--- Log opened Fri Apr 17 00:00:15 2009
20090417 00:00:38< corn> how do I disable a panel at runtime?
20090417 00:01:06< corn> lets say for example that I want to create a panel similar to that for replays except for observers
20090417 00:01:21< corn> and the panel should only be displayed if the player is an observer in an MP match and not a regular player
20090417 00:05:21-!- stikonas [n=quassel@wesnoth/translator/stikonas] has quit [Read error: 110 (Connection timed out)]
20090417 00:08:36-!- borgrel [n=borgrel@196-210-181-229-snnd-esr-1.dynamic.isadsl.co.za] has quit [Read error: 110 (Connection timed out)]
20090417 00:09:12-!- wesbot changed the topic of #wesnoth-dev to: Summer of Code info: http://www.wesnoth.org/wiki/SummerOfCodeIdeas | 49 bugs, 234 feature requests, 14 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090417 00:22:26-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit ["To err is human.. but to really foul things up requires a computer"]
20090417 00:23:06< boucman> back (but not for long)
20090417 00:28:07< Blueblaze> Does Wesnoth actually process LUA code faster, or it is just a matter of ease in writing and understanding it?
20090417 00:30:30< boucman> Blueblaze: it's very hard to compare to WML, but it's really worth it, if only for the ease of use
20090417 00:30:43< boucman> esp since you can now define WML tags entirely in lua
20090417 00:31:40< Crab_> Blueblaze: lua is *compiled*, so complicated bits of code will be faster in it than in WML
20090417 00:33:41< Blueblaze> Crab_: Is it a significant difference?
20090417 00:34:11< Crab_> Blueblaze: depends on workload, needs benchmarking.
20090417 00:35:26< Blueblaze> boucman: Right, but those custom tags are just glorified macros?
20090417 00:35:48< Crab_> Blueblaze: if you use lua to serialize/deserialize wml object, then i think that it won't be significantly faster. but if you use lua to perform some operations on lua objects, it *should be* faster..
20090417 00:36:05< Crab_> s/object/objects
20090417 00:39:15< boucman> Blueblaze: no, it's a procedure as in "a call to a procedural language" it's not a macro, it results in call to the wesnoth engine, no translation to WML
20090417 00:39:25< ilor_> to me lua seems easier to write and potentially faster, what's not to like
20090417 00:43:11-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 00:46:40< boucman> night all
20090417 00:46:47< Crab_> night, boucman
20090417 00:46:47-!- boucman [n=rosen@wesnoth/developer/boucman] has quit [Read error: 104 (Connection reset by peer)]
20090417 00:53:21-!- karl_m [n=u3468067@dubhe.anu.edu.au] has joined #wesnoth-dev
20090417 00:58:02-!- loonycyborg [n=sergey@wesnoth/developer/loonycyborg] has quit ["Zzzzzzzzzzzzzzzzzzzzzzzzzzzz"]
20090417 00:59:10-!- loonybot [n=loonybot@wesnoth/bot/loonybot] has quit [Remote closed the connection]
20090417 01:03:37-!- molgrum [n=molgrum@c83-249-53-163.bredband.comhem.se] has quit ["Lämnar"]
20090417 01:06:32< Blueblaze> ilor_: How long have you been using lua?
20090417 01:06:46< Blueblaze> If at all
20090417 01:09:32-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has quit ["ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]"]
20090417 01:11:33-!- Gnutoo [n=gnutoo@host13-11-dynamic.31-79-r.retail.telecomitalia.it] has quit [Remote closed the connection]
20090417 01:17:26-!- cjhopman [n=chris@68-118-159-7.dhcp.mdsn.wi.charter.com] has joined #wesnoth-dev
20090417 01:17:26-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20090417 01:27:47-!- DDR [n=chatzill@66.183.125.196] has joined #wesnoth-dev
20090417 01:29:13-!- BenUrban [n=benurban@c-68-49-10-243.hsd1.md.comcast.net] has joined #wesnoth-dev
20090417 01:43:08-!- Sirp [n=me@wesnoth/developer/dave] has joined #wesnoth-dev
20090417 01:45:05< alink> silene: in your new add_localized_overlay(), you use SDL_BlitSurface  to "compose it with the original surface". If these surfaces uses transparency, it's better to use our own blit_surface() which handle alpha correctly when composing image.
20090417 01:47:15-!- Polarina [n=polarina@wesnoth/translator/Polarina] has quit ["Leaving."]
20090417 01:58:23-!- Crab_ [n=Crab_@wesnoth/developer/crab] has quit ["Leaving."]
20090417 02:07:23-!- ikarius [n=ross@smtp.gridironsystems.com] has quit []
20090417 02:14:42-!- ancestral [n=masterpr@97-116-124-173.mpls.qwest.net] has joined #wesnoth-dev
20090417 02:15:06< corn> hi, I want to add a caption ala the kind that shows up when you hover over a button in the main menu. how do I generate the text box and how do I draw it to the screen?
20090417 02:16:27< alink> corn in game or in a dialog ?
20090417 02:16:28< corn> I was trying to find it in the gui::button code but I only found the code that generates the textbox
20090417 02:16:32< corn> in game
20090417 02:16:49-!- Appleman1234 [n=Appleman@131.181.47.7] has joined #wesnoth-dev
20090417 02:16:53< corn> the goal is to create an indication for the play/pause thing that I implemented yesterday
20090417 02:17:12< corn> so the observer doesn't have to go to the menu to figure out what state they're in
20090417 02:17:30< alink> in game, we use tooltip which are defined in theme WML
20090417 02:18:44< corn> that wouldn't accomplish what I need though because I don't want it to be triggered on mouse-over
20090417 02:19:03< corn> is there a way to make tooltips appear without waiting for mouse-over?
20090417 02:19:12-!- noy [n=Noy@70.70.128.147] has joined #wesnoth-dev
20090417 02:19:36< alink> yeah a tooltip with a trigger zone covering all the screen ;-)
20090417 02:20:12< alink> but maybe you can use it on the map area with the theme WML
20090417 02:20:16< corn> lol
20090417 02:20:27< corn> alternatively
20090417 02:20:37< corn> I could make a menu ala the kind in replay mode
20090417 02:20:48< corn> but observers use the same playmp_controller code as regular mp players
20090417 02:20:50< alink> however the tooltip code is independent, so you can maybe hack something
20090417 02:21:05< corn> and afaik there is no way to turn off panels in wml
20090417 02:21:54-!- ancestral_ [n=ancestra@97-116-124-173.mpls.qwest.net] has joined #wesnoth-dev
20090417 02:22:24< Soliton> i think there is some is_observer() function to figure out if you're an observer.
20090417 02:22:59< corn> yes, there is
20090417 02:23:05< corn> I use that function already
20090417 02:23:45< corn> right now though I am looking for some way to display a textbox in game
20090417 02:23:52< corn> for observers
20090417 02:23:57< corn> so
20090417 02:24:15< alink> corn: add some "pause" icon near/on the turn flag field in the top menu bar
20090417 02:24:25< corn> if(is_observer() && _is_networking_paused) { draw("pause") }
20090417 02:24:30-!- ancestral_ [n=ancestra@97-116-124-173.mpls.qwest.net] has quit [Client Quit]
20090417 02:24:37< corn> ah, that's a good idea
20090417 02:24:52< corn> add a pause icon in the wml for the top panel and then control the on/off setting in the current loop
20090417 02:24:55< corn> er
20090417 02:24:56< corn> current logic
20090417 02:24:57< corn> good call
20090417 02:26:10< alink> the image already change each turn, so it's already a dynamic thing not hardcoded in theme WML
20090417 02:27:59< alink> and you can define tootltip in themeWML for the icons there
20090417 02:28:20< corn> cool, thanks
20090417 02:28:54< alink> ah the tooltip for the timer icon seems again broken
20090417 02:29:30 * alink wonders how much times he fixed that
20090417 02:29:55-!- Morian [i=romain@bezut.info] has joined #wesnoth-dev
20090417 02:30:14< alink> that's the kind of sensible spot of themeWML
20090417 02:35:32< alink> ah ok there was never tooltip for this icon, I guess that the previous bugs were for the other icons
20090417 02:38:01< alink> and there is no tooltip because the time indicated is either the current time or the time left for playing your turn
20090417 02:39:11< alink> but that's paradoxical, because the only icon having a variable meaning could use a tooltip
20090417 02:39:43< alink> or better: change the icon to some clock when it indicate hour and minutes
20090417 02:39:57< alink> and use sandglass for time left
20090417 02:43:09-!- Morian__ [i=romain@bezut.info] has joined #wesnoth-dev
20090417 02:43:10< alink> corn: when you will have some experience with themeWML+variable icon, and if you like the feature, maybe you could check how doable it is ^
20090417 02:43:32-!- res|laptop [n=res@pdpc/supporter/student/res2k] has quit [Read error: 110 (Connection timed out)]
20090417 02:44:37-!- res|laptop [n=res@erft-4d07d9ba.pool.mediaWays.net] has joined #wesnoth-dev
20090417 02:44:43-!- Morian_ [i=romain@bezut.info] has quit [Read error: 110 (Connection timed out)]
20090417 02:51:26< alink> a "Time" tooltip works in English ("time left" and "what time is it") but at least in french I don't see a word having both meaning ("temps restant" and "quelle heure est-il ?"
20090417 02:52:27< alink> that's why English is always better for UI. It has all these short words with broad meaning :-)
20090417 02:53:46< alink> translate a English menu in French and you have a nice (but long) poem explaining the precise meaning of each item ;)
20090417 02:57:20< alink> Middle-click scrolling = défilement avec le bouton central de la souris
20090417 02:58:10-!- Morian [i=romain@bezut.info] has quit [Read error: 110 (Connection timed out)]
20090417 02:58:11< alink> which is so long that it create a glitch there
20090417 03:05:57< corn> alink: how do I manipulate the rect= param?
20090417 03:06:09< corn> I want to place a pause icon after the location
20090417 03:06:17< corn> after id=status-position
20090417 03:06:31< corn> or alternatively after the time of day box
20090417 03:08:13< corn> not the time of day box, I mean the box at the top where it shows the actual time of day
20090417 03:08:37< alink> or the time left, see my point above ;) ^
20090417 03:10:09< alink> there is also the observer icon there, maybe next to that
20090417 03:11:13< corn> right
20090417 03:12:18< alink> wait, does your pause feature will work during AI turn too, or just when observing network actions ?
20090417 03:12:39< corn> what do you mean
20090417 03:12:47< corn> the icon will show up whenever the observer pauses the game
20090417 03:12:52< corn> regardless of what player's turn it is
20090417 03:13:49< alink> so also during a local AI turn ?
20090417 03:14:08< corn> I wasn't aware that you could observe your own game
20090417 03:14:19< corn> an AI will never be local to an observer
20090417 03:15:15< alink> i mean i play a SP campaign, I would love be able to pause AI moves.
20090417 03:15:36< corn> no, it won't do that
20090417 03:15:39< alink> and I am not real obsever as we use the word in MP stuff
20090417 03:15:52< corn> although the functionality could be implemented
20090417 03:16:02< corn> where is the tie in for a WML id and source code?
20090417 03:16:15< corn> right now I am searching for 'observers' in the source code and I get a bunch of matches
20090417 03:17:17< alink> just search an other similar id but having a less used name
20090417 03:17:35< alink> with a bit of luck, it will be in the same area
20090417 03:17:48< alink> generate_report probably
20090417 03:18:33< alink> ah now that's where we generate the content   (see  case OBSERVERS:  there)
20090417 03:18:54< alink> it's in reports.cpp
20090417 03:19:36< corn> ah I see
20090417 03:20:13< alink> corn: grep tip: search "observers" not just observers when you want to find where we use a WML key
20090417 03:20:42< corn> what is the difference?
20090417 03:21:12< corn> is "" case insensitive?
20090417 03:21:17< corn> just curious
20090417 03:21:20< alink> you want to find the string = "observers", not the c++ variable called observer-something
20090417 03:21:32< corn> oh
20090417 03:22:28< corn> the thing that generate_report is switching on
20090417 03:22:30< corn> the TYPE
20090417 03:22:34< corn> is that the wml tag?
20090417 03:22:41< alink> no
20090417 03:23:24< alink> see the enum in reports.hpp
20090417 03:24:02< alink> which define the number associated to string in report_names[] defined in reports.cpp
20090417 03:24:46< corn> ah, I see
20090417 03:27:37< corn> hm.
