--- Log opened Thu Mar 12 00:00:20 2009
--- Day changed Thu Mar 12 2009
20090312 00:00:20-!- YogiHH [i=YogiHH@wesnoth/developer/yogihh] has quit ["Bye folks"]
20090312 00:02:03< ilor> wesbot: topic
20090312 00:02:05< wesbot> ilor: Don't be impatient, wait until something changes...  *tsk*
20090312 00:05:42-!- BenUrban [n=benurban@c-68-49-10-243.hsd1.md.comcast.net] has joined #wesnoth-dev
20090312 00:07:52< Shadow_Master> um there's an interesting change in the s11n dir between 1.5.10 and 1.5.11
20090312 00:11:15< Shadow_Master> ah. it isn't. I'm out of ideas. :-|
20090312 00:12:04-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 104 (Connection reset by peer)]
20090312 00:12:19-!- Turuk [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has joined #wesnoth-dev
20090312 00:14:45< Shadow_Master> I should probably create another user in my laptop, clone the git directory and start bisecting with help of a fresh ccache dir :-|
20090312 00:15:51-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has joined #wesnoth-dev
20090312 00:19:15-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Remote closed the connection]
20090312 00:24:40< Shadow_Master> well, I was already going to clone my user account anyway so it doesn't matter much.
20090312 00:26:34-!- Shadow_Master is now known as shadowmaster
20090312 00:27:20-!- shikadibot [n=1001@wesnoth/bot/shikadibot] has joined #wesnoth-dev
20090312 00:27:22-!- Azkar_ is now known as Azkar
20090312 00:31:01-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 00:31:24-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has quit [Client Quit]
20090312 00:39:38-!- Tigge [n=tigge@bacchus.olf.sgsnet.se] has joined #wesnoth-dev
20090312 00:42:10-!- ikarius [n=ross@smtp.gridironsystems.com] has quit []
20090312 00:48:38< Dragonking> Elvish_Pillager: Yes, it was my_moves.moves. I thought you know that... since you have it clearly stated when you use my_moves or write dir(my_moves)
20090312 00:49:13< Elvish_Pillager> Dragonking: I did know that at one point. It's easy to forget.
20090312 00:49:25< Elvish_Pillager> Why is there that extra level of complication?
20090312 00:50:22< Dragonking> Elvish_Pillager: No clue. It was like that since forever (since Sirp wrote it)
20090312 00:50:49< Elvish_Pillager> if there's not a particular reason for it, now's the best time to simplify it.
20090312 00:55:50-!- m4v3r1c [n=m4v3r1c@67.110.216.147.ptr.us.xo.net] has joined #wesnoth-dev
20090312 00:56:28-!- m4v3r1c [n=m4v3r1c@67.110.216.147.ptr.us.xo.net] has left #wesnoth-dev []
20090312 01:03:13-!- [Relic] [n=[Relic]@76.229.202.137] has joined #wesnoth-dev
20090312 01:04:08< [Relic]> Hello :)
20090312 01:05:26-!- shadowmaster is now known as Shadow_Master
20090312 01:05:42 * Shadow_Master throws a smoke bomb at [Relic] 
20090312 01:06:05 * [Relic] breaks out the strobe lights for a proper party
20090312 01:06:49< Shadow_Master> grr.
20090312 01:12:51-!- Sirp [n=me@wesnoth/developer/dave] has quit ["leaving"]
20090312 01:13:23-!- busfahrer` [n=busfahre@ulmg-5f70da84.pool.einsundeins.de] has joined #wesnoth-dev
20090312 01:22:23-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit [Read error: 145 (Connection timed out)]
20090312 01:22:36-!- Elvish_Pillage2 [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090312 01:28:39-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090312 01:30:54-!- Elvish_Pillage2 [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has quit ["Hi! I'm a quit message virus vaccine. If you see a quit message virus, don't replace your quit message with it!"]
20090312 01:30:57-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["night all"]
20090312 01:34:32< Ivanovic> i am off to bed now, n8
20090312 01:36:03-!- busfahrer` [n=busfahre@ulmg-5f70da84.pool.einsundeins.de] has quit [Read error: 145 (Connection timed out)]
20090312 01:36:37< Shadow_Master> su -c "cp /home/shadowm/src/wesnoth/.scons-options-cache ."
20090312 01:36:56< Shadow_Master> is .scons-option-cache supposed to be more intuitive than .scons-options-cache or what. :-|
20090312 01:39:05-!- Elvish_Pillager [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)]
20090312 01:40:35< loonycyborg> That was esr's idea to rename it to that. In all other my scons scripts it's called options.cache
20090312 01:40:46< Shadow_Master> aaaaaaaaaahhh.
20090312 01:50:30-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzz"]
20090312 01:51:11-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20090312 01:52:07-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090312 01:55:12-!- Dragonking [n=dk@wesnoth/developer/dragonking] has quit []
20090312 01:57:09< Shadow_Master> Running a bisect for bug  $13123 .
20090312 01:57:29< Shadow_Master> I don't know if I'll make it to the guilty commit before I end up using distcc through wifi.
20090312 02:19:26< Shadow_Master> pah. Wesnoth looks like crap on DirectFB.
20090312 02:21:25-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has quit ["switching to broken laptop"]
20090312 02:24:09-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Remote closed the connection]
20090312 02:29:30-!- shadowm_broken [n=ignacior@190.22.74.182] has joined #wesnoth-dev
20090312 02:31:08-!- Shadow_Master_ [n=ignacior@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 02:31:29-!- Shadow_Master_ is now known as shikadibot_
20090312 02:31:41-!- shikadibot_ [n=ignacior@wesnoth/developer/shadowmaster] has quit [Client Quit]
20090312 02:31:49-!- shikadibot_ [n=ignacior@190.22.74.182] has joined #wesnoth-dev
20090312 02:32:05-!- shikadibot_ [n=ignacior@190.22.74.182] has quit [Client Quit]
20090312 02:32:14-!- shadowm_broken [n=ignacior@190.22.74.182] has quit [Client Quit]
20090312 02:32:41-!- Blueblaze [n=Blueblaz@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090312 02:34:42-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 02:37:19-!- shikadibot [n=1001@wesnoth/bot/shikadibot] has quit [Read error: 104 (Connection reset by peer)]
20090312 02:37:42-!- shikadibot [n=1001@wesnoth/bot/shikadibot] has joined #wesnoth-dev
20090312 02:38:03< Shadow_Master> Turuk: btw, you can set ban lengths in the ACP if you didn't know.
20090312 02:38:27< Shadow_Master> or deadlines.
20090312 02:39:08< Turuk> I'm well aware
20090312 02:39:33< Turuk> I just did his without setting it besides permanent
20090312 02:39:42< Turuk> I figured I'd come back to it when he'd cooled his heels
20090312 02:45:36-!- Netsplit verne.freenode.net <-> irc.freenode.net quits: erl, cjhopman, Turuk, law_, ilor, dfranke
20090312 02:46:41-!- Netsplit over, joins: cjhopman, Turuk, law_, dfranke, erl
20090312 02:47:56-!- BenUrban [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
20090312 02:50:34-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090312 02:52:51-!- BenUrban [n=benurban@c-68-49-10-243.hsd1.md.comcast.net] has joined #wesnoth-dev
20090312 02:56:57-!- ilor_ [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090312 03:07:22-!- ilor [n=user@wesnoth/developer/ilor] has quit [Connection timed out]
20090312 03:14:46-!- Zen_Clark [n=user@99-136-80-191.lightspeed.rcsntx.sbcglobal.net] has quit ["ERC Version 5.3 (IRC client for Emacs)"]
20090312 03:20:27-!- happygrue [n=George@wesnoth/developer/wintermute] has quit ["http://xkcd.com/91/"]
20090312 03:23:59-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has quit ["n8"]
20090312 03:24:38-!- andreluiz [i=bd47030e@gateway/web/ajax/mibbit.com/x-380ff4e8797b7e46] has joined #wesnoth-dev
20090312 03:42:17-!- shikadibot [n=1001@wesnoth/bot/shikadibot] has quit [Read error: 110 (Connection timed out)]
20090312 03:44:42-!- andreluiz [i=bd47030e@gateway/web/ajax/mibbit.com/x-380ff4e8797b7e46] has quit [Read error: 110 (Connection timed out)]
20090312 04:18:37-!- Sirp [n=me@wesnoth/developer/dave] has joined #wesnoth-dev
20090312 04:26:06-!- Ivanovic_ [n=ivanovic@dtmd-4db2a8a3.pool.einsundeins.de] has joined #wesnoth-dev
20090312 04:27:22-!- Turuk_ [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has joined #wesnoth-dev
20090312 04:35:57-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 110 (Connection timed out)]
20090312 04:43:05-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 110 (Connection timed out)]
20090312 04:44:04-!- Ivanovic_ is now known as Ivanovic
20090312 05:18:38-!- noy [n=Noy@S01060016d3eab391.vs.shawcable.net] has joined #wesnoth-dev
20090312 05:32:27-!- Sapient [n=sapien-x@wesnoth/developer/sapient] has joined #wesnoth-dev
20090312 05:32:32< Sapient> yo ilor
20090312 05:33:22< Sapient> I assume you solved that little formula problem so you don't need to ask me anything at this point?
20090312 05:34:18< Sapient> FYI... I will be on vacation till Monday so I won't be around this weekend
20090312 05:34:21-!- [Relic] [n=[Relic]@76.229.202.137] has quit ["Leaving"]
20090312 05:34:58< Sapient> If there are any bugs assigned to me and it needs to get fixed soon then you might want to have someone else look at it
20090312 05:36:56< Sapient> as far as I know, there's only that one that Soliton recently assigned to me
20090312 05:45:32< Sapient> cya later, have fun.
20090312 05:45:32-!- Sapient [n=sapien-x@wesnoth/developer/sapient] has left #wesnoth-dev []
20090312 05:58:26-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090312 06:10:34-!- noy [n=Noy@S01060016d3eab391.vs.shawcable.net] has joined #wesnoth-dev
20090312 06:11:06-!- noy [n=Noy@wesnoth/developer/noy] has quit [Connection reset by peer]
20090312 06:11:15-!- noy [n=Noy@S01060016d3eab391.vs.shawcable.net] has joined #wesnoth-dev
20090312 06:19:49-!- xonev [n=chatzill@59.92.46.175] has joined #wesnoth-dev
20090312 06:40:44-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090312 06:45:21-!- vivri [n=chatzill@CPE001839c50068-CM00080d747884.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20090312 06:47:09< vivri> hi, i'm new here. wanted to see whether you'll be taking part in GSoC this year
20090312 06:48:25< Blueblaze> vivri, I believe so
20090312 06:49:13< vivri> Blueblaze, cool. I looked at some of your ideas from last year, there's one I'm particularly interested in
20090312 06:50:06< vivri> Blueblaze, i'm wondering in what condition your ai player is? I read some documentation, but i'm suspecting it's dated
20090312 06:50:13< vivri> *outdated
20090312 06:50:26< Sirp> vivri: we are hoping to participate. Organizations who have been accepted by Google to participate will be announced next Wednesday
20090312 06:51:11< vivri> Sirp, thanks. i'm aware of that, just doing some heads-up before the flood comes ;)
20090312 06:53:36< vivri> it's kinda late here... anyone wants to chat bout ai for a bit or should i drop in later?
