--- Log opened Sun Feb 22 00:00:01 2009
20090222 00:01:27< Shadow_Master> Soliton: what's exactly the purpose of the explictly declared branches in your .git/config for the Wesnoth git-svn tree?
20090222 00:02:12< Shadow_Master> I'm trying to fix a different git-svn tree (my campaign from the wesnoth-umc-dev repository) in which master == remote 1.4 branch for some reason
20090222 00:05:00< Soliton> you mean the local tracking branches?
20090222 00:05:02-!- alink_ [n=alink@78.129.18.215] has quit [Read error: 104 (Connection reset by peer)]
20090222 00:05:12< Shadow_Master> yes.
20090222 00:07:31< Soliton> well, they're there.. so you have some local tracking branches..
20090222 00:07:38< Shadow_Master> my branches are: master, 1.4, 1.4@1, trunk, wmllint-test, wmllint-test@148 . And the 'master' refers to the real SVN /branches/1.4
20090222 00:07:56< Shadow_Master> so, um, I am wondering how to "fix" that
20090222 00:08:15< Soliton> fix what?
20090222 00:08:27< Shadow_Master> that master != SVN trunk  :-|
20090222 00:09:23< Shadow_Master> um, "git branch trunk" is creating a new "trunk" branch.
20090222 00:09:31< Soliton> what does your fetch line say?
20090222 00:10:00< Shadow_Master>         fetch = trunk/Invasion_from_the_Unknown:refs/remotes/trunk
20090222 00:10:28< Shadow_Master> above that,  url = https://wesnoth-umc-dev.svn.sourceforge.net/svnroot/wesnoth-umc-dev
20090222 00:11:56< Soliton> delete master and recreate it as a tracking branch from trunk?
20090222 00:12:32< silene> git branch -D master; git branch refs/remotes/trunk master
20090222 00:12:49< silene> (which is what Soliton just suggested in fact)
20090222 00:13:24< Shadow_Master> for first I'd have to checkout a different branch
20090222 00:13:46< Shadow_Master> I tried to do that in other copy of the tree and the operation left me at a branch "named" "(no branch)"
20090222 00:14:35< silene> usually, you get "(no branch)" when you checkout a non-local branch
20090222 00:15:35< zookeeper> boucman, filed that bug for you.
20090222 00:16:17< silene> oops, do not type what i typed, i inverted both names; i don't what would happen, but probably not a good thing
20090222 00:16:35< Shadow_Master> I used it already in a copied tree
20090222 00:16:49 * Shadow_Master removes the copy and restores from a backup
20090222 00:17:27-!- Sapient [n=sapien-x@c-24-126-133-155.hsd1.ga.comcast.net] has joined #wesnoth-dev
20090222 00:18:10< zookeeper> hey sapient
20090222 00:19:15< Shadow_Master> um
20090222 00:19:36< Shadow_Master> this doesn't look any good..
20090222 00:21:48-!- grzywacz [n=grzywacz@moinmoin/developer/karol] has quit [":wq"]
20090222 00:22:19< Sapient> hi zookeeper
20090222 00:22:39< Sapient> I'm cooking some squash so a bit AFK
20090222 00:22:41< zookeeper> esr, i presume no objections to removing the silly unmounted deoran from TSG's into the depths?
20090222 00:23:02< zookeeper> i thought squash was a sport.
20090222 00:24:04< Shadow_Master> git branch foo ; git checkout foo
20090222 00:24:08< esr> zookeeper: Actually,  thought the unmounted Deoran was a nice bit of attention to detail.  Why do you want to remove it?
20090222 00:24:10< Shadow_Master> git branch -D master; git branch master refs/remotes/trunk
20090222 00:24:20< Shadow_Master> is this what you meant silene ?
20090222 00:24:51< Sapient> zookeeper: http://en.wikipedia.org/wiki/Squash_(plant)
20090222 00:25:07< Sapient> I would expect a vegetarian like yourself to know of squash ;)
20090222 00:26:04< zookeeper> esr, hideous graphics, annoying hacky WML to deal with, conceptually IMO lame, since no one else can unmount
20090222 00:26:54< esr> zookeeper: If you really didn't know, "to squash" is also an English transitive verb combining the meanings of "crush" and "smash", but with a lighter, almost homorous intention.
20090222 00:27:07< zookeeper> Sapient, my english plant vocabulary isn't that good
20090222 00:27:28< zookeeper> of course i know that, otherwise i wouldn't understand what every other wesnoth troll is talking about.
20090222 00:27:32< esr> s/homorous /humorous/
20090222 00:28:27< Shadow_Master> silene: thanks.
20090222 00:29:08< esr> About unmounted Deoran...I won't try to stop you if you think he graphics are really too ugly to live, but I'd prefer if the feature were kept and the WML cleaned up.
20090222 00:29:32< Shadow_Master> he looks ridiculous.
20090222 00:29:49< Shadow_Master> he looks like a young child, less than 8 years old. Not worthy of being the leader of the South Guard!
20090222 00:30:05-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has quit ["bbl"]
20090222 00:32:10-!- BenUrban [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
20090222 00:32:21-!- BenUrban [n=benurban@c-68-49-10-243.hsd1.md.comcast.net] has joined #wesnoth-dev
20090222 00:34:44-!- loonycyborg [n=sergey@79.139.247.150] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzz"]
20090222 00:34:45< zookeeper> esr, btw, the resemblance between the second TB map and return to kerlath is uncanny.
20090222 00:34:59< zookeeper> the little lake on the left is at the same spot, the enemy castle too, etc
20090222 00:35:09< zookeeper> i presume you edited the latter out of the former or something ;)
20090222 00:35:26-!- loonybot [n=loonybot@79.139.247.150] has quit [Remote closed the connection]
20090222 00:35:34-!- silene [n=plouf@ASte-Genev-Bois-152-1-37-233.w82-121.abo.wanadoo.fr] has quit [Read error: 110 (Connection timed out)]
20090222 00:36:10< esr> zookeeper: Yes, feel free to mutate it further.  That was my first-ever map and I was cribbing shamelessly; I've gotten better since.
20090222 00:37:35< CIA-70> ilor * r32987 /trunk/src/ (terrain_translation.cpp terrain_translation.hpp): fix bug #13031 by implementing a hard limit on map size. The cap is set to 200 tiles (so 200*200 is the max map area), and as of now is a hardcoded constant in terrain_translation.cpp.
20090222 00:38:33< Sapient> ilor: you do know there was already a MAX_MAP_AREA constant defined, right?
20090222 00:38:46< ilor> it was?
20090222 00:38:52< Sapient> yes
20090222 00:39:26< Sapient> terrain_filter uses it, among other code
20090222 00:40:07< ilor> well, the critical bit of code, namely map deserialization, did not in any way
20090222 00:40:10-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev
20090222 00:40:48< Sapient> that may be true, but it doesn't make sense to have two different definitions of the max
20090222 00:41:28< Sapient> #define MAX_MAP_AREA	65536
20090222 00:41:37< Sapient> map_location.hpp
20090222 00:43:08< Sapient> so the max dimension is 256*256
20090222 00:44:05< ilor> just to clarify, that constant was not used and a huge map was crashing wesnoth
20090222 00:44:27< Sapient> right, I read the discussion
20090222 00:44:35-!- zookeeper [n=l@dsl-tkubrasgw1-fe9ff800-215.dhcp.inet.fi] has quit []
20090222 00:45:13< ilor> oh sweet, it's defined twice
20090222 00:45:20< ilor> in map.hpp too
20090222 00:45:35< Sapient> lol, so it was already in violation
20090222 00:45:51< ilor> it was the same value so probably not
20090222 00:53:45< Sapient> so who made this oversived map?
20090222 00:54:18< Sapient> guess it was a hand-edited file
20090222 00:54:33< Shadow_Master> or script-generated
20090222 00:55:43< Sapient> I mean, how big was it? 
20090222 00:56:39< ilor> Sapient: to make it crash on my system, I used a 3*3M
20090222 00:58:28< ilor> the map data takes up some 30MB uncompressed, but it gzips to mere kilobytes
20090222 01:00:10-!- crimson_penguin [n=ben@CPE0013f7fa2460-CM0013f7fa245c.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20090222 01:01:55< Shadow_Master> http://www.wesnoth.org/forum/viewtopic.php?p=341260#p341260  ?
20090222 01:02:11< Shadow_Master> pfff, someone adds the [spoiler] tag and the first thing people do is abuse it?
20090222 01:03:31< Sirp_> why is that abusing it?
20090222 01:03:56< Shadow_Master> because the whole thread is a giant spoiler! :-D
20090222 01:03:58< ilor> Shadow_Master: that's the point of the spoiler tag right there
20090222 01:07:04-!- thespaceinvader [n=chatzill@91.108.177.73] has quit ["ChatZilla 0.9.84 [Firefox 3.0.6/2009011913]"]
20090222 01:07:12< Sirp_> well yeah I guess if you label the thread as (SPOILERS!) then you don't need to use spoiler tags inside it; but that thread isn't labelled like that
20090222 01:07:36< Shadow_Master> now it is ;-)
20090222 01:08:54-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has quit ["Food."]
20090222 01:12:57-!- cib0 [n=cib@p4FD0EEC2.dip.t-dialin.net] has quit [Remote closed the connection]
20090222 01:15:46-!- Noyga [n=noyga@AVelizy-151-1-27-8.w82-124.abo.wanadoo.fr] has quit ["Quitte"]
20090222 01:26:16-!- nital is now known as nital_
20090222 01:26:52-!- nital_ is now known as nital
20090222 01:29:02< Dragonking> good night everyone
20090222 01:29:08< Sapient> gn, Dragonking
20090222 01:30:56-!- Dragonking [n=dk@dedikerad/dragonking] has quit []
20090222 01:32:15-!- nital [n=nital@unaffiliated/nital] has quit ["Konversation terminated!"]
20090222 01:36:05< ilor> gn
20090222 01:36:09-!- ilor [n=user@213-238-122-211.adsl.inetia.pl] has quit []
20090222 01:36:13-!- Sapient [n=sapien-x@c-24-126-133-155.hsd1.ga.comcast.net] has left #wesnoth-dev []
20090222 01:37:26-!- Elvish_Pillage2 is now known as Elvish_Pillbot
20090222 01:42:19-!- nital_away [n=ldobrogo@unaffiliated/nital] has joined #wesnoth-dev
20090222 01:45:02-!- Elvish_Pillage2 [n=eli@66-189-14-42.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090222 01:46:48< CIA-70> ai0867 * r32988 /trunk/data/campaigns/Descent_Into_Darkness/scenarios/03_A_Haunting_in_Winter.cfg: Get the python ai in DiD:03 to work again. (bug #13039)
20090222 01:49:46-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev
20090222 01:56:00< AI0867> wesbot: topic
20090222 01:57:17-!- boucman [n=rosen@140.117.71-86.rev.gaoland.net] has quit [Remote closed the connection]
20090222 02:01:49-!- Elvish_Pillbot [n=eli@66-189-14-42.dhcp.oxfr.ma.charter.com] has quit [Read error: 110 (Connection timed out)]
20090222 02:02:15-!- nital [n=nital@public12656.xdsl.centertel.pl] has joined #wesnoth-dev
20090222 02:03:34-!- nital [n=nital@unaffiliated/nital] has quit [Remote closed the connection]
20090222 02:05:05-!- Elvish_Pillage2 is now known as Elvish_Pillager
20090222 02:07:57-!- Appleman1234 [n=Appleman@131.181.103.69] has joined #wesnoth-dev
20090222 02:08:11-!- Elvish_Pillager is now known as Elvish_Pillage2
20090222 02:09:58-!- nital_away [n=ldobrogo@unaffiliated/nital] has quit ["leaving"]
20090222 02:13:34-!- nital [n=nital@mion.elka.pw.edu.pl] has joined #wesnoth-dev
20090222 02:13:52-!- nital is now known as nital_away
20090222 02:14:29-!- [Relic] [n=[Relic]@adsl-76-229-202-137.dsl.milwwi.sbcglobal.net] has joined #wesnoth-dev
20090222 02:15:28< [Relic]> Hello :)
20090222 02:17:11< Shadow_Master> f
20090222 02:17:14< Shadow_Master> er
20090222 02:17:29< Shadow_Master> irssi is acting up
20090222 02:18:16-!- Elvish_Pillage2 [n=eli@66-189-14-42.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!"]
20090222 02:24:42< AI0867> ugh, the pythonAI finding code is abusing get_binary_file_location
20090222 02:25:07< AI0867> the proper way to fix this seems to be to move the whole data/ai directory to data/core/ai
20090222 02:25:47< AI0867> possibly breaking formulaAI in the process...
20090222 02:28:09< CIA-70> dave * r32989 /trunk/src/ (4 files): attempt to fix crash bug when cancelling a download
20090222 02:30:35-!- AndreLuiz [n=AndreLui@189.71.43.112] has quit [Read error: 101 (Network is unreachable)]
20090222 02:44:16-!- Shadow_Master_ [n=imorelle@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev
20090222 02:44:24-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has quit [Nick collision from services.]
20090222 02:44:34-!- Shadow_Master_ is now known as Shadow_Master
20090222 02:45:54-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has quit [Client Quit]
20090222 02:47:13-!- Thrawn [n=chatzill@pool-72-93-191-102.bstnma.east.verizon.net] has joined #wesnoth-dev
20090222 02:51:50-!- Sirp_ [n=me@c-71-198-3-104.hsd1.ca.comcast.net] has quit ["leaving"]
20090222 02:57:19-!- Sirp_ [n=me@c-71-198-3-104.hsd1.ca.comcast.net] has joined #wesnoth-dev
20090222 03:10:04-!- happygrue [n=Wintermu@c-67-176-145-41.hsd1.in.comcast.net] has joined #wesnoth-dev
20090222 03:20:42-!- nital [n=nital@public12833.xdsl.centertel.pl] has joined #wesnoth-dev
20090222 03:31:28< CIA-70> soliton * r32990 /trunk/src/server/ (game.cpp simple_wml.cpp simple_wml.hpp): allow printing of a node and use it for [side] debugging
20090222 03:31:59< CIA-70> soliton * r32991 /trunk/src/server/game.cpp: 
20090222 03:31:59< CIA-70> only change scenario WML once the game started
20090222 03:31:59< CIA-70> fixes a bug with open slots when a player leaves before the game started
20090222 03:33:25< nital> what does this "NETWORK MANAGED CALLED!" in console is supposed to mean?
20090222 03:33:53< nital> it shows whenever I choose a map for "local game"
20090222 03:34:28< Soliton> just debugging stuff. i'll make it only show on the proper level in a bit.
20090222 03:35:31< nital> thanks, was afraid that maybe my changes were causing that
20090222 03:45:49-!- nital [n=nital@public12833.xdsl.centertel.pl] has quit [Read error: 60 (Operation timed out)]
20090222 03:48:14-!- nital [n=nital@public13409.xdsl.centertel.pl] has joined #wesnoth-dev
20090222 03:49:52-!- nital [n=nital@public13409.xdsl.centertel.pl] has quit [Remote closed the connection]
20090222 04:09:53-!- AndreLuiz [n=AndreLui@189.71.41.237] has joined #wesnoth-dev
20090222 04:33:42-!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has quit [Read error: 60 (Operation timed out)]
20090222 04:36:49-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090222 04:37:12-!- cjhopman [n=chris@68-118-159-7.dhcp.mdsn.wi.charter.com] has joined #wesnoth-dev
20090222 04:45:36-!- Ivanovic_ [n=ivanovic@dtmd-4db26ddc.pool.einsundeins.de] has joined #wesnoth-dev
20090222 05:02:56-!- Ivanovic [n=ivanovic@dtmd-4db26e4e.pool.einsundeins.de] has quit [Read error: 110 (Connection timed out)]
20090222 05:03:34-!- Ivanovic_ is now known as Ivanovic
20090222 05:11:24-!- Turuk [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has quit [Read error: 104 (Connection reset by peer)]
20090222 05:11:38-!- Turuk [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has joined #wesnoth-dev
20090222 05:14:58-!- Appleman1234 [n=Appleman@131.181.103.69] has quit ["Leaving"]
20090222 05:17:37-!- Corvvs [n=michaelm@pool-71-163-211-230.washdc.east.verizon.net] has quit ["Cheers."]
