#gemrb@irc.freenode.net logs for 10 Aug 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:02:51] --> edheldil_ has joined #GemRb
[00:27:58] <-- edheldil_ has left IRC (Ping timeout: 252 seconds)
[03:57:42] --> raevol has joined #GemRb
[04:45:38] <-- raevol has left IRC (Quit: Leaving.)
[05:17:43] --> edheldil_ has joined #GemRb
[05:39:01] <-- edheldil_ has left IRC (Ping timeout: 276 seconds)
[05:54:42] --> edheldil_ has joined #GemRb
[06:17:04] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[06:18:22] <-- CIA-26 has left IRC (Ping timeout: 276 seconds)
[06:43:42] --> CIA-26 has joined #GemRb
[07:25:50] <-- |Cable| has left IRC (Remote host closed the connection)
[07:29:07] --> lynxlynxlynx has joined #GemRb
[07:29:08] --- ChanServ gives channel operator status to lynxlynxlynx
[07:31:22] --> SiENcE has joined #GemRb
[07:34:16] <lynxlynxlynx> that was fast :)
[07:35:48] <fuzzie> just a proof-of-concept
[07:47:10] --> edheldil_ has joined #GemRb
[07:49:37] <-- edheldil_ has left IRC (Client Quit)
[07:53:30] <fuzzie> although i think i will fix it up because i am fed up with openal
[07:55:38] <Edheldil> fuzzie: you get 1000XP and a Box of unending tunes ;-)
[07:55:44] <fuzzie> is Balor really meant to say "Balor was immune to my damage"?
[07:56:48] <Edheldil> now if only we had standalone players for MVE and ACM ;-)
[07:57:02] <fuzzie> i think totem-gstreamer should play both, now
[07:57:16] <Edheldil> ah, ok
[07:57:44] <fuzzie> i actually rewrote some of our MVE code to use the gstreamer code, to avoid possible copyright issues (no-one was quite sure whether the original core MVE stuff was legally licensed)
[07:58:13] <Edheldil> which rewrite do you mean? :-D
[07:58:40] <fuzzie> may 2009, i guess
[07:59:37] <Edheldil> I remeber rewrite from FMod, then a rewrite to ffmpeg code ... :)
[07:59:50] <fuzzie> hehe, this was at a lower level :)
[08:00:12] <Edheldil> not the ffmpeg one, I believe
[08:00:34] <fuzzie> i think this one was labelled with some ffmpeg copyright notices
[08:00:34] <Edheldil> anyway, thank you for all the embedded users :)
[08:01:05] <fuzzie> but ffmpeg/mplayer couldn't play most MVE files when i checked, gstreamer people hadn't ported it back :|
[08:01:37] <fuzzie> i just wish we had any sort of plan for pst
[08:01:55] <fuzzie> whenever gemrb is mentioned in these forum posts, everyone is always most excited about the possibility of playing pst on their device
[08:02:44] <Edheldil> it will be hard, but I hope :)
[08:02:58] <fuzzie> i still haven't finished it
[08:03:09] <fuzzie> one time i got all the way to Curst! :p
[08:03:30] <Edheldil> it's one of those few games I finished :)
[08:04:12] <fuzzie> it's just so time-consuming
[08:05:39] <fuzzie> i guess i will see if i can get scient's attention, get the ida pro db
[08:07:04] <Edheldil> that would be nice, although then it seems so pointless to do anything myself
[08:07:36] <fuzzie> i think i am stretched thinly enough trying to fix core gemrb things as it is
[08:08:06] <fuzzie> but having his db would be a much better starting point if you or someone else ever had time
[08:08:17] <Edheldil> I was thinking that perhaps the IDA's database could be pushed to some private Git repository and worked upon together, but I haven't looked to it's unpacked format yet
[08:08:59] <fuzzie> you can make it export a script which has (thereotically) only your changes in it
[08:09:19] <Edheldil> hmmm, interesting
[08:09:26] <fuzzie> "add comment here", "change this var to type X", etc
[08:09:56] <Edheldil> btw, nonfree can (with varying success) create C sources :)
[08:10:46] <fuzzie> well, if you pay for the hideously expensive Hex Rays bit :( and i really don't want to start pirating stuff, myself
[08:11:12] <Edheldil> that sounds like it could work with git, if the binary was the same
[08:12:54] <fuzzie> it is .. €629 for IDA Pro + €1721 for x86 Hex Rays
[08:13:10] <Edheldil> ah, I thought that IDA's price covered it. Too bad. And I stumbled on some warez forum where the folks were grumbling that it was too hard to un-watermark :)
[08:13:10] <fuzzie> really depressing prices :)
[08:14:32] <Edheldil> it costs much more than I seemed to remember :). Too bad that the free version is several years old
[08:14:53] <fuzzie> the scummvm people negotiated a really good discount at some point for some of them, i hear
[08:15:34] <Edheldil> good for them, we will push IE on them when they are done ;-)
[08:16:22] <fuzzie> but i think the old IDA itself is pretty good for IE work
[08:17:34] <fuzzie> i am just irritated that it doesn't load LX files :(
[08:17:44] <Edheldil> what's LX?
[08:17:53] <fuzzie> i have a prototype scummvm engine for an old game, Star Trek: A Final Unity
[08:18:30] <fuzzie> and i had to manually create all the segments etc in IDA Pro, because even after i unpacked the engine, it is in LX format, some old DOS extended mode (32-bit) file format, and the free IDA Pro doesn't come with a loader
[08:18:58] <fuzzie> v.irritating, you'd think that there would be no commercial value in something so old :P
[08:19:52] <Edheldil> is not it some runtime exe compressor?
[08:20:12] <fuzzie> no, i decompressed it first, and removed the Watcom overlays too :)
[08:20:33] <fuzzie> internet has a lot of helpful tools for that sort of thing, i guess written when they were still relevant
[08:21:23] <Edheldil> it's like pilfering on an abandoned battlefield, sometimes. I get sad, when that happens :)
[08:21:55] <fuzzie> the IE community starts to feel like that :(
[08:23:03] <Edheldil> it does for several yers, no? It just gets worse as time goes
[08:25:08] <Edheldil> and I have a nagging feeling that it's true for many game modding communities - like Tamriel Rebuilt etc.
[08:25:44] <Edheldil> on the other hand, IF is reduced to core ppl who will have to die to leave :)
[08:26:18] <lynxlynxlynx> the ie community is aging very well
[08:26:32] <fuzzie> mhm
[08:27:20] <Edheldil> fuzzie, is it possible to split a procedure in two in IDA?
