--- Log opened Sat Dec 17 00:00:32 2011
20111217 00:06:35-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds]
20111217 00:10:20-!- mjs-de [~mjs-de@g224177047.adsl.alicedsl.de] has quit [Remote host closed the connection]
20111217 00:14:12-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20111217 00:22:28-!- Vorpal [~AnMaster@unaffiliated/anmaster] has quit [Ping timeout: 276 seconds]
20111217 00:27:16-!- LordNasty [~NaSTy@93-43-141-188.ip92.fastwebnet.it] has quit [Ping timeout: 240 seconds]
20111217 00:27:43-!- LordNasty [~NaSTy@93-43-141-188.ip92.fastwebnet.it] has joined #wesnoth-dev
20111217 00:28:25-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20111217 00:29:18-!- Johannes13 [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 255 seconds]
20111217 00:29:27-!- Johannes13 [~Johannes@p4FDEB259.dip.t-dialin.net] has joined #wesnoth-dev
20111217 00:29:28-!- Johannes13 [~Johannes@p4FDEB259.dip.t-dialin.net] has quit [Changing host]
20111217 00:29:28-!- Johannes13 [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20111217 00:33:18-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20111217 00:39:45< shadowmaster> anonymissimus: 1 is perfectly fine in my book
20111217 00:39:56< shadowmaster> 2 doesn't apply
20111217 00:40:03< shadowmaster> 3 is the point
20111217 00:40:49-!- LordNasty [~NaSTy@93-43-141-188.ip92.fastwebnet.it] has quit [Ping timeout: 252 seconds]
20111217 00:42:53-!- LordNasty [~NaSTy@93-43-141-188.ip92.fastwebnet.it] has joined #wesnoth-dev
20111217 00:45:02-!- Sytyi1 [~Sytyi@40-6-133-95.pool.ukrtel.net] has joined #wesnoth-dev
20111217 00:46:11-!- Sytyi [~Sytyi@150-26-133-95.pool.ukrtel.net] has quit [Ping timeout: 252 seconds]
20111217 00:46:36< anonymissimus> shadowmaster: I dunno about your CPU (probably XXXXXL) but for some of use a rebuilt lasts more than just some minutes
20111217 00:47:07< anonymissimus> and because of 2) I'd need to do that every time I dcommit or rebase
20111217 00:47:10-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 276 seconds]
20111217 00:47:25< shadowmaster> not really
20111217 00:48:02< shadowmaster> just change your build configuration to pass -DDEBUG_IN_NETWORKED always
20111217 00:48:16< shadowmaster> that doesn't require changing any versioned files unless your build system sucks
20111217 00:48:45< anonymissimus> and 3) means that I could use it, but the many mp authors which dion't compile can't
20111217 00:49:45< shadowmaster> yeah, that's not a hueg change in regards to 1.8, is it?
20111217 00:51:35< anonymissimus> and I guess that my build system sucks then, since that requires modifying the project files
20111217 00:51:50 * shadowmaster nods
20111217 00:52:05< anonymissimus> although I could probably force the change upon everyone else using it :)
20111217 00:52:53-!- MeccaGod [~majs@host189-199.bornet.net] has quit []
20111217 00:52:59< shadowmaster> cmake -DCXXFLAGS='-DDEBUG_IN_NETWORKED' or so, and scons extra_flags_release='-DDEBUG_IN_NETWORKED'
20111217 00:53:10< shadowmaster> no changes to the build scripts whatsoever
20111217 00:53:43< shadowmaster> either way, don't worry, I wasn't serious about fixing it myself
20111217 00:53:48< anonymissimus> sure, but I dunno how to do that with the project files
20111217 00:53:50< shadowmaster> I don't fix things for MP users in general
20111217 00:53:58< anonymissimus> but it would be useful at times
20111217 01:05:14-!- Sytyi1 [~Sytyi@40-6-133-95.pool.ukrtel.net] has quit [Quit: Leaving.]
20111217 01:07:04< anonymissimus> after some more investigation I get debug mode in networked 1.8 as well
20111217 01:07:35< anonymissimus> one needs just to use the new mp looby, since the old one sets game_config::debug to false in its constructor
20111217 01:07:55< anonymissimus> by using the new one this is bypassed
20111217 01:07:59< shadowmaster> ah, that thing no serious player would use
20111217 01:08:19< anonymissimus> doesnt matter, I can cheat!!! I can cheat!!!
20111217 01:08:58 * shadowmaster raises eyebrow
20111217 01:12:42-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20111217 01:13:39-!- PolarPanda [~quassel@unaffiliated/peterporty] has quit [Read error: Connection reset by peer]
20111217 01:14:21-!- PolarPanda [~quassel@unaffiliated/peterporty] has joined #wesnoth-dev
20111217 01:15:26-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20111217 01:21:47-!- horon [~horon@nttkyo316066.tkyo.nt.ngn2.ppp.infoweb.ne.jp] has joined #wesnoth-dev
20111217 01:26:00-!- anonymissimus [~chatzilla@HSI-KBW-078-042-163-105.hsi3.kabel-badenwuerttemberg.de] has quit [Quit: done building targets]
20111217 01:34:24-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20111217 01:42:05< shadowmaster> AI0867: this thing with the saves index generation is rather frustrating, and I heard you were going to look into it the other day yourself. Have you discovered anything that I haven't?
20111217 01:42:56< shadowmaster> I only discovered the criterion used to determine whether a save is a replay isn't valid anymore (savegame.cpp:886), and instead of looking into the [replay_start] node contents the code should be looking into [replay]
20111217 01:43:52< shadowmaster> the correct criterion is handled in gamestatus.cpp:369 onwards  under the implementation of extract_summary_from_config()
20111217 01:44:19< shadowmaster> it is correct there, but that code is only used when the saves index file isn't available
20111217 01:45:07< shadowmaster> I'm too tired to check now whether I can make the code in savegame.cpp aware of the [replay] node since it doesn't seem to receive the full save WML
20111217 01:45:35-!- atomicbomb [~quassel@61.94.183.107] has joined #wesnoth-dev
20111217 01:46:23< shadowmaster> I think I fixed the case where a start-of-scenario save is reported as being at Turn /" though
20111217 01:48:45-!- Johannes13 [~Johannes@unaffiliated/johannes13] has quit [Ping timeout: 268 seconds]
20111217 01:49:23< shadowmaster> except I can't reproduce it now
20111217 01:50:30< shadowmaster> hm, I¿ll postpone the fix for "Turn /" until I can reproduce the bug again
20111217 01:51:08< shadowmaster> A better question would be what do we need the saves index for.
20111217 01:51:26< shadowmaster> it's been broken repeatedly in the past
20111217 01:52:11< loonycyborg> I wonder that too. Considering they appear in load dialog without problems even when I copy them manually.
20111217 01:54:06< shadowmaster> A few bits of code would suggest that at one point it was planned to manage the saves list in a tree-like fashion, assigning turn saves to a parent start-of-scenario save whenever possible
20111217 01:54:34< shadowmaster> The .parent attribute is stored in the children nodes in the saves index for that purpose at least.
