--- Log opened Tue Oct 20 00:00:40 2009
20091020 00:04:36-!- Sapient [n=patrickp@wesnoth/developer/sapient] has joined #wesnoth-dev
20091020 00:04:55< Sapient> Soliton: thou hast summoned me?
20091020 00:04:58 * Soliton hails Sapient.
20091020 00:05:16< Soliton> Sapient: do you know how unstore_unit works in replays?
20091020 00:05:47< Soliton> Sapient: it seems to me nothing is stored so it is executed the same as while playing?
20091020 00:06:02< Sapient> yes, the replay event such as "moveto" firing will trigger a WML event handler
20091020 00:06:16< Sapient> then it executes the WML [event] which contains [unstore_unit] etc
20091020 00:07:12< Soliton> do you know why [advance_unit] (and subsequent [choose]) is saved though?
20091020 00:07:28< Soliton> (when advance=yes)
20091020 00:07:44< Sapient> right, because advancing often involves a choice
20091020 00:07:54< Sapient> the same with messages containing options
20091020 00:08:10< Sapient> it has to be saved explicitly in the replay
20091020 00:08:20< Soliton> right, i wasn't so much asking about the [choose], it's clear that that has to be saved. :-)
20091020 00:08:32< Soliton> but that chould be enough, no?
20091020 00:08:36< Soliton> should*
20091020 00:09:08< Sapient> I think unit advancement is handled differently than other things for some particular reason I don't recall exctly
20091020 00:09:26< Sapient> details are in replay.cpp but I'd have to look over it again
20091020 00:10:15< Sapient> YogiHH is the replay master, though ;)
20091020 00:10:29-!- Sapient [n=patrickp@wesnoth/developer/sapient] has left #wesnoth-dev []
20091020 00:13:02-!- elynia [n=shyde@wesnoth/umc-dev/misc/elynia] has joined #wesnoth-dev
20091020 00:14:23-!- Arc [n=arc@pysoy/developer/ArcRiley] has quit [Read error: 110 (Connection timed out)]
20091020 00:16:42< Soliton> ok, i see now that the advance handling is disabled in replays.
20091020 00:18:33< Soliton> that seems to just ask for trouble.
20091020 00:20:26-!- elynia [n=shyde@wesnoth/umc-dev/misc/elynia] has quit ["nyu"]
20091020 00:21:02< CIA-45> alink * r39363 /trunk/src/ (actions.cpp menu_events.cpp unit.cpp): 
20091020 00:21:02< CIA-45> Fix bug #14540 "debug-generated units with HP-modifying traits don't have full HP".
20091020 00:21:02< CIA-45> Probably also fix an old similar bug with unit created by "plague".
20091020 00:21:02< CIA-45> Unify these creation calls and give full MP to unit spawned by ":create".
20091020 00:27:11-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20091020 00:29:55-!- SonIcco [n=SonIcco@pD95132D4.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091020 00:35:13-!- shikadibot [n=shikadi@wesnoth/umc-dev/bot/shikadibot] has quit ["manual override, desu!"]
20091020 00:35:37-!- shikadibot [n=shikadi@wesnoth/umc-dev/bot/shikadibot] has joined #wesnoth-dev
20091020 00:48:49-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20091020 00:55:15-!- mordocai [n=mordocai@66.119.9.243] has joined #wesnoth-dev
20091020 00:56:33-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit ["WRYYYYYYYYYYYYYYYYYYYY!"]
20091020 01:03:28< mordocai> Okay, this is way off-topic but #math seems to be ignoring me... I took a college algebra test today and completely blanked on how to solve 2^(3x + 1) = 3^(x - 2). I looked in my book and they had how to do problems like 3^(3x + 1) = 5... just not with x on both sides. Can someone help?
20091020 01:04:26< alink> uses logarithm probably
20091020 01:04:36< mordocai> Yeah, i know that much... 
20091020 01:04:42< mordocai> I just can't figure out how... lol
20091020 01:05:01< mordocai> I'm just being dumb basically... i'll kick myself once i figure it out
20091020 01:05:08< alink> let me find some old paper and pen :)
20091020 01:08:31< mordocai> I mean... i THINK i'm supposed to end up at some point with (3x + 1)log2 = (x - 2)log3... but then i'm stuck...
20091020 01:09:55< alink> i have x =-2.94686537
20091020 01:10:12< mordocai> Yeah, that's the answer in the book... can you explain how you got it?
20091020 01:10:28< alink> that is x = (-2*log3 - log2) / (3*log2-log3)
20091020 01:10:50< alink> do you know properties of log ?
20091020 01:10:51< loonycyborg> mordocai: That is correct. Though I have no idea what your problem is. log x are just numbers like 2 or 3
20091020 01:11:37< loonycyborg> So (3x + 1)log2 is the same as 3log2 x + log2
20091020 01:11:58< mordocai> Oh... duh.. distribute...
20091020 01:12:01< alink> i just use log on 2 members of the equality and use the formula of log for exponenent 
20091020 01:12:22< alink> ah yes and distribution law, obviously :-)
20091020 01:13:37< mordocai> I think i just haven't been sleeping enough... lol.