[08:28:26] <fuzzie> if it's really two functions then you can manually do it, i think
[08:29:20] <fuzzie> by starting with the second half and creating a function there as a hint, or similar
[08:30:00] <Edheldil> it's because IDA believes RaiseException returns :)
[08:30:30] <fuzzie> ah, you can teach it that
[08:31:15] <fuzzie> (it knows it already for functions like exit())
[08:31:21] <fuzzie> not sure how though :)
[08:32:02] <Edheldil> but then it does not do a pretty function tail and converts the other half to DB garbage, even though it clearly begins with push EBP, MOV EBP ESP
[08:32:33] <fuzzie> i guess you can't just right-click and do 'create function' right there? :)
[08:32:36] <Edheldil> but I will try to play with that
[08:33:15] <fuzzie> but i am avoiding my Windows VM today, must study for compiler exam
[08:33:31] <Edheldil> I haven't seen 'create function' menu item anywhere :)
[08:33:46] <Edheldil> I won't bother you then :)
[08:35:00] <fuzzie> it is what you use to manually create functions, but i think it only appears when you don't already have one
[08:35:08] <fuzzie> maybe you can just right-click, edit function, and change the params..
[08:50:59] <Edheldil> found it, thank you
[08:59:04] <fuzzie> the source code (in C++) for ToBEx might be helpful if you want to do serious analysis
[08:59:21] <fuzzie> it has comments on a lot of the bg2 structures; not directly helpful for pst, but not unhelpful
[08:59:38] <fuzzie> did you get Avenger's notes?
[09:20:19] --> SiENcE_ has joined #GemRb
[09:20:39] <-- SiENcE has left IRC (Read error: No route to host)
[09:51:45] <Edheldil> yes, but so far I was unable to understand them much
[09:52:03] <fuzzie> well, in pst_code/opcodes there's a starting point, is all i was thinking
[09:52:14] <Edheldil> looks like what I just found is a main malloc+exception
[10:14:50] <fuzzie> :)
[12:20:35] <SiENcE_> hey fuzzie, the git build for uClibc/mipsel was successful
[12:20:42] <fuzzie> cool
[12:21:10] <SiENcE_> some makefile and config.h modifications where needed
[12:21:19] <SiENcE_> but no that much
[12:21:21] <fuzzie> this is still using cmake?
[12:21:26] <SiENcE_> no
[12:21:30] <fuzzie> i mean, still not using
[12:21:37] <SiENcE_> ah yes
[12:21:46] <fuzzie> i guess you have to work out how to make SDL_mixer work with autotools, then
[12:22:11] <SiENcE_> you mean SDL_Audio
[12:22:17] <fuzzie> no
[12:22:23] <fuzzie> that's easy, if you can link it against SDL_mixer
[12:22:30] <lynxlynxlynx> the sdl plugin also wasn't added to our makefiles
[12:22:43] <fuzzie> but someone has to do that
[12:22:49] <CIA-26> GemRB: 03fuzzie * r01a0b775da49 10gemrb/gemrb/plugins/SDLAudio/SDLAudio.cpp: SDLAudio: handle music queueing
[12:22:49] <SiENcE_> yes i know
[12:22:55] <SiENcE_> i will do
[12:23:20] <fuzzie> but i guess, if you port this stuff, then you probably know another project to copy from :)
[12:23:54] <SiENcE_> yep
[12:24:14] <SiENcE_> i had to remove several -WError options from makefiles
[12:24:19] <fuzzie> alignment warnings?
[12:24:27] <SiENcE_> yes
[12:24:50] <SiENcE_> because otherwise the gcc quits its compilation
[12:24:52] <fuzzie> i suggest just removing the -Wcast-align from configure.in
[12:25:25] <SiENcE_> ok
[12:25:25] <fuzzie> but either way works
[12:26:36] <fuzzie> the SDLAudio isn't ready yet, btw, it is just the start
[12:27:12] <fuzzie> so if you want to fix it too, go ahead
[12:28:51] <fuzzie> i just wrote enough that it works for testing
[12:29:02] <fuzzie> (so hopefully you can see if it is better than openal..)
[12:31:08] <SiENcE_> i hope so :)
[12:32:12] <fuzzie> it is pointless me continuing with it otherwise, so i would like to know
[12:37:58] <SiENcE_> ok
[12:39:28] <SiENcE_> but i think an sdlaudio plugin is not pointless, just because one less dependency
[12:43:24] <fuzzie> it removes openal and adds sdl_mixer :P
[12:46:45] <fuzzie> heh, audio takes up a lot of RAM
[12:47:03] <fuzzie> SDLAudio just leaks loaded audio files right now, it used 50mb in one cutscene :)
[12:48:04] <fuzzie> i note that clicking on portrait icons seems to not run the selection code
[12:48:56] <SiENcE_> is it possible to convert the audio to ogg and play this?
[12:49:03] <SiENcE_> maybe usign a small script
[12:49:05] <fuzzie> sure, but it's no better RAM-wise
[12:49:20] <SiENcE_> but its faster for cpu?
[12:49:23] <fuzzie> this is the annoying part about SDL, it is too simple
[12:49:34] <fuzzie> i think it doesn't help with anything except disk space :)
[12:50:40] <CIA-26> GemRB: 03fuzzie * ra19aef95383c 10gemrb/gemrb/plugins/SDLAudio/SDLAudio.cpp: SDLAudio: handle speech channel
[12:56:10] <SiENcE_> how can i set cross_compiling=yes for configure ?
[12:56:26] <fuzzie> i think none of us have much of an idea about autotools
[12:56:51] <SiENcE_> and i have only little :)
[12:56:53] <SiENcE_> ok
[12:56:56] <SiENcE_> i find out
[12:57:29] <fuzzie> but typically, you just pass --build with your target
[12:58:32] <fuzzie> oh, no, --host
[12:58:53] <fuzzie> confusing
[12:59:24] <lynxlynxlynx> or --target hehe
[12:59:46] <lynxlynxlynx> you're in for a world of pain
[13:00:34] <fuzzie> but it should just work if you set the right combination
[13:00:52] <fuzzie> except i seem to remember it doesn't actually work, because our python detection is dumb
[13:01:09] <fuzzie> it tries actually running the python binary to get the path, or something similar
[13:03:53] <fuzzie> but it is surely the kind of thing which must be possible to take from the acinclude.m4 of other projects
[13:30:23] <Edheldil> it WAS taken from another project, I think :)
[13:30:34] <fuzzie> well, yes, but i mean, multiple other projects :)
[13:30:51] <fuzzie> i think all autotools files are a mix of borrowed code and some changes
[13:31:10] <Edheldil> maybe I could change it so it's possible to specify --with-python or st. like that
[13:31:24] <Edheldil> all, including the first one :)
[13:32:03] <fuzzie> well, i hope that no-one was crazy enough to write it :)
[13:32:25] <fuzzie> so it is safer to assume there is an infinite chain of borrowed autotools code
[13:42:46] <Edheldil> more or less :)
[13:42:49] <Edheldil> dnl Configure paths for SDL
[13:42:49] <Edheldil> dnl Edheldil & Subvertir - 12/09/03
[13:42:49] <Edheldil> dnl Sam Lantinga 9/21/99
[13:42:49] <Edheldil> dnl stolen from Manish Singh
[13:42:49] <Edheldil> dnl stolen back from Frank Belew
[13:42:50] <Edheldil> dnl stolen from Manish Singh
[13:42:51] <Edheldil> dnl Shamelessly stolen from Owen Taylor
[13:45:49] <fuzzie> haha
[13:46:57] <Edheldil> btw, SDL check tests for sdl-config as well, so is the python one such a problem?
