--- Log opened Mon Aug 23 00:00:06 2010
--- Day changed Mon Aug 23 2010
20100823 00:00:06< TheBuzzSaw> I was wondering why it was running so much slower
20100823 00:00:13< TheBuzzSaw> and then I discovered the lack of hardware acceleration
20100823 00:00:31< shadowmaster> we aren't even sure yet that opengl rendering would help with that
20100823 00:00:41-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100823 00:00:42< TheBuzzSaw> Um, I 200% guarantee it would.
20100823 00:00:54< shadowmaster> at least it would make the code much simpler since we wouldn't need all those adaptative refresh hacks anymore
20100823 00:00:55< Ivanovic> it runs slower because a lot more is redrawn periodically
20100823 00:00:57< TheBuzzSaw> software rendering is a monumental task, especially with all these images loaded
20100823 00:01:02< Ivanovic> where it was a rather static image before
20100823 00:01:09< shadowmaster> and allow much more possibilities with special effects, etc.
20100823 00:01:27< TheBuzzSaw> The point is that a GPU redrawing the whole scene is flat out faster than the CPU redrawing only pieces
20100823 00:01:27< shadowmaster> and spritesheets. Can't forget the sprite sheets.
20100823 00:01:32< shadowmaster> yeah
20100823 00:01:43< TheBuzzSaw> I mean, look at 3D games. Every pixel on the screen changes frame to frame
20100823 00:02:08< Ivanovic> TheBuzzSaw: i am confident that it can help when done in a good way
20100823 00:02:15< TheBuzzSaw> I have a game project right now that uses OpenGL + sprite sheets
20100823 00:02:18< Ivanovic> though it is a monumental effort that someone has to start
20100823 00:02:18< TheBuzzSaw> 2D game
20100823 00:02:32< Ivanovic> and stay involved so that the port is eventually completed
20100823 00:02:33< TheBuzzSaw> I know. Wesnoth is a behemoth of a project, but it looks fairly well abstracted.
20100823 00:03:04< TheBuzzSaw> I would have to fully understand the high level logic of drawing the scene, but once I do, I guarantee I could do it
20100823 00:03:17< shadowmaster> involvement is particularly important even after completion of the prject.
20100823 00:03:22< TheBuzzSaw> Lemme show you a quick sample of my work
20100823 00:03:29< TheBuzzSaw> http://www.youtube.com/watch?v=unwPu4c888E
20100823 00:03:50< Ivanovic> anyway, it is time for me to head off to bed, n8
20100823 00:03:57< TheBuzzSaw> see ya Ivanovic
20100823 00:04:15< shadowmaster> (at least until us troglodites get the handle of an ogl rendering subsystem for wesnoth)
20100823 00:04:15< TheBuzzSaw> shadowmaster: Yeah, I know. I wouldn't leave everyone hangin' if I indeed did get involved.
20100823 00:04:27< shadowmaster> :)
20100823 00:04:52-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20100823 00:06:45< TheBuzzSaw> It's good that Wesnoth is built in such a way that it supports much older hardware, but it's starting to have the reverse effect: more modern machines cannot accelerate the graphics at all. :(
20100823 00:07:19< TheBuzzSaw> and I cannot afford a new CPU :P
20100823 00:07:25< shadowmaster> I personally think that support for older hardware isn't really important anymore since dropping support for Windows 9x.
20100823 00:07:38< shadowmaster> the AI has also become quite CPU-intensive
20100823 00:07:47< TheBuzzSaw> Yeah, it's a prime CPU job
20100823 00:08:02< TheBuzzSaw> Graphics has a shiny new hotel next door, but it refuses to leave XD
20100823 00:08:20< shadowmaster> and the rendering issues become really obvious in large scenarios with many AI-controller units, such as Northern Rebirth's.
20100823 00:08:24< shadowmaster> controlled
20100823 00:08:42< TheBuzzSaw> Is the game multi-threaded?
20100823 00:08:53< TheBuzzSaw> I guess that's a dumb question given its boost::thread requirement
20100823 00:09:07< shadowmaster> I'm not sure where threads are used. At least they are in the newer networking code.
20100823 00:09:42< shadowmaster> but I don't think the AI is threaded. At least I can't find any mention of "thread" in its source code
20100823 00:10:19< TheBuzzSaw> I am fascinated with AI, but I know next to nothing about it XD
20100823 00:11:55< loonycyborg> TheBuzzSaw: Personally instead of direct use of opengl I'd prefer using gl-enabled cairo lib, considering that we already depend on it.
20100823 00:12:40< TheBuzzSaw> I have little experience with cairo, but if it uses GL, it's a good start
20100823 00:12:49< TheBuzzSaw> What part of the game uses cairo?
20100823 00:13:11< loonycyborg> For now not much. It was pulled in as dependency of pango.
20100823 00:14:09< TheBuzzSaw> I'm not sure where cairo would help exactly... Based on what I know of OpenGL, it'd be rather straightforward to replace the SDL drawing mechanism
20100823 00:14:31< TheBuzzSaw> aside from the intelligent redraw logic currently there
20100823 00:15:34< TheBuzzSaw> But it seems several others have already died trying :P
20100823 00:15:41< TheBuzzSaw> I see some forum discussion from 2006
20100823 00:16:10< shadowmaster> there's been more recent discussion on the devs mailing list around February-March this year
20100823 00:16:34< Soliton> it was more of an issue of good 3d drivers being available back then.
20100823 00:17:12< TheBuzzSaw> MESA was pretty bad until recently
20100823 00:17:20< TheBuzzSaw> But it works great with my 2D projects now
20100823 00:17:22< shadowmaster> the last talk began here: https://mail.gna.org/public/wesnoth-dev/2010-02/msg00003.html
20100823 00:17:35< loonycyborg> TheBuzzSaw: I don't know much about 2d rendering and application of opengl to it. I only know that cairo is specialized on 2d graphics while gl isn't.
20100823 00:17:37< TheBuzzSaw> I have a laptop with no dedicated GPU. It runs on MESA, and it runs great now
20100823 00:17:46< shadowmaster> and yes, Mesa has been improving. I can't say much about their support for newer hardware, though
20100823 00:17:57< loonycyborg> TheBuzzSaw: It's also a chance to keep non-gl rendering path :P
20100823 00:18:14< shadowmaster> loonycyborg: to me it seems like ogl *is* specialized on 2d graphics
20100823 00:18:17< shadowmaster> and that 3d models, etc., are built on top of that
20100823 00:18:20< TheBuzzSaw> Well, what newer hardware? ATI and Nvidia have competent drivers now. MESA works great with intel acceleration
20100823 00:18:34< TheBuzzSaw> OpenGL technically specializes in nothing
20100823 00:18:38< TheBuzzSaw> whole frameworks are built on top of it
20100823 00:18:41< shadowmaster> TheBuzzSaw: well, it took two years to get Mesa 3D support for my ATI Radeon HD 3200
20100823 00:18:57< shadowmaster> although I hear they've bene progressing faster on the newer Evergreen-based GPUs
20100823 00:19:09< TheBuzzSaw> Aren't there good ATI drivers?
20100823 00:19:14< shadowmaster> I mean one year, bw
20100823 00:19:14< TheBuzzSaw> My buddy runs Ubuntu + ATI just fine.
20100823 00:19:25< shadowmaster> eh, the proprietary drivers don't like me :(
20100823 00:19:30< TheBuzzSaw> ah
20100823 00:19:47< TheBuzzSaw> I have a system76 laptop that uses MESA. It runs great now.
20100823 00:19:54< shadowmaster> they make my screen buzz, cause general unstability, and have really poor 2D performance compared to the free radeon DDX + Mesa DRI2 (KMS)
20100823 00:20:04< TheBuzzSaw> ouch
20100823 00:20:17 * TheBuzzSaw has an Nvidia GTX 275.
20100823 00:21:06< shadowmaster> all I've heard about NVIDIA suggests that their proprietary drivers work much better than ATI's on Linux, even if they aren't willing to release their specs unlike AMD/ATI
20100823 00:21:20< TheBuzzSaw> Yeah, it's an interesting tradeoff
20100823 00:21:35< TheBuzzSaw> As long as they keep up on driver updates, I'm happy
20100823 00:21:49< TheBuzzSaw> But once they become lazy, I'll be mad that they don't trust the open community.
20100823 00:24:52< loonycyborg> TheBuzzSaw: I have nvidia geforce 6800GS and had much issues with it.
20100823 00:24:58-!- mjs-de [~mjs-de@vpw.wh.Uni-Dortmund.DE] has quit [Remote host closed the connection]
20100823 00:25:10< TheBuzzSaw> hit and miss, I guess
20100823 00:25:21< loonycyborg> If they become lazy there's nouveau :P
20100823 00:25:35< loonycyborg> *hadn't much issues with it.
20100823 00:25:49< TheBuzzSaw> Regardless, I hate running Wesnoth so slowly. Someone mail me a Core i7 running at 4 GHz. :P
20100823 00:26:32-!- Johannes13 [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 276 seconds]
20100823 00:26:51< loonycyborg> iirc you can disable water animations..
20100823 00:27:51< TheBuzzSaw> another method is not maximizing the window in my 1080p monitor
20100823 00:29:06< shadowmaster> loonycyborg: nouveau is developed based on reverse engineering.
20100823 00:29:15< shadowmaster> I suspect their support rate will be very low.
20100823 00:29:38< shadowmaster> (i.e. how much time passes before support for X new chip is implemented)
20100823 00:29:42< TheBuzzSaw> or I'll just wait for Wesnoth II
20100823 00:29:59< shadowmaster> and yes, you can disable animated terrains
20100823 00:31:04< TheBuzzSaw> or the Wesnoth MMO
20100823 00:31:35< shadowmaster> augh, the Linux kernel bugzilla is timing out :(
20100823 00:41:32-!- zookeeper [~l@wesnoth/developer/zookeeper] has quit []
20100823 00:42:39-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Quit: Leaving.]
20100823 01:06:27-!- Elvish_Pillage2 [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has quit [Ping timeout: 246 seconds]
20100823 01:06:52-!- Elvish_Pillage2 [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20100823 01:07:16-!- TheBuzzSaw [~buzz@75-174-85-170.bois.qwest.net] has left #wesnoth-dev []
20100823 01:24:38-!- PeterPorty [~Pete@pc-30-248-120-200.cm.vtr.net] has joined #wesnoth-dev
20100823 01:33:58-!- ancestral [~ancestral@206.55.180.8] has quit [Ping timeout: 272 seconds]
20100823 01:34:40-!- ancestral [~ancestral@206.55.180.8] has joined #wesnoth-dev
20100823 01:37:36-!- knotwork [~markm@142.177.234.80] has quit [Ping timeout: 246 seconds]
20100823 01:37:53-!- knotwork [~markm@142.177.234.80] has joined #wesnoth-dev
20100823 02:04:28-!- Gallaecio [~quassel@225.189.60.213.dynamic.mundo-r.com] has quit [Remote host closed the connection]
20100823 02:06:17-!- ancestral [~ancestral@206.55.180.8] has quit [Ping timeout: 240 seconds]
20100823 02:11:12< Gambit> Soliton: whilst jogging/walking I came up with a perfect system for protecting persistent variables: http://wesnoth.pastebin.com/y41Hxe0r
20100823 02:11:27< Gambit> And it can all be done in WML :)
20100823 02:12:23< Gambit> It can further be protected by scrambling up what order you handle generating random letters for 1-9
20100823 02:12:53< Gambit> It's not perfect, but it's a huge pain to cheat.
20100823 02:15:59-!- ilor__ [~ilor@netbajt.pl] has joined #wesnoth-dev
20100823 02:16:09-!- ilor_ [~ilor@netbajt.pl] has quit [Ping timeout: 246 seconds]
20100823 02:16:10< Gambit> And this gives another use to the [chat] tag. While doing really heavy WML like that you can send [chat] messages occasionally to let the players no that the game isn't frozen, it's just busy. And they don't have to be closed like [message]es.