20090312 07:05:20-!- vivri [n=chatzill@CPE001839c50068-CM00080d747884.cpe.net.cable.rogers.com] has quit ["cheers"]
20090312 07:12:01-!- Blueblaze [n=Blueblaz@c-98-199-143-139.hsd1.tx.comcast.net] has quit []
20090312 07:23:57-!- xonev_ [n=chatzill@59.92.46.175] has joined #wesnoth-dev
20090312 07:42:17-!- xonev [n=chatzill@59.92.46.175] has quit [Read error: 110 (Connection timed out)]
20090312 08:13:34-!- Sirp [n=me@wesnoth/developer/dave] has quit ["leaving"]
20090312 08:36:41-!- ilor_ [n=user@wesnoth/developer/ilor] has quit [Read error: 104 (Connection reset by peer)]
20090312 08:36:53-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090312 08:40:11-!- Netsplit verne.freenode.net <-> irc.freenode.net quits: nital
20090312 08:48:29-!- nital [n=nital@mion.elka.pw.edu.pl] has joined #wesnoth-dev
20090312 08:51:13-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20090312 08:55:46-!- ilor [n=user@wesnoth/developer/ilor] has quit [Read error: 113 (No route to host)]
20090312 09:01:08-!- Tigge [n=tigge@bacchus.olf.sgsnet.se] has quit [Read error: 60 (Operation timed out)]
20090312 09:01:19-!- Tigge [n=tigge@bacchus.olf.sgsnet.se] has joined #wesnoth-dev
20090312 09:11:39-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has quit [Remote closed the connection]
20090312 09:12:58< CIA-70> dragonking * r33566 /trunk/src/formula.cpp: Fixed formula . operator bug
20090312 09:43:15-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090312 09:48:02-!- xonev [n=chatzill@59.92.43.27] has joined #wesnoth-dev
20090312 09:51:15-!- kawoi [n=chatzill@public13614.xdsl.centertel.pl] has joined #wesnoth-dev
20090312 09:51:56< kawoi> hey, if anyone of devs is online, you may want to check wesnoth 1.5 lobby
20090312 09:52:20< kawoi> there is a game "survival DVDR9" or something, that I can't join without assertion fail
20090312 09:52:37< kawoi> using Windows Vista, wesnoth 1.6-rc2
20090312 09:52:52< kawoi> fresh install, no addons installed
20090312 09:53:18< Soliton> what does the assertion say?
20090312 09:53:32< kawoi> something about gui, let me check again
20090312 09:54:45< Soliton> you get the assert right when you join?
20090312 09:55:08< Soliton> did you install over a previous version?
20090312 09:56:06< kawoi> http://img7.imageshack.us/img7/5066/errordtl.png
20090312 09:56:21< kawoi> nope, fresh laptop, fresh windows, fresh wesnoth
20090312 09:56:28-!- xonev_ [n=chatzill@59.92.46.175] has quit [Read error: 145 (Connection timed out)]
20090312 09:56:54< kawoi> just bought a laptop yesterday and thought that i can test wesnoth on it before removing it and installing linux
20090312 09:57:01< Soliton> so was it asserts right when you join? i can't reproduce that.
20090312 09:57:09< Soliton> s/was //
20090312 09:57:48< kawoi> well, it's reproducible for me, what can I do to provide useful information?
20090312 09:58:07< Soliton> that error message is good enough.
20090312 09:58:37< Soliton> please make a bug report with it and tell that it happens with the dark forecast scenario.
20090312 09:59:16< Soliton> i guess you're screen resolution might be relevant, too.
20090312 10:10:39< kawoi> hey, Soliton, still around?
20090312 10:10:58< kawoi> since I think I've found one more relevant information
20090312 10:11:17< kawoi> the bug doesn't reproduce for me if I turn the "English [US]" language on
20090312 10:11:29< Soliton> ah, good to know.
20090312 10:11:33< kawoi> if I switch it back to my mother tongue, "Polski", that is
20090312 10:11:37< kawoi> it reproduces again
20090312 10:11:47< Soliton> probably an issue with line length of some text.
20090312 10:14:32< CIA-70> ivanovic * r33567 /trunk/ (8 files in 7 dirs): updated Finnish translation
20090312 10:14:33< Ivanovic> moin
20090312 10:16:25< kawoi> is it good enough? https://gna.org/bugs/index.php?13180
20090312 10:17:15< Soliton> excellent, thanks.
20090312 10:17:30< kawoi> np, cya
20090312 10:17:31-!- kawoi [n=chatzill@public13614.xdsl.centertel.pl] has quit ["ChatZilla 0.9.84 [Firefox 3.0.7/2009021910]"]
20090312 10:22:30-!- xonev_ [n=chatzill@59.92.57.234] has joined #wesnoth-dev
20090312 10:25:15< Soliton> cjhopman: is this something you would know about? https://gna.org/bugs/index.php?13174
20090312 10:25:49< Soliton> cjhopman: i think this is about a unit in the unit_map that is not on the map.
20090312 10:28:30-!- xonev [n=chatzill@59.92.43.27] has quit [Read error: 110 (Connection timed out)]
20090312 10:28:43-!- xonev_ is now known as xonev
20090312 10:37:19-!- stikonas [n=stikonas@wesnoth/translator/stikonas] has joined #wesnoth-dev
20090312 10:41:25-!- loonybot [n=loonybot@79.139.247.203] has joined #wesnoth-dev
20090312 10:42:02-!- loonycyborg [n=sergey@79.139.247.203] has joined #wesnoth-dev
20090312 11:10:23-!- stikonas [n=stikonas@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20090312 11:24:36< Soliton> does someone else also get odd "press_enter.gz" savegames from time to time?
20090312 11:25:13 * zookeeper hasn't
20090312 11:25:53< Soliton> just now it's a simple text file with "böö" in it... o.O
20090312 11:26:27< zookeeper> sapient, it seems that [effect] apply_to=attack special=magical doesn't work, but you must use special=chance_to_hit instead. i think the special id should be used there instead of the tag name (that's the way it works or is supposed to work elsewhere)
20090312 11:26:47< zookeeper> huh?
20090312 11:27:31< zookeeper> that's the equivalent of "boo!" in finnish, maybe someone haxed your wesnoth :P
20090312 11:28:11< Soliton> well, clearly a finnish guy then!
20090312 11:28:16< zookeeper> uh oh
20090312 11:28:18< zookeeper> nevermind
20090312 11:28:44< zookeeper> don't you think suokko would be more likely ;)?
20090312 11:29:03< Soliton> more likely than who?
20090312 11:29:53< Soliton> Shadow_Master: see bug #13177 for iftu. you probably know already though.
20090312 11:29:57< zookeeper> no one
20090312 11:30:52< Soliton> "I need in STATISTICS the numbers (especially losses) of the current scenario and of the whole campain."
20090312 11:32:13-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 11:32:36-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 11:39:24< Ivanovic> Soliton: we talked about bug #13174 in here yesterday
20090312 11:39:35< Ivanovic> will paste the respecitve part of the log in the report
20090312 11:43:44< loonycyborg> Soliton: I have a 'press enter' save game too
20090312 11:47:27< Soliton> does it have suspicious finnish words in it?
20090312 11:48:06< loonycyborg> Nope.
20090312 11:52:59-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090312 12:12:22-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090312 12:14:24-!- Dragonking [n=dk@wesnoth/developer/dragonking] has joined #wesnoth-dev
20090312 12:44:03-!- fabi [n=fabi@wesnoth/developer/fendrin] has quit [Remote closed the connection]
20090312 12:57:40< Rhonda> dfranke: http://lists.debian.org/debian-security-announce/2009/msg00047.html :)
20090312 13:05:15-!- silene [n=plouf@nap13-6-88-179-16-10.fbx.proxad.net] has joined #wesnoth-dev
20090312 13:12:04< Soliton> seems the AI is quite happy to attack 1XP-to-level units now even with no chance to kill and no one else around. maybe for the better in campaigns though.
20090312 13:17:59< Dragonking> Soliton: Any info about how AI recruitment in MP games is doing?
20090312 13:18:33< Soliton> Dragonking: nope. haven't tested it.
20090312 13:19:32< Dragonking> ok
20090312 13:33:04< Soliton> well, actually in a recent test game i was surprised the undead AI recruited no adepts against me.
20090312 13:33:29< Soliton> i was knalgan alliance i think.
20090312 13:33:30-!- esr [n=chatzill@static-71-162-243-5.phlapa.fios.verizon.net] has quit ["ChatZilla 0.9.84 [Firefox 3.0.5/2008121623]"]
20090312 13:36:22< Soliton> anyobdy mind if i remove dfool and sample ai from the available ais? they're not really useful for anyone except they want to write their own ai and then they#ll find them anyway i hope..
20090312 13:41:32< Soliton> Dragonking: how can i tell if formulaAI is recruiting? from the log it looks to me like it's not.
20090312 13:42:21< Dragonking> Soliton: Really?
20090312 13:42:49-!- esr [n=chatzill@static-71-162-243-5.phlapa.fios.verizon.net] has joined #wesnoth-dev
20090312 13:44:29< Soliton> on a map with statues the AI checks new recruits against the statues...
20090312 13:44:48< Soliton> i guess i can forget testing recruitment on caves of the basilisk.
20090312 13:45:27< Soliton> undead vs drakes and it got 2 ghost and the rest skeletons.
20090312 13:45:31< Dragonking> heh... yes.. I forgot about that when writing it... somehow we should make statues as friendlu side for AI purposes I think
20090312 13:45:57-!- sk- [n=sk@pool-71-242-206-125.phlapa.east.verizon.net] has joined #wesnoth-dev
20090312 13:46:32< Soliton> we should get rid of them or add more checks for stoned units.
20090312 13:49:31< Soliton> ok, similar recruitment on onis except some skeletons are archers. no adepts.
20090312 13:50:52< Dragonking> Seems that AI overestimates ulfserkers
20090312 13:51:21< Soliton> i'm drake.
20090312 13:51:27< Soliton> i recruited all fighters.
20090312 13:51:38< Dragonking> oh
20090312 13:51:41< Dragonking> hmr
20090312 13:52:30< Soliton> is it only checking against what i can recruit or also against what i actually did recruit?
20090312 13:53:04< Soliton> i'm still not convinced the formulaAI is even recruiting though.
20090312 13:53:07< Dragonking> Depends on how much gold you have
20090312 13:53:27< Dragonking> I'm checking that
20090312 13:56:51< Soliton> the ai seems quite aggressive, btw.
20090312 13:57:15< Dragonking> In terms of gameplay?
20090312 13:57:43< Soliton> yeah.
20090312 13:57:58< Soliton> not related to formulaAI just saying.
20090312 13:58:21< Dragonking> ok
20090312 13:59:35< Dragonking> Well, I don't think that this recruting is done by formula AI...
20090312 14:00:00< Dragonking> Just added some debug message lines to formula and nothing is dispayed
20090312 14:01:42< Dragonking> ahahah
20090312 14:01:44< Dragonking> I think I know
20090312 14:02:17< Dragonking> There should be formula_ai besides default to chose from, then formula recruting should work 
20090312 14:02:29< Soliton> there isn't.
20090312 14:02:35< Dragonking> Yup
20090312 14:02:42-!- mjs-de [n=mjs-de@vpw.wh.Uni-Dortmund.DE] has joined #wesnoth-dev
20090312 14:02:50< Soliton> info ai: ai_manager::get_ai() for algorithm: formula_ai
20090312 14:03:01< Soliton> but it doesn't seem to do anything.
20090312 14:04:43< Dragonking> ok, I will make simple test
20090312 14:08:29< Dragonking> Soliton: formula AI does not recruit - I replaced recruting formula with formula that ends turn - and it just recruited as usual
20090312 14:09:20< Dragonking> That just proves 2 things:  first we need to make recruitment work to test it  second: default AI recruting really sux 
20090312 14:10:55< Soliton> yes, adept and skeleton archer are the same usage so it should have picked the adept there.
20090312 14:11:10< Soliton> maybe it values resistances too much.
20090312 14:11:47< Soliton> either way as said before you really can't do much wrong with letting formulaAI do the recruitment. :-)
20090312 14:12:24< Soliton> Dragonking: src/ai.cpp:1900 seems to suggest that FAI is called.