20091020 01:15:45-!- ilor_ [n=ilor@wesnoth/developer/ilor] has quit [Read error: 113 (No route to host)]
20091020 01:19:54-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20091020 01:20:19-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20091020 01:21:15-!- wesbot changed the topic of #wesnoth-dev to: 1.7.7 planned for Wednesday | string/feature freeze active! | 96 bugs, 238 feature requests, 10 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091020 01:23:17< CIA-45> alink * r39364 /trunk/src/pathfind.cpp: Optimize "paths" pathfinding function
20091020 01:25:30-!- SonIcco [n=SonIcco@pD95132D4.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091020 01:34:29-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091020 02:01:29< shadowmaster> alink: I'm sure I had something to tell you some days ago :/
20091020 02:21:31< alink> shadowmaster: ok, noted :)
20091020 02:23:26-!- Appleman1234 [n=Appleman@131.181.102.177] has joined #wesnoth-dev
20091020 02:27:05 * shadowmaster has just remembered but he now has to recompile to check if it still applies.
20091020 02:27:36< alink> maybe i know if still applies
20091020 02:30:02-!- mordocai [n=mordocai@66.119.9.243] has quit ["Leaving"]
20091020 02:32:57< shadowmaster> the Level column colors used in the Recall dialog don't match the sidebar/unit preview thingy
20091020 02:33:03< shadowmaster> s/Level/XP/
20091020 02:33:31< shadowmaster> that was as of 39350
20091020 02:33:40< alink> checking..
20091020 02:36:03< alink> XP colors matches but not level, is it what you mean?
20091020 02:36:48-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091020 02:38:02-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091020 02:38:37< alink> mmh level will be a problem to change there, because the old markup will color the whole line with the "level" word which is probably not wanted
20091020 02:38:45< shadowmaster> no, XP.
20091020 02:39:01< shadowmaster> s/foo/bar/ is a sed-like replacement operation.
20091020 02:39:23-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit [Client Quit]
20091020 02:39:30< alink> shadowmaster: seems to be the same color, are you sure that it's not because of the different background ?
20091020 02:40:47< alink> yes i know about s//, but i see the oppesite here (level has no color, and XP has same colors)
20091020 02:41:11< alink> both use font::color2markup(u.xp_color()) 
20091020 02:41:15< shadowmaster> or maybe I have to check against HEAD. THat's a 3-days old build
20091020 02:42:40< alink> color was introduced in XP column in r39204
20091020 02:43:04-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091020 02:44:52< alink> shadowmaster: just to be sure, you see some blue-ish colors in both unit preview and XP column ? and you mean it's not exactly the same?
20091020 02:46:26< shadowmaster> they are the same.
20091020 02:46:33< shadowmaster> I tink my memory is just horribly messed up.
20091020 02:47:43< alink> at least you remembered what you wanted to say ;)
20091020 03:03:06-!- elynia [n=shyde@wesnoth/umc-dev/misc/elynia] has joined #wesnoth-dev
20091020 03:10:00< CIA-45> alink * r39365 /trunk/src/dialogs.cpp: Fix broken indentation for race in unit-preview
20091020 03:14:45-!- ardesh [n=ardesh@port-92-206-57-145.dynamic.qsc.de] has quit [Read error: 113 (No route to host)]
20091020 03:15:28-!- ardesh [n=ardesh@port-92-206-62-226.dynamic.qsc.de] has joined #wesnoth-dev
20091020 03:22:33< shadowmaster> shikadibot: dir /branches/1.0
20091020 03:22:33< shikadibot> Web interface URL to directory /branches/1.0: http://svn.gna.org/viewcvs/wesnoth/branches/1.0
20091020 03:25:47-!- Blueblaze [n=nick@99.171.161.30] has joined #wesnoth-dev
20091020 03:30:55< shadowmaster> shikadibot: dir /tags
20091020 03:30:55< shikadibot> Web interface URL to directory /tags: http://svn.gna.org/viewcvs/wesnoth/tags
20091020 03:34:13< Aethaeryn> Ivanovic: Any estimation as to the release date of 1.8? I need a somewhat-certain deadline because I was intending to do some wiki cleanup, especially with regards to the campaign documentation.
20091020 03:36:22< shadowmaster> The following words in your search query were ignored because they are too common words: wesnoth 1.
20091020 03:36:35< shadowmaster> stupid forum search 
20091020 03:37:03< shadowmaster> Aethaeryn: when the bug count is below 90.
20091020 03:37:19< shadowmaster> oh, you mean the *definitive* release? :P
20091020 03:39:14< Aethaeryn> shadowmaster: I mean, people work best with deadlines. It'd be great if I know when I need to do the wiki thing, or if I have time to do it at all.
20091020 03:41:03< alink> "people work best with deadlines" [citation needed]
20091020 03:41:28< shadowmaster> it has happened to e at least.
20091020 03:42:02< shadowmaster> I set a deadline for a release, I slacked for months, and spent the last week skipping sleep to accomplish the job.
20091020 03:42:35< alink> ok maybe people work more, but not sure if the work itself is better
20091020 03:42:36< shadowmaster> do not try that at home if you are proud of your face, though.