20090417 03:27:44< corn> maybe I should just make the whole thing a button
20090417 03:27:50< corn> instead of an icon
20090417 03:28:14< corn> well, I'll make it an icon for now and see what yogihh says tomorrow
20090417 03:30:09< alink> yes indeed, i initially suggested a icon after your permanent tooltip question, but i suppose that a pause/play button having both the functionality and displaying the info could be nicer
20090417 03:31:01< alink> how do you currently activate the pause in your patch ?
20090417 03:31:02-!- noy [n=Noy@wesnoth/developer/noy] has quit ["GO, GET TO THE CHOPPAH!!!"]
20090417 03:31:10< corn> via menu
20090417 03:31:11< corn> via game menu
20090417 03:31:24< alink> and hotkey, i imagine
20090417 03:31:28< corn> yeah
20090417 03:32:31< alink> maybe you can just steal the play/pause button from replay themeWML and add it in the top bar ?
20090417 03:32:54< alink> mmh and hide it when the replay bar is visible
20090417 03:33:28< corn> in macros.cfg?
20090417 03:34:16< alink> lol, there is no play/pause button in replay bar, but two buttons play and pause, sorry
20090417 03:34:22< corn> well
20090417 03:34:23< corn> same idea
20090417 03:34:42< alink> however their pressed status indicate the playing status
20090417 03:35:24< corn> all I need to do is make it one button and create a method to modify the button img in button.cpp
20090417 03:36:13< alink> or just change "pressed" version of the button
20090417 03:36:40< corn> I wasn't aware that it used seperate button images for different states
20090417 03:36:46< corn> I thought you only specified once
20090417 03:36:49< corn> one*
20090417 03:36:56< alink> all ui buttons image have different version for hovering and pressing
20090417 03:37:18< corn> ok
20090417 03:37:21< corn> yeah, that works
20090417 03:37:28< corn> what is the [add] wml tag?
20090417 03:37:37< corn> it's around the button definitions in macros.cfg
20090417 03:39:14< alink> it seems that the replay mode only modify the current theme see   [replay]  [change] [add] .. in the same file
20090417 03:39:59< alink> sorry  [replay]  [change] [/change ][add] .. but you get the idea
20090417 03:41:05< alink> themeWML is obscure for most dev, I fear that you will probably need to understand it by yourself
20090417 03:41:20< corn> heh
20090417 03:41:39< corn> this is shaping up to be a substantial fix for a simple FR
20090417 03:42:30< corn> what is the item paramter
20090417 03:42:40< corn> I am hacking in a new menu item into default.cfg
20090417 03:42:49< corn> a new menu*
20090417 03:43:00< corn> hm
20090417 03:43:02< corn> however
20090417 03:43:18< corn> if I put in a new menu, how will I be able to control visibility so that it only shows up to observers and not regular MP players?
20090417 03:43:53< corn> it doesn't look like reports.hpp controls menus
20090417 03:44:29< alink> like I said, themeWML is obscure, I have no idea, but as I hinted previously, try to find a place doing a similar thing and copy/adapt it
20090417 03:44:37< corn> yep
20090417 03:44:38< corn> righting
20090417 03:44:40< corn> trying*
20090417 03:44:47< corn> which dev is most into themewml?
20090417 03:44:49< corn> yogihh?
20090417 03:45:00< alink> for example, IIRCk the "end turn" button change in linger mode
20090417 03:45:10< corn> yeah, I found that
20090417 03:45:28< corn> that's in playmp_controller.cpp
20090417 03:48:20< alink> there is no themeWML dev, only people fighting to add the thing needed by their new feature, and when done, they never touch ThemeWML again and just tell stories how horrible it was <:o)
20090417 03:48:28< corn> great
20090417 03:49:27< alink> which as all stories is often exaggerated ;) 
20090417 03:50:00< corn> heh
20090417 03:50:10< corn> well right now I am trying to translate "+2,=,+100,=" into english
20090417 03:50:39< corn> lets see if the wiki has anything on theme wml
20090417 03:50:40< alink> i will give you a link, if themeWML in wiki is not clear enough
20090417 03:51:11< alink> http://www.wesnoth.org/forum/viewtopic.php?p=213708#p213708
20090417 03:51:43-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit ["The man who smiles when things go wrong has thought of someone to blame it on."]
20090417 03:52:13< alink> btw DarthFool seems to know it, but he's not active since some time
20090417 03:53:08< Blueblaze> What are you trying to do with the ThemeWML?
20090417 03:53:12< alink> anyway, I need to go now, good night
20090417 03:53:16< corn> cya
20090417 03:53:21< corn> Blueblaze: trying to add a new button
20090417 03:53:24-!- alink [n=alink@wesnoth/developer/alink] has quit [Read error: 104 (Connection reset by peer)]
20090417 03:53:35< Blueblaze> Where at?
20090417 03:53:40< corn> top-panel
20090417 03:53:49< corn> it will be to the right of the actions menu
20090417 03:53:50< Blueblaze> Next to action?
20090417 03:53:54< corn> yeah
20090417 03:56:27< Blueblaze> corn: That shouldn't be too hard
20090417 03:56:37< corn> trying to figure out what to put for the rect= tag
20090417 03:56:47< Blueblaze> Just copy what the action menu uses
20090417 03:57:00< corn> tried that and it broke completely
20090417 03:57:01< Blueblaze> Then, you'll have to adjust the other icons like gold and so forth
20090417 03:57:17< corn> gah
20090417 03:57:20< corn> no relative placement?
20090417 03:57:54< Blueblaze> Yes there is, that's what the + is for
20090417 03:58:30< corn> what about for tags with no ref=?
20090417 03:58:41< corn> all of the panels on the top panel have no ref tag
20090417 03:59:04< Blueblaze> Main Menu does
20090417 03:59:10< Blueblaze> And the others are built upon that
20090417 03:59:19< corn> no, check out gold-panel for example
20090417 03:59:25< Blueblaze> Main menu refers to top-panel
20090417 03:59:51< Blueblaze> Check turn-panel
20090417 03:59:54< Blueblaze> That refers to action panel
20090417 03:59:58< Blueblaze> Gold panel refers to the turn panel
20090417 04:00:03< corn> ah, I understand
20090417 04:00:33< Blueblaze> If it doesn't say specifically what the reference is, then it is the object above it
20090417 04:00:53< Blueblaze> Basically, to do what you want to do, you need to do two things, make your new panel and add it directly under action menu
20090417 04:01:08< Blueblaze> Then edit turn panel so that it refers to the new menu you created
20090417 04:01:20< corn> yep, done
20090417 04:01:26< Blueblaze> So, that should be all you have to change :)
20090417 04:01:31< corn> how do I change the dimensions?
20090417 04:01:40< corn> I hover over the icon and it gives me the tooltip
20090417 04:01:44< corn> but a large area to the right also does
20090417 04:02:10< Blueblaze> Hmm? menus don't give tooltips
20090417 04:02:35< Blueblaze> Oh, the icon for what?
20090417 04:02:46< corn> pause icon
20090417 04:02:48< corn> I just put it in
20090417 04:02:55< corn> I think I need to change the c,d param of rect
20090417 04:03:46< Blueblaze> The tooltip is controled by [label]
20090417 04:04:01< Blueblaze> That's just a matter of editting the rect as you said though :p
20090417 04:06:27< corn> Blueblaze: thanks
20090417 04:06:53< corn> now, I need to be able to hide this menu at times
20090417 04:06:57< corn> any idea how to do that?
20090417 04:07:04< corn> it should only show up to observers for now
20090417 04:08:13< Blueblaze> So, this new menu should only show to observers?
20090417 04:08:18< corn> yep
20090417 04:08:19< Blueblaze> I don't think that's possible :p
20090417 04:08:29< corn> heh. I'll have to hack it in then
20090417 04:08:39< Blueblaze> The theme is the same for all players/observers
20090417 04:09:34< Blueblaze> If you really wanted to, it might be a good idea to make a command like observer_theme=, which if left blank will use theme=
20090417 04:09:53< corn> is there any way to transition from an observer to a player in-game?
20090417 04:10:17< Blueblaze> Yes
20090417 04:10:29< Blueblaze> Observers can become players if the host assigns it to them
20090417 04:10:34< corn> ok
20090417 04:10:45< Blueblaze> :control <side> <player>
20090417 04:11:29< corn> I wonder how much effort it would take to make the menu removable in-game
20090417 04:13:56< corn> can you reload theme in game?
20090417 04:14:09< corn> no, you can;t
20090417 04:14:09< Blueblaze> probably a lot lol
20090417 04:14:09< corn> ?/
20090417 04:14:10< corn> :/
20090417 04:14:30< Blueblaze> Yeah, it would be nifty if you could change themes during a game
20090417 04:14:40< corn> I wonder what's preventing it now
20090417 04:14:42< corn> I'll check it later
20090417 04:15:07< Blueblaze> Well themes are loaded before the scenario starts
20090417 04:20:12< corn> true
20090417 04:20:15< corn> ok, I'll be back in an hour
20090417 04:23:10-!- csaunders [n=christop@ip216-239-85-191.vif.net] has joined #wesnoth-dev
20090417 04:35:20-!- Shadow_Master [n=ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20090417 04:41:00-!- Ivanovic_ [n=ivanovic@dtmd-4db2d613.pool.einsundeins.de] has joined #wesnoth-dev
20090417 04:45:56-!- Netsplit niven.freenode.net <-> irc.freenode.net quits: AndreLuiz, JW1, turin, yann
20090417 04:47:06-!- Netsplit niven.freenode.net <-> irc.freenode.net quits: Dragonking, ABCD, CIA-30
20090417 04:49:06-!- Smar [n=smar@freenet/translator/finnish/Smar] has quit [Remote closed the connection]
20090417 04:49:17-!- Smar [n=smar@88.113.60.192] has joined #wesnoth-dev
20090417 04:52:42-!- johani [n=johani@c83-250-162-206.bredband.comhem.se] has quit [Read error: 104 (Connection reset by peer)]
20090417 04:52:50-!- johani [n=johani@c83-250-162-206.bredband.comhem.se] has joined #wesnoth-dev
20090417 04:54:02-!- Netsplit over, joins: JW1, yann, turin, AndreLuiz, Dragonking, ABCD, CIA-30
20090417 04:56:37-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 110 (Connection timed out)]
20090417 04:56:58-!- Ivanovic_ is now known as Ivanovic
20090417 05:13:51< corn> Ivanovic: I have updated my app
20090417 05:15:50-!- ABCD is now known as ABCD_away
20090417 05:22:09-!- Morian [i=romain@bezut.info] has joined #wesnoth-dev
20090417 05:30:01-!- ilor_ [n=user@wesnoth/developer/ilor] has quit []
20090417 05:36:21-!- Morian__ [i=romain@bezut.info] has quit [Read error: 110 (Connection timed out)]
20090417 05:40:48-!- Chusslove [n=caslav@brsg-d9bef9e0.pool.mediaWays.net] has quit [Read error: 110 (Connection timed out)]
20090417 05:46:34-!- Netsplit niven.freenode.net <-> irc.freenode.net quits: teaser, Espreon, DDR
20090417 05:47:22-!- Netsplit over, joins: DDR, Espreon, teaser
20090417 05:49:41-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090417 05:55:44-!- JW1 [n=X@c-71-57-85-168.hsd1.il.comcast.net] has quit []
20090417 05:56:32-!- crimson_penguin [n=ben@wesnoth/developer/crimsonpenguin] has quit ["Leaving"]
20090417 06:14:30-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090417 06:20:22-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090417 06:32:00-!- csaunders [n=christop@ip216-239-85-191.vif.net] has quit ["Grab life by the BitchX"]
20090417 06:36:30-!- xonev [n=chatzill@59.92.70.67] has joined #wesnoth-dev
20090417 06:41:34< esr> fabi: I've just had a wonderfully evil idea for DM.
20090417 06:46:10-!- Appleman1234 [n=Appleman@131.181.47.7] has quit [Read error: 60 (Operation timed out)]
20090417 06:53:02-!- Appleman1234 [n=Appleman@131.181.47.13] has joined #wesnoth-dev
20090417 06:59:47-!- DDR [n=chatzill@66.183.125.196] has quit ["ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]"]
20090417 07:05:33-!- silene [n=plouf@ASte-Genev-Bois-152-1-13-176.w82-121.abo.wanadoo.fr] has joined #wesnoth-dev
20090417 07:22:05-!- xonev_ [n=chatzill@59.92.70.67] has joined #wesnoth-dev
20090417 07:23:47-!- xonev [n=chatzill@59.92.70.67] has quit [Read error: 110 (Connection timed out)]
20090417 07:24:00-!- xonev_ is now known as xonev
20090417 07:27:42-!- BenUrban [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
20090417 07:34:47-!- BenUrban [n=benurban@c-68-49-10-243.hsd1.md.comcast.net] has joined #wesnoth-dev
20090417 07:42:49< dfranke> esr: Is there API documentation anywhere for wmlparser.py?  Or do I just RTFS?