20090312 14:13:26< Dragonking> hmmm
20090312 14:13:39< Dragonking> Let me check something...
20090312 14:14:02< Soliton> well, the log message there isn't printed though so the call returns false.
20090312 14:16:57< Dragonking> ok, something happened
20090312 14:17:01< Soliton> default AI does recommend archer and adept which is fine so i guess the AI is just unlucky.
20090312 14:19:30< Dragonking> Soliton: I managed to make formula run once and create maps with unit values 
20090312 14:19:42< Dragonking> Drake vs elf (nothing recruited) result: http://wesnoth.pastebin.com/m2787dcc6
20090312 14:20:56< nital> http://wesnoth.pastebin.com/m2787dcc6
20090312 14:21:09< nital> sorry, missclick
20090312 14:21:10< nital> http://wesnoth.pastebin.com/m2787dcc6
20090312 14:21:21 * Soliton blinks.
20090312 14:21:47< Soliton> Dragonking: clasher is a bad defender, huh? :-)
20090312 14:21:56< Dragonking> Yea, I wonder about that too ;)
20090312 14:22:03< Dragonking> Prolly cause of lack of retaliation damage
20090312 14:22:20< Dragonking> Against archers
20090312 14:22:21< Soliton> i don't htink that's very important for a defender.
20090312 14:22:28< Dragonking> Yup
20090312 14:22:47< Dragonking> Still, I see where the problem is with formula :/
20090312 14:23:03< Soliton> you do?
20090312 14:23:18< Dragonking> Problem why it is not recruting
20090312 14:23:29< Dragonking> Not why clasher is scored poorly ;p
20090312 14:23:43< Dragonking> I made this formula to be part of bigger AI
20090312 14:24:15< Soliton> so the recruitment can't be run on its own?
20090312 14:24:53< Dragonking> Thing is: it can, if whole sde is controlled by formula_ai, thne it recruits, and fallbacks to default ai
20090312 14:25:15< Dragonking> But currently it is messed.
20090312 14:26:29-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Connection reset by peer]
20090312 14:26:30< Dragonking> I added for myself formula AI to the AI list but still no luck with running it
20090312 14:26:45-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 14:29:07< Dragonking> I'm afraid that it is still fault of this AI mess that took place some time ago... :/
20090312 14:30:32< Soliton> well, what seems to be the problem?
20090312 14:31:03< Soliton> afaict formula ai is called so i'm not sure how that relates to any mess in the default ai.
20090312 14:33:08< Dragonking> That's what a proboel is :/
20090312 14:33:17< Dragonking> Soliton: Run *any* campaign
20090312 14:33:18< Soliton> i guess that master_ bool might not be set. checking that.
20090312 14:33:21< Dragonking> And you get "ai_manager::get_ai() for algorithm: formula_ai"
20090312 14:34:04< Dragonking> I suspect when suokko tryed to connect formula AI with default one he commited all this..
20090312 14:34:19< Dragonking> But I am not sure, it could be that it was there all the time
20090312 14:34:24< Soliton> the ai manager is not from suokko.
20090312 14:34:37< Dragonking> Yes, it is form barbarianhero AFAIR
20090312 14:35:35< Dragonking> The thing is, if we now specify [ai] recruit=formula [/]   it will be called every turn
20090312 14:35:52< Dragonking> And that's how I ran now my formula
20090312 14:36:12< Dragonking> But like I said, it is not how it is supposed to be ran, so it doesn't work this way
20090312 14:37:15< Soliton> well, how is it supposed to work?
20090312 14:37:49< Dragonking> As a controller of a whole side
20090312 14:38:00< Dragonking> And I'm trying to make it work like that now
20090312 14:41:13-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 14:41:56-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 14:42:36-!- happygrue [n=George@wesnoth/developer/wintermute] has joined #wesnoth-dev
20090312 14:46:24-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 14:46:52-!- Netsplit verne.freenode.net <-> irc.freenode.net quits: loonycyborg
20090312 14:47:03-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 14:49:05-!- Netsplit over, joins: loonycyborg
20090312 14:51:01-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 14:51:52-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 14:52:52< Dragonking> Seems that default AI is always calling formula AI and looking if it wants to recruit
20090312 14:53:45< Soliton> yep, but the recrui_formula_ pointer is NULL.
20090312 14:53:57< Dragonking> Yes
20090312 14:53:59-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 14:54:19< Dragonking> Because it is used incorrectly in the cfg
20090312 14:54:36-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 14:57:04< Soliton> so how should it look like?
20090312 14:57:27-!- BenUrban [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
20090312 14:57:32< Soliton> [ai] recruit=recruitment.fai?
20090312 14:58:27< Dragonking> recruit="{ai/formula/recruitment.fai}"    but Like I said - thi formula is not made to be run like that, so it will be evaluated, but no recruiting will be made
20090312 15:01:34< Dragonking> The problem is that apparently I can't choose other AI than default in MP combo box, even after modyfing gat_available_ais()
20090312 15:01:39< Dragonking> Or I do sometihng wrong
20090312 15:03:14< Soliton> well, i just disabled dfool and sample ai and that worked. :-)
20090312 15:04:57< zookeeper> yay
20090312 15:05:05< Dragonking> Even if I add formula_ai to the list, and then chose it, still default AI is requested first
20090312 15:05:31< Dragonking> Soliton: I'll need to disappear soon (classes)
20090312 15:05:47< Dragonking> But I'll be back around evening, and ofr sure this thing has priority for now
20090312 15:05:51< Soliton> well, that's quite a different proble then.
20090312 15:06:29< Dragonking> IMO it should work like that:
20090312 15:06:33< Soliton> i chose formula ai now and the recruit formula pointer is still missing.
20090312 15:06:46< Soliton> i guess that is expected?
20090312 15:07:08< Dragonking> If recruit= does not exist - then yes
20090312 15:07:27< Soliton> right, ok.
20090312 15:08:19< Dragonking> Soliton: current way of how recruitment is defined in faction config should make it work if we somehow manage to make fomrula AI control side in MP game
20090312 15:08:41< Dragonking> So if "formula AI" option form menu list work.
20090312 15:09:42< Soliton> yeah, i got that now.
20090312 15:10:05< Dragonking> Thne it just creates formula AI object, parse everyting, nicely run recruitment.fai as team formula rulebase, do recruitment, and calls default AI to take over.
20090312 15:10:19< Dragonking> That how it is working in formula_recruitment test scenario
20090312 15:10:51< Soliton> and there you have ai_algorithm=formula_ai?
20090312 15:11:20< Dragonking> Yes.
20090312 15:11:45< Soliton> ok, maybe i can work it out then.
20090312 15:13:29< Dragonking> Soliton: heh
20090312 15:13:34< Dragonking> YOu're genious.
20090312 15:13:39< Dragonking> I totally forgot.
20090312 15:13:52< Soliton> ok...
20090312 15:13:55< Dragonking> Just add ai_algorithm=formula_ai   to every faction config 
20090312 15:14:04< Soliton> well, yes.
20090312 15:14:19< Dragonking> And it works.
20090312 15:14:33< Soliton> but.. yeah, i suppose that's what we really want.
20090312 15:14:53-!- Netsplit verne.freenode.net <-> irc.freenode.net quits: loonycyborg
20090312 15:15:04-!- Netsplit over, joins: loonycyborg
20090312 15:15:04< Soliton> ok, i'll do that.
20090312 15:15:20< Soliton> and i'll remove dfool and sample ai but not add formula ai yet.
20090312 15:15:29< Dragonking> Sure, don't add it
20090312 15:15:51< Dragonking> It was my mistake to try to add it instead of directly put this one lin into cfg :/
20090312 15:15:55< Dragonking> line'
20090312 15:18:08< Dragonking> Soliton: Now you should see some dark adepts. :)
20090312 15:18:33< Soliton> ok, i'll test in a bit.
20090312 15:18:48-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20090312 15:20:07 * Dragonking should prolly get some sleep instead of going to classes.. 
20090312 15:20:51< Soliton> sleep in class?
20090312 15:21:15< Dragonking> Laboratories in front of computers, difficult to sleep there :)
20090312 15:21:39< Dragonking> Soliton: Before I leave I'll add some changes to make formula print out scores that it assigns to units, it may be helpful to determine how "good" it is
20090312 15:21:55< Soliton> yep, that'd be nice.
20090312 15:24:04-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 15:24:50-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 15:27:11< CIA-70> soliton * r33568 /trunk/src/ai.cpp: uncomment dfool and sample ai since they're not really useful to the user
20090312 15:27:48-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 15:28:00-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090312 15:28:35-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 15:30:14< CIA-70> dragonking * r33569 /trunk/data/ai/formula/recruitment.fai: Some info for debugging purposes
20090312 15:30:19< Dragonking> Soliton: ok, commited, this is how it shoud look like and what does it mean: http://wesnoth.pastebin.com/m3c6ecc35
20090312 15:32:59< sk-> Okay I submitted a bug about the AI's lack of movement. The saved game I attached is a sight to behold. But I have one other question...
20090312 15:34:20< sk-> How come when I add special tiles from Under the Burning Suns the map won't load?
20090312 15:34:39< sk-> Using the map editor, that is.
20090312 15:34:51< zookeeper> because those are UtBS tiles
20090312 15:34:53< Soliton> are you playing utbs?
20090312 15:34:53< Ivanovic> because it is a special terrain that is *only* available in UTBS
20090312 15:35:08< sk-> oh
20090312 15:35:34< sk-> Why are they included in the general map editor then?
20090312 15:35:48< sk-> It isn't obvious that they aren't usable except in special cases.
20090312 15:35:54< Soliton> because the general map editor is used to edit utbs maps.
20090312 15:35:58< Ivanovic> to allow editing utbs maps
20090312 15:36:46< sk-> Okay.
20090312 15:37:47-!- Turuk_ [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has quit [Read error: 110 (Connection timed out)]
20090312 15:41:26< Soliton> Dragonking: now the ai is not moveing though.
20090312 15:41:37< Dragonking> um, really?
20090312 15:42:02-!- stikonas [n=stikonas@ctv-213-164-100-13.vinita.lt] has joined #wesnoth-dev
20090312 15:42:37< Soliton> well, or it put all the units nicely back into the castle when i wasn't looking.
20090312 15:42:48< Dragonking> It works for me
20090312 15:44:14< Dragonking> ok sec
20090312 15:44:28< Soliton> and you added ai_algorithm to the factions you're playing?
20090312 15:44:32< Dragonking> No ;p
20090312 15:44:45< Dragonking> Just to one
20090312 15:44:58< Soliton> i see "info formula_ai: do move..." but nothing happens.
20090312 15:45:21< Soliton> then it tries to recruit in a full castle and that's that.
20090312 15:46:33< Dragonking> Does it run in infinite loop?
20090312 15:46:35< Dragonking> Maybe a bug
20090312 15:46:45< Soliton> nope it ends its turn.
20090312 15:47:57< Dragonking> Seriously, it works for me
20090312 15:48:21< Soliton> hmm, ok.
20090312 15:48:53< Dragonking> Here, AI recruits, fallbacks and moves
20090312 15:49:32< Ivanovic> sk-: what happens when not setting any of the [AI] params?
20090312 15:50:08< sk-> I'll try that right now.
20090312 15:50:13< Ivanovic> with other words: most campaigns do not use [AI] blocks and work, maybe something is going on in this direction?
20090312 15:50:21< Soliton> *sigh* [bug #13185] Wrong units in default Fractions
20090312 15:50:25< Ivanovic> as in "if you want it to work, just don't try to control the AI"
20090312 15:51:31< sk-> the problem is that if this is the case, survivals kind of demand maximum aggression.