20091020 03:42:44< shadowmaster> *face's looks
20091020 03:43:24< alink> The "it will be ready when it will be ready" thing
20091020 03:50:30-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091020 03:54:36-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091020 04:06:15-!- Doppp [n=Doppp@unaffiliated/doppp] has quit [Connection timed out]
20091020 04:07:32-!- elynia [n=shyde@wesnoth/umc-dev/misc/elynia] has quit ["nyu"]
20091020 04:20:58-!- Kenpachi [n=chatzill@CPE-58-170-83-81.sa.bigpond.net.au] has quit ["ChatZilla 0.9.85 [Firefox 3.5.3/20090824101458]"]
20091020 04:30:14-!- Max2008 [n=chatzill@188.23.4.124] has joined #wesnoth-dev
20091020 04:35:57-!- Ivanovic_ [n=ivanovic@77.178.169.243] has joined #wesnoth-dev
20091020 04:37:22-!- Max2008_ [n=chatzill@188-23-11-105.adsl.highway.telekom.at] has quit [Read error: 104 (Connection reset by peer)]
20091020 04:44:56-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 145 (Connection timed out)]
20091020 04:45:55-!- Ivanovic_ is now known as Ivanovic
20091020 05:03:43-!- Blueblaze [n=nick@99.171.161.30] has quit [Remote closed the connection]
20091020 05:20:33< Zarel> I have a citation for "people work best with deadlines"!
20091020 05:21:23< Zarel> http://www.amazon.com/dp/0061854549/?tag=codinghorror-20
20091020 05:39:39< noy> congratulations
20091020 05:41:51< noy> is a work of your cited or did you help someone who wrote it?
20091020 05:59:47-!- Sirp [n=user@wesnoth/developer/dave] has quit [Read error: 113 (No route to host)]
20091020 06:07:54-!- Desciero [n=chatzill@cpe-024-074-122-087.carolina.res.rr.com] has quit ["Chatzilla 0.9.75.1 [SeaMonkey 1.1.18/2009082513]"]
20091020 06:16:37-!- rtfb [n=read-the@78-56-0-35.static.zebra.lt] has quit [Read error: 110 (Connection timed out)]
20091020 06:22:16< Aethaeryn> The problem with "it will be ready when it will be ready" is that if I don't even know a *general* date, it could be in a week or it could be in a month and a half, and I won't know when to stop my add-on work, prepare for 1.8 releases, and 
20091020 06:22:50< Aethaeryn> (1.8 releases of the add-ons I maintain... I hit enter instead of backspace, a sign of tiredness.)
20091020 06:22:57< Aethaeryn> and when to start the wiki work.
20091020 06:23:26< Aethaeryn> No one can even say "sometime in November"?
20091020 06:25:07-!- [Relic] [n=[Relic]@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit ["Leaving"]
20091020 06:25:15< Aethaeryn> Also, I am operating on the assumption that the 1.8 add-on server will be started sometime before the release of 1.8 (probably around the release candidate phase), thus making me prioritize add-ons over wiki work...
20091020 06:32:43-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20091020 06:41:13-!- Aethaeryn [n=Michael@wesnoth/umc-dev/developer/aethaeryn] has quit ["meh."]
20091020 06:56:11-!- Doppp [i=Doppp@CMU-330744.WV.CC.CMU.EDU] has joined #wesnoth-dev
20091020 07:24:00< CIA-45> silene * r39366 /trunk/src/pathfind.cpp: 
20091020 07:24:00< CIA-45> Revert "Optimize "paths" pathfinding function" (revision 39364).
20091020 07:24:00< CIA-45> Visited nodes are either on the closed or the open list. Nodes on the open list are not optimal; they have to keep being modified until they switch to the closed list.
20091020 07:45:08-!- rtfb [n=read-the@78-56-0-35.static.zebra.lt] has joined #wesnoth-dev
20091020 08:04:11-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091020 08:32:32-!- Appleman1234 [n=Appleman@131.181.102.177] has quit [Remote closed the connection]
20091020 08:42:22< CIA-45> jetryl * r39367 /trunk/data/core/images/units/undead-skeletal/ (13 files): Updated shadows and bow attack for the bone shooter.
20091020 08:58:25-!- Appleman1234 [n=Appleman@131.181.102.177] has joined #wesnoth-dev
20091020 09:03:28-!- eradicus [n=eradicus@unaffiliated/eradicus] has joined #wesnoth-dev
20091020 09:32:05-!- eradicus [n=eradicus@unaffiliated/eradicus] has quit ["Leaving"]
20091020 09:44:15-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20091020 09:48:55-!- stikonas [n=and@bcm-131-111-216-119.girton.cam.ac.uk] has joined #wesnoth-dev
20091020 10:20:03-!- mjs-de [n=mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20091020 10:57:34-!- loonybot [n=loonybot@79.139.137.59] has joined #wesnoth-dev
20091020 10:58:22-!- loonycyborg [n=sergey@79.139.137.59] has joined #wesnoth-dev
20091020 11:02:39-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091020 11:05:47< Ivanovic> moin
20091020 11:06:55< Ivanovic> Aethaeryn: there *are* no real estimates
20091020 11:07:13< Ivanovic> Aethaeryn: it won't be out till the bug list is significantly smaller
20091020 11:07:17< Ivanovic> that is *all* i can say
20091020 11:18:20-!- Crab_ [i=crab@wesnoth/developer/crab] has joined #wesnoth-dev
20091020 11:30:41-!- Crab_ [i=crab@wesnoth/developer/crab] has quit ["Leaving."]