[13:47:11] <fuzzie> yes
[13:47:31] <fuzzie> sdl-config is a shell script, no problem to run that on the host
[13:47:36] <fuzzie> python is a binary, which you can't run on the host
[13:49:05] <fuzzie> i mean, i guess you mean "SDL tries running a test app", but it doesn't :)
[13:49:20] <fuzzie> hence the ""cross compiling; assumed OK..." fallback for cross-compile mode
[13:52:26] <lynxlynxlynx> victory
[13:53:32] <Edheldil> I admit I have never tried cross-compilation, so the problem is not clear to me. Running python during crosscompilation to ARM on e.g. i386 would fail or what?
[13:53:33] <lynxlynxlynx> proper kit removal on dualclassing is working (at least the learned ability stuff)
[13:53:55] <fuzzie> Edheldil: yes, because the python is an ARM binary
[13:54:28] <fuzzie> you can run the native python, of course, but that is not very helpful to the configure script
[13:54:33] <Edheldil> ah
[13:55:08] <fuzzie> lynxlynxlynx: you fixed it?
[13:55:26] <lynxlynxlynx> yep
[13:55:30] <fuzzie> :)
[13:55:49] <fuzzie> very cool next release
[13:56:02] <lynxlynxlynx> the gained ability stuff needs the Source fix though
[13:56:16] <fuzzie> did you work out what went wrong there?
[13:57:10] <lynxlynxlynx> haven't looked at it yet
[13:57:23] <fuzzie> not the guiscript calling ApplyEffect or something?
[13:57:40] <fuzzie> because i notice it seems to do that in LUHLASelection
[13:57:45] <Edheldil> can I rely on python-config to exist, then?
[13:58:37] <fuzzie> but i assume that is only for this weird internal storage effect thing
[13:59:03] <CIA-26> GemRB: 03lynxlupodian * r4d1525dd0e8f 10gemrb/gemrb/GUIScripts/GUICommon.py: fixed the comment of RemoveClassAbilities
[13:59:06] <CIA-26> GemRB: 03lynxlupodian * ra7dbcbee0599 10gemrb/gemrb/plugins/GUIScript/GUIScript.cpp: fixed the GemRB_HasResource parameter doc
[13:59:08] <CIA-26> GemRB: 03lynxlupodian * r789121837ea8 10gemrb/gemrb/GUIScripts/ (DualClass.py LevelUp.py bg2/CharGenEnd.py): check if the mage clabs exists instead of hardcodingly ignoring them - they could be added by a modder
[13:59:14] <CIA-26> GemRB: 03lynxlupodian * r8a4044b38f74 10gemrb/gemrb/GUIScripts/DualClass.py: don't run pcf_level twice when setting up a dualclass or the pc will get too many clabs
[13:59:15] <CIA-26> GemRB: 03lynxlupodian * re01707233809 10gemrb/gemrb/core/Scriptable/Actor.cpp: don't apply/remove the old kit clab if the kit is disabled (=pc dualclassed)
[13:59:16] <CIA-26> GemRB: 03lynxlupodian * r64299fd8d660 10gemrb/gemrb/ (2 files in 2 dirs): added an optional parameter to GemRB_SetPlayerStat for disabling the pcf
[14:00:31] <fuzzie> oh boy, these pcf hacks are going to be fun with the previous-state effect code
[14:01:41] <fuzzie> Edheldil: i'm not sure
[14:02:04] <fuzzie> i can only find copies of python-config which call into python themselves, so i doubt it
[14:05:14] <fuzzie> i imagine the best idea is to allow users to provide ones themselves
[14:05:51] <fuzzie> i'm not getting an annoying guy approaching me after the dungeon exit
[14:06:38] <fuzzie> that 'everyone dies anew on area load' thing should be a simple fix, too
[14:08:02] <fuzzie> and magic missile is *way* too fast
[14:08:15] <fuzzie> i know i implemented it too simply, but what are other projectile speeds like?
[14:09:15] <lynxlynxlynx> differing
[14:10:24] <fuzzie> ah, if we ActionOverride someone with Dialog([PC]) in gemrb, gemrb will simply give up when it can't find a target
[14:12:37] <fuzzie> and we've misimplemented DisplayStringWait
[14:17:50] <lynxlynxlynx> core->GetGame()->GameTime%AI_UPDATE_TIME <-- this is true each second, right?
[14:18:18] <fuzzie> no
[14:18:20] <lynxlynxlynx> "true"
[14:18:22] <fuzzie> oh
[14:18:25] <fuzzie> hmm
[14:18:39] <lynxlynxlynx> it would be better if i said false
[14:19:08] <fuzzie> nope
[14:19:35] <fuzzie> but it should be true each second, so you should use that snippet and we should fix it
[14:20:06] <lynxlynxlynx> uh ok
[14:20:15] <fuzzie> GameTime is a bit broken
[14:20:35] <fuzzie> i mean, it should be false each second, but you covered that :-)
[14:20:47] <SiENcE_> hey fuzzie
[14:21:02] <fuzzie> summary: GlobalTimer calls Game::AdvanceTime(count), which then does GameTime+=count, and this is bad because 'count' is not necessarily 1.
[14:21:06] <SiENcE_> please create a file named "Makefile.am" in the SDLAudio dir
[14:21:13] <SiENcE_> and add this to the file
[14:21:34] <lynxlynxlynx> ;)
[14:21:34] <SiENcE_> http://pastebin.com/itS6BksF
[14:21:53] <fuzzie> SiENcE_: does that build?