20090312 15:51:36< sk-> but I'll try it out
20090312 15:53:16-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit [Read error: 110 (Connection timed out)]
20090312 15:53:16< zookeeper> Soliton, lolz, great report :P
20090312 15:53:43< Dragonking> Soliton: :D
20090312 15:54:08< Dragonking> ok, I disappear for now, bye
20090312 15:56:53< AI0867> wesbot: bug #13185
20090312 15:56:54< wesbot> Bug #13185  Assigned to: None  Status: Invalid  Priority: 5 - Normal
20090312 15:56:57< wesbot> Summary: Wrong units in default Fractions
20090312 15:57:00< wesbot> Original submission: While the dwarfs miss out on the Copter, Runesmith, and s
20090312 15:57:03< wesbot> teamwheel they have got thiefes in their Fraction.
20090312 15:57:06< wesbot> URL: https://gna.org/bugs/?13185
20090312 15:57:44< Soliton> seriously, who messed up the factions like that!?
20090312 15:57:48< sk-> Ivanovic, That does not resolve the issue.
20090312 15:57:57< sk-> I get the same result as in the saved game attached to the bug.
20090312 15:58:17< Ivanovic> interesting...
20090312 15:58:25< Soliton> that's quite weird since in a normal mp game the ai attacks as usual.
20090312 15:58:51< sk-> That's why I thought it might have to do with can_recruit=1 but that isn't it.
20090312 16:00:08-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 104 (Connection reset by peer)]
20090312 16:00:37-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:00:37< Soliton> mordante: http://www.wesnoth.org/forum/viewtopic.php?p=344773#p344773
20090312 16:07:03< esr> zookeeper: I tried to fix the missing EI journey tracks yesterday by replacing the [switch] parts with [if]/[then]/[else], but they're still not showing.  Would you please look at those bits in The_Eastern_Invasion/utils/journey.cfg and tell me if anything looks obviously wrong in the WML?
20090312 16:08:30< Soliton> Dragonking: well, it doesn't work here. maybe you just reloaded a game and didn't start a new one?
20090312 16:09:59< sk-> btw, I added the MP scenario "Skullduggery" to the server for testing, if needed.
20090312 16:11:55< Ivanovic> WOW
20090312 16:11:58< Soliton> the smaller the test case the better.
20090312 16:12:09 * Ivanovic sees something that looks like real bullshit in the file team.cpp
20090312 16:13:20< Ivanovic> someone please have a look eg at lines 235 to 246
20090312 16:14:05< esr> Ivanovic: Looking...
20090312 16:15:01< Ivanovic> short guess: the cause for some of the "AI strangeness" might come from this area...
20090312 16:15:41< esr> Ivanovic: None of it looks obviiously wrong to me.  What are you seeing?
20090312 16:15:51< Ivanovic> http://pastebin.com/m2459a411
20090312 16:16:01< Ivanovic> oh, not really wrong per se, but "dead code"
20090312 16:16:23< Ivanovic> short guess: revision 29168 is to be blamed for some bogus!
20090312 16:16:45< zookeeper> esr, frankly all that journey WML is pretty obscure to me.
20090312 16:17:03< zookeeper> i don't see anything wrong with the [if]s though
20090312 16:17:13< Ivanovic> http://svn.gna.org/viewcvs/wesnoth?rev=29168&view=rev
20090312 16:17:29< esr> So, all it means is the defaults get set from some global_params item rather than hardcoded constants.  Have you tried to find out where that ting is set?
20090312 16:17:55< Ivanovic> that is one of the problems, i have not really found those global defaults
20090312 16:18:22< esr> zookeeper: Thanks, that means there's some subtle bad interaction with [part] or [story] going on.  Dammit.
20090312 16:18:48< esr> There might be an engine bug lurking here.
20090312 16:19:09< zookeeper> esr, hmm, i think [if] only works in [story], not in [part].
20090312 16:19:30< esr> zookeeper: Checking...
20090312 16:21:43< happygrue> Ivanovic: what development version first saw that revision?  I am not sure how to find that easily...
20090312 16:21:58< happygrue> ah, 1.5.4?
20090312 16:21:59< Ivanovic> happygrue: have a look at the webinterface for the tags
20090312 16:22:25< Ivanovic> jupp, 1.5.4
20090312 16:23:50< happygrue> that seems to be consistant with what I found testing a save file, since 1.5.3 is the first place (that I saw) where the AI starts acting 'ok'
20090312 16:24:07-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 16:24:12< happygrue> though I didn't check 1.5.4, by 1.5.5 and later things are not so good
20090312 16:24:20< happygrue> *but by 1.5.5
20090312 16:24:29< Ivanovic> good, so this *might* be the cause
20090312 16:24:38< Soliton> what exactly is the problem with that diff?
20090312 16:24:45-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:24:53< happygrue> there were some other changes in 1.5.5 that probably didn't help, but yeah - could be!
20090312 16:25:01< Ivanovic> Soliton: eg this line:  caution_ = 1.0;
20090312 16:25:16< Ivanovic> though it is set to a completely different value further up
20090312 16:25:25< Ivanovic> caution_(0.25),
20090312 16:25:43-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 16:26:04-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:27:12-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Remote closed the connection]
20090312 16:28:02< Soliton> that would explain the complete opposite of what we seem to see.
20090312 16:28:32< Soliton> but people complain about scenarios where they do that all those parameters.
20090312 16:28:40< Soliton> s/that/set/
20090312 16:29:08< Ivanovic>  // If are starting new scenario overide settings from [ai] tags
20090312 16:29:10< Ivanovic> ;)
20090312 16:30:04< Soliton> no idea what you're trying to say.
20090312 16:30:21< Ivanovic> part of the commit is to "always use defaults"
20090312 16:30:25< Ivanovic> at least it reads like this
20090312 16:32:09-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:32:23< Soliton> i can't see that.
20090312 16:32:45< Soliton> it tries to find the settings at various places and if they're all empty goes back to defaults.
20090312 16:33:06< Soliton> it's odd that those are different from the initial values though.
20090312 16:34:39-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 16:35:30-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:37:39-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 16:38:19-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:40:19-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 16:40:41-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20090312 16:42:41< Ivanovic> okay, lets see if i will be able to set the values to the real defaults (and have them in #define blocks for easier editing later on)
20090312 16:43:01-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:44:12-!- Blueblaze [n=Blueblaz@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090312 16:47:21< Soliton> what are the real defaults?
20090312 16:47:50-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 16:48:52-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 16:53:36< Ivanovic> the ones listed in the wiki i'd assume
20090312 16:53:40< Ivanovic> http://www.wesnoth.org/wiki/AiWML
20090312 16:54:15-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 16:54:28< Ivanovic> yes, they differ significantly from what is listed in the source files
20090312 16:55:25< Shadow_Master> hi
20090312 16:57:33-!- xonev [n=chatzill@59.92.57.234] has quit ["ChatZilla 0.9.84 [Firefox 3.0.7/2009021910]"]
20090312 16:58:34< Shadow_Master> Soliton: thanks. I didn't know but I was going to run a search for those mispellings today or later anyway
20090312 16:58:56-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:00:02< Soliton> run wmllint?
20090312 17:00:04-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has quit [Connection timed out]
20090312 17:00:48-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:01:16< Shadow_Master> perhaps. :-/
20090312 17:01:48-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:02:20-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:02:43-!- Sirp [n=me@wesnoth/developer/dave] has joined #wesnoth-dev
20090312 17:05:49-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has quit [Client Quit]
20090312 17:07:18-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 17:08:49< crimson_penguin> Shadow_Master: looked into font merging yet? ;)
20090312 17:09:05< crimson_penguin> oh yay! looks like my real internet is back!
20090312 17:09:08 * crimson_penguin goes to switch
20090312 17:09:13-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit ["Leaving"]
20090312 17:09:55-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20090312 17:09:57 * Shadow_Master has not looked into it yet.
20090312 17:10:04-!- ikarius [n=ross@smtp.gridironsystems.com] has joined #wesnoth-dev
20090312 17:11:00< Ivanovic> esr: please have a look at the scattered items in THOT
20090312 17:11:06< Ivanovic> they look bad in the new forests!
20090312 17:11:30< esr> Checking...
20090312 17:11:38< Shadow_Master> remove them?
20090312 17:13:37< CIA-70> esr * r33570 /trunk/data/campaigns/The_Hammer_of_Thursagan/scenarios/01_At_The_East_Gate.cfg: Remove leaning oaks from the new forests.
20090312 17:15:54-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit []
20090312 17:16:34< Shadow_Master> crimson_penguin: installed fontforge. will experiment with it later
20090312 17:16:55< Shadow_Master> it gives some warnings about DejaVu btw
20090312 17:16:59< crimson_penguin> cool
20090312 17:17:02< crimson_penguin> hmm
20090312 17:17:27< Shadow_Master> oh. it has a "merge font" command.
20090312 17:19:23-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:20:10-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:20:49< Shadow_Master> fontforge[29513]: segfault at 20 ip 00007f6c859a85f1 sp 00007fff8e251900 error 4 in libfontforge.so.1.0.0[7f6c85768000+2e3000]
20090312 17:20:53< Shadow_Master> pfff.
20090312 17:21:08-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:21:59< CIA-70> ivanovic * r33571 /trunk/src/team.cpp: 
20090312 17:21:59< CIA-70> put the default parameters into variables so that they are handled at a single locations
20090312 17:21:59< CIA-70> the currently set defaults are taken from the wiki and from old 1.4.x times, not the "revised" ones by suokko (which were never documented in the wiki!)
20090312 17:21:59< CIA-70> several commits by suokko touched those and the commits are likely to be the cause for the current "ai strangeness", probable commits might be rev29168 and rev29531
20090312 17:22:02-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:22:11< Ivanovic> sk-: are you using svn?
20090312 17:22:31< Ivanovic> if you are, please try with revision 33571 how the behaviour changed, if it got better/worse/no change at all
20090312 17:23:34< zookeeper> Soliton, silene: the weapon special stuff seems to be working perfectly ok now
20090312 17:24:40< Shadow_Master> BUG: soft lockup - CPU#0 stuck for 263s! [s2disk:27549]
20090312 17:24:42< Shadow_Master> pfff too.
20090312 17:24:52< Shadow_Master> crimson_penguin: apparently fontforge doesn't like DejaVu :-)
20090312 17:25:09< crimson_penguin> fun
20090312 17:25:13< Shadow_Master> it segfaulted while exporting a merged font to TTF
20090312 17:25:20< crimson_penguin> I'm trying to look into the actual problem with pango/cairo
20090312 17:30:07< Ivanovic> esr: i know you have done some work on resolving the AI problems introduced by suokko some time ago
20090312 17:30:32< Ivanovic> question: is rev29531 still active or might this be particially to be blamed for the problems? http://svn.gna.org/viewcvs/wesnoth?rev=29531&view=rev
20090312 17:31:02< Soliton> Dragonking: after save-loading it moves two turns later. o.O
20090312 17:32:03< sk-> I'm not using svn, haven't ever used it. Is it pretty simple to figure out?
20090312 17:32:20< Ivanovic> sk-: on which operating system are you and are you compiling yourself?
20090312 17:32:35< sk-> Windows and no.
20090312 17:32:55< Shadow_Master> too bad.
20090312 17:32:57< Ivanovic> then: wait for the next release, no way for you to check...
20090312 17:33:02< sk-> darn
20090312 17:33:03-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:33:09< Shadow_Master> someone else will have to check this tho.
20090312 17:33:24< Ivanovic> yes
20090312 17:33:34< Ivanovic> that is: Shadow_Master, have you seen those problems, too?
20090312 17:33:45< sk-> In the mean time, I'm going to move my 1.4 survival map pack over to 1.5 and see how the AI behaves the same way.
20090312 17:33:46< Shadow_Master> can't say. I haven't played lately.