20090222 05:33:24< CIA-70> soliton * r32992 /trunk/src/ (network.cpp network_worker.cpp): fix a warning and a debug message
20090222 05:43:12-!- DDR [n=chatzill@66.183.125.196] has quit ["ChatZilla 0.9.84 [Firefox 3.0.6/2009011913]"]
20090222 05:44:35-!- Sapient [n=sapien-x@c-24-126-133-155.hsd1.ga.comcast.net] has joined #wesnoth-dev
20090222 05:45:20< Sapient> it looks like the code to set mutated_ as true was lost when game_event command handling was refactored
20090222 05:45:55< Sapient> I'm surprised it wasn't noticed sooner
20090222 05:51:27< Sapient> returning mutated_ as a non-const reference seems like bad style as well
20090222 05:56:55< Sapient> looks like this was a suokko change
20090222 05:59:59< Sapient> the old saying about premature optimizations comes to mind... ~_~
20090222 06:00:24 * Turuk feels that old sayings would be about things that existed before computers ;)
20090222 06:02:16< Sirp_> ummm....have we revoked suokko's svn access?
20090222 06:02:44< Sapient> I dunno... he hasn't been around for a long time
20090222 06:02:52< Sapient> wesbot: seen suokko
20090222 06:02:53< wesbot> Sapient: The person with the nick suokko 33d 13h ago was here and on the channel #wesnoth with the message: Remote closed the connection
20090222 06:02:58< Sirp_> yeah; just seems like a good idea. Would be symbolically satisfying if nothing else. :p
20090222 06:03:58< Sapient> ha, if you're going that route then take him out of the about.cfg ;)
20090222 06:04:13< Sapient> that would be a stronger blow
20090222 06:07:05< Sapient> wow, he's actually returning three private state variables as non-const reference
20090222 06:10:49-!- happygrue [n=Wintermu@c-67-176-145-41.hsd1.in.comcast.net] has quit ["grok 'noth!"]
20090222 06:14:27-!- Thrawn [n=chatzill@pool-72-93-191-102.bstnma.east.verizon.net] has quit [Read error: 110 (Connection timed out)]
20090222 06:16:08-!- [Relic] [n=[Relic]@adsl-76-229-202-137.dsl.milwwi.sbcglobal.net] has quit ["Leaving"]
20090222 06:21:03< CIA-70> sapient * r32993 /trunk/src/ (game_events.cpp game_events.hpp): remove the needless encapsulation violations of class event_handler
20090222 06:31:14< Sirp_> Sapient: I don't think taking him out of about.cfg would be fair. If you contribute things and make it into about.cfg you remain even if we later decide that you don't have the skills to continue development.
20090222 06:32:00< Sirp_> by symbolic I meant more "whew! We won't have to put up with cleaning up bugs after this person anymore!" not "let's teach them a lesson!"
20090222 06:32:31< Sapient> k
20090222 06:34:00< Sapient> ahhh... this probably also was the source of the ESC key bug which I recently pestered Mordante about
20090222 06:34:58< Sapient> I see that when the code that set mutated as true was removed, the code to set skip messages as false was also removed
20090222 06:36:31< Sapient> yup
20090222 06:38:01< Sapient> this was broken back in may/june of last year
20090222 07:24:15-!- Soliton [n=Soliton@rzdspc10.informatik.uni-hamburg.de] has quit [Remote closed the connection]
20090222 07:24:47-!- Soliton [n=Soliton@rzdspc10.informatik.uni-hamburg.de] has joined #wesnoth-dev
20090222 07:24:48-!- Soliton [n=Soliton@rzdspc10.informatik.uni-hamburg.de] has quit [Remote closed the connection]
20090222 07:40:28< CIA-70> sapient * r32994 /trunk/ (changelog src/game_events.cpp): 
20090222 07:40:28< CIA-70> * WML engine:
20090222 07:40:28< CIA-70> * Fixed bug #13024: Conditional [allow_undo] not always working right
20090222 07:45:45< Sapient> cya later, have fun.
20090222 07:45:45-!- Sapient [n=sapien-x@c-24-126-133-155.hsd1.ga.comcast.net] has left #wesnoth-dev []
20090222 07:52:08-!- Mordante [n=chatzill@roadie.xs4all.nl] has joined #wesnoth-dev
20090222 07:52:19< Mordante> mornng
20090222 07:52:40< Mordante> morning*
20090222 07:54:04< dfranke> morning Mordante 
20090222 07:54:17< Mordante> hi dfranke 
20090222 08:23:19< Mordante> soliton Ivanovic save loading used to work, you clicked after the animation the menu opened and you could save load
20090222 08:23:31< Mordante> and I like that feature it helped me a lot to fix some features
20090222 09:15:15< dfranke> I think use of Python AIs is broken in multiplayer.  It seems to use the default AI no matter what you set it to.
20090222 09:15:25< dfranke> Would someone please check my sanity on this before I report it?
20090222 09:15:46< dfranke> I've never used this feature until now so I might be doin it wrong.
20090222 09:18:00< CIA-70> mordante * r32995 /trunk/ (8 files in 5 dirs): 
20090222 09:18:00< CIA-70> Added experimental new addon download dialog.
20090222 09:18:00< CIA-70> This dialog is only available when started with --new-widgets and shouldn't
20090222 09:18:00< CIA-70> add new strings. The dialog is added to ease the testing the listbox
20090222 09:18:00< CIA-70> alignment feature that still needs to be added before 1.6. This dialog won't
20090222 09:18:03< CIA-70> be finished before 1.6.
20090222 09:18:04-!- zookeeper [n=l@dsl-tkubrasgw1-fe9ff800-215.dhcp.inet.fi] has joined #wesnoth-dev
20090222 09:29:50-!- Baufo [n=thomas@62-47-156-228.adsl.highway.telekom.at] has joined #wesnoth-dev
20090222 09:32:08< CIA-70> mordante * r32996 /trunk/ (po/wesnoth-lib/POTFILES.in src/CMakeLists.txt src/SConscript): Add my previous commit to the other build systems.
20090222 09:36:47< zookeeper> hmh, i wonder if the no-scroll-when-mouse-exits-window thing should be removed after all :/ http://gna.org/bugs/?13046 or maybe made an advanced option
20090222 09:43:00-!- boucman [n=rosen@140.117.71-86.rev.gaoland.net] has joined #wesnoth-dev
20090222 09:43:39< Sirp_> boucman: I committed an attempt to fix the receive data bug
20090222 09:43:44< Sirp_> were you able to repro that?
20090222 09:43:45< Sirp_> because I couldn't.
20090222 09:43:48< Sirp_> so if you could test it....
20090222 09:44:13< boucman> I was, I will test
20090222 09:47:26 * dfranke gesticulates wildly at https://gna.org/bugs/index.php?13048
20090222 09:50:03< boucman> ouch, this is bad...
20090222 09:52:05< dfranke> yah.
20090222 09:53:31-!- nital_away is now known as nital
20090222 09:56:31< nital> boucman: hello, patch not breaking the string freeze submitted :)
20090222 09:56:44< boucman> ok, will look at it today
20090222 10:06:10-!- AndreLuiz_ [n=AndreLui@201008164251.user.veloxzone.com.br] has joined #wesnoth-dev
20090222 10:07:12-!- AndreLuiz [n=AndreLui@189.71.41.237] has quit [Read error: 110 (Connection timed out)]
20090222 10:11:15< Ivanovic> moin
20090222 10:13:58< boucman> Sirp_: your code looks good after proofreading, but did you intent to leave the debug printing in (it was changed to DBG so it's fine, but I'm not sure you intended to keep it)
20090222 10:15:11< Sirp_> boucman: ahhh no I didn't.
20090222 10:15:17< Sirp_> boucman: we need to change the send code to
20090222 10:15:18< Sirp_> too
20090222 10:15:32< Sirp_> boucman: my code is not very pretty bug I intend to re-write this networking code eventually
20090222 10:16:32< boucman> Sirp_: it's less of a problem in send, send will timeout eventually, receive can block forever
20090222 10:17:44< Sirp_> but doesn't send have some really huge timeout?
20090222 10:17:59< boucman> yeah :(
20090222 10:23:57< Ivanovic> AI0867, AI0867_: around?
20090222 10:24:23< Ivanovic> AI0867_: dfranke spotted some problem with pythonais but to test if his fixes work, he will need them working in multiplayer, too
20090222 10:24:30< Ivanovic> can you have a look at this?
20090222 10:25:37-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090222 10:29:04-!- grzywacz [n=grzywacz@moinmoin/developer/karol] has joined #wesnoth-dev
20090222 10:29:11< Ivanovic> cf https://gna.org/bugs/?13047
20090222 10:32:33-!- Sirp_ [n=me@c-71-198-3-104.hsd1.ca.comcast.net] has quit ["leaving"]
20090222 10:33:40-!- loonybot [n=loonybot@79.139.247.150] has joined #wesnoth-dev
20090222 10:34:28< Ivanovic> esr: looking at the latest reports by dfranke we will have a 1.4.x release today and with lots of luck (and security issues fied) a 1.5.x release tomorrow
20090222 10:36:35< zookeeper> boucman, i got the filtering for making drakes not fly over fords and such, but the problem is that due to their flies=true movetype attribute, they don't get submerged at all on those terrains.
20090222 10:37:11< boucman> zookeeper: hmm
20090222 10:37:21< boucman> let me look at the doc for a second
20090222 10:37:23< Mordante> do we really need the flies = true?
20090222 10:37:34< zookeeper> that's what i want to find out
20090222 10:37:40< Mordante> better said do we still need it
20090222 10:37:42< zookeeper> it's only for visuals
20090222 10:37:50< zookeeper> i'm gonna test how it looks without it
20090222 10:38:07< boucman> zookeeper: remove the flies=true, and use submerge=0 in the standing animation, you should have the result you want
20090222 10:39:04< boucman> Mordante: it's needed as a shortcut, but probably could be replaced by a macro
20090222 10:39:13< boucman> post 1.6 cleanups, I'd say
20090222 10:40:11-!- loonycyborg [n=sergey@79.139.247.150] has joined #wesnoth-dev
20090222 10:40:33< Mordante> I prefer a macro so we can remove another variable from C++, agreed that it's post 1.6 material
20090222 10:40:36-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20090222 10:41:49-!- silene [n=plouf@ASte-Genev-Bois-152-1-33-133.w83-114.abo.wanadoo.fr] has joined #wesnoth-dev
20090222 10:42:33-!- AndreLuiz_ [n=AndreLui@201008164251.user.veloxzone.com.br] has quit [Read error: 101 (Network is unreachable)]
20090222 10:43:21< zookeeper> boucman, no luck
20090222 10:43:38< boucman> hmm
20090222 10:43:47< zookeeper> tried submerge=0 both in [frame] and as an animation-level one, neither did anything
20090222 10:44:07< boucman> lemme check the code
20090222 10:46:04< boucman> it seems that flying can also be put in the movement type, maybe it's there too
20090222 10:46:21< zookeeper> uh, what?
20090222 10:46:42< boucman> just repeating what I see in the code :P
20090222 10:48:27< boucman> ok, I was not clear
20090222 10:49:14< boucman> could you remove flying from the mvt type, or would that lead to other problems ?
20090222 10:49:26< zookeeper> that's exactly what i've done
20090222 10:49:53< zookeeper> i commented out the flies=true, added submerge=0 to the flying anim, and it still submerges in water (when using that flying anim).
20090222 10:50:35< boucman> hmm
20090222 10:52:33< zookeeper> also tried submerge=0.1, and it changes nothing
20090222 10:52:47< boucman> zookeeper: ok, it's engine side
20090222 10:52:51< boucman> looking a bit deeper
20090222 10:53:10< zookeeper> i'll commit the WML fix anyway
20090222 10:53:46< boucman> yeah, please set submerge to 0.01 not 0
20090222 10:56:03-!- Dragonking [n=dk@dedikerad/dragonking] has joined #wesnoth-dev
20090222 10:56:57< CIA-70> zookeeper * r32997 /trunk/data/core/ (macros/animation-utils.cfg units.cfg): Made flying drakes not use the flying standing animation on fords, reefs and water villages, and removed the flies=true movetype attribute to allow normal submersion on those terrains.
20090222 11:02:32-!- jorda [n=jorda@114.206.219.87.dynamic.jazztel.es] has joined #wesnoth-dev
20090222 11:04:40< Mordante> loonycyborg which version of scons did you want to make the minimum version
20090222 11:05:46< Mordante> ?
20090222 11:06:47-!- stikonas [n=and@ctv-213-164-101-51.vinita.lt] has joined #wesnoth-dev
20090222 11:07:29< loonycyborg> Mordante: You mean in 1.7? 0.98.3
20090222 11:08:12< Mordante> loonycyborg yes thanks, working on a email to get the build system evaluation on track again
20090222 11:16:51-!- ettin [n=jorda@199.207.219.87.dynamic.jazztel.es] has quit [Read error: 110 (Connection timed out)]
20090222 11:19:26 * Ivanovic currently looks at the group registration at freenode
20090222 11:19:42< Ivanovic> that is: giving us the ability to "really" manage the chans and stuff like this
20090222 11:19:49< Ivanovic> including cloaking and whatever
20090222 11:19:50< Mordante> :-)
20090222 11:19:58< Ivanovic> i am authorized to do so, right?
20090222 11:19:59< Ivanovic> ;)
20090222 11:20:06< Ivanovic> cf http://freenode.net/group_registration.shtml
20090222 11:20:20< Ivanovic> "Informal, project-oriented groups.  Informal groups vary considerably in their internal organization. If the group is run by a single project leader or developer, that person should submit an approving contact form. If the group is run by two equal participants, each should submit a approving contact form to make it clear that both wish to have the group registered. If it's run by a larger core group or by voting across the 
20090222 11:20:21< Ivanovic> project, the voting group should make a collective decision to register and should appoint one of their number as approving contact. "
20090222 11:20:28< loonycyborg> Ivanovic: Yes. You can't expect Sirp to bother with that, anyway :)
20090222 11:21:11< Mordante> well you're our general admin so why not here ;-)
20090222 11:22:14< boucman> yup, go for it
20090222 11:24:44< dfranke> I'd rather not use a cloak.  I like waving my IPv6 address around :-)
20090222 11:25:22< loonycyborg> dfranke: You're using a tunnel broker or you ISP switched? :)
20090222 11:25:30< Mordante> Sapient nice commit 32993 :-)
20090222 11:25:30< dfranke> loonycyborg: tunnel broker
20090222 11:26:18< loonycyborg> Which one?
20090222 11:26:25< dfranke> Hurricane Electric.
20090222 11:26:54< dfranke> I get an 11ms ping time to my tunnel endpoint, and 10ms of that is the hop from my house to the DSL switching office.
20090222 11:27:32< dfranke> and my linode is in the *same datacenter* as the tunnel endpoint.
20090222 11:28:10< dfranke> so, modulo configuration hassles, good as native.
20090222 11:28:14< zookeeper> umm, if suokko committed that much bad code, how come it was never noticed before? when he was still around i got the impression he was doing lots of useful stuff
20090222 11:31:50< silene> i'm just guessing here, but it may be something like that: there are some parts of the code where even useful stuff necessarily turns bad, so as long as he was working on the safe parts, everything went well, but once he went to the darker part of wesnoth, murphy's law striked in
20090222 11:32:02< loonycyborg> zookeeper: But his new useful stuff ended up breaking old useful stuff :) 
20090222 11:32:50< Ivanovic> okay, sent the registration form
20090222 11:33:19< Mordante> yep a lot of breakage went undetected for a long time
20090222 11:33:46< Mordante> and some things we expected him to fix it and waited too long to revert it
20090222 11:34:17< Mordante> Ivanovic is the language switch problem also a blocker?