20090417 07:48:01< dfranke> esr: oh... I assumed it was your code, but svn blame says otherwise.
20090417 07:51:55-!- YogiHH [n=chatzill@c203051.adsl.hansenet.de] has joined #wesnoth-dev
20090417 07:56:26-!- YogiHH [n=chatzill@c203051.adsl.hansenet.de] has left #wesnoth-dev []
20090417 08:00:21-!- Sirp [n=me@wesnoth/developer/dave] has quit ["leaving"]
20090417 08:13:34-!- mordante [n=mordante@87.215.201.26] has joined #wesnoth-dev
20090417 08:13:37< mordante> hi
20090417 08:16:05< Blueblaze> hey mordante
20090417 08:16:10< mordante> hi Blueblaze 
20090417 08:16:22< Blueblaze> Do you know what image file is used for the background of the message box?
20090417 08:17:31< mordante> which message box we have several of them most are in images/dialogs
20090417 08:17:52< mordante> YogiHH, Ivanovic no idea why the tiny-gui box pops up
20090417 08:18:13< Blueblaze> The standard [message] box, where regular dialogue is used
20090417 08:18:16< mordante> but I'm busy fixing cmake with tiny-gui so I can do more testing with it
20090417 08:18:34< Blueblaze> Which image from images/dialogs does it use?
20090417 08:18:53< Blueblaze> Since there are about 10 different similar possibilities :p
20090417 08:21:03< mordante> about 7 of them ;-) :-P
20090417 08:21:33< mordante> oh wait that one only 3 ;-)
20090417 08:21:53< Blueblaze> So it doesn't use any of the translucent-background_xx?
20090417 08:22:26< mordante> dialogs/translucent65[-border-top.png|-border-bottom.png|-background.png]
20090417 08:22:34< Blueblaze> ah okay
20090417 08:22:37< Blueblaze> thanks
20090417 08:24:15< mordante> esr, digits are illegal in terrain names, it might work it might fail it's documented in the wiki ;-)
20090417 08:25:11< Blueblaze> mordante: Oh, are there any plans on creating (or resurrecting the old one accessible via wml) a [message] box that is more fitting for non-dialogue messages/options? The new messages look great for dialogue, but not so flattering for everything else :P
20090417 08:26:24< mordante> Blueblaze, not sure yet, maybe I leave them since they look bad (which is a pre in this case) ...
20090417 08:26:41< mordante> silene, wants to look at allowing lua to show dialogs with the new widgets
20090417 08:27:09< mordante> in that case the old option parts of message are not that important anymore and we can try to phase them out
20090417 08:27:59-!- Reisiger [n=Reisiger@adsl-62-167-41-96.adslplus.ch] has joined #wesnoth-dev
20090417 08:28:14< Blueblaze> mmm, I'm so unexcited about the incorporation of lua :P
20090417 08:28:26-!- Netsplit niven.freenode.net <-> irc.freenode.net quits: johani, silene
20090417 08:28:42-!- Netsplit over, joins: silene, johani
20090417 08:30:32-!- Guest61089 [n=synergiz@c-75-73-225-141.hsd1.mn.comcast.net] has quit ["Leaving"]
20090417 08:31:11< Blueblaze> Not fun from having a good handle on wml to having no clue what lua is :p
20090417 08:32:31< mordante> we still need to look what to do with those message boxes, but in order to show them in different dialogs they should be different wml items
20090417 08:32:58< mordante> and somehow I expect once silene and I finish the lua dialogs some people eg bob might switch to them
20090417 08:33:08< mordante> and a new era of addons will appear
20090417 08:33:38< Blueblaze> mordante: any idea how difficult it is to switch to them if you have no coding experience except wml?
20090417 08:34:05< Blueblaze> lua for Wesnoth in general that is
20090417 08:35:52< mordante> not sure not really looked at lua and I've too much coding experience to really judge it
20090417 08:36:23< Blueblaze> mmm, maybe zookeeper would be a good judge of it, since afaik all he codes is wml, though very well
20090417 08:37:05< mordante> I think indeed he is a better judge, or just become one yourself ;-)
20090417 08:37:32< Blueblaze> I barely understand much of it, so I need a second opinion :P
20090417 08:38:11< mordante> corn, not sure what you want to do with the gui but we have two gui engines, so if you write an entirely new dialog I prefer it if you use the new code
20090417 08:39:09< corn> mordante: I don't want a dialog, I want to put a new button into top-panel
20090417 08:39:23< Blueblaze> mordante: I am curious how people who are the real contributors via wml will take to lua (ie, people who make the campaigns and mp scenarios) since I haven't heard much about it/concerning it from prominent wml users
20090417 08:39:33< corn> however, there should be an ability to remove or add the button to view in the code
20090417 08:39:56< Blueblaze> Since, I've just heard about lua from people who really don't code much of anything in wml :p
20090417 08:43:01< mordante> corn, not sure whether that can be done with the old widgets, it can't with the new one but you can make the button invisible
20090417 08:43:40< mordante> Blueblaze, several people have asked for lua in the forums...
20090417 08:43:55< corn> mordante: that's fine. the button will be used right now for observers but hopefully will be a pause button that can be used in all contexts (sp, observer in mp, regular mp)
20090417 08:44:16< mordante> but we'll see whether lua gets used when the development branch gets more releases
20090417 08:44:26< corn> how do I use the new widget code?
20090417 08:44:46< mordante> you can't mix old and new so you have to do it entirely in new
20090417 08:45:10-!- Reisiger [n=Reisiger@adsl-62-167-41-96.adslplus.ch] has quit [Read error: 104 (Connection reset by peer)]
20090417 08:45:21< corn> how do I even access the new?
20090417 08:45:23< mordante> more info here http://www.wesnoth.org/wiki/GUIToolkit
20090417 08:45:26< corn> I just defined the button in themewml
20090417 08:45:30-!- Reisiger [n=Reisiger@adsl-62-167-41-96.adslplus.ch] has joined #wesnoth-dev
20090417 08:45:30< corn> ok
20090417 08:45:45< mordante> and the code is in src/gui and the definitions in data/gui
20090417 08:45:51< mordante> got to go now
20090417 08:46:00< corn> ah, ok
20090417 08:46:04< corn> is the existing ui ported?
20090417 08:46:21< mordante> in progress, which is the reason the old stuff is still there
20090417 08:46:26< corn> ok
20090417 08:46:28< corn> thanks
20090417 08:46:29< corn> cya later
20090417 08:46:49-!- turin [n=turin@168.215.250.1] has quit [Read error: 110 (Connection timed out)]
20090417 08:47:26< corn> I am going to go to bed now as well. I'll do an easycoding task for the gui code and then I'll talk to yogi about how to approach this problem. Maybe there is no need for a pause button, but we'll see
20090417 08:47:59< Blueblaze> mordante: I've seen that too, but I can only recall that most of the comments were from people who I haven't seen release anything with WML, though I may be quite wrong
20090417 08:49:24< mordante> the future will tell, but I think lua offers more exciting features, but no idea whether the learning is higher or not
20090417 09:03:00< silene> Blueblaze: you can't compare lua and wml with respect to learning; as a language, lua is much easier to learn: after a few minutes (an hour if it's your first time seeing a programming language), you will have some code that does something useful, while it would take days for wml; but as an interface to wesnoth, it is not easier, since you have to know a bit about wml anyway
20090417 09:04:07< silene> Blueblaze: the main difference will appear once you are proficient with both language, you will waste a lot less time trying to implement things that if you knew only wml
20090417 09:05:32< silene> Blueblaze: to give you an example, a few days ago, while he had never used lua in wesnoth, crab had wrote a dynamic interface for controlling the ai in about one hour; it would have been impossible in such a short time in wml
20090417 09:10:48-!- Appleman1234 [n=Appleman@131.181.47.13] has quit [Read error: 110 (Connection timed out)]
20090417 09:15:39-!- Synergy [n=synergiz@c-75-73-225-141.hsd1.mn.comcast.net] has joined #wesnoth-dev
20090417 09:16:06-!- Synergy is now known as Guest68724
20090417 09:17:45-!- silene [n=plouf@wesnoth/developer/silene] has quit ["Leaving."]
20090417 09:18:10 * Rhonda laughs at Shadow_Master
20090417 09:18:30< Rhonda> Shadow_Master: You really try to tell me that a robots.txt hinders you from downloading something?  8-)
20090417 09:18:53< Rhonda> Learn to understand the software you use, dood.
20090417 09:20:38< Rhonda> ... sorry, the nick in the query was a different one.
20090417 09:40:03-!- Espreon [n=espreon@wesnoth/developer/espreon] has quit [Remote closed the connection]
20090417 09:44:43< Ivanovic> moin
20090417 09:49:34-!- Chusslove [n=caslav@brsg-d9bef01a.pool.mediaWays.net] has joined #wesnoth-dev
20090417 09:51:26-!- Blueblaze2 [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090417 09:52:40-!- Reisiger [n=Reisiger@adsl-62-167-41-96.adslplus.ch] has quit ["Verlassend"]
20090417 09:53:38-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Nick collision from services.]
20090417 09:53:42-!- Blueblaze2 is now known as Blueblaze
20090417 09:57:17-!- euschn [n=chatzill@turing.logic.tuwien.ac.at] has joined #wesnoth-dev
20090417 09:57:24< euschn> good morning
20090417 09:57:42-!- Guest68724 [n=synergiz@c-75-73-225-141.hsd1.mn.comcast.net] has quit [Remote closed the connection]
20090417 10:02:40-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090417 10:05:22-!- Morian [i=romain@bezut.info] has quit ["Leaving"]
20090417 10:10:47-!- YogiHH [i=c3f5f398@wesnoth/developer/yogihh] has joined #wesnoth-dev
20090417 10:13:40-!- zookeeper [n=l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20090417 10:18:23< zookeeper> corn, there isn't much space left in the top panel for new buttons. it's pretty much full except on resolutions wider than 1024/1280
20090417 10:42:08-!- stikonas [n=quassel@ctv-213-164-120-170.vinita.lt] has joined #wesnoth-dev
20090417 10:57:06-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090417 11:04:30-!- stikonas [n=quassel@wesnoth/translator/stikonas] has quit [Remote closed the connection]
20090417 11:06:31-!- edb [n=edb@58.12.95-79.rev.gaoland.net] has joined #wesnoth-dev
20090417 11:07:33< Ivanovic> zookeeper: maybe he wants to add another row like it is done for replays?
20090417 11:07:50< Ivanovic> corn: in general yogihh probably knows most about what he has done to achive this
20090417 11:08:26< Ivanovic> YogiHH: for you as reference, this nights logs, right at midnight
20090417 11:18:42-!- ancestral [n=masterpr@97-116-124-173.mpls.qwest.net] has quit ["And that's the end of THAT chapter"]
20090417 11:41:50-!- edb is now known as EdB
20090417 11:41:56-!- fabi [n=fabi@wesnoth/developer/fendrin] has quit [Remote closed the connection]
20090417 11:47:02-!- loonybot [n=loonybot@79.139.246.210] has joined #wesnoth-dev
20090417 11:48:13-!- loonycyborg [n=sergey@79.139.246.210] has joined #wesnoth-dev
20090417 11:54:43-!- Ivanovic_ [n=ivanovic@dtmd-4db2cdcf.pool.einsundeins.de] has joined #wesnoth-dev
20090417 11:59:04-!- Ivanovic__ [n=ivanovic@dtmd-4db2d04c.pool.einsundeins.de] has joined #wesnoth-dev
20090417 12:05:28-!- Ivanovic [n=ivanovic@wesnoth/developer/ivanovic] has quit [Nick collision from services.]
20090417 12:05:34-!- Ivanovic__ is now known as Ivanovic
20090417 12:09:12-!- wesbot changed the topic of #wesnoth-dev to: Summer of Code info: http://www.wesnoth.org/wiki/SummerOfCodeIdeas | 50 bugs, 234 feature requests, 14 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090417 12:12:15-!- Ivanovic_ [n=ivanovic@wesnoth/developer/ivanovic] has quit [Read error: 110 (Connection timed out)]
20090417 12:40:15-!- mordante [n=mordante@87.215.201.26] has quit ["Leaving"]
20090417 12:56:32< Ivanovic> fendrin, zookeeper: any idea why utbs has copies of the strings for healing and whatever that already exist in mainline?