20111217 01:54:50< loonycyborg> Hm. That would be nice.
20111217 01:55:15< loonycyborg> There are often lots of per-turn saves polluting the view..
20111217 01:56:58< shadowmaster> the code to show the leader baseframe on the left panel int he Load Game dialog is still there too
20111217 01:58:50< shadowmaster> the saves index still stores a .leader (the leader unit id, not unit type id) and .leader_image attribute
20111217 01:59:47< shadowmaster> hm, when and why could that have been broken...
20111217 02:01:10< shadowmaster> it's probably not worth it to resurrect it at this point since there's support for multiple leaders for a single side now, but it's concerning that there's still code handling this feature -- that means someone inadvertently broke it and didn't *care*
20111217 02:04:06< shadowmaster> and that was probably during the 1.7.x cycle
20111217 02:06:42< loonycyborg> I wonder why does it show minimap previews only for some saves in the load dialog for me.
20111217 02:07:25< shadowmaster> it's supposed to show them only if the map is valid, and the scenario doesn't have shroud or fog enabled for the currently playing side in the current turn
20111217 02:08:13< shadowmaster> the map is valid only if the terrain definition and symbol_image for every single hex can be found; if you are playing a UMC campaign with custom terrains, save the game and quit, you won't see the minimap on the Load Game dialog unless the data for that campaign is loaded already
20111217 02:08:49< shadowmaster> which usually involves starting a new game with that campaign or loading one of your saved games, so that's a little problematic
20111217 02:10:04< shadowmaster> I forgot I wanted to make map validation optional in the Load Game dialog this cycle, and the required changes are too intrusive to perform them at this point
20111217 02:16:06-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20111217 02:20:09-!- vultraz [~chatzilla@124.109.10.221] has joined #wesnoth-dev
20111217 02:33:58-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has joined #wesnoth-dev
20111217 02:37:39-!- lansiir [~oldtopman@71-33-144-66.hlrn.qwest.net] has joined #wesnoth-dev
20111217 02:37:39-!- lansiir [~oldtopman@71-33-144-66.hlrn.qwest.net] has quit [Changing host]
20111217 02:37:39-!- lansiir [~oldtopman@unaffiliated/oldtopman] has joined #wesnoth-dev
20111217 02:37:56-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has quit [Read error: Connection reset by peer]
20111217 02:39:11-!- lansiir is now known as oldtopman
20111217 02:41:26-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has quit [Client Quit]
20111217 02:41:45-!- oldtopman [~oldtopman@71-33-144-66.hlrn.qwest.net] has joined #wesnoth-dev
20111217 02:41:50-!- oldtopman [~oldtopman@71-33-144-66.hlrn.qwest.net] has quit [Changing host]
20111217 02:41:50-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has joined #wesnoth-dev
20111217 02:46:08-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has quit [Read error: Connection reset by peer]
20111217 02:46:13-!- lansiir [~oldtopman@71-33-144-66.hlrn.qwest.net] has joined #wesnoth-dev
20111217 02:46:13-!- lansiir [~oldtopman@71-33-144-66.hlrn.qwest.net] has quit [Changing host]
20111217 02:46:13-!- lansiir [~oldtopman@unaffiliated/oldtopman] has joined #wesnoth-dev
20111217 02:49:12-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has quit [Quit: crimson_penguin]
20111217 02:50:04-!- alink [~alink@wesnoth/developer/alink] has quit [Remote host closed the connection]
20111217 03:03:26-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20111217 03:10:49-!- Elvish_Pillager [~eli@dhip-149.coburn.residences.colby.edu] has quit [Ping timeout: 276 seconds]
20111217 03:22:25-!- loonybot [~loonybot@wesnoth/bot/loonybot] has quit [Remote host closed the connection]
20111217 03:30:58-!- Appleman1234 [~Appleman1@ppp59-167-222-56.qld.adsl.internode.on.net] has quit [Ping timeout: 276 seconds]
20111217 03:51:00-!- horon [~horon@nttkyo316066.tkyo.nt.ngn2.ppp.infoweb.ne.jp] has quit [Quit: Leaving...]
20111217 03:56:52-!- enchilado [~enchilado@defocus/yummy/enchilado] has quit [Quit: leaving]
20111217 04:05:25-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 248 seconds]
20111217 04:07:48-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20111217 04:53:58-!- Gambit [~quassel@wesnoth/developer/grickit] has quit [Remote host closed the connection]
20111217 05:28:05-!- oldtopman [~oldtopman@71-33-144-66.hlrn.qwest.net] has joined #wesnoth-dev
20111217 05:28:05-!- oldtopman [~oldtopman@71-33-144-66.hlrn.qwest.net] has quit [Changing host]
20111217 05:28:05-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has joined #wesnoth-dev
20111217 05:28:41-!- lansiir [~oldtopman@unaffiliated/oldtopman] has quit [Read error: Connection reset by peer]
20111217 05:32:50-!- noy [~Noy@wesnoth/developer/noy] has quit [Ping timeout: 252 seconds]
20111217 05:38:03-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20111217 05:52:19-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has quit [Quit: oldtopman has left the house]
20111217 05:54:12-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Ping timeout: 255 seconds]
20111217 05:55:38-!- enchilado [~enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20111217 05:56:33-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20111217 06:11:02-!- enchilado [~enchilado@defocus/yummy/enchilado] has quit [Ping timeout: 244 seconds]
20111217 06:15:51-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit []
20111217 06:20:52-!- PolarPanda [~quassel@unaffiliated/peterporty] has quit [Read error: Connection reset by peer]
20111217 06:24:44-!- enchilado [~enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20111217 06:35:03-!- Blueblaze [~Blueblaze@adsl-99-148-247-0.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20111217 06:35:38-!- vultraz [~chatzilla@124.109.10.221] has quit [Ping timeout: 240 seconds]
20111217 06:50:38-!- vultraz [~chatzilla@124.109.10.221] has joined #wesnoth-dev
20111217 06:57:50-!- Blueblaze [~Blueblaze@adsl-99-148-247-0.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20111217 07:57:22-!- Appleman1234 [~Appleman1@ppp59-167-222-56.qld.adsl.internode.on.net] has joined #wesnoth-dev
20111217 08:44:16< Ivanovic> moin
20111217 08:45:09-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev
20111217 08:53:24-!- loonycyborg [loonycybor@wesnoth/developer/loonycyborg] has quit [Quit: ZNC - http://znc.in]
20111217 08:55:40-!- loonycyborg [loonycybor@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20111217 09:15:17-!- loonycyborg [loonycybor@wesnoth/developer/loonycyborg] has quit [Ping timeout: 248 seconds]
20111217 09:22:42-!- loonycyborg [loonycybor@wesnoth/developer/loonycyborg] has joined #wesnoth-dev
20111217 09:57:09-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has quit [Quit: Leaving]
20111217 09:58:28-!- csarmi [csarmi@217.197.182.29] has joined #wesnoth-dev
20111217 10:06:25-!- Sytyi [~Sytyi@40-6-133-95.pool.ukrtel.net] has joined #wesnoth-dev
20111217 10:24:54-!- MeccaGod [~majs@host189-199.bornet.net] has joined #wesnoth-dev
20111217 10:29:11-!- atomicbomb [~quassel@61.94.183.107] has quit [Remote host closed the connection]
20111217 10:37:33-!- crimson_penguin [~ben@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev
20111217 10:40:48-!- mjs-de [~mjs-de@d185133.adsl.hansenet.de] has joined #wesnoth-dev
20111217 10:46:38< CIA-85> espreon * r52301 /trunk/data/lua/wml-tags.lua: Made it so that [store_side] also stores current_player.