20100823 02:16:26< Gambit> *know
20100823 02:21:10-!- anonymissimus [~chatzilla@HSI-KBW-095-208-088-086.hsi5.kabel-badenwuerttemberg.de] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.2/20100316074819]]
20100823 02:30:57< Sapient> Gambit: that would be a really weak protection
20100823 02:33:34< Sapient> since the code to generate the checksum is plainly observable, you could just copy and paste it into a test scenario, then pass in your altered state
20100823 02:34:16< Gambit> Yes it doesn't protect against people with even a tiny bit of WML knowledge.
20100823 02:34:31< Gambit> But it protects against just opening up a text editor and changing numbers.
20100823 02:35:51< Gambit> But nothing will protect against people know know WML sadly.
20100823 02:36:10< Gambit> Apart from not saving persistent variables in plain text.
20100823 02:36:17< Gambit> But that is a terrible solution.
20100823 02:37:25< Sapient> the best protection would be something that is very unobtrusive and only people with extensive knowledge of the protection system would even realize the alert had been triggered
20100823 02:38:28< Sapient> so that way you can silently monitor the cheaters without them being aware of how you detected them
20100823 02:40:50< Gambit> Or after you've detected a cheater set a variable with a very unrelated looking name
20100823 02:41:07< Gambit> And then elsewhere in your code check that variable and reset them to a new player if it's a certain value.
20100823 02:41:24< Sapient> too obvious
20100823 02:41:33< Gambit> Actually no that still doesn't prevent someone making a cheating add-on.
20100823 02:41:41< Gambit> [20:35:51] <Gambit> But nothing will protect against people know know WML sadly.
20100823 02:42:36-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has quit [Quit: Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz]
20100823 02:42:48< Sapient> people won't be able to work around something if they are unaware of it. that's the best strategy to catch WML proficient cheaters
20100823 02:42:57< AI0867> how about this: use [store_global_variable] both globally and on the side itself, then compare the two
20100823 02:43:25-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection]
20100823 02:43:34< AI0867> and a normal variable as a third thing to compare with
20100823 02:44:49< Gambit> Sapient: But silently monitoring them does not prevent them from cheating.
20100823 02:45:11< Gambit> I mean yeah I don't have to literally scream out that they're a cheater.
20100823 02:45:18< Gambit> That gives them soemthing to search for.
20100823 02:45:29< Sapient> to prevent cheateing requires social engineering
20100823 02:45:33< Gambit> Rather I'd just reset their variables and treat them like I would someone who had never played the addon before.
20100823 02:46:18< Sapient> imagine if you have a reputation of being able to analyze anyones replays and determine if they were a cheater, but nobody knows how you determine it
20100823 02:46:46< Sapient> that would have a negative impact on the frequency of cheating, especially among established players
20100823 02:47:33< Gambit> But that requires that I actually examine replays for cheaters.
20100823 02:47:40< Gambit> At least a few times...
20100823 02:47:47< Sapient> right, social engineering
20100823 02:47:48< Gambit> Also if nobody figures out how I do it, how do they do it themselves?
20100823 02:48:08< Sapient> anyways, I need to go eat
20100823 02:48:12< Sapient> cya later
20100823 02:48:16-!- Sapient [~sapient@wesnoth/developer/sapient] has quit [Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.0.17/2009122204]]
20100823 03:35:04-!- alink [~alink@wesnoth/developer/alink] has joined #wesnoth-dev
20100823 03:36:50-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has quit [Read error: Connection reset by peer]
20100823 03:45:26-!- ancestral [~ancestral@mobile-166-137-142-113.mycingular.net] has joined #wesnoth-dev
20100823 04:00:00-!- ancestral_ [~ancestral@206.55.180.8] has joined #wesnoth-dev
20100823 04:00:23-!- ancestral_ [~ancestral@206.55.180.8] has quit [Read error: Connection reset by peer]
20100823 04:21:34< ancestral> Hey anyone know if I can plop entire images on the screen for UI with Theme WML?
20100823 04:22:00< alink> plop?
20100823 04:22:05< ancestral> Like, medium-large sized decals/ornaments
20100823 04:23:55< ancestral> Like the banner in the top left corner http://i19.tinypic.com/8eptyqq.jpg
20100823 04:24:31< alink> ah, not possible yet
20100823 04:24:53< ancestral> Darn
20100823 04:25:09< alink> i see a simple implementation to allow it, but I am not sure that I want to touch themeWML
20100823 04:25:44< ancestral> Seems no one does — no offense
20100823 04:26:15< alink> the simple implementation is: use floating label or halo, but with fixed screen coordinates instead of being attached to the map
20100823 04:26:31< shadowmaster> someone would touch it once the GUI2 transition is completed.
20100823 04:26:40< shadowmaster> in at least five years.
20100823 04:27:05< alink> ancestral: well I did some minor improvement there in 1.9.0
20100823 04:27:40< ancestral> So GUI2 will come out with 2.0? :-p
20100823 04:27:46< alink> but yeah themeWML should die, it doesn't evolve anymore
20100823 04:27:51< shadowmaster> hopefully that, 3.0 otherwise
20100823 04:28:00< ancestral> Haha, 3.0
20100823 04:28:09< alink> GUI2 is already there, I think the plan is to progressively replace GUI1
20100823 04:28:21< shadowmaster> yeah, but GUI2 doesn't implement all of GUI1 completely right yet
20100823 04:28:46< shadowmaster> and there are some widgets in the queue such as tab controls, progress bars, etc.
20100823 04:28:55< shadowmaster> actually I could probably code a progress bar widget myself. It's not rocket science.
20100823 04:32:45< alink> ancestral: mmh if you only want to add such banner in the map view, then such new code would not need to touch themeWML. But then it will collide with other floating text like MP chat or command line. This is mainly what prevented me to try this already 
20100823 04:33:16< alink> that and the fact the floating label currently only allow text and halo allow only simple images
20100823 04:34:06-!- PeterPorty [~Pete@pc-30-248-120-200.cm.vtr.net] has quit [Quit: Saliendo]
20100823 04:34:11< alink> mmh but I suppose that combining several halo and floating text could hack something
20100823 04:34:48< ancestral> alink: Yeah I was just curious :) There's nothing then to specify whether elements show during a game or in the multiplayer lobby, for example?
20100823 04:35:14< ancestral> It's everywhere or nowhere, mostly?
20100823 04:36:08< alink> "There's nothing then to specify whether elements show during a game or in the multiplayer lobby" ?? I don't understand, you mean GUI element ?
20100823 04:36:44< ancestral> Yeah, or a banner picture say
20100823 04:36:55< alink> the map view and lobby have no common elements (other than chat maybe)
20100823 04:37:06-!- Ivanovic_ [~ivanovic@dtmd-4db2f650.pool.mediaWays.net] has joined #wesnoth-dev
20100823 04:37:27< alink> oh, ok, then no, such things should be specify in both GUI I suppose
20100823 04:37:45< alink> which make sense since they are very different
20100823 04:38:54< alink> but maybe you mean the general "UI look", then there is 3 source, GUI1 (dialog look is hardcoded), GUI2(in nice WML) and themeWML (in complex and limited WML)
20100823 04:39:36< ancestral> Actually I was thinking originally the map view, but I guess it could extend to the others later
20100823 04:39:46< alink> AFAIK themeWML is only used for map view
20100823 04:40:15< alink> but GUI2 is planned to unify all of that, not sure yet how this will work for map view
20100823 04:40:29< ancestral> Sure
20100823 04:40:50-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Ping timeout: 276 seconds]
20100823 04:41:05-!- Ivanovic_ is now known as Ivanovic
20100823 04:42:13< alink> but the main problem with such big banner or floating UI element on top of map view, is that it may slow scrolling. Or if using halo-like implemention, it may cause a lot of redraw
20100823 04:42:47< alink> (current scrolling use the fact that the map view is rectangular)
20100823 04:44:59< alink> but these days haloes seems to be a less important performance cost, so few small UI elements using haloes should be ok. But better avoid use it for the whole sidebar info
20100823 04:55:16-!- Greywhind [~Greywhind@pool-96-238-43-241.prvdri.fios.verizon.net] has quit [Quit: leaving]
20100823 05:07:54-!- rusty [~rusty2@110.23.11.84] has joined #wesnoth-dev
20100823 05:08:42-!- un214 [~quassel@adsl-75-45-0-228.dsl.scrm01.sbcglobal.net] has joined #wesnoth-dev
20100823 05:09:01< un214> at some point someone was complaining about the prevalence of anceint liches
20100823 05:09:27< un214> to my mind, it's used in lots of places where demiliches should have been used but demilich was removed before 1.2
20100823 05:26:41-!- chungy [~mike@ip24-253-68-14.lv.lv.cox.net] has joined #wesnoth-dev
20100823 05:26:45-!- Daltx [~~@unaffiliated/daltx] has quit []
20100823 05:27:15< chungy> Hey, typo on the front page of the website, the Mac OS X version is said to be 290.5B (not MB)
20100823 05:27:19-!- Daltx [~~@unaffiliated/daltx] has joined #wesnoth-dev
20100823 05:28:19-!- wesbot changed the topic of #wesnoth-dev to: 136 bugs, 290 feature requests, 16 patches | logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100823 05:31:17< ancestral> chungy: That would download Soo fast
20100823 05:46:35< chungy> hehe, pretty good compression :p
20100823 05:47:55< ancestral> Yeah I want THAT compression method
20100823 05:49:31-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20100823 05:50:13-!- Blueblaze2 [~Blueblaze@adsl-76-202-21-219.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100823 05:50:54-!- un214 [~quassel@adsl-75-45-0-228.dsl.scrm01.sbcglobal.net] has quit [Remote host closed the connection]
20100823 05:52:59-!- Blueblaze [~Blueblaze@adsl-99-182-55-97.dsl.hstntx.sbcglobal.net] has quit [Ping timeout: 276 seconds]
20100823 05:53:03-!- Blueblaze [~Blueblaze@adsl-76-202-22-240.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20100823 05:54:39-!- Blueblaze2 [~Blueblaze@adsl-76-202-21-219.dsl.hstntx.sbcglobal.net] has quit [Read error: Connection reset by peer]
20100823 05:59:47-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection]
20100823 05:59:51-!- Gambit [~Gambit@unaffiliated/gambit] has quit [Read error: Connection reset by peer]
20100823 06:02:25-!- TheBuzzSaw [~buzz@75-174-85-170.bois.qwest.net] has joined #wesnoth-dev
20100823 06:03:26-!- TheBuzzSaw [~buzz@75-174-85-170.bois.qwest.net] has left #wesnoth-dev []
20100823 06:07:26-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100823 06:18:58-!- chungy [~mike@ip24-253-68-14.lv.lv.cox.net] has left #wesnoth-dev ["Leaving"]
20100823 06:21:11< shadowmaster> mordante: maybe if you have time you could help these people with the [portrait] tag: http://forums.wesnoth.org/viewtopic.php?f=21&t=31322
20100823 06:21:55< CIA-35> alink * r45953 /trunk/src/ (builder.cpp builder.hpp): Remove the WiP probability in [variant], since a better way is needed
20100823 06:21:56 * shadowmaster awaits some amusing guesswork in that thread for the time being
20100823 06:22:00< CIA-35> alink * r45954 /trunk/ (6 files in 2 dirs): 
20100823 06:22:00< CIA-35> Optimization of cache rebuild (>4x faster, for both cache building and "building terrain rules")
20100823 06:22:00< CIA-35> by introducing a new method for terrain variations, which stop relying on the WML preprocessor.
20100823 06:22:00< CIA-35> Still WiP. WML syntax is not final, and thus also WML macros and WML parsing (not optimized yet)
20100823 06:22:13< alink> \o/
20100823 06:22:47< alink>  I was only expecting a 2x boost, so I am happily surprised :-)
20100823 06:22:49< shadowmaster> alink: is that supposed to break trunk? :P
20100823 06:22:58< shadowmaster> or can I report whatever issue I stumble upon?
20100823 06:23:31< alink> it is not supposed to have visible impact
20100823 06:24:35< alink> except maybe that I still need to tune the RNG used to pick terrain variations.
20100823 06:27:12< alink> The current WiP one doesn't seem to have visible pattern but i just used a big prime number to fix a new binary pattern. I will check the math later to be sure that any input can have one.