20090417 12:56:45< Ivanovic> (yes, they are put into the main textdomain, but are they *really* needed?)
20090417 12:57:57< Ivanovic> cases for this are berserk, firststrike and magical in 03_Stirring_in_the_Night.cfg
20090417 12:58:22< Ivanovic> slowed in 12_The_Final_Confrontation.cfg
20090417 12:58:36< Ivanovic> as well as healing stuff in utils/abilities.cfg
20090417 12:59:54-!- cjhopman [n=chris@wesnoth/developer/cjhopman] has quit ["Konversation terminated!"]
20090417 13:00:22< zookeeper> berserk, firststrike and magical don't need those strings there, they could be used through the standard macros
20090417 13:01:07< zookeeper> the duplicated healing strings _might_ be possible to remove
20090417 13:01:24< zookeeper> i'll have to take a look at that later, i'm not entirely sure
20090417 13:01:36< Ivanovic> hmm, the healing is a change to have the dehydration in
20090417 13:01:41< zookeeper> yep
20090417 13:01:42< Ivanovic> might be difficult there...
20090417 13:01:58< Ivanovic> but at least the stuff in the scenarios is "bad style" IMO
20090417 13:02:02< zookeeper> {ABILITY_HEALS} [+heals] [affect_adjacent] ...
20090417 13:02:08< zookeeper> (might work)
20090417 13:02:20< zookeeper> really? what's the problem with them?
20090417 13:04:37< esr> Morning, all.  I have two cool ways I could end the boss fight in DM and am undecided between them.
20090417 13:04:39< Ivanovic> in general using stuff that is a plain copy of stuff that already is available in mainline
20090417 13:04:57< Ivanovic> esr: good that you are around, a request i posted some days ago that you seem to have missed:
20090417 13:05:21< Ivanovic> what about the commit by mordante meant to fix a type ("and" at the beginning of a string to "And")
20090417 13:05:43< Ivanovic> do you think this changes the meaning of the string significantly, or should this be done by pofix.py?
20090417 13:06:02< esr> (1) Delfador must strike the final blow on Iliah-Malal using his knowledge from the Book of Crelanu; any other deathblow on himm triggers a full-heal.
20090417 13:06:11< Ivanovic> http://svn.gna.org/viewcvs/wesnoth/branches/1.6/data/campaigns/Legend_of_Wesmere/scenarios/04_Elvish_Treasury.cfg?rev=34822&r1=34821&r2=34822&view=diff
20090417 13:06:18< Ivanovic> (the same change in trunk)
20090417 13:06:57< esr> That's not a change of mweaning.
20090417 13:07:11< CIA-30> ivanovic * r34975 /trunk/ (825 files in 33 dirs): 
20090417 13:07:11< CIA-30> pot-update (yes, trunk, translators should not work on it, but at least the latest DM stuff should be in...)
20090417 13:07:11< CIA-30> regenerated doc files
20090417 13:07:25< Ivanovic> ahh, for some langs it might appear as such since the start of the string is included and with composed strings...
20090417 13:07:26< Ivanovic> ;)
20090417 13:07:38< Ivanovic> that is why i would like to hear your opinion
20090417 13:08:01< esr> (2) Delf gets to summonm his veterans from the Land of the Dead in 20 only, and one of *them* must strike the final blow...Iliah-Malal can only be killed by an undead.
20090417 13:08:14< esr> An undead fighting for the living.
20090417 13:08:14< Ivanovic> at least in the german translation the string was started with a non capital letter
20090417 13:08:41< esr> WSo, which ending is cooler?  I am undecided.
20090417 13:08:47< Ivanovic> i think 2) mixed with 1) is cooler
20090417 13:08:53< esr> ?
20090417 13:08:59< Ivanovic> if he is killed by a non undead unit the enemy is leveled+healed
20090417 13:10:30< zookeeper> Ivanovic, the slowed string in the UtBS finale needs to stay unless it gets macroified in core
20090417 13:11:39< esr> Ivanovic: I must be slow this morning, I don't understand eitherr of your responses.
20090417 13:11:43< Ivanovic> zookeeper: would slowed as a macro make sense?
20090417 13:12:07< zookeeper> Ivanovic, i think i'd find it slightly silly to do that, frankly. it's not much different from any other core or hardcoded little string like that.
20090417 13:12:31< zookeeper> "Ambush!", "poisoned", "Enemy sighted", etc etc etc
20090417 13:12:37< Ivanovic> esr: re translation: screw my request since it *is* a change in meaning (a string starting with a non capital letter means (to many translators not reading the sources) that it was used in some composed string as 2nd part
20090417 13:12:49< Ivanovic> so when it is not with a capital start, it is a different meaning
20090417 13:13:11< Ivanovic> tiem for breakfast, back "soon"
20090417 13:13:19< Ivanovic> esr: re what to do about the ending:
20090417 13:13:26< zookeeper> hmh, maybe we could have a core macro for something like that...like {STRING_FROM_DIFFERENT_TEXTDOMAIN wesnoth _"slowed"}, but i have no idea if it could work.
20090417 13:13:35< Ivanovic> have several units allowed but only an undead unit can strike the killing blow
20090417 13:14:16< Ivanovic> if a non undead unit tries to do so, have some "muhaha, no living being can kill me" stuff said and level+heal the enemy leader
20090417 13:14:41< Ivanovic> zookeeper: this would require a change to wmlxggettext and, even worse, some checking if the string *really* exists in the other textdomain
20090417 13:14:43< Ivanovic> bad one...
20090417 13:16:53< esr> Ivanovic: So how is your propsal different from (2) Whayt happens if Deldrives him to negative HP?
20090417 13:17:18-!- Alesis-Novik [n=alesis@78.60.188.59] has quit [""Gone for the weekend""]
20090417 13:17:48< zookeeper> Ivanovic, well, i'd suggest don't-fix in this case then...
20090417 13:25:15< Ivanovic> esr: if delf does so the unit is revived and gains a level
20090417 13:25:52< Ivanovic> esr: is delf also able to recruit living units in this scenario?
20090417 13:26:00< Ivanovic> if he is, those can't kill the boss either
20090417 13:26:25< Ivanovic> zookeeper: the stuff with existing macros in mainline can and should be fixed though, right?
20090417 13:26:33< Ivanovic> (where exactly the same effect is wanted)
20090417 13:27:24< esr> Ivanovic: It sounds like you're just describing my proposal (2).  How is (1) "mixed with" it?
20090417 13:27:45< Ivanovic> any other deathblow on himm triggers a full-heal.
20090417 13:27:55< Ivanovic> you have not had the full heal in 2)
20090417 13:28:07< Ivanovic> it could just be "back to the status right before the last hitting attack"
20090417 13:35:38< CIA-30> ivanovic * r34976 /branches/1.6/ (198 files in 5 dirs): 
20090417 13:35:38< CIA-30> pot-update (one altered string in wesnoth-low: "and (...)" -> "And (...)" (string is a message of its own, not part of some composed string and should thus start with a capital letter))
20090417 13:35:38< CIA-30> regenerated doc files
20090417 13:37:51-!- gabm [n=Gabriel@72.0.215.2] has left #wesnoth-dev []
20090417 13:42:32< zookeeper> Ivanovic, berserk, firststrike and magical? yes, as i said, i'll get around to it sometime soon
20090417 13:42:53< zookeeper> and healing
20090417 13:55:38-!- xonev [n=chatzill@59.92.70.67] has quit [Read error: 110 (Connection timed out)]
20090417 14:00:06< CIA-30> esr * r34977 /trunk/data/campaigns/Delfadors_Memoirs/ (scenarios/05_council_in_weldyn.cfg utils/sides.cfg): 
20090417 14:00:06< CIA-30> Fix a reference to King Garard's portrait.
20090417 14:00:06< CIA-30> Make Deglfador's undead veterans available for scenario 20
20090417 14:04:10-!- AndreLuiz [n=AndreLui@unaffiliated/andreluiz] has quit [Remote closed the connection]
20090417 14:04:22-!- Crab_ [n=Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20090417 14:04:25< Crab_> hi
20090417 14:08:45< Shadow_Master> Rhonda: that is what I get for not reading the "see also" section in the wget man page, I guess :(
20090417 14:09:22-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 14:09:50< Shadow_Master> thanks
20090417 14:17:14-!- boucman [n=rosen@2.117.71-86.rev.gaoland.net] has joined #wesnoth-dev
20090417 14:17:46< boucman> hey all
20090417 14:17:54< Crab_> hi boucman
20090417 14:26:32-!- EdB [n=edb@58.12.95-79.rev.gaoland.net] has quit [Remote closed the connection]
20090417 14:35:43< boucman> corn: around ?
20090417 14:38:33-!- AndreLuiz [n=AndreLui@unaffiliated/andreluiz] has joined #wesnoth-dev
20090417 14:39:24< Ivanovic> boucman: he went to bed about 5h ago
20090417 14:39:27< Ivanovic> so probably not
20090417 14:39:35< Ivanovic> boucman: but maybe he will read the logs
20090417 14:39:37< Ivanovic> ;)
20090417 14:41:50< boucman> np, I'll comment in the patch instead
20090417 14:45:58< Ivanovic> good ida
20090417 14:46:01< Ivanovic> s/ida/idea
20090417 14:49:25< CIA-30> ivanovic * r34978 /trunk/po/wesnoth-low/de.po: updated German translation
20090417 14:49:26< CIA-30> ivanovic * r34979 /branches/1.6/po/wesnoth-low/de.po: updated German translation
20090417 14:56:47-!- AndreLuiz [n=AndreLui@unaffiliated/andreluiz] has quit ["Leaving"]
20090417 15:02:16< CIA-30> boucman * r34980 /trunk/ (changelog src/log.cpp src/log.hpp src/upload_log.cpp): put stat upload logs in a separate logdomain, patch by corn
20090417 15:05:14-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 15:05:51< CIA-30> ivanovic * r34983 /branches/1.6/po/wesnoth-low/de.po: updated German translation
20090417 15:05:52< CIA-30> ivanovic * r34981 /trunk/po/wesnoth-low/de.po: updated German translation
20090417 15:05:52< CIA-30> esr * r34982 /trunk/data/campaigns/Delfadors_Memoirs/scenarios/19_the_portal_of_doom.cfg: Suppress a wmllint message by reordering some code.
20090417 15:06:45-!- ettin [n=jorda@wesnoth/developer/ettin] has quit [Read error: 60 (Operation timed out)]
20090417 15:11:22-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 110 (Connection timed out)]
20090417 15:11:46-!- elias [n=allefant@allegro/developer/allefant] has joined #wesnoth-dev
20090417 15:29:10< esr> zookeeper: Is there a convenient way to count the number of units matching a specified SUF?
20090417 15:31:19-!- BenUrban [n=benurban@unaffiliated/benurban] has quit ["Power failu"]
20090417 15:37:28< zookeeper> esr, store them and you'll find the number as $units.length
20090417 15:38:01-!- happygrue [n=quassel@wesnoth/developer/wintermute] has joined #wesnoth-dev
20090417 15:47:35< corn> boucman: now I am here
20090417 15:47:55< boucman> morning :)
20090417 15:48:16< boucman> so, I've commited your logging patch (congratulation)
20090417 15:48:19< corn> thanks
20090417 15:48:24< boucman> and I was looking at your sorting patches
20090417 15:48:30< corn> I read your comments on the sorting patch
20090417 15:48:33< corn> yep
20090417 15:49:00< boucman> ok, I havn't looked at the other one yet, does it apply on top of the first one, or separately ?
20090417 15:49:06< corn> it is more useful to have a sort method that uses 2 columns in this case because level and exp are not independent to each other
20090417 15:49:20< corn> the 2nd one (which your comments apply to) also applies to the 1st in a way
20090417 15:49:32< corn> the 1st is only applied to exp column but takes in 2 columns as parameters
20090417 15:50:03< corn> the 2nd one takes in 2 columns and can be used on both the exp column and the level column
20090417 15:51:49< Shadow_Master> cycholka's DNS blackhole idea makes previewing and editing/posting new posts particularly uncomfortable and tedious
20090417 15:52:00< Shadow_Master> and we are still banning more than 2 spambots a      day
20090417 15:52:22< Shadow_Master> Ivanovic: perhaps you could help me bug him to disable it ;)
20090417 15:52:38-!- molgrum [n=molgrum@c83-249-53-163.bredband.comhem.se] has joined #wesnoth-dev
20090417 15:52:42< boucman> corn: ok, well, could you post a single post for both FR, then ? (just tell me which one I should review)
20090417 15:52:48< corn> sure
20090417 15:53:02< corn> wait, can you tell me what you mean by post
20090417 15:53:18< corn> patch on the tracker?