[14:21:55] <fuzzie> it doesn't for me
[14:21:58] <SiENcE_> yep
[14:22:13] <fuzzie> well, ok, it builds for me, but it doesn't work
[14:22:15] <SiENcE_> but some other makefile.am must also be modified
[14:22:25] <fuzzie> because no @SDL_MIXER_LIBS@
[14:22:38] <SiENcE_> yep laters
[14:22:50] <fuzzie> but if you make it not break, no problem adding it :)
[14:23:09] <SiENcE_> this does not break it
[14:23:14] <SiENcE_> its only the first step
[14:23:26] <fuzzie> oh, right, i didn't put the plugins/Makefile.am change into git
[14:23:42] <SiENcE_> yep and also in configure.in
[14:24:00] <lynxlynxlynx> <lynxlynxlynx> the sdl plugin also wasn't added to our makefiles ‹-- in the morning :)
[14:24:10] <fuzzie> well, that could mean anything :p
[14:24:23] <SiENcE_> :)
[14:24:32] <fuzzie> i mean, obviously i'm not going to make it build w/autotools if autotools is broken
[14:24:45] <fuzzie> and i am slow in mornings
[14:25:14] <fuzzie> but if i add a Makefile.am into SDLAudio, we just have to change it later
[14:26:01] <SiENcE_> ok than wait
[14:26:12] <fuzzie> hehe, too late :)
[14:26:46] <CIA-26> GemRB: 03fuzzie * r446f89d17469 10gemrb/gemrb/plugins/SDLAudio/Makefile.am: add dummy Makefile.am for SDLAudio, for SiENcE
[14:27:13] <fuzzie> that is the one i was using, but i guess it is identical to the one you wanted
[14:28:22] <fuzzie> tomorrow, i'll steal the cache stuff from OpenALAudio, make it not leak and fix video
[14:29:00] <fuzzie> then it should be ok as a low-cpu alternative, even if it doesn't do everything
[14:30:45] <fuzzie> and maybe stop people complaining about gemrb's stuttery audio, everywhere :)
[14:32:21] <SiENcE_> do you also thought about changing the rendering to screen not every frame, just if it's needed (something changed) ?
[14:33:07] <fuzzie> that is not easy
[14:33:33] <fuzzie> i mean, i don't care about the gui because no-one ever spends a lot of time there
[14:33:44] <fuzzie> so the main game screen is the problem
[14:34:08] <fuzzie> and like wjp said, we probably have to profile and work out what takes most of the time
[14:34:59] <lynxlynxlynx> NB. The timings are relative to the movement rate of the affected creature(s). Effects 40 (Slow) and 78 (Disease) doubles the timing rate, effect 16 (Haste) halves the timing rate.
[14:35:01] <lynxlynxlynx> NB. If two or more effects of this type affect a creature, only the shortest duration effect will occur.
[14:35:13] <lynxlynxlynx> pretty handy, they don't stack
[14:35:45] <fuzzie> that is an original engine bug, though?
[14:36:09] <lynxlynxlynx> i'd call it a feature
[14:36:22] <fuzzie> their timed-effect/recurrent-damage queues manage to wipe themselves after the first item in the queue fires
[14:36:22] <lynxlynxlynx> it is clear the effect would be too powerful otherwise
[14:37:02] <fuzzie> well, i don't know what queue you discuss :-)
[14:37:33] <fuzzie> and it does sound like manually implementing that behaviour would make a lot of sense
[14:37:51] <lynxlynxlynx> all the spells i know that use this also deal damage with it
[14:37:55] <fuzzie> but that effect i was looking at the other day, they had to fixpack that out of a bunch of spells
[14:38:04] <fuzzie> the one used in Earthquake etc
[14:38:14] <fuzzie> because it was completely useless due to this stacking bug
[14:38:31] <fuzzie> something would do a visual effect with it, and you'd get no damage
[14:38:59] <lynxlynxlynx> that one is apply effect on condition
[14:39:11] <fuzzie> oh, maybe not that one :-) i forget which one
[14:39:54] <lynxlynxlynx> earthquake just uses delays to make it appear it is doing something each round
[14:40:07] <fuzzie> "apply repeating eff"
[14:40:29] <lynxlynxlynx> same effect
[14:40:36] <lynxlynxlynx> ups, hex vs dev
[14:40:38] <lynxlynxlynx> dec
[14:40:38] <fuzzie> ofc, in the original engine, it uses the same queue as everything else
[14:41:17] <fuzzie> so "shortest duration effect" includes regeneration, poison, etc
[14:42:39] <fuzzie> > What happens at a lower level is that the engine recreates the entire stats (called DerivedStats) on some changes to the creature. The application of an effect to the creature is one such change.
[14:42:46] <fuzzie> > Since the repeating effects are stored in a list inside those stats, they are being recreated every ai update, if there has been an effect application.
[14:42:57] <fuzzie> > The repeating effect itself internally uses a counter to store how often it has been called. And when this counter equals the period it fires of the effect. When the list is being recreated all those counters are lost.
[14:43:08] <fuzzie> ^- ok, so Taimon explains this much better than me :)
[14:44:01] <lynxlynxlynx> the first two are the same as we have it, but the last is confusing me
[14:44:18] <fuzzie> the list is stored *in the DerivedStats*
[14:44:40] <fuzzie> so you re-make the DerivedStats, and the counters for everything delayed (poison, regeneration, repeating effects, etc) goes back to 0
[14:44:46] <lynxlynxlynx> the recreation happens every ai cycle, so how would anything with a greater duration be able to use it?
[14:44:59] <fuzzie> every ai cycle where an effect is applied
[14:45:21] <fuzzie> (we are lazy and just recreate every ai cycle)
[14:45:26] <lynxlynxlynx> yeah
[14:47:36] <fuzzie> but this is the reason behind the 'only the shortest duration effect will occur', so it is a bit more complicated than just resisting stacking, especially since the longer-repeat effects will work as soon as the shorter-repeat ones expire
[14:48:01] <lynxlynxlynx> so what do you propose? I wouldn't tackle the big issue and just add a guard to the start of the effect: if the same effect with the same params and shorter time is applied, skip
[14:48:28] <fuzzie> well, putting the above (from Taimon) in a comment there would be good
[14:48:39] <fuzzie> and if you want to reproduce behaviour, i think you have to check for the same effect full stop
[14:49:15] <fuzzie> (i hope we don't need to care about the weird poison/regeneration/etc behaviour)
[14:49:34] <lynxlynxlynx> i'm not sure it is true - can you have only one active fireshield? I could swear i saw them both, but maybe this is the fixpack part
[14:49:53] <lynxlynxlynx> i'll go check the readme
[14:50:20] <fuzzie> identical repeat times will work fine
[14:50:45] <fuzzie> since they'll both get applied, and then DerivedStats will reset their counter to zero
[14:53:18] <lynxlynxlynx> so you wouldn't limit the check to the same resource/params?