20091020 12:20:46-!- ilor_ [n=ilor@wesnoth/developer/ilor] has joined #wesnoth-dev
20091020 12:27:37-!- Doppp [i=Doppp@unaffiliated/doppp] has quit [Read error: 110 (Connection timed out)]
20091020 12:30:05-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has joined #wesnoth-dev
20091020 12:30:29-!- Max2008_ [n=chatzill@188-23-99-127.adsl.highway.telekom.at] has joined #wesnoth-dev
20091020 12:38:21-!- Max2008 [n=chatzill@188.23.4.124] has quit [Read error: 145 (Connection timed out)]
20091020 12:38:59-!- Doppp [n=Doppp@c-67-171-96-240.hsd1.pa.comcast.net] has joined #wesnoth-dev
20091020 12:54:36-!- Appleman1234 [n=Appleman@131.181.102.177] has quit [Remote closed the connection]
20091020 13:09:44-!- Nekhro [i=kvirc@ip59936c2c.rubicom.hu] has joined #wesnoth-dev
20091020 13:09:53-!- Nekhro [i=kvirc@ip59936c2c.rubicom.hu] has left #wesnoth-dev []
20091020 13:49:20-!- grzywacz [n=grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20091020 13:55:37< grzywacz> mordante: and so SDL 1.2.14 has been released ;)
20091020 14:22:58< stikonas> grzywacz: do you know if SDL 1.3 will have support for Multi Pointer X and XInput 2?
20091020 14:37:13< grzywacz> stikonas, I know they are talking about it :)
20091020 14:41:19-!- SonIcco [n=SonIcco@217.81.43.53] has joined #wesnoth-dev
20091020 14:44:51-!- ilor_ [n=ilor@wesnoth/developer/ilor] has quit [Read error: 113 (No route to host)]
20091020 15:09:29-!- Arc [n=arc@pysoy/developer/ArcRiley] has joined #wesnoth-dev
20091020 15:45:49-!- ilor_ [n=ilor@wesnoth/developer/ilor] has joined #wesnoth-dev
20091020 15:48:47-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091020 15:49:17-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit [Client Quit]
20091020 16:16:23-!- stikonas [n=and@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20091020 16:18:05-!- Sirp [n=user@wesnoth/developer/dave] has joined #wesnoth-dev
20091020 16:18:53-!- stikonas [n=and@131.111.216.119] has joined #wesnoth-dev
20091020 16:24:16-!- knotwork_ [n=markm@hlfxns01bbg-142177232207.pppoe-dynamic.ns.aliant.net] has quit [Read error: 104 (Connection reset by peer)]
20091020 16:24:19-!- knotwork__ [n=markm@hlfxns01bbg-142177232207.pppoe-dynamic.ns.aliant.net] has joined #wesnoth-dev
20091020 16:49:29-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20091020 17:09:33-!- blarumyrran [n=minaise@81-20-159-197.levira.ee] has joined #wesnoth-dev
20091020 17:16:46-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has quit ["Saurian Augur - I'll heal you by 4 hp if you leave next to me"]
20091020 17:23:15-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091020 17:23:54-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit [Client Quit]
20091020 17:28:09-!- SonIcco [n=SonIcco@217.81.43.53] has quit [Read error: 131 (Connection reset by peer)]
20091020 17:28:10< AI0867> fendrin: what's the idea behind the team_color_overrides in LoW? Because right now, only uradredia gets white and the gold carrier from scenario 5 blinks from white to purple a few times (before staying purple)
20091020 17:48:11-!- noy [n=Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20091020 17:49:36< Soliton> how do i write 'if ((snapshot) && (const config &vars = snapshot.child("variables")))' so the compiler likes it?
20091020 17:50:18< Soliton> it's complaining about "error: expected primary-expression before ‘const’".
20091020 17:51:24-!- allefant [n=elias@allegro/developer/allefant] has joined #wesnoth-dev
20091020 17:52:01< Sirp> Soliton: how about const config* vars = snapshot ? &snapshot.child("variables") : NULL; if(vars) { ... }
20091020 17:54:33< Soliton> yeah, that'd work but i wonder what is wrong with the above? it works when i write it as a nested if.
20091020 17:55:11< Soliton> does the compiler not understand to do a boolean conversion for snapshot?
20091020 17:56:45< loonycyborg> const config &vars = .. is a declaration, not an expression.
20091020 17:56:51< Sirp> Soliton: you can't initialize a variable in the middle of an expression.
20091020 17:57:17< Sirp> Soliton: there is a special case for if() where it can take a declaration and treat it as an expression but it's not a general thing.
20091020 17:57:31< ilor_> anyone, do we have some figures on what load the addon server has, say, daily?
20091020 17:57:35< Soliton> Sirp: ah, i see.
20091020 17:57:59< Soliton> ilor_: what kind of load?
20091020 17:58:07< Soliton> number of connections?
20091020 17:58:07< Sirp> ilor_: well we'd be able to look at number of downloads
20091020 17:58:15< Soliton> bandwidth usage?
20091020 17:58:16< ilor_> Soliton: any number really, downloads per day or something
20091020 17:59:29< ilor_> a rough estimate with one significant digit will be enough for me
20091020 18:00:30< Soliton> ilor_: http://nopaste.com/p/aJu0lKrOeb
20091020 18:00:41< Soliton> number of downloads per day.