20090222 11:34:36< Ivanovic> Mordante: which one do you mean, the "does not work on windows" thingie?
20090222 11:34:38< Ivanovic> no, it is not
20090222 11:34:48< Ivanovic> since loonycyborg can not reproduce it with current dev versions
20090222 11:34:56< Mordante> no I switch to German and my menus remain English
20090222 11:35:11< Ivanovic> the problem there is that we have *no* idea where the fault could be
20090222 11:35:32< Ivanovic> but in general i would at least call it an important bug
20090222 11:35:36< dfranke> r32997 fixes #13045, rigt?
20090222 11:35:41< loonycyborg> Mordante: Which menus?
20090222 11:35:42< dfranke> zookeeper: ^^
20090222 11:36:03< Mordante> uhm well that we have no idea, shouldn't affect the blocker state :-/
20090222 11:36:25< zookeeper> dfranke, yeah, i'll handle the bug..
20090222 11:36:36< Ivanovic> the "real" problem is that we have noone who is probably able to look at it
20090222 11:36:38< Mordante> loonycyborg the in game menu, the menu button is in German but the menu items are still English
20090222 11:36:40< Ivanovic> though maybe silene can
20090222 11:37:03< Ivanovic> since he already fixed lots of suokkos breakages (this is another one)
20090222 11:37:24< silene> which report?
20090222 11:37:26< Ivanovic> https://gna.org/bugs/?12568
20090222 11:37:32< Mordante> :-)
20090222 11:38:13< dfranke> I suppose #12992 is a suokkoism then?
20090222 11:38:41< silene> Ivanovic: i will take a look
20090222 11:39:00< zookeeper> dfranke, i'm not 100% sure, but i think that's been around as long as the whole back to turn x shortcuts
20090222 11:39:08< zookeeper> (i don't think suokko implemented that?)
20090222 11:39:26< dfranke> zookeeper: 'Back to start' has only been around since 1.5.something.
20090222 11:39:44< dfranke> before that it always just showed a turn number.
20090222 11:39:46< Ivanovic> it *might* be related to suokko changes but i am not sure at all about this one
20090222 11:40:14< zookeeper> dfranke, oh, ok.
20090222 11:40:22< Ivanovic> that some stuff ingame is sometimes/often shown untranslated is more likely caused by suokko
20090222 11:40:29< Ivanovic> due to caching changed
20090222 11:40:53< Ivanovic> afk for a while, good luck silene and thanks for your efforts on fixing the breakage suokko left us with
20090222 11:41:03< boucman> zookeeper: the hurricane drake seems still to be considered a flying unit...
20090222 11:41:29< zookeeper> boucman, right, i'll handle that..
20090222 11:41:43< boucman> ok, which one did you already do so I can test ?
20090222 11:42:19< zookeeper> the burner and fighter lines
20090222 11:42:23< boucman> k
20090222 11:42:49< zookeeper> i forgot that sky and hurricane have the fly movement type, not drakefly
20090222 11:43:38< zookeeper> ah, excellent, the flies= attribute can be overridden in [unit_type]
20090222 11:44:21< CIA-70> boucman * r32998 /trunk/src/ (formula_ai.cpp formula_ai.hpp unit_frame.cpp): fix a slight problem with the submerge animation parameter always being overridden by the engine.
20090222 11:46:05< CIA-70> boucman * r32999 /trunk/src/formula_ai.hpp: revert accidental commit of debug code
20090222 11:46:17< CIA-70> boucman * r33000 /trunk/src/formula_ai.cpp: revert accidental commit of debug code
20090222 11:49:14< CIA-70> zookeeper * r33001 /trunk/data/core/units/drakes/ (Hurricane.cfg Sky.cfg): Fixed submersion problem.
20090222 12:03:28-!- Noyga [n=noyga@AVelizy-151-1-27-8.w82-124.abo.wanadoo.fr] has joined #wesnoth-dev
20090222 12:15:03< Dragonking> boucman: I looked a bit at formulas
20090222 12:15:13< boucman> and ?
20090222 12:15:18< Dragonking> boucman: I need to fix move and move_partial functions
20090222 12:15:32< Dragonking> I found few ways when wesnoth crashes when using these
20090222 12:15:36< boucman> what's wrong with them ?
20090222 12:15:42< boucman> oh, ok
20090222 12:15:55< boucman> this whole thing is less  mature than we thought
20090222 12:16:04< Dragonking> Much less
20090222 12:16:16< Dragonking> Well, it is mine code, that crashes :)
20090222 12:16:57< Dragonking> I could commit a simple fix, but I prefer to find better solution
20090222 12:17:23< boucman> take your time, that not blocking for 1.5.11
20090222 12:17:31< Dragonking> Yup
20090222 12:17:51-!- ilor [n=user@213-238-122-211.adsl.inetia.pl] has joined #wesnoth-dev
20090222 12:18:23< Dragonking> ok, I disappear for now, bye
20090222 12:25:41-!- nital is now known as nital_away
20090222 12:37:21-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has quit []
20090222 12:51:46< ilor> Ivanovic: I can fix https://gna.org/bugs/?13046 by making it a advanced preference, this requires a new string however
20090222 12:52:26< boucman> ilor: I don't understand the bug at all
20090222 12:53:12< ilor> boucman: it's about scrolling when Wesnoth is windowed, and likely not maximized
20090222 12:53:37< boucman> ok
20090222 12:53:45< ilor> the "new" way is better for someone who switches between apps a lot and doesn't want the game to scroll
20090222 12:53:59< ilor> the "old" way seems to be better for actual gameplay
20090222 12:54:10< zookeeper> yeah
20090222 12:54:29< zookeeper> both ways are potentially annoying depending on what you tend to want to do
20090222 12:55:30< ilor> I'd defult to the "old" method and keep the new one around via advanced preferences
20090222 12:55:54< zookeeper> sure, the old is probably a better default if we're gonna have an option
20090222 12:57:04< zookeeper> Mordante, btw, one thing which is rather annoying is that scrolling upwards activates even if you have your pointer on a top bar button (going to access the menus) or some status panel (to check tooltips)
20090222 12:58:08< zookeeper> i'm not sure how that could be fixed...for buttons it'd be simple (just don't activate scrolling if the pointer is on a button, even if it's also near the screen edge), but the status panels cover most of the top bar width
20090222 12:58:54< Ivanovic> ilor: then add it
20090222 13:00:28< ilor> Ivanovic: roger
20090222 13:00:36 * zookeeper tries to think up a string for that
20090222 13:03:05< zookeeper> "stop mouse scrolling if pointer exits window" is a bit long :p
20090222 13:03:07< CIA-70> ivanovic * r33002 /trunk/ (7 files in 6 dirs): updated Lithuanian translation
20090222 13:03:40< ilor> zookeeper: "Keep scrolling when mouse leaves window"?
20090222 13:03:47< Ivanovic> zookeeper: how much are python AIs used in mainline and in addons?
20090222 13:04:00< zookeeper> Ivanovic, not at all in mainline except for that one DiD scenario
20090222 13:04:02< Ivanovic> that is: we got a *serious* security issue with the python code run from wesnoth
20090222 13:04:22< zookeeper> in add-ons...eh, i'm not sure. surely not in more than a couple, max.
20090222 13:04:33< Ivanovic> a clean option of fixing this would be completely removing ingame python support since then no arbitratry code can be executed
20090222 13:04:41< zookeeper> no one knows how to write one anyway :p
20090222 13:04:53< Ivanovic> that is *GOOD*
20090222 13:05:03< zookeeper> i wouldn't oppose, the whole feature has been about a dozen times more trouble than what it's been worth AFAICT :)
20090222 13:05:37< boucman> nital_away: ping me when you're back
20090222 13:05:43< Ivanovic> esr, dfranke, AI0867, AI0867_, elias: do you oppose completely removing the option to have python AI active for 1.6?
20090222 13:06:00< Ivanovic> zookeeper: is there a "short term fix" for DiD?
20090222 13:06:14< loonycyborg> Ivanovic: There's no security issue with python, since it's run in 'crippled' mode.
20090222 13:06:17< Ivanovic> so that i can right off deactivate python support in our build systems
20090222 13:06:22< Ivanovic> loonycyborg: there IS one
20090222 13:06:26< Ivanovic> https://gna.org/bugs/index.php?13048
20090222 13:06:42< CIA-70> silene * r33003 /trunk/src/serialization/tokenizer.cpp: Ensures the tokenizer starts with the same domain the parser is initialized with. Fix for bug #12568.
20090222 13:06:46< Ivanovic> because the sandbox has some pretty problematic wholes
20090222 13:06:58< zookeeper> Ivanovic, well, i was hoping on dragonking on doing his formula ai fixes on that. but if he won't make it in time, i can write some simple WML substitute for that too
20090222 13:07:00 * Ivanovic cheers for the fix silene commited
20090222 13:07:11< zookeeper> ilor, yeah, that sounds ok to me
20090222 13:07:26< Ivanovic> zookeeper: would it be very problematic if this part if just commented out for the moment?
20090222 13:07:49< Ivanovic> loonycyborg: you agree that there is a serious security issue with python?
20090222 13:08:03 * stikonas gives cookie to silene.
20090222 13:08:10< stikonas> Ivanovic: strings in data/campaigns/Legend_of_Wesmere/scenarios/17_Breaking_the_siege.cfg:325 and data/campaigns/Legend_of_Wesmere/utils/deaths.cfg:116 are very similar
20090222 13:08:12< ilor> gah, I triggered a full rebuild by modifying preferences.hpp
20090222 13:08:35< stikonas> Ivanovic: maybe it is possible to remove 1 string.
20090222 13:08:37< Ivanovic> stikonas: what is the string?
20090222 13:08:49< stikonas> Before you die, know that you lost in Wesmere, you lost at Tath, you will lose here and that your Great Chief is dead!
20090222 13:08:54< stikonas> Before you die, know that you lost in Wesmere, you lost in Tath, you will lose here and that your Great Chief is dead!
20090222 13:09:32< Ivanovic> the question is: what is "Tath", is it a town, a river, a battlefield close to town, ...
20090222 13:09:50< Ivanovic> short guess: it is a town and it was a battle close to that town
20090222 13:10:06< Ivanovic> so that the string with 'in' can be removed, the one with 'at' be kept
20090222 13:10:23< stikonas> Ivanovic: the fortified human settlement at Tath
20090222 13:10:46< Ivanovic> okay
20090222 13:11:06-!- thespaceinvader [n=chatzill@91.108.177.73] has joined #wesnoth-dev
20090222 13:12:46< zookeeper> Ivanovic, no, not problematic
20090222 13:13:18< zookeeper> i can remove it if you want
20090222 13:13:43< Ivanovic> zookeeper: please do so
20090222 13:13:49< CIA-70> ivanovic * r33004 /trunk/data/campaigns/Legend_of_Wesmere/scenarios/17_Breaking_the_siege.cfg: unify the string with the one available in utils/death, thanks stikonas for spotting this one
20090222 13:16:51< loonycyborg> Ivanovic: Personally, I think that sandboxing python is empty waste of time. I'm not even sure why would you want to.
20090222 13:17:18< Ivanovic> loonycyborg: to not be able to place some code on the addon server that can go bogus in your system
20090222 13:17:45< loonycyborg> .py files already cannot be placed on the addon server.
20090222 13:18:35< zookeeper> (without them having been deemed safe some way)
20090222 13:18:38< zookeeper> (right?)
20090222 13:19:02< loonycyborg> They require moderator validation.
20090222 13:19:34< esr> Ivanovic: I've been looking at the IRC scrollback, but I don't understand the Python security issue.  What is it?
20090222 13:19:47< Ivanovic> https://gna.org/bugs/index.php?13048
20090222 13:20:01< Ivanovic> login to gna to be able to see this report, it is marked private for obvious reasons
20090222 13:20:10< CIA-70> silene * r33005 /trunk/src/ (8 files in 4 dirs): Removed now unused parameter.
20090222 13:21:19< esr> Looking...
20090222 13:21:37< zookeeper> well, if the moderator validation thing is still working, then there's not much of a security problem
20090222 13:21:45< zookeeper> but i still wouldn't mind the feature being dropped ;)
20090222 13:23:23< esr> I'm opposed to droopping the feature because iit would effectively kill off the possibility of writing custom AIs.  No way that's gonna happen in C++.
20090222 13:24:18< esr> I do see the point that sandboxing is reaky, however.  dfranke made that one well.
20090222 13:24:24< Ivanovic> esr: and for me it is more important to have WML not able to touch anything outside of the wesnoth path than the option to write own AIs
20090222 13:24:25< esr> s/reaky/leaky/
20090222 13:24:53< esr> The question becomes, can we trust out moderation system?
20090222 13:25:03< boucman> esp since the feature has been around for a long time and has been almost unused
20090222 13:25:12< Ivanovic> the problem is that there are still lots of other ways like eg attachments in the forums
20090222 13:27:18< esr> That's kind of a "You deserve to loose" situation,. though.
20090222 13:27:36< esr> s/loose/lose...gahh, not being able to tyope...
20090222 13:27:40< loonycyborg> Well.. I think that you could sandbox python by embedding it in C++ and strictly controlling what you expose to it..
20090222 13:28:37< Ivanovic> for 1.6 i prefer a short term solution and would welcome some better embedded scripting ways for ai and other scripting in 1.7+
20090222 13:28:39< esr> loonycyborg: Do you think you can do that?
20090222 13:29:01< boucman> zookeeper: how do you reproduce the item on water bug ?
20090222 13:29:03< Ivanovic> that is: lua seems to be good for sandboxing and formula ai should mature and be "better" than instead of python for the ais
20090222 13:30:31< loonycyborg> esr: I'd like to work on some python/C++ integration for wesnoth sometime in the future, but I already got enough other things to do for now.
20090222 13:30:34< zookeeper> boucman, uh...it's in the report
20090222 13:31:05< boucman> k
20090222 13:31:09< boucman> :blush:
20090222 13:31:33< esr> Eh?  How do we know the same problem won't recur with lua?  Seems to me it's generic to any language with import facilities and name handling enough like Python.  Lue probably qualifies.
20090222 13:31:45< loonycyborg> Ivanovic: Are you *sure* that lua doesn't have similar problems?
20090222 13:31:52< Ivanovic> loonycyborg: no, i am not
20090222 13:32:09< Ivanovic> loonycyborg: we will have to make sure that a scripting language inside wesnoth is *really* sandboxed well
20090222 13:33:03< loonycyborg> I think that we should find a way to avoid strict requirements on sandboxing.
20090222 13:33:17< esr> loonycyborg: Agreed.
20090222 13:34:00< esr> Basically, I don't think sandboxing to the level we're imagining is really possible with a Turing-complete language.
20090222 13:34:12< Ivanovic> i do see it as requirement that nothing not really wesnoth related can be touched from wml
20090222 13:34:30< esr> chrooting, maybe?
20090222 13:34:33< Ivanovic> IMO that is a reasonable expectation players can have
20090222 13:34:51< Mordante> zookeeper scrolling upwards where?
20090222 13:35:20< Ivanovic> silene: i think your fix for the untranslated text clearly deserves a changelog entry
20090222 13:35:22< Ivanovic> ;)
20090222 13:35:40< esr> Ah.  Any reason not to chroot to the Wesnoth data tree before game execution?  That would be pretty ironbound.
20090222 13:35:58< loonycyborg> esr: Th reason is windows :P
20090222 13:36:00< Ivanovic> esr: yeah: where are the saves?
20090222 13:36:13< Ivanovic> and the usermade content
20090222 13:36:24< esr> Ivanovic: Good point.  Hmmm..
20090222 13:36:26< zookeeper> Mordante, when mouse scrolling upwards. you know, mouse to the top edge of the screen :P
20090222 13:36:26-!- malumalu [n=malu@hnvr-4dbbf85e.pool.einsundeins.de] has joined #wesnoth-dev
20090222 13:36:29< boucman> esr: is it possible to sandbox in windows ?