[14:54:29] <fuzzie> well
[14:54:38] <fuzzie> it's kinda your choice
[14:54:58] <fuzzie> if you do it your way, you get very different behaviour from the original
[14:57:13] <fuzzie> i'd prefer it as flexible as possible
[14:57:23] <fuzzie> just not sure it'll actually work :)
[14:57:32] <fuzzie> we need a gemrb fixpack, clearly
[14:57:45] <lynxlynxlynx> heh
[14:59:03] <fuzzie> well, quite a lot of the fixpack spell changes are to work around things like this..
[14:59:52] --> lafille has joined #GemRb
[15:02:11] <-- lafille has left IRC (Client Quit)
[15:36:07] --> |Cable| has joined #GemRb
[16:17:16] <SiENcE_> fuzzie, when i want to test the new SDLAudio plugin, i have to set AudioDriver = sdlaudio in cfg file?
[16:17:33] <fuzzie> yes
[16:17:36] <SiENcE_> ok
[16:25:12] <SiENcE_> ah no
[16:25:15] <SiENcE_> its sdl
[16:25:16] <SiENcE_> VideoDriverName = "sdl";
[16:25:33] <fuzzie> they are not related
[16:25:36] <SiENcE_> k
[16:25:42] <fuzzie> the audio driver name is SDLAudio.
[16:25:48] <SiENcE_> where is the check?
[16:26:05] <lynxlynxlynx> case sensitive?
[16:26:10] <fuzzie> plugins register a name via the macro PLUGIN_DRIVER when they're loaded
[16:26:21] <SiENcE_> ok
[16:26:27] <fuzzie> lynx might be right
[16:26:58] <fuzzie> i wanted tomprince to finish the driver config code before he committed it, but it didn't get done: you can't provide lists, either, just the one preferred driver
[16:27:06] <SiENcE_> ok
[16:27:12] <fuzzie> so it is all on the "will be fixed someday" list
[16:28:23] --> Avenger has joined #GemRb
[16:28:23] --- ChanServ gives channel operator status to Avenger
[16:28:29] <Avenger> yo!
[16:28:33] <fuzzie> aha, an Avenger
[16:28:43] <fuzzie> i have action/message questions, when you have a little time
[16:28:45] <Avenger> hey, this ida pro tool is a wonder
[16:28:52] <fuzzie> we did tell you :-)
[16:28:58] <Avenger> i found out what 'restrictequipitemtype' do!
[16:29:07] <Avenger> didn't test it yet
[16:29:21] <Avenger> but, it is not itemtype, but usability flags
[16:29:50] <Avenger> it is not too useful
[16:30:11] <Avenger> but it is one of the mysteries i couldn't figure out by experiments or simple disassembly
[16:30:37] <Avenger> with ida, i can 'see' the engine much more
[16:31:07] <fuzzie> :)
[16:31:08] <Avenger> and i didn't even run it yet!
[16:31:34] <Avenger> i guess, the write/read debugging and such can help too
[16:31:45] <Avenger> if it doesn't crash like for you
[16:32:00] <fuzzie> i was running ToB in full-screen mode, it is my own fault :P
[16:32:11] <fuzzie> obviously that will never work well
[16:32:15] <Avenger> i should have tried this tool 2 years ago
[16:32:38] <fuzzie> but it's also really useful to set a break in a function and then trace the logic, seeing what it does and being able to look at the structures as it goes
[16:33:04] <fuzzie> much more useful when the debugger knows the structures
[16:33:16] <Avenger> yep, i'm busy teaching it
[16:33:37] <Avenger> i found the clab apply stuff
[16:33:48] <fuzzie> but my question of the day is: do you know if ActionOverride is instant, or if it sends a message? :)
[16:33:52] <Avenger> it is truly applying clabrn05/clabpa05 for fallen guys
[16:34:00] <Avenger> i knew i remembered that
[16:34:06] <lynxlynxlynx> :)
[16:34:06] <fuzzie> that's cool
[16:34:40] <fuzzie> i guess you will know the whole engine soon :)
[16:34:40] <Avenger> no i don't know
[16:34:45] <Avenger> i can look later
[16:34:48] <fuzzie> ok
[16:35:31] <fuzzie> i saw in ToBEx that they say that one of the messages is 'Action', with an action to run inside
[16:35:52] <fuzzie> i would be very happy if that was used for ActionOverride, that's all )
[16:36:04] <fuzzie> kit stuff is cool but i know nothing about it so, actions are more interesting :)
[16:37:24] <Avenger> i will look at that soon
[16:37:35] <Avenger> bye for now!
[16:37:36] <-- Avenger has left IRC (Quit: bye!)
[16:49:26] <-- SiENcE_ has left IRC (Quit: @all: cya)
[16:49:55] <lynxlynxlynx> fuzzie: where do you think we should clear LastDamage and LastHitter?
[16:50:41] <lynxlynxlynx> i wanted to see if it would do any good and put them in initround (bad) and it didn't dampen the fireshield effect at all
[16:51:25] <fuzzie> um
[16:51:43] <fuzzie> complicated :)
[16:53:19] <fuzzie> i think the answer is probably "we shouldn't clear them at all", but another question for Avenger
[16:54:26] <lynxlynxlynx> LastDamage is enough for this to block, but it is also used by triggers
[16:54:29] <fuzzie> the correct system to use is to have both LastDamage/LastHitter/LastHelp/etc, which i think never get cleared, and a list of triggers for things like the fireshield effect, which say "X happened last frame"
[16:55:26] <fuzzie> and i think moving the clearing code around will just break other things
[16:56:29] <fuzzie> but if you don't mind breaking other things, just put it somewhere after RefreshEffects is called on each actor, i guess
[16:56:50] <lynxlynxlynx> nah, i'll leave it be
[16:57:34] <fuzzie> ToBEx/src/OSprite.h has the most commentary i've seen so far
[16:57:38] <CIA-26> GemRB: 03lynxlupodian * r6ce2192b7f17 10gemrb/gemrb/ (2 files in 2 dirs): fixed the GemRB_ApplyEffect doc
[16:57:40] <CIA-26> GemRB: 03lynxlupodian * r78db9c6a68d7 10gemrb/gemrb/plugins/FXOpcodes/FXOpcodes.cpp:
[16:57:40] <CIA-26> GemRB: fx_apply_effect_repeat: don't apply the effect if a similar one is already
[16:57:40] <CIA-26> GemRB: applied with a shorter duration
[16:57:40] <CIA-26> GemRB: plus a note since this is a bit different in the original
[16:57:45] <CIA-26> GemRB: 03lynxlupodian * rdc3219dcce43 10gemrb/gemrb/plugins/FXOpcodes/FXOpcodes.cpp: fixed the once per second modes of fx_apply_effect_repeat
[16:58:18] <fuzzie> hm, where do we do contingency?