20091020 18:00:48< ilor_> thanks
20091020 18:01:38< Soliton> wait that also includes the list of addons.
20091020 18:02:02< ilor_> okay
20091020 18:03:06< Soliton> real downloads: http://nopaste.com/p/aSzEqqxlk
20091020 18:04:28< Soliton> you can also see the number of connections at collectd.wesnoth.org but you can also see there that we have a big issue with ghost connections. :-P
20091020 18:05:18-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091020 18:05:22< ilor_> Soliton: thanks that'll be enough, I just need a performance figure for the project doc and I don't want to be off by many magnitudes ;)
20091020 18:09:10< Sirp> ilor_: well after our plans for World Domination are complete you will be off by many orders of magnitude. ;)
20091020 18:14:19-!- SonIcco [n=SonIcco@pD9512B35.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091020 18:24:20-!- SonIcco [n=SonIcco@pD9512B35.dip0.t-ipconnect.de] has quit [Read error: 104 (Connection reset by peer)]
20091020 18:25:58-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20091020 18:46:15-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20091020 18:47:49-!- SonIcco [n=SonIcco@pD9512B35.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091020 18:50:15< alink> silene: hum about r39366, i still think i am right. Did you noticed that it was not the A* code? In "paths", things are simpler (no heurisitc), making this optimization possible there
20091020 18:52:27< alink> In fact, I think this code can be simplified more, we never need to update an already visited node, making the use of 2 lists a bit useless
20091020 18:57:36< alink> One possible explanation is: we always pick node with smallest cost first (and only use cost, no heuristic). So it's impossible to find a new better path to a location. The only way to be better would be if the previous location of this path has a smaller cost than the previous location of the older path
20091020 18:58:34< alink> but in that case this new path should have be picked first. So it never happen
20091020 19:03:38< alink> To be complete, this rely on some simple properities of the cost function, like ignoring direction and having more MP before entering a hex is always better (move_left will not be lower than if I had less MP)
20091020 19:04:57-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20091020 19:05:26< mordante> servus
20091020 19:05:32-!- EdB [n=edb@79.88.117.178] has joined #wesnoth-dev
20091020 19:05:34< alink> hi mordante
20091020 19:05:51< mordante> hi alink
20091020 19:06:37< mordante> thanks grzywacz 
20091020 19:07:23< alink> silene: also, that's not proof, but my empirical tests show that in_list always fails and next_visited is always false there
20091020 19:09:28< alink> mordante: btw how do you detect/clean trailing spaces? astyle?
20091020 19:09:52< mordante> ilor_, wescamp update in manually at a (ir)regular interval
20091020 19:10:31< ilor_> mordante: uh, come again?
20091020 19:10:50< mordante> alink, find . -type f \( -name '*pp' -o -name '*i' \) -exec sed  -i 's/[ \t]*$//' \{} \;
20091020 19:11:08< mordante> ilor_, you asked about the wescamp updating yesterday
20091020 19:11:24< alink> ah regex, of course, thanks
20091020 19:11:38< ilor_> mordante: I got a parse error reading your answer ;)
20091020 19:11:56< mordante> alink, and I have a script that I run (ir)regular to remove them all
20091020 19:12:08-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has joined #wesnoth-dev
20091020 19:12:14< ilor_> mordante: who runs the manual update?
20091020 19:12:40< mordante> ilor_, I do download all addons and update them at berlios
20091020 19:13:04< ilor_> berlios is where wescamp is?
20091020 19:13:08< mordante> the way back had some problems I never got around to fix
20091020 19:13:11< mordante> yes
20091020 19:13:30< ilor_> you integrate the translations or just upload addons to wescamp?
20091020 19:13:53< mordante> once I upload them at wescamp Torangan does the pot-update
20091020 19:14:17< mordante> the plan was to upload it back to the addon server and I have some scripts but these are kind of untested
20091020 19:14:27< alink> sorry for the noob question, but what is wescamp exactly ?
20091020 19:14:41< mordante> I thought it worked but seems to be commented out
20091020 19:15:07< mordante> alink, the way to coordinate translations of addons
20091020 19:15:43< ilor_> mordante: so who puts the translations in the addons?
