfaller> well, here's where I admit that I don't actually understand how poker bots work
20101231 20:03:12< Crab_> wesnoth's ai works by evaluating a serie of 'if situation looks like THIS, then do THAT' directives.
20101231 20:03:25< skyfaller> obviously they do a good job of calculating odds, but I'm guessing that the possibility space in poker is still smaller than in Wesnoth
20101231 20:03:50< skyfaller> http://www.codingthewheel.com/archives/how-i-built-a-working-poker-bot is an interesting-looking article
20101231 20:04:02< Crab_> 'if it's possible to recruit, recruit; if there are good attacks available,attack; if there are units which should retreat or heal, make them retreat or heal;  move units closer to enemy;'
20101231 20:04:27< skyfaller> well, I feel that one huge deficit in the Wesnoth AI is strategy
20101231 20:04:32< skyfaller> which I'm sure is extremely hard to program
20101231 20:05:23< skyfaller> but what I'm thinking is that a lot of Wesnoth is like poker in that you have to apply pressure to your opponent, with some degree of advantage on your part, and then your opponent has to respond
20101231 20:05:36< skyfaller> which I think is a little bit like the "check/raise/fold" decision that a poker bot goes through
20101231 20:06:27< skyfaller> in poker, uncertainty partly comes from unrevealed cards, so that part is different, usually you can see all of the relevant units in Wesnoth if you have scouts performing adequately
20101231 20:06:33< skyfaller> (although I wonder whether the AI actually scouts)
20101231 20:07:06< skyfaller> but I think the similarity is that usually you don't have an extremely strong hand
20101231 20:07:16< Crab_> (no, it doesn't, it knows all)
20101231 20:07:31< skyfaller> (oh, that makes things easier I guess)
20101231 20:07:48< skyfaller> sometimes you are straight out bluffing and you don't have the ability to defeat your opponent if they stay in the hand
20101231 20:08:11< skyfaller> similarly, sometimes you can apply pressure to a position without sufficient force to take it if your opponent remains, but if he's a wuss he'll run away anyway
20101231 20:08:26< skyfaller> sometimes you have a decent hand, but it's a coin flip whether you'll win
20101231 20:08:48< skyfaller> and the same thing can happen in Wesnoth when you pit relatively equal forces against each other
20101231 20:09:48< skyfaller> I think maybe the AI should have a strategic consideration of whether to "raise", that is attack full-force, "check", that is to move forward and apply pressure but not actually attack, or "fold", that is to retreat all of your forces
20101231 20:10:03< skyfaller> the AI never retreats its whole army, as far as I can tell
20101231 20:10:44< skyfaller> does that make sense to you?
20101231 20:12:26< skyfaller> as far as I can tell the AI doesn't take into account Time Of Day, and a thousand things like that, because it doesn't consider strategy at all
20101231 20:13:10< skyfaller> but once you have the skeleton of a strategy engine, it can be improved gradually, you don't have to consider all of those factors at once
20101231 20:13:33< Crab_> yes, makes sense. unfortunately, there are some complications. 1) sometimes ai needs to do different things in different places     2) not all parts of done 'raise' is possible, ai can attack quite well;  'check' is not implemented - current ai cannot defend efficiently, 'retreat' is somewhat implemented, but there's no global retreat.
20101231 20:14:05< Crab_> I guess that it's pretty easy to make the AI which will either do a full attack or full retreat depending on ToD.
20101231 20:14:13< skyfaller> there are other obvious considerations that the AI doesn't take into account, such as XP conservation / giving XP to intelligent units / leveling units
20101231 20:14:49< skyfaller> but I think the top priority is giving it the ability to choose "raise/check/fold" on a given front as a strategic consideration
20101231 20:15:17< skyfaller> I guess the AI doesn't have the concept of a "front", is what you're saying
20101231 20:15:19< Crab_> AI partially considers leveling/intelligence in combat, it's accounted for. it doesn't consider the LvLup threat enough
20101231 20:15:43< Crab_> yes, exactly, it has a poor understanding of Zones of Control and 'front's.