20090417 15:53:18< boucman> add to one of the currently opened patches
20090417 15:53:21< boucman> yes
20090417 15:53:23< corn> ah, ok
20090417 15:57:17< Ivanovic> Shadow_Master: today we even had one in with ~10 porn posts
20090417 15:57:36< Ivanovic> Shadow_Master: and yeah, i think we need some different way for login
20090417 15:57:45< Ivanovic> s/login/registration
20090417 15:57:57< Ivanovic> eg some "do your math" stuff
20090417 15:58:09< Ivanovic> or "answer a wesnoth question"
20090417 15:58:10< Ivanovic> ^^
20090417 15:58:35< Soliton> Shadow_Master: if you can not make it only check new accounts then turn it off.
20090417 15:59:01< Shadow_Master> I plan to try to upgrade the forums to phpBB 3.0.4 this weekend or during next week
20090417 15:59:37< Shadow_Master> if someone would be so kind to find some better CAPTCHA script while I work on making our mods work with phpBB 3.0.4 on my laptop...
20090417 16:02:51< corn> boucman: I've updated the patch https://gna.org/patch/index.php?1165
20090417 16:07:32< boucman> corn: ok, I'll close 1161 in that case so we only have one patch for both related features
20090417 16:07:55< corn> 1161 is already commited afaik, just to make sure we're on the same page
20090417 16:08:46< boucman> corn: you're probably right, I've only looked at it, not applied yet
20090417 16:08:58< corn> ok
20090417 16:08:59< boucman> corn: do you know who comitted it ? (he forgot to mark as such)
20090417 16:09:03< corn> dfranke
20090417 16:09:21< boucman> ok dfranke around ?
20090417 16:10:29< boucman> corn: ok, let's discuss your patch here instead of using the patch interface if it's only discussion
20090417 16:10:36< corn> sure
20090417 16:11:27< boucman> I don't like the idea of a sorter that use another column for sorting... it should be something else that takes care of the sub-sorting
20090417 16:12:01< boucman> moreover, knowing mordante, chances are high that there is something like that elsewhere in the GUI code, so I won't commit, i'll wait for his commands
20090417 16:12:03< corn> as in, a generalization of the current sorting code so that all the columns can have a 2nd column to sub sort by?
20090417 16:12:27< boucman> usually, the second order sort is the previous sort order
20090417 16:12:48< boucman> i.e if you click level then click xp it will be sorted by xp then by level
20090417 16:12:59< boucman> anyway, mordante is the one taht would know that
20090417 16:13:02< corn> ok
20090417 16:13:08< corn> thanks for looking at the patch
20090417 16:13:15< boucman> np
20090417 16:18:45-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection reset by peer]
20090417 16:19:00-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 16:20:35-!- happygrue [n=quassel@wesnoth/developer/wintermute] has quit [Remote closed the connection]
20090417 16:20:50-!- teaser [n=tsr@c213-89-114-91.bredband.comhem.se] has quit [Read error: 60 (Operation timed out)]
20090417 16:24:41-!- euschn [n=chatzill@turing.logic.tuwien.ac.at] has quit ["ChatZilla 0.9.84 [Iceweasel 3.0.6/2009032803]"]
20090417 16:26:51-!- Morian [n=romain@som62-1-82-238-56-126.fbx.proxad.net] has joined #wesnoth-dev
20090417 16:34:48-!- crimson_penguin [n=ben@64.201.60.216] has joined #wesnoth-dev
20090417 16:37:57-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 60 (Operation timed out)]
20090417 16:38:16-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 16:47:10-!- Reisiger [n=Reisiger@adsl-62-167-41-96.adslplus.ch] has joined #wesnoth-dev
20090417 16:48:51-!- EdB [n=edb@119.117.88-79.rev.gaoland.net] has joined #wesnoth-dev
20090417 16:52:45-!- Gnutoo [n=gnutoo@host13-11-dynamic.31-79-r.retail.telecomitalia.it] has joined #wesnoth-dev
20090417 16:56:39-!- ikarius [n=ross@smtp.gridironsystems.com] has joined #wesnoth-dev
20090417 17:01:58< boucman> dfranke: ping me when you're around
20090417 17:07:46-!- Noyga [n=lame-z@wesnoth/developer/noyga] has joined #wesnoth-dev
20090417 17:08:00-!- Noyga [n=lame-z@wesnoth/developer/noyga] has quit [Client Quit]
20090417 17:08:53-!- turin [n=turin@168.215.250.1] has joined #wesnoth-dev
20090417 17:16:42-!- ettin [n=jorda@wesnoth/developer/ettin] has joined #wesnoth-dev
20090417 17:31:48-!- EdB [n=edb@119.117.88-79.rev.gaoland.net] has quit [Remote closed the connection]
20090417 17:39:02-!- Sirp [n=me@wesnoth/developer/dave] has joined #wesnoth-dev
20090417 17:46:21-!- Psyche^ [n=Psyche@f054003208.adsl.alicedsl.de] has joined #wesnoth-dev
20090417 17:47:47-!- Patterner [n=Psyche@d137031.adsl.hansenet.de] has quit [Read error: 60 (Operation timed out)]
20090417 17:47:47-!- Psyche^ is now known as Patterner
20090417 17:49:16-!- YogiHH [i=c3f5f398@wesnoth/developer/yogihh] has left #wesnoth-dev []
20090417 18:02:36< boucman> lumiere
20090417 18:02:43< boucman> oops wrong window
20090417 18:03:56-!- noy [n=Noy@d75-157-52-251.bchsia.telus.net] has joined #wesnoth-dev
20090417 18:08:18< CIA-30> loonycyborg * r34984 /trunk/scons/install.py: scons wasn't installing locales and translations if those dirs didn't exist at SConscript parse time. Fixed.
20090417 18:09:12-!- wesbot changed the topic of #wesnoth-dev to: Summer of Code info: http://www.wesnoth.org/wiki/SummerOfCodeIdeas | 50 bugs, 234 feature requests, 11 patches | logs: http://irclogs.wesnoth.org/ | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20090417 18:23:48-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection reset by peer]
20090417 18:24:03-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 18:26:50-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090417 18:27:27< Shadow_Master> let's see what happens if I manage forum mods with git :p
20090417 18:28:16< Soliton> seems like a good idea especially when you plan to do local modifications.
20090417 18:36:44-!- Blueblaze is now known as Blueblaze2
20090417 18:37:17< Shadow_Master> yeah, why didn't I think of it before?
20090417 18:37:21-!- Blueblaze2 is now known as Blueblaze
20090417 18:38:56-!- Blueblaze is now known as Blueblaze2
20090417 18:39:18-!- Blueblaze2 is now known as Blueblaze
20090417 18:52:03-!- Sapient [n=patrickp@wesnoth/developer/sapient] has joined #wesnoth-dev
20090417 18:52:10-!- alink [n=alink@wesnoth/developer/alink] has joined #wesnoth-dev
20090417 18:58:21-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has quit [Remote closed the connection]
20090417 18:58:22-!- mordante [n=mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20090417 18:58:37< mordante> hi
20090417 18:58:51< Crab_> hi mordante
20090417 18:58:57< mordante> hi Crab_ 
20090417 19:00:21< Crab_> mordante: a few days ago we talked about "disabling  -Wno-unused by default". What's the perspective of that being done ?
20090417 19:01:00< mordante> Crab_, we didn't we talked about -Wno-sign-compare ;-)
20090417 19:01:23< mordante> I want to ask Sirp but haven't done so
20090417 19:01:26< Shadow_Master> seriously, why didn't I think of managing phpBB3 modifications with git before?
20090417 19:01:35< Crab_> mordante: ok)
20090417 19:01:45< Shadow_Master> .... later I can just use patch and diff on non versioned trees  o.O
20090417 19:02:04< mordante> but no unused params can be handy as well
20090417 19:02:20 * mordante goes to fix another compiler warning ;-)
20090417 19:02:30< Shadow_Master> why didn't cycholka think about it either?
20090417 19:02:31< Crab_> mordante: yes. since this warning can be fixed by /**/ really unused parameters...
20090417 19:02:57< Shadow_Master> oh wait, right, he _didn't think about it_, therefore I have a bonus for delayed cluefulness
20090417 19:03:49< mordante> Crab_, it can also be that the parameter is really no longer needed at all
20090417 19:04:11< Crab_> mordante: yes, and then it can be removed 
20090417 19:04:22< mordante> yup
20090417 19:08:39-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 19:10:56< boucman> mordante: hello
20090417 19:11:05< boucman> mordante: I have assigned a patch by corn for you
20090417 19:11:16< mordante> hi boucman I just read it in the log
20090417 19:11:21< boucman> oh, ok
20090417 19:11:45< boucman> Crab_: any new big debate regarding AI ? :P
20090417 19:12:31-!- stikonas [n=quassel@ctv-213-164-120-170.vinita.lt] has joined #wesnoth-dev
20090417 19:12:35< mordante> however I'm not sure whether I want to see the patch it's the old code ;-)
20090417 19:12:41< Crab_> boucman: not yet. now I'm working on making "AI movement" function that will reuse the replay/human movement code (which is in actions.cpp)
20090417 19:13:24< boucman> mordante: well, the patch is clean, the question I have is mainly about design, and since it has impact on a feature you might have to handle in the new gui too... you might as well have a look
20090417 19:13:42< boucman> Crab_: a worthy adventure..
20090417 19:14:15< mordante> boucman, I'll have a look, but I had a look at that code not too long ago and it feels very c-ish
20090417 19:14:54< Crab_> boucman: well, It will only require adding additional pointer parameter to that function (which will point to an object which will "collect" info for the ai)
20090417 19:16:16-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection timed out]
20090417 19:16:24< Crab_> boucman: basically, an analog for "display" object. events during the movement (ambushes, units seen, and such) are shown to human on display. but the ai can't read, so I need to tell it 'what was the result of the move' by using a special object which will contain all that info.
20090417 19:16:54< boucman> yeah, still your ai_result structure thing...
20090417 19:17:34< Crab_> boucman: I've already done a new version of ai_result 'recruit unit' routine, so I need to add three more - movement, attack, and setting unit movement/attacks to zero (a way for the ai to say 'i've done with this unit')
20090417 19:18:13< boucman> yes, recruiting was probably the simplest
20090417 19:18:37< Crab_> boucman: 'setting movement to zero' is simpler, of course ) but the recruitment is simpler than attacking and movement.
20090417 19:19:41< Crab_> boucman: and, I was thinking about combining default ai with formula ai. there's two approaches: 1) "let default ai phases be a candidate action  2) "let evaluate-candidate-actions routine be a default ai phase.
20090417 19:20:35< boucman> Crab_: my knowledge of default_ai is quite rudimentary
20090417 19:20:54< CIA-30> mordante * r34985 /trunk/src/ai/ai_manager.cpp: Fix a compiler warning.
20090417 19:21:17< mordante> Crab_, I guess you love that^ one  ;-)
20090417 19:21:34< Crab_> mordante: I'll take a look )
20090417 19:22:25< Crab_> boucman: see http://wesnoth.pastebin.com/m250714b7
20090417 19:22:56< Crab_> boucman: very simplified explanation of current situations and two ways to combine
20090417 19:23:40< boucman> I personally find approch one better
20090417 19:23:58< boucman> it gives us a scale in which we can insert our own actions
20090417 19:25:07< Crab_> mordante: thanks :) that's why I want to add "variable is unused" warning as default ) (I'll remove that parameter a bit later, it is no longer needed there)
20090417 19:26:11< Crab_> boucman: to do (1), we need to assign a 'evaluation score' to default_ai phase
20090417 19:26:18< mordante> I don't feel to strong about that warning, but the sign comparison a bit stronger
20090417 19:26:36< Crab_> boucman: noting that default ai phase can consist of several moves.
20090417 19:27:23< boucman> well, if you could have it be one move it would be cool
20090417 19:27:36< boucman> as for assigning value to each AI phase, it doesn't seem a bad idea to me
20090417 19:28:06< boucman> we need to start having a scale somewhere anyway, and we intend to phase out c++ ai, that's a good way to do it...
20090417 19:28:13-!- noy [n=Noy@wesnoth/developer/noy] has quit []
20090417 19:28:38< Crab_> boucman: well, we do not have t 'phase out' the c++ ai
20090417 19:28:39< dfranke> boucman: pong, but I just have a couple minutes.