20111217 10:48:34< CIA-85> espreon * r52302 /trunk/changelog: Updated the changelog.
20111217 10:57:14-!- Vorpal [~AnMaster@unaffiliated/anmaster] has joined #wesnoth-dev
20111217 11:27:26-!- atomicbomb [~quassel@125.160.84.66] has joined #wesnoth-dev
20111217 11:54:11-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Quit: ...]
20111217 12:19:45-!- Sytyi [~Sytyi@40-6-133-95.pool.ukrtel.net] has quit [Quit: Leaving.]
20111217 12:22:21-!- atomicbomb [~quassel@125.160.84.66] has quit [Remote host closed the connection]
20111217 12:40:36-!- vultraz [~chatzilla@124.109.10.221] has quit [Read error: Connection reset by peer]
20111217 12:48:02-!- vultraz [~chatzilla@124.109.10.221] has joined #wesnoth-dev
20111217 12:49:18-!- vultraz [~chatzilla@124.109.10.221] has quit [Client Quit]
20111217 12:49:45-!- vultraz [~chatzilla@124.109.10.221] has joined #wesnoth-dev
20111217 12:56:39-!- vultraz [~chatzilla@124.109.10.221] has quit [Quit: ChatZilla 0.9.87 [Firefox 8.0.1/20111120135848]]
20111217 12:57:01-!- vultraz [~chatzilla@124.109.10.221] has joined #wesnoth-dev
20111217 13:03:55-!- loonybot [~loonybot@46.138.81.125] has joined #wesnoth-dev
20111217 13:03:55-!- loonybot [~loonybot@46.138.81.125] has quit [Changing host]
20111217 13:03:55-!- loonybot [~loonybot@wesnoth/bot/loonybot] has joined #wesnoth-dev
20111217 13:09:07-!- atomicbomb [~quassel@125.164.89.170] has joined #wesnoth-dev
20111217 14:09:09-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20111217 14:09:29-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has left #wesnoth-dev []
20111217 14:09:45-!- Alarantalara [~alarantal@CPEc0c1c09e8055-CM00252eac6d62.cpe.net.cable.rogers.com] has joined #wesnoth-dev
20111217 14:27:36< CIA-85> mordante * r52303 /trunk/src/image.cpp: Fix some compiler warnings.
20111217 14:36:46-!- mordante [~mordante@roadie.xs4all.nl] has joined #wesnoth-dev
20111217 14:36:52-!- mordante [~mordante@roadie.xs4all.nl] has quit [Changing host]
20111217 14:36:52-!- mordante [~mordante@wesnoth/developer/mordante] has joined #wesnoth-dev
20111217 14:37:00< mordante> servus
20111217 14:38:04< mordante> anonymissimus does the sorting bug only happen -std=c++0x?
20111217 14:42:21< Alarantalara> I need some help with SDL concerning https://gna.org/bugs/?18319
20111217 14:43:31< Alarantalara> There is a fix for the upstream problem that starts it, but the new version of SDL sets a different video mode that causes is_neutral() to return false (the RGB mask is exactly the opposite of what is expected)
20111217 14:45:10< mordante> Alarantalara, I might be able to have a look later, first want to catch up with the log
20111217 14:46:26< Ivanovic> mordante: also make sure to catch up with the ML(s)
20111217 14:46:38< Ivanovic> (regarding fosdem)
20111217 14:47:07< mordante> I will, just wanted to read the log first
20111217 14:51:01< Ivanovic> :)
20111217 14:52:23< mordante> Ivanovic, no objections against a release tomorrow, hope to fix the issue Espreon found with the portraits
20111217 14:55:15 * mordante wonders why we still get a lot of feature like commits :-(
20111217 14:58:09-!- Elvish_Pillager [~eli@dhip-149.coburn.residences.colby.edu] has joined #wesnoth-dev
20111217 15:02:48-!- Gambit [~quassel@wesnoth/developer/grickit] has joined #wesnoth-dev
20111217 15:09:57< mordante> Espreon, did you check whether the old player_name is not cached anywhere and thus causes bugs
20111217 15:10:15< mordante> also what prevents two players to use the same name?
20111217 15:27:34< mordante> Ivanovic, still thinking about a topic for the FOSDEM :-(
20111217 15:32:58-!- happygrue [~quassel@wesnoth/developer/wintermute] has joined #wesnoth-dev
20111217 15:34:45< Ivanovic> why not talk about the steps and amount of work required with a change in the "gui engine"?
20111217 15:34:55< Ivanovic> like the transition from gui1 to gui2
20111217 15:35:05< Ivanovic> steps you undertook, problems you encountered, stuff like this
20111217 15:35:13< Ivanovic> might be interesting and relevant for other projects and devs, too
20111217 15:37:39-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20111217 15:44:11< mordante> yeah have been thinking about that topic as well, just thinking whether it's interesting enough
20111217 15:45:21-!- horon [~horon@nttkyo316066.tkyo.nt.ngn2.ppp.infoweb.ne.jp] has joined #wesnoth-dev
20111217 15:45:47-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Remote host closed the connection]
20111217 15:45:56-!- Johannes13 [~Johannes@p4FDEE802.dip.t-dialin.net] has joined #wesnoth-dev
20111217 15:45:56-!- Johannes13 [~Johannes@p4FDEE802.dip.t-dialin.net] has quit [Changing host]
20111217 15:45:56-!- Johannes13 [~Johannes@unaffiliated/johannes13] has joined #wesnoth-dev
20111217 15:46:13< mordante> added myself to the FOSDEM wiki
20111217 15:46:22< Ivanovic> mordante: i think so
20111217 15:46:51-!- atomicbomb [~quassel@125.164.89.170] has quit [Remote host closed the connection]
20111217 15:47:14< mordante> will ponder a bit longer about it
20111217 15:47:32< Alarantalara> mordante: that would be very nice. It's probably the most complained about Apple bug at the moment
20111217 15:47:37-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20111217 15:47:41< Alarantalara> and I don't know the entire graphics system
20111217 15:48:08< mordante> Ivanovic, is there still time to add the NativeClient patch?
20111217 15:48:55< Ivanovic> i'd say so
20111217 15:51:37-!- oldtopman [~oldtopman@unaffiliated/oldtopman] has joined #wesnoth-dev
20111217 15:58:39-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20111217 16:00:59-!- alink [~alink@wesnoth/developer/alink] has joined #wesnoth-dev
20111217 16:01:07< alink> hi
20111217 16:07:11< CIA-85> alink * r52304 /trunk/src/ (7 files): 
20111217 16:07:11< CIA-85> Big optimization of terrain rendering by removing all empty images.
20111217 16:07:11< CIA-85> This make screen update (zooming, ToD changes, etc) several times faster,
20111217 16:07:11< CIA-85> gain some fps (esp. for complex map), optimize some animated terrains
20111217 16:07:11< CIA-85> (stop refreshing empty parts) and spare some memory (sometimes, not clear yet)
20111217 16:07:11< CIA-85> Also remove the noise in :layers.
20111217 16:07:12< CIA-85> Note: This is an improved version of r46936 from the OGL branch
20111217 16:13:41< alink> mordante: thanks for r52303. I didn't have warning here, what flag did you use to catch that ?
20111217 16:18:51-!- timotei [~timotei@wesnoth/developer/timotei] has joined #wesnoth-dev
20111217 16:27:53< mordante> loonycyborg, could you have a look at the scons part of this patch? https://github.com/eugenis/wesnoth-nacl-build/blob/master/wesnoth-1.9.11-nacl.patch
20111217 16:28:49< mordante> alink, no problem, -Wmissing-declarations catches it
20111217 16:28:55< loonycyborg> mordante: I already did. Or did it change since I posted about that on the ML?
20111217 16:29:02< alink> mordante: ok thanks
20111217 16:29:39< CIA-85> ivanovic * r52305 /trunk/ (12 files in 11 dirs): updated Hungarian translation
20111217 16:29:57< mordante> loonycyborg, not sure, I can also just commit it and then you look afterwards
20111217 16:36:00< loonycyborg> mordante: I noticed some changes, but it's still fine afaict.
20111217 16:36:20< fendrin> hi
20111217 16:36:34< mordante> loonycyborg, ok thanks
20111217 16:36:37< mordante> hi fendrin 
20111217 16:41:45< shadowmaster> alink: These optimizations really make FADE_TO_BLACK sequences (which uses [color_adjust]) much more bearable in my campaign! :p
20111217 16:42:10< shadowmaster> (Comparing against 1.9.12, the effect is now very smooth.)
20111217 16:42:55< alink> shadowmaster: good :-) I was aiming for that (and other things)
20111217 16:43:04-!- Aethaeryn [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev
20111217 16:44:09< alink> I didn't test FADE_TO_BLACK, but the timing improvement for simple ToD turn change was very promising
20111217 16:53:25< mordante> Alarantalara, there are no issues in windowed mode?
20111217 16:53:45< Alarantalara> using SDL 1.2.13, that is correct
20111217 16:54:14< Alarantalara> the upstream fix is in 1.2.14, which I need to get to eventually because of deprecated libraries
20111217 16:54:38< mordante> so using 1.2.14 saves all your issues?
20111217 16:54:59< Alarantalara> unfortunately, there SDL_SetVideoMode returns a surface that does not match the expected neutral surface
20111217 16:55:10< Alarantalara> so I get neutral surface assertions everywhere
20111217 16:55:17< mordante> ah ok
20111217 16:55:34< mordante> I think I might have a patch for that...
20111217 16:57:25< mordante> Alarantalara, can you test this patch http://paste.debian.net/149475/ ?
20111217 16:59:44< Alarantalara> certainly
20111217 17:00:06< zookeeper> Alarantalara, do you think you'll want to handle committing mefisto's tropical forest(s)?
20111217 17:00:20< shadowmaster> mordante: is that sure to not create much runtime overhead?
20111217 17:00:42< Alarantalara> zookeeper: sure, no problem
20111217 17:01:20< zookeeper> Alarantalara, ah, great. i'll gladly leave it to you then.
20111217 17:01:21< alink> shadowmaster: none
20111217 17:02:05< zookeeper> Alarantalara, jetrel already gave those the green light, in case you're wondering.
20111217 17:02:27< mordante> shadowmaster, why do you expect too much runtime overhead?
20111217 17:02:40< Alarantalara> mordante: patch has no effect
20111217 17:03:32< alink> mmh SDL_SetVideoMode shouldn't return something matching a neutral surface, maybe we forgot a conversion somewhere and it just triggers with Mac
20111217 17:04:09< alink> Alarantalara: what error/glitch do you have without the asserts?
20111217 17:04:17< mordante> Alarantalara, is there GDB on your Mac ?
20111217 17:04:25< Alarantalara> yes
20111217 17:04:27< shadowmaster> mordante: dunno, one extra call.
20111217 17:04:40< alink> shadowmaster: see the 'static'
20111217 17:04:59< Alarantalara> without the assert everything is very slow due to SDL doing conversion everywhere between formats and blurred surfaces are miscoloured
20111217 17:05:10< Alarantalara> and yes, I have GDB
20111217 17:05:15< mordante> can you run wesnoth in gdb and post a backtrace from the crash (then leave the debugger open)
20111217 17:05:53< Alarantalara> http://pastebin.com/sGNTeV4h
20111217 17:06:04-!- Appleman1234 [~Appleman1@ppp59-167-222-56.qld.adsl.internode.on.net] has quit [Ping timeout: 252 seconds]
20111217 17:06:24< Alarantalara> Rmask is 0xff00
20111217 17:06:24< shadowmaster> alink: yeah, I know the test result is cached
20111217 17:07:30< Alarantalara> mordante: backtrace is at link above
20111217 17:08:03< mordante> yeah I saw it, looking at it
20111217 17:09:26< mordante> shadowmaster, I don't think one call for a static causes too much overhead also don't think the overhead of accessing a class member instead of a hardcoded const will be too much
20111217 17:10:47< shadowmaster> I said "much", not "too much". :p
20111217 17:11:56< shadowmaster> It just seemed suspicious to add an extra (probably non-inlineable) function call to a function that's probably already used too much.
20111217 17:13:36< alink> even if it was not static, it's not important anyway, it's only done before image manipulation (at least a 72x72 loop) and some image conversion/copy, which always dwarfs everything else ;)
20111217 17:15:03< alink> btw does someone already try if disabling all our asserts give some perfomance improvement? maybe there is a slow one somewhere
20111217 17:17:12< mordante> Alarantalara, which svn version are you running exactly (there have been some recent commits and want to use the same line numbers)
20111217 17:18:43< mordante> never tried without asserts, but might help the speed
20111217 17:18:43< Alarantalara> 52304 + the patch you sent
20111217 17:19:10< shadowmaster> alink: there seem to be a few custom assert-like constructs floating around and I'm not sure those can be disabled. Otherwise, I doubt anyone has tried (I've been wanting to do that, but lengthy rebuilds aren't particularly pleasing)
20111217 17:20:01-!- happygrue [~quassel@wesnoth/developer/wintermute] has quit [Ping timeout: 252 seconds]
20111217 17:20:06< shadowmaster> I imagine that alternatively you could redefine assert and friends to add benchmarking information to them.
20111217 17:20:40< alink> yeah I was thinking that, like having a counter for each one or something like that
20111217 17:21:58< shadowmaster> (of course, people keep coming up with excuses to alter frequently used header files and I end up compiling Wesnoth from scratch pretty much every single day...)
20111217 17:23:18< shadowmaster> mordante: have you tried running the test suite lately, btw?
20111217 17:25:06-!- Blueblaze [~Blueblaze@adsl-99-148-247-0.dsl.hstntx.sbcglobal.net] has joined #wesnoth-dev
20111217 17:29:44< mordante> shadowmaster, yes you mean that somebody broke them again
20111217 17:30:07< shadowmaster> I don't know. I haven't checked in detail, but one test fails.
20111217 17:30:29< shadowmaster> I keep forgetting exactly how to do that.
20111217 17:32:21-!- enchilado [~enchilado@defocus/yummy/enchilado] has quit [Quit: BBL->]
20111217 17:32:56-!- MeccaGod [~majs@host189-199.bornet.net] has quit []
20111217 17:37:29< CIA-85> mordante * r52306 /trunk/ (29 files in 6 dirs): 
20111217 17:37:29< CIA-85> Added the NativeClient port code.
20111217 17:37:29< CIA-85> Thanks to Evgeniy Stepanov for providing the patch.
20111217 17:40:26-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has quit [Read error: Connection reset by peer]
20111217 17:44:28-!- happygrue [~quassel@24-216-213-91.dhcp.stls.mo.charter.com] has joined #wesnoth-dev
20111217 17:44:29-!- happygrue [~quassel@24-216-213-91.dhcp.stls.mo.charter.com] has quit [Changing host]
20111217 17:44:29-!- happygrue [~quassel@wesnoth/developer/wintermute] has joined #wesnoth-dev
20111217 17:46:09-!- shadowm_laptop [ignacio@wesnoth/developer/shadowmaster] has joined #wesnoth-dev
20111217 18:00:04< mordante> Alarantalara, can you type the following two lines?
20111217 18:00:06< mordante> f 6
20111217 18:00:09< mordante> p /x *surf.surface_.resource->format
20111217 18:04:09< Alarantalara> mordante: http://pastebin.com/p99V5SHW
20111217 18:05:26-!- anonymissimus [~chatzilla@HSI-KBW-078-042-163-105.hsi3.kabel-badenwuerttemberg.de] has joined #wesnoth-dev
20111217 18:09:13< mordante> Alarantalara, and also the output of:  p /x get_neutral_pixel_format()
20111217 18:10:39< Alarantalara> $1 = 0x110e1c0
20111217 18:10:58< anonymissimus> mordante: "does the sorting bug only happen -std=c++0x?": YES! if you have a c++0x binary and some addons installed it'd be nice you make a quick check whether you can reproduce it
20111217 18:12:19< mordante> Alarantalara, hmm did you also add the parens at the end? (I would expect a similar output as before)
20111217 18:12:38< Alarantalara> I did
20111217 18:12:55< fendrin> alink: zooming is indeed faster now. nice.
20111217 18:14:48< mordante> anonymissimus, I only have a c++0x capable compiler in a virtual machine without X so I can't run Wesnoth compiled with c++0x
20111217 18:15:27< mordante> Alarantalara, odd, can you then try to move the line "static SDL_PixelFormat neutral = get_neutral_pixel_format();"
20111217 18:15:48< mordante> it should be before the function definition, then please recompile and run wesnoth
20111217 18:16:41< Alarantalara> I'm not quite sure what you're asking with "move the line"
20111217 18:16:45< Alarantalara> move it where?
20111217 18:17:34< mordante> Alarantalara, before the line  "inline bool is_neutral(const surface& surf)"
20111217 18:17:52< Alarantalara> so you want it out of the function?
20111217 18:18:18< mordante> yup, then I can see how a neutral surface looks on your system
20111217 18:19:04-!- Sytyi [~Sytyi@46.211.193.126] has joined #wesnoth-dev
20111217 18:19:28< Alarantalara> http://pastebin.com/8wT464TA
20111217 18:21:53< Alarantalara> mordante: For your terror, SDL_SetVideoMode returns a surface matching the first one in windowed mode on an intel processor, and one matching the neutral surface in fullscreen and on ppc processors
20111217 18:23:19< mordante> joy :-(
20111217 18:23:44< Alarantalara> so of course the crash only happens in windowed mode
20111217 18:25:03-!- Sytyi [~Sytyi@46.211.193.126] has quit [Ping timeout: 252 seconds]
20111217 18:36:11-!- [Relic] [~relic@99-58-54-211.lightspeed.milwwi.sbcglobal.net] has joined #wesnoth-dev
20111217 18:41:33-!- MeccaGod [~majs@host189-199.bornet.net] has joined #wesnoth-dev
20111217 18:42:25< loonycyborg> anonymissimus: What exactly is the sorting bug again?
20111217 18:43:32< loonycyborg> I do get an empty item in campaigns list in c++0x mode
20111217 18:44:02< loonycyborg> gcc 4.5.3/linux
20111217 18:45:26< loonycyborg> Didn't get around to test it before because uncommitted changes are getting in the way..
20111217 18:53:02-!- Vorpal [~AnMaster@unaffiliated/anmaster] has quit [Ping timeout: 268 seconds]
20111217 19:07:10< anonymissimus> loonycyborg: yes, sounds like the bug
20111217 19:08:35< anonymissimus> ok thanks so its not limited to me and since std::stable_sort is supposedly used in a lot of projects I can wait until it gets reported and fixed by someone else :)
20111217 19:10:03< shadowm_laptop> theoretically that kind of bugs should be caught by the C++ STL tests
20111217 19:10:17< shadowm_laptop> (because libstdc++ has its own regression tests, right?)
20111217 19:10:39< CIA-85> anonymissimus * r52307 /trunk/projectfiles/CodeBlocks/ (liblua.cbp wesnoth.cbp wesnothd.cbp): 
20111217 19:10:39< CIA-85> Revert "testwise enable --std=c++0x in CodeBlocks"
20111217 19:10:39< CIA-85> This reverts commit r52275. There appears to be at least one fatal c++0x bug yet;
20111217 19:10:39< CIA-85> std::stable_sort, messing up the campaign selection menu.
20111217 19:16:56< mordante> anonymissimus, I think in general it's a bad idea to enable c++0x anywhere, to avoid accidental stuff being committed using C++11 features
20111217 19:19:05< CIA-85> soliton * r52308 /trunk/src/network.cpp: 
20111217 19:19:05< CIA-85> made sockets close-on-exec
20111217 19:19:05< CIA-85> avoids leaking FDs, for example on graceful wesnothd restarts
20111217 19:23:49-!- negusnyul [~negusnyul@1F2E9BC5.dsl.pool.telekom.hu] has joined #wesnoth-dev
20111217 19:25:16-!- stikonas [~and@wesnoth/translator/stikonas] has quit [Ping timeout: 240 seconds]
20111217 19:27:59< anonymissimus> shadowmaster: perhaps the devs of these libraries tenf to forget that they have tests, use them only rather occasionally so that their test sometimes have compilation errors, and so use them even less than they should and don't catch their bugs ?
20111217 19:28:47< shadowmaster> or they are too lazy to check with gcc's C++11 support
20111217 19:29:55< anonymissimus> mordante: regarding our tests; visual leak detecter showed *a lot* of memory leaks for test_gui2.cpp
20111217 19:30:24< anonymissimus> it didnt for hte other tests (or only a few)
20111217 19:30:43< shadowmaster> libstdc++ requires tests for new code: http://gcc.gnu.org/onlinedocs/gcc-4.6.2/libstdc++/manual/manual/appendix_contributing.html#list.patches
20111217 19:32:57< mordante> or read, especially the bold text http://gcc.gnu.org/projects/cxx0x.html
20111217 19:33:31< mordante> anonymissimus, I know it leaks a lot, I don't care about leaks in the unit tests
20111217 19:38:26-!- Octalot [~noct@host109-145-74-254.range109-145.btcentralplus.com] has quit []
20111217 19:39:38-!- arnem [~quassel@host-89-242-107-181.as13285.net] has joined #wesnoth-dev
20111217 19:42:12-!- arnem [~quassel@host-89-242-107-181.as13285.net] has left #wesnoth-dev []
20111217 19:42:39-!- arnem [~quassel@host-89-242-107-181.as13285.net] has joined #wesnoth-dev
20111217 19:49:34< mordante> Alarantalara, can you try this patch instead of the previous one? http://paste.debian.net/149494/
20111217 19:54:34< Alarantalara> that patch produces the same error
20111217 19:54:42-!- negusnyul [~negusnyul@1F2E9BC5.dsl.pool.telekom.hu] has quit [Remote host closed the connection]
20111217 19:54:50-!- negusnyul [~negusnyul@1F2E9BC5.dsl.pool.telekom.hu] has joined #wesnoth-dev
20111217 19:59:17< mordante> Alarantalara, :-( then I'm out of ideas at the moment
20111217 20:00:01< alink> Alarantalara: if i understand correctly, window and fullscreen have different formats on the same platform?
20111217 20:00:06< Alarantalara> yes
20111217 20:00:31< alink> and when it doesn't crash/assert you have a slow down?
20111217 20:01:00< Alarantalara> It slows down immediately and crashes the first time is_neutral_surface is checked
20111217 20:01:24< alink> because of the assert or a true crash ?
20111217 20:01:30< Alarantalara> because of an assert
20111217 20:02:04< alink> and only in fullscreen mode?
20111217 20:02:12< Alarantalara> only in windowed mode
20111217 20:02:40< alink> when starting in windowed mode ?
20111217 20:04:06< Alarantalara> no, since the only surface that is not neutral is the game window, the problem only occurs when trying to check that surface
20111217 20:04:26< Alarantalara> which happens with gui2 blur_surface
20111217 20:04:56< Alarantalara> I'm assuming that the objectives dialog is gui2
20111217 20:05:11< shadowmaster> correct
20111217 20:06:07< alink> IMHO the blur is less important we can always convert format only there
20111217 20:06:52< alink> my main question was is it caused by the switch from fullscreen to window, or does it happen even when starting in windowed mode (-w flag)
20111217 20:07:10< Alarantalara> it happens when starting in windowed mode
20111217 20:07:23< Alarantalara> starting in windowed mode then switching to fullscreen avoids the problem
20111217 20:07:50< Alarantalara> since there is no dialog that triggers the assertion before the game starts
20111217 20:09:02< alink> ok :(  because if it was from the switch, i was going to suggest a :refresh to reload image in a proper display format to try fixing the slowdown
20111217 20:10:28< Alarantalara> The best I've come up with is to redefine neutral surface in terms of the current video mode
20111217 20:10:34-!- stikonas [~and@ctv-213-164-110-84.vinita.lt] has joined #wesnoth-dev
20111217 20:10:34-!- stikonas [~and@ctv-213-164-110-84.vinita.lt] has quit [Changing host]
20111217 20:10:34-!- stikonas [~and@wesnoth/translator/stikonas] has joined #wesnoth-dev
20111217 20:10:52< Alarantalara> but they're all over the place
20111217 20:11:18< Alarantalara> and it looks like some of the algorithms depend on it
20111217 20:11:21< alink> yes that's probably not possible
20111217 20:11:54< alink> it it was known at compile time, we could use some macro voodoo
20111217 20:12:07< alink> but since it can change at runtime...
20111217 20:12:27< Alarantalara> exactly
20111217 20:13:13< alink> i think we should silence your blur problem so we could check about the slowdown (the real problem IMO)
20111217 20:13:46< Alarantalara> It's a format conversion by SDL between surface types
20111217 20:14:18< Alarantalara> it can't copy anything directly to the window but has to convert every pixel of every image that gets blitted
20111217 20:14:47< alink> yes but we are supposed to cache the  surface converted to the display format, so it should not be a problem
20111217 20:15:23< alink> unless that conversion is incorrect, then indeed SDL will reconvert at each blit
20111217 20:15:55< alink> see create_optimized_surface()
20111217 20:16:38< alink> after that the surface should have the same format as display
20111217 20:16:48< alink> it not, then it's the problem
20111217 20:16:52< alink> s/it/if
20111217 20:18:07< mordante> that's also why I'm somewhat surprised that reloading the format every time doesn't fix the issue
20111217 20:18:48< mordante> Alarantalara, could you move neutral again out of the function and post the debug information regarding neutral and surf?
20111217 20:19:24< alink> mordante: I think you were focusing on the neutral surface format, to make blur works. I am more focusing on the display format, to make blit fast
20111217 20:20:10< mordante> alink, true, I first like to figure out why the thing fails
20111217 20:20:32< mordante> if you want to silence the blur just comment out the code in canvas.cpp
20111217 20:21:10< mordante> ... maybe I have a solution for the blur as well
20111217 20:21:26< alink> do you blur directly in screen surface or in another surface?
20111217 20:22:02< mordante> Alarantalara, could you also test the original code, but change canvas.cpp
20111217 20:22:29< mordante> line 1437, change the 32 to 33 
20111217 20:22:31< alink> mmh get_surface_portion probably take a portion of screen, with the screen format, that should be converted to neutral I think
20111217 20:22:56< alink> oh it is later nvm
20111217 20:23:07< mordante> alink, yes but in that case I call another blur function that does so
20111217 20:23:19< Alarantalara> starting with the canvas one
20111217 20:23:28< mordante> ok
20111217 20:23:44< alink> mordante: yes just saw it, two functions with the same name is a bit confusing ;)
20111217 20:24:17-!- Sytyi [~Sytyi@31.129.96.212] has joined #wesnoth-dev
20111217 20:24:24< Alarantalara> the alternate blur algorithm avoids the assertion
20111217 20:24:24< mordante> alink, somehow I always thought neutral was the format of the screen, but that seems not to be true
20111217 20:24:59< mordante> Alarantalara, ok thanks, they I think I can cook a patch
20111217 20:25:10< alink> mordante: no it isn't, at least it's not always like that
20111217 20:25:55< alink> wait:    if(surf->format->BitsPerPixel == 32) {blur_surface(surf, rect, blur_depth_);    this calls the other blur function without making it neutral
20111217 20:26:12< mordante> Alarantalara, I'll use an #ifdef, do you know which set I exactly need if not I just use __APPLE__
20111217 20:26:25< mordante> alink, yes that one always directly operates on the screen
20111217 20:26:43-!- Sytyi [~Sytyi@31.129.96.212] has quit [Client Quit]
20111217 20:27:06< alink> without checking its format
20111217 20:28:22< mordante> like I said I assumed that neutral was the screen version
20111217 20:28:29< alink> ok
20111217 20:28:44< mordante> sounds logical and nobody bothered to document what neutral meant
20111217 20:29:11< Alarantalara> Alarantalara __APPLE__ covers all mac os platforms
20111217 20:29:12< mordante> maybe I should test for neutral, that might work cross-platform
20111217 20:29:19< alink> mordante: yes that also confused me when I joined the project
20111217 20:29:26< alink> mordante: so we should make "if(surf->format->BitsPerPixel == 32)" check the whole format
20111217 20:29:46< mordante> Alarantalara, I think I'll cook up another patch
20111217 20:30:10< mordante> alink, I think is_neutral() should do the trick
20111217 20:30:38< alink> yes seems enough
20111217 20:31:00< Alarantalara> and that leaves just the very slow speed
20111217 20:31:23< alink> Alarantalara: ok good now we can focus on that :)
20111217 20:33:46< Alarantalara> for comparison: SDL 1.2.13 uses ~10% of CPU while 1.2.14 uses ~75%
20111217 20:34:36-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy]
20111217 20:34:57< alink> indeed much slower
20111217 20:39:00< alink> mmh even here the frameBuffer->format seems not correct
20111217 20:43:44< alink> Alarantalara: maybe try this http://pastebin.com/ZfKF4T1A
20111217 20:45:07< alink> ah no sorry, will probably not help
20111217 20:46:47< Alarantalara> yeah, the patch fails
20111217 20:47:40-!- PolarPanda [~quassel@unaffiliated/peterporty] has joined #wesnoth-dev
20111217 21:04:25< mordante> Alarantalara, can you test this patch http://paste.debian.net/149503/ ? 
20111217 21:04:55-!- horon [~horon@nttkyo316066.tkyo.nt.ngn2.ppp.infoweb.ne.jp] has quit [Quit: Leaving...]
20111217 21:07:32< CIA-85> soliton * r52309 /trunk/src/server/server.cpp: avoid auto-kicking unregistered users
20111217 21:10:12< Alarantalara> mordante: this one is taking longer to build, it will be a few minutes
20111217 21:10:50< mordante> yeah I expected that, it touches one of the `recompile everything headers'
20111217 21:14:34< Alarantalara> The patch functions fine.
20111217 21:15:17-!- happygrue [~quassel@wesnoth/developer/wintermute] has quit [Ping timeout: 248 seconds]
20111217 21:15:32< mordante> ok, thanks for testing
20111217 21:16:39< mordante> that means https://gna.org/bugs/?18319 is fixed?
20111217 21:17:41< Alarantalara> I will say no, since the extremely low frame rate likely still keeps me from making the change to the SDL version
20111217 21:18:13< Alarantalara> It's one step closer, though :)
20111217 21:18:25< mordante> ok
20111217 21:19:12< mordante> and which SDL version is the fix exactly for 1.2.14?
20111217 21:20:02< Alarantalara> Actually, the fix is part of the SDL nightly builds
20111217 21:20:25< mordante> ok
20111217 21:21:59< Alarantalara> and it's more than one changeset, and I don't know what all of them are
20111217 21:22:29< mordante> ok, just mention it's a nightly post 1.2.14
20111217 21:25:52< alink> Alarantalara: a last shot: http://pastebin.com/uiN9mDG2   to keep the fast fullscreen format when switching to windowed (in case the reported format is wrong)
20111217 21:26:31< alink> need to start in fullscreen with "-f -t", scrolls around to load all images then switch to windowed
20111217 21:26:52< CIA-85> anonymissimus * r52310 /trunk/src/ (time_of_day.cpp time_of_day.hpp): mute some MSVC warnings
20111217 21:28:00< CIA-85> mordante * r52311 /trunk/ (4 files in 3 dirs): (log message trimmed)
20111217 21:28:00< CIA-85> Fixes crash when using fullscreen on Mac OS.
20111217 21:28:00< CIA-85> Fixes a problem when using the SDL nightly builds (post 1.2.14).
20111217 21:28:00< CIA-85> The direct blurring of the screen fails on OS X, since the surface is
20111217 21:28:00< CIA-85> not a neutral surface, so only testing for 32 bpp isn't enough. Exports
20111217 21:28:00< CIA-85> the neutral surface test from sdl_utils to have a proper test in the
20111217 21:28:01< CIA-85> canvas code.
20111217 21:29:43-!- Sytyi [~Sytyi@31.129.96.212] has joined #wesnoth-dev
20111217 21:31:35-!- Sytyi [~Sytyi@31.129.96.212] has quit [Client Quit]
20111217 21:34:41< Alarantalara> alink: with that patch it continues to behave correctly in fullscreen and poorly in windowed mode, regardless of which mode it starts in
20111217 21:35:57< alink> Alarantalara: ok, I have no more idea for now, thanks for the testing :)
20111217 21:36:51< alink> the fatc that it's maybe not fixable without patching SDL is also not very motivating
20111217 21:37:42< Alarantalara> I'm running it through a profiler right now
20111217 21:44:27< CIA-85> alarantalara * r52312 /trunk/projectfiles/Xcode/Wesnoth.xcodeproj/project.pbxproj: Update Xcode project file
20111217 21:46:27< mordante> I'm off bye
20111217 21:46:55< mordante> oh Ivanovic I've no commits pending for the next release
20111217 21:47:09-!- mordante [~mordante@wesnoth/developer/mordante] has quit [Quit: Leaving]
20111217 21:49:02-!- Appleman1234 [~Appleman1@ppp59-167-222-56.qld.adsl.internode.on.net] has joined #wesnoth-dev
20111217 21:52:04-!- Upthorn [~ogmar@c-98-208-38-181.hsd1.ca.comcast.net] has quit [Ping timeout: 240 seconds]
20111217 21:54:36< Espreon> mordante: 1. Nothing has gone wrong in the areas I've tested (the turn dialog and the status table). 2. Nothing
20111217 21:55:11-!- Blueblaze [~Blueblaze@adsl-99-148-247-0.dsl.hstntx.sbcglobal.net] has quit [Quit: Blueblaze]
20111217 21:56:51< Espreon> mordante: As far as I know, current_player is only used in those two areas.
