#gemrb@irc.freenode.net logs for 16 Apr 2011 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage

[01:46:06] <-- Bo_Thomsen has left IRC (Quit: Leaving.)
[03:01:12] <-- |Cable| has left IRC (Remote host closed the connection)
[03:02:37] --> |Cable| has joined #gemrb
[03:32:31] <-- Maighstir has left IRC (Quit: ~ Trillian Astra - www.trillian.im ~)
[03:55:17] <-- fuzzie has left IRC (Remote host closed the connection)
[03:55:23] --> fuzzie has joined #gemrb
[07:31:29] --> lynxlynxlynx has joined #gemrb
[07:31:29] --- ChanServ gives channel operator status to lynxlynxlynx
[08:22:18] --> Bo_Thomsen has joined #gemrb
[08:31:54] <-- lynxlynxlynx has left IRC (Read error: Operation timed out)
[09:05:32] <fuzzie> g'morning
[09:07:52] <dhewg> yo
[09:17:28] <dhewg> pardon my noobness, but memories are blurry. that torgal troll dude, is he supposed to heal himself?
[09:17:53] <dhewg> or steal hp, vampire style?
[09:19:57] <fuzzie> heh
[09:20:03] <fuzzie> god i hate torgal
[09:20:12] <dhewg> im getting there
[09:20:16] <fuzzie> one of the fixpack releases made him immortal
[09:20:53] <fuzzie> by means of being a bit over-anxious to apply a troll HP fix to every troll
[09:21:27] <dhewg> if i keep the cursor on him, so the tooltip is visible at all times, i see him go from "uninjured" to "barely injured", something happens and he's "ununjured" again
[09:21:43] <fuzzie> do you have the cheat keys on?
[09:22:03] <fuzzie> if you hit ctrl-m while hovering over torgal, you should get a debug dump on the console
[09:22:07] <dhewg> dont think so, at least i dont know about that
[09:22:20] <fuzzie> unfortunately i think it is hooked into the cheat keys mechanism
[09:22:27] <dhewg> ah, i get output
[09:22:40] <fuzzie> look for 'monhp1' in his inventory
[09:22:54] <dhewg> not there
[09:22:58] <fuzzie> oh good
[09:23:21] <dhewg> i dont think my install has any mods btw
[09:23:34] <fuzzie> people say that a lot :-P
[09:23:36] <dhewg> extension+official updates
[09:23:57] <dhewg> well, its an ancient install, im not 100% sure myself, but i like staying on vanilla
[09:23:59] <fuzzie> torgal *does* regenerate pretty quickly
[09:24:08] <dhewg> k
[09:24:26] <fuzzie> if i'm actually playing bg2, i will install the fixpack, because there are so many bugs in vanilla
[09:24:43] <dhewg> well, i get lots of warnings, and i see that invisible animation thing too, i thought its maybe related
[09:25:28] <fuzzie> maybe we regenerate too fast
[09:25:33] <dhewg> i thought you guys fix that shit too with the overwrites?
[09:26:17] <fuzzie> we have only a very small number of fixes to original game data
[09:26:33] <dhewg> hm
[09:26:37] <fuzzie> mostly our override/ has new tables, with data which is hardcoded in the original
[09:26:48] <dhewg> im really not sure what this exactly is
[09:26:53] <dhewg> can i find out?
[09:27:06] <fuzzie> the stuff you have installed, you mean?
[09:27:11] <dhewg> yeah
[09:27:26] <fuzzie> put a directory listing of your game dir's override/ somewhere
[09:27:29] <dhewg> i see no weidu fingerprints
[09:27:48] <fuzzie> but anyway, no monhp1 means no buggy torgal
[09:28:02] <fuzzie> he does regenerate quickly, maybe we just fuck up the speed a bit more
[09:28:16] <dhewg> aprospro
[09:28:19] <fuzzie> oh god, yes, this code
[09:28:31] <dhewg> the ticking thingy on the lower left
[09:28:40] <dhewg> that is faster compared to the original, right?