20090417 19:29:11< boucman> dfranke: did you commit a patch from corn regarding column sorting ?
20090417 19:29:16< dfranke> boucman: yes
20090417 19:29:27< boucman> ok, you forgot to mark it as done, I'll do that
20090417 19:29:35< Crab_> boucman: consider this: if we allow to 'expose a c++ routine that proposes an action as a 'candidate action', then we can have those routines written in fai,c++,lua, etc. - it won't matter to the candidate action evaluator
20090417 19:29:44< dfranke> boucman: oh... that's because it didn't exist there when I committed it.  I took the attachment from the bug.
20090417 19:30:10< boucman> Crab_: yes, that's another plan
20090417 19:30:12< Crab_> boucman: so, we won't "phase out" c++ ai, we'll just make ai centered about a c++-based candidate move evaluator
20090417 19:30:21< dfranke> boucman: also, mordante has seen it.  We discussed it in #wesnoth-mentor
20090417 19:30:27< boucman> k
20090417 19:30:35< dfranke> bbiab
20090417 19:30:46< boucman> Crab_: ok, make sense
20090417 19:31:06< Crab_> boucman: it is possible to do. do do this without changing default ai behaviour, I'll modify its phases to return an ai_action_result (not executed, just "checked" for possibility)
20090417 19:31:08-!- stikonas [n=quassel@wesnoth/translator/stikonas] has quit [Read error: 104 (Connection reset by peer)]
20090417 19:31:48< boucman> yes, we have to cut default AI in chunks and wrap the chunks in RCA
20090417 19:32:14< Crab_> boucman: yes, then those chunks can be wrapped in RCA, you're right.
20090417 19:32:42< boucman> Crab_: ok, that sounds like a plan
20090417 19:32:49< Crab_> boucman: then the RCA evaluator will be 'ai-independent' of sorts.
20090417 19:32:59< boucman> be carefull though, you need a way to tell the RCA "no move left of that type"
20090417 19:33:19< Crab_> boucman: return score<0  ?
20090417 19:33:35< boucman> yes, but that's mean the score evaluation has to be pluged in C++ too...
20090417 19:34:09< Crab_> boucman: yes. it can be hardcoded, though. no need to be fancy in the beginning.
20090417 19:34:11-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 104 (Connection reset by peer)]
20090417 19:34:26-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 19:34:57< Crab_> boucman: to be able to bug-for-bug replicate default_ai, we only need hardcoded scores - to establish relative priorities of each default_ai phases
20090417 19:35:37< Crab_> so, '-1 if nothing to do, hardcoded value otherwise' will be ok for the beginning
20090417 19:36:08< Crab_> boucman: and, there's (imho) not a reason to allow only one "action" per candidate action.
20090417 19:36:18< boucman> Crab_: yes, but the "if nothing to do" part was the one I was wondering about
20090417 19:36:24< Crab_> boucman: return -1
20090417 19:36:36< boucman> nonononono
20090417 19:36:42< Crab_> boucman: why ?
20090417 19:37:26< boucman> I mean : "we have to have a way for the c++ AI to tell the scoring part of RCA that there are no more actions of that type to do" returning -1 is fine, it's the comunication with scoring I was worried about
20090417 19:40:28< Crab_> I was thinking about http://wesnoth.pastebin.com/m40aca110 
20090417 19:42:06< boucman> Crab_: so you actually do only one call at evaluation time, which returns the move (or no move at all) and you memorize the move and execute it if/when we are called later
20090417 19:42:10< boucman> yes, that could work
20090417 19:42:13< Crab_> so, from the point of view of candidate action evaluator, it will be just a candidate action. it'll evaluate it, get a score (HARDCODED_VILLAGE_GRABBING_PRIORITY or -1), and then, after sorting all evaluated actions, it'll pick the best
20090417 19:42:31< Crab_> boucman: yes.  "execute it if/when we are called later"
20090417 19:42:45< Crab_> boucman: ai_action_result supports this usage, btw
20090417 19:42:59< Crab_> since it wraps all action parameters at construction time
20090417 19:43:34< Crab_> and it can be executed later (provided that the locations are unchanged, and they'll be unchanged if all candidate actions have no side effect during evaluation)
20090417 19:43:35< boucman> yes, indeed
20090417 19:44:20< Crab_> boucman: and so, we'll be able to replicate default_ai as an RCA-based ai
20090417 19:44:31< Crab_> boucman: and then we'll test it against default_ai
20090417 19:45:05< Crab_> boucman: that is, if we seriously break something during this transformation, then default ai will win more against our RCA-based default ai
20090417 19:45:20< Crab_> boucman: and since they're supposed to be equivalent, we'll get the message ))
20090417 19:45:45-!- ABCD_away [n=ABCD@wikipedia/ABCD] has quit [Read error: 113 (No route to host)]
20090417 19:45:52< Crab_> and then we can start augmenting that RCA-based default ai with formulas
20090417 19:45:56< boucman> Crab_: makes sense
20090417 19:46:26< Crab_> and each ai configuration will be stored in a .cfg file
20090417 19:46:48< Crab_> and if someone commits that karl_m's patch for --ai_configN=.. , we'll be even able to test those ai's in a batch mode )
20090417 19:47:57< Crab_> but before that transformation of default_ai int RCA-based default ai, i need to encapsulate all game-state-changing calls in default_ai.
20090417 19:48:00-!- noy [n=noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090417 19:48:12-!- elias [n=allefant@allegro/developer/allefant] has quit ["Leaving"]
20090417 19:48:20< Crab_> and that is what I'm doing now
20090417 19:51:52< boucman> Crab_: i'll let you commit the patch if that's fine with you...
20090417 19:52:12< Crab_> boucman: ok
20090417 19:52:21< boucman> ok, commented and reassigned
20090417 19:53:06< Crab_> thanks
20090417 19:56:42< Crab_> boucman: and note that candidate move evaluator doesn't need to *understand* the action that he will do.
20090417 19:57:05< boucman> Crab_: yes, that was part of the original idea
20090417 19:57:20-!- noy [n=noy@wesnoth/developer/noy] has quit ["Get Colloquy for iPhone! http://mobile.colloquy.info/"]
20090417 19:57:23< Crab_> boucman: since all that 'candidate action evaluator' needs is a score and a 'trigger to push'.
20090417 19:57:49< Crab_> boucman: so, it's really not a problem if candidate action wishes to play the entire turn :)
20090417 19:58:07< Crab_> boucman: it doesn't matter to candidate action evaluator.
20090417 19:58:49< Crab_> (that is regarding the fact that some default_ai routines do several actions at a time -it won't stop our plans)
20090417 19:59:45< boucman> Crab_: you can only do multiple actions at a time if you are sure that they will all succeed, ( at least at this point) 
20090417 19:59:56< boucman> or else there mught be some nasty suprises
20090417 20:00:02-!- noy [n=noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090417 20:00:57< Crab_> boucman: no, this won't be a problem...because even if you do one action, you're not sure if it succeeds or not
20090417 20:01:27< Crab_> boucman: so, if current code tries to disagree with that, it will be changed )
20090417 20:02:51< Crab_> boucman: since you can imagine candidate action as a "entity who wants to play for this side a bit, and tells about 'how good it will be', and candidate action evaluator picks one of the candidate actions and lets them 'play a bit'
20090417 20:03:08< Crab_> s/them/it
20090417 20:03:55< boucman> Crab_: we currently support an evaluator returning a set of actions, in which case we play them all, but it might be a good idea to stop that whenever an action is not successfull (think ambush)
20090417 20:04:04< boucman> or the whole set my go crazy
20090417 20:04:20< Crab_> boucman: imo, evaluator needs to return one action
20090417 20:04:44< boucman> Crab_: there are some cases where it could be a big limitation
20090417 20:04:55< boucman> for instance leader move+unit attack
20090417 20:05:01< Crab_> boucman: it is not. this one action can be a wrapper around a 'set of actions done in sequence' or around 'play the entire turn'
20090417 20:05:04< boucman> it's hard to code that through two formulas
20090417 20:05:18< boucman> i'm not sure what you mean
20090417 20:05:32< Crab_> from the point of view of the candidate action evaluator, it should be *one* action.
20090417 20:05:58< Crab_> even if it is a set of actions, an entire ai play_turn routine, or something else
20090417 20:06:22< boucman> Crab_: ok, we disagree on semantic, but not on what we mean
20090417 20:07:38< Crab_> and I agree that when an a 'candidate action' wants to execute something, it *should* return control to the evaluate action evaluator in a case when it is not sure what to do, for example, if one of underlying actions fails or is interrupted.
20090417 20:08:35< Crab_> basically "evaluator chooses and grants control to candidate action" -> "candidate action plays till it has done all that it wanted, or until something unexpected by it happens" -> "control is returned to evaluator"
20090417 20:09:04< Crab_> what is important is the interface
20090417 20:09:40< Crab_> example implementation can be "get a set of actions, do them all one-by-one, but stop and return control if any fails"
20090417 20:10:15< boucman> Crab_: yes, makes sense, the AI is slightly lower right now, since the evaluators don't "do the action" but "return the action to the caller, which will execute it" but it's a slight sliding in semantic that might be worth doing... but later
20090417 20:10:37< Crab_> they will still  "return the action to the caller, which will execute it"
20090417 20:10:45< Crab_> but an "execute" should be a virtual method
20090417 20:11:11< Crab_> so, the action itself has to decide "how-it-wants-to-be-execute-itself"
20090417 20:12:02< Crab_> imagine a crowd of people standing around a computer running wesnoth. one of the people is the boss - he is called "candidate move evaluator", and he sits near a computer.
20090417 20:12:20< Crab_> the boss ask all those people in a crowd : "here's the situation in the game, who wants to play?"
20090417 20:12:36< boucman> Crab_: that's fine, I completely understood your idea, no worry,
20090417 20:12:52< boucman> I was just assessing the impact of the change, but it's ok, it's only a small change
20090417 20:13:04< Crab_> all the people, one by one, say "I want to play, it will be +20 good",  another "I want to play, it will be +30 good", ... and the boss picks one of them and says : sit down and play till you're satisfied.
20090417 20:13:18< boucman> though I think integrating default_ai in the RCA is not something to tackle right away, AI testing should come first
20090417 20:13:19< Crab_> so, as you see, it is not the boss who is actually playing
20090417 20:13:37-!- noy [n=noy@wesnoth/developer/noy] has quit ["Get Colloquy for iPhone! http://mobile.colloquy.info/"]
20090417 20:13:47-!- noy [n=noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090417 20:14:08-!- noy [n=noy@wesnoth/developer/noy] has quit [Remote closed the connection]
20090417 20:14:20< Crab_> boucman: testing is "parallel" to this project so far (since I don't touch default ai so far)
20090417 20:14:40< Crab_> boucman: I "add" to it, but i do not yet use it
20090417 20:14:53< boucman> Crab_: I was more worried about disturbing other people working in that area (DK) and loosing focus on what is important
20090417 20:16:00< Crab_> boucman: "dividing up work with DK" will be after 20th, when the cards will open and we'll see who (none of us, one of us, both of us) are accepted.
20090417 20:16:25< boucman> Crab_: no, that's not how it works
20090417 20:16:46< Crab_> boucman: then tell me how it should work
20090417 20:16:49< boucman> whether you stay for SoC or not, you are currently a wesnoth dev, it comes with all the powers etc...
20090417 20:17:08< boucman> and that includes looking at other devs and what they are doing, in that case DK
20090417 20:17:27< Crab_> boucman: yes. but I'm sure that won't get to DKs area before 20th
20090417 20:17:45< boucman> now that's just minor, don't worry about it, but currently DK is working on RCA, so just wait a bit or synchronize with him if you want to work on that
20090417 20:17:53< boucman> ok, fine then
20090417 20:17:54< Crab_> boucman: so, I'm sure that my current work (making ai const-correct in regard to game state) won't impact his
20090417 20:18:03< AnMaster> what does it mean when wesnoth spits message like this to stderr:
20090417 20:18:05< AnMaster> (process:16366): Pango-WARNING **: pango_layout_set_markup_with_accel: Error on line 3 char 4: '1' is not a valid character following a '<' character; it may not begin an element name
20090417 20:18:19< AnMaster> noticed them when playing IftU
20090417 20:19:51< Crab_> boucman: and, i'm also sure that, no matter how 'candidate-action-evaluator' will look like, it will have candidate actions with an (implicit or explicit) interface of "get_score" and "execute", so my plan of 'integrating both default_ai and formula ai under the same roof' will work, i'll just have to discuss and adapt.
20090417 20:20:12< Soliton> accidental (invalid) pango markup somewhere i guess.
20090417 20:20:15-!- maxy [n=maxy@80-219-0-199.dclient.hispeed.ch] has joined #wesnoth-dev
20090417 20:20:26-!- noy [n=noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090417 20:20:39< boucman> Crab_: yes, I have good knowledge of RCA, and I am not too worried about that, it should go fine
20090417 20:21:29< Crab_> boucman: so, basically, i'm doing two projects now - "testing" and "ai refactoring without ai behavior changes", and they're parallel
20090417 20:21:56< boucman> Crab_: ok, that's fine
20090417 20:22:54< AnMaster> Soliton, ah
20090417 20:23:33< Crab_> and with testing, karl_m has done a good thing in creating this --ai_config option (we discussed it with him a week ago), and we've talked with him about doing a 'batch testing framework' on the basic on the same tests which are used in ai_arena_small ''
20090417 20:24:25< Crab_> boucman: since I've used lua to 'extract' each test from ai_arena_small, we can re-use it from another scenario (for example, without a human observer), which will be run from a command-line
20090417 20:24:50< Crab_> boucman: and since most tests do not involve leaders, we can do multiple tests without reloading the scenario itself.
20090417 20:29:45< Crab_> boucman: and even if the tests involve leaders, we can "intercept" the killing blow (as it is done in DM), and just 'increment the win counter' and 'go to next test'
20090417 20:38:30< Crab_> boucman: and about 'loosing focus on what is important'. can you specify 'what is important' in our ai changes, from your point-of-view ?
20090417 20:39:30< boucman> Crab_: I was more worried about starting coding on all sort of stuff instead of doing changes one at a time...
20090417 20:39:42< boucman> not about the order of stuff itself, 
20090417 20:41:03< CIA-30> ivanovic * r34986 /trunk/po/ (4 files in 4 dirs): updated Chinese (simplified and traditional) translations
20090417 20:41:07< CIA-30> ivanovic * r34987 /branches/1.6/po/ (4 files in 4 dirs): updated Chinese (simplified and traditional) translations
20090417 20:42:16-!- noy [n=noy@wesnoth/developer/noy] has quit ["Get Colloquy for iPhone! http://mobile.colloquy.info/"]
20090417 20:43:39-!- Blueblaze [n=nick@c-98-199-143-139.hsd1.tx.comcast.net] has joined #wesnoth-dev
20090417 20:43:51< Crab_> boucman: don't worry, it's quite structured, and will be done gradually. I should maybe write a dependency graph to make the sequence clearer
20090417 20:44:07< boucman> nah that's fine...
20090417 20:44:17< boucman> you 
20090417 20:44:23< boucman> nvm
20090417 20:44:47< Crab_> boucman: but, basically, I've started-from-the-bottom and want to make ai code modular and structured, to be able to insert things I want later (such as ai perception, etc)
20090417 20:49:47< Crab_> (afk ~20 minutes)
20090417 20:51:15-!- noy [n=noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090417 20:51:27< mordante> Ivanovic, do you know what the port number of the addon server should be, sources contain both 15003 and 15004
20090417 20:52:27< Soliton> 15003 is 1.6 and 15004 is trunk.
20090417 20:52:29-!- noy [n=noy@wesnoth/developer/noy] has quit [Client Quit]
20090417 20:52:44-!- noy [n=noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090417 20:53:08< mordante> ok thanks I was looking for trunk, the commit for the server looked suspicious
20090417 20:54:39-!- noy [n=noy@wesnoth/developer/noy] has quit [Client Quit]
20090417 20:55:16-!- noy [n=noy@wesnoth/developer/noy] has joined #wesnoth-dev
20090417 21:00:55-!- maxy [n=maxy@80-219-0-199.dclient.hispeed.ch] has quit [Remote closed the connection]
20090417 21:07:46-!- Elvish_Pillager [n=eli@66-189-40-122.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20090417 21:11:27< CIA-30> crab * r34988 /trunk/ (5 files in 4 dirs): 
20090417 21:11:27< CIA-30> AI Testing. Adds a command line parameter --ai_config<number>=value (patch #1164
20090417 21:11:27< CIA-30> by karlm). This patch allow to specify the exact path to ai configuration for
20090417 21:11:27< CIA-30> each of the sides from the command line, which is very important for
20090417 21:11:27< CIA-30> batch-testing of various AIs.
20090417 21:12:10< Crab_> karl_m: ^^^ congratulations and thanks.
20090417 21:15:55< AnMaster> 20090417 21:13:06 error display: could not open image 'maps/mainmenu.png' <-- err happened when I was watching credits after finishing IftU. And a black/yellow image of someone digging showed up. Resized to fill window size (1374x976).
20090417 21:16:45< boucman> I think mainmenu.pnh has been moved
20090417 21:16:58< boucman> IftU probably needs an upgrade, you should contact the author
20090417 21:17:03< AnMaster> ah
20090417 21:17:15< AnMaster> wasn't sure if credits was handled in core or in scenario
20090417 21:17:23< boucman> both
20090417 21:17:28< AnMaster> I see.
20090417 21:17:30< Blueblaze> Sapient: Is there any way to delete a [set_menu_item]? If not, can you completely override an old one?
20090417 21:17:30< boucman> iftu specific parts are in iftu
20090417 21:17:42< alink> AnMaster: so, you play in debug mode, hehe ;)
20090417 21:17:54< Sapient> Blueblaze: you can change the filters so that it never shows
20090417 21:17:58< Sapient> this is how to delete it
20090417 21:18:10< Sapient> [show_if] [not]
20090417 21:18:10< AnMaster> alink, I did because I was checking why after credits I was dumped back into level before last
20090417 21:18:37< AnMaster> was using -d --log-debug=all
20090417 21:18:51< AnMaster> which is the standard line I type when something goes very odd
20090417 21:19:00< alink> (the black/yellow image is only visible in debug mode, normal users have a black image when it's missing)
20090417 21:19:23< alink> but glad that this debug feature really helps to spot missing image
20090417 21:19:45< AnMaster> alink, aha. But still pretty odd I got dumped back into last turn of scenario before epilogue after the credits
20090417 21:19:56< Blueblaze> Sapient: Ah, is there any real danger of it taking too much memory or slowing down other processes if you have too many present?
20090417 21:20:14< Sapient> nope
20090417 21:21:16< AnMaster> alink, maybe that black/yellow one should always be used, since obviously no images should be missing, so users shouldn't ever see it anyway
20090417 21:21:18< alink> AnMaster: maybe the credits code doesn't handle well missing image. It does some stretching etc. So with no image dimensions could probably cause problems if not watched
20090417 21:21:18< CIA-30> soliton * r34989 /branches/1.6/src/server/game.cpp: small tweak for the case of an empty history
20090417 21:21:20< Blueblaze> Sapient: Even if you have, let's say 100 disabled?
20090417 21:21:40< Sapient> Blueblaze: it would need to be hundred of thousands
20090417 21:22:19< alink> AnMaster: not sure, image path are often broken, if this doesn't cause visible glitch, better don't annoy the user with it
20090417 21:22:34< AnMaster> alink, but that explains the black "blinks" I have seen when some units attack in some campaigns. Missing images I guess.
20090417 21:22:50< Sapient> Blueblaze: there's really no need to create that many, since you can just re-use a previous ID and set the new command
20090417 21:22:58< AnMaster> alas I don't remember *which* campaign it was
20090417 21:23:21< Blueblaze> Sapient: Can you override it though? I tried to do that and it just broke the entire thing :x
20090417 21:23:52< alink> AnMaster: mmh, no, I was imprecise, missing image don't cause black image. It doesn't  draw anything. I said black background because the credits use that behind the image
20090417 21:23:57< Sapient> [set_menu_item] will set the menu item
20090417 21:24:18< CIA-30> soliton * r34990 /trunk/src/server/game.cpp: small tweak for the case of an empty history
20090417 21:24:19< Sapient> kinda goes without saying, I thought
20090417 21:24:21< AnMaster> alink, no, the normal map was there
20090417 21:24:25< AnMaster> this happened after a while
20090417 21:24:29< AnMaster> and was put on top of it
20090417 21:24:36-!- thespaceinvader [n=chatzill@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20090417 21:24:46< AnMaster> as in: for once I was actually reading the credits
20090417 21:25:08< alink> AnMaster: credits have a list of images and cycle them 
20090417 21:25:11< AnMaster> ah
20090417 21:25:36< Blueblaze> Sapient: Indeed, but it just meshed the two items together and broke it :|
20090417 21:25:39< Blueblaze> I'll go and play on it some more
20090417 21:25:43< AnMaster> alink, never looked at credits long enough before to notice that :D
20090417 21:25:49-!- noy [n=noy@wesnoth/developer/noy] has quit ["Get Colloquy for iPhone! http://mobile.colloquy.info/"]
20090417 21:25:51< alink> that's funny, i just saw the commit introducing this today
20090417 21:26:03< Sapient> Blueblaze: you have to set the filter parts of it too, if you want them to change
20090417 21:26:12-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 60 (Operation timed out)]
20090417 21:26:16< Sapient> any aspect that is missing will not be "set"
20090417 21:26:25< Sapient> and will thus remain the way it was
20090417 21:26:35< alink> wesbot log 10715
20090417 21:26:38< wesbot> darthfool * r10715 : Allow multiple background images in about, customizable by campaign, and put in a white drop-shadow on the text.
20090417 21:26:41< wesbot> URL: http://svn.gna.org/viewcvs/wesnoth?view=rev&rev=10715
20090417 21:27:14< Blueblaze> Sapient: Well, the filter conditions are the same (though I did declare them again in the second occurance of the [set_menu_item])
20090417 21:27:30< alink> AnMaster: so that feature is there since march 2006 ;)
20090417 21:27:34< CIA-30> soliton * r34991 /trunk/src/ (10 files): applied patch #1168: move delay shroud update to game display
20090417 21:27:44-!- fendrin [n=fabi@wesnoth/developer/fendrin] has joined #wesnoth-dev
20090417 21:27:59< AnMaster> alink, heh
20090417 21:28:23< AnMaster> alink, well I'm usually more interested in playing than reading credits
20090417 21:28:25< Blueblaze> Would that cause a problem that the filter conditions are the same?
20090417 21:30:30< alink> AnMaster: do I add a vertical shoot 'em up into the scrolling credits ? <:o)
20090417 21:31:06< AnMaster> alink, hah
20090417 21:31:14< AnMaster> nah
20090417 21:31:26< Soliton> boucman: sorry hadn't seen your comment before.
20090417 21:31:36< crimson_penguin> alink: that'd be awesome :D
20090417 21:31:37< AnMaster> but maybe make it into a map? With units representing the different people
20090417 21:31:37< boucman> Soliton: what comment ?
20090417 21:31:46< AnMaster> grouped by sides of programmers, translators and such
20090417 21:31:52< Soliton> boucman: for the patch i just committed.
20090417 21:31:52< AnMaster> and some auto scrolling thing over it
20090417 21:32:06< AnMaster> alink, that could be interesting :)
20090417 21:32:14< boucman> Soliton: :P
20090417 21:32:25< boucman> well, you're up for a revert I guess :)
20090417 21:32:41< Soliton> you really think your argument is any realistic?
20090417 21:33:25< Soliton> i made a comment with a workaround if there are really people that want that.
20090417 21:33:29< boucman> ??
20090417 21:33:31< AnMaster> alink, anyway there are better things to spend time on in wesnoth I think :D
20090417 21:33:54-!- voris [n=voris@c-24-16-107-149.hsd1.wa.comcast.net] has joined #wesnoth-dev
20090417 21:34:13< boucman> Soliton: in that case we should remove local games entirely
20090417 21:34:13-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 21:34:14< AnMaster> what is "delay shroud update"? I think I saw it in some menu
20090417 21:34:17< Soliton> boucman: who's playing hotseat (to begin with?) and wants different shroud delay settings for each side...
20090417 21:34:22< AnMaster> never bothered checking what it did
20090417 21:34:30< alink> AnMaster: note that there is already keys to control this scrolling. I even started a patch to allow variable speed and direction, but i dropped the idea
20090417 21:34:37< boucman> people which have only access to one computer and can't run a server
20090417 21:34:44< AnMaster> alink, yes I know you can control scrolling speed
20090417 21:34:48< boucman> (like me and my collegues on a plane to the us)
20090417 21:35:09< boucman> as for different settings, it's a matter of taste (or else we wouldn't have the option)
20090417 21:35:12< Soliton> and you'd be annoyed if you can not use different shroud delay settings?