20091020 19:16:04< alink> mordante: ah ok, thanks, I see that is not my area at all :-)
20091020 19:16:57< mordante> ilor_, the addon author
20091020 19:17:23< ilor_> mordante: so he has to download the translation files from wescamp and put them in the addon, okay
20091020 19:17:53< ilor_> I'm surprised one person does the uploads to wescamp, thought authors do that on their own too
20091020 19:19:53< mordante> I think most authors use the addon server for that nowadays
20091020 19:20:22< mordante> and the goal is that the author doesn't need to care about translations at all
20091020 19:20:51< mordante> the plan was to store them separately and also allow the user to only download language X
20091020 19:21:00< mordante> (of course by setting some preferences)
20091020 19:21:15-!- wesbot changed the topic of #wesnoth-dev to: 1.7.7 planned for Wednesday | string/feature freeze active! | 96 bugs, 238 feature requests, 11 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20091020 19:21:38< alink> silene: i ll go afk for some time. I will check this stuff again later but if you notice something obvious that i missed, feel free to drop a hint ;-)
20091020 19:21:46< alink> bbl
20091020 19:35:17-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091020 19:44:04-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091020 19:57:44< CIA-45> soliton * r39368 /trunk/src/gamestatus.cpp: fixed the copy constructor to use the initialization list and not forget next_scenario
20091020 19:57:48< CIA-45> soliton * r39369 /trunk/src/ (dialogs.cpp dialogs.hpp): comment tweaks
20091020 19:57:53< CIA-45> soliton * r39370 /trunk/src/gamestatus.cpp: moved game_classification defaults into its constructor
20091020 19:57:56< CIA-45> soliton * r39371 /trunk/src/gamestatus.cpp: fixed an assert()
20091020 19:58:00< CIA-45> soliton * r39372 /trunk/src/gamestatus.cpp: only check [snapshot] for [variables] if we're not loading a replay
20091020 19:58:04< CIA-45> soliton * r39373 /trunk/src/mp_game_settings.cpp: also check [replay_start] for [multiplayer]
20091020 20:06:32< CIA-45> soliton * r39374 /trunk/src/game.cpp: consistently return 2 on commandline argument errors
20091020 20:14:16-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit []
20091020 20:30:13-!- Max2008 [n=chatzill@188-23-13-67.adsl.highway.telekom.at] has joined #wesnoth-dev
20091020 20:32:39< mordante> Ivanovic, I wonder about https://gna.org/bugs/index.php?14506, this is bound to happen again I can draw the text unformatted in the screen, looks ugly and might cause bug report
20091020 20:32:43< mordante> reports*
20091020 20:33:15< mordante> or I can postpone it post 1.8 and add a nice warning about it (which can be translated)
20091020 20:33:28< mordante> what do you think is wise?
20091020 20:34:09< mordante> (I thought I already added protection against it and I kind of did only not at the best place...)
20091020 20:39:38< Ivanovic> re
20091020 20:40:04< Ivanovic> mordante: i'd guess that you can use a general wml error warning, right?
20091020 20:41:14< mordante> Ivanovic, no unless I start to throw exceptions, which is what I do at the moment
20091020 20:41:28< Ivanovic> not good, really not good...
20091020 20:41:31< mordante> the best place to fix it is inside the renderer
20091020 20:41:50< mordante> and that has no ideas about screens 
20091020 20:42:29< mordante> so I can keep it as is, draw the text but show the raw markup and hope nobody complains
20091020 20:44:12< Ivanovic> sounds okay
20091020 20:44:14< CIA-45> mordante * r39376 /trunk/data/ai/utils/default_config.cfg: Fix alignment.
20091020 20:44:19< CIA-45> mordante * r39375 /trunk/src/ai/default/ai.hpp: Fix alignment.
20091020 20:44:26< CIA-45> mordante * r39377 /trunk/src/gui/dialogs/wml_message.cpp: 
20091020 20:44:26< CIA-45> Directly set the markup mode for the listbox.
20091020 20:44:26< CIA-45> Recently added code to directly do it so also use it here.
20091020 20:44:35< CIA-45> mordante * r39378 /trunk/src/text.cpp: Fix the warning level.
20091020 20:45:18< mordante> Ivanovic, oops that should have been "so I can keep it as is, _or_ draw the text but show the raw markup and hope nobody complains"
20091020 20:46:00< mordante> option 2 is ugly but at least allows the player to play on instead of breaking the game
20091020 20:47:11-!- Max2008_ [n=chatzill@188-23-99-127.adsl.highway.telekom.at] has quit [Read error: 110 (Connection timed out)]
20091020 20:50:59-!- SonIcco [n=SonIcco@pD9512B35.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091020 20:52:38< AI0867> mordante: when starting wesnoth with --nogui, a std::bad_cast is thrown
20091020 20:52:50< CIA-45> ai0867 * r39379 /trunk/src/game.cpp: Catch std::exception too, so we get more information than "unknown error"
20091020 20:52:58-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has quit ["This computer has gone to sleep"]
20091020 20:53:52< mordante> AI0867, odd I throw a wml_exception
20091020 20:54:18< mordante> but the real question why is that code used with --no-gui
20091020 20:55:40< AI0867> trying to get a backtrace
20091020 20:55:59< mordante> ok
20091020 20:56:39-!- Nayela [n=Nayela@82.132.136.193] has joined #wesnoth-dev
20091020 20:57:40< AI0867> a quick grep through the code showed the throwing dozens of structs that don't inherit from any common base class though =/
20091020 20:58:22< AI0867> I added catch(std::exception& e) and got lucky
20091020 20:58:55< AI0867> it was "unknown error" before that, caught by catch(...)
20091020 20:59:11-!- SonIcco [n=SonIcco@pD9512B35.dip0.t-ipconnect.de] has joined #wesnoth-dev
20091020 20:59:56< AI0867> #7  0xb79a5845 in __cxa_bad_cast () from /usr/lib/libstdc++.so.6
20091020 20:59:56< AI0867> #8  0x08642565 in theme::modify_label ()
20091020 21:00:41< Soliton> is there any reason to have the catch(...)?