20090312 17:33:48< sk-> I have a feeling it will.
20090312 17:33:51-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:33:51< Ivanovic> and can you check if my commit changes *anything* (should only have an impact when default values are used)
20090312 17:34:05< Shadow_Master> I'm rather excessively busy trying to track other bug.
20090312 17:34:19< Ivanovic> okay
20090312 17:34:53< sk-> If someone can start a game with my addon on the server with the revision, I will test it out.
20090312 17:35:40< Shadow_Master> it's not easy to start a game on the 1.5.13 server with a 1.5.13+svn build afaik.
20090312 17:35:52< sk-> ah, okay nm then.
20090312 17:36:27-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:37:03< Soliton> just change the version to test.
20090312 17:37:15-!- ABCD [n=ABCD@pool-173-71-204-106.clppva.fios.verizon.net] has joined #wesnoth-dev
20090312 17:39:21-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has joined #wesnoth-dev
20090312 17:39:34-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:39:47< Shadow_Master> may try later, but at the moment I cannot do anything. The "rocket launch sequence started".
20090312 17:39:54< Ivanovic> sk-: what is the name of your addon and is it on the server?
20090312 17:40:25< Shadow_Master> a.k.a. my CPU is extremely hot and it doesn't have much time for anything else than a debug build
20090312 17:41:19< Ivanovic> (yes, in the bugreport it would be nice to name which addon you tried with)
20090312 17:41:35-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:42:41< zookeeper> sk-, you can also ask loonycyborg to make and upload a new build to http://www.wesnoth.org/files/hidden/win/ and then you can use that to test with your existing installation
20090312 17:43:34-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 17:43:46< Ivanovic> sk-: around?
20090312 17:43:56< Ivanovic> i got a build and can start the game on the server so that you can check
20090312 17:44:03< Ivanovic> i just need to know the name of the addon!!!
20090312 17:44:21< Shadow_Master> um, I take back my offer of trying it later.
20090312 17:45:31< Soliton> Ivanovic: as he said skullduggery.
20090312 17:46:22-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:48:28< Ivanovic> sk-: i opened the game, just got to join and do the playing
20090312 17:55:25< Shadow_Master> anyone mind to share processing power with me ?  :-P
20090312 17:55:52-!- ivan_i [n=chatzill@ppp91-77-205-206.pppoe.mtu-net.ru] has joined #wesnoth-dev
20090312 17:56:14< Shadow_Master> ... forgot that it'd be pretty useless without enough bandwidth
20090312 17:56:20-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 104 (Connection reset by peer)]
20090312 17:57:24< ivan_i> hi
20090312 17:57:49-!- ABCD_ [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 17:58:00< Shadow_Master> ho
20090312 18:00:01-!- ABCD_ [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 18:05:43-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 18:07:48-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 18:08:43-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 18:09:12< sk-> joining
20090312 18:10:53-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 18:11:10-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has quit [Read error: 110 (Connection timed out)]
20090312 18:11:48< Ivanovic> sk-: please join the new game
20090312 18:11:50-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 18:17:26-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 18:18:08< Ivanovic> okay, looks good, the error does not appear anymore now that the "real" default values are used
20090312 18:18:23< Soliton> cool.
20090312 18:18:29< sk-> Would this explain why this error does not happen in the campaigns?
20090312 18:18:54< Shadow_Master> possibly. Most of them use their own values.
20090312 18:19:43< Soliton> standard mp doesn't.
20090312 18:20:07< sk-> ah
20090312 18:22:17-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 18:22:55-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090312 18:27:54< CIA-70> ivanovic * r33572 /website/start/1.6/po/ru.po: update Russian translation of announcement (only 76% ready, 85% needed for generating the html file)
20090312 18:29:27-!- ivan_i [n=chatzill@ppp91-77-205-206.pppoe.mtu-net.ru] has quit ["ChatZilla 0.9.84 [Firefox 3.0.7/2009030422]"]
20090312 18:31:48< Soliton> Dragonking: pff, you're formula only falls back when it doesn't want to recruit anymore!
20090312 18:32:57< Shadow_Master> since DK is formula, I'll try to debug him.
20090312 18:35:30-!- Netsplit verne.freenode.net <-> irc.freenode.net quits: Sirp, stikonas
20090312 18:36:06-!- Netsplit over, joins: Sirp, stikonas
20090312 18:37:02-!- Sirp_ [n=me@c-71-198-3-104.hsd1.ca.comcast.net] has joined #wesnoth-dev
20090312 18:38:13-!- stikonas [n=stikonas@wesnoth/translator/stikonas] has quit [Connection reset by peer]
20090312 18:38:23-!- stikonas [n=stikonas@ctv-213-164-100-13.vinita.lt] has joined #wesnoth-dev
20090312 18:40:46-!- ABCD_ [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 18:40:56-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 18:41:44-!- ABCD_ [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 18:43:12-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 18:44:57< Soliton> funny how formul ai can give a filename and line on errors when the formula comes from a savegame.
20090312 18:45:47-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 18:46:06< happygrue> Ivanovic: great! another good test scenario on the add-on server is Temple of the Nagas
20090312 18:46:10< Shadow_Master> so that's the real purpose of the faifile directive ..
20090312 18:46:58< happygrue> if you recruit and sit in your keeps and the AI has attacked you by around turn 8-10 then that would be great
20090312 18:47:12< Ivanovic> happygrue: would be great if you could test if there is a change in behavior in revision 33571
20090312 18:47:17< Soliton> if that is on purpose then it is just aiming to confuse the user since changing the referenced file obvioulsy does nothing.
20090312 18:47:34-!- foo1002 [n=1002@190.22.74.182] has joined #wesnoth-dev
20090312 18:47:50< happygrue> Ivanovic: ah, I hadn't seen that you committed 
20090312 18:47:53< happygrue> *it
20090312 18:47:56< happygrue> I'll test it
20090312 18:48:05< Ivanovic> i commited it since it made sense to have
20090312 18:48:06< Ivanovic> ;)
20090312 18:48:13-!- foo1002 [n=1002@190.22.74.182] has left #wesnoth-dev ["(null)"]
20090312 18:49:37-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 18:50:44< sk-> Would wesnoth.exe at http://www.wesnoth.org/files/hidden/win/ have the revision?
20090312 18:51:15< happygrue> mmm, no
20090312 18:51:17< happygrue> good point
20090312 18:51:25< Ivanovic> not sure
20090312 18:51:27< Ivanovic> not sure at all
20090312 18:51:28-!- Sirp [n=me@wesnoth/developer/dave] has quit [Read error: 110 (Connection timed out)]
20090312 18:51:35< crimson_penguin> http://pangopdf.sourceforge.net/docs/pangopdf-Fonts.html#pango-font-description-set-family <- "In some uses of PangoFontDescription, it is also possible to use a comma separated list of family names for this field."
20090312 18:51:46< Ivanovic> loonycyborg: which revision is the "daily built" exe?
20090312 18:52:16< Soliton> new enough.
20090312 18:52:16< crimson_penguin> loonycyborg: so probably you just can't use a comma separated list with Quartz rendering
20090312 18:52:21< Soliton> it's not daily built.
20090312 18:52:55< Soliton> just look at the modification times.
20090312 18:54:10< sk-> I have half a dozen other survivals of mine I'll test too for good measure when the revision gets up there.
20090312 18:55:01< Soliton> it is up.
20090312 18:56:23< sk-> So the exe at http://www.wesnoth.org/files/hidden/win/ will have it? okay, I'll check it out.
20090312 18:56:39< Ivanovic> yes
20090312 18:57:55< Ivanovic> okay, will submit the SoC application today
20090312 18:58:47< Soliton> hrm, now the formula error only goes to stderr and not the ingame chat.
20090312 19:06:44-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has joined #wesnoth-dev
20090312 19:07:05< sk-> are all the addons that are on the 1.5 server going to be transferred to the 1.6 or is it going to be a fresh start?
20090312 19:07:08< sk-> out of curiousity
20090312 19:07:26< Shadow_Master> they won't be transferred.
20090312 19:07:30< Shadow_Master> the 1.5 server will be the 1.6 server.
20090312 19:07:41< sk-> so they'll stay then
20090312 19:11:53< Soliton> yep, the recent cleanup was all that is needed.
20090312 19:12:02< Soliton> everything now should be compatible.
20090312 19:13:35< CIA-70> zookeeper * r33573 /trunk/data/core/terrain-graphics.cfg: Wired in the new desert plants.
20090312 19:13:57 * Soliton now managed to crash wesnoth with an invalid formula.
20090312 19:15:06< loonycyborg> crimson_penguin: I suspected as much..
20090312 19:15:57 * crimson_penguin is having a really hard time finding out much of anything about this
20090312 19:24:57-!- Blueblaze [n=Blueblaz@c-98-199-143-139.hsd1.tx.comcast.net] has quit []
20090312 19:25:12< Shadow_Master> Turuk: around?
20090312 19:26:39-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20090312 19:27:22-!- Azkar [n=Azkar@pool-96-235-20-27.pitbpa.fios.verizon.net] has quit []
20090312 19:28:32< Dragonking> Soliton: So you found a bug?
20090312 19:30:02< Soliton> Dragonking: somewhat yes. you assume recruits always work when you want to recruit which is not the case.
20090312 19:30:22< Dragonking> Yes
20090312 19:30:22< Soliton> like when there is no space to recruit.
20090312 19:30:28< Dragonking> I know
20090312 19:30:35< Soliton> and then the fallback is never reached,
20090312 19:30:39< Dragonking> I was suppsed to fix that - now I remember
20090312 19:30:46< Dragonking> It is this infinite loop of recruting
20090312 19:31:01< Soliton> well, it's not infinite.
20090312 19:31:05< sk-> everything else I tested (all my old survivals from 1.4) works great with the revision. Seems the problem is solved.
20090312 19:31:06< Dragonking> But I found some bugs to fix before fixing that, and then new bugs came out and..
20090312 19:31:13< Ivanovic> good
20090312 19:31:22< Dragonking> Soliton: It's not?
20090312 19:31:38< Soliton> nope it just returns false.
20090312 19:31:41< Dragonking> ok
20090312 19:31:49< Dragonking> I'll try to fix that
20090312 19:32:03< Soliton> you just have to recruit in the formula in another if.
20090312 19:32:10< Soliton> +wrap
20090312 19:32:24< Soliton> you just have to wrap the recruit in the formula in another if.
20090312 19:33:17< Ivanovic> sk-: this sounds promising, looks like my "unification" of stuff (really simple change!!!) had some positive effect...
20090312 19:34:07< zookeeper> i'm gonna try to test the AI out a bit too
20090312 19:34:12< sk-> it radically changed the behavior. Before I could end turn 30 times and the AI would start filling up on both sides of that map, spilling over into the void.
20090312 19:34:14< Ivanovic> zookeeper: perfect
20090312 19:34:55< Dragonking> Soliton: I don't think it will work
20090312 19:35:04< Ivanovic> zookeeper: and especially have a look at "what should be changed" like making it go more after villages and such
20090312 19:35:15< Dragonking> I need to check for free space first somehow, then if I have space, recruit
20090312 19:35:17< Ivanovic> atm i reset the stuff to really old default values
20090312 19:35:29< Dragonking> This invlolves adding new function :)
20090312 19:35:53< Soliton> what's wrong with checking what recruit() returns=
20090312 19:36:01< Soliton> s/=/?/
20090312 19:36:05< Dragonking> recruit() in formula language?
20090312 19:36:24< Dragonking> It does not return anything - it is just passed and executed
20090312 19:36:37< zookeeper> Ivanovic, was your change only about changing what values are used by default or also something about scenario-specified values getting overwritten?
20090312 19:36:55< Dragonking> Or different - recruit() in fomrula language always return recruit_action - and it is executed
20090312 19:36:55< Soliton> hmm, i see.