20100823 06:28:09< alink> Note that the old way was probably just using modulo 100 there (for probability%), so maybe nothing new here
20100823 06:30:35< alink> shadowmaster: but I am interested in new terrain glitches. I don't expect one, but it's always possible that I make a typo or miss a WML macro
20100823 06:31:17-!- ancestral [~ancestral@mobile-166-137-142-113.mycingular.net] has quit [Quit: Colloquy for iPhone - http://colloquy.mobi]
20100823 06:34:58< alink> btw should also reduce cache size (about the same order, maybe less), but mainly for the raw WML, because the gz compression probably compressed well the terrain rules repetition
20100823 06:35:24-!- Blueblaze [~Blueblaze@adsl-76-202-22-240.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20100823 06:37:34< alink> and I still need to use this new feature to also optimize the few uneven random terrains. But I wait to have the final syntax before touching all these special cases
20100823 06:41:02< alink> boucman: btw these macro changes are only for test (committed to make trunk faster). I suppose that we should handle variations is a nicer way. Either introducing new macro parameter (yet another one :-/ ) or define some hardcoded default (like 1,2,3...).
20100823 06:41:20< alink> *in a nice way
20100823 06:43:09< alink> currently always defined in WML random rules and only used when '@V' is in the image name. But that's just a convention used to easily test this without rewriting all the macros
20100823 06:52:04-!- stikonas [~and@ctv-213-164-99-190.vinita.lt] has joined #wesnoth-dev
20100823 06:52:04-!- stikonas [~and@ctv-213-164-99-190.vinita.lt] has quit [Changing host]
20100823 06:52:04-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100823 06:58:56-!- noy [~Noy@wesnoth/developer/noy] has quit [Ping timeout: 260 seconds]
20100823 07:10:22-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100823 07:11:17-!- zookeeper [~l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20100823 07:11:56-!- silene [~plouf@bau91-1-82-239-244-109.fbx.proxad.net] has joined #wesnoth-dev
20100823 07:11:56-!- silene [~plouf@bau91-1-82-239-244-109.fbx.proxad.net] has quit [Changing host]
20100823 07:11:56-!- silene [~plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20100823 07:20:33-!- Kestenvarn [~ie@c-68-40-169-207.hsd1.mi.comcast.net] has joined #wesnoth-dev
20100823 07:21:11< shadowmaster> hi Kestenvarn 
20100823 07:23:21< Kestenvarn> hello
20100823 07:25:57-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 272 seconds]
20100823 07:50:40-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100823 07:50:58-!- Octalot [~noct@cpc2-hitc1-0-0-cust727.9-2.cable.virginmedia.com] has quit [Remote host closed the connection]
20100823 07:52:23-!- zookeeper [~l@wesnoth/developer/zookeeper] has quit [Ping timeout: 276 seconds]
20100823 07:52:28-!- MeccaGod [~majs@h14n4fls31o279.telia.com] has joined #wesnoth-dev
20100823 07:55:33-!- zookeeper [~l@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20100823 07:59:37-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20100823 08:01:13< CIA-35> alink * r45955 /trunk/data/core/ (terrain-graphics/internal-generic.cfg terrain-graphics.cfg): 
20100823 08:01:13< CIA-35> Experimental optimization of water animation (invalidation are several times better, depending of geometry)
20100823 08:01:13< CIA-35> Actually use an old forgotten single-hex terrain rule syntax, instead of the usual multihex.
20100823 08:01:13< CIA-35> Only done for the 4 basic water tiles, no visible change and just added a new WML macro.
20100823 08:01:13< CIA-35> This is also a test to check how an optimized handling of multihex could help.
20100823 08:02:18< alink>  \o/
20100823 08:02:46< alink> I was recently annoyed by the handling of this unused single-hex-image syntax, but that seems usefull :)
20100823 08:05:47< alink> but since the multihex syntax is more handy, I wish to try optimizing multihex stuff by autoremoving empty image, so any unnecessarily multihex will be equivalent to single hex
20100823 08:06:29< alink> of course specifying in WML which image is single hex will always be more efficient, maybe just need good macro
20100823 08:06:46< alink> boucman:^
20100823 08:08:32< alink> this is mainly useful to help animation, but a more general use of this could also reduce the number of empty images
20100823 08:09:09< alink> assuming they have a significant cost somewhere and that my future plan to auto-prune them fails
20100823 08:10:31-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Quit: ...]
20100823 08:13:48< alink> boucman: btw I called a "internal" macro in terrain-graphics.cfg. I suppose that I should have added another macro layer, but since it's just an experiment to show how this work, I made it simple and direct.
20100823 08:15:22-!- Kestenvarn [~ie@c-68-40-169-207.hsd1.mi.comcast.net] has quit [Quit: who knows what delicate wonders have died out of the world, for want of the strength to survive]
20100823 08:18:57-!- alink [~alink@wesnoth/developer/alink] has quit [Remote host closed the connection]
20100823 08:28:48-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: GO, GET TO THE CHOPPAH!!!]
20100823 08:54:04-!- euschn [~eugen@wesnoth/developer/euschn] has joined #wesnoth-dev
20100823 09:01:16-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100823 09:04:23< shadowmaster> shikadibot: repo
20100823 09:04:23< shikadibot> Repository URI: http://svn.gna.org/svn/wesnoth
20100823 09:07:44-!- silene [~plouf@wesnoth/developer/silene] has quit [Quit: Leaving.]
20100823 09:26:07-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit []
20100823 09:28:24-!- Ivanovic [~ivanovic@dtmd-4db2f650.pool.mediaWays.net] has quit [Changing host]
20100823 09:28:24-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20100823 09:32:03< Ivanovic> moin
20100823 09:44:56< CIA-35> zookeeper * r45956 /trunk/data/core/macros/utils.cfg: Converted the MOVE_UNIT macro to use the new [move_unit] tag.
20100823 09:45:52< Ivanovic> zookeeper: can you have a look at the last comment in this patch (about SOTBE not working): https://gna.org/patch/?1879
20100823 09:47:18< zookeeper> yeah, looking...
20100823 09:49:40-!- Netsplit *.net <-> *.split quits: euschn, Daltx
20100823 09:50:54< zookeeper> Ivanovic, no idea, it works for me and the macro is there too; i added it on the 16th and the file hasn't been touched since.
20100823 09:52:36< Ivanovic> okay
20100823 09:52:57< CIA-35> ivanovic * r45957 /trunk/data/campaigns/ (183 files in 25 dirs): 
20100823 09:52:57< CIA-35> apostrophe conversion, ellipses conversion, stuff like this (patch #1877 by Simon Smith)
20100823 09:52:57< CIA-35> TODO: apply a fixed up version of patch #1879, currently the pofix rules in it will not work
20100823 09:53:57-!- Netsplit over, joins: Daltx
20100823 09:59:37< Ivanovic> zookeeper: can you have a look at converting these (after the last commit) to using the whisper macro?
20100823 09:59:50< Ivanovic> http://pastebin.com/ukFaLrAu
20100823 09:59:55-!- euschn [~eugen@wesnoth/developer/euschn] has joined #wesnoth-dev
20100823 10:00:21< zookeeper> sure
20100823 10:00:26< Ivanovic> i grepped on the diff i had, so the files might not always be accurate, the strings and unit names should work nicely though
20100823 10:12:12< CIA-35> zookeeper * r45958 /trunk/data/campaigns/ (7 files in 4 dirs): Made some whispers use the WHISPER macro instead of having them prefixed with "(<i>Whispered</i>)".
20100823 10:13:58< zookeeper> it's a bit ackward to try to use regular speech and a whisper on the same line, so i split one long TRoW message into two
20100823 10:18:33< Ivanovic> mkay
20100823 10:18:39< zookeeper> hmm, well, maybe it's not that ackward after all.
20100823 10:18:54< Ivanovic> then do it via string concatenation
20100823 10:19:06< zookeeper> yeah
20100823 10:19:42< CIA-35> zookeeper * r45959 /trunk/data/campaigns/ (2 files in 2 dirs): Converted a remaining whisper in DW to use the WHISPER macro, and reverted an unnecessary splitting of a message into two in TRoW.
20100823 10:20:15< zookeeper> it's a bit ugly since you need an extra space in the beginning of the second string..
20100823 10:54:42< CIA-35> zookeeper * r45960 /trunk/data/campaigns/The_Rise_Of_Wesnoth/ (2 files in 2 dirs): 
20100823 10:54:42< CIA-35> In 'Temple in the Deep', opened up the chokepoints somewhat, unsymmetrized the
20100823 10:54:42< CIA-35> map and beautified it a bit, removed the healing monoliths which were
20100823 10:54:42< CIA-35> confusingly just normal villages, added more regular villages and adjusted both
20100823 10:54:42< CIA-35> sides' income accordingly.
20100823 11:06:39< CIA-35> zookeeper * r45961 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/06_Temple_of_the_Deep.cfg: Lenvan's income should actually have been decreased a bit, not increased...
20100823 11:15:28-!- Ivanovic_ [~ivanovic@dtmd-4db2db4d.pool.mediaWays.net] has joined #wesnoth-dev
20100823 11:15:28-!- Ivanovic_ [~ivanovic@dtmd-4db2db4d.pool.mediaWays.net] has quit [Changing host]
20100823 11:15:28-!- Ivanovic_ [~ivanovic@wesnoth/developer/ivanovic] has joined #wesnoth-dev
20100823 11:15:44-!- Ivanovic [~ivanovic@wesnoth/developer/ivanovic] has quit [Disconnected by services]
20100823 11:15:47-!- Ivanovic_ is now known as Ivanovic
20100823 11:28:19-!- wesbot changed the topic of #wesnoth-dev to: 136 bugs, 289 feature requests, 15 patches | logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100823 11:42:18-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Remote host closed the connection]
20100823 11:52:10-!- rusty [~rusty2@110.23.11.84] has quit [Ping timeout: 258 seconds]
20100823 11:58:38-!- loonybot [~loonybot@ppp95-165-186-202.pppoe.spdop.ru] has joined #wesnoth-dev
20100823 11:58:38-!- loonybot [~loonybot@ppp95-165-186-202.pppoe.spdop.ru] has quit [Changing host]
20100823 11:58:38-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-dev
20100823 11:59:28-!- loonycyborg [~sergey@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20100823 12:03:30-!- Appleman1234 [~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au] has quit [Ping timeout: 272 seconds]
20100823 12:03:35-!- Appleman1234 [~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au] has joined #wesnoth-dev
20100823 13:03:15-!- Johannes13 [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20100823 13:04:59-!- Appleman1234 [~Appleman1@CPE-60-226-176-19.qld.bigpond.net.au] has quit [Ping timeout: 265 seconds]
20100823 13:06:55-!- mjs-de [~mjs-de@vpw.wh.uni-dortmund.de] has joined #wesnoth-dev
20100823 13:18:28< CIA-35> zookeeper * r45962 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/09_Fallen_Lich_Point.cfg: Made getting the early finish bonus require killing all enemy leaders but not moving Haldric to the sewer. Also simplified some of the WML a bit.
20100823 13:36:02< CIA-35> zookeeper * r45963 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/10_Sewer.cfg: Made the early finish bonus be given as the objectives state. Also cleaned up the WML a bit.
20100823 13:50:39< CIA-35> zookeeper * r45964 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/08_Clearwater_Port.cfg: Made the loot from the enemy leaders be given to the player even if the killer is the AI ally. Also simplified the WML for the orc gold boosts.
20100823 13:52:59< CIA-35> zookeeper * r45965 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/12_A_Final_Spring.cfg: Made merman hunters recruitable in addition to fighters, made Typhon arrive with a loyal fighter and a loyal hunter, and added a message telling that merfolk are now recruitable.
20100823 13:56:18< CIA-35> zookeeper * r45966 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/15_New_Land.cfg: Disallow the recruitment of hunters as well as fighters when Typhon leaves.
20100823 13:59:24-!- Gambit [~Gambit@pa-67-234-126-121.dhcp.embarqhsd.net] has joined #wesnoth-dev
20100823 13:59:24-!- Gambit [~Gambit@pa-67-234-126-121.dhcp.embarqhsd.net] has quit [Changing host]
20100823 13:59:24-!- Gambit [~Gambit@unaffiliated/gambit] has joined #wesnoth-dev
20100823 14:01:08< CIA-35> eleazar * r45967 /trunk/data/core/terrain.cfg: added a few castles to additional groups.