[16:58:33] <fuzzie> in the effect, or somewhere sensible?
[16:58:49] <fuzzie> (not sure where it's done originally)
[17:00:03] <lynxlynxlynx> which part? :)
[17:00:12] <fuzzie> ok, i guess there's an effect
[17:00:27] <lynxlynxlynx> it's the same effect
[17:00:38] <fuzzie> and the original version does it right
[17:01:40] <fuzzie> the ToB engine constructs triggers: HitBy, See(NearestEnemyOf), etc
[17:03:51] <fuzzie> which is to say, it is just the same code as the trigger stuff we don't do right atm
[17:05:15] <fuzzie> wow, >4 years old
[17:08:34] --> Micru has joined #GemRb
[17:08:47] --- Micru is now known as micru
[17:09:37] <micru> hi, one question regarding "don't apply the effect if a similar one is already applied with a shorter duration"
[17:10:00] <lynxlynxlynx> sure
[17:10:05] <micru> what kind of effects are meant? magical? all?
[17:10:17] <lynxlynxlynx> only one specific effect
[17:11:11] <micru> so, if i understood correctly, if a character is under the effect of charm, and then another charm is applied, then the second one doesn't count?
[17:11:44] <lynxlynxlynx> no
[17:12:17] <micru> it doesn't make much sense to me... why is like this?
[17:12:29] <fuzzie> this is just for one single effect, as lynx says
[17:13:00] <lynxlynxlynx> this is limited to something very specific
[17:13:08] <micru> like....
[17:13:18] <lynxlynxlynx> like fireshields
[17:13:36] <lynxlynxlynx> or insect plague
[17:13:44] <micru> so you cannot "renew" a fireshield by casting another one?
[17:14:10] <lynxlynxlynx> you can
[17:14:22] <lynxlynxlynx> once the old one expires, the new one will pass that check
[17:14:36] <fuzzie> doesn't 'duration' here refer to one of the params, anyway?
[17:14:39] <micru> aham
[17:15:00] <micru> now i understand, so the effect is not applied, but the cause of the effect remains
[17:15:23] <lynxlynxlynx> the effect stays in the effect queue until expiry
[17:15:25] <fuzzie> hm, that is difficult to make work
[17:15:46] <micru> ok, now it makes sense
[17:15:47] <fuzzie> i hadn't checked your commit
[17:16:09] <lynxlynxlynx> ohh
[17:16:20] <lynxlynxlynx> i don't check the actual time
[17:16:23] <fuzzie> yes :)
[17:16:41] <fuzzie> i guess the 'switch' statement should put the time to check somewhere?
[17:17:32] <fuzzie> and then a check below it for your thing, a check for not being the right time, and then another switch statement for the actual application
[17:18:02] <fuzzie> but you see why i think this is a pain, because this is going to have to be calculated for every effect
[17:18:30] <fuzzie> maybe we should repurpose one of the params as the counter
[17:19:17] <fuzzie> but it ends up being quite a mess
[17:19:56] <lynxlynxlynx> maybe that unknown was used for this
[17:20:09] <fuzzie> no, as Taimon's comment says, they keep a seperate list
[17:20:15] <fuzzie> which is why it all goes wrong
[17:20:52] <fuzzie> the other trouble with your commit is that HasEffect is only going to return the first effect it finds
[17:21:42] <lynxlynxlynx> i'll revert it
[17:21:55] <fuzzie> keep the comment and a TODO, at least?
[17:22:07] <lynxlynxlynx> the timing issue is better solved globally
[17:22:09] <fuzzie> but it's not exactly any more broken now
[17:24:31] <lynxlynxlynx> mhm
[17:25:28] <fuzzie> the original engine keeps this list so that it doesn't have to reapply the effects at all
[17:25:55] <fuzzie> i get the impression that the intention was that re-applying all the effects and regenerating the stats was meant to be rare
[17:26:24] <fuzzie> and most of the time it could just quickly make sure that there were no new/removed effects, go down the list of timers which expired this frame, and carry on
[17:26:44] <fuzzie> but it doesn't seem us worth trying to copy this, it's a hack and clearly didn't work well for them
[17:26:58] <lynxlynxlynx> maybe a performance gain
[17:27:25] <fuzzie> yes, i think it probably is for bg1, where there aren't many effects
[17:27:34] <fuzzie> i mean, not many changes
[17:28:20] <micru> actually to have 2 lists is not so bad, in one list you can keep track of the timing and the other list has no duplicates, just the effects being applied
[17:28:28] <micru> or i am missing something?
[17:28:42] <fuzzie> we don't need the timing list
[17:28:58] <micru> ok
[17:29:17] <fuzzie> but i'm not going to think about it too much now :) i'm no expert on the effects
[17:30:54] <fuzzie> i just wish that slayer death thing was easier to fix :(
[17:31:25] <lynxlynxlynx> what's wrong there? too fast?
[17:31:53] <fuzzie> damage immunity checks the wrong set of stats and you take 1500 damage
[17:32:48] <fuzzie> because the damage is in the queue *before* the immunity :)
[17:33:13] <fuzzie> and damage immunity is meant to check the previous stats (from the last effect queue run), but we don't keep that around yet
[17:33:41] <lynxlynxlynx> ah, yes
[17:35:41] <fuzzie> i guess you can thereotically do everything without needing to be the slayer
[17:35:52] <fuzzie> but people seem to often resort to it in the mind flayer lair
[17:36:40] <lynxlynxlynx> mhm
[17:36:52] <lynxlynxlynx> if you're evil, the reputation penalty even comes handy
[17:37:17] <fuzzie> well, or if you're just trying to keep people in your party without a script hack :)
[17:37:55] <lynxlynxlynx> there are mods for that too
[17:38:11] <fuzzie> ToB seems to actively work against this, i think you get one rep point for each of the three prisoners in the saradush castle
[17:38:21] <fuzzie> yes, i just installed the bg2tweaks one myself
[17:38:42] <lynxlynxlynx> you do get those, yes
[17:38:59] <fuzzie> it's just mean! :-)
[17:39:02] <lynxlynxlynx> just like in the slums ship/building
[17:39:08] <fuzzie> you do one good deed, 3 rep!