20090222 13:37:41< esr> Windoes is so intrinsically fuckred, security-wise, that I don't consider we can make the problem any worse, really. no matter what we do or fail to do.
20090222 13:37:58< esr> But save location is a real issue.
20090222 13:38:48< Mordante> zookeeper ah ok, I never did anything with that code, so wanted to make sure what you meant ;-)
20090222 13:39:17< boucman> esr: though I agree about windows security, I don't think this kind of attitude is responsible
20090222 13:40:02< esr> OK, let's define the challenge here.  Would it be sufficient to control custom AIs' access to the file system, or do we think their ability to run processes and open network sockets is a serious attack vector?
20090222 13:41:13< boucman> launching a process which can access network seems an easy way to bring in a nastier trojan in...
20090222 13:41:26< Mordante> esr my biggest problem with python is that it wasn't meant to be used as a sandboxed plugin language and lua was designed for that goal
20090222 13:41:53< Mordante> bolting security afterwards is often a bad idea, which won't work properly
20090222 13:43:06 * dfranke can't sleep
20090222 13:43:16< esr> Mordante: Wheras I strongly suspect there will be analgous vulnerabilities in lua or any other Turing-complete language, so we can't look to the design of the language itself as a solution.
20090222 13:43:16 * dfranke reads scrollback
20090222 13:43:50< dfranke> IIRC, PyPy has a "total sandboxing" mode that compiles Python into a program that does nothing but read from stdin and write to stdout...
20090222 13:44:07-!- cib0 [n=cib@p4FD0F94F.dip.t-dialin.net] has joined #wesnoth-dev
20090222 13:44:16< dfranke> when it tries to make a system call, it prints a request to stdout.
20090222 13:44:18< cib0> hello
20090222 13:44:24< dfranke> and a wrapper program is expected to handle it for it.
20090222 13:44:44< Mordante> hi cib0 
20090222 13:44:51< esr> I googled "Sandboxing Python".  It;s a research area.  No good solutions.
20090222 13:44:54< cib0> hi Mordante
20090222 13:44:58< loonycyborg> esr: What does this have to do with Turing-completeness? Even a Turing-complete language can't do any harm if it has no functions to access files, run subprocesses, etc.
20090222 13:45:00< dfranke> there's the JVM, which doesn't necessarily imply Java.
20090222 13:45:07< esr> Googling "Sandboxing lua"...
20090222 13:45:12< cib0> esr, what about pypy?
20090222 13:45:16< Mordante> esr yes it was also in that state a few years ago
20090222 13:45:18< dfranke> and PLT Scheme has a security system similar to the JVM.
20090222 13:45:40< Mordante> esr lua is designed to be used sandboxed, not even sure you can use it standalone
20090222 13:47:00< silene> esr: lua is fine from a sandboxing point of view
20090222 13:47:58< cib0> it seems to be made for that stuff after all
20090222 13:47:58-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has quit [Read error: 110 (Connection timed out)]
20090222 13:48:01 * esr is reading...
20090222 13:48:56< cib0> http://codespeak.net/pypy/dist/pypy/doc/sandbox.html
20090222 13:49:19< esr> OK, lua sandboxing looks better.  There may be some very subtle attacks, but...
20090222 13:49:23< silene> (and just in case someone points it out, i know that the "dofile" function is still available in my lua patch, it's just for testing purpose and will be disabled later)
20090222 13:50:52< boucman> please let's try to focus on the problem at hand
20090222 13:51:08< boucman> what can we do about this in the 1.6 time frame 
20090222 13:51:14< esr> Hm.  I didn't know about pypy.  That looks interesting too, though it sounds like there might be serious performance issues.
20090222 13:51:41< dfranke> compiled pypy is faster than CPython.  The issue would be dependencies.
20090222 13:51:48< dfranke> We'd have to require GCC.
20090222 13:53:06< dfranke> and it'd need a great deal of wrapper code to work.  You basically have to write your own POSIX emulation layer.
20090222 13:53:26< boucman> not realistic under windows...
20090222 13:53:51< esr> There are research-level problems here.  We're not going to solve them before 1.6 release.
20090222 13:54:30< loonycyborg> boucman: If python ais weren't a serious issue for 1.4 they won't for 1.6 :)
20090222 13:54:46< dfranke> for 1.6 I propose just declaring all Python AIs unsafe.  Let anything in the system data directory run unrestricted.  Don't let anything in the user data directory run at all unless the user enables it.
20090222 13:55:02< Mordante> loonycyborg they weren't since we were unaware of the problem
20090222 13:55:07< boucman> loonycyborg: except that we are going to kill 1.4 pretty soon, whereas 1.6 is not born yet
20090222 13:55:15 * cib0 agrees with dfranke.
20090222 13:55:56< silene> how many uses of python are there in the system data? why not disable python globally unless the user asks for it?
20090222 13:56:06< cib0> are there any cases where python AIs are used by default?
20090222 13:56:12< boucman> zookeeper: I have no idea how to solve the [item] problem realistically
20090222 13:56:22< dfranke> cib0: yes, DiD scenario 3.
20090222 13:56:32< dfranke> nowhere else currently.
20090222 13:56:37< esr> dfranke: That sounds like a good way to avoid getting lost in a technical jungle of how to sandbox perfectly.
20090222 13:56:45< cib0> can't that scenario use formula instead?
20090222 13:57:09< silene> then let's just remove this scenario temporarily; why bother with a whole subsystem for a single scenario?
20090222 13:57:09< loonycyborg> cib0: Someone will have to make that formula :)
20090222 13:57:27< Ivanovic> cib0: it can and Dragonking will work on it "soon"
20090222 13:57:47< Ivanovic> silene: i already asked zookeeper to have this python include be removed
20090222 13:57:52< cib0> sounds good
20090222 13:58:02< Ivanovic> in general python provides us with the "option to write own ais"
20090222 13:58:11< Ivanovic> so far this option is not well adopted at all, correct?
20090222 13:58:24< cib0> Ivanovic, pretty much
20090222 13:58:34< Ivanovic> so in general i think it won't hurt if we completely deactivate python for the whole 1.6 series
20090222 13:58:55< Ivanovic> and come up with a good and nicely working scripting solution that allows fancy stuff for the 1.7 series
20090222 13:58:59< loonycyborg> Ivanovic: Nope. Some people were interested in using python.
20090222 13:59:04< Ivanovic> including a way of making it "reasonable" safe
20090222 13:59:16< Ivanovic> loonycyborg: so far python is *only* useable for AI anyway
20090222 13:59:23< Dragonking> zookeeper, Ivanovic, cib0 working on some fixes to make that formula work
20090222 13:59:32< esr> Ivanovic: It bothers me on a philosophical level, though -- it's a retreat from making the game language more flexible and powerful.
20090222 13:59:42< Ivanovic> loonycyborg: so people who want to use it in general will have to wait for the 1.7 tree anyway
20090222 13:59:48< Dragonking> It may take some time tho, and I doublt it'll be ready today
20090222 14:00:08< Ivanovic> esr: the reason simply is that security is more important than flexibility
20090222 14:00:17< Mordante> esr true but finding a good solution will take a while and quite some testing, not something we can do in three weeks
20090222 14:00:33< Ivanovic> esr: and i am *all* for having the thing flexible, especially with more advanced scripting in general
20090222 14:00:40< Mordante> so is Python important enough to delay 1.6 for quite a bit
20090222 14:00:53< Ivanovic> that is why i can live with the step back for 1.6 on AIs and having a good and full fledged scripting support in 1.7
20090222 14:00:59< loonycyborg> Ivanovic: Though personally I'm disappointed with the way python is currently used. I.e. only of ais.
20090222 14:01:25< Ivanovic> loonycyborg: like i said: further stuff on scripting is clearly 1.7 material and really should be done there
20090222 14:01:38< Mordante> well personally I'm happy with it due to this security issue
20090222 14:02:01< esr> loonycyborg: I agree, but for displacing C++ it might be that lua is a better option.  I'm beginning to suspect that might be true even though I like Python.
20090222 14:02:03< Ivanovic> loonycyborg: more python scripting than the current ai stuff won't be enabled for 1.6.x anyway
20090222 14:02:54< esr> ...like Python and don't know lua yet, I meant.
20090222 14:03:01< Ivanovic> so the question at the moment is: is python AI stuff (more we don't have) important enough that it has to be included in 1.6 and thus will lead to a probably bigger delay of it?
20090222 14:03:10< loonycyborg> esr: I more care for displacing WML than C++ since I still didn't quite got the hang of WML :)
20090222 14:03:48< esr> Ah, WML is easy. Spiky in spots, but easy.
20090222 14:03:50< Ivanovic> for the then to be started trunk we are free to experiment with scripting languages
20090222 14:04:20< dfranke> Is anyone opposed to what I suggested above?
20090222 14:04:43< loonycyborg> dfranke: I am not.
20090222 14:04:49< esr> dfranke: I think I'm with you, but repeat the suggestion please.
20090222 14:05:05< loonycyborg> [15:54:44] <dfranke> for 1.6 I propose just declaring all Python AIs unsafe.  Let anything in the system data directory run unrestricted.  Don't let anything in the user data directory run at all unless the user enables it.
20090222 14:05:18< esr> +1
20090222 14:06:07-!- boucman [n=rosen@140.117.71-86.rev.gaoland.net] has quit [Read error: 104 (Connection reset by peer)]
20090222 14:07:00< zookeeper> boucman, ok, no problem. as i said it's not an especially important thing to fix.
20090222 14:08:10-!- boucman [n=rosen@140.117.71-86.rev.gaoland.net] has joined #wesnoth-dev
20090222 14:08:41-!- silene [n=plouf@ASte-Genev-Bois-152-1-33-133.w83-114.abo.wanadoo.fr] has quit [Read error: 60 (Operation timed out)]
20090222 14:12:10< cib0> loonycyborg, even a scripting language cant get you around learning WML, since you still need to work with WML variables and all
20090222 14:12:46< dfranke> I would love to have Lua bindings that can do everything WML can do, though.
20090222 14:12:58-!- silene [n=plouf@ASte-Genev-Bois-152-1-97-226.w86-218.abo.wanadoo.fr] has joined #wesnoth-dev
20090222 14:13:28< esr> I tyhink we're trying toi decide between dfranke's proposal and dropping Python AIs out of 1.6 entirely.  Can we reach a conclusion here?
20090222 14:13:55< Ivanovic> i think those are the two options we got at the moment
20090222 14:14:05< esr> loonycyborg: cib0 is correct, WML is probably unavoidable.
20090222 14:14:28< Mordante> IMO we should disable it for 1.6
20090222 14:14:37< Ivanovic> i prefer to disable it, too
20090222 14:14:47< Mordante> if we decide to keep it at least warn the packagers so they can decide for themselves
20090222 14:14:47< cib0> i vote for dfranke's proposal
20090222 14:14:51< Ivanovic> just to be 100% sure and since it is basically unused anyway it should not hurt
20090222 14:14:51< silene> i would rather suggest that we disable python by default in 1.6 (and we something about this lone scenario) and if the user wants it, it can enable
20090222 14:15:31< cib0> having python atleast available will encourage people to experiment with it
20090222 14:15:36< zookeeper> isn't the current add-on server python moderation enough for 1.6?
20090222 14:15:50< Ivanovic> cib0: it already was "just available" for two stable releases
20090222 14:15:54< esr> So: dfranke, cib0, and me back dfranke's proposal.  Mordante, Ivanovic,m and silene want to drop Pyton. 
20090222 14:15:55< Ivanovic> and people did not experiment with it
20090222 14:16:15< esr> We're kind of split down the middle.
20090222 14:16:18< cib0> Ivanovic, i did ;) well, i tried, but it didnt work >.<
20090222 14:16:22< silene> esr: i don't want to drop it, i want to disable it by default
20090222 14:16:30< esr> Ah.
20090222 14:16:35< Mordante> esr disable, the code can remain just won't be compiled
20090222 14:16:47< zookeeper> how is it _not_ effectively disabled by default atm?
20090222 14:16:49< Ivanovic> as in "built system default is no python"
20090222 14:17:04< dfranke> Mordante: that's not good enough. You shouldn't be able to make something insecure by setting build options.
20090222 14:17:09< dfranke> It has to be a runtime option.
20090222 14:17:12< silene> zookeeper: you mean that the DiD campaign does not work?
20090222 14:17:32< zookeeper> silene, there's no security issue with DiD since it's a mainline campaign.
20090222 14:17:52< loonycyborg> Ivanovic: I think that python should be enabled by default in build systems, but disabled by default with a runtime option.
20090222 14:17:52< Mordante> dfranke you can add a warning with the build flag
20090222 14:17:54< zookeeper> the only problem we have is malicious add-ons, is it not?
20090222 14:18:07< zookeeper> i don't see why anyone would care about malicious mainline campaigns
20090222 14:18:07< Mordante> but of course you can also make it a runtime switch if you enable it in the build
20090222 14:18:11< dfranke> Mordante: the binary might be used by people other than the person who built it.
20090222 14:18:28< Mordante> dfranke hence my second line ;-)
20090222 14:18:28< dfranke> Mordante: it shouldn't be incumbent upon the person who build the binary to pass along that warning.
20090222 14:18:41< silene> zookeeper: no, that's not what i meant, i wanted to understand why you say it is currently disabled
20090222 14:18:51< Mordante> still I prefer to disable it in the build by default
20090222 14:19:21< loonycyborg> Mordante: I don't share this sentiment :)
20090222 14:19:27< Mordante> I expect some packagers not to enable the build option
20090222 14:19:30< dfranke> I guess the default build setting should depend on what we plan to do with it in 1.7.
20090222 14:19:41< ilor> if we disable it in the default build and this gets shipped as the windows binary, for most of our users python will be disabled for good
20090222 14:19:46< dfranke> If we plan to keep it supported, it should be enabled.  If we plan to deprecate it and replace it with Lua, it should be disabled.
20090222 14:19:52< boucman> let's disable python for tomorow's 1.5 and say in the release note that DID is known to be broken
20090222 14:19:52< Mordante> that's silly since we don't know what we do in 1.7 with it
20090222 14:19:53< zookeeper> silene, ok, i take it back and rephrase: is it not true that those problems which disabling is supposed to help with don't currently even exist (because of the server moderation)?
20090222 14:20:00< boucman> that leaves us some more time to discuss and decide
20090222 14:20:08< stikonas> zookeeper: can you add quote marks to data/core/units/nagas/Myrmidon.cfg:86 in order to correctly show portrait
20090222 14:20:26< dfranke> boucman: we don't have to ship a broken DiD.  We can just switch that one scenario back to the default AI.
20090222 14:20:35< boucman> oh, ok
20090222 14:20:42< zookeeper> stikonas, no time right now
20090222 14:20:52< boucman> why does it have a specific ai in the first place ? what's the need ?
20090222 14:21:13< dfranke> I haven't looked at the code, but here's my guess:
20090222 14:21:23< zookeeper> the need is that the default ai parameters can't make a side attack a specific other unit
20090222 14:21:25< dfranke> One of the two enemies has its keep right next to the signpost you have to reach...
20090222 14:21:30< stikonas> Ivanovic: maybe you have time to add quote marks to data/core/units/nagas/Myrmidon.cfg:86 in order to correctly show myrmidon+female portrait
20090222 14:21:32< dfranke> The AI is called 'kamikaze'.
20090222 14:21:35< zookeeper> i just tried a WML workaround, and it just can't be done AFAICT, which is retarded
20090222 14:21:41< loonycyborg> boucman: It's needed to make rebelious Malin's servant attack preferably him.
20090222 14:21:50< loonycyborg> *servants
20090222 14:21:51< zookeeper> when sirp drops by i'm going to ask him whether he knows how [target] actually works
20090222 14:22:01< zookeeper> and whether it could be changed to work in a non-stupid way
20090222 14:22:02< dfranke> oh, num.
20090222 14:22:05< dfranke> nm*
20090222 14:22:16< dfranke> I was thinking of the following scenario.