20100823 14:20:33-!- Smar [smar@freenet/translator/finnish/Smar] has quit [Ping timeout: 245 seconds]
20100823 14:21:54-!- esr [~chatzilla@wesnoth/developer/esr] has quit [Ping timeout: 240 seconds]
20100823 14:31:05-!- Smar [smar@freenet/translator/finnish/Smar] has joined #wesnoth-dev
20100823 14:33:24-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev
20100823 14:36:54< CIA-35> zookeeper * r45968 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/13_Peoples_in_Decline.cfg: To reduce the tomatoeness of Bitey the Serpent, made it appear on the next turn start instead of on moveto and also tweaked the random location area.
20100823 14:37:08< Soliton> hrm, dev server crash because of an admin command...
20100823 14:39:23< Soliton> infact possibly also triggerable by a normal client.
20100823 14:40:52-!- rusty [~rusty2@110.23.3.180] has joined #wesnoth-dev
20100823 14:41:09-!- rusty [~rusty2@110.23.3.180] has quit [Client Quit]
20100823 14:47:07-!- euschn [~eugen@wesnoth/developer/euschn] has left #wesnoth-dev []
20100823 14:57:17< Gambit> Oh cool. Apparently [place_shroud] and [remove_shroud] can both be fixed at the same time.
20100823 14:57:40< Gambit> (You guys think of everything.)
20100823 15:09:26< Gambit> I've heard people say that the AI ignores shroud. Do they literally ignore it, or is it just not set for them ever?
20100823 15:09:39< Gambit> And if it's the latter do I have to avoid placing shroud on AI teams?
20100823 15:10:16< Soliton> it's the former.
20100823 15:10:48< Soliton> you should place shroud same as for a human player.
20100823 15:11:16< Soliton> in case we ever get an AI that works with shroud.
20100823 15:16:37-!- krotop [~kvirc@smj33-1-82-233-64-163.fbx.proxad.net] has joined #wesnoth-dev
20100823 15:16:59-!- Elvish_Pillage2 [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has quit [Ping timeout: 276 seconds]
20100823 15:18:03-!- DesertPanther [~Khalid@unaffiliated/desertpanther] has joined #wesnoth-dev
20100823 15:23:51-!- Greywhind [~Greywhind@fw-34-17.cs.brown.edu] has joined #wesnoth-dev
20100823 15:38:34< Gambit> Hmmm
20100823 15:38:48< Gambit> The minimap is correct, but the shroud on screen is not affected.
20100823 15:39:19< Soliton> missing a redraw?
20100823 15:40:05< Gambit> You mean in WML?
20100823 15:41:24< Soliton> i have no idea but i assumed you're writing C++ atm.
20100823 15:42:16< Gambit> Yes I am
20100823 15:42:35< Gambit> And yeah adding [redraw] in WML didn't do anything.
20100823 15:42:44 * Gambit looks at the original code to make sure he didn't delete anything
20100823 15:43:51-!- thespaceinvader [~chatzilla@wesnoth/artist/thespaceinvader] has joined #wesnoth-dev
20100823 15:44:08-!- stikonas [~and@ctv-213-164-99-190.vinita.lt] has joined #wesnoth-dev
20100823 15:44:09-!- stikonas [~and@ctv-213-164-99-190.vinita.lt] has quit [Changing host]
20100823 15:44:09-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20100823 15:45:04-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20100823 15:45:12< Gambit> Nope that's not it.
20100823 15:51:36< Gambit> Well this is interesting.
20100823 15:51:43< Gambit> I undid all my changes back to the original code.
20100823 15:51:48< Gambit> It does the same thing.
20100823 15:51:59< Gambit> Is [remove_shroud] broken in 1.9 then?
20100823 15:52:20< zookeeper> woah. this condition evaluates to false: [variable] name=side_$side_number|_serpent_triggered boolean_equals=no [/variable]
20100823 15:52:26< zookeeper> (the variable hasn't been set)
20100823 15:55:17< Gambit> Or maybe it's a problem of when I'm removing shroud.
20100823 15:56:38< Gambit> Nope. name=recruit also only updates the minimap even with the original code.
20100823 15:57:48< Gambit> zookeeper: As you're experimenting with WML already anyway, can you please quickly confirm for me if [remove_shroud] is broken?
20100823 15:58:23< zookeeper> sure, give me a minute...
20100823 16:01:55< zookeeper> nope, worked for me
20100823 16:02:02< zookeeper> (in TRoW:Sewer)
20100823 16:02:08< Gambit> Weird...
20100823 16:02:17< Gambit> Running svn diff shows no differences :(
20100823 16:02:35< Soliton> how do you compile?
20100823 16:02:40< Gambit> Scons
20100823 16:04:42< Gambit> Could it be that zookeeper is (more|less) up to date than me and it was (fixed|broken) in that time difference?
20100823 16:05:15< Soliton> potentially.
20100823 16:05:59< zookeeper> my build is 2 days old
20100823 16:06:23< Soliton> (new build coming up)
20100823 16:06:46< zookeeper> yay
20100823 16:07:10< Gambit> I just did svn update and game_events does have changes
20100823 16:07:42< Gambit> So I guess I'll see when it's done
20100823 16:09:42< Soliton> if you build wesnothd too it'd be nice to try "/query status" on a local server and see if it crashes.
20100823 16:15:02< Gambit> Nope. Still not working.
20100823 16:15:21< Gambit> Minimap is updated. On the playing field is not :(
20100823 16:15:57 * Gambit will try a mainline campaign that uses it
20100823 16:22:59< zookeeper> arr. a WML dilemma. rare but annoying.
20100823 16:23:34< Gambit> Strange according to kate's find in files tool only UtBS's Test Scenario uses it.
20100823 16:23:58< Gambit> Err nevermind.
20100823 16:25:29< Gambit> It works there. So apparently it's a problem with my WML somehow
20100823 16:25:45< Gambit> And by there I mean DM:19 which clears shroud for player 1 at prestart.
20100823 16:28:45< Gambit> Okay it seems to be only when I remove it at a specific location.
20100823 16:28:58< Gambit> Clearing all shroud for a side works.
20100823 16:29:27< Gambit> Clearing at x,y=15,15 only affects the minimap
20100823 16:30:20< Gambit> Can anyone else test that for me?
20100823 16:30:27< Soliton> maybe shroud transitions are so you can't see one unshrouded hex?
20100823 16:30:47< Gambit> Ah
20100823 16:31:53< Gambit> Yeah that's exactly it!
20100823 16:32:01< Gambit> Well that is... really bad art.
20100823 16:32:06< fendrin> hello world
20100823 16:34:08< Gambit> http://imagebin.org/110972
20100823 16:34:16< Gambit> The hex that is selected in that screenshot is 15,15
20100823 16:34:24< Gambit> That's with 15-18,15-18 unshrouded
20100823 16:34:29< Gambit> You can barely see any of 15,15
20100823 16:35:30< Gambit> Well it's always nice to know that your code isn't broken. ^_^
20100823 16:35:48 * Gambit puts his new shroud function back
20100823 16:37:54-!- Upthorn [ogmar@adsl-75-26-204-39.dsl.scrm01.sbcglobal.net] has quit [Ping timeout: 240 seconds]
20100823 16:48:10-!- Gallaecio [~quassel@225.189.60.213.dynamic.mundo-r.com] has joined #wesnoth-dev
20100823 17:07:55< Gambit> The controller of a side not in play would be null?
20100823 17:08:05< Gambit> For example if they hosted a multiplayer game with side 2 empty.
20100823 17:08:41< Gambit> Do I need to worry about placing/removing shroud on a side that doesn't exist?
20100823 17:08:46< CIA-35> zookeeper * r45969 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/14_Rough_Landing.cfg: 
20100823 17:08:46< CIA-35> Made the three serpents less of a tomato surprise: each now appears on the next
20100823 17:08:46< CIA-35> turn start after their respective enemy side has taken their first loss, at a
20100823 17:08:46< CIA-35> randomly picked location from which they cannot reach any of the player's units
20100823 17:08:46< CIA-35> in one turn.
20100823 17:08:50< zookeeper> phew. that took a bit of effort to do somewhat cleanly.
20100823 17:10:08-!- dtiger [~dtiger@dynamic-vpdn-93-125-68-39.telecom.by] has joined #wesnoth-dev
20100823 17:12:34< Gambit> I'm trying to figure out how to handle placing/removing shroud for all
20100823 17:14:21< Soliton> same as for one side but for each?
20100823 17:14:44< Gambit> Soliton: Right, but how to know who each is?
20100823 17:15:02< Gambit> As in the case of side 2 not playing.
20100823 17:15:44< Gambit> I guess that's handled at a higher level?
20100823 17:15:56< Gambit> The same place where it checks if shroud is enabled and what not?
20100823 17:17:00< Soliton> well, what happens if side=2 should be unshrouded and it doesn't play?
20100823 17:17:12< Gambit> So is it acceptable to just place/remove for 1 through resources::teams->size() ?
20100823 17:17:38< Soliton> why not?
20100823 17:17:54< Gambit> Well... 0 through resources::teams->size() I guess it would be, but anyway...
20100823 17:18:14< Gambit> Soliton: I don't know why not. That's why I'm asking so many questions. :)
20100823 17:18:40< Gambit> I have no idea what is acceptable to you.
20100823 17:19:24< Soliton> well, you have 1) the current code to check for the current behaviour and can 2) test, test, test.
20100823 17:31:25< fendrin> Hmmm, every flat terrain in the editor is named "flat (flat)" with the exception of the wood floor which is named "flat ()". I wonder if this is intentional.
20100823 17:38:46-!- _jbx_ [~jbailey@12.190.80.225] has joined #wesnoth-dev
20100823 17:56:26< fendrin> Who is maintaining Death Water?
20100823 17:59:23< zookeeper> no one, esr, me, whoever :P
20100823 18:03:08< Gambit> I semi-volunteered to fix bugs I guess.
20100823 18:03:30< Gambit> Also — it's working :D And you can now use [remove_shroud][/remove_shroud] to clear everyone's shroud.
20100823 18:03:54< Gambit> (or [place_shroud][/place_shroud] to do the opposite of course)
20100823 18:04:25< fendrin> zookeeper: Do you agree with replacing some of the httt elensefar wood bridges with stone ones?
20100823 18:04:29< CIA-35> zookeeper * r45970 /trunk/data/campaigns/The_Rise_Of_Wesnoth/scenarios/17b_A_Beach.cfg: Made the nagas appear a bit sooner and a bit further away.
20100823 18:06:22-!- elvish_sovereign [~elvish_so@pool-108-2-86-247.phlapa.east.verizon.net] has joined #wesnoth-dev
20100823 18:06:27< zookeeper> fendrin, umm, i guess that's ok, even though i haven't really gotten used to those yet
20100823 18:06:50< zookeeper> don't forget to copy those changes to liberty
20100823 18:07:19-!- elvish_sovereign [~elvish_so@pool-108-2-86-247.phlapa.east.verizon.net] has quit [Read error: Connection reset by peer]
20100823 18:07:21-!- elvish_sovereign [~elvish_so@pool-108-2-86-247.phlapa.east.verizon.net] has joined #wesnoth-dev
20100823 18:07:31< fendrin> zookeeper: Which map does feature elensefar in liberty?
20100823 18:08:23< zookeeper> 4th
20100823 18:09:23< fendrin> zookeeper: Let's merge the maps into one, select only a subset on the scenarios and do differences with masks.
20100823 18:10:14< Soliton> you have to put the base map into core then though.
20100823 18:13:07< fendrin> Soliton: Somthin like data/core/places/maps
20100823 18:15:55< zookeeper> fendrin, uh, i can't see how that could be worth the trouble
20100823 18:16:06< zookeeper> sounds like a severe case of overengineering
20100823 18:16:24< fendrin> zookeeper: Okay
20100823 18:16:59< fendrin> zookeeper: Note that the maps are already out of sync regarding the different rode types for example.
20100823 18:17:51< zookeeper> there's almost 20 years in between so small differences don't matter much
20100823 18:20:18< fendrin> zookeeper: Which one is the map taking place before?