20090312 19:37:04< Ivanovic> zookeeper: i only changed the "default that is loaded" and have no idea if the params set might be overridden by those values
20090312 19:37:15< Dragonking> Soliton: But I'll fix it, most likely today
20090312 19:37:20< Ivanovic> so it *might* be that you will see a difference in all scenarios
20090312 19:37:21< Soliton> i mean check whether the move returns true.
20090312 19:37:28< Dragonking> Just give me some time to put myself together and rest :)
20090312 19:37:31< Soliton> i guess you can't do that.
20090312 19:37:33-!- thespaceinvader_ [n=chatzill@91.108.177.192] has joined #wesnoth-dev
20090312 19:37:37< Ivanovic> as in "ai params not working at all"
20090312 19:37:39< Dragonking> Yup, I can't.
20090312 19:37:59-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit [Nick collision from services.]
20090312 19:38:01< Dragonking> I'll just check if I have free space in castle and if not, fallback to default AI
20090312 19:38:10-!- thespaceinvader_ is now known as thespaceinvader
20090312 19:38:29< Soliton> what if some weird scenario makes the recruit fail anyway?
20090312 19:38:41< Dragonking> How?
20090312 19:38:58< Soliton> no idea, the problem i see is that you reimplement the checks in recruit().
20090312 19:39:39< Dragonking> What is the problem? I don't understand to be honest.
20090312 19:39:55< Dragonking> Mind I'm tired, I may fail to comprehend. :)
20090312 19:40:01< Soliton> the next time recruit() changes your formula might beak.
20090312 19:40:10< Soliton> break too.
20090312 19:40:21< Dragonking> recruit() formula function?
20090312 19:40:27< Soliton> no.
20090312 19:40:53< Dragonking> recruit() AI function?
20090312 19:41:22< Soliton> yes.
20090312 19:41:59< Dragonking> Well, just we (people messing with AI) has to make sure it works
20090312 19:42:35< Dragonking> Formula API should work, it someone modifies recruit() so it does not work again with formula, then we have bug.
20090312 19:42:53< Dragonking> And someone needs to fix that.
20090312 19:43:09< Soliton> i'm not talking about the API but just the specific recruitment ai we're talking about.
20090312 19:43:39 * Dragonking honestly fails to comprehend the problem :/
20090312 19:44:17< Soliton> nevermind, fix it.
20090312 19:45:08< Soliton> the next problem will appear when there is an invalid unit type in the recruit list.
20090312 19:45:39< Dragonking> It is taken form "my_recruits", so there should not be a problem
20090312 19:45:50< Dragonking> Unless someone messes formula - then yes, we have a problem
20090312 19:46:10< Soliton> i don't know what my_recruits is.
20090312 19:46:34< CIA-70> jorda * r33574 /website/start/1.6/: Ignore stamp directory
20090312 19:46:37< Dragonking> Ai member, that keeps all info about units available to recruit
20090312 19:46:49< Dragonking> So it is not really a problem.
20090312 19:46:51< Soliton> and that validates input?
20090312 19:47:06< Dragonking> No, input is based on that.
20090312 19:47:13< Dragonking> In my formula
20090312 19:47:24< Dragonking> Someone may write recruit('wolf')
20090312 19:47:27< Dragonking> ANd it will fail
20090312 19:47:37< Soliton> so where does the recruit list come from?
20090312 19:47:43< Dragonking> But that is not a problem with recruit(), it is problem with usage
20090312 19:48:14< Dragonking> from formula_ai.cpp:2050
20090312 19:48:26< Dragonking> So form the game info itself.
20090312 19:48:48< Soliton> "if ( current_team().is_enemy( i+1 ) )"
20090312 19:49:15< Dragonking> Yes?
20090312 19:49:32< Soliton> that's line 2050 but i think i see what line you mean.
20090312 19:50:28< Dragonking> For me 2050 is "} else if(key == "my_recruits")"
20090312 19:51:18< Soliton> ok, what's that supposed to tell me=
20090312 19:51:21< Dragonking> Soliton: So I don't think tha recruitment is going to fail so easily
20090312 19:51:49< Dragonking> It is answer for
20090312 19:51:53< Dragonking> <Soliton> the next problem will appear when there is an invalid unit type in the recruit list.
20090312 19:52:06< Soliton> i don't know what my_recruits is.
20090312 19:52:22< Soliton> but nevermind rest or whatever i'll look into it.
20090312 19:52:37-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has quit ["."]
20090312 19:52:40< Dragonking> my_recruit is formula.. object
20090312 19:52:52< Dragonking> It contains all recruits current side can do
20090312 19:53:19< Dragonking> I use it, so there won't be such thing like "impossible" recruit in terms of recruting unit that cannot be recruited.
20090312 19:53:26< Soliton> that doesn't tell me anything about its validity.
20090312 19:53:56< Dragonking> Then look into code and check if it is not buggy.
20090312 19:54:03< Dragonking> Then you will make sure it is valid.
20090312 19:54:18< Dragonking> I can only tell you it is based on what AI has access to.
20090312 19:54:46< Soliton> the ai has access to anything.
20090312 19:55:02< Dragonking> Yup.
20090312 19:56:08< Dragonking> I don't really understand your concerns regarding what we pass to recruit function. If it is incorrect, it just won't recruit. I tried to make sure it will be correct. ONly thing so far left to fix is this no place for recruit thing.
20090312 19:57:24< Soliton> it won't recruit and that's the end of the move.
20090312 19:57:38< Soliton> doesn't sound like that will likely be wanted.
20090312 19:57:50< Dragonking> If formula is broken, then yes.
20090312 19:57:54< Dragonking> If not, then it will work.
20090312 19:58:06< Dragonking> It's all about formula now, not about how does recruit behave
20090312 19:58:27< Soliton> no, formula needs to know exactly how recruit() behaves.
20090312 19:58:39< Soliton> or it will be "broken".
20090312 19:58:50< Turuk> Shadow_Master, what are your current goals on Administration then and what would you like to handle so I know not to cross into your lane?
20090312 19:58:54< Dragonking> How it can know about it?
20090312 19:59:03< Dragonking> Formula can't check for return value.
20090312 19:59:11< Dragonking> It is pure functional language.
20090312 19:59:37< Soliton> so afaict there is no check on the recruit list.
20090312 19:59:54< Soliton> so that will indeed be the next point where it will fail.
20090312 20:00:00< Dragonking> How?
20090312 20:00:07< Dragonking> Give me example please.
20090312 20:00:20< Soliton> you recruit an invalid unit and don't fallback.
20090312 20:00:31< Dragonking> How you get info about invalid unit?
20090312 20:00:47< Soliton> you write an invalid unit into the recruit key.
20090312 20:00:59-!- moi [n=moi@153.12.95-79.rev.gaoland.net] has joined #wesnoth-dev
20090312 20:01:03< Dragonking> Then it is human problem. And it will fail then.
20090312 20:01:14-!- moi is now known as EdB
20090312 20:01:20< Soliton> yes.
20090312 20:01:57< Dragonking> How is that releated to my recruitment formula?
20090312 20:02:49< Soliton> it will fail for no apparent reason.
20090312 20:03:22-!- Tigge [n=tigge@bacchus.olf.sgsnet.se] has quit [Read error: 104 (Connection reset by peer)]
20090312 20:03:29< Dragonking> It will
20090312 20:04:12< Dragonking> Your reasoning is like "if we write "2+'a'" it should not fail". 
20090312 20:04:23< Dragonking> Writing incorrect formula is a reason to fail.
20090312 20:04:38< Soliton> the formula is not incorrect.
20090312 20:04:42< Soliton> its input is.
20090312 20:04:48< Dragonking> Yes
20090312 20:05:01< Dragonking> And it resuults in that kind of behaviour
20090312 20:05:13< Soliton> anyway it was one example of how recruitment can fail.
20090312 20:05:19< Dragonking> Formula is not evaluated to error and no info about error is displayed.
20090312 20:05:27< Dragonking> No it can't. :)
20090312 20:05:49< Dragonking> Bacause in my formula impossible recruitment won't be passed to recruit() :)
20090312 20:05:57< Soliton> my original point was that you duplicate existing checks in your formula which is bad.
20090312 20:06:09< Soliton> indeed the point.
20090312 20:06:10-!- Tigge [n=tigge@bacchus.olf.sgsnet.se] has joined #wesnoth-dev
20090312 20:06:48< Dragonking> Soliton: OK, now I got your point.
20090312 20:07:15< Dragonking> Thing is, there is no other way of doing that.
20090312 20:07:43< Dragonking> We can't check what recruit() returned, wheter it failed or not
20090312 20:07:51< Dragonking> We just order AI to recruit something
20090312 20:09:24< Soliton> how does if work in formula?
20090312 20:09:54< Soliton> i mean there must be a away to return truth values somewhere?
20090312 20:11:13< Dragonking> Kay thing is to understan how formula functions work.
20090312 20:11:23< Dragonking> Key even
20090312 20:11:50< Dragonking> There are 2 types of function  in general - these that return action object, and these that return data type
20090312 20:11:58< Dragonking> recruit returns action object
20090312 20:12:21< Dragonking> And to make unit recruit, *whole* formula has to return this object
20090312 20:12:49< Dragonking> If result of recruitment is correct, we re-evaluate whole formula again
20090312 20:12:56< Dragonking> If not, we stop processing formula
20090312 20:13:28< Dragonking> Other examples of function that return action obects are for example move or attack
20090312 20:13:39< Soliton> ok, i understand.
20090312 20:13:54< Soliton> so do not fail the move and instead set some variable?
20090312 20:14:31< Dragonking> Thing is - formula will be executed till it provides valid result.
20090312 20:14:45< Dragonking> That's how Dave designed it, and I have been sticking to it.
20090312 20:14:57< Soliton> i'm not suggesting to deviate from it.
20090312 20:16:23< Dragonking> So how would it work?
20090312 20:16:28< Dragonking> Setting this variable?
20090312 20:17:02< Dragonking> You want to suggest a way to check if something failed or not, yes?
20090312 20:17:05< Soliton> don't fail the recruit as before and instead set some variable (recruitment_failed) which can be checked in the formula.
20090312 20:17:09< Soliton> yes.
20090312 20:17:44< Dragonking> Like if( ai.vars.error = 'recruitment' , .. )
20090312 20:17:49< Soliton> the formula needs to decide to end the move not wesnoth.
20090312 20:17:58< Soliton> sure, whatever.
20090312 20:19:08 * Ivanovic is going to submit the SoC application now
20090312 20:19:28< Soliton> seems like a better design to me. i know the invalid unit thing wasn't the best example but it's the principle.
20090312 20:20:36< Dragonking> I'll think about it, thanks for idea.
20090312 20:20:41< Soliton> anytime a valid action is produced by the formula it should be called again whether that action really succeeded or not.
20090312 20:21:06< Soliton> and to avoid infinite loops the formula needs to be able to check for errors.
20090312 20:21:12< Dragonking> Yes.
20090312 20:21:35< Soliton> otherwise you basically have to reimplement all the game rules.
20090312 20:21:51< Soliton> admittedly they are unlikely to change but... :-)
20090312 20:21:55< Dragonking> :)
20090312 20:21:59-!- YogiHH [i=YogiHH@c169090.adsl.hansenet.de] has joined #wesnoth-dev
20090312 20:22:12< Dragonking> Good idea I tihnk. <- now one I understood your point ;p
20090312 20:22:14< Dragonking> once'
20090312 20:22:21< Soliton> ok, i'm glad.
20090312 20:22:35< Sirp_> hi YogiHH 
20090312 20:22:45< YogiHH> hi Sirp_
20090312 20:23:12< Dragonking> Sirp_: Opinion on what is written above?