20091020 21:01:06< AI0867> not sure
20091020 21:01:34< mordante> AI0867, it's no luck bad_cast is std::bad_cast, which derives from std::exception
20091020 21:01:52< mordante> but having debug info would be helpful
20091020 21:01:55< AI0867> mordante: yes, but whatever catch(...) was catching could've been anything
20091020 21:02:02< AI0867> I'm compiling a debug build now
20091020 21:02:43< mordante> well theme.cpp:803
20091020 21:02:46< mordante> throws it
20091020 21:03:15< mordante> yup catch(...) catches everything
20091020 21:05:46< mordante> I'm off bye
20091020 21:06:00-!- lizard_r [n=Miranda@wesnoth/umc-dev/developer/lizard] has quit ["Saurian Augur - I'll heal you by 4 hp if you leave next to me"]
20091020 21:06:52-!- mordante [n=mordante@wesnoth/developer/mordante] has quit ["Leaving"]
20091020 21:10:30< AI0867> r39283 broke things
20091020 21:14:39-!- noy [n=noy@24.114.237.33] has joined #wesnoth-dev
20091020 21:15:53-!- Zarel [n=Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20091020 21:23:50< CIA-45> soliton * r39380 /trunk/src/server/server.cpp: return 2 on commandline argument errors like wesnoth
20091020 21:23:59< CIA-45> soliton * r39381 /trunk/src/game.cpp: consistently return 1 on initialization errors
20091020 21:24:03< CIA-45> soliton * r39382 /trunk/doc/man/ (wesnoth.6 wesnothd.6): documented the exit status conventions
20091020 21:24:07< CIA-45> soliton * r39383 /trunk/src/ (campaign_server/campaign_server.cpp server/server.cpp): use common exit codes for signal caused exits
20091020 21:24:18-!- noy [n=noy@24.114.237.33] has quit []
20091020 21:24:37< CIA-45> ivanovic * r39384 /trunk/ (14 files in 14 dirs): updated Russian translation
20091020 21:25:58< Soliton> oh, was the man page addition a string freeze violation?
20091020 21:26:53< Ivanovic> jupp
20091020 21:27:34< Soliton> i guess you want it reverted then? :-)
20091020 21:27:36-!- SonIcco [n=SonIcco@pD9512B35.dip0.t-ipconnect.de] has quit [Remote closed the connection]
20091020 21:27:44< Ivanovic> jupp
20091020 21:29:49-!- shadowmaster_ [n=ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20091020 21:31:00< CIA-45> soliton * r39385 /trunk/doc/man/ (wesnoth.6 wesnothd.6): 
20091020 21:31:00< CIA-45> Revert "documented the exit status conventions"
20091020 21:31:00< CIA-45> string freeze...
20091020 21:31:22< Soliton> i suppose i should figure out how to have a local 1.9 branch.
20091020 21:31:37< shadowmaster> I don't recall we were this strict with the string freezes last time?
20091020 21:31:53< Soliton> pretty sure we were.
20091020 21:32:13< shadowmaster> I remember I still got newfstrings to translate before the first RC
20091020 21:32:30-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20091020 21:32:45< Soliton> it gets lifted once at least before release, i think.
20091020 21:33:07< Soliton> or maybe that was just for typo fixes and we don't need that anymore now.
20091020 21:33:14< Ivanovic> Soliton: ehm, those fixes really required get in directly
20091020 21:33:23< Ivanovic> the typo fixes are done via pofix.py anyway
20091020 21:33:32< Soliton> exactly my point.
20091020 21:53:09< shadowmaster> zookeeper, AI0867 : please do not move the second spam post from a single user to the hidden forum, the first is enough, the rest can be deleted on sight.
20091020 21:53:21< shadowmaster> it makes it more annoying for me to have to go to the forum MCP.
20091020 21:57:30-!- Nayela [n=Nayela@82.132.136.193] has quit ["Leaving"]
20091020 22:00:59-!- silene [n=plouf@ASte-Genev-Bois-152-1-69-248.w86-198.abo.wanadoo.fr] has joined #wesnoth-dev
20091020 22:01:06< silene> hi
20091020 22:05:15< silene> alink: i agree that, due to the cost being the same on all the incoming edges, the second test can be skipped; but i disagree with actually removing it from the code; first, the gain will be negligible as it is predicted by the processor; second, if we decide to change the cost function (for instance, making teleportation costs 1mp), the pathfinder will break bad
20091020 22:11:05< alink> silene: I did not benchmark it yet, but it seems to allow skipping 30-40% of expensive move_cost and ZoC checks. (I don't see how the processor can detect that, but i don't know well this kind of things)
20091020 22:12:11< silene> alink: how can you skip 30-40%? the only thing you skip is !(n < next)
20091020 22:13:42< alink> the point is the (next_visited && !(t < next))
20091020 22:14:30< alink> by removing the (&& !(t < next)) there we can skip all that part
20091020 22:14:44-!- allefant [n=elias@allegro/developer/allefant] has quit ["Leaving"]
20091020 22:15:04< alink> (just checking next_visited allow to move it on top)
20091020 22:16:59-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
20091020 22:17:20-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20091020 22:19:50< Soliton> AI0867: have you tested the command you posted?
20091020 22:20:08< Soliton> looks to me like no gui doesn't work anymore.
20091020 22:20:26< AI0867> I have
20091020 22:20:30< AI0867> I just fixed it
20091020 22:21:01< CIA-45> ai0867 * r39386 /trunk/src/play_controller.cpp: Don't try to change the theme if there is no gui
20091020 22:21:02< AI0867> it was broken 4 days ago
20091020 22:22:07< Soliton> i just get "unknown option: --no-gui" when i put it before -m.