20100823 18:20:37< Gambit> Maybe only the latter map should use stone bridges and fancier roads
20100823 18:21:44< zookeeper> fendrin, liberty
20100823 18:22:45 * zookeeper wonders how to un-tomatoize the chocobone ambushes in TRoW 'the vanguard'
20100823 18:25:03< Gambit> Man there were a lot of places using [place_shroud] and [remove_shroud] in data/. Luckily like 80% of them were already using side=1 :D
20100823 18:25:17< zookeeper> i could place them at the edge of the player's vision, but the chocobones are still twice as fast as, say, the loyal HI
20100823 18:26:26< fendrin> zookeeper: You can bring them in but eat some or all their movepoint left. So they can't move at the first sight.
20100823 18:26:58< zookeeper> that's exploitable and would seem pretty odd
20100823 18:27:15< zookeeper> if there wasn't fog then i could just place them far enough away that they can't reach the player in one turn
20100823 18:27:48< zookeeper> but i can't place them in fog because then anyone who doesn't realize that they're chocobones (and knows what they are) can be screwed easily
20100823 18:27:50< fendrin> zookeeper: Bring them in "stoned" and unstone them ?
20100823 18:28:20< fendrin> zookeeper: Would also fit to the unit.
20100823 18:28:40< zookeeper> you get ambushed by petrified units which then suddenly get unpetrified and charge you? -.-
20100823 18:28:53< zookeeper> sounds like a worse kind of surprise to me
20100823 18:30:19< fendrin> zookeeper: Script their appearance and show the player how the stone a creature.
20100823 18:46:18-!- krotop [~kvirc@smj33-1-82-233-64-163.fbx.proxad.net] has quit [Ping timeout: 272 seconds]
20100823 18:46:45-!- Zarel [~Zarel@warzone2100/developer/Zarel] has joined #wesnoth-dev
20100823 18:46:55< Zarel> "But for now, nobody has done this, and we don't see this as a priority since it would require writing some automated mechanism for generating these packages in conformance with the format used by the full distributions such as the Windows version installer."
20100823 18:47:31< Zarel> This doesn't sound that hard.
20100823 18:47:43-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100823 18:49:51< boucman> hey all
20100823 18:49:55< CIA-35> eleazar * r45971 /trunk/data/core/terrain.cfg: fixed the move/def of the wood floor.
20100823 18:50:02< eleazar> boucman: morning
20100823 18:52:51< Gambit> eleazar: does terrain art include shroud?
20100823 18:52:58< eleazar> yeah
20100823 18:53:36< eleazar> a single non-shroud hex doesn't show up well
20100823 18:53:38< eleazar> hmm
20100823 18:53:42< Gambit> Okay then: http://imagebin.org/110972
20100823 18:53:44< eleazar> i think i have an easy fix for that
20100823 18:53:50< Gambit> Yeah a single hex, or the top left corner of a box
20100823 18:54:09< eleazar> the top left corner there isn't a "real" hex
20100823 18:54:25< Gambit> It is unshrouded.
20100823 18:54:40< eleazar> it looks like a "half hex"
20100823 18:54:49< eleazar> like on the edge of a screen
20100823 18:54:50< eleazar> is it?
20100823 18:54:52< Gambit> Nope
20100823 18:55:05< Gambit> in a box from 15-18,15-18 that is 15,15
20100823 18:57:12< eleazar> Gambit: are you messing with the shroud code?
20100823 18:57:25< Gambit> Just the tags that place it.
20100823 18:58:23-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100823 18:58:23< Gambit> I have a patch (unuploaded yet) that lets the tags do multiple sides or all at once.
20100823 19:00:27< eleazar> i made those 6 years ago -- i could do better now
20100823 19:09:34< boucman> wesbot: seen Gabba
20100823 19:09:34< wesbot> boucman: The person with the nick gabba last spoke 1d 19h ago.  1d 19h ago person left:
20100823 19:12:11-!- krotop [~kvirc@smj33-1-82-233-64-163.fbx.proxad.net] has joined #wesnoth-dev
20100823 19:12:14-!- krotop [~kvirc@smj33-1-82-233-64-163.fbx.proxad.net] has quit [Client Quit]
20100823 19:30:15-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Read error: Connection reset by peer]
20100823 19:30:37-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev
20100823 19:30:46< CIA-35> caslav_ilic * r45972 /branches/1.8/ (10 files in 3 dirs): Updated translations for Serbian (intentionally not applying to trunk too).
20100823 19:40:44-!- Greywhind [~Greywhind@fw-34-17.cs.brown.edu] has quit [Read error: Connection reset by peer]
20100823 19:41:40-!- Greywhind [~Greywhind@fw-34-18.cs.brown.edu] has joined #wesnoth-dev
20100823 19:44:10-!- Johannes13_ [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20100823 19:44:17-!- Johannes13 [~Johannes@unaffiliated/johannes13] has quit [Read error: Connection reset by peer]
20100823 19:51:45-!- Latsni [~Latsni@p54B627A5.dip.t-dialin.net] has joined #wesnoth-dev
20100823 19:55:22-!- alink [~alink@wesnoth/developer/alink] has joined #wesnoth-dev
20100823 19:55:28< alink> hi
20100823 19:57:12< alink> Gambit, eleazar: that WML function clearing shroud allows to create case impossible to get when moving units, maybe not important to support it
20100823 19:58:02< eleazar> alink: so i expected
20100823 19:58:04< boucman> alink: thx for your work on random, it looks good
20100823 19:58:12< eleazar> but i can still make it better
20100823 19:58:16< boucman> I won't use it until I get your "go" though
20100823 19:58:48< Ivanovic> alink: have you seen that 1.9.0 uses significantly more ram when eg starting a campaign?
20100823 19:58:59< alink> Gambit: I see several ways to fix it: 1) ask WML author to smooth their shroud map 2) provide an editor tool only creating smooth map 3) auto smooth them at runtime (but that will remove some shourd which may not be wanted)
20100823 19:59:01< Ivanovic> (not running on the pandora anymore due to using >256MB ram)
20100823 19:59:36< alink> boucman: yes I still need to work and decide some details
20100823 20:00:05< alink> Ivanovic: when building the cache ?
20100823 20:00:08< Ivanovic> yes
20100823 20:00:41< alink> Ivanovic: then maybe try trunk, uncompressed cache is much smaller now
20100823 20:01:14< alink> something like test-scenario going from 10MB to 3MB (from memory)
20100823 20:01:25< Ivanovic> have not tried trunk so far
20100823 20:01:34< alink> *from my memory
20100823 20:02:55< Ivanovic> it looks like ram usage jumps up for a short time
20100823 20:03:14< Ivanovic> when it is writing the compressed cache onto the disc (at least that seems to be the time regarding the log info)
20100823 20:03:18< alink> Ivanovic: before drawing anything ? 
20100823 20:03:24< CIA-35> torangan * r45973 /branches/1.8/ (changelog data/hardwired/fonts.cfg fonts/DejaVuSans.ttf): font update
20100823 20:04:03< Ivanovic> http://wesnoth.org/irclogs/2010/08/%23wesnoth-dev.2010-08-18.log
20100823 20:04:11< Ivanovic> starting at 20100818 18:41:17
20100823 20:04:30< CIA-35> torangan * r45974 /trunk/ (changelog data/hardwired/fonts.cfg fonts/DejaVuSans.ttf): font update
20100823 20:04:35< alink> I think I saw it, but will read it again
20100823 20:04:39< Ivanovic> started the game, tried to start "two brothers" (really tiny wml wise) on easy, out of memory
20100823 20:06:28< alink> ok, just checked trunk with test-scenario, cache is ~2x smaller (both compressed and uncompressed). maybe it will help this 
20100823 20:06:56< alink> "(really tiny wml wise)" the biggest WML part was from terrain_graphics which is used in all campaigns
20100823 20:08:49< Ivanovic> that is correct
20100823 20:09:07< alink> Ivanovic: have you tried creating the cache on PC and copy it to the pandora, to see if it's just cache creation, not cache itself ?
20100823 20:09:11< Ivanovic> though eg northern rebirth, trow and the likes got a lot more wml (including huge amounts of macros), too
20100823 20:09:26< Ivanovic> yes, i tried to create it on the pc though it seems to not have used that one
20100823 20:09:46-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20100823 20:09:47< Ivanovic> currently building a pandora binary of trunk, want log output with --log-info=all or not?
20100823 20:09:51< Ivanovic> hi mordante
20100823 20:09:53< alink> you need to use the same configure switch like small-gui or tiny
20100823 20:09:55< Ivanovic> mordante: noy was looking for you
20100823 20:09:58< mordante> servus
20100823 20:10:02< mordante> hi Ivanovic 
20100823 20:10:02< alink> hello mordante
20100823 20:10:06< noy> oh hey mordante 
20100823 20:10:06< mordante> hi alink 
20100823 20:10:10< mordante> hi noy 
20100823 20:10:20< Ivanovic> alink: this is my cmake configuration line:
20100823 20:10:22< Ivanovic> CFLAGS="-O3 -mcpu=cortex-a8 -mfpu=neon -ftree-vectorize -mfloat-abi=softfp -ffast-math -fsingle-precision-constant" CXXFLAGS="-O3 -mcpu=cortex-a8 -mfpu=neon -ftree-vectorize -mfloat-abi=softfp -ffast-math -fsingle-precision-constant" cmake -DCMAKE_BUILD_TYPE=release -DENABLE_STRICT_COMPILATION=off -DCMAKE_TOOLCHAIN_FILE=../PandoraToolchain.cmake -DPKG_CONFIG_EXECUTABLE=/home/nils/pandora-dev/arm-2009q3/bin/arm-none-linux-gnueabi-pkg-
20100823 20:10:24< Ivanovic> config -DSDL_CONFIG=/home/nils/pandora-dev/arm-2009q3/usr/bin/sdl-config -DPREFERENCES_DIR=wesnoth-1.9_userdata ../wesnoth-1.9.x
20100823 20:10:41< Ivanovic> as you see, basically a perfectly normal linux build
20100823 20:10:46< Ivanovic> i start the game via a startscript
20100823 20:11:23-!- fangism [~davidfang@gilbert.achronix.com] has joined #wesnoth-dev
20100823 20:11:24< Ivanovic> http://pastebin.com/mqJrWt98
20100823 20:11:39< alink> Ivanovic: ok, sry, you probaby also need to run with "--validcache", otherwise it will detected file timestamp change and try to refresh the cache
20100823 20:11:54< alink> this assumes that --validcache works as advertised
20100823 20:12:18< Ivanovic> ugh, validcache would be a bad option for the pandora build "in general" since users can download updated versions
20100823 20:12:38< alink> Ivanovic: this is just for testing this issue
20100823 20:12:52< alink> the part <alink> Ivanovic: have you tried creating the cache on PC and copy it to the pandora, to see if it's just cache creation, not cache itself ?
20100823 20:13:14< Ivanovic> lets first see if it starts with trunk now
20100823 20:13:19< alink> Ivanovic: to better identify what need to be optimized
20100823 20:13:24< _jbx_> does anyone know who is working on the tool-tips / sidebar and how this slipped under the radar?     * Damage in sidebar now also take account of ToD and leadership
20100823 20:13:46< Ivanovic> alink: my current speculation is/was that the problem might be the part where the cache is written to the disc
20100823 20:13:53< alink> _jbx_: me
20100823 20:13:54< Ivanovic> that there is a complete copy created to do so
20100823 20:14:28< _jbx_> alink: ok, I'm not really liking this idea.  It doesn't take into account resists anyway, and will only create more confusion
20100823 20:14:52< _jbx_> alink: here is the first sample of it http://forums.wesnoth.org/viewtopic.php?f=6&t=31335
20100823 20:15:49< alink> _jbx_: resists is opponent specific, that seems different enough
20100823 20:17:31< alink> _jbx_: the point was to spare that, for example, on an underground map or during the whole ToD turn, damage info always need to be translated by the human
20100823 20:17:33< _jbx_> alink: well, you also make it difficult to calculate what your damage value will be when ToD changes back to normal (on your next turn for example).  It would be better to keep the values consistent though all ToD's and let the player do the math
20100823 20:18:37< _jbx_> alink: I can clearly see you are trying to make it easier for players, but sometimes that doesn't work as planned :)
20100823 20:19:05< Ivanovic> alink: creating the cache atm, this will take a while...