20090312 20:23:44< Soliton> hey YogiHH, seen the bug i assigned to you?
20090312 20:23:49< Dragonking> I could add internal var "error" that will be empty if no error occurs, and has something set within otherwise.
20090312 20:23:57< cjhopman> Dragonking: you could just have something else that is called when it fails... like a recruit() formula and a recruit_failed() formula
20090312 20:24:01< YogiHH> Soliton: Yes, i have. But i didn't quite understand your comment about the unit key.
20090312 20:24:34< cjhopman> then recruit_failed() could try to fix something and then return something that says whether or not to continue the recruit loop
20090312 20:25:05< Soliton> YogiHH: in your commit you merged two functions that read unit and unit_type. one was for backwards compatibility and the only one that you left.
20090312 20:25:21< Dragonking> cjhopman: you can define it easly with error too
20090312 20:25:35< YogiHH> Soliton: can you tell me which functions that were?
20090312 20:25:52< Dragonking> cjhopman: if( ai.vars.error != 'recruitment',  normal_formula,  recruit_failed_formula )
20090312 20:27:45< Soliton> YogiHH: actually they weren't seperate functions just two for loops.
20090312 20:28:20< YogiHH> Soliton: ok, i think i know what you are referring to
20090312 20:28:23< Soliton> YogiHH: it's hard to explain just look at the diff.
20090312 20:29:04< Soliton> basically there was logic to read unit_type and for backwardscompaitibility unit and you remove the unit_type part in your commit.
20090312 20:29:16< Soliton> now it's too late anyway.
20090312 20:29:35< Soliton> well, we can allow both maybe.
20090312 20:29:53< YogiHH> i'll check that
20090312 20:29:59-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has quit [Remote closed the connection]
20090312 20:30:11< Soliton> really it's just a point about not making huge commits. ;-)
20090312 20:30:32< Soliton> or being really careful not to introduce regressions because no one will check you.
20090312 20:30:45< Soliton> ..unless something breaks. :-)
20090312 20:31:31< YogiHH> i know, i will spend more time to think about how to provide smaller steps next time :-)
20090312 20:32:04< Soliton> using something like git helps a lot since you don't have to push commits out when you make them.
20090312 20:32:17< CIA-70> esr * r33575 /trunk/data/campaigns/Eastern_Invasion/ (12 files in 2 dirs): Fix broken journey paths in Eastern Invasion.
20090312 20:33:12< Soliton> you can just implement your feature step by step and keep committing and only need to push it out once it is finished and tested.
20090312 20:33:39< Soliton> with svn you really just can't do that unfortunately.
20090312 20:34:15< YogiHH> I am not sure if that is all that different. The only point i see is when you stumble about some unexpected behaviour, you can revert stuff more easily.
20090312 20:34:53< Soliton> you can't make local commits with svn.
20090312 20:35:11-!- Elvish_Pillager [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090312 20:35:30< Soliton> so if you're online and want to commit part of a feature fine otherwise you're screwed.
20090312 20:36:06-!- sk- [n=sk@pool-71-242-206-125.phlapa.east.verizon.net] has quit ["Leaving"]
20090312 20:36:11< YogiHH> Soliton: well, ok, that is different to a branch then. It eases the merge part a lot, i guess.
20090312 20:36:55-!- ilor [n=user@wesnoth/developer/ilor] has joined #wesnoth-dev
20090312 20:36:57< Soliton> well, it's a local branch.
20090312 20:39:12< Soliton> if you develop your feature on a branch in svn that's indeed similar.
20090312 20:48:04-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has left #wesnoth-dev []
20090312 20:49:46< nital> hello, is the glitch with overlapping the top of a big tree during attack from behind of it already known (maybe as a part of #13041) or should I submit it?
20090312 20:51:11< nital> or maybe it's intended behaviour? looks strange, though
20090312 20:54:14-!- ABCD_ [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 20:55:21< Ivanovic> everybody: submitted the current version of http://www.wesnoth.org/wiki/SoC_Information_for_Google to google
20090312 20:55:39< Ivanovic> if you have any edits there, ping me ASAP so that i can update the application with the info
20090312 20:55:39< Soliton> how does it look for games?
20090312 20:55:41-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 104 (Connection reset by peer)]
20090312 20:58:15-!- ABCD_ [n=ABCD@wikipedia/ABCD] has quit [Client Quit]
20090312 20:58:24-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20090312 20:58:47< YogiHH> Soliton: this is probably the for-loops you were talking about? http://wesnoth.pastebin.com/d7de0afca
20090312 20:59:22< Soliton> yep.
20090312 20:59:37-!- ABCD [n=ABCD@wikipedia/ABCD] has joined #wesnoth-dev
20090312 20:59:44< YogiHH> Soliton: but the unit_key loop survived, wasn't that what we intended to do?
20090312 20:59:56< YogiHH> oops
20090312 21:00:05< YogiHH> unit_key => unit_type
20090312 21:01:58< Soliton> ok, let me look at it again.
20090312 21:02:22< YogiHH> sure
20090312 21:03:33< Soliton> -                       const std::string &from = (**af.first)["unit_type"];
20090312 21:03:37< Soliton> +        const std::string &from = (**af.first)["unit"];
20090312 21:03:44< Soliton> that's what i meant.
20090312 21:04:54< Soliton> so the loop is still about unit_type but the key is reverted to unit.
20090312 21:06:00-!- Turuk_ [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has joined #wesnoth-dev
20090312 21:06:19< Soliton> there was a loop over [unit] with unit= in the [advancefrom] and a loop over [unit_type] with unit_type in the [advancefrom].
20090312 21:06:50< Soliton> and after the commit it was a loop over [unit_type] with unit= in the [advancefrom].
20090312 21:07:26< Soliton> since there are no advancefroms in mainline nobody cares though and it's too late to change now.
20090312 21:08:48< Soliton> well, actually there are.
20090312 21:09:08< Soliton> but they use unit= now of course.
20090312 21:09:31< Soliton> they probably were never converted.
20090312 21:14:29< YogiHH> Soliton: i see. I remember that i messed up advancefrom when i started lazy loading and restored it later on when i was fixing a UMC campaign bug that was caused by it. But the key issue obviously slipped through.
20090312 21:16:31< Soliton> right and if it wouldn't have all been in one big commit maybe it would have been spotted earlier. :-)
20090312 21:17:49< Soliton> do you know why the experience value gets overwritten?
20090312 21:18:32< Soliton> i would have thought the type is already loaded when such and advancefrom is processed?
20090312 21:19:47< Soliton> Dragonking: btw, how do you test formula changes like when you wrote recruitment.fai?
20090312 21:20:25< Dragonking> Soliton: YOu mean how I debug and test formulas while I'm writing them?
20090312 21:21:15< Soliton> yep.
20090312 21:21:39< Dragonking> Mostly with debug_print
20090312 21:21:43< Soliton> i did a bit of trial and error to fix the formula but that got old quick.
20090312 21:22:11< Dragonking> I make sure separate parts work properly and put these together.
20090312 21:22:24< Soliton> i mean like you have to restart wesnoth all the time, no?
20090312 21:22:27< Dragonking> That's why I split recruitment into many functions
20090312 21:22:49< YogiHH> Soliton: No, i am not sure why xp gets overwritten. I know for sure that advancements are loaded in the very beginning, so they should be available.
20090312 21:23:01< Dragonking> Mostly, yes. 
20090312 21:23:10< Soliton> i guess one could actually develop them in the savegame and then save-load. :-)
20090312 21:23:54< Dragonking> I develop these in formula scenario
20090312 21:23:54< Soliton> i made a command line to start a multiplayer game which helped but always restarting is still annoying.
20090312 21:24:06< Dragonking> So I just ./wesnoth -t formula_recruitment -d
20090312 21:24:14< Soliton> yeah, that's similar.
20090312 21:25:08< Soliton> YogiHH: but for an advancefrom the unit in question doesn't need to be used anywhere else.
20090312 21:25:13< Dragonking> Soliton: I'll make it so we can specify on_fail= formula, and besides we will have some error flags set to help to avoid infinite loops.
20090312 21:25:33< Soliton> Dragonking: ok, sounds good.
20090312 21:25:50< Dragonking> And thanks for a tip. :)
20090312 21:25:54< Dragonking> That was good idea.
20090312 21:26:03< Soliton> you're welcome. :-)
20090312 21:27:18< Soliton> YogiHH: especially i that campaign a cuttlefish was probably never encountered before so when the kraken references it in its advancefrom that#s the first occurence.
20090312 21:28:13< YogiHH> Soliton: if i remember correct, that is the reason why i processed every advance stuff before any access happens. This is also needed for the help.
20090312 21:28:40< Soliton> ok, i see.
20090312 21:29:30< Soliton> well, the logic there looked all fine to me so i can only assume the experience value gets overwritten somewhere later on.
20090312 21:30:30< YogiHH> Soliton: yes, i suspect the same
20090312 21:32:21< Soliton> hmm, i wonder how it works with the ancient lich. there we actually do use the experience key as well.
20090312 21:35:48< Ivanovic> esr: http://www.wesnoth.org/forum/viewtopic.php?f=4&t=24460&start=0&st=0&sk=t&sd=a
20090312 21:38:37-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090312 21:40:32-!- YogiHH_ [i=YogiHH@c169090.adsl.hansenet.de] has joined #wesnoth-dev
20090312 21:41:50< YogiHH_> Soliton: well, in a way this is logical. Otherwise you could define different higher level units with different xp needed
20090312 21:42:30< Soliton> you can and the lowest xp will be needed.
20090312 21:42:46< YogiHH_> Soliton: hmm, i see
20090312 21:43:20< Soliton> which is why the advancefrom of the ancient lich actually doesn't make a lot of sense.
20090312 21:43:26-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090312 21:44:06-!- Turuk is now known as ghost
20090312 21:44:14< Soliton> we could make it always use the value from advancefrom though. but i guess then what happens when there are several....
20090312 21:44:16-!- ghost is now known as Turuk
20090312 21:48:40-!- EdB [n=moi@153.12.95-79.rev.gaoland.net] has quit [Remote closed the connection]
20090312 21:48:49-!- Turuk_ [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has quit [Read error: 110 (Connection timed out)]
20090312 21:51:09< YogiHH_> Soliton: if we have different xp needed, does the lowest count for all paths or do we only allow leveling into the lowest?
20090312 21:54:10< Soliton> only the lowest counts for all paths.
20090312 21:54:19< YogiHH_> ok
20090312 21:55:03-!- YogiHH [i=YogiHH@wesnoth/developer/yogihh] has quit [Read error: 110 (Connection timed out)]
20090312 21:55:45-!- YogiHH_ is now known as YogiHH
20090312 22:04:43< Soliton> zookeeper: should malin need 150 or 250 XP to become an ancient lich?
20090312 22:06:46< Soliton> hmm, i wonder if we even need advancefrom anymore with base_unit.
20090312 22:09:01< crimson_penguin> Ivanovic: Turns out the alias should've been a symlink, and the DMG Canvas guy changed it and updated the program already :D
20090312 22:09:54< Ivanovic> crimson_penguin: ah, okay
20090312 22:10:46< zookeeper> Soliton, i don't know. 250 is probably better.
20090312 22:10:51< crimson_penguin> Aliases store an inode number and a path, symlinks... well, you know what they do
20090312 22:11:17< crimson_penguin> aliases seem to include the volume name in the path though, which could be the problem
20090312 22:11:20< Soliton> zookeeper: do you think there is still a need for advancefrom?
20090312 22:13:37< Soliton> hmm, maybe for mp addons that want to avoid a download requirement.
20090312 22:14:29-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has joined #wesnoth-dev
20090312 22:18:21< zookeeper> well, not a strict need besides that MP stuff i guess.