20090417 21:35:21< alink> AnMaster: use "delay shroud update" and explore fog/shroud you will understand
20090417 21:35:56< AnMaster> alink, mhm. Since I just finished IftU (what a marathon of several days of playing) I think I will take a pause from wesnoth for a few days now.
20090417 21:35:59< boucman> Soliton: do as you please, I don't care much
20090417 21:36:18< boucman> but we'll get the bug "changing seting for player A also changes it for player B eventually
20090417 21:36:19< voris> Soliton, I would be.
20090417 21:36:30< boucman> that's not the kind of UI design one expects
20090417 21:36:57< boucman> maybe move the delay shroud update as a persistant seting with all the other ones would make more sense
20090417 21:37:07< voris> I play hotseat often and like to delay shroud but my gf hates it because she always forgets to update it as she moves. Me, I prefer to be able to undo because I'm a sloppy clicker. :)
20090417 21:37:15< Soliton> that's clearly orthogonal.
20090417 21:38:50< Soliton> well, then implement the proposed improvement in the patch but that is what i think is bad design.
20090417 21:39:49< Soliton> i guess it is people who play hotseat with several people vs someone who plays a local game controlling several sides.
20090417 21:41:22< Soliton> i don't really see a good way to integrate both.
20090417 21:42:12< voris> Preferences dialog level setting to control what, by default, is set up for any unit created on this computer and context menu to toggle it per unit.
20090417 21:43:15< Soliton> well, i guess i'll revert and wait for a persistent setting.
20090417 21:43:40< Soliton> since i suppose what annoys me is when i save-load and have to set it again for all sides.
20090417 21:43:54< voris> That is annoying, I agree.
20090417 21:44:02< voris> Persistence would be nice.
20090417 21:44:38< alink> mmh a visual indication to what mode you are could help both cases. A special move cursor when you use delay shroud update is very easy to do and seems useful
20090417 21:45:02 * alink thinks that he will try that soon
20090417 21:46:00< CIA-30> soliton * r34992 /trunk/src/ (10 files): 
20090417 21:46:00< CIA-30> Revert "applied patch #1168: move delay shroud update to game display"
20090417 21:46:00< CIA-30> There are people that actually want different settings per side
20090417 21:47:15< alink> alternatively a visual indication that fog or shroud tiles are "frozen", but that's more complicated and intrusive
20090417 21:47:36-!- Alesis-Novik [n=alesis@78-61-128-45.static.zebra.lt] has joined #wesnoth-dev
20090417 21:47:47< boucman> alink: animated shroud ? :P
20090417 21:48:02< alink> lol
20090417 21:50:41-!- dfranke [i=dfranke@wesnoth/developer/dfranke] has quit [Connection timed out]
20090417 21:51:51-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 60 (Operation timed out)]
20090417 21:52:36-!- silene [n=plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20090417 21:52:43< silene> hi
20090417 21:52:49< alink> hi silene
20090417 21:58:11< alink> when I create a local MP game, i can use the combo to assign the same name/player to different sides, instead to let them be Local Player. Maybe use this to indicate that one control these sides and want to share his delay shroud option between them ?
20090417 21:58:40< Soliton> yeah, i thought about that too.
20090417 21:58:50-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 21:58:59< Soliton> not realy intuitive to understand though, maybe.
20090417 22:00:33< Soliton> what is the proper fix for https://gna.org/bugs/?13383 ?
20090417 22:00:43< alink> one way to force to use it, would be to change the default of the combo and use something else than Local Player
20090417 22:01:00< Soliton> i guess an "#include <bits/fcntl.h>" is missing but then also O_RDONLY should be unknown?
20090417 22:01:00< alink> (empty, AI, or my login)
20090417 22:01:38< Soliton> (i'd prefer a menu entry for each of those)
20090417 22:01:51< Soliton> (like we used to have)
20090417 22:02:06< Soliton> (well, except for empty probably)
20090417 22:02:28< alink> Soliton: i don''t remember that, it was not a combo ?
20090417 22:02:54< Soliton> no, there was not just local game but hotseat and vs ai.
20090417 22:03:09< alink> ah yes indeed
20090417 22:03:16< Soliton> depending on what you choose the default is set.
20090417 22:04:51< Soliton> vs AI, hotseat, play alone? (defaults: default AI, local player, login)
20090417 22:05:27< Soliton> maybe that should rather be a drop down menu in the create game dialog though.
20090417 22:06:43< silene> Soliton: the header is called <fcntl.h> only; that being said, i'm not even sure this is the issue, it may just be that his system is a bit old; it's a bit unfortunate he didn't write the version of his suse
20090417 22:07:17< Soliton> silene: 9.0
20090417 22:07:32< Soliton> he cleverly used the release field.
20090417 22:07:42< alink> Soliton: or replace the observer switch (useless for local game) by hotseat (multi-humans) switch
20090417 22:07:57< Soliton> silene: i was wondering about that, so a won't fix?
20090417 22:09:01< Soliton> silene: out of curiousity why is it only <fcntl.h>? (it's in /usr/include/bits/fcntl.h on my system)
20090417 22:09:07< alink> plus, about this side specific shroud update in hotseat, one can defend that accelerated speed has the same problem
20090417 22:10:09< alink> but i am afk now, bbl
20090417 22:10:19< silene> Soliton: because posix says so? (are you sure you haven't a /usr/include/fcnhl.h file?)
20090417 22:11:41< silene> Soliton: 9.0 looks pretty old, so it may just from a time NOATIME didn't exist
20090417 22:12:04< Soliton> silene: yeah, but that does not have O_NOATIME but i see it includes <bits/fcntl.h>.
20090417 22:13:14< Ivanovic> suse 9.0 is ancient IIRC
20090417 22:14:05< Soliton> do we have a build option to disable sendfile?
20090417 22:14:11< Ivanovic> it was released end of october 2003
20090417 22:14:35< Soliton> also when we advise users to use some build option what build system do we refer to? :-)
20090417 22:14:35< Ivanovic> kernel 2.4.21 as default, outsch...
20090417 22:14:52< Soliton> oh, i didn't think it's that old.
20090417 22:15:10< Ivanovic> regarding this list it is: http://de.wikipedia.org/wiki/SuSE_Linux#Versionen
20090417 22:16:17-!- Sapient [n=patrickp@wesnoth/developer/sapient] has left #wesnoth-dev []
20090417 22:16:22< Soliton> guess they are releaseing slower than debian.
20090417 22:16:44< silene> Soliton: no, it is automatically enabled if it is detected (i'm only talking for autoconf here, don't know what the other build systems do)
20090417 22:17:04< Soliton> ok, suppose i can just tell him to upgrade then.
20090417 22:17:52< Soliton> or just remove the NOATIME there, i doubt it's important.
20090417 22:19:12< silene> i suppose it is, otherwise noone would have thought about putting it there
20090417 22:20:15< Soliton> it looks just like a performance tweak to me. i wouldn't be surprised if that bit is just copied from somewhere else.
20090417 22:23:39< silene> haha, it's again a patch from suokko
20090417 22:24:05< silene> (when will that end...)
20090417 22:24:21-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Read error: 104 (Connection reset by peer)]
20090417 22:24:21< Ivanovic> is anyone suprised?
20090417 22:24:26 * Ivanovic looks around
20090417 22:24:31< loonycyborg> Soliton: sendfile check works in scons just like in autoconf. ie adds a define to config.h if sendfile is detected so this function will be used only if available..
20090417 22:24:35-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 22:25:16< Ivanovic> the same for cmake post 1.6...
20090417 22:25:23< Soliton> loonycyborg: ok, that makes sense of course. it's just a special case where disabling it would help.
20090417 22:27:26< loonycyborg> Perhaps add NOATIME check to buildsystems? :P
20090417 22:27:52< Soliton> i'd rather take a look what the point of using it there is.
20090417 22:32:08< Soliton> well, really i can only imagine it's a performance tweak so a system that honors atime doesn't do useless updates of it.
20090417 22:33:10-!- YogiHH [n=chatzill@c211231.adsl.hansenet.de] has joined #wesnoth-dev
20090417 22:34:41< loonycyborg> open() manpage says that O_NOATIME was added in Linux 2.6.8
20090417 22:35:45-!- Reisiger [n=Reisiger@adsl-62-167-41-96.adslplus.ch] has quit [Read error: 104 (Connection reset by peer)]
20090417 22:36:10< loonycyborg> Using it should matter only if it does a *lot* of read()s.
20090417 22:37:22< Soliton> it probably just does one.
20090417 22:42:04< silene> Soliton: if this code is used by the campaign server, noatime may make a noticeable difference if the filesystem is atime
20090417 22:43:25< Soliton> yes, certainly. that's the other thing that i doubt that guy wants to run a campaign server. :-)
20090417 22:44:21< Soliton> sendfile support was implemented for the campaign server but i can't tell if it isn't used elsewhere too.
20090417 22:46:22-!- dfranke [i=dfranke@wesnoth/developer/dfranke] has joined #wesnoth-dev
20090417 22:47:27< voris> alink: the difference with accelerated speed is you don't get stuck in a "I can't undo" moment if it's not what you expected. You just, at worst, miss out on some animation display and then argue it out with your mate what speed you can both stand. :)
20090417 22:49:04< Soliton> pfff, that just means you turn delayed shroud on if either of the players wants it. :-)
20090417 22:49:12-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 22:53:08< YogiHH> corn: around?
20090417 22:53:36-!- noy [n=Noy@70.70.224.183] has joined #wesnoth-dev
20090417 22:55:55-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection timed out]
20090417 23:03:08< YogiHH> afk
20090417 23:11:48-!- happygrue [n=quassel@c-67-176-145-41.hsd1.in.comcast.net] has joined #wesnoth-dev
20090417 23:12:01-!- Espreon [n=espreon@wesnoth/developer/espreon] has joined #wesnoth-dev
20090417 23:16:49-!- busfahrer [n=busfahre@unixboard/user/busfahrer] has joined #wesnoth-dev
20090417 23:20:52< voris> Soliton: sure, that's what I do now. But if changing for one changed for all I'd have to change back every time. That's what I understood was being suggested...
20090417 23:22:51< Soliton> voris: i mean if you can compromise on the animation speed setting you can compromise on the delayed shroud setting.
20090417 23:26:49< voris> I will admit to being willing to deal with no shroud delay if I were forced to conform to a game-wide setting instead of per-player.
20090417 23:27:24< voris> Which is what I'd do if that's how it were set up. My gf just cannot deal with the delay option being set at all. :)
20090417 23:33:49-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
20090417 23:36:17< mordante> Sirp, I've been talking with Crab_ about the -Wno-sign-compare switch in trunk and we like to remove that option and make sign-compares fatal as well
20090417 23:36:46< mordante> Sirp, trunk is clean of those warnings at the moment and I was wondering what your opinion is
20090417 23:37:27< loonycyborg> Should -Wno-unused be removed too?
20090417 23:38:07< mordante> I feel less strong about that one, since it hides less bugs and people might get annoyed by ut
20090417 23:38:09< mordante> it*
20090417 23:38:24< mordante> but I have both of them as fatal
20090417 23:40:00< Crab_> mordante, loonycyborg: imo, 'make the compiler catch everything that is possibly not nice/wanted', is a good idea.
20090417 23:41:13< mordante> Crab_, too bad, but that won't be possible there are some gcc switches which do too much :-(
20090417 23:41:40< mordante> -Weff-c++ for example is nice but too intrusive (several headers also choke on it)
20090417 23:41:40< loonycyborg> Maybe also add -Wextra? :P
20090417 23:41:59-!- Turuk_ [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection timed out]
20090417 23:42:36< Crab_> mordante: well, its better to move slowly. if the trunk has no places which trigger the specific warning, why not turn it on ?
20090417 23:42:42< mordante> is -Wextra not enabled ?
20090417 23:43:09< mordante> Crab_, agreed just too bad -Weff-c++ does too much which can't be splitted
20090417 23:43:31< loonycyborg> Ah. -Wextra is the same as -W.
20090417 23:43:44< mordante> if warns about not initialized members but also about missing copy constructors when there are pointer members
20090417 23:43:47< mordante> it*
20090417 23:59:20-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has quit [Connection reset by peer]
20090417 23:59:35-!- Turuk [n=Turuk@wesnoth/forumsith/turuk] has joined #wesnoth-dev
--- Log closed Sat Apr 18 00:00:23 2009