20100823 20:19:17< alink> _jbx_: good point, but note that bas value is still visible in tooltip
20100823 20:19:40< Ivanovic> alink: looking good, it went over the problematic part, currently building terrain rules
20100823 20:19:48< _jbx_> alink: perhaps switching the tooltip to reveal current ToD damage might make more sense?
20100823 20:20:20< alink> _jbx_: well it's also to make things easier for new players (who often forget/miss the ToD effect), but I expected old players to be a bit disturbed by the change
20100823 20:21:01< alink> it always hard to please both groups
20100823 20:21:10< _jbx_> alink: I actually think it might hurt new players more, who see 8-3 damage, then it changes on them
20100823 20:21:16-!- Mat[1] [~Mat@d-72-9-4-145.cpe.metrocast.net] has joined #wesnoth-dev
20100823 20:22:37< Mat[1]> What tags must I incorporate to make an ai-controlled leader invisible, especially so that the screen doesn't scroll annoyingly to him each ai turn?  Think survival.
20100823 20:22:42< alink> it changes when ToD changes, seems to make sense to me. And seeing it changes once, will make it clear that it change with ToD
20100823 20:24:02< alink> _jbx_: also not the damage info always used trait effect, that's mainly why I wanted to unify this
20100823 20:24:45-!- Mat[1] [~Mat@d-72-9-4-145.cpe.metrocast.net] has quit [Client Quit]
20100823 20:25:27< _jbx_> alink: sure, as I said I understand what you are trying to do.  I think swapping the tool-tip to reveal current damage and keep the actual damage fixed is a simple solution that pleases everyone. 
20100823 20:27:20< alink> _jbx_: mmh not really, I think the most common case is to need the current value, so we shouldn't need to use the tooltip each time for getting it
20100823 20:27:58< alink> also it defect the purpose to make the change more visible (a new user should check the tooltip twice between turns to notice it)
20100823 20:29:26< alink> _jbx_: but it's true that easily get the value for next ToD turn is a problem :-/
20100823 20:30:07< Ivanovic> alink: any reason for not having "local time of day light" active by default?
20100823 20:30:32-!- silene [~plouf@wesnoth/developer/silene] has joined #wesnoth-dev
20100823 20:31:07< alink> Ivanovic: yes, experimental stuff, early prototype with hacked code, may more memory if you play a lot of cave map
20100823 20:31:19< Ivanovic> hmm, okay
20100823 20:31:38< Ivanovic> currently i play with the idea of turning off "animated map" since the water really has a speed impact
20100823 20:31:47< alink> I only added in preferences because it was improved enough shortly before the release
20100823 20:32:01< CIA-35> mordante * r45975 /trunk/src/ai/composite/goal.cpp: Initialize all members.
20100823 20:32:06< CIA-35> mordante * r45976 /trunk/src/game_display.cpp: Initialize all members.
20100823 20:32:08< CIA-35> mordante * r45977 /trunk/src/unit_animation.cpp: Added constructor to initialize all members.
20100823 20:32:11< alink> Ivanovic: trunk also optimize water anim a lot, and it's a simple WML change
20100823 20:32:11< CIA-35> mordante * r45978 /trunk/src/ (image.cpp persist_context.cpp persist_var.cpp): Strip trailing whitespace.
20100823 20:33:34< Ivanovic> alink: this *is* trunk and there is a significant slowdown on the pandora when scrolling
20100823 20:34:34< alink> Ivanovic:  animate map and scrolling are not really related, but indeed they add up
20100823 20:36:12< alink> ah and now that I think of it, a huge patch of water could cause a useless double redraw, from animation and copy&pasting screen for scrolling. Maybe we can improve this
20100823 20:37:27< Ivanovic> would be really good
20100823 20:37:47< _jbx_> alink: my final thought on this would be that new players need to learn the game anyway (because they are new) and older players are not going to like this change.  I would rather cater to our veterans than to our newbies, and this is the type of change that will cause a lot of debate.   So please respond to that thread with your reasoning behind the idea.  That way you get all the hate mail :P
20100823 20:39:48< alink> _jbx_: ok thanks, I will do that later today. I need to think more about your point and must go afk now (I just joined to warn about the shroud stuff :-/).
20100823 20:40:53< alink> _jbx_: but not that early 1.9.x are for experimenting such changes and they need to be released when users need to try it for some time (always needed when changing habits)
20100823 20:41:54< alink> Ivanovic: ok so the cache stuff works now? good, that recent trunk change helps. But unfortunately, this is WiP  WML syntax, so I don't think it's good for release :-/
20100823 20:42:07< Ivanovic> yes, it helps
20100823 20:42:22< alink> I mean it will need to wait 1.9.1
20100823 20:42:23< Ivanovic> i am about to test the cave stuff, trying to :n through httt
20100823 20:43:29< alink> on cave map is ok, just use few more terrain image, my concern is that it doesn't clear yet the lightmaps between different cave map
20100823 20:44:07-!- rigved [~rigved@116.72.163.243] has joined #wesnoth-dev
20100823 20:44:18-!- rigved [~rigved@116.72.163.243] has left #wesnoth-dev []
20100823 20:44:21< alink> so any heavy case played during a session may waste memory
20100823 20:44:33< alink> :refresh fix that, though
20100823 20:45:00< zookeeper> alink, for what it's worth, i agree with jb on the sidebar damage values
20100823 20:46:17< alink> zookeeper: yes he has valid points, and I initially expected some opposition to that, but it never came up and then i forget about it (it was in really early 1.9.0 dev)
20100823 20:47:01< alink> anyway, I will try to improve that later, afk now, bbl
20100823 20:47:29< zookeeper> sure, i don't mind it in an early dev release. anyway, i think it's mostly confusing for newbies since there's no obvious indicator as to why their units' damage display keeps changing.
20100823 20:49:44< Ivanovic> *this* is facinating
20100823 20:50:02< Ivanovic> wesnoth uses a whole less memory (residentual) if the cache was already existing
20100823 20:50:18< Ivanovic> 180MB when the cache has to be created vs 110MB when it already exists
20100823 20:56:28< mordante> I'm off again, bye
20100823 20:56:49-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20100823 21:03:19-!- gabba [~gabba@wesnoth/developer/gabba] has joined #wesnoth-dev
20100823 21:03:38< boucman>  hey gabba
20100823 21:03:53< gabba> hey boucman
20100823 21:03:54< boucman> I couldn't find the thread with feedback about WB, could you point it to me plz ?
20100823 21:04:51< gabba> forgot to set it up :P, guess I got lost in the forum looking for early wb reports
20100823 21:05:03< gabba> think MP development is the right place for that thread?
20100823 21:07:11< boucman> probably, yes...
20100823 21:07:32< gabba> ok, I'll make that thread now
20100823 21:07:39< boucman> anything major on the WB front ? I didn't see any comments at all (which is suprising I would have expected some reactions)
20100823 21:07:50< boucman> gabba: post a link in the announcement thread after that
20100823 21:09:23< gabba> boucman: sure. Well, there was one major bug reported (which I'm very annoyed not to have fixed before 1.9.0): you can still cheat by planning multiple attacks w/ the whiteboard and executing them
20100823 21:09:43< Ivanovic> alink: i think for the pandora build of 1.9.1 i will ship a default config that does turn animated map off and the local tod stuff one
20100823 21:09:45< Ivanovic> s/one/on
20100823 21:09:55< boucman> gabba: don't worry too much it's annoying but that's what dev versions are for...
20100823 21:10:01< boucman> any comments on the feature itself ?
20100823 21:10:06< Ivanovic> alink: but yeah, current trunk seems to work and hopefully longer gameplay is possible, too
20100823 21:10:29< Ivanovic> (animated maps cause a cpuload of almost 85% on isle of the damned
20100823 21:10:38< Ivanovic> when watching a mainly water area
20100823 21:10:50< gabba> boucman: I still have to check for new posts from yesterday, but up to now commenters said it was "cool" and not much else :P
20100823 21:11:08< Ivanovic> when watching a non animated area it is just 35%
20100823 21:11:32< boucman> gabba: that's a rather good sign for a revolutionary feature :P
20100823 21:18:37-!- Latsni [~Latsni@p54B627A5.dip.t-dialin.net] has quit [Quit: Leaving.]
20100823 21:23:57< eleazar> boucman: is this name proper for transitions?
20100823 21:23:58< eleazar> void-n-ne-se-nw.png
20100823 21:24:16< eleazar> I'm confused about the order
20100823 21:24:24< eleazar> that transitions must be listed
20100823 21:24:28< boucman> seems good
20100823 21:24:41< boucman> it's supposed to be clockwise starting north...
20100823 21:24:46< boucman> is it a transition ?
20100823 21:25:22< eleazar> yes
20100823 21:25:30< zookeeper> well that nw doesn't make any sense there
20100823 21:25:51< eleazar> it's 4 contiguous sides
20100823 21:26:07< zookeeper> hmh, right
20100823 21:26:22< eleazar> i'm having trouble getting it to show up in game
20100823 21:26:40< zookeeper> i'd have expected it to be void-nw-n-ne-se.png but i don't really know about that stuff
20100823 21:27:11< eleazar> but it's not common, so i can't tell if it's broken or just isnt' needed at the moment
20100823 21:27:34< zookeeper> having the directions list un-contiguous looks suspicious to me, i don't recall seeing that before (but i might remember wrong, too)
20100823 21:27:48< boucman> zookeeper: good point, let me check
20100823 21:28:43< eleazar> there are few 4-sided transitions
20100823 21:29:33< eleazar> and i couldn't find an example terrain that had all 6 contiguous posibilities
20100823 21:31:06< boucman> eleazar: zookeeper is right, transitions only work for contiguous edges...
20100823 21:31:21< eleazar> that is contiguous
20100823 21:31:28< boucman> in your case it would put bot a n-ne-se transition and a nw transition
20100823 21:31:36< zookeeper> no it's not, nw should be s in that list
20100823 21:31:57< eleazar> zookeeper: ?
20100823 21:31:57< boucman> zookeeper: eleazar is right it's nw-n-ne-se
20100823 21:32:09< boucman> aaargh
20100823 21:32:16< boucman> I got it wrong again :P
20100823 21:32:39< zookeeper> eleazar, void-n-ne-se-nw.png isn't contiguous because nw isn't clockwise from se
20100823 21:32:57< eleazar> so it doesn't neccesarily start with "n"?
20100823 21:33:08< zookeeper> i guess. i don't know, ask boucman ;P
20100823 21:33:09< boucman> void-nw-n-ne-se.png should work
20100823 21:33:19 * eleazar testing...
20100823 21:33:51< boucman> eleazar: no, transitions are based on rotations, so it's ordered clockwise, contiguously
20100823 21:33:55< boucman> my bad for earlier
20100823 21:33:59< eleazar> k
20100823 21:34:23< boucman> bridges are clockwise starting by n, but I had to use completely different system internally
20100823 21:34:48< boucman> hopefully, once alink has added his new system I'll finish the bridge macros and document all that properly
20100823 21:38:17< eleazar> arrg
20100823 21:38:42< eleazar> can't get a 4 sided transition in the game
20100823 21:39:29< boucman> still not working ?
20100823 21:39:48< eleazar> i can't get it on the screen
20100823 21:40:03< eleazar> it is rarely used
20100823 21:40:43< boucman> anybody good at the log system ? I think it's time to add a log-domain for terrain :P
20100823 21:41:23< boucman> alink: around ?
20100823 21:41:55< eleazar> well, i can test the name thing on 3 sided terrains
20100823 21:42:10< gabba> Whiteboard feedback: http://forums.wesnoth.org/viewtopic.php?f=15&t=31338
20100823 21:43:45< Soliton> gabba: is there a reason switching to wb is shift+p and not p?
20100823 21:45:56< gabba> Soliton: good question, let me check
20100823 21:47:16< gabba> Soliton: it's because p is already assigned to open preferences. you can easily remap that though
20100823 21:47:50< Soliton> preferences is ctrl+p
20100823 21:47:50< gabba> at some point I need to submit a proposal to remap some keys so the WB has hotkeys that make sense
20100823 21:48:15< Soliton> yeah.