20091020 22:22:11-!- shadowmaster_ [n=ignacio@wesnoth/developer/shadowmaster] has quit [Read error: 110 (Connection timed out)]
20091020 22:22:41< alink> silene: when t is exactly known, next_visited && (t < next) is always false, allowing to simplify (next_visited && (t < next)) to next_visited. and thus moving it on top. It's less clear(and probably not the case) for (next_visited && !(n < next)) , so just ignore that part
20091020 22:23:02< alink> * allowing to simplify (next_visited && !(t < next)) to next_visited
20091020 22:24:39< AI0867> --nogui
20091020 22:24:41< AI0867> no -
20091020 22:25:21< Soliton> ok, so you didn't test the command you posted. :->
20091020 22:25:39< AI0867> I tested parts
20091020 22:25:48< AI0867> but his command was mostly broken =P
20091020 22:26:01< Soliton> indeed.
20091020 22:26:22< Soliton> btw, most things are printed to stderr.
20091020 22:27:13< shadowmaster> and that's pretty nonstandard and annoying !
20091020 22:27:32< shadowmaster> (yes, I guess it's only because cerr isn't buffered)
20091020 22:27:39< AI0867> the actual gameresults are printed to stdout
20091020 22:27:43< alink> silene: as for teleport, it could already have special handling and will probably need it anyway if/when it become more "special".
20091020 22:27:45< AI0867> though the faction selection is on stderr
20091020 22:28:18< silene> alink: i see; still, i don't feel too confident about moving away from dijkstra's algorithm: it works for any cost function, while the pathfinder would be stuck with the current one; i think there are other optimizations that would have a much more important impact
20091020 22:28:30< Soliton> yeah, at some point i guess we should switch to stdout. should be a simple change.
20091020 22:28:35-!- Desciero [n=chatzill@cpe-024-074-122-087.carolina.res.rr.com] has joined #wesnoth-dev
20091020 22:30:07< Soliton> though it makes sense to use stderr for the error level at least.
20091020 22:31:22< alink> silene: well that one was easy (it simplify code). and the cost function is not even a function here, the code is in the body of the loop. But ok I will try some benchmark to see how useful it is 
20091020 22:40:23-!- Blueblaze [n=nick@adsl-99-171-161-30.dsl.hstntx.sbcglobal.net] has quit [Remote closed the connection]
20091020 22:43:26< alink> silene: One possibility to cleanly do this is to use cost_calculator like for A* and then add code to get cost_calculator's properties and use it like "if(cost_calc->has_property_X) skip this"
20091020 22:43:51< alink> or similarly, use such trick in the A* code (if possible)
20091020 22:45:53< silene> no, i don't think it is possible in A*; but for this one, if you notice a speed-up, please reapply your patch, just add a comment like "since all the incoming edges have the same cost, we don't need to use Dijksta's algorithm"
20091020 22:46:04-!- EdB [n=edb@79.88.117.178] has quit [Remote closed the connection]
20091020 22:48:11< alink> ok, and indeed the exact same trick doesn't work for A*, i was thinking about this because most cost function ignore the so_far parameter, but I don't remember right now how i wanted to use that :-/
20091020 22:50:29< alink> maybe it was something like caching the local cost into the node
20091020 22:51:55< silene> yes, and impassable tiles, and zoc too
20091020 22:53:32< alink> yes that was the idea, but now I remember that i had problem to both stay general and store specific info
20091020 22:53:33< silene> no, my mistake, impassable tiles are not needed, since they are scanned only once
20091020 22:55:05< alink> mmh not sure about impassable only scanned once, i noticed that it was not true at some point, not sure if i fixed it or not
20091020 22:55:23< silene> hmm... no again... i guess it's getting too late for me to think straight
20091020 22:57:26 * alink must find again and clean his debug-tool showing on map-view how A* expores nodes
20091020 22:57:34< alink> *explores
20091020 23:09:08< Max2008> hi - i've recently updated to trunk (after about a month) and rebuilt (using vs2008 express) and get a crash whenever i try to start a campaign
20091020 23:09:31< Max2008> here's callstack, build warnings: http://wesnoth.pastebin.com/m65d26771
20091020 23:10:09< Max2008> and a --log-debug=gui/* debug log (skipped quite a few lines in between) - http://wesnoth.pastebin.com/m6eedf6a
20091020 23:10:24< ilor_> mordante is sure gonna enjoy this ;)
20091020 23:12:05< Max2008> i'd really appreciate if somebody could have a look - i've reverted to a revision where somebody committed stuff who is also using vs and get the same problem - so it might rather be an issue at my side...
20091020 23:12:54< ilor_> I have near zero time for actual wesnoth work now :/
20091020 23:15:25-!- zookeeper [n=l@wesnoth/developer/zookeeper] has quit []
20091020 23:20:02-!- Nayela [n=Nayela@82.132.136.189] has joined #wesnoth-dev
20091020 23:20:08-!- Nayela [n=Nayela@82.132.136.189] has quit [Client Quit]
20091020 23:52:59-!- alink [n=alink@wesnoth/developer/alink] has quit [Remote closed the connection]
--- Log closed Wed Oct 21 00:00:48 2009