20090312 22:18:33< zookeeper> but i think it's probably still a nice feature to have
20090312 22:18:58< zookeeper> since it is easier to just use that instead of [base_unit]ing stuff a lot in order to for example wire in a new advancement to an existing unit line or whatever
20090312 22:22:21< Soliton> you can not set xp requirements higher than the existing one though. like the ancient lich has experience=250 in his advancefrom which does nothing.
20090312 22:25:10< zookeeper> right
20090312 22:25:13< zookeeper> has it ever worked?
20090312 22:25:53< Soliton> i don't think so. let me check...
20090312 22:27:47< Soliton> nope, that way since its implementation.
20090312 22:31:05< zookeeper> i guess it's not incredibly useful then, what with [base_unit] and all. want to remove it in 1.7 or what?
20090312 22:32:41< happygrue> Ivanovic: Can't really say for sure that it's 'fixed', but your changes seem to be a big improvement AFAICT
20090312 22:32:47< happygrue> good work :)
20090312 22:34:12< Soliton> zookeeper: sounds like a good idea to me since it's really a bit of a hacky feature.
20090312 22:35:44< zookeeper> that's the impression i've gotten
20090312 22:38:44< CIA-70> jhinrichs * r33576 /trunk/ (changelog src/unit_types.cpp): Fixes bug #13176 (advancefrom tag ignores experience requirement) and removes some no longer needed debug messages.
20090312 22:40:31< YogiHH> Soliton: done
20090312 22:41:15< Soliton> sweet.
20090312 22:43:31< Soliton> YogiHH: would be nice to convince your editor to stop inserting a BOM into the changelog.
20090312 22:44:05-!- boucman [n=rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20090312 22:44:29< YogiHH> Soliton: I can't, that editor is too stupid (notepad)! But i will use another one next time ;-)
20090312 22:44:52-!- cjhopman_ [n=chris@wesnoth/developer/cjhopman] has joined #wesnoth-dev
20090312 22:45:17-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has quit [Connection timed out]
20090312 22:45:24< YogiHH> night everyone
20090312 22:45:43-!- YogiHH [i=YogiHH@c169090.adsl.hansenet.de] has quit ["Bye folks"]
20090312 22:47:52< CIA-70> soliton * r33577 /trunk/changelog: removed BOM
20090312 22:57:00-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 22:57:53-!- Shadow_Master_ [n=imorelle@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 22:58:07-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has quit [Nick collision from services.]
20090312 22:58:19-!- Shadow_Master_ is now known as Shadow_Master
20090312 22:58:49< Shadow_Master> I cannot read the thermal zone stats, but I guess my CPU is way above 75°C
20090312 22:59:24< Shadow_Master> the guilty commit for the cache regeneration issue must be one of the last ones before 1.5.11
20090312 23:00:20< Shadow_Master> now at 1.5.10+svn revision 33040 (removal of Python AI)
20090312 23:02:41< Dragonking> night everyone
20090312 23:02:46< Shadow_Master> as I suspected, dfranke broke it :-(
20090312 23:02:52< Shadow_Master> wesbot: log 33040
20090312 23:02:57< wesbot> dfranke * r33040 : Remove Python AI, resolving bug #13047 and #13048.
20090312 23:03:00< wesbot> URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=33040
20090312 23:03:14-!- Dragonking [n=dk@wesnoth/developer/dragonking] has quit []
20090312 23:03:16< Shadow_Master> anyone can notice what could have broken add-on cache regeneration there?
20090312 23:04:17< CIA-70> shadowmaster * r33578 /trunk/data/scenario-test.cfg: Remove a WML event test case.
20090312 23:04:35< CIA-70> shadowmaster * r33579 /trunk/src/leader_list.cpp: 
20090312 23:04:35< CIA-70> Don't apply the RC image functor on random-dice.png, which is a TC-less
20090312 23:04:35< CIA-70> image.
20090312 23:04:50< CIA-70> shadowmaster * r33580 /trunk/src/ (unit.cpp unit.hpp): 
20090312 23:04:50< CIA-70> Small hack to allow to override a unit's gender in one of the unit class
20090312 23:04:50< CIA-70> constructors with a last parameter.
20090312 23:04:55< Ivanovic> http://svn.gna.org/viewcvs/wesnoth/trunk/src/multiplayer_connect.cpp?rev=33040&view=diff&r1=33040&r2=33039&p1=trunk/src/multiplayer_connect.cpp&p2=/trunk/src/multiplayer_connect.cpp
20090312 23:05:14< CIA-70> shadowmaster * r33581 /trunk/ (changelog src/menu_events.cpp): 
20090312 23:05:14< CIA-70> Late feature: make it possible to spawn units with random genders in
20090312 23:05:14< CIA-70> debug mode (Create Unit dialog).
20090312 23:05:14< CIA-70> This does not introduce any new strings and cannot be used unless in
20090312 23:05:14< CIA-70> debug mode, therefore it is not big deal to implement it now IMHO. Feel
20090312 23:05:17< CIA-70> free to revert it if you really believe such a tiny and usually
20090312 23:05:17< Ivanovic> ups, sorry, wrong
20090312 23:05:19< CIA-70> invisible chaneg is bad...
20090312 23:06:30-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has quit [Remote closed the connection]
20090312 23:07:03< CIA-70> shadowmaster * r33582 /trunk/data/core/about.cfg: Add Girgistian to the mainline credits.
20090312 23:08:58< Ivanovic> Shadow_Master: are you 100% sure about the commit breaking things?
20090312 23:09:18< Ivanovic> somehow i see *nothing* in there that might break it that was not broken before when there was no python support activated
20090312 23:09:44< Shadow_Master> yeah, I know. I'm examining the immediately earlier commit in detail.
20090312 23:10:16< Shadow_Master> (and burning my legs, since the laptop is overheating and I can't put it on a desk without disconnecting...)
20090312 23:11:24< boucman> Ivanovic: thx for the update of the GSoC page
20090312 23:11:35< Ivanovic> boucman: did so right with submitting
20090312 23:11:43< boucman> ok, cool
20090312 23:11:57< Ivanovic> boucman: if you got further improvements till tomorrow ~5pm ping me
20090312 23:12:19< boucman> should be fine
20090312 23:12:23< Shadow_Master> um.
20090312 23:12:52< Ivanovic> that is: you saw what i removed/added in the "template to fill" part?
20090312 23:13:02< Ivanovic> since most changes i have done were there
20090312 23:13:21< Shadow_Master> 33040 seems to be the breakpoint.
20090312 23:13:44< boucman> yes, I read the diff
20090312 23:13:45< Shadow_Master> that is, I built wesnoth from the immediately earlier commit and cache regeneration when installing/upadting add-ons worked.
20090312 23:13:51< Ivanovic> Shadow_Master: which build system and have you tried with --disable-python, too?
20090312 23:13:55< Shadow_Master> scons.
20090312 23:16:36< Ivanovic> will check it here, too
20090312 23:17:36< Shadow_Master> surprise, surprise.
20090312 23:17:50-!- stikonas [n=stikonas@wesnoth/translator/stikonas] has quit [Read error: 145 (Connection timed out)]
20090312 23:17:54< Shadow_Master> bringing back the PYTHON WML macro alone fixes it.
20090312 23:18:14< Shadow_Master> e.g. insert defines_map_["PYTHON"] = preproc_define(); on src/config_cache.cpp line 79 at HEAD.
20090312 23:18:15< Ivanovic> WOW
20090312 23:19:00< Ivanovic> so basically an additional forced update
20090312 23:20:12< Shadow_Master> I believe I have stumbled upon a suokko bug actually.
20090312 23:20:17< Ivanovic> :)
20090312 23:20:33< Ivanovic> as i have with the AI stuff i solved today by clearly setting defaults?
20090312 23:20:35< Ivanovic> ;)
20090312 23:20:39< Shadow_Master> if you define FOOBAR in that line instead of PYTHON, cache generation will work again as well
20090312 23:20:45< Shadow_Master> so, sorry dfranke  :-P
20090312 23:21:58< Shadow_Master> now I have no idea why it believes that if the defines_map_ object <condition here>, it isn't neceessary to regenerate the cache
20090312 23:22:12< Shadow_Master> <condition here> seems to be <defines_map_ is empty>
20090312 23:22:16< Ivanovic> shall we add a "#define SUOKKO_BUG_WORKAROUND"?
20090312 23:22:34< Ivanovic> and have this active there
20090312 23:22:36< Ivanovic> ;)
20090312 23:22:52-!- BenUrban [n=benurban@c-68-49-10-243.hsd1.md.comcast.net] has joined #wesnoth-dev
20090312 23:23:24< Shadow_Master> if I don't figure out what the hell is going on there, I guess that workaround will have to go in 1.6.0...
20090312 23:23:36< Shadow_Master> e.g. if I don't figure out a fix for sunday
20090312 23:23:59< Ivanovic> i have no idea how much my changes in the ai to use default params help
20090312 23:24:23< Ivanovic> that is: if they don't help "at all", then the release will have to be delayed a little...
20090312 23:24:39< Ivanovic> unless Sirp_ magically finds the time and fixes the problems
20090312 23:24:41< Ivanovic> ;)
20090312 23:24:51< Ivanovic> zookeeper: you wanted to test the AI a little, any results on this?
20090312 23:25:17< Soliton> after the last big reversion there shouldn't be many differences left to the 1.4 ai so now that the defaults are fine too i think there is slim chances for actual regressions.
20090312 23:25:59< Ivanovic> Soliton: i would prefer to have some more feedback on this to be "really sure"
20090312 23:26:20< Soliton> fun ai fact though: it shoots at me with for hours until it is 1XP-to-level and then stops.
20090312 23:26:41< Soliton> and that at 1HP poisoned...
20090312 23:26:43< Ivanovic> yeah, it wants to leave you with the choice of "what will it be"
20090312 23:26:50< Ivanovic> ;)
20090312 23:26:55< Soliton> very kind indeed.
20090312 23:27:00-!- cjhopman_ [n=chris@wesnoth/developer/cjhopman] has quit [Read error: 60 (Operation timed out)]
20090312 23:27:02-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has quit ["I hate suokko D:"]
20090312 23:27:19< zookeeper> Ivanovic, well, i've never had much testcases myself. i tried the third LoW scenario but didn't see much change there.
20090312 23:28:54< zookeeper> i'm not sure how i'd go about testing it more than that, since i don't recall stumbling upon especially retarded behaviour myself lately
20090312 23:29:24< Ivanovic> maybe check how aggressive the AI approaches you in the first httt scenario?
20090312 23:29:35< Soliton> for(preproc_map::const_iterator i = defines_map_.begin(); i != defines_map_.end(); ++i) {
20090312 23:29:50< Soliton> so if there is no define the loop doesn't run at all.
20090312 23:31:40< Soliton> and without it it seems the cache will be considered valid.
20090312 23:32:12< Soliton> hmm..
20090312 23:34:51-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Remote closed the connection]
20090312 23:35:00< Soliton> and there is a set_force_not_valid_cache() function which is never used except in tests.
20090312 23:35:37< Ivanovic> uhm, this sounds like a function that should be used after successfully downloading an addon
20090312 23:35:52< Soliton> yep.
20090312 23:51:41-!- valer [n=s@parsa.yok.utu.fi] has joined #wesnoth-dev
20090312 23:52:28-!- Shadow_Master [n=imorelle@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090312 23:54:55< zookeeper> in EB, blue advances really slowly. i've recruited a bunch of elvish fighters and am just sitting in my keep.
20090312 23:55:06< zookeeper> r33571
20090312 23:55:56< zookeeper> turn 5 and blue's northernmost unit is a pillager on 19,40
20090312 23:58:22< zookeeper> turn 6, green has 7 units just hanging around a couple hexes east of his keep
--- Log closed Fri Mar 13 00:00:17 2009