20090222 14:23:40< dfranke> Anyway, the whole rebellion thing can go as far I'm concerned.  It's a nice plot device but tactically I find it annoying.
20090222 14:24:09< zookeeper> dfranke, i'm pretty confident it could be made to work in a less annoying manner
20090222 14:25:58< zookeeper> the only mainline python ai was required simple because there was no friggin way to make the default ai want to prefer attacking a certain unit. -.-
20090222 14:26:06< zookeeper> simply, even
20090222 14:27:04< zookeeper> as i've said before, if anyone wants to really really help scenario authors to make better AI's, then please help fix the existing default AI parameters :p
20090222 14:27:31< zookeeper> but luckily i'm about to go for a while now, so i don't need to go on my usual rant about the subject
20090222 14:27:37< boucman> zookeeper: what you want seems to be a good job for a formula AI..
20090222 14:27:46< dfranke> ok, I reeeeeally need to get back to sleep now.
20090222 14:28:09< dfranke> see you all in the morning, or more likely the afternoon at this point.
20090222 14:28:35< zookeeper> boucman, yeah, that's what DK is working on now
20090222 14:29:17< zookeeper> anyways, afk ->
20090222 14:31:15-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20090222 14:34:44-!- Soliton [n=Soliton@rzdspc10.informatik.uni-hamburg.de] has joined #wesnoth-dev
20090222 14:36:59< boucman> Soliton: you've added the output method to simple_wml, havn't you
20090222 14:37:12< boucman> would it be ok to rename it to Debug() ?
20090222 14:37:20< boucman> so it's named like the normal wml method
20090222 14:39:47< Soliton> sure. i named it like the method to output a complete document though.
20090222 14:41:57< Dragonking> boucman: Commited first fix, but I need to go now 
20090222 14:42:13< Dragonking> Still need to improve this pathfinding thing...
20090222 14:42:16< Dragonking> bye
20090222 14:42:28-!- Dragonking [n=dk@dedikerad/dragonking] has quit []
20090222 14:50:17< esr> Ivanovic: Still there?
20090222 14:50:48< Ivanovic> esr: currently busy, i need at least 15mins
20090222 14:51:42< esr> Ivanovic: OK, it's not urgent.
20090222 15:01:50< ilor> loonycyborg: what is the status of scons's comment ignoring feature?
20090222 15:02:04< Ivanovic> esr: just write your question/task and i'll tackle it when i find a free moment
20090222 15:02:25-!- BroodKiller [n=BroodKil@chello087207235243.chello.pl] has joined #wesnoth-dev
20090222 15:03:18< BroodKiller> boucman: are multiple alpha infos allowed in the frame-specific tags or just the global ones?
20090222 15:05:38< esr> Ivanovic: I think we should offer dfranke dev access.  His security analysis and participation in the design discussion have both been top-quality.
20090222 15:06:11< Ivanovic> yes, sounds reasonable
20090222 15:07:39< esr> Thanks. I know himm ftrom elsewhere; his performance didn't surprise me much.
20090222 15:08:42< loonycyborg> ilor: I haven't heard anything new about it :P
20090222 15:08:57< Ivanovic> dfranke: so if you want to, i can add you in the list of people with gna access
20090222 15:10:14< esr> I think he's afk.
20090222 15:10:49< Ivanovic> esr: i know, but he will read the backlog
20090222 15:12:04< esr> Food enough.
20090222 15:12:13< esr> Good enough, I mean.
20090222 15:12:28 * Mordante looks for the food ;-)
20090222 15:18:04< crimson_penguin> Ivanovic: btw, I asked, and none of the computers didn't have Office
20090222 15:18:22< crimson_penguin> I'll try asking in #macosx later
20090222 15:18:26< crimson_penguin> I have to catch a bus now though
20090222 15:18:32< crimson_penguin> hejdå
20090222 15:18:40-!- crimson_penguin [n=ben@CPE0013f7fa2460-CM0013f7fa245c.cpe.net.cable.rogers.com] has quit ["Leaving"]
20090222 15:20:27-!- BroodKiller [n=BroodKil@chello087207235243.chello.pl] has quit ["Leaving."]
20090222 15:43:22-!- nital [n=kvirc@unaffiliated/nital] has joined #wesnoth-dev
20090222 15:44:21-!- nital [n=kvirc@unaffiliated/nital] has quit [Client Quit]
20090222 15:45:14-!- nital_away is now known as nital
20090222 15:45:26< nital> boucman: ping
20090222 15:46:53< CIA-70> mordante * r33007 /trunk/CMakeLists.txt: Add the low mem option.
20090222 15:47:11< CIA-70> mordante * r33008 /trunk/src/campaign_server/campaign_server.cpp: Removed some unused parameters forgotten in r33005.
20090222 15:48:03< CIA-70> dragonking * r33009 /trunk/src/ (formula_ai.cpp formula_ai.hpp): Attampt to fix formula move and move_partial functions - still work in progress
20090222 15:48:17-!- YogiHH [i=YogiHH@c225088.adsl.hansenet.de] has joined #wesnoth-dev
20090222 15:48:19< CIA-70> ilor * r33010 /trunk/changelog: changelog update
20090222 15:48:35< Soliton> Mordante: re "you clicked after the animation the menu opened" that's how it still works.
20090222 15:48:55-!- YogiHH [i=YogiHH@c225088.adsl.hansenet.de] has quit [Client Quit]
20090222 15:49:18< Mordante> Soliton ok 
20090222 15:49:45< Soliton> so do we mark bug #12997 as won't fix?
20090222 15:50:23< Soliton> we could allow opening the menu and only disable saving of course.
20090222 15:50:44< Mordante> I prefer to keep saving
20090222 15:50:46-!- YogiHH_ [i=YogiHH@c225088.adsl.hansenet.de] has joined #wesnoth-dev
20090222 15:50:59-!- YogiHH_ [i=YogiHH@c225088.adsl.hansenet.de] has quit [Client Quit]
20090222 15:51:06< Mordante> I see no reason to remove it and it helps to fix certain bugs
20090222 15:51:16< Soliton> so won't fix then?
20090222 15:52:11< Mordante> yes, maye the new gui will fix it but I see no reason to keep it open
20090222 15:52:20< CIA-70> ilor * r33011 /trunk/projectfiles/VC9/wesnoth.vcproj: add a missing header to the vc9 project file
20090222 15:52:20< CIA-70> ilor * r33012 /trunk/projectfiles/VC9/wesnoth.vcproj: another VC9 projectfile update
20090222 15:53:18< Soliton> well, the only fix is to disable saving unless you want to fix saving while attacking.
20090222 15:53:44< Mordante> is that broken then? it used to work
20090222 15:54:03< CIA-70> zookeeper * r33013 /trunk/data/campaigns/Descent_Into_Darkness/ (ais/ scenarios/03_A_Haunting_in_Winter.cfg): Removed kamikaze.py and replaced it with a crude WML workaround for now.
20090222 15:54:07< Soliton> in 1.0?
20090222 15:54:27< Soliton> it's the reason why the menu is disabled while attacking...
20090222 15:54:34< Mordante> no later not sure when exactly
20090222 15:54:57< Mordante> I used it to fix some bugs which happened on certain death events
20090222 15:55:06< CIA-70> loonycyborg * r33014 /trunk/projectfiles/ (3 files in 3 dirs): Comitted McShark's projectfiles update.
20090222 15:55:11< loonycyborg> Wow! git's three-way merge rocks :)
20090222 15:55:21< Mordante> in which case the unit in question just didn't want to die
20090222 15:56:59< Soliton> i wouldn't quite call that "used to work" ;-)
20090222 15:57:12< zookeeper> umm...regarding 12997: couldn't you make it so that if the player clicks on a menu during an animation, the animation immediately gets accelerated to 16x or whatever, so the menu opens almost instantaneously?
20090222 15:57:30< zookeeper> (and once the menu is open, of course revert back to whatever the speed was previously)
20090222 15:57:37< Soliton> the exploit with saving while attacking has existed as long as it was possible to save while attacking i'd imagine.
20090222 15:57:45< nital> why a patch may be reported as "malformed", by what, and why?
20090222 15:58:15< nital> since boucman told me that and I have no idea what to do: https://gna.org/patch/?1124
20090222 15:58:20< Mordante> I know it can be abused for save loading, but if that makes people happy I don't mind it
20090222 15:59:27< zookeeper> nital, maybe he just clicked on the link in your previous comment, which doesn't work, and not the link in the attachment list, which works.
20090222 16:00:49-!- Baufo [n=thomas@62-47-156-228.adsl.highway.telekom.at] has quit [Remote closed the connection]
20090222 16:02:37< Soliton> Mordante: https://gna.org/bugs/?4233
20090222 16:02:51< Soliton> Mordante: i'm not sure what exactly you're suggesting.
20090222 16:03:00< Soliton> Mordante: you didn't like that fix or what?
20090222 16:04:48< Mordante> no don't really like it
20090222 16:05:03< Mordante> well I might reenable it in debug mode when I feel the need for it
20090222 16:05:45< Soliton> why would you need to save mid-attack?
20090222 16:07:08< Mordante> sometimes a bug only happens under rare conditions
20090222 16:07:25< Soliton> somehow i think we're talking past each other.
20090222 16:07:41< Ivanovic> yes, you two clearly are
20090222 16:07:47< Soliton> if you save mid-attack the attack is over when you load.
20090222 16:07:59< Soliton> it's impossible that that helps you in any way.
20090222 16:08:16< Ivanovic> i think Mordante means "in the middle of an AI turn"
20090222 16:08:19< silene> cib0: concerning the issue with lua and tutorial savegame, it's just that i had forgotten to mark the event as "first_time_only=no"
20090222 16:08:27< Soliton> so there is no benefit i can see for saving mid-attack as opposed to after the attack.
20090222 16:08:28< silene> cib0: concerning the creation of events on the fly, it doesn't work, contrarily to what i thought: the game has already destroyed the event by the time it is called, hence causing a segfault (if lucky)
20090222 16:08:41< Ivanovic> Soliton: i like the idea zookeeper had
20090222 16:09:01< Soliton> Ivanovic: sure, if someone wants to spend the effort. :-)
20090222 16:09:07< Ivanovic> to increase game speed to 16x when clicking on menu and an attack or another movement currently takes place
20090222 16:09:16< Soliton> maybe it's easy. no idea.
20090222 16:09:17< Ivanovic> it would be rather clean
20090222 16:11:07< Mordante> I mean saving after a unit attack during the ai turn so I can redo what that unit did and get another result
20090222 16:11:30< Soliton> that works as always.
20090222 16:11:51< Mordante> ok good
20090222 16:17:07< cib0> silene, why will the game destroy the event?
20090222 16:20:16< cib0> what gui toolkit would you guys use for a cross-platform tool?
20090222 16:20:34< Ivanovic> qt
20090222 16:22:07< cib0> eeww, qt :/
20090222 16:22:27< silene> cib0: the game assumes that events created on-the-fly physically exist in the original wml, so it only keeps a reference to the original location; in the lua case, there is none, the reference is dangling (or it could be made to point to a memory-leaked object, hence not destroyed)
20090222 16:22:54< Mordante> cib0 there's also the wesnoth widget toolkit :-)
20090222 16:23:22< silene> i thought my effort to rewrite the config system was orthogonal to the lua patch; i guess this is no longer true...
20090222 16:23:30< cib0> Mordante, a great toolkit indeed =)
20090222 16:23:44< Mordante> the best :-)
20090222 16:24:45< cib0> silene, i think it would make more sense to dump the event into the scenario state once it's loaded
20090222 16:25:19< silene> cib0: as a matter of fact, that's what happens once you save and reload a scenario ;-)
20090222 16:29:43< CIA-70> ilor * r33015 /trunk/src/network_worker.cpp: fix compilation error on win32 (hello stupid min/max macros in windows.h)
20090222 16:32:23-!- Alesis-Novik_ [n=alesis@78.60.188.59] has quit ["Back later"]
20090222 16:32:33< Mordante> ilor you can disable these macros 
20090222 16:32:47< ilor> Mordante: that's what I did in this commit :)
20090222 16:33:40< Mordante> ah oke didn't look to close at the fix
20090222 16:34:01< Mordante> only really sad microsoft didn't fix that ages ago
20090222 16:34:19< stikonas> Mordante: can you add quote marks to data/core/units/nagas/Myrmidon.cfg:86 in order to correctly show Myrmidon portrait? It is broken now.
20090222 16:34:54< Mordante> stikonas yes 
20090222 16:40:25< Mordante> esr around?
20090222 16:40:32-!- Soliton [n=Soliton@rzdspc10.informatik.uni-hamburg.de] has quit ["Lost terminal"]
20090222 16:41:40< CIA-70> mordante * r33016 /trunk/ (changelog data/core/units/nagas/Myrmidon.cfg): 
20090222 16:41:40< CIA-70> Fix a broken reference to the Myrmidon portrait.
20090222 16:41:40< CIA-70> The problem was spotted by stikonas.
20090222 16:42:24-!- AndreLuiz [n=AndreLui@189.71.66.24] has joined #wesnoth-dev
20090222 16:48:36-!- Soliton [n=Soliton@rzdspc10.informatik.uni-hamburg.de] has joined #wesnoth-dev
20090222 17:04:54-!- silene [n=plouf@ASte-Genev-Bois-152-1-97-226.w86-218.abo.wanadoo.fr] has quit [Read error: 110 (Connection timed out)]
20090222 17:04:58-!- silene [n=plouf@ASte-Genev-Bois-152-1-63-127.w83-114.abo.wanadoo.fr] has joined #wesnoth-dev
20090222 17:06:45-!- zookeeper2 [n=l@dsl-tkubrasgw1-fe9ff800-215.dhcp.inet.fi] has joined #wesnoth-dev
20090222 17:11:37-!- zookeeper [n=l@dsl-tkubrasgw1-fe9ff800-215.dhcp.inet.fi] has quit [Nick collision from services.]
20090222 17:11:43-!- zookeeper2 is now known as zookeeper
20090222 17:16:51 * Soliton just finished reading the log.
20090222 17:17:03< Soliton> so is there another 1.4 release coming soon?
20090222 17:17:18 * loonycyborg hopes not :(
20090222 17:17:47< Soliton> that much trouble to package?
20090222 17:18:33< loonycyborg> Yep. It involves tweaking paths in multiple projectfiles.
20090222 17:18:44< Ivanovic> Soliton: no, there won't be a 1.4 release
20090222 17:18:50< Ivanovic> since the "fix" is serverside only
20090222 17:19:10< Ivanovic> and basically has already happened since nothing with python is sent to "end users"
20090222 17:19:25< Soliton> ok.
20090222 17:19:38< Ivanovic> (you know, in general moderation required and now that we know the problem we won't allow anything through...)
20090222 17:20:00< Soliton> yeah, the log wasn't clear on whether you realised that. :-)
20090222 17:23:54< Mordante> esr around?
20090222 17:29:28-!- Corvvs [n=michaelm@pool-71-163-211-230.washdc.east.verizon.net] has joined #wesnoth-dev
20090222 17:38:35-!- elias [n=elias@allegro/developer/allefant] has joined #wesnoth-dev
20090222 17:41:38< Ivanovic> elias: good to see you around
20090222 17:41:43< Ivanovic> please read todays irclog
20090222 17:41:55< Ivanovic> and have a look at the bugtracker, you will see what i mean
20090222 17:43:18 * elias knew he should not start up IRC and think there won't be distractions...
20090222 17:44:09-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has quit [Read error: 110 (Connection timed out)]
20090222 17:44:28< Mordante> in that case you really picked the wrong day
20090222 17:45:19< elias> hm, the log is long... what should i be looking for?