20101231 20:15:53< skyfaller> it does?  news to me, maybe it just seems to make bad XP decisions to me, but that's not my biggest concern
20101231 20:16:16< skyfaller> I just feel that the AI's inability to consider defense and retreat is crippling against a competent player
20101231 20:16:41< skyfaller> any ladder player can make mincemeat of the AI, and I think the primary reason is its ignorance of strategy
20101231 20:16:59< skyfaller> it just attacks until it is dead
20101231 20:17:17< skyfaller> every other problem with the AI is secondary IMHO
20101231 20:17:30< Crab_> not 'ignorance of strategy', but 'the unability to defend efficiently', but, more or less, correct.
20101231 20:18:34< skyfaller> so for retreating, you're saying it is smart enough to retreat individual units that are injured, but it will still send out lone healthy units to get trapped and killed?
20101231 20:18:36< Crab_> 'just attacks until it is dead' is just the consequence - if the AI can't defend well, then things will be better if it should attack, then the ai will attack even on unfavourable odds, then, it'll need around 2:1 gold advantage
20101231 20:18:46< Crab_> skyfaller: yes
20101231 20:19:13< skyfaller> well, what makes defense hard?  Is it possible to at least implement "folding" i.e. full retreats?
20101231 20:19:25< Crab_> yes, implementing full retreat is easy enough
20101231 20:19:27< skyfaller> even if we can't teach it to "check"/defend?
20101231 20:19:44< skyfaller> having two possible strategic postures is better than only having one
20101231 20:19:51< Crab_> yes, ' I guess that it's pretty easy to make the AI which will either do a full attack or full retreat depending on ToD.'
20101231 20:19:55< skyfaller> even if it would be better to have at least the 3 I proposed
20101231 20:19:56-!- dtiger [~dtiger@dynamic-vpdn-93-125-65-104.telecom.by] has quit [Read error: Connection reset by peer]
20101231 20:20:18< skyfaller> I mean, TOD is vitally important but it can't be the only consideration... 
20101231 20:20:23< Crab_> yes, it'll be better. we wanted to make a GSoC dedicated to 'make the AI defend better', but unfortunately we had no good applications for that one
20101231 20:20:30< skyfaller> for instance, if you can figure out what the "front" is
20101231 20:20:37< skyfaller> then you could count the units on the "front"
20101231 20:20:53< Crab_> ai has the concept of 'power projection' - how many dmg can be on avg inflicted by enemy on a particular hex
20101231 20:21:01< skyfaller> and if the AI has lots more units available on a given front then it could advance or even attack despite bad TOD
20101231 20:21:28< skyfaller> but it seems that the AI can't "advance" well while restraining itself from attacking?
20101231 20:21:37< Crab_> so it is possible to figure out areas where it has an advantage, and areas where the enemy has an advantage
20101231 20:21:48< skyfaller> ah, that's interesting
20101231 20:21:50< Crab_> yes, because advancing requires considering several units at a time
20101231 20:22:20-!- noy_ [~Noy@74.198.151.79] has joined #wesnoth-dev
20101231 20:22:20< skyfaller> ah, I see
20101231 20:22:20-!- noy_ [~Noy@74.198.151.79] has quit [Changing host]
20101231 20:22:20-!- noy_ [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20101231 20:23:36< skyfaller> well, maybe it is best to go for the low hanging fruit
20101231 20:23:45< skyfaller> and at least implement proper retreats when necessary
20101231 20:23:53< skyfaller> even if it can't advance / defend
20101231 20:23:59< Crab_> and handling that requires somewhat different algorithms, which will try to determine the best set of positions to hold to make a good 'front' with units protecting each other, minimizing the damage enemy can do and maximizing the enemy vulnerability if he attacks
20101231 20:24:29< Crab_> but making a 'global multi-unit retreat if needed' is certainly possible
20101231 20:24:36< skyfaller> is anyone working on that? defending / advancing?