[17:39:33] <lynxlynxlynx> if you're evil, why are you even bothering with them?
[17:39:43] <lynxlynxlynx> pathetic weaklings
[17:39:47] <fuzzie> hehe
[17:39:55] <fuzzie> no neutrality in the bg2 world :)
[17:47:16] <fuzzie> need to remember: EvaluateObject passes ga_flags when retrieving an actor by name, i'm almost sure this is wrong
[17:52:46] <lynxlynxlynx> added it to your page
[17:53:14] <-- micru has left IRC (Remote host closed the connection)
[17:54:43] --> budlust has joined #GemRb
[17:57:44] <fuzzie> thanks
[18:10:12] --> Avenger has joined #GemRb
[18:10:12] --- ChanServ gives channel operator status to Avenger
[18:11:21] <Avenger> lynx: about checking the effect, what about doing a better effect finding
[18:11:35] <Avenger> see bool EffectQueue::RemoveEffect(Effect* fx)
[18:11:41] <fuzzie> it needs to be able to check all of them
[18:12:12] <fuzzie> but i'm sure it can be made to work
[18:12:16] <Avenger> it needs to be able to check more than opcode
[18:12:36] <fuzzie> just iterating over the whole list would work fine, i'm sure
[18:12:42] <Avenger> we can also make a short list of repeating effects
[18:13:16] <Avenger> we can build the list with some extra pointers, or maybe some safe reference
[18:13:24] <fuzzie> i'm not sure we really need one
[18:13:28] <Avenger> ok
[18:13:38] <fuzzie> we run on faster computers, and we only want to check this one effect type
[18:13:43] <fuzzie> but i guess lynx can decide :)
[18:13:56] <Avenger> i would hope that the decision is optional
[18:14:06] <Avenger> err, optimal
[18:14:09] <Avenger> O_o
[18:14:22] <fuzzie> you have some idea how the triggers work in the original?
[18:14:25] <lynxlynxlynx> fast computers? maybe with regard to the bg1 minimal specs
[18:14:29] <Avenger> lists
[18:14:37] <Avenger> triggers are added to a list
[18:14:39] <fuzzie> i know there is a list, but i wonder if entries are deleted if checked, or at the end of a frame, or something
[18:14:48] <fuzzie> i was thinking about making a forum post asking, but i guess if you're here, i ask first :)
[18:15:18] <Avenger> checked - definitely not
[18:15:38] <fuzzie> well, i mean, our system is that we 'mark for deletion' when it is checked
[18:15:41] <Avenger> either: checked AND result block executed
[18:15:48] <fuzzie> lynxlynxlynx: well, even the bg2 min spec was only 233mhz :)
[18:15:48] <Avenger> or, maybe at every ai cycle
[18:15:51] <Avenger> or such
[18:16:03] <fuzzie> ok
[18:16:11] <lynxlynxlynx> fuzzie: yes, but i doubt it ran on that even with minimal settings
[18:16:34] <fuzzie> lynxlynxlynx: i ran it originally on a 300mhz machine i think, it worked ok
[18:16:49] <fuzzie> although i wasn't very interested in it at the time, honestly :)
[18:17:33] <lynxlynxlynx> oh
[18:17:53] <fuzzie> my brother bought the collector's edition or something?
[18:18:45] <fuzzie> all i remember is that it had a cloth map and i thought it would be really boring, like Baldur's Gate was
[18:19:43] <Avenger> which one you like the best? bg2?
[18:20:10] <fuzzie> well, i am all grown up now, 10 years later :)
[18:20:25] <lynxlynxlynx> http://pastebin.ca/1914587 <-- how do you feel about this? Classmask is iwd2 only and i'm about to fix it for what it is supposed to do, while those triggers are used elsewhere and Class was already fixed
[18:20:27] <fuzzie> but yes, i think bg2 is the best computer RPG i've played
[18:20:55] <lynxlynxlynx> i'm undecided between bg2 and pst
[18:21:16] <fuzzie> lynxlynxlynx: doesn't that break iwd2?
[18:21:44] <fuzzie> well, i guess it was broken before, too
[18:21:50] <lynxlynxlynx> because of the order?
[18:23:03] <lynxlynxlynx> those triggers need further fixing too, since IE_LEVEL may be the wrong stat
[18:23:40] <fuzzie> well, if i have some levels in cleric in iwd2, then it seems that no longer necessarily works
[18:23:58] <fuzzie> i mean, i know it doesn't right now anyway, but now it'll never work
[18:24:36] <lynxlynxlynx> the game data isn't helpful, because all of the uses have single classes
[18:25:35] <fuzzie> yes, but they should match the multiclass
[18:25:48] <fuzzie> but if you use ID_Class, this will never work?
[18:26:09] <lynxlynxlynx> iwd2 doesn't have multiclasses like bg2
[18:26:23] <fuzzie> i know
[18:26:54] <lynxlynxlynx> IE_CLASS holds only one, probably the last one leveled
[18:27:19] <fuzzie> so if i levelled my fighter and then something checks for 10 cleric levels, your patch breaks, no?
[18:27:42] <fuzzie> because now it'll always decide i'm not a cleric and return false
[18:28:40] <lynxlynxlynx> yes
[18:29:09] <lynxlynxlynx> that can be ammended there though
[18:29:24] <fuzzie> i'm just confused as to what the patch gains
[18:29:27] <lynxlynxlynx> if i fix classmask, those triggers will be broken everywhere except for iwd2
[18:29:43] <fuzzie> well, i would just add the iwd2 triggers
[18:29:54] <fuzzie> apply your patch, and add LevelInClass seperately
[18:30:36] <fuzzie> i'm just wondering if i'm missing something
[18:31:10] <lynxlynxlynx> classmask is broken now and it will get worse for non-iwd2 games, so just adding iwd2 triggers wouldn't help
[18:31:31] <lynxlynxlynx> or did you mean this patch? :s
[18:31:37] <fuzzie> yes
[18:31:46] <fuzzie> i mean, if you apply your patch, you break iwd2
[18:31:54] <lynxlynxlynx> ok, makes sense
[18:31:56] <fuzzie> so i'm wondering why not just add the iwd2 triggers
[18:32:07] <fuzzie> if they don't do the same thing, adding more complexity seems silly
[18:33:33] --> Maighstir has joined #GemRb
[18:53:20] --> edheldil_ has joined #GemRb
[18:56:19] --> Maighstir_laptop has joined #GemRb
[18:56:19] <-- Maighstir_laptop has left #GemRb
[19:18:57] <-- budlust has left IRC (Quit: Lost terminal)
[19:39:42] <-- Avenger has left IRC (Quit: bye!)
[19:44:40] --> SiENcE has joined #GemRb
[19:44:51] <fuzzie> SiENcE: you made it work yet? :)
[19:48:34] <fuzzie> hm, http://gomf.de/taimon/bg_effs.txt is worth remembering
[20:00:57] <lynxlynxlynx> "Windows determines executable by the file extension .EXE not by any per-file mode, and I haven't found any archive format that allows setting that from Windows. So, you'll probably have to build the MacOSX distribution on some Unix system. (You may want to check if your phone can do it.)" <-- nice advice :D
[20:09:18] <fuzzie> if silly :)
[20:41:59] --> Micru has joined #GemRb
[20:53:07] <SiENcE> fuzzie, i only compiled the binary, not yet tested
[20:53:33] <fuzzie> i want to know if sdlaudio helps, darnit :p
[20:59:30] <SiENcE> testing
[21:03:40] <-- edheldil_ has left IRC (Ping timeout: 276 seconds)
[21:03:58] <Micru> :::drumroll:::
[21:27:50] <fuzzie> silence, i see :)
[21:40:21] <Micru> i'm falling asleep...
[21:40:28] <Micru> good night!
[21:40:37] <-- Micru has left IRC (Remote host closed the connection)
[21:59:40] <SiENcE> fuzzie, can't get the current git to work
[21:59:47] <SiENcE> always segfault
[22:00:36] <lynxlynxlynx> where?
[22:00:45] <SiENcE> when i start
[22:00:52] <SiENcE> no output
[22:01:06] <SiENcE> i think its the libgemrb_core.so
[22:01:19] <SiENcE> i have to go into it
[22:01:28] <lynxlynxlynx> i guess gdb is out of the question?
[22:01:36] <SiENcE> yep :(
[22:01:44] <lynxlynxlynx> what about strace?
[22:01:56] <SiENcE> i have to port it
[22:02:05] <lynxlynxlynx> not sure if it would be useful anyway
[22:02:25] <lynxlynxlynx> do you know git well?
[22:02:48] <SiENcE> no
[22:02:51] <fuzzie> when was 0.6.1?
[22:03:03] <fuzzie> oh, 2006-06-16
[22:03:07] <lynxlynxlynx> no
[22:03:13] <fuzzie> erm, 2010
[22:03:17] <fuzzie> tired :P
[22:03:23] <lynxlynxlynx> let's try a bisect
[22:03:29] <fuzzie> we didn't break anything core since then
[22:04:15] <lynxlynxlynx> is this the first time you tried the git version?
[22:04:27] <lynxlynxlynx> run gemrb with AudioDriver = None
[22:04:36] <SiENcE> i did
[22:04:38] <SiENcE> same
[22:04:44] <fuzzie> i think it is, so i would be suspicious of build differences
[22:05:02] <SiENcE> maybe
[22:05:07] <SiENcE> i'm testing
[22:05:27] <CIA-26> GemRB: 03lynxlupodian * r3cc52d96d182 10gemrb/gemrb/core/GameScript/ (GameScript.cpp GameScript.h Objects.cpp):
[22:05:27] <CIA-26> GemRB: removed ID_AVClass and linked avclass.ids to ID_CLASS
[22:05:27] <CIA-26> GemRB: these identifiers are only present in iwd2 and ignored
[22:05:30] <CIA-26> GemRB: 03lynxlupodian * rc8b498d312fa 10gemrb/gemrb/core/GameScript/ (GameScript.cpp GameScript.h Triggers.cpp): added LevelInClass iwd2 triggers, since they're different from the present ClassLevel ones
[22:05:31] <CIA-26> GemRB: 03lynxlupodian * r607edc8fd473 10gemrb/gemrb/core/GameScript/Triggers.cpp:
[22:05:31] <CIA-26> GemRB: changed the ClassLevel triggers to check the class instead of the classmask,
[22:05:31] <CIA-26> GemRB: since it is only available in iwd2
[22:05:33] <CIA-26> GemRB: 03lynxlupodian * r8139f5976c20 10gemrb/gemrb/core/ (3 files in 2 dirs): implemented iwd2 class mask matching
[22:05:35] <SiENcE> but i only did some changes for sdlvideo
[22:05:56] <fuzzie> i mean, you using different configure settings this time
[22:06:38] <SiENcE> no
[22:06:52] <SiENcE> only disabled the -werror
[22:07:26] <fuzzie> i mean, -werror was there in 0.6.1 too
[22:09:08] <lynxlynxlynx> this is simple to test, just build 0.6.1 from the tag + your local changes
[22:09:47] <lynxlynxlynx> does gemrb with your patches work on your pc?
[22:09:58] <SiENcE> not tested
[22:10:16] <fuzzie> if there's no console output, it probably fails at load time
[22:10:26] <SiENcE> but i think the plugins where not loaded at this stage
[22:10:35] <SiENcE> i start and wops---segfault
[22:11:33] <SiENcE> haha
[22:11:54] <SiENcE> i took GemRb 0.61 and only changed the lib to the git version
[22:11:59] <SiENcE> and i starts
[22:12:07] <SiENcE> but fails than later
[22:12:14] <SiENcE> so it isn't the lib
[22:13:12] <fuzzie> probably misbuilt, then
[22:15:23] <fuzzie> but, well, try putting some printf at the top of GemRB.cpp
[22:23:59] <SiENcE> ah it's the
[22:24:04] <SiENcE> mipsel-linux-uclibc-gemrb
[22:26:08] <SiENcE> i took the mipsel-linux-uclibc-gemrb of 0.61 and it works
[22:26:23] <SiENcE> strange
[22:28:47] <SiENcE> now i works...but my mainmenu is black and i get this http://pastebin.com/723DV1By
[22:31:54] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[22:35:44] <SiENcE> the pluginloader doesn't load the SDLAudio plugin
[22:38:20] <SiENcE> ah you should change the gemrb version to something other than v0.6.1
[22:38:37] <SiENcE> mybe git xxxx
[22:38:49] <SiENcE> this is my output http://pastebin.com/Gnevm1m1
[22:48:01] --> SiENcE_ has joined #GemRb
[22:49:11] <SiENcE_> cya
[22:49:15] <-- SiENcE_ has left IRC (Client Quit)
[22:51:37] <-- SiENcE has left IRC (Ping timeout: 252 seconds)
[23:28:24] <-- Maighstir has left IRC (Quit: Maighstir)
[23:50:55] --> nickdaly has joined #GemRb
[23:50:58] --> edheldil_ has joined #GemRb
[23:54:29] <-- nickdaly has left IRC (Client Quit)
[23:54:55] --> nickdaly has joined #GemRb