20090222 17:45:21< elias> heh
20090222 17:47:23< elias> about the Python security, the threading module should not have been whitelisted
20090222 17:47:56< Ivanovic> elias: the sandbox we got seems to be, uhm, not really secure anyway
20090222 17:48:05< elias> yes, in general, it's a flawed idea maintaining our own "secure python"
20090222 17:48:06< Ivanovic> i am currently writing a mail to the ml about todays issues
20090222 17:48:32< elias> I'd say remove the security layer, and disable Python AIs by default
20090222 17:48:48< elias> and the few (if any) uses should be rewritten with formula AI for the 1.6 release
20090222 17:49:32< elias> removing threading module from the whitelist will fix all *known* security holes again - but some others probably remain
20090222 17:50:26< Ivanovic> elias: let me write the mail and then just answer it
20090222 17:51:09< malumalu> mh, is it normal the help text at the bottom of the main menu is not anti-aliased or is it a bug?
20090222 17:51:32-!- happygrue [n=Wintermu@c-67-176-145-41.hsd1.in.comcast.net] has joined #wesnoth-dev
20090222 18:00:01-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has quit [Read error: 104 (Connection reset by peer)]
20090222 18:00:14-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has joined #wesnoth-dev
20090222 18:08:24< Ivanovic> okay, sent a mail to the dev-ml
20090222 18:08:31< Ivanovic> everyone: please read it and reply
20090222 18:08:45< zookeeper> you really want everyone to reply ;)?
20090222 18:08:49< Ivanovic> i hope i have not forgotten anything
20090222 18:08:59< Ivanovic> zookeeper: that is at least when they got something to say in the matter
20090222 18:14:08-!- stikonas_ [n=and@ctv-213-164-101-51.vinita.lt] has joined #wesnoth-dev
20090222 18:16:16-!- stikonas [n=and@ctv-213-164-101-51.vinita.lt] has quit [Read error: 110 (Connection timed out)]
20090222 18:16:54-!- Turuk [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has quit [Read error: 104 (Connection reset by peer)]
20090222 18:16:56< loonycyborg> Ivanovic: Getting rid of python dependency won't help packagers much :)
20090222 18:17:22< loonycyborg> E.g. python dependency is no-brainer on Gentoo.
20090222 18:17:30-!- Turuk [n=Turuk@FUSE-WS1-208-102-202-188.fuse.net] has joined #wesnoth-dev
20090222 18:18:19< Ivanovic> loonycyborg: on mac, yeah
20090222 18:18:26< Ivanovic> but what about windows or osx?
20090222 18:18:27 * loonycyborg originally decided to learn python because Portage is written in it
20090222 18:18:43-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has quit []
20090222 18:18:52< Ivanovic> on osx it is clearny *not* a no brainer, you have to ship a fairly big framework to have 2.4 available in osx 10.4
20090222 18:19:24< loonycyborg> On windows you just ship an additional dll and some .py/.pyd files
20090222 18:20:12< Mordante> zookeeper does this properly remove the python dependency in DiD? http://svn.gna.org/viewcvs/wesnoth?rev=33013&view=rev
20090222 18:20:30< Mordante> and can it be ported to 1.4?
20090222 18:20:42< Ivanovic> Mordante: in 1.4 there is no python needed in did
20090222 18:20:52< Mordante> Ivanovic it is
20090222 18:20:52< loonycyborg> It was somewhat hard to work out what should be shipped at first, but this work is already done..
20090222 18:21:57< zookeeper> Mordante, yes, that should remove all of it
20090222 18:21:59-!- yann [n=dwitch@nan92-1-81-57-214-146.fbx.proxad.net] has joined #wesnoth-dev
20090222 18:22:17< Ivanovic> Mordante: i don't want to release anoth 1.4.x version if i don't have to
20090222 18:22:35< zookeeper> you could easily port those changes to 1.4 if you'd want
20090222 18:23:00< Mordante> Ivanovic here it is http://svn.gna.org/svn/wesnoth/branches/1.4/data/campaigns/Descent_Into_Darkness/ais/kamikaze.py
20090222 18:23:19< Mordante> Ivanovic I assume Rhonda might want to do another release
20090222 18:23:33< Rhonda> Another release of ...?
20090222 18:23:43< Mordante> wesnoth1.4
20090222 18:23:53< Ivanovic> Mordante: in 1.4.x we are "reasonably safe" as in nothing can come in from the official addon server
20090222 18:24:34< Rhonda> That's not reasonable enough, I'm afraid.
20090222 18:25:19< Rhonda> People are known to download stuff from around the net. I am happy that it's possible to keep the wesnoth community quite tight together, but even then it should get addressed.
20090222 18:25:44< Rhonda> You can't solve an outlook problem with claiming that the bug in exchange got fixed. :)
20090222 18:26:03< Mordante> yes you can Microsoft does that quite often :-P
20090222 18:26:14 * Mordante hides
20090222 18:26:26< Ivanovic> Rhonda: so what kind of fix can be done in 1.4.x beside completely removing python support?
20090222 18:26:47-!- Thrawn [n=chatzill@pool-72-93-191-102.bstnma.east.verizon.net] has joined #wesnoth-dev
20090222 18:27:28< loonycyborg> Ivanovic: Removing whatever whitelisted python module that caused it.
20090222 18:27:59< Rhonda> loonycyborg: It requires checking all whitelisted python modules, not just the reported one.
20090222 18:28:37< Mordante> imo the difference is we terminate support for 1.4 in about a month, debian in about 2.5 years
20090222 18:29:30< elias> Ivanovic: I replied to the mail
20090222 18:30:22< Ivanovic> so the option for debian is to disable it in the binaries
20090222 18:30:42< loonycyborg> Is it enabled currently?
20090222 18:30:54< Mordante> yes but it would then be nice to have a patch for DiD, hence my question
20090222 18:30:56< Rhonda> Ivanovic: And thus break the -did campaign?
20090222 18:30:57< Ivanovic> most likely it is
20090222 18:31:21< Ivanovic> Rhonda: the patch for did: http://svn.gna.org/viewcvs/wesnoth?rev=33013&view=rev
20090222 18:31:26< Mordante> Rhonda http://svn.gna.org/viewcvs/wesnoth?rev=33013&view=rev this patch might fix it
20090222 18:31:27< Rhonda> loonycyborg: If it wouldn't I definitely would have received an MK bugreport about the broken campaign. :P
20090222 18:31:36< Mordante> or better said might apply
20090222 18:32:36< Rhonda> How to disable it in the compile? --disable-python?
20090222 18:34:03< Mordante> yes
20090222 18:34:30< Rhonda> Does any ai work then? The default ai isn't in python, is it?
20090222 18:34:57< loonycyborg> No, it isn't.
20090222 18:35:07< loonycyborg> It's in C++
20090222 18:38:58< Rhonda> Good.
20090222 18:40:57-!- Sirp_ [n=me@c-71-198-3-104.hsd1.ca.comcast.net] has joined #wesnoth-dev
20090222 18:41:22< loonycyborg> Rhonda: esr wouldn't agree with you :)
20090222 18:42:11< Ivanovic> loonycyborg: i think it is a "good, since removing python support won't break the game itself"
20090222 18:42:33< loonycyborg> Ivanovic: I understand. That's why I added :)
20090222 18:42:51< Rhonda> loonycyborg: I couldn't care less, to be honest. And I could misread your statement as "ESR doesn't care for safety" like you misread mine.
20090222 18:44:14< loonycyborg> Rhonda: I didn't misread it. I delibirately twisted it for humorous purposes.
20090222 18:44:25< Rhonda> And even ESR wouldn't go as far as to push python everwhere where it causes more troubles than gain.
20090222 18:44:30< elias> i really was hoping for the Python people to re-introduce a sandboxing module with Python 1.6
20090222 18:45:59< elias> but since they didn't, I think now something like the scripting language starting with L actually would be more appropriate for Wesnoth
20090222 18:46:05< Mordante> are there any concrete plans to introduce such a module
20090222 18:46:41< elias> doesn't seem so, it must be about 2 years now since safe.py was added
20090222 18:47:03< elias> and there was no change in this time at all
20090222 18:47:18< Mordante> doesn't sound hopeful
20090222 18:50:10-!- stikonas_ is now known as stikonas
20090222 18:52:26< loonycyborg> Btw there's a Boost.Python look alike for Lua. Perhaps silene should consider using it..
20090222 18:53:56< Mordante> in the vault?
20090222 18:54:15< elias> I wonder if there's a Python look alike implementation for Lua
20090222 18:54:15< loonycyborg> No. AFAIK it's a separate project.
20090222 18:54:27< elias> so it would use Lua, but scripts could use a Python-like syntax :)
20090222 18:55:03< loonycyborg> http://www.rasterbar.com/products/luabind.html
20090222 18:55:51< Mordante> isn'tt that what allacrost also uses
20090222 18:55:58< Mordante> isn't*
20090222 18:56:22< loonycyborg> Last time I checked they used lua directly.
20090222 18:56:41< Mordante> they used some binding but were very unhappy with it
20090222 18:57:06< elias> using Lua directly isn't much of a problem, it's way more high level than the Python-C API :)
20090222 18:57:20< elias> no reference counting or endless filling out of structures required
20090222 18:58:00< loonycyborg> How do you expose C++ classes with lua api?
20090222 19:01:13< Mordante> yep allacrost uses luabind
20090222 19:04:47< silene> loonycyborg: by providing a proxy object with overloaded __index and __newindex methods; in other words, whenever someone accesses a field of your proxy object, you can transmit the modifications to the real C++ object
20090222 19:06:18< loonycyborg> silene: And you do so manually? like switch(method) { case method_1 ... ?
20090222 19:08:46< silene> loonycyborg: at the algorithmic level, this is what you eventually get at
20090222 19:08:53-!- Corvvs [n=michaelm@pool-71-163-211-230.washdc.east.verizon.net] has quit []
20090222 19:09:18< loonycyborg> Then I'm not surprised that allacrost uses luabind
20090222 19:12:42< CIA-70> ilor * r33017 /trunk/projectfiles/VC9/ (. wesnoth.sln wesnoth.vcproj): add a Debug (fast) build type in the VC9 project file that creates a debug binary without run-time checks. This makes MSVC debugging more feasible as starting the game with the run-time checks takes several minutes.
20090222 19:13:37< cib0> from what ive seen, lua seems to be the first language i really like
20090222 19:14:10< cib0> it doesn't seem to be doing pointless black magic and stuff like that
20090222 19:16:53-!- EdB [n=EdB@228.12.95-79.rev.gaoland.net] has joined #wesnoth-dev
20090222 19:17:11< CIA-70> zookeeper * r33018 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/ (4 files): Some visual tweaks to cutscenes and disabled carryover reports and linger mode in story scenarios.
20090222 19:18:38-!- Alesis-Novik [n=alesis@78.60.188.59] has joined #wesnoth-dev
20090222 19:19:43< silene> loonycyborg: luabind is nice, but it is so heavyweight that it only becomes of interest when you have lots of C++ class you want to use in lua; this is hardly the case in wesnoth
20090222 19:22:15-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090222 19:33:17< Mordante> I'm off bye
20090222 19:33:33-!- Mordante [n=chatzill@roadie.xs4all.nl] has quit ["ChatZilla 0.9.84 [Iceweasel 3.0.6/2009020409]"]
20090222 19:38:18< silene> is there any reason for having the boost foreach.hpp file directly in the source? (Sirp_?)
20090222 19:39:05< loonycyborg> IIRC It wasn't available in 1.33.1
20090222 19:40:21< loonycyborg> Besides, it also provides handy #define foreach BOOST_FOREACH
20090222 19:40:37< loonycyborg> Unlike boost's own foreach.hpp
20090222 19:42:53-!- Elvish_Pillager [n=eli@66-189-14-42.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090222 19:43:37-!- ikarius [n=ross@c-98-207-225-178.hsd1.ca.comcast.net] has quit []
20090222 19:45:43< silene> sure, we can keep this define if needed, but i would like to get ride of the file and up the requirement to boost 1.34, if possible
20090222 19:45:51-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090222 19:47:46< loonycyborg> I'm not against boost version requirement bump, but be sure to ask others.
20090222 19:52:33< zookeeper> sheesh
20090222 19:52:55 * zookeeper doesn't appreciate the undead tomato surprise in "the vanguard"
20090222 19:55:27< zookeeper> in fact, it's not just unappreciatable, it's retarded -.-
20090222 19:56:48< boucman> nital: around ?
20090222 19:58:00-!- Corvvs [n=michaelm@pool-71-163-211-230.washdc.east.verizon.net] has joined #wesnoth-dev
20090222 20:00:18< nital> boucman: yeah
20090222 20:00:41< boucman> nital: I'm not sure what's wrong with your patch, how did you create it ?
20090222 20:00:49< nital> what did you mean by malformed?
20090222 20:01:13< nital> svn diff > hidden4.diff
20090222 20:01:21< boucman> patching file src/menu_events.cpp
20090222 20:01:21< boucman> patch: **** malformed patch at line 71:  			result = slist.show();
20090222 20:01:44< nital> all right, let me re-create it
20090222 20:01:45< boucman> not sure what happened, could you try svn up then redo the patch with svn diff
20090222 20:02:03< boucman> post it to pastebin, no need to add it to the gna patch
20090222 20:02:18< nital> maybe that's because I tried to make some updates to it manually
20090222 20:03:02-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev
20090222 20:03:07< boucman> hehe
20090222 20:03:10< boucman> yeah, likely
20090222 20:04:48< Shadow_Master> hi
20090222 20:05:49< CIA-70> shadowmaster * r33019 /trunk/src/ (tests/test_version.cpp version.cpp version.hpp): 
20090222 20:05:49< CIA-70> Completely rewrote the version_info parsing algorithm for the
20090222 20:05:49< CIA-70> string-based constructor to allow extreme freedom for the special
20090222 20:05:49< CIA-70> version suffix. This fixes bug #13017.
20090222 20:05:49< CIA-70> As a side effect, version_info's are now "sane" even if they are built from
20090222 20:05:52< CIA-70> strings such as "crapcrap" (it turns into 0.0.0crapcrap).
20090222 20:06:09< CIA-70> shadowmaster * r33020 /trunk/src/addon_management.cpp: Clarify a non-translatable error message.
20090222 20:06:32< Shadow_Master> wesbot: topic
20090222 20:08:23-!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has joined #wesnoth-dev
20090222 20:12:38-!- Sapient [n=sapien-x@c-24-126-133-155.hsd1.ga.comcast.net] has joined #wesnoth-dev
20090222 20:15:45< grzywacz> http://www.bugseverywhere.org/be/show/HomePage
20090222 20:17:27< Sapient> Supported version control backends: Arch, Bazaar, GIT, Mercurial, RCS
20090222 20:18:33-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has joined #wesnoth-dev
20090222 20:18:51< grzywacz> Sure, some people use git. Mentioned as an interesting thing to look at.
20090222 20:19:10< Ivanovic> silene: so were you looking into https://gna.org/bugs/index.php?13037 ?
20090222 20:19:12< jorda> 20:04 < Shadow_Master> hi
20090222 20:19:25< Shadow_Master> um?
20090222 20:19:34< Ivanovic> hi jorda, nice to see you around
20090222 20:19:41< jorda> hmm, -EBUTTON
20090222 20:20:09-!- jorda is now known as ettin
20090222 20:20:24< Ivanovic> ettin: it would be nice to have a "clean" list of the steps needed for generating po files from the docfiles and for generating docfiles from po files
20090222 20:20:45< Ivanovic> ettin: that is: currently the implementation of this in cmake feels strange and it would be nice to know what really is needed
20090222 20:21:40< silene> Ivanovic: no, i thought dfranke was
20090222 20:21:59< Ivanovic> silene: he wanted to look into the stuff with python, that is right
20090222 20:22:09< Ivanovic> no idea about the "lots of data" stuff
20090222 20:22:11< ettin> Ivanovic: I'm still using autotools, I don't know much about cmake
20090222 20:22:21< Ivanovic> ettin: no need for you to know about it
20090222 20:22:42< Ivanovic> ettin: i would just like to have a "normal list" of commands like you would type them in a shell
20090222 20:22:45< cib0> does anyone know of a good lua tutorial?
20090222 20:23:02< Ivanovic> as in "to achive ABC, you have to do X, Y and then Z"
20090222 20:23:14< ettin> Ivanovic: oh, I see
20090222 20:23:14< loonycyborg> Ivanovic: mog did a good job of implementing that stuff. You probably won't achieve anything better.
20090222 20:23:35< Ivanovic> loonycyborg: i would at least like to cleanly understand it and comment the single steps
20090222 20:23:39< cib0> im currently reading the reference manual and you know, learning a language by reading grammar expressions isn't really that effecient =)
20090222 20:24:29< loonycyborg> Ivanovic: Look in po/SConscript for a cleaner implementation.
20090222 20:24:33< CIA-70> shadowmaster * r33021 /trunk/po/wesnoth-editor/es.po: Spanish translation update.
20090222 20:24:59< Ivanovic> loonycyborg: i don't trust this implementation either...
20090222 20:25:00< Ivanovic> ;)
20090222 20:25:36< CIA-70> shadowmaster * r33022 /trunk/ (changelog players_changelog): Forgot to update the changelogs with r33021...
20090222 20:26:02< silene> Ivanovic: i can take a look at it tomorrow; for tonight, i already have enough on my hands (deadlines deadlines...)
20090222 20:26:04< Ivanovic> i prefer to have a clean explaination of all the black magic somewhere
20090222 20:26:12< loonycyborg> Ivanovic: Well.. If you'll get a useful list of steps, be sure to share it with me so I could check it.
20090222 20:26:16< ettin> Ivanovic: what exactly do you mean with "docfiles"? are you only interested in po4a stuff?
20090222 20:26:41< Ivanovic> ettin: the stuff used to handle the manpages as well as the manual
20090222 20:26:47< Ivanovic> so basically the dark po4a stuff
20090222 20:27:17< Ivanovic> the "normal" po stuff is handled via wmlxgettext is "reasonable simple" or does not involve much black magic
20090222 20:28:19< Ivanovic> silene: sure, take your time
20090222 20:28:23< Shadow_Master> zookeeper: sound.
20090222 20:28:28< Shadow_Master> zookeeper: sources.
20090222 20:28:44< zookeeper> Shadow_Master, roger
20090222 20:28:48< Ivanovic> silene: before releasing i would like to have the mess with python sorted out as well as this one
20090222 20:29:02< Ivanovic> (so that the two CVE are done in 1.5.11)
20090222 20:29:39 * Shadow_Master threatens zookeeper with releasing a bunch of hungry lions
20090222 20:30:08-!- boucman1 [n=rosen@2.117.71-86.rev.gaoland.net] has joined #wesnoth-dev
20090222 20:30:18< cib0> man.. i have the choice between reading grammar expressions or reading tutorials aimed at newbies to programming.. v.v
20090222 20:30:21< loonycyborg> Ivanovic: both scons and cmake implementation of po4a stuff differ from autotools by calling po4a tools directly instead of using po4a driver.
20090222 20:31:15< ettin> loonycyborg: po4a driver?
20090222 20:31:28< nital> boucman1: http://wesnoth.pastebin.com/d38bd085b
20090222 20:31:34< Ivanovic> loonycyborg: like i said, i would like to hear what "really is needed" since i do not understand this part at all
20090222 20:31:54< loonycyborg> ettin: po4a program which calls po4a-gettextize, etc
20090222 20:31:55-!- boucman [n=rosen@140.117.71-86.rev.gaoland.net] has quit [Read error: 110 (Connection timed out)]
20090222 20:32:24< loonycyborg> New build systems call po4a-gettextize and po4a-translate directly.
20090222 20:32:35< boucman1> nital: looking, thx
20090222 20:32:46-!- EdB [n=EdB@228.12.95-79.rev.gaoland.net] has quit ["Konversation terminated!"]
20090222 20:34:56< loonycyborg> It also helps with avoiding duplication between the buildsystems and po4a config files.
20090222 20:37:26-!- Corvvs [n=michaelm@pool-71-163-211-230.washdc.east.verizon.net] has quit []
20090222 20:38:25< Ivanovic> loonycyborg: and for the po4a stuff it would IMO be fine if this was just done in a plain shellscript
20090222 20:38:37< Ivanovic> since it is not needed to have it done in any build system
20090222 20:39:20< Ivanovic> this is stuff i run every now and then to have the doc files up to date (both regarding po files, as well as resulting manpages and html manual files)
20090222 20:39:46< CIA-70> boucman * r33023 /trunk/ (6 files in 3 dirs): add a hidden= attribute to make a side invisible in the status table, patch by nital
20090222 20:41:01< Ivanovic> personally a plain .sh shell skript would be fine for me (if it is possible to parallize the creation of the files in there)
20090222 20:41:45< loonycyborg> How can it be parallelized? Does bash support threading? :)
20090222 20:41:55-!- Corvvs [n=michaelm@pool-71-163-211-230.washdc.east.verizon.net] has joined #wesnoth-dev
20090222 20:41:59< Shadow_Master> git svn rebase
20090222 20:41:59< boucman1> yeah, with the & character
20090222 20:42:07< Shadow_Master> er.
20090222 20:42:41< Shadow_Master> boucman1: why the patch number isn't mentioned in the 33023 commit message ?
20090222 20:43:37< boucman1> cause I forgot to :P
20090222 20:43:40-!- Corvvs [n=michaelm@pool-71-163-211-230.washdc.east.verizon.net] has quit [Client Quit]
20090222 20:43:41-!- boucman1 is now known as boucman
20090222 20:44:51< Shadow_Master> of course. I'm just reminding you that it's better to mention it so the commit message goes to the patch report automatically
20090222 20:45:03< Shadow_Master> then you don't need to post a comment by hand before marking it Done.
20090222 20:45:08< Shadow_Master> (I know you know this)
20090222 20:46:09< boucman> Shadow_Master: I didn't want to close the report right away
20090222 20:48:58< Shadow_Master> nital:  (*teams)[team_index].set_hidden( utils::string_bool(hidden, true) );
20090222 20:49:05< Shadow_Master> why is it 'true' by default?
20090222 20:49:36< Ivanovic> loonycyborg: like i said, i don't care much how it is done, i would also go for a plain, hardcoded makefile if that paralizes things nicely
20090222 20:49:42-!- Thrawn [n=chatzill@pool-72-93-191-102.bstnma.east.verizon.net] has quit [Read error: 110 (Connection timed out)]
20090222 20:49:51< Ivanovic> but atm the autotools solution is not really clean and it does not allow paralization
20090222 20:50:13< Shadow_Master> game_events.cpp, I suspect [modify_side].
20090222 20:50:20< loonycyborg> Ivanovic: SCons does allow parallelization
20090222 20:50:27< Ivanovic> loonycyborg: cmake does, too
20090222 20:50:30< Ivanovic> ;)
20090222 20:51:07< Ivanovic> though i would prefer some solution that is clean and not really a build system thingie if possible, since it is only "useless" clutter in there
20090222 20:51:18< loonycyborg> But at a cost of crapping its stuff everywhere :)
20090222 20:52:28< loonycyborg> Po4a and pot-update stuff is implemented in one po/SConscript file.
20090222 20:52:47-!- elias [n=elias@allegro/developer/allefant] has quit ["Leaving"]
20090222 20:52:55< boucman> Shadow_Master: when it's an event to modify the hidden status of a side, it makes as much sense to default to yes than to default to no, I guess
20090222 20:53:01< loonycyborg> While with cmake this ended up being several generated shell scripts and innumerable CMakeLists.txt files.
20090222 20:53:14< Shadow_Master> boucman: noted
20090222 20:53:19< Shadow_Master> nital: otherwise, why [store_side] doesn't reflect [modify_side]'s behavior?
20090222 20:54:37< nital> Shadow_Master: because I didn't have any idea of [store_side+
20090222 20:54:46< nital> before you told me now )
20090222 20:55:12< Shadow_Master> static void store_gold_side(bool store_side, const vconfig& cfg), game_events.cpp line 859
20090222 20:55:27< boucman> nital: complementary patch ? :P
20090222 20:55:34< nital> thanks, I'll read it now :)
20090222 20:56:21< Ivanovic> loonycyborg: and yes, it is possible to parallize in bash scripts
20090222 20:56:32< Ivanovic> cf http://ubuntuforums.org/showthread.php?t=382330
20090222 20:56:43< Ivanovic> post #7 has some nice pseudo code that just goes the right way
20090222 21:01:10-!- Baufo [n=thomas@62-47-145-230.adsl.highway.telekom.at] has joined #wesnoth-dev
20090222 21:02:26< loonycyborg> Ivanovic: I'm sure that it's possible to parallelize with bash, but IMO in this particular case using an insanely flexible buildsystem is more appropriate ;)
20090222 21:03:09< Ivanovic> peronsally i think this is not really what should be part of a build system since it just clutters things with a "rarely used special thing"
20090222 21:03:48< Ivanovic> better leave the build system for what it is meant for: building the program
20090222 21:03:49< Ivanovic> ;)
20090222 21:05:10< loonycyborg> Ivanovic: You're wrong. Building po/html files is as complex as building programs, and in this case the use of buildsystem is fully appropriate.
20090222 21:05:22< zookeeper> isn't menu_events.cpp:424 missing the units/ prefix from the path?
20090222 21:06:33< Shadow_Master> it's a name, not a file path
20090222 21:06:49< Shadow_Master> why it isn't translatable, though?
20090222 21:07:12< zookeeper> well, on line 522 the units/ is there
20090222 21:07:42< Shadow_Master> zookeeper: I just noticed that you are reading different code ;-)
20090222 21:07:45< Shadow_Master> try to report against HEAD.
20090222 21:07:45< CIA-70> boucman * r33024 /trunk/data/core/about.cfg: third stupid mistake commit of the day : forgot to remove the merge markers in about.cfg
20090222 21:07:57< zookeeper> i'm just wondering, since i got an "error display: could not open image 'unknown-unit.png'" at some point when playing a fogged scenario (used the status table etc)
20090222 21:08:27< Shadow_Master> zookeeper: I can't find a prefix-less unknown-unit.png in my copy anyway
20090222 21:08:27< boucman> wesbot: topic
20090222 21:08:36-!- Sapient [n=sapien-x@c-24-126-133-155.hsd1.ga.comcast.net] has quit []
20090222 21:08:59< zookeeper> Shadow_Master, nice, someone fixed it in the last couple of hours :|
20090222 21:09:07< Shadow_Master> yep.
20090222 21:10:57< Shadow_Master> your problem must be coming from other file
20090222 21:11:46< Shadow_Master> ... silly me, I was reading the fixed source code, but forgot to recompile.
20090222 21:12:02-!- Zen_Clark [n=user@99-136-80-191.lightspeed.rcsntx.sbcglobal.net] has joined #wesnoth-dev
20090222 21:17:08< Ivanovic> loonycyborg: ha, an even easier way to parallize in shell: just use xarg
20090222 21:40:48-!- ikarius [n=ross@216.27.182.3] has joined #wesnoth-dev
20090222 21:43:40< nital> boucman: http://wesnoth.pastebin.com/d22a9a636
20090222 21:44:51-!- kitty_ [n=kitty@e180214086.adsl.alicedsl.de] has joined #wesnoth-dev
20090222 21:44:52< nital> it took me about 100 times longer to check if it's really the only place to be changed that actually writing that one line 
20090222 21:45:07< boucman> save/reload the state, I guess ?
20090222 21:51:01< CIA-70> ivanovic * r33025 /trunk/utils/wesnoth-optipng: change optipng script to allow parallization
20090222 21:51:37< CIA-70> boucman * r33026 /trunk/src/game_events.cpp: add save/load of hidden side status
20090222 21:51:56< Ivanovic> loonycyborg: as you see it is easy enough to parallize in shellscripts, nothing speaks against doing so in a possible "plain and easy" script of doc file creation
20090222 21:52:58-!- Noyga [n=noyga@AVelizy-151-1-27-8.w82-124.abo.wanadoo.fr] has quit ["Quitte"]
20090222 21:53:13-!- Elvish_Pillage2 [n=eli@66-189-14-42.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090222 21:53:53-!- Elvish_Pillager [n=eli@66-189-14-42.dhcp.oxfr.ma.charter.com] has quit [Read error: 60 (Operation timed out)]
20090222 21:54:16< boucman> thespaceinvader: your serpent looks great, can't wait to see it ingame
20090222 21:54:34< thespaceinvader> thanks boucman
20090222 21:54:47< thespaceinvader> i just need to persuade some maintainers to put it in campaigns...
20090222 21:55:36< Shadow_Master> Ivanovic: when... next... release?
20090222 21:55:41< boucman> kitty_: darn, you're good http://www.wesnoth.org/forum/viewtopic.php?p=341540#p341540
20090222 21:55:48< boucman> thespaceinvader: shouldn't be too hard
20090222 21:55:51< kitty_> :)
20090222 21:56:05< Ivanovic> Shadow_Master: when security issues are handled
20090222 21:56:11< Ivanovic> at least the two with CVEs
20090222 21:57:35< boucman> Ivanovic: let's disable python for next release, not supress until a decision is taken
20090222 21:57:44< Shadow_Master> CVE?
20090222 21:57:48< boucman> but I really think we can't slip with our planning
20090222 21:57:51< Ivanovic> Shadow_Master: cf wesnoth-dev
20090222 21:57:57< Shadow_Master> ah, k
20090222 21:58:07< Shadow_Master> no, the acronym. But I just looked it up
20090222 21:58:10< Ivanovic> boucman: short guess: tomorrow it is time for a new beta with those two issues fixed
20090222 21:58:19< boucman> ok, sounds good
20090222 22:00:37< Ivanovic> and if noone comes up with a good argument on the -dev ml for leaving python in the next ~14h (just to say a number) i will deactivate it in the build systems and not allow configuring it in there
20090222 22:01:26< Shadow_Master> what will happen to DiD scenario 3?
20090222 22:01:34< Shadow_Master> it uses a simple Python AI
20090222 22:01:46< boucman> it has been replaced by C++ AI
20090222 22:02:00< boucman> dragonking is looking into a FAI replacement
20090222 22:02:07< Shadow_Master> good.
20090222 22:02:11< Shadow_Master> C++ wins.
20090222 22:02:19< Shadow_Master> Python loses. ;-)
20090222 22:02:40< Ivanovic> ahhh, parallization of utils/wesnoth-optipng does *really* help
20090222 22:03:02< Ivanovic> this way finally i can use all my 4 cores and those are now at >95% usage
20090222 22:03:04< Shadow_Master> esr hasn't replied to the PythonAI thread yet? :-(
20090222 22:03:31< Shadow_Master> thespaceinvader: can you let me handle that portrait btw? ;-)
20090222 22:03:39< esr> Shadow_Master: I made my opinion knmown on IRC.
20090222 22:03:42< Ivanovic> jupp, esr has not replied on the dev ml so far
20090222 22:03:45< thespaceinvader> Shadow_Master: sure
20090222 22:03:56< thespaceinvader> as an aside, though, what for?
20090222 22:04:03< Ivanovic> Sirp_: please send a mail to the dev ml with your opinion, too
20090222 22:04:04< Shadow_Master> esr: roger.
20090222 22:04:33< Shadow_Master> thespaceinvader: because I have always handled kitty_'s contributions in the past when I have been around?
20090222 22:04:41< thespaceinvader> ah ok
20090222 22:05:51< Shadow_Master> :-)
20090222 22:06:24< Ivanovic> okay, with parallization in the script the stats about "amount saved" are really screwed
20090222 22:06:40< Ivanovic> but to be honest: i don't care much since the speed improvement is *wow*
20090222 22:07:00< boucman> Ivanovic: do a "df" before and after :P
20090222 22:07:18< Sirp_> Ivanovic: very well.
20090222 22:07:19< Ivanovic> boucman: i think i will got for a more crude approach
20090222 22:07:22< zookeeper> Sirp_, do you recall how [ai] [target] is supposed to actually do exactly? it certainly can't be used to make a side for example always prefer attacking a certain unit out of all possible targets.
20090222 22:07:31< zookeeper> s/how/what
20090222 22:09:44< zookeeper> i mean, even if i specify targets with high values, the AI just won't target them over other units.
20090222 22:11:05< Sirp_> zookeeper: IIRC [target] is ONLY used by the AI when the target is not immediately in range.
20090222 22:11:20 * Shadow_Master . O o
20090222 22:11:30< Sirp_> it's used when the AI is thinking "hey there are no villages in range to grab, no units to attack. What do I actually *do* with my units?"
20090222 22:11:46< Sirp_> (this being a kinda hard thing for the AI to decide, because it's now thinking at the strategic rather than tactical level)
20090222 22:11:57< Shadow_Master> no wonder why my enemy AIs are so dumb then.
20090222 22:12:00< Sirp_> and then it gets a list of targets -- by default unclaimed/enemy villages are targes, as is the enemy leader...
20090222 22:12:02< cib0> i can actually hear campfires now? goo, wesnoth! =D
20090222 22:12:10< zookeeper> Sirp_, ah, interesting
20090222 22:12:12< Sirp_> and it values each o the targets and moves units toward the targets
20090222 22:12:13< Shadow_Master> cib0: ?
20090222 22:12:33< cib0> Shadow_Master, its the first time i stumble upon that feature
20090222 22:12:36 * Sirp_ would STILL like to re-write the AI to be much much better.... :(
20090222 22:12:57< Shadow_Master> cib0: IftU provides campfires with sound sources, but I don't think the sound file is in mainline yet
20090222 22:13:00< grzywacz> cib0, ;-)
20090222 22:13:02< zookeeper> Sirp_, so i wonder if there's a way to do the same thing for the tactical level (when there are units in range)
20090222 22:13:19< CIA-70> ivanovic * r33027 /trunk/utils/wesnoth-optipng: some more optimizations to the script, so that stats actually do work again
20090222 22:13:39< cib0> Shadow_Master, im playing IftU :p
20090222 22:13:58< cib0> aww, it should be put in mainline =(
20090222 22:14:19< Shadow_Master> um I think I can do it
20090222 22:14:37< grzywacz> Shadow_Master, make sure to put it in the ambient dir
20090222 22:14:40< cib0> it adds atmosphere to Wesnoth
20090222 22:15:02< Shadow_Master> grzywacz: it is already in that dir for IftU, why would I break the rules now? ;-)
20090222 22:15:13< grzywacz> Shadow_Master, just saying... :)
20090222 22:15:33 * cib0 is looking forwards to hearing the sound of animals when hovering over forests, the sound of flowing water when hovering over rivers, the sound of blowing wind when hovering over large patches of mountains.
20090222 22:15:38< cib0> that gotta be so cool =)
20090222 22:17:04< cib0> why does Elvish Fighter advance from 15 units with the type elvish civilian?
20090222 22:17:06< CIA-70> shadowmaster * r33028 /trunk/data/core/sounds/ambient/campfire.ogg: 
20090222 22:17:06< CIA-70> New campfire sound file for use with [sound_source]s, from
20090222 22:17:06< CIA-70> wesnoth-umc-dev trunk/Invasion_from_the_Unknown/sounds/ambient .
20090222 22:17:27-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has quit []
20090222 22:17:47< cib0> yay =)
20090222 22:18:14< grzywacz> cib0, it would be cool if you felt like providing more samples... ;)
20090222 22:18:53< cib0> grzywacz, i don't even know how to create sounds =(
20090222 22:19:16< CIA-70> ivanovic * r33029 /trunk/utils/wesnoth-optipng: reduce output when no change happend
20090222 22:19:27< CIA-70> shadowmaster * r33030 /trunk/changelog: Update changelog.
20090222 22:19:37< Shadow_Master> grzywacz: the campfire sound source is courtesy of Blueblaze actually
20090222 22:20:12< grzywacz> Shadow_Master, cool. I have to re-create my lost lava sound. :(
20090222 22:22:38< CIA-70> ivanovic * r33031 /trunk/utils/wesnoth-optipng: add some output to mark in which part of the script we currently are
20090222 22:23:59-!- Baufo [n=thomas@62-47-145-230.adsl.highway.telekom.at] has quit [Remote closed the connection]
20090222 22:24:22< Ivanovic> Shadow_Master: now i think the wesnoth-optipng script is a *lot* more useable
20090222 22:24:28< Shadow_Master> k
20090222 22:24:43< Ivanovic> though the overhead is >100 MB now since the temp files are kept until stats are gathered
20090222 22:25:16< Ivanovic> but the speedup due to being able to parallize things is gigantic (if you got enough cores to be used)
20090222 22:25:16< Shadow_Master> and I was going to work wesnoth-optipng for super-optipng
20090222 22:25:33< Shadow_Master> s/work/fork/
20090222 22:25:45< Ivanovic> ???
20090222 22:25:56< Shadow_Master> I was going to experiment with Perl threading
20090222 22:26:05< Ivanovic> ah, okay
20090222 22:26:13< Ivanovic> as you see it is easy enough in plain shell
20090222 22:26:30< Shadow_Master> I wasn't going to do it in mainline anyway, but in wesnoth-umc-dev's maintenance workshop
20090222 22:27:08< boucman> Shadow_Master: threading is pretty easy
20090222 22:28:31< Ivanovic> rev33025 shows how simple it can be
20090222 22:28:33< Ivanovic> ;)
20090222 22:31:13-!- Mordante [n=chatzill@roadie.xs4all.nl] has joined #wesnoth-dev
20090222 22:31:21< ilor> hi Mordante
20090222 22:31:25< Mordante> hi
20090222 22:31:35< Mordante> hi ilor 
20090222 22:36:47< Shadow_Master> hi Mordante 
20090222 22:39:37-!- martyfuhry [i=marty@kddb1.kddb.cs.kent.edu] has joined #wesnoth-dev
20090222 22:42:37< Mordante> hi Shadow_Master 
20090222 22:44:51< Mordante> silene I also have no problem with switching to boost 1.34 post 1.6
20090222 22:46:17 * Shadow_Master neither
20090222 22:47:15-!- ABCD [n=ABCD@wikipedia/ABCD] has quit [Read error: 110 (Connection timed out)]
20090222 22:59:36< cib0> i have been owned by AI.. twice.. i suck =(
20090222 23:01:58-!- zookeeper [n=l@dsl-tkubrasgw1-fe9ff800-215.dhcp.inet.fi] has quit []
20090222 23:02:46-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090222 23:03:33-!- malumalu [n=malu@hnvr-4dbbf85e.pool.einsundeins.de] has quit ["Verlassend"]
20090222 23:05:16< Mordante> cib0 you start to become a real developer sucking at playing wesnoth is the first step :-P
20090222 23:06:25< CIA-70> ivanovic * r33032 /trunk/data/ (118 files in 15 dirs): 
20090222 23:06:25< CIA-70> ran (the new version of) utils/wesnoth-optipng
20090222 23:06:25< CIA-70> Overall statistics (only for files with a smaller recompressed size):
20090222 23:06:25< CIA-70> Original size: 2419 KiB on 118 files
20090222 23:06:25< CIA-70> Optimized size: 2256 KiB
20090222 23:06:26< CIA-70> Total saving: 162 KiB = 6% decrease
20090222 23:06:34< cib0> Mordante, i think i haven't laughed this happily in a long time =)
20090222 23:06:46 * Shadow_Master grabs r33032
20090222 23:07:20< Mordante> cib0 you start to progress, you now lay at huge commits ;-)
20090222 23:07:26< Mordante> laugh*
20090222 23:07:49< Ivanovic> hey, this were just 2.2MB, it is not *this* large for my normal dimensions...
20090222 23:08:19< Mordante> the number of files is huge ;-)
20090222 23:11:08 * cib0 shuts up now.
20090222 23:11:49< Shadow_Master> it didn't take very long to fetch it anyway ;-)
20090222 23:12:22< Mordante> Shadow_Master you're used to it and think it didn't take long ;-)
20090222 23:13:18< Ivanovic> the most important part: creating the compressed files only took a reasonable time and not >1h on my rather powerfull box
20090222 23:13:28< Shadow_Master> cib0: the key is: :debug :unit hitpoints=10000
20090222 23:13:58< Ivanovic> with a big enough value a DA could survive a berserk
20090222 23:14:06< Ivanovic> for this case we really need the DA to laugh...
20090222 23:14:14< boucman> hmm
20090222 23:14:19< Ivanovic> ;)
20090222 23:14:26< Espreon> Ivanovic: Yeah... I ran the new version of wesnoth-optipng, and I keep getting instances of these: jobs: 201: Illegal option -r
20090222 23:14:31< Shadow_Master> I think I can provide a DA laugh "niahaniahaniaha"
20090222 23:14:40< Ivanovic> Espreon: interesting
20090222 23:14:42< boucman> Ivanovic: you probably just figured out a case where there is no support in the animation engine
20090222 23:14:47< Ivanovic> what was your command for calling it, Espreon?
20090222 23:15:08< Espreon> Um... wesnoth-optipng
20090222 23:15:20< Ivanovic> run from where?
20090222 23:15:22< Shadow_Master> it works for me.
20090222 23:15:38< Shadow_Master> though I cannot use 6 threads like Ivanovic, I guess
20090222 23:15:46< Ivanovic> that is: in which folder were you, what was your exact command used and what was the error message?
20090222 23:15:51< Ivanovic> that is: retry with --threads 1
20090222 23:16:08< Espreon> Hold on and such.
20090222 23:16:16< Shadow_Master> wesnoth-optipng --threads 2 is what I used
20090222 23:16:22< Ivanovic> Shadow_Master: even with 6 threads i end at "only" approximate usage of 75% of the cpu cycles...
20090222 23:17:03< Ivanovic> so the half second wait time is even too much on my box already...
20090222 23:17:03< Shadow_Master> Ivanovic: um, it got stuck at a single file.
20090222 23:17:05< Ivanovic> ;)
20090222 23:17:31< Ivanovic> Shadow_Master: it should not get stuck, but there are in general some files that are bigger and thus need more time
20090222 23:17:31< Shadow_Master> optipng uses 100% CPU at nice 19 and it isn't going to come back alive
20090222 23:17:35< Ivanovic> eg all the big portraits
20090222 23:17:48< Ivanovic> what is the current file?
20090222 23:17:57< Shadow_Master> a portrait :-P
20090222 23:18:10< Shadow_Master> it came back after a lot of time. That demonstrates threading optipng isn't very useful here
20090222 23:18:10< Ivanovic> like i said, those need a "little" longer
20090222 23:18:36< Ivanovic> it is damn usefull here since this way i can do several of those at the same time saving some waiting
20090222 23:18:56< Espreon> Ivanovic: I have a symlink in /usr/bin/ that points to it. The actual script lies in ~/wesnoth/utils/
20090222 23:19:14< Espreon> And I tried tunning it with that switch and this problem still occurs.
20090222 23:19:15< Ivanovic> Espreon: and before my changes it was working nicely?
20090222 23:19:22< Espreon> Yup.
20090222 23:19:30< Ivanovic> *strange*
20090222 23:19:32< Shadow_Master> Espreon: a symlink from /usr/bin to your home dir!?
20090222 23:19:33< Shadow_Master> are you MAD?
20090222 23:19:51< Espreon> YES!
20090222 23:19:54< Shadow_Master> adjust your .profile to allow ${HOME}/bin as part of your PATH and create the symlink there instead...
20090222 23:20:14-!- kitty_ [n=kitty@e180214086.adsl.alicedsl.de] has quit ["good night"]
20090222 23:21:15< Ivanovic> Espreon: here it works perfectly when i got a symlink to the script somewhere in my path
20090222 23:21:24< Mordante> how much memory does optigen eat?
20090222 23:21:29< Shadow_Master> in other news, I have discovered that most of the overhead here when linking wesnoth-debug is at the HDD rather than the RAM or CPU
20090222 23:21:42< Ivanovic> Mordante: not much at all
20090222 23:21:44< Ivanovic> some MB
20090222 23:21:47< Mordante> at a lowmem dual core it was faster to use one core instead of two and swapping
20090222 23:21:48< wesbot> Mordante: Sometimes we are fast
20090222 23:22:02< Shadow_Master> wesbot: I assume we are slow most of the time ? ;-)
20090222 23:22:12< Mordante> then memory is not a problem
20090222 23:22:15< loonycyborg> Shadow_Master: If you had gkrellm open, you'd learn it much earlier :)
20090222 23:22:25< Mordante> Shadow_Master we got our CVEs fast
20090222 23:22:59< Ivanovic> Espreon: okay, i think i got it
20090222 23:23:08< Ivanovic> you are using some different version of "jobs"
20090222 23:28:58-!- noy [n=Noy@70.70.128.133] has joined #wesnoth-dev
20090222 23:31:34< Ivanovic> Espreon, Shadow_Master: okay, removed the -r option from jobs (should somehow work anyway...) and dropped the default to only one thread
20090222 23:31:54< Shadow_Master> k
20090222 23:31:57< Shadow_Master> brb
20090222 23:32:25-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has quit ["in ~ 3 minuets"]
20090222 23:33:42-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has quit ["leaving"]
20090222 23:34:00< boucman> night all
20090222 23:34:08-!- boucman [n=rosen@2.117.71-86.rev.gaoland.net] has quit [Remote closed the connection]
20090222 23:34:38-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev
20090222 23:40:48-!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has quit [Remote closed the connection]
20090222 23:46:20-!- Espreon [n=espreon@c-76-117-239-33.hsd1.pa.comcast.net] has joined #wesnoth-dev
20090222 23:47:05-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has quit ["be right back"]
20090222 23:48:30-!- Shadow_Master [n=imorelle@unaffiliated/shadowmaster/x-275924] has joined #wesnoth-dev
20090222 23:53:23< Ivanovic> time for me to head off to bed
20090222 23:53:25< Ivanovic> n8
20090222 23:53:34< Espreon> Good night.
20090222 23:53:51< Ivanovic> Espreon: so is the latest version of wesnoth-optipng working for you?
20090222 23:54:04-!- alink [n=alink@78.129.18.215] has joined #wesnoth-dev
20090222 23:54:13< Mordante> night Ivanovic 
20090222 23:54:17< alink> good evening
20090222 23:54:29< Mordante> hi Al
20090222 23:54:32< Mordante> alink *
20090222 23:54:53< Shadow_Master> hi alink , nice optimizations in the GUI2 stuff
20090222 23:54:54< alink> 'lo MMoorrddaannttee
20090222 23:54:58< Shadow_Master> it seems to help a bit here.
20090222 23:55:00< Shadow_Master> er?
20090222 23:55:38< alink> Shadow_Master: was a small thing , but good to know that it helped a bit
20090222 23:55:41< Mordante> alink did you test with the 1 ms delay before push event? if so did it help?
20090222 23:56:11< Espreon> Ivanovic: It seems so.
20090222 23:56:32< alink> Mordante: not yet  tested that but i tested the draw_timer(0,0) which seems to works and help
20090222 23:57:06< CIA-70> ivanovic * r33034 /trunk/utils/wesnoth-optipng: make waiting a little more aggressive
20090222 23:57:14< CIA-70> ivanovic * r33033 /trunk/utils/wesnoth-optipng: 
20090222 23:57:14< CIA-70> remove '-r' option from the 'jobs' call since at least Espereon has problems with it
20090222 23:57:14< CIA-70> reduce default number of threads to use to just one
20090222 23:57:21< Mordante> ok then I think I just test it tomorrow and commit it afterwards
20090222 23:57:39< alink> Mordante: btw for this delay the gain for the worst case scenario is 40ms, because the timer will often fire during the SDL_delay(10)
20090222 23:57:50< Mordante> I know ;-)
20090222 23:58:28< Mordante> but before I had some problems and there always seemed to be an event which caused the cpu run at 100% usage
20090222 23:59:05< Mordante> but I haven't tested much whether the issue was solved (like I wrote in the log of yesterday)
20090222 23:59:27< Mordante> but I think it's a bit too tricky to do it before 1.6
20090222 23:59:52< Mordante> I think it should work, but rather want to avoid new features if not needed
--- Log closed Mon Feb 23 00:00:00 2009