20101231 20:24:39-!- noy [~Noy@wesnoth/developer/noy] has quit [Ping timeout: 260 seconds]
20101231 20:24:40-!- noy_ is now known as noy
20101231 20:25:07< skyfaller> ... I dunno if a truly "global" retreat would be desirable, that might make the AI too timid... usually a full retreat is necessary on one front but not the other
20101231 20:25:16< skyfaller> although if it taught the AI to conserve its units better it might be worthwhile anyway
20101231 20:25:43< Crab_> well, it can be 'global' but only to affect areas where the ai is at disadvantage
20101231 20:25:50< skyfaller> I guess it's an empirical question as to whether it is better to have an AI that attacks all the time or one that either fully attacks or fully retreats
20101231 20:26:06< skyfaller> ah, I see
20101231 20:26:10< Crab_> skyfaller: no. during the GCI, some ai bugs were fixed, and recruitment was improved, but there is no work on defence/holding.
20101231 20:26:47< Crab_> my work made in 2009 makes it possible to maintain multiple different AIs constructed from components, or even to change the AI configuration on the fly.
20101231 20:27:03< skyfaller> hm, the AI still does silly things like prefer HI when fighting Undead, so I wonder how much better recruitment has gotten, but I'm willing to accept that it was even sillier before
20101231 20:27:22< Crab_> skyfaller: it's not enabled yet
20101231 20:27:30< skyfaller> I guess HI make more sense when you don't have the concept of retreating
20101231 20:27:58< skyfaller> the advantage of Cavalry in that matchup is that they can run away when HI will be stuck at night and get nuked by DAs
20101231 20:28:23< skyfaller> well, anyway, I guess that's why I haven't seen the improvement, since it hasn't been turned on :)
20101231 20:28:32< Crab_> and it's possible to resolve the question empirically, (there's a script to test ai-vs-ai matchups to see which one is better)
20101231 20:28:56< Crab_> error handling needs to be improved in there, before it can be enabled. but it's not much work.
20101231 20:29:06< skyfaller> well, I guess my immediate suggestion is to try giving the AI the ability to do full retreats and see whether that makes it better :)
20101231 20:29:10< Crab_> the new code considers potential enemy counters when recruiting
20101231 20:30:04< Crab_> yes, I agree
20101231 20:30:53-!- ancestral [~ancestral@mobile-166-137-143-008.mycingular.net] has joined #wesnoth-dev
20101231 20:34:58-!- GNUtoo|laptop [~gnutoo@host7-149-dynamic.51-79-r.retail.telecomitalia.it] has quit [Quit: Program received signal SIGSEGV, Segmentation fault.]
20101231 20:46:54-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has joined #wesnoth-dev
20101231 20:46:59< skyfaller> Crab_: I'm a total programming noob and have little chance of being able to actually contribute code to this, but if I were interested in doing so, what ought I do to to help out?  are there particular bugs open for this?
20101231 20:49:51< Crab_> well, for one, even this conversation helped a bit. about code... it's better to help with non-coding questions like 'when we should retreat?' how far away we should retreat ?
20101231 20:50:11< Crab_> coding is not hard, in comparison to answering those questions
20101231 20:50:21< skyfaller> I see
20101231 20:50:40< skyfaller> well, the obvious low hanging fruit is time of day
20101231 20:50:45< skyfaller> look at your opponent's recruits
20101231 20:50:55< skyfaller> and consider your own units
20101231 20:51:06< skyfaller> and whether you will have an advantage at a given time of day
20101231 20:51:18< skyfaller> it's not good enough to consider factions
20101231 20:51:55< skyfaller> HODOR does much worse in daytime than dwarves, a saurian force is great at night while drakes are great at day
20101231 20:52:17< skyfaller> there are too many factions that can be flexible across different times of day depending on their recruiting patterns
20101231 20:52:40< Crab_> ok. if we're are all lawful and the enemy is all chaotic, when we should retreat and how far away? at dusk ? at first/second night ? a dawn ?
20101231 20:53:16< skyfaller> well, that's the trick... you really need the intermediate step of advancing without attacking
20101231 20:53:35< skyfaller> let's say you have two equal forces facing each other, neither has attacked yet
20101231 20:53:56< skyfaller> the lawful AI is 1p
20101231 20:54:43< skyfaller> the AI should advance at 2nd watch, but not attack... if the 2p chaotic enemy attacks at 2nd watch, then you get to respond at dawn
20101231 20:54:51< skyfaller> which is bad for the chaotic enemy
20101231 20:55:15< Crab_> we can do advance without attacking, it will be ok if we are stronger and will attack next turn
20101231 20:55:39< skyfaller> so what you're trying to do with an advance is tempt him into attacking when you'll be able to respond at a better time of day
20101231 20:55:54< Crab_> that is, we will advance not in the best defensive position possible, but it won't matter much because if the enemy responds he'll be in full combat in unfavourable ToD...
20101231 20:56:04< skyfaller> if the lawful AI is 2p, then you want to attack at 2nd watch, because if he hangs around to fight his response will be at dawn
20101231 20:56:22< skyfaller> so he either has to retreat despite your attack, or fight at a bad TOD
20101231 20:57:26< skyfaller> as for retreating, it's similar, it depends on whether you are 1p or 2p
20101231 20:58:16< Crab_> ok, we can assume we can code a function which will handle that (we know the ToD cycle and its bonuses, we know current turn and units)
20101231 20:58:21< Crab_> but how far to retreat ?
20101231 20:58:43< skyfaller> well, that's the thing, you need to be able to defend
20101231 20:59:07< skyfaller> because you want to be able to pull back to a position where if he attacks, he will have to do so from bad terrain, and from someplace closer to your keep so that you can develop a unit advantage
20101231 20:59:28< Crab_> we can't :) we can pick best hexes for particular units, and we know the enemy reach and firepower
20101231 20:59:32< skyfaller> if there is no good terrain around, you may just want to flee as hard as you can
20101231 20:59:45< skyfaller> if there is good terrain, you might want to pull a defensive line back to that
20101231 20:59:56< Crab_> we don't know about defensive lines )
20101231 21:00:21< skyfaller> because if you flee as hard as you can, you may run out of room to flee before TOD becomes favorable for you to attack again
20101231 21:00:42< Crab_> let's pretend that we know (1): it's time to retreat because its dusk, we're lawful p1, and our enemy is chaotic p2
20101231 21:01:02< Crab_> and we know all the units, their reach and firepower and potential (night) firepower.
20101231 21:01:24< Crab_> but we don't know anything at all about 'defensive lines'
20101231 21:01:39< Crab_> so, if we want the AI to form a 'defensive line', we need to teach it, how.
20101231 21:02:51< Crab_> simple thing like 'for each unit, retreat to safest place ignoring all other our units' might not work.
20101231 21:03:18< Crab_> we know everything about terrain and unit defenses on it
20101231 21:04:07< skyfaller> but we don't know about ZoC, right?
20101231 21:04:39< Crab_> we know how ZoC restricts enemy moves at any particular position of units.
20101231 21:04:53< skyfaller> you can't defend without understanding ZoC, it's hopeless... and you can't do ZoC well if you can't figure out what happens if you lose a crucial defending unit, but that's an ambitious project
20101231 21:05:05< skyfaller> even humans are bad at thinking that far ahead
20101231 21:05:21< Crab_> 'take this position, show me how enemy can move and attack' is implemented
20101231 21:06:01< skyfaller> so why is the AI famously bad at understanding ZoC?
20101231 21:06:19< Crab_> because it doesn't understand it, it just knows the current situation
20101231 21:06:34< Crab_> it doesn't think in 'unit groups'
20101231 21:07:42< Crab_> but to find a good defensive position you need to take several units and think 'ok, that one is there, the second one is here, the third one is here, let's see how the enemy can respond.. oh, we need to move this one farther away', like that.
20101231 21:08:51< Crab_> so, let's return to 'how far we should retreat'
20101231 21:09:15< Crab_> if we want to make retreat possible without investing much time into making defence work, what we can do ?
20101231 21:10:31-!- ancestral [~ancestral@mobile-166-137-143-008.mycingular.net] has quit [Quit: Smell ya later!]
20101231 21:13:57< skyfaller> sorry, multiple conversations
20101231 21:14:03< Crab_> all's ok
20101231 21:14:28< skyfaller> lemme think...
20101231 21:15:12-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has quit [Read error: Connection reset by peer]
20101231 21:15:46< skyfaller> I mean, the problem is that you want to be able to accept some minimum level of damage
20101231 21:16:02< skyfaller> it is frequently unnecessary to retreat out of the range of all enemy attackers
20101231 21:16:29< skyfaller> b/c if he comes in with only, say, one unit you may be able to massacre that one and then continue retreating after that
20101231 21:16:39< skyfaller> or shrug off the damage and keep running
20101231 21:16:46-!- Appleman1234 [~Appleman1@CPE-60-226-179-130.qld.bigpond.net.au] has joined #wesnoth-dev
20101231 21:17:02< skyfaller> retreating out of the range of all attackers may be too dramatic
20101231 21:17:22< skyfaller> and retreating as hard as you can, using all your move points, is probably too dramatic in most situations
20101231 21:17:38< skyfaller> you'll run out of room to run generally if you always retreat like that
20101231 21:18:07< skyfaller> I guess maybe you want to retreat hard enough that your enemy can attack you, but you probably won't lose any units or get any units trapped?
20101231 21:18:39< skyfaller> I'm not sure how to make the decision clear if we don't know how to defend
20101231 21:18:43< Crab_> ok, how to achieve that by moving 1 unit at a time ?
20101231 21:18:55< skyfaller> ah, I see
20101231 21:18:57< skyfaller> that's awful
20101231 21:19:04< skyfaller> I'm not sure it's feasible
20101231 21:19:08< Crab_> lets start with slowest units first..
20101231 21:19:37< Crab_> and pick a direction to flee (towards the keep or away from enemy or whatever)
20101231 21:19:59< skyfaller> can the AI figure out if a unit can be trapped?
20101231 21:20:20< Crab_> in a particular situation, yes
20101231 21:20:26< skyfaller> ok
20101231 21:20:43< skyfaller> you want your slowest units to run as slowly as necessary to avoid getting killed or trapped
20101231 21:21:05-!- noy [~Noy@wesnoth/developer/noy] has quit [Read error: Connection reset by peer]
20101231 21:21:07-!- fabi [~fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20101231 21:21:10< skyfaller> if they can be killed or trapped, then you need to decide to defend or abandon them
20101231 21:21:52< skyfaller> I mean, if they can always be killed or trapped no matter how hard they run, then you have that decision
20101231 21:22:23< skyfaller> we don't know how to defend, so you just have to abandon them and try to save the rest of your forces
20101231 21:22:41< Crab_> what about this approach : take a slow unit, assume you retreat it to some place, then try to form a defensive line 1 unit at a time retreating other units to sides or behind that line
20101231 21:23:01< Crab_> then, at the end, check how good our line is.
20101231 21:23:09< skyfaller> that sounds plausible
20101231 21:23:17< Crab_> if we don't like it, repeat from the start, assuming we run further away
20101231 21:23:33< skyfaller> that sounds almost intelligent
20101231 21:23:48-!- fendrin [~fabi@wesnoth/developer/fendrin] has quit [Ping timeout: 246 seconds]
20101231 21:23:49< skyfaller> how many times can the AI do this before the human gets tired of waiting and deletes Wesnoth? :)
20101231 21:24:12< Crab_> well, it doesn't look so expensive so far
20101231 21:24:46-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20101231 21:24:55< skyfaller> for the record, I'm "nelson" on the ladder/forums http://ladder.subversiva.org/profile.php?name=nelson
20101231 21:25:28< skyfaller> I'm an OK player but not a genius, someone like Dauntless might be able to give better AI training advice
20101231 21:25:30< Crab_> since, we are basically  trying to retreat 'around 1 hex, around 2 hexes, around 3 hexes, ... , around N hexes' where N is the movement rate, which is limited
20101231 21:26:14< Crab_> basically, constructing the line in a dump way and then evaluating the vulnerability of the end result
20101231 21:26:18< Crab_> *dumb
20101231 21:26:42< skyfaller> that sounds pretty good to me
20101231 21:26:51< skyfaller> it sounds like a huge improvement over whatever the AI is doing now
20101231 21:27:06< Crab_> ok, then this looks like the first thing to try - code that as a 'oops, everyone fall back' routine.