[09:28:46] <fuzzie> yes
[09:29:01] <fuzzie> i am working on the game speed thing, recently
[09:29:07] <dhewg> nice
[09:29:13] <fuzzie> the game is *meant* to aim for 30fps
[09:29:29] <dhewg> http://pastie.org/private/8nfxmupohikjpfv9jbfdjw
[09:29:35] <dhewg> erm, wait
[09:29:38] <fuzzie> with the idea that everything is synced to the frame timer
[09:30:00] <dhewg> http://pastie.org/private/xw0drqkxzhaym7kjykcfkw
[09:30:29] <fuzzie> that is fully-patched ToB :)
[09:30:50] <dhewg> as is bioware or as in fixpack?
[09:30:54] <fuzzie> bioware
[09:31:07] <dhewg> k, memory was right for once \o/
[09:31:15] <fuzzie> it almost matches my override
[09:31:24] <fuzzie> i have one mod installed, DungeonBeGone
[09:31:54] <dhewg> so i still 'should' use the fixpack with gemrb?
[09:31:56] <fuzzie> because really, once i spent some weeks debugging the dungeon to work, i never wanted to see it again
[09:32:11] <fuzzie> well, it's always personal choice :) vanilla works
[09:32:33] <fuzzie> but yes, the fixpack fixes are just as relevant to gemrb as they are to the original
[09:32:47] <dhewg> ok cool, thx
[09:32:53] <fuzzie> they're really helpful documentation actually
[09:33:07] <fuzzie> http://www.gibberlings3.net/bg2fixpack/docs_core.php is really helpful when working out whether game data is just plain broken
[09:33:21] <dhewg> i prefer vanilla with only fixes, there're so many community patches that add random crap
[09:33:38] <dhewg> and the fixpack had tons of options to confuse me
[09:33:42] <fuzzie> the PS:T fixpack is not good, though
[09:33:43] <dhewg> also, omg ocalm
[09:34:03] <fuzzie> it hacks the exe too much
[09:34:18] <dhewg> heh
[09:34:21] <fuzzie> and adds all kinds of arguable game-changing stuff
[09:34:34] <dhewg> i dislike that so much
[09:34:47] <fuzzie> but the bg2 fixpack's 'core fixes' and 'game text update' = good
[09:34:54] <dhewg> do you know the co8 pack for ToE?
[09:34:58] <fuzzie> nope
[09:35:03] <fuzzie> ToE is still in my pile of stuff to play
[09:35:08] <fuzzie> i think my brother gave me a copy, somewhere..
[09:35:10] <dhewg> you didnt yet?
[09:35:24] <fuzzie> but i also have a copy of Portal lying around from him, even
[09:35:33] <fuzzie> i am not very good at finding time to play games :)
[09:35:42] <dhewg> omg, you must at least try it
[09:35:56] <dhewg> skip the stupid initial village and try a little combat
[09:36:11] <fuzzie> i like to try and persuade people to implement an engine for it instead
[09:36:12] <dhewg> that game has the most awesome combat system
[09:36:20] <fuzzie> that is my story for KoTOR, too
[09:36:26] <dhewg> :)
[09:36:29] <fuzzie> DrMcCoy will make it work someday, and i shall wait for then :P
[09:36:31] <dhewg> but really, try it
[09:36:37] <fuzzie> hehe, someday
[09:36:44] <dhewg> anyway, its buggy as hell too
[09:36:52] <dhewg> and there's this co8 fixpack for it
[09:37:02] <dhewg> except it adds random new crap
[09:37:39] <dhewg> which is really silly and doesn't fit at all sometimes
[09:37:42] <fuzzie> but bg2 fixpack is really just fixes, if you don't start picking things labelled 'cool' :P
[09:38:02] <dhewg> and then there's a fixpack for the fixpack to rm the new content crap
[09:38:15] <fuzzie> ah, cute
[09:38:36] <dhewg> maybe thats different now, but its a nice "wtf moment" if you read about it
[09:39:01] <dhewg> anyway
[09:39:39] <dhewg> my torgal is non-bugged, i just fail it and hate him
[09:40:05] <fuzzie> :-)
[09:40:14] <fuzzie> our regeneration code sucks
[09:40:38] <fuzzie> i looked at the original engine's regeneration code and it is quite amazing in some way, but i couldn't work out what it was trying to do either
[09:41:03] <dhewg> i like challenging scenes like this, but this regeneration looks insane, at least at my party level
[09:41:56] <fuzzie> hmm
[09:42:01] <fuzzie> what's the Regeneration line in the debug output?
[09:42:47] <dhewg> http://pastie.org/1800191
[09:43:47] <fuzzie> so 2 HP every second
[09:43:50] <dhewg> i see damages of 15-20 per round, which he heals instantly
[09:44:02] <dhewg> one char that is
[09:45:06] <dhewg> i cherry-picked your tick fix, does that backfire maybe?
[09:45:12] <fuzzie> no
[09:45:16] <fuzzie> RealTime is used very little
[09:45:37] <fuzzie> and if you picked the AdvanceTime line too, i think that would only improve this
[09:46:54] <dhewg> no, i left that out as you said
[09:47:03] <DrMcCoy> fuzzie: lol
[09:47:06] <dhewg> used just the 2 realtime++ lines
[09:47:10] <fuzzie> the responsible code is fx_set_regenerating_state in plugins/FXOpcodes/FXOpcodes.cpp
[09:47:23] <DrMcCoy> fuzzie: I'd need more people, though :P
[09:47:27] <fuzzie> you could try adding a debug print at the end, print damage, see if it's insaning
[09:47:47] <fuzzie> DrMcCoy: my religious anti-boost vows prohibit me from participating :P
[09:48:03] <DrMcCoy> :o
[09:48:09] <DrMcCoy> I love boost!
[09:48:41] <fuzzie> but mostly i tried it and even the poor string class is full of compile-slowing templates
[09:48:45] <fuzzie> i hate compilers so much :P
[09:49:18] <DrMcCoy> :P
[09:49:21] <fuzzie> but someday i will buy a nice multi-core machine, i'm sure
[09:54:18] <dhewg> fuzzie: how often is fx_set_regenerating_state supposed to run?
[09:54:46] <fuzzie> it's meant to get to the end once per second
[09:55:05] <fuzzie> i have a rather dreadful feeling that in 0.6.4, it will *run* at 60fps
[09:55:26] <fuzzie> but this is why someone has carefully added sabotaging code for that, rather than fixing the actual bug at source
[09:55:39] <dhewg> hehe
[09:56:20] <dhewg> how do i filter that by monster?
[09:56:26] <dhewg> the 'target' var?
[09:56:45] <fuzzie> yeah
[09:57:18] <fuzzie> target->GetName(1) probably
[09:57:26] <fuzzie> for nice name
[09:57:28] <dhewg> was just about to ask :)
[09:59:07] <dhewg> this is roughly 1s: http://pastie.org/1800212
[09:59:28] <fuzzie> ugh
[09:59:50] <dhewg> giant troll seems good, but is torgal getting 2hp per frame?
[10:00:28] <fuzzie> it does look like it, or 2hp per AI round at least
[10:00:34] <fuzzie> could you print fx->FirstApply too?
[10:01:05] <dhewg> %d?
[10:01:19] <fuzzie> yes
[10:01:56] <dhewg> always 0
[10:02:01] <fuzzie> oh good
[10:03:33] <dhewg> this is RPD_POINTS for torgal?
[10:03:44] <fuzzie> i hope so
[10:04:07] <dhewg> so he's supposed to get 2hp for one round
[10:04:26] <fuzzie> print fx->Parameter3 and gameTime, i guess
[10:04:49] <fuzzie> probably they've drifted out of sync
[10:05:38] <fuzzie> since Parameter3 is read/written, sigh
[10:05:55] <dhewg> random snippet:
[10:05:55] <dhewg> x_set_regenerating_state (98): Mod: 1, Type: 2 P3: 1756994 FA: 0 GT: 1756990
[10:05:56] <dhewg> fx_set_regenerating_state (98): Mod: 1, Type: 2 P3: 1756994 FA: 0 GT: 1756990
[10:05:56] <dhewg> fx_set_regenerating_state (98): Mod: 2, Type: 2 P3: 1720820 FA: 0 GT: 1756990
[10:06:03] <dhewg> P3 and GT
[10:06:22] <fuzzie> presumably that last one is torgal
[10:06:37] <fuzzie> which is 2411s behind
[10:06:45] <dhewg> upon function end i see fx_set_regenerating_state end: TorGal damage: 2
[10:07:14] <fuzzie> do s/nextHeal/gameTime/ on "fx->Parameter3 = nextHeal + AI_UPDATE_TIME;"
[10:07:51] <dhewg> not for RPD_SECONDS?
[10:08:29] <fuzzie> for all of them, i guess
[10:08:36] <fuzzie> i'll have to be more awake to know if it's the right fix
[10:08:47] <fuzzie> so i figured just RPD_POINTS would fix your immediate issue :)
[10:09:27] <dhewg> weeeee
[10:10:03] <dhewg> now i see "fx_set_regenerating_state end: TorGal damage: 2" around once per sec like for his two buddies
[10:10:16] * dhewg goes of to kill that asshole now
[10:10:23] <dhewg> and i will enjoy that :P
[10:15:04] <dhewg> thats much better
[10:15:16] <dhewg> still hard at this level, but doesnt look insane
[10:15:51] <fuzzie> we should be checking fx->FirstApply anyway
[10:15:52] <fuzzie> sigh
[10:29:15] <dhewg> should i keep that hack or does that break other stuff?
[10:29:43] <fuzzie> keep it
[10:29:57] <fuzzie> usually i talk these things over with lynx
[10:31:44] <dhewg> k
[11:08:08] <dhewg> is it save to apply the fixpack and not starting a new game?
[11:09:25] <fuzzie> i don't know
[11:12:43] <fuzzie> not sure i'd trust it myself
[11:44:03] <fuzzie> ok, let's take a look at the animations, maybe
[11:44:14] <fuzzie> apparently i didn't check google calendar quite closely enough today
[11:47:44] --> Maighstir has joined #gemrb
[12:16:19] <fuzzie> hmm
[12:16:22] <fuzzie> sigh
[12:18:49] <fuzzie> CGameAnimationType::SetNeckOffsets(short, short, short, short, short, short, short, short, short, short, short, short, short, short, short, short)
[12:18:52] <fuzzie> ^- i daren't ask
[12:19:27] <wjp> a 16 dimensional neck space?
[12:19:52] <fuzzie> avenger had it marked as something to do with terrain
[12:23:26] --> pupnik has joined #gemrb
[12:23:33] <pupnik> happy weekend to gemrb!
[12:54:23] <fuzzie> our anim code is really quite incomplete :/
[12:55:28] <fuzzie> there's so many special-cases which you can't handle in a plain table
[13:41:20] <fuzzie> http://fuzzie.org/nfs/gemrb/cgameanim_vtable.txt <- that took way too long to work out
[13:50:24] <fuzzie> so far i am looking at the one anim type corresponding to dhewg's last error
[13:51:08] <fuzzie> and i see fields for infravision detection, movement speed, sound walk frequencies, path smoothing, etc etc
[13:58:02] <fuzzie> and one for whether weapon anims are present or not..
[15:00:06] <xrogaan> hey hey
[15:01:07] <xrogaan> i've got some "Couldn't assign function to trigger:" and "Couldn't assign function to action:" line on startup. Is it "safe" or will I have a problem in game ? :)
[15:01:13] <xrogaan> lines*
[15:01:29] <fuzzie> it depends which ones :P
[15:01:33] <xrogaan> http://dpaste.com/532697/
[15:02:09] <fuzzie> bg2, i guess? it's fine
[15:02:14] <xrogaan> okay
[15:02:18] <xrogaan> yeah, bg2
[15:03:04] <fuzzie> the Detected() one should be implemented
[15:03:13] <fuzzie> and the Killed()
[15:03:52] <xrogaan> and some instantaction too, I didn't see them
[15:04:05] <fuzzie> LayHands is iwd2 only
[15:05:57] <fuzzie> i guess we should probably have a list of actions/triggers to ignore though, so we don't have junk there
[15:08:49] <xrogaan> I don't understand why it try to assign actions/triggers that don't exists in the game files
[15:10:00] <fuzzie> they *do* exist in the game files
[15:10:12] <fuzzie> that is the whole problem :P
[15:10:21] <xrogaan> it's just unused then
[15:10:30] <xrogaan> uh ?
[15:10:31] <fuzzie> for example the 'deathmatch' stuff is some feature that Bioware didn't finish
[15:10:59] <fuzzie> and the 'nidspecial' stuff is some kind of internal/debug hacks, i guess
[15:11:14] <xrogaan> yeah, but about > 17:04 < fuzzie> LayHands is iwd2 only
[15:11:30] <fuzzie> *that* is our bad, we didn't implement it yet because it's only used in iwd2
[15:12:44] <fuzzie> but we should add it
[15:16:46] <xrogaan> okay
[15:17:45] <xrogaan> one last dumb question: will you implement/does it exists a tool to create games with this engine ?
[15:18:34] <fuzzie> sure, people have made some simple stuff already
[15:18:53] <fuzzie> but it's difficult, because it requires a lot of artwork
[15:20:13] <fuzzie> something isometric like FIFE is much easier artwork-wise
[15:21:29] <xrogaan> much simpler and ugly too
[15:21:39] <xrogaan> simple*
[15:21:57] <fuzzie> yep
[15:24:57] <xrogaan> yeah, gemrb is kinda a nerdy project. Sort of impossible to start a OpenSource project with the lack of assets and commercial games wouldn't risk to use that "old" style of gameplay
[15:25:27] <fuzzie> so really it is best to think of it as just an engine for playing the original games
[15:29:13] <xrogaan> mh, gemrb is no longer an "emulator", right ?
[15:29:40] <xrogaan> nevermind
[15:29:49] <fuzzie> most people use it for that though :)
[15:30:28] <xrogaan> well, it's in the source code : "/* GemRB - Infinity Engine Emulator"
[15:31:26] <fuzzie> hehe
[15:31:34] <fuzzie> someone should deal with that, i expect
[15:32:24] <xrogaan> bah, can't a sed to the trick ?
[15:33:23] <tomprince> What is it, if it isn't an emulator?
[15:33:37] <fuzzie> usually 'reimplementation' is used
[15:33:59] <fuzzie> although we've deviated a lot more than reimplementations usually do, too
[15:35:02] <xrogaan> "something like a emulator-reimplementations-other stuff from the infinity engine"
[15:35:47] <tomprince> That sounds about right.
[15:37:37] <tomprince> re: tools, all the tools used for modding ie games can also be used for creating new games.
[15:37:53] <tomprince> The only thing that there aren't really tools for is new animations, i think.
[15:38:20] <tomprince> That was, I think, even the original point of dltcep
[15:41:21] <fuzzie> mm
[15:41:25] <fuzzie> anyway, dinnertime, bbiab
[15:47:06] --> lynxlynxlynx has joined #gemrb
[15:47:06] --- ChanServ gives channel operator status to lynxlynxlynx
[16:03:11] <-- lynxlynxlynx has left IRC (Read error: Operation timed out)
[16:04:00] --> lynxlynxlynx has joined #gemrb
[16:04:01] <-- lynxlynxlynx has left IRC (Changing host)
[16:04:01] --> lynxlynxlynx has joined #gemrb
[16:04:01] --- ChanServ gives channel operator status to lynxlynxlynx
[16:21:39] <fuzzie> hey lynx
[16:22:32] <lynxlynxlynx> oj
[16:22:45] <lynxlynxlynx> going afk again
[16:23:00] <lynxlynxlynx> reading the backlogs though
[16:24:06] <fuzzie> ah, well
[16:24:11] <fuzzie> i am poking at the animation stuff
[16:24:44] <fuzzie> it is not *so* hard to pull the data out of the exe, just time-consuming
[16:25:05] <fuzzie> so for now i'll just dump it in .txt files and maybe we can discuss some solution at another point
[16:39:29] <dhewg> somewhat in the middle of a fight: http://pastie.org/1801033
[16:40:18] <fuzzie> odd
[16:41:05] <fuzzie> line 881 is the one commented "this was checked before we got here"?
[16:45:46] <-- |Cable| has left IRC (Read error: Connection reset by peer)
[16:47:18] --> |Cable| has joined #gemrb
[16:47:29] <-- lynxlynxlynx has left IRC (Ping timeout: 240 seconds)
[16:55:07] --> barra_home has joined #gemrb
[17:17:12] <dhewg> oh
[17:17:14] <dhewg> uhm
[17:17:33] <dhewg> no
[17:17:47] <dhewg> level = actor->GetCasterLevel(spl->SpellType);
[17:18:11] <fuzzie> all that code is on rewrite list
[17:18:17] <fuzzie> there's a race condition i think
[17:27:46] <dhewg> hm maybe
[17:27:56] <dhewg> just got it again, upon running into a trap
[17:28:59] <dhewg> heh, reproducible :P
[17:29:30] <fuzzie> that is not good
[17:30:02] <fuzzie> http://fuzzie.org/nfs/gemrb/monster.txt <- animation data, btw :P
[17:30:29] <fuzzie> i spent some hours pulling that out of disasm, hopefully we can fix at least some invisible anims with it
[17:31:15] <dhewg> nice
[17:32:09] <dhewg> doesn't mean much to me, but that looks like it has been painful to assemble :P
[17:32:27] <fuzzie> they just have a hardcoded switch table, predictably :/
[17:33:43] <fuzzie> bit too fed up with it now to look at prodding it into gemrb
[17:34:26] <fuzzie> but all of that seems not too much trouble to stuff into our table, at least
[17:35:18] <dhewg> nice
[17:37:32] <fuzzie> the monster ones are 106 out of 302 total, it seems
[17:38:18] <fuzzie> you reported MBER, CHMT, CEFB, MGIB, so i guess 50% of those :P
[17:40:22] <dhewg> :)
[17:52:16] <fuzzie> i forgot coffee though, so someone else should fix it
[17:52:35] <fuzzie> we have 16 people in here too, that's not so bad
[18:03:26] <dhewg> heh
[18:03:35] <dhewg> now i get that segfault alot
[18:05:09] <fuzzie> if it's racing, then it's because some actor is trying to do two spell-type things at once
[18:05:14] <fuzzie> possibly modal actions?
[18:05:21] <fuzzie> (find traps, etc)
[18:05:35] <dhewg> now it was that gem portal thingy
[18:05:41] <dhewg> where the monsters come out
[18:14:00] <dhewg> hm, something is wrong on that scene
[18:20:33] <fuzzie> v.conclusive :)
[18:22:52] <dhewg> well :P
[18:23:23] <fuzzie> i think i have to start at the beginning of the dungeon on master and bugfix everything again
[18:23:27] <dhewg> same issue on master, Scriptable.cpp:949
[18:23:36] <fuzzie> but, yes, the spell thing is probably doomed
[18:23:53] <dhewg> if i protect that with "if (spl)", the cutscene breaks somehow
[18:24:26] <fuzzie> yeah, it must be already broken there
[18:24:47] <fuzzie> the problem is that the spell action shouldn't be storing state in LastTarget
[18:25:38] <fuzzie> or so i assume
[18:25:47] <dhewg> hm
[18:25:52] <fuzzie> i think actually *nothing* should be storing state in LastTarget
[18:26:29] <dhewg> btw
[18:26:49] <dhewg> if i run master i see "Searching for bag03.sto...[Cache]" a _lot_ more
[18:26:54] <fuzzie> yes
[18:26:57] <fuzzie> i did notice that
[18:27:06] <fuzzie> i'm not entirely sure why
[18:27:42] <fuzzie> i think probably the global/area scripts
[18:38:01] <fuzzie> also master is super broken
[18:38:02] <fuzzie> sigh
[20:06:00] --> lynxlynxlynx has joined #gemrb
[20:06:00] <-- lynxlynxlynx has left IRC (Changing host)
[20:06:01] --> lynxlynxlynx has joined #gemