#gemrb@irc.freenode.net logs for 26 Aug 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:03:02] --> raevol has joined #GemRb
[00:44:05] <-- spike411 has left IRC (Quit: Manga & anime pokec na Jabberu: manga.cz@conf.netlab.cz)
[04:17:29] --> pupnik_ has joined #GemRb
[04:22:08] <-- pupnik has left IRC (Ping timeout: 276 seconds)
[05:30:23] <-- raevol has left IRC (Quit: Leaving.)
[06:47:53] --> spike411 has joined #GemRb
[06:49:29] --> lynxlynxlynx has joined #GemRb
[06:49:30] --- ChanServ gives channel operator status to lynxlynxlynx
[08:50:05] --> SiENcE has joined #GemRb
[08:56:26] <-- SiENcE has left IRC (Ping timeout: 264 seconds)
[08:59:09] --> SiENcE has joined #GemRb
[09:03:24] <SiENcE> hey fuzzie, i checked sdlaudio again and it works. but we need to add some options (-lsdlmixer and so on) to makefile.am
[09:17:07] <lynxlynxlynx> make a patch
[09:20:35] <fuzzie> i haven't done it because, as i said, you have to put it in acinclude.m4 to do it properly - but since we don't support autotools anyway, i guess it really doesn't matter whether it works properly or not, whatever works for you is good
[10:02:36] <SiENcE> so you dont going to merge a patch for this?
[10:04:12] <fuzzie> if you have something which works for you, make a patch :)
[10:04:29] <SiENcE> http://github.com/SiENcE/gemrb/commit/f80ee78e6eed88cc1f6da5a577305c129952c4bc
[10:05:30] <SiENcE> this flags are also needed, i dont know if you already added them
[10:05:31] <SiENcE> http://github.com/SiENcE/gemrb/commit/a5ea819b213772ad6f962d8a5290279ef499c1f7
[10:06:19] <SiENcE> i also have to change the following, http://github.com/SiENcE/gemrb/commit/b22f3f6f5cf32aac2c6c47aa6eb93bd8119f1a48
[10:06:59] <SiENcE> but it depends on you if you wanna add the autogen_dingux script or change configure.in. but the two last Makefile.am changes are really needed
[10:07:20] <SiENcE> and dont break anything
[10:07:22] <fuzzie> i don't have the code here, but i thought i already applied the INCLUDES one
[10:08:34] <SiENcE> ok
[10:08:51] <SiENcE> but you didn't merged my code changes using defines?
[10:09:21] <lynxlynxlynx> no
[10:09:37] <SiENcE> i mean this http://github.com/SiENcE/gemrb/commit/1220d8f89ab4f93bfd40d9fd50f3fc0adb9b7477
[10:09:42] <SiENcE> ok
[10:09:53] <fuzzie> given we don't support autotools anyway, merging the manual link flags one doesn't seem like a problem
[10:15:47] <lynxlynxlynx> the hardcoded one doesn't look so nice though
[10:30:26] <SiENcE> maybe you can take my downscaler and use the normal way to support a lower resolution @320x240
[10:30:35] <SiENcE> specified in config file
[10:50:54] <-- SiENcE has left IRC (Ping timeout: 265 seconds)
[11:30:28] <-- spike411 has left IRC (Quit: Manga & anime pokec na Jabberu: manga.cz@conf.netlab.cz)
[11:54:43] --> SiENcE has joined #GemRb
[12:48:18] <pupnik_> SiENcE: for pandora, how about a floating opengl map view
[12:48:28] <pupnik_> that zooms out to show whole party if desired ;)
[12:49:10] <pupnik_> could implement soft-pan and zoom with acceleration too
[13:21:29] <SiENcE> pupnik_, nice idea but i don't have a pandora
[14:02:04] <pupnik_> who does...
[14:09:20] <SiENcE> before i get a pandora, i have 100 caanoo's
[14:09:22] <SiENcE> ;)
[14:21:08] <pupnik_> also: amplifier tests - what to look for etc http://www.avsforum.com/avs-vb/showthread.php?t=855865
[15:38:26] --> avenger has joined #GemRb
[15:38:26] --- ChanServ gives channel operator status to avenger
[15:38:45] <avenger> hello
[15:39:52] <avenger> fuzzie is there any game you want me to look at with IDA?
[15:40:39] <fuzzie> oh
[15:40:52] <fuzzie> i don't think so
[15:40:58] <fuzzie> i think i was wrong about the avatar removal stuff
[15:41:09] <fuzzie> there's some weirdness, but i think maybe it's just the spells
[15:42:51] <avenger> ok, then i just continue to dig in tob :)
[15:42:57] <fuzzie> that would be great :)
[15:45:44] <fuzzie> i am not managing to get things done, lots of things taking my time, maybe until Saturday, but i guess there is no rush
[15:47:14] <fuzzie> i am still happy about the sparkles, too
[15:51:31] <-- pupnik_ has left IRC (Remote host closed the connection)
[15:59:47] <avenger> now that it works, maybe you can even understand what i intended with it :)
[16:23:59] <edheldil> avenger: you have probably checked already, but JZ jumps if Z=1, e.g. in case of cmp eax, ecx it jumps if they are equal
[16:24:39] <avenger> yeah, thanks :)
[16:26:17] <edheldil> I am poring through PS:T in IDA, but it's so time-consuming to understand even a not-so-large stupid procedure and PS:T is sooo big, when viewed on the ASM level
[16:30:55] <avenger> you got some of my lists, start with those
[16:31:49] <avenger> for example, understanding the opcodes
[16:32:00] <avenger> they are easy start
[16:32:54] <avenger> btw, i wanted to go with HoW or PST as a second run
[16:33:16] <avenger> iwd2 is trickier, it has no stack frames/debug info
[16:38:40] <fuzzie> not any of the binary versions?
[16:57:27] <avenger> what do you mean?
[16:59:38] <avenger> iwd2 is compiled with msvc just like the other games, but it is compiled in 'release' mode not like all the rest. it is noticeably better code (less redundant), that is sometimes helpful with readability, but, msvc doesn't provide any call chain
[17:00:21] <avenger> and i'm pretty sure the lack of debug info will give ida a harder time
[17:00:48] <avenger> ahh, on the other hand, i think all functions start on a 16 byte boundary, that's also a plus for iwd2 reversing :)
[17:01:31] <-- SiENcE has left IRC (Quit: @all: cya)
[17:28:45] <CIA-26> GemRB: 03lynxlupodian * rcef97bfdffe9 10gemrb/gemrb/ (includes/opcode_params.h plugins/FXOpcodes/FXOpcodes.cpp): handle more conditions for fx_cast_spell_on_condition
[17:29:23] <lynxlynxlynx> now the blade barrier deals damage, but just to the caster
[17:30:12] <lynxlynxlynx> i'd like to add a check in the effect to bail out if the target==actor in these COND_NEAR* cases
[17:30:25] <lynxlynxlynx> do you see anything like that in the opcode avenger?
[17:31:11] <avenger> i thought, the bladebarrier has an ignore center projectile
[17:31:23] <lynxlynxlynx> i'll check
[17:31:39] <lynxlynxlynx> i was just starting dltcep to find which one it uses, since it doesn't hurt anyone else
[17:31:41] <avenger> it is quite the same as fireshield
[17:31:56] <avenger> would fireshield harm the caster?
[17:32:19] <lynxlynxlynx> rarely, never in the original
[17:32:40] <lynxlynxlynx> sppr603 has no projectile (1 - None)
[17:32:43] <avenger> 216-SmlArea Ignore Center (SMLLARNC)
[17:32:48] <avenger> look at sppr603d
[17:33:29] <lynxlynxlynx> bah, another original one
[17:33:53] <avenger> what do you mean :)
[17:34:15] <lynxlynxlynx> that means the projectile is fine and our handling is not so fine
[17:34:28] <lynxlynxlynx> it does have the ignore center bit set
[17:34:36] <avenger> well, ignore center should be a pretty straightforward thing
[17:35:05] <lynxlynxlynx> it works for other stuff
[17:35:06] <avenger> i guess, the problem is the original caster is not preserved
[17:35:20] <avenger> so sppr603d doesn't get the original caster
[17:43:53] <avenger> i hate it when ida is 'helpful' and screws up an already good stack map with its weird guesses
[18:06:59] <lynxlynxlynx> another question for the same effect if you don't mind
[18:07:39] <lynxlynxlynx> i know it is used for many different things, but at least half the cases should only be applied once per round
[18:17:50] <fuzzie> oh?
[18:18:49] <avenger> hmm do we play weapon sounds?
[18:19:37] <lynxlynxlynx> these near ones and the everyround one
[18:20:11] <avenger> what is the question lynx?
[18:20:23] <lynxlynxlynx> last attacked also suggests that, but i don't know if it matters, it's probably only used by contingencies
[18:21:14] <lynxlynxlynx> do you see anything like this in your disassembly?
[18:21:38] <avenger> well, you didn't question anything, but i guess the answer is : this opcode uses triggers
[18:21:48] <lynxlynxlynx> eg. fx_cast_spell_on_condition being only applied each round in some cases
[18:22:19] <lynxlynxlynx> i know it does and that we have them broken, but since they don't get cleared, i don't see how this would help here
[18:22:40] <fuzzie> the triggers used here do get cleared
[18:23:25] <lynxlynxlynx> but how often?
[18:23:34] <fuzzie> every frame, i think
[18:23:45] <lynxlynxlynx> oh, that'd work then
[18:23:59] <fuzzie> well, not for 'everyround'
[18:25:14] <lynxlynxlynx> yeah, but that can be handled separately, i don't know of any spell that uses it of the top of my head
[18:27:25] <fuzzie> but there are 'triggers', like HitBy(), and 'objects', like LastAttacked, and we should be using the triggers, which get cleared quickly, i hope every tick
[18:28:27] <fuzzie> Avenger has them all commented by name in the diassembly of that effect, i think
[18:29:56] <fuzzie> it is maybe simplest to just do the same, actually construct a Trigger object and call it
[18:31:56] <lynxlynxlynx> something for you
[18:49:49] --> SiENcE has joined #GemRb
[18:59:22] <avenger> lynx: the IE does this about lore: when it constructs an item, and its lore to ID is 0, it sets it automatically identified
[18:59:31] <avenger> i think you recently did that :)
[19:02:00] <lynxlynxlynx> yes, i think we handle it in the guiscript
[19:03:12] <lynxlynxlynx> yep
[19:03:55] <lynxlynxlynx> the inventory just checks against the pc lore with greater-or-equal (so it is always true unless you have massive penalties)
[19:04:06] <lynxlynxlynx> the container code checks specifically for 0
[19:04:34] <lynxlynxlynx> do you think this would be better off in the importer?
[19:11:05] <avenger> it is where it assigns an item to an item slot
[19:11:31] <avenger> because the identified bit is part of the slot
[19:55:40] <-- avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
[20:07:05] <lynxlynxlynx> heh, found a regression (my fault) from may
[20:07:19] <lynxlynxlynx> combat output, yet nobody noticed
[20:07:46] <fuzzie> well, there's so many issues with the combat anyway
[20:08:05] <Lightkey> wtf
[20:08:22] <Lightkey> fuzzie: did you hear aboot Firefox 4 not for PowerPC?
[20:08:34] <fuzzie> is that a surprise?
[20:08:40] <Lightkey> to me
[20:08:50] <fuzzie> i mean, they didn't support it on anything except OS X anyway
[20:09:02] <fuzzie> and ppc macs are almost as dead as 32-bit x86 macs
[20:09:18] <Lightkey> what
[20:09:59] <Lightkey> it's a difference of supporting or won't run
[20:10:33] <fuzzie> people always manage to make it run :p
[20:10:51] <lynxlynxlynx> it's a difference between running and working
[20:11:12] <lynxlynxlynx> every second or third time i try to use firefox, it crashes
[20:11:28] <Lightkey> the out-of-process-plugins and JIT compiler beg to differ
[20:11:54] <lynxlynxlynx> it crashes with empty tabs, so there's no plugin excuse there
[20:11:59] <fuzzie> you can see that it's not things which are worth the engineering effort to them, though
[20:12:00] <Lightkey> I mean fuzzie ;-)
[20:13:29] --> spike411 has joined #GemRb
[20:22:02] <Lightkey> dobre.. jitro?
[20:22:15] <Lightkey> I'm not sure which one it was again
[20:26:04] <spike411> So… I've tried to document somehow my Mac OS X building escapade… http://paste.jabbim.cz/4812
[20:26:12] <lynxlynxlynx> Lightkey: that'd be the morning
[20:26:21] <spike411> Good morning
[20:26:23] <Lightkey> dang
[20:26:28] <Lightkey> :-PPP
[20:26:35] <spike411> What's my native language doing here? :)
[20:26:38] <Lightkey> rano neither, I guess
[20:26:51] <lynxlynxlynx> rano == early
[20:27:03] <Lightkey> then I forgot which
[20:27:16] <lynxlynxlynx> something like vecer
[20:27:23] <Lightkey> ah
[20:27:28] <lynxlynxlynx> at least if it is also so similar to slovenian
[20:27:38] <spike411> Dobry vecer
[20:27:46] <Lightkey> yes
[20:27:47] <lynxlynxlynx> večer?
[20:27:53] <spike411> lynxlynxlynx: Yes
[20:27:56] <lynxlynxlynx> :)
[20:27:57] <spike411> Dobrý večer
[20:28:04] <CIA-26> GemRB: 03lynxlupodian * r61448407b28a 10gemrb/gemrb/core/Scriptable/Actor.cpp:
[20:28:04] <CIA-26> GemRB: DisplayCombatFeedback: fixed a regression from 3b8e1b87 that made bg2 always show
[20:28:04] <CIA-26> GemRB: the bg1 combat string
[20:28:06] <CIA-26> GemRB: 03lynxlupodian * recc77ed8228d 10gemrb/gemrb/ (core/Scriptable/Actor.cpp override/bg2/strings.2da): DisplayCombatFeedback: don't hardcode the bg1/iwd/trap/self-inflicted string
[20:28:42] <spike411> BTW I've tried running gemrb in gdb, but I doubt the output is actually useful at all…
[20:29:10] <lynxlynxlynx> let me guess, you have gcc 4.2?
[20:29:18] <spike411> Yes
[20:29:32] <spike411> You don't have anything like make debug target, do you?
[20:29:35] <lynxlynxlynx> nothing newer or older available?
[20:29:50] <lynxlynxlynx> the 0 initialisations aren't really needed
[20:29:58] <lynxlynxlynx> you should just disable -Werror
[20:30:37] <spike411> OK, I guess that's the least problem though.
[20:31:21] <lynxlynxlynx> yeah, the sdl thing kills it
[20:31:33] <lynxlynxlynx> btw, are you using cmake?
[20:31:40] <spike411> Yes
[20:32:11] <lynxlynxlynx> pass -DDISABLE_WERROR next time or change it in the cache
[20:32:53] <lynxlynxlynx> what's your cmake version?
[20:33:02] <spike411> 2.8.2
[20:33:45] <lynxlynxlynx> ok
[20:36:23] <spike411> I guess I could/should just build everything myself instead of relying on system- and macports-supplied libraries. There goes my ‘clean’ system… >_<
[20:37:52] <lynxlynxlynx> you can have a small chroot for such things
[20:38:04] <spike411> Good point.
[20:40:46] <spike411> Would for example my CMakeCache.txt help shed some light on any issue?
[20:42:25] <lynxlynxlynx> not really
[20:42:48] <lynxlynxlynx> i know vorbisfile is missing and am looking into better detection
[20:44:37] <fuzzie> spike411: did you try just making the include "SDL.h"?
[20:45:45] <spike411> Can't remember, I've tried various things… let me see…
[20:46:58] <fuzzie> and i think the 0 initialisations are needed for error cases
[20:48:02] <fuzzie> e.g. code calling GetSpeakerColor and then uses the result without checking if speaker is non-NULL..
[20:48:19] <spike411> Well /Users/spike411/gemrb/gemrb/GemRB.cpp:32:17: error: SDL.h: No such file or directory
[20:48:50] <fuzzie> spike411: oh, #include <SDL.h> maybe
[20:49:09] <fuzzie> actually, i guess we're missing a INCLUDE_DIRECTORIES( ${SDL_INCLUDE_DIR} ) in CMakeLists.txt
[20:49:14] <fuzzie> but maybe better idea: try hard-coding the path :)
[20:49:55] <fuzzie> i'm not sure if this is even your bug, but i do remember it's important that SDLmain get called into, and as far as i know that won't happen if the include is commented
[20:51:35] <spike411> I *think* I've tried hard-coding the path to #include, <>…
[20:52:39] <spike411> Or maybe not.
[20:54:55] <spike411> Awesome.
[20:55:10] <spike411> I say wow.
[20:55:17] <fuzzie> does it work now?
[20:55:23] <spike411> Yes it does.
[20:55:26] <spike411> Lovely.
[20:55:40] <fuzzie> cool :) i guess we need a big warning comment there
[20:56:35] <spike411> No awesome titlescreen music though. :'(
[20:56:45] <spike411> But button click sounds are OK.
[20:56:51] <fuzzie> strange
[20:56:58] <fuzzie> it should complain on the console about what's going wrong there
[20:57:24] <spike411> Yep, I'm looking through it.
[20:58:38] <spike411> Red/error messages: http://paste.jabbim.cz/4814
[20:59:26] <spike411> Are these not-yet-implemented features? http://paste.jabbim.cz/4815
[21:00:42] <spike411> I can hear the narrator, wind, howling…
[21:03:37] <lynxlynxlynx> the errors are fine and yes, those are unimplemented
[21:08:42] <lynxlynxlynx> http://pastebin.ca/1926256 <-- does this help with your sdl problem?
[21:09:15] <fuzzie> that looks correct
[21:09:38] <spike411> Pssst, don't disturb, I'm playing…!
[21:09:44] <fuzzie> although if you apply that, a 'WARNING: commenting this out will cause SDL 1.2.x to crash" would be appreciated
[21:09:45] <spike411> Let me try. :)
[21:16:52] <lynxlynxlynx> are you sure you're not still playing?
[21:17:47] <spike411> Well I was just reading an interesting article on how to handle patches with git… :)
[21:18:57] <spike411> As you've probably noticed by now I'm totally clueles in these areas.
[21:19:16] <spike411> Anyways, it's building… and building…
[21:19:45] <lynxlynxlynx> just do patch -p1 < $(wget -O - http://pastebin.ca/raw/1926256)
[21:19:50] <lynxlynxlynx> ok
[21:20:06] <spike411> Oh nice trick with wget.
[21:20:08] <lynxlynxlynx> it would actually be <<<
[21:20:21] <lynxlynxlynx> or with a pipe
[21:20:47] <spike411> Of course I could have simply used patch… :p
[21:22:21] <spike411> What's with the grey rectangle disfiguring the beautiful character portraits…?
[21:22:23] <spike411> http://img834.imageshack.us/img834/7096/gemrb062iwd2001.png
[21:22:28] <spike411> *rectangles
[21:23:22] <lynxlynxlynx> could be from the portait icon font
[21:23:40] <lynxlynxlynx> iwd2 is one of the bastard childs
[21:23:56] <spike411> There's [TextEdit]: Invalid font set!
[21:23:59] <fuzzie> i think it is, but i thought i fixed that one
[21:24:24] <fuzzie> but i got distracted by gemrb's tendency to hang as soon as you leave that boat
[21:25:06] <lynxlynxlynx> so my patch enabled you to compile it without further sdl hackery?
[21:26:05] <fuzzie> oh, and the barrels in the warehouse :)
[21:26:12] <fuzzie> i really don't understand the barrels
[21:26:21] <fuzzie> but they are funny
[21:26:33] <spike411> lynxlynxlynx: Yes. I have yet to try to run the binary. :)
[21:27:33] <spike411> OK, seems to be running.
[21:27:56] <spike411> Aha!
[21:27:56] <spike411> [ResourceManager]: Searching for THEME/THEMEA... THEME/THEMEA.acm...[Music]
[21:27:56] <spike411> [OpenAL]: Unable to create music buffers: 40963 [ERROR]
[21:27:57] <spike411> Playing: THEME/THEMEA
[21:27:57] <spike411> [OpenAL]: Unable to query music source state: 40963 [ERROR]
[21:28:32] <lynxlynxlynx> three too many
[21:30:24] <fuzzie> change the #define MUSICBUFFERS in plugins/OpenALAudio/OpenALAudio.h i guess
[21:32:18] <lynxlynxlynx> is /opt/local/include a standard path on macs?
[21:32:32] <spike411> lynxlynxlynx: macports
[21:32:55] <CIA-26> GemRB: 03lynxlupodian * r57a218a29e16 10gemrb/gemrb/core/ (DisplayMessage.cpp LRUCache.cpp): fixed a few possible uninitialised var uses, patch from spike411
[21:32:57] <CIA-26> GemRB: 03lynxlupodian * r18c2ddcdab95 10gemrb/gemrb/ (CMakeLists.txt GemRB.cpp plugins/SDLAudio/SDLAudio.cpp): search the sdl include dir when compiling gemrb, fixing a bad apple problem
[21:32:58] <lynxlynxlynx> why do you need OV_EXCLUDE_STATIC_CALLBACKS?
[21:34:17] <spike411> Let's search logs… :)
[21:34:46] <lynxlynxlynx> avoid unused symbol warnings from vorbisfile.h
[21:34:50] <fuzzie> /opt/local/include isn't standard enough to hardcode it
[21:35:12] <spike411> lynxlynxlynx: Yeah, that rings a bell.
[21:35:54] <lynxlynxlynx> fuzzie: it can't hurt adding it, cmake already looks in plenty of places
[21:36:14] <fuzzie> sure
[21:36:40] <lynxlynxlynx> spike411: try this http://pastebin.ca/raw/1926276
[21:46:13] <spike411> fuzzie: 55 #define MUSICBUFFERS 10 <- what seems like a ‘good’ value? 13? 20?
[21:46:39] <fuzzie> i would try 1.
[21:53:14] <spike411> lynxlynxlynx: That doesn't seem to have worked: http://paste.jabbim.cz/4817
[21:53:29] <spike411> not found
[21:53:38] <lynxlynxlynx> did cmake run again?
[21:54:18] <spike411> I only deleted CMakeCache.txt and ran make.
[21:55:10] <fuzzie> the FIND_LIBRARY won't deal with the include paths, maybe?
[21:55:41] <lynxlynxlynx> maybe, but something for tommorow
[21:55:43] <lynxlynxlynx> good night
[21:56:03] <spike411> Good night and thanks!
[21:56:15] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[21:59:11] <spike411> fuzzie: I still get [OpenAL]: Unable to create music buffers: 40963 [ERROR] with MUSICBUFFERS 1
[22:02:36] <fuzzie> could be all kinds of things :( not sure right now
[22:02:49] <spike411> Yeah, better get some sleep.
[22:23:04] <spike411> Good night
[22:23:13] <-- spike411 has left IRC (Quit: Manga & anime pokec na Jabberu: manga.cz@conf.netlab.cz)
[22:35:07] --> edheldil_ has joined #GemRb
[23:21:43] <-- SiENcE has left IRC (Quit: cya @all)
[23:33:53] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)