20101231 21:27:28< Crab_> and once it's ready it can be integrated in the turn sequence by using the ToD as a trigger
20101231 21:28:12< skyfaller> I think that numerical superiority also really needs to be considered
20101231 21:28:22< Crab_> well, that's the easy part compared to first
20101231 21:28:27< skyfaller> ok :)
20101231 21:28:43< skyfaller> so you won't end up with a situation where a couple grunts walk in at night and get creamed by 5 mages? :)
20101231 21:29:08< Crab_> ok, what is needed now is a good test case or two - situations (saves are ok) where the AI has a simple&good way to retreat.
20101231 21:29:52< skyfaller> hm, ok
20101231 21:29:59< Crab_> that is, I would be able to open up this save as the player who is supposed to retreat, droid the side and watch if the algorithm works correctly or not.
20101231 21:30:05< Crab_> can you help with that ?
20101231 21:30:25< skyfaller> sure, I need to go eat lunch now, but if you'll be around in an hour or two I could work with you on that
20101231 21:30:51< skyfaller> or if you want to draw up specs for what you want in a bug or something, I can work on it asynchronously
20101231 21:31:17< Crab_> ok. it's new year time here (1.5 h till new year), but, overall, I've got some free time till 6th January, so potentially I can code a prototype
20101231 21:31:24< skyfaller> ok, cool :)
20101231 21:31:33< Crab_> you can PM me (Crab) on wesnoth's forums
20101231 21:31:40< skyfaller> sounds good
20101231 21:32:22< skyfaller> It would be great to have an AI that is actually challenging to ladder players without a 2:1 unit advantage, that can train newbies to think more strategically
20101231 21:32:23< Crab_> the easiest way, probably, is to do a mock human-vs-human battle, and then save during the turn of the player who is supposed to retreat.
20101231 21:33:08< Crab_> to be a good example, the situation needs to have a 'very clean' solution, it should be obvious where to retreat
20101231 21:33:40< Crab_> and no units left behind, at first
20101231 21:34:37< skyfaller> hm, ok :)  Can you write this down on a forum post or bug report or something?  Somewhere where I can upload the saves in response?
20101231 21:35:03< Crab_> Forum PM allows attachments
20101231 21:35:24< Crab_> or I (or you) can make a topic, if you think that it'll be better that way.
20101231 21:35:54< skyfaller> I suppose it would be nice to make a topic, and invite public discussion
20101231 21:36:01< skyfaller> in case other people have ideas that we haven't considered
20101231 21:36:23< Crab_> yes, right
20101231 21:37:43< skyfaller> the interesting thing about your proposed solution is that it gets closer to implementing my "check" strategic position... if it turns out that the best solution is to retreat only zero spaces or no spaces, that is kind of like defending without retreating
20101231 21:38:13< skyfaller> if you can do something similar for advancing without attacking, that would fully implement "checking" I think
20101231 21:38:31< skyfaller> as long as the AI has somewhat competent defensive lines
20101231 21:38:33< Crab_> we'll see, it's better to move 1 thing at a time
20101231 21:38:44< skyfaller> however it arrives at them
20101231 21:39:10-!- fstltna [~fstltna@184-195-121-75.pools.spcsdns.net] has joined #wesnoth-dev
20101231 21:40:18< skyfaller> well, I nominate you to make a quick forum post about this, and I can expand on it for you when I get back from lunch :)
20101231 21:40:36< Crab_> doing it at the moment.
20101231 21:40:47< skyfaller> I'll hang out in this channel, and you can find me as nelson on the forum
20101231 21:40:48< skyfaller> groovy
20101231 21:40:50< skyfaller> ttyl :)
20101231 21:40:56-!- StealthyCoin [~StealthyC@adsl-99-35-218-30.dsl.pltn13.sbcglobal.net] has joined #wesnoth-dev
20101231 21:41:10< skyfaller> link me to the post when you have it
20101231 21:41:34< Crab_> http://forum.wesnoth.org/viewtopic.php?f=10&t=32688
20101231 22:02:23-!- StealthyCoin_ [~StealthyC@adsl-99-35-218-30.dsl.pltn13.sbcglobal.net] has joined #wesnoth-dev
20101231 22:02:23-!- StealthyCoin [~StealthyC@adsl-99-35-218-30.dsl.pltn13.sbcglobal.net] has quit [Read error: Connection reset by peer]
20101231 22:02:23-!- StealthyCoin_ is now known as StealthyCoin
20101231 22:06:13-!- PetePorty [~Pete@pc-8-253-120-200.cm.vtr.net] has quit [Quit: Saliendo]
20101231 22:22:56-!- neb1 [neb1@f79-118-255-149.pitesti.rdsnet.ro] has quit [Ping timeout: 265 seconds]
20101231 22:25:28-!- noy [~Noy@wesnoth/developer/noy] has quit [Read error: Connection reset by peer]
20101231 22:27:12-!- StealthyCoin [~StealthyC@adsl-99-35-218-30.dsl.pltn13.sbcglobal.net] has quit [Quit: StealthyCoin]
20101231 22:53:05< skyfaller> Crab_: I'm back from lunch, if you're still around :)  I'll see if I can create a few good savegames for you
20101231 23:00:17-!- Cookiee [~quassel@unaffiliated/cookiee] has joined #wesnoth-dev
20101231 23:03:18< Crab_> great, thanks. 1 or 2 will be enough
20101231 23:21:10-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20101231 23:33:00-!- Cookiee [~quassel@unaffiliated/cookiee] has quit [Read error: Connection reset by peer]
20101231 23:33:23< skyfaller> hm, it's actually hard creating a situation where the correct retreating pattern is obvious
20101231 23:34:03< skyfaller> frequently there will be a chance you can accomplish something by hanging around, such as a 14% ctk on the enemy leader
20101231 23:34:29< skyfaller> and you have to weigh the potential benefits of hanging around against the losses you'll take if you miss
20101231 23:36:04< Crab_> yes, that's why I've asked for help :)
20101231 23:36:14< Crab_> because it's a thing you'll do better than I.
20101231 23:36:17< skyfaller> hahaha I just did a leaderkill on myself
20101231 23:36:22< Crab_> :))
20101231 23:36:24< skyfaller> let's try this again
20101231 23:38:04< skyfaller> I guess what I want is a situation where you can't get your units in place for a proper attack before TOD turns against you
20101231 23:38:19< skyfaller> so you advance as far as you can, but then you are forced to retreat without having done anything
20101231 23:39:03< skyfaller> so I was trying Fallenstar Lake with HODOR vs. Loyalists, but my HODOR is actually able to do an OK job of rushing the enemy position
20101231 23:39:24< skyfaller> before daybreak
20101231 23:39:25< skyfaller> maybe orcs will be slower?
20101231 23:40:50< Crab_> depending on what you recruit, I guess.
20101231 23:41:34< Crab_> and yes, a 'natural' situation (arising naturally in a particular matchup on a particular map) is better
20101231 23:45:57-!- Upth [ogmar@adsl-75-26-166-84.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20101231 23:46:12-!- Upthorn [ogmar@adsl-75-26-166-84.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 255 seconds]
20101231 23:54:09-!- shadowmaster is now known as shadowm_deaf
20101231 23:55:37-!- shadowm_deaf is now known as shadowmaster
20101231 23:58:10< CIA-84> grickit * r48150 /trunk/data/lua/wml-tags.lua: Made [chat] no longer look up variables that match speaker. It's trivial to make speaker=$unit.name instead of speaker=unit. The previous implemention prevented users from using the string "unit" as a speaker name.
20101231 23:58:42< skyfaller> dammit, I'm too good at rushing
20101231 23:59:03< skyfaller> my orcs took a village
20101231 23:59:28< skyfaller> maybe I'll retreat them at day, but often it's worth losing a grunt to hold a village for a couple of turns
--- Log closed Sat Jan 01 00:00:05 2011