20111217 21:58:30-!- Sytyi [~Sytyi@31.129.96.212] has joined #wesnoth-dev
20111217 22:00:57-!- Octalot [~noct@host109-145-74-254.range109-145.btcentralplus.com] has joined #wesnoth-dev
20111217 22:04:06-!- Sytyi [~Sytyi@31.129.96.212] has quit [Quit: Leaving.]
20111217 22:05:24< CIA-85> espreon * r52313 /trunk/ (12 files in 3 dirs): Set svn:eol-style.
20111217 22:16:10< Alarantalara> alink: SDL_DisplayFormatAlpha doesn't take into account the format used by the Mac OS window mode
20111217 22:16:40< Alarantalara> so when it's called in display_format_alpha(), it doesn't set it to match the video
20111217 22:16:45< Alarantalara> in windowed mode
20111217 22:17:07< Alarantalara> I'm going to try the ConvertSurface call in all cases
20111217 22:28:00< Alarantalara> alink: well that's exciting: I now get the neutral surface assertion again, this time in cut_surface as called from load_image_sub_file
20111217 22:30:23< Alarantalara> is there any reason that a locked surface must be neutral?
20111217 22:34:09< Alarantalara> alink: if I remove the assertions in lock_surface, and in const_lock_surface, and patch SDL so that SDL_DisplayFormatAlpha knows about the different video mode, then everything appears to work well
20111217 22:36:19< Alarantalara> the only problem I see at the moment is that overlays in the minimap display improperly
20111217 22:38:10< Alarantalara> and that only happens if I start a game in windowed mode
20111217 22:45:30< CIA-85> alarantalara * r52314 /trunk/data/core/ (39 files in 2 dirs): 
20111217 22:45:30< CIA-85> Replace tropical forest with Mefisto's new version
20111217 22:45:30< CIA-85> There is an unused forested-tropical-hills-tile.png image
20111217 22:45:30< CIA-85> It allows a hill tile in the editor after string freeze is over
20111217 22:48:47-!- Sytyi [~Sytyi@31.129.96.212] has joined #wesnoth-dev
20111217 22:52:09-!- Sytyi [~Sytyi@31.129.96.212] has quit [Client Quit]
20111217 22:58:13-!- timotei [~timotei@wesnoth/developer/timotei] has quit [Read error: Connection reset by peer]
20111217 23:06:02-!- gabba [~gabba@wesnoth/developer/gabba] has joined #wesnoth-dev
20111217 23:10:00< shadowmaster> ugh
20111217 23:10:21< shadowmaster> my maps have been ruined
20111217 23:10:56< shadowmaster> or alternatively, I'll need a few months to get used to the new super-dense tropical forest.
20111217 23:12:48< Alarantalara> shadowmaster: Mefisto also did a nice light palm forest, but I can't put it in until after the feature freeze
20111217 23:14:16-!- Upth [~ogmar@108-85-91-228.lightspeed.frokca.sbcglobal.net] has joined #wesnoth-dev
20111217 23:14:16-!- Upth is now known as Upthorn
20111217 23:19:05< Espreon> mordante: Of course, if you feel that such a thing should be prevented, feel free to implement the mechanism.
20111217 23:19:08< Gambit> Why are they a replacement instead of a new terrain? :(
20111217 23:21:14< Gambit> The old ones were good for beaches, and these... aren't.
20111217 23:21:20< Espreon> Alarantalara: You know, the string freeze has been thawed, so...
20111217 23:21:38< shadowmaster> Espreon: not the feature freeze
20111217 23:21:51< Gambit> Can the replacement wait until 1.11 when you can also do palms (or better yet, keep all three)
20111217 23:22:03< Espreon> shadowmaster: I thawed that.
20111217 23:22:16< Espreon> ... too... so... blargh.
20111217 23:22:18< shadowmaster> Espreon: no, you haven't
20111217 23:22:24< shadowmaster> you aren't the release manager
20111217 23:22:38< Gambit> These are too jungley to be perfect replacements for the old ones Alarantalara.
20111217 23:22:54< Espreon> Blargh...
20111217 23:23:00< shadowmaster> the agreement is that we should not add any new features or functionality unless it's strictly necessary to fix a bug. You pushed it a little too far yesterday.
20111217 23:27:17< Alarantalara> Gambit: I know. It's hard to argue with this, though: http://forums.wesnoth.org/viewtopic.php?f=23&t=23595&start=270#p513760
20111217 23:27:36-!- enchilado [~enchilado@defocus/yummy/enchilado] has joined #wesnoth-dev
20111217 23:27:42< Gambit> " these are a sound replacement for the current jungles"
20111217 23:28:06< Gambit> Right, but not for sparse beaches :)
20111217 23:28:29< Gambit> Yeah I guess there's no arguing with the art director.
20111217 23:29:06< Alarantalara> We'll all have to wait for 1.11 for some beachfront forst
20111217 23:31:26< shadowmaster> :\
20111217 23:31:49< Espreon> ... well, he did say "Naturally these don't go in until after the feature freeze thaws".
20111217 23:31:53< Espreon> So...
20111217 23:34:17< shadowmaster> I really find this rather insulting to us who make maps for add-ons.
20111217 23:34:24-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev
20111217 23:34:48< Alarantalara> do you want to revert and wait until we can put both in at once?
20111217 23:34:58< zookeeper> i can't tell if you're referring to the commit or espreon's and gambit's complaining.
20111217 23:35:47< shadowmaster> You'd probably not care either way, so it doesn't matter.
20111217 23:36:24< Espreon> Alarantalara: I do.
20111217 23:37:15< zookeeper> if for some reason the jungles really can't replace the tropical forest as it is, then i suggest using the sparse palms instead, when he finishes them
20111217 23:38:49< Gambit> But then those won't work for tropical jungles.
20111217 23:39:23< Gambit> The current terrain did a decent job at both roles.
20111217 23:39:42< zookeeper> err, i dunno why i said "sparse palms". i meant just palms, and it looks like he finished them already.
20111217 23:43:23< Alarantalara> what happened to the fog on Isle of the Damned?
20111217 23:44:27< Alarantalara> ah, it's hard only now
20111217 23:45:33-!- gabba [~gabba@wesnoth/developer/gabba] has left #wesnoth-dev []
20111217 23:55:45< Espreon> mordante: Hmmm, yeah, I guess it could be abused thanks to [message][text_input]
20111217 23:58:33< CIA-85> alarantalara * r52315 /trunk/src/sdl_utils.cpp: 
20111217 23:58:33< CIA-85> Remove assertions in lock_surface functions
20111217 23:58:33< CIA-85> A surface may be safely locked without being neutral
20111217 23:58:33< CIA-85> They also prevent bug #18319 from being resolved
20111217 23:59:10< Alarantalara> I should probably have added that the functions that need neutral surfaces should test that themselves. blur already does
--- Log closed Sun Dec 18 00:00:32 2011