20100823 21:49:05< gabba> Soliton: ah you're right, it has the IF_APPLE_CMD_ELSE_CTRL macro in hotkeys.cfg
20100823 21:49:21< gabba> I'll move it to P then, why not.
20100823 21:49:29< Soliton> i thought that might have confused you. :-)
20100823 21:51:45< alink> eleazar: if it's for shroud, then shroud transition uses their own transition system. I can look into it if needed. But as I said before, a lot of shroud cases are impossible to get without WML hacks. IMHO, we should improve the WML part before checking if new images are needed
20100823 21:52:37< alink> eleazar: basically normal units move always clear "flower" of 7 hexes, so you can't have isolated cleared hex
20100823 21:53:11< eleazar> 4 sided shroud isn't my only concern
20100823 21:53:30< eleazar> alink: also i plan to repurpose these for your "light" transitions
20100823 21:53:55< eleazar> assuming you add support for more than the basic 6 transitions
20100823 21:53:55< alink> eleazar: Gambit had these weird cases because he manually edited the shroud via WML
20100823 21:55:04< eleazar> ahh, shroud is always +1 move distance
20100823 21:55:05< alink> eleazar: yes I wish to try to allow that, but I must clean the current system first
20100823 21:55:10< boucman> eleazar: that would make terrain macros even heavier... so not right away :P
20100823 21:55:20< CIA-35> gabba * r45979 /trunk/data/core/hotkeys.cfg: Whiteboard: moved toggle hotkey to p instead of shift+p
20100823 21:55:39< Gambit> Yes, I was clearing a box, and a signle hex with [remove_shroud]
20100823 21:55:41< alink> eleazar: yes and the algo is: take all reachable hexes and clear all its adjacent hexes
20100823 21:56:00< alink> s/its/their
20100823 21:56:15< alink> which smooth things 
20100823 21:56:28< eleazar> So, if i'm visualizing this right 4 sides trans are imposible
20100823 21:57:08< alink> eleazar: you need WML event, test-scenario has one IIRC
20100823 21:57:10< eleazar> boucman: it would make the macros heavier to use the normal transition rules for light and shroud?
20100823 21:57:29< alink> eleazar: light and shroud will never use macros
20100823 21:57:34< eleazar> alink: I don't care about the 4 sided trans any more
20100823 21:57:41< boucman> to make more rules for non-contiguous transitions (I thought that was what you were thinking of)
20100823 21:58:02< eleazar> boucman: no that's not what i mean
20100823 21:58:21< boucman> k, forget my comment, then
20100823 21:58:34< eleazar> IIRC light currently uses just n, nw, sw, s, se, and ne transtions
20100823 21:58:51< eleazar> can't handle n+ne for instance
20100823 21:59:16< alink> eleazar: ok then yes 4 sided shroud trans are impossible to get by moving unit
20100823 21:59:41< Soliton> what about impassable terrain?
20100823 22:00:20< alink> (all reachable hex are clear, and their adjacent hexes all have 3 clear neighbours (the reachable hex and 2 other adjacent hexes) 
20100823 22:00:46< alink> Soliton: the adjacent system also clear impassable terrain 
20100823 22:01:03< alink> to allow to see that there is an impassable terrain ;)
20100823 22:02:15< Soliton> yeah, i guess the +1 vision guarantees it.
20100823 22:03:13< alink> in case it was not clear the +1 vision comes from the algo : clear all reachable hexes and clear all their adjacent hexes
20100823 22:04:11< eleazar> also BTW next time someone is messing around in that area, i'd appreciate it if void, fog, and shroud images could each be moved into their own subfolder-- since IIRC the file location is hardcoded, i can't move the files
20100823 22:05:04< alink> eleazar: I think i defined that in game_config.cfg for 1.9.0
20100823 22:05:24< alink> shroud_prefix="terrain/void"
20100823 22:05:44< alink> then transition suffixes are added to that
20100823 22:05:52< eleazar> ah, you are right
20100823 22:05:52< alink> +.png
20100823 22:05:55-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has joined #wesnoth-dev
20100823 22:06:03-!- dtiger [~dtiger@dynamic-vpdn-93-125-68-39.telecom.by] has quit [Remote host closed the connection]
20100823 22:06:04< eleazar> i'll can do it then...
20100823 22:06:09< alink> glad it was not useless :-)
20100823 22:06:42< eleazar> all the hours i spend scrolling through the terrain folders
20100823 22:06:52< eleazar> it's worth my time to make em tidy
20100823 22:07:24< alink> boucman: needed me for something? (even if I need to eat something soon)
20100823 22:08:21< boucman> alink: not that I recall of, I think I mentionned your name once but not specifically to attact your attention.
20100823 22:08:47< alink> [21:42:19] <boucman> alink: around ?
20100823 22:09:08< alink> and [21:35:44] <boucman> hopefully, once alink has added his new system I'll finish the bridge macros and document all that properly
20100823 22:09:29< alink> does new system is the variations stuff ?
20100823 22:09:44-!- mysticX [gdwunl@rbi0332.giga-dns.com] has joined #wesnoth-dev
20100823 22:10:13< boucman> aaah yes :)
20100823 22:10:45< alink> because to be clear, the single hex trick used for water, is old never used WML syntax
20100823 22:10:58< boucman> I was wondering if you had a logdomain for terrain rules.... it might be a good idea to have one to log missing images. Usually not usefull but a great tool to debug terrain images
20100823 22:11:24< alink> boucman: agreed I often used that to debug it
20100823 22:11:27< boucman> I don't understand your question...
20100823 22:12:38< alink> I mean I introduced 2 things yesterday : new variations syntax to optimize random terrains, and a new WML macro to optimize water anim
20100823 22:12:42< boucman> I don't think it's an old syntax... I've use the other one to be able to merge bigger than hex and normal ones easily, but if we know that it fits in hex, I see no problem to use the image in tile syntax
20100823 22:13:21< boucman> I was supprised it made such a difference though... maybe I should rething my macros to use it more
20100823 22:14:13-!- thespaceinvader [~chatzilla@wesnoth/artist/thespaceinvader] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]]
20100823 22:14:43< alink> for animations it make a difference because we animate 7 hexes instead of 1 and since all anim are randomly shifted a lot of hexes are invalidated at a high rate (7x times more if surrounded by water)
20100823 22:15:30< boucman> alink: hmm, in unit animation we had a "trick" to prevent had
20100823 22:15:50< boucman> I had hardcoded that if an image is exactly 72x72 then I assume it fits in hex.
20100823 22:16:26< alink> boucman: yes but we improved that later with the in_hex stuff, scanning the image
20100823 22:16:55-!- fangism1 [~davidfang@gilbert.achronix.com] has joined #wesnoth-dev
20100823 22:17:12-!- Sapient [~patrickp@wesnoth/developer/sapient] has joined #wesnoth-dev
20100823 22:17:26< alink> but it's different with terrains, since terrains are hex-cutted
20100823 22:17:35< boucman> true... could we port that in terrain ? in_hex is a property of the image that could be filled at load time, I guess
20100823 22:17:57< alink> I plan to try somehting there soon
20100823 22:18:17< CIA-35> eleazar * r45980 /trunk/data/core/images/terrain/ (20 files): new shroud trans which are more consistant in the degree they cover a hex.
20100823 22:18:22< alink> my plan is to check it at the hex-cutting phase
20100823 22:18:28-!- fangism [~davidfang@gilbert.achronix.com] has quit [Ping timeout: 245 seconds]
20100823 22:18:52< alink> but then I need to update the animation so dummy image doesn't trigger update
20100823 22:19:11-!- Johannes13__ [~Johannes@pD950116B.dip0.t-ipconnect.de] has joined #wesnoth-dev
20100823 22:19:32< alink> the main problem will be that we load image at first draw or when scrolling, but we build terrain rules and animation at loading phase
20100823 22:19:58< alink> so I may need to preload them, which also could improve the first (slower) scrolling
20100823 22:20:15< boucman> hmm
20100823 22:20:47< boucman> loading all images at rule parsing time might be costy... 
20100823 22:21:14< boucman> i'm a bit confused... we don't need the image size before we invalidate IIRC, which is at first draw... and at this point we need to load the image anyway
20100823 22:22:42-!- Johannes13_ [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 240 seconds]
20100823 22:23:12< alink> there are 3 phases: load rules, apply rules on map, draw hexes. We load images at phase 3, but phase 2 already build animation with dummy image
20100823 22:24:15< boucman> what do you mean with dummy image, I thought that building animations only created locators, and the images were only loaded on draw... or is it different for terrains
20100823 22:25:28< alink> I call dummy image and image only having transparent pixels after hex-cutting the multi-hex images. the engine currently has no way to detect it, and thus consider them part of animation
20100823 22:25:35< alink> s/and/an
20100823 22:25:53< boucman> aaah, ok.
20100823 22:26:03< alink> anyway preload some image before the first scrolling will not change much (except maybe for shrouded map) and will allow other optimization that I always wanted to try
20100823 22:26:10< boucman> got it, indeed the way we animate multi hex is completely different
20100823 22:26:49< alink> mainly clear the base multi-hex image when we finished to hex-cutting in pieces
20100823 22:28:52< alink> all these things are related, so I wish to find a way helping most of them
20100823 22:29:18-!- Crab_ [~Crab_@wesnoth/developer/crab] has joined #wesnoth-dev
20100823 22:29:39< Crab_> Gambit: hi, I can somehow help ?
20100823 22:29:49< alink> afk now, really too hungry, bbl
20100823 22:30:35< CIA-35> eleazar * r45981 /trunk/data/ (95 files in 5 dirs): moved fog and void images into eponymous subfolders.
20100823 22:31:12< Gambit> Crab_: I think I had a question for you a while ago, but I've since learned that it was rather nonsensical anyways.
20100823 22:31:20< Crab_> what question ?
20100823 22:31:31< zookeeper> silene or anyone else interested: [variable] name=some_unset_variable boolean_equals=no [/variable] evaluates to false to me, although unset variables should boolean equal false, not true
20100823 22:31:53< zookeeper> (that was with a build a few days old, though, in case it's been fixed since...)
20100823 22:32:07< zookeeper> Sapient too ^
20100823 22:32:46-!- alink [~alink@wesnoth/developer/alink] has quit [Remote host closed the connection]
20100823 22:33:25< Sapient> maybe there should be some boolean tests added to the test-scenario
20100823 22:34:06< Sapient> [if] [variable] name=is_it_broken
20100823 22:35:02< Sapient> zookeeper: I also confirmed that setting canrecruit to "1" no longer works
20100823 22:36:37-!- MeccaGod [~majs@h14n4fls31o279.telia.com] has quit []
20100823 22:36:47< Sapient> I'm not sure whether to consider that a bug or not, but at least there should be a note made in the WML Forum about backwards incompatibility
20100823 22:37:26< zookeeper> the test scenario is so bloated that i don't really use it for anything else than testing single events i can just paste in temporarily
20100823 22:37:41< Soliton> maybe vaguely related: village gold is always 1 no matter the settings.
20100823 22:37:43< zookeeper> i bet it's just full of all sorts of ancient cruft
20100823 22:41:24-!- Daltx [~~@unaffiliated/daltx] has quit [Changing host]
20100823 22:41:24-!- Daltx [~~@unaffiliated/op] has joined #wesnoth-dev
20100823 22:46:56-!- Elvish_Pillage2 [~eli@71-10-224-192.dhcp.oxfr.ma.charter.com] has joined #wesnoth-dev
20100823 22:49:49-!- gabba [~gabba@wesnoth/developer/gabba] has quit [Quit: Leaving.]
20100823 22:50:45-!- gabba [~gabba@wesnoth/developer/gabba] has joined #wesnoth-dev
20100823 22:54:07< zookeeper> Sapient, btw, one thing i very often find myself wanting to do is pick locations (or units) which are closest to some location, or farthest away east, or something like that. it's pretty ackward to do.
20100823 22:55:00< zookeeper> like, currently i'd like to find the location closest to x,y which is 12 hexes away from any player unit
20100823 22:55:21< zookeeper> but to do that i have to use at least one loop
20100823 22:56:14< Sapient> yeah, we need more location utilities
20100823 22:56:34< Sapient> distance(loc1,loc2)   direction(loc1,loc2)
20100823 22:56:55< Sapient> then we need a list sorting utility
20100823 22:57:09< Sapient> those two features combined would give what you want
20100823 22:58:25< Crab_> Sapient: note that there are multiple distances, e.g. distance with 'dummy' cost calculator, or distance by pathfinding of some unit
20100823 22:58:48-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20100823 22:59:23< zookeeper> Sapient, i don't have much suggestions, but one idea i had was to allow some kind of a [closest_to] tag, which, if used in [store_locations], would only store the matches which are closest to the given location
20100823 22:59:56< boucman> hmm, would lua be abetter tool than wml for that ? maybe usung lua to implement the tags...
20100823 23:01:07< eleazar> boucman: can you tell me the difference between the three types of whiteboard arrows?
20100823 23:01:19< Crab_> boucman: note that the 'find first location FOO satisfying conditions A and B' is quite tricky to calculate fast\
20100823 23:02:33< boucman> eleazar: gabba would be better than me at that... IIUC you have the one that follows the cursor, the one for already planned actions that are unselected and the one for the planned action that is selected
20100823 23:02:44< boucman> but since there are two sets that are similar i'm not 100% sure
20100823 23:03:10-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20100823 23:03:35< Crab_> boucman: since there are a lot of corner cases depending on the conditions. for example, in 'find me locations closest to X,Y which satisfy A', it might make sense to get the locations 'closest to x,y' first, one, by one, check A, and stop at first location which satisfies A
20100823 23:03:42< gabba> eleazar: well, you have the standard arrows, and two type of highlighted ones
20100823 23:03:54< boucman> Crab_: true, so c++ helpers might be usefull. However combining such formulas would probably be simpler in lua, and doing a complete SQL like syntax to join these would be hell
20100823 23:04:15< eleazar> gabba: what's the intended difference between the two highlighted ones
20100823 23:04:26< eleazar> i mean under what conditions are they used?
20100823 23:04:44-!- fangism1 is now known as fangism
20100823 23:05:00< gabba> eleazar: when you mouseover a planned move, it's arrow receives the "focus" arrow, while the other moves of the same unit only receive the "highlighted" one which is a bit less bright/conspicuous
20100823 23:06:09< Crab_> boucman: the problem is not in the language, c++ or lua (lua implementation can be faster than C++ one if the algorithm is better suited to a task). I see the problem in defining a language extension(s) which will allow doing the things that zookeeper wants to while being easy to use correctly
20100823 23:06:09< eleazar> ok
20100823 23:06:45< boucman> Crab_: yes, that's what I meant by combining c++ functions would be easier in lua than wml
20100823 23:06:54< Crab_> yes, I agree
20100823 23:07:01< eleazar> gabba: do they use the "magenta is the team color" list or something different to color-shift?
20100823 23:07:31< gabba> eleazar: yes, it uses "magenta is the team color"
20100823 23:10:05< gabba> eleazar: feel free to use other colors, even though I guess there are not many colors left after you exclude magenta and all the team colors
20100823 23:10:42< Sapient> Crab_, boucman: I see no reason why it can't be exposed to WML
20100823 23:11:04< Sapient> the WML language has been getting left behind on some of the newer features for no good reason, IMO
20100823 23:11:21< Crab_> Sapient: the simplest of those possible extensions is the 'slf+custom_sort' extension. e.g. , sort all locations returned by SLF according to a custom comparator
20100823 23:11:30< Sapient> a good example is the ability to write to the chat log
20100823 23:11:55< Crab_> Sapient: where 'custom comparator' is, for example, distance()  function you've talked about
20100823 23:12:03< Sapient> Crab_: WML doesn't use functions, it uses events
20100823 23:12:17-!- grzywacz [~grzywacz@wesnoth/developer/grzywacz] has quit [Remote host closed the connection]
20100823 23:12:22< Crab_> Sapient: the problem is, it's a lot harder to write a 'custom comparator' using wml, while it's easy in lua.
20100823 23:12:37< Sapient> not at all
20100823 23:13:02< Crab_> Sapient: ok, what might be the WML syntax to specify 'the location should be sorted by the distance to location (10,15) ascending' ?
20100823 23:13:11< Soliton> Sapient: there is a [chat] patch waiting to get applied.
20100823 23:13:13< Sapient> in particular it's quite simple to handle the most common case, sorting by an attribute
20100823 23:13:48< Sapient> the case of sorting by a comparator event is only slightly more complex
20100823 23:14:00< Crab_> Sapient: how it might look like ?
20100823 23:14:47< Crab_> for comparison, it can be  comparator="distance(x,y,10,15)" using embedded lua code.
20100823 23:14:53< eleazar> hey, startup is FAST!
20100823 23:15:11 * eleazar just compiled his first post release build
20100823 23:15:13< Sapient> {SORT_LIST my_list rank_value ()}
20100823 23:15:21< Sapient> that would be an attribute sort
20100823 23:15:37< Sapient> {SORT_LIST my_list rank_value rank_event}
20100823 23:15:48< Sapient> that would be a comparator sort
20100823 23:16:10< Sapient> where rank_event sets a variable rank_value
20100823 23:16:37-!- Greywhind [~Greywhind@fw-34-18.cs.brown.edu] has quit [Quit: leaving]
20100823 23:17:00< Sapient> so in the simple case, you have written one line
20100823 23:17:20< Sapient> hardly what I would consider "complex"
20100823 23:19:17< zookeeper> basically the functionality i'd like is [closest_to](SUF)[/closest_to] and [farthest_from](SUF)[/farthest_from] in SLF, and which would match only the locations closest to or farthest from the given locations out of all the locations that otherwise would match.
20100823 23:19:36< zookeeper> of course that doesn't need to be the syntax, but you get the idea (but i guess that was clear already)
20100823 23:19:39< Crab_> Sapient: for that to work, we need to expose more stuff which is available in lua only to wml
20100823 23:20:21< Crab_> Sapient: for example, there is find_path which allows lua code to get a distance by pathfinding
20100823 23:21:01< Crab_> Sapient: which is, roughly, a proxy for pathfind::a_star_search
20100823 23:21:32< Sapient> finding the shortest path is inherently complex, there's no getting around it
20100823 23:21:59< Sapient> but it shouldn't be hard to come up with sensible defaul behavior
20100823 23:22:00< Crab_> Sapient: yes. and you're right in saying that WML lags behind lua for such features
20100823 23:22:03< Gambit> Crab_: For getting globals from a particular side; that works for later joiners right?
20100823 23:22:33< Gambit> For example in an RPG if someone quits you want to kill their character, give an observer the team, and load the new player's character?
20100823 23:22:47< Crab_> Gambit: yes, it loads by side number
20100823 23:23:06< Sapient> Crab_: the main reason for that is because I haven't been active lately, so I can blame myself
20100823 23:23:29< Sapient> but it is starting to annoy me when I hear talk about WML being outdated and unusable
20100823 23:24:23< Sapient> maybe that is a good motivation for me to finish setting up my development environment ;)
20100823 23:24:30-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100823 23:24:33< Crab_> Sapient: well, with each day, more and more c++-based wml tags are replaced with lua-based wml tags. so, wml keeps its place even in a lua-based world
20100823 23:24:35-!- noy [~Noy@wesnoth/developer/noy] has quit [Client Quit]
20100823 23:25:30-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20100823 23:25:48< Crab_> Sapient: and I think that it's best to not duplicate the functionality in wml and lua, but to allow easier cases to be specified by wml tags (possibly, with lua function names referenced in it), and let lua code handle complex things which are harder to write in wml.
20100823 23:26:31< Crab_> for example, it's easy enough to make a [find_path] wml tag, backed by lua find_path function, which will save path and/or cost to wml variables
20100823 23:27:45< Crab_> Sapient: what do you think ?
20100823 23:27:56< Sapient> I don't really have much interest in writing Lua at the moment
20100823 23:28:19-!- wesbot changed the topic of #wesnoth-dev to: 135 bugs, 289 feature requests, 15 patches | logs: http://irclogs.wesnoth.org | Don't paste on IRC! Use a pastebin: http://wesnoth.pastebin.com | http://imagebin.org
20100823 23:28:34< zookeeper> i've only heard some clueless UMC people say that WML is outdated and unusable :p
20100823 23:28:36< Sapient> but I see your point about duplication of efforts
20100823 23:29:47< Sapient> Crab_: I think the needs of the Lua users and the needs of the WML users are different, although there is some similarity
20100823 23:30:24< Sapient> so a WML version of find_path shouldn't necessarily be a dumb wrapper for the C++ method
20100823 23:30:35< Sapient> it should provide ease and convenience
20100823 23:31:02< Sapient> (just making an example)
20100823 23:31:03< Crab_> yes, I agree with that. but, more and more users start to use both WML and lua.
20100823 23:31:36< Crab_> maybe it's because they want to do something in WML, ask in the forums, and are told 'use lua', yes.
20100823 23:31:45< zookeeper> lua seems great for implementing new tags and other WML functionality, but i don't like to use it in scenarios as such
20100823 23:32:17-!- fangism1 [~davidfang@gilbert.achronix.com] has joined #wesnoth-dev
20100823 23:33:03-!- fangism [~davidfang@gilbert.achronix.com] has quit [Ping timeout: 245 seconds]
20100823 23:34:20< zookeeper> the fact that i can implement new (action) tags myself using a sane language and that i don't need to be able to compile to do so is totally great
20100823 23:34:54< Sapient> Crab_: have no fear, if I write something in C++ which is accessible to WML but not Lua there is no doubt that it will be added to Lua by some other developers who are currently active ;)
20100823 23:35:18< Crab_> and your C++ code might be removed in the process, as it happened with some C++ wml tags.
20100823 23:35:52< Sapient> yes
20100823 23:36:12< Sapient> that area isn't really my concern
20100823 23:36:23< Crab_> is't it better to discuss with silene beforehand, to see what should be done in C++ and what is better to code in lua ?
20100823 23:36:25-!- silene [~plouf@wesnoth/developer/silene] has quit [Quit: Leaving.]
20100823 23:36:31< Crab_> (to avoid the need to recode things)
20100823 23:36:42< Crab_> zookeeper: ok, returning to implementation.. .say, we have a SLF with [slf] [closest_to]A[/closest_to] B [/slf], then, we evaluated A and got, say, 10 locations
20100823 23:36:53< Sapient> not really, different developers like to work on different areas
20100823 23:36:56< Crab_> ok
20100823 23:37:42< Sapient> as long as neither of us introduces bugs we should get along fine with each other ;)
20100823 23:37:43< zookeeper> Crab_, if you got a question, try to ask it fast :P
20100823 23:38:12< Crab_> zookeeper: now, the problem - the most straightforward way of continuing the process is 'evaluate B, then check AxB pairs for being closest. so, it'll be slow. will it be ok if that [closest_to] is slow ?
20100823 23:38:35< zookeeper> Crab_, yes, if it's documented to be slow
20100823 23:38:40< Crab_> ok, good.
20100823 23:39:04< Crab_> then I think that that straightforward and dump implementation can be coded fairly quickly
20100823 23:39:22< zookeeper> for example, i already figured that it'd probably be pretty slow if you used it in a moveto event filter for example: when checking if the SLF matches, the game would need to actually check every matching location and see if it's the closest one...
20100823 23:40:00< Crab_> zookeeper: there are faster ways to make it work, for those particular cases. but they are harder to code.
20100823 23:40:07< Crab_> another issue - closest-by-coordinates works strange in maps-with-caves. is this ok ?
20100823 23:40:30< Crab_> e.g., you can be 3 hexes away from enemy unit, but 30 turns away by pathfinding, because of the cave walls.
20100823 23:42:31< zookeeper> Crab_, yeah, it's fine
20100823 23:42:41< Crab_> ok, good, no more questions then.
20100823 23:43:52< zookeeper> if you need to treat impassable terrains differently then you're probably using some kind of a radius thing already and can use [filter_radius]
20100823 23:44:31< zookeeper> but i really gotta go now, bye
20100823 23:44:34-!- zookeeper [~l@wesnoth/developer/zookeeper] has quit []
20100823 23:46:18-!- Crab_ [~Crab_@wesnoth/developer/crab] has quit [Quit: Leaving.]
20100823 23:51:21-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Quit: Leaving.]
--- Log closed Tue Aug 24 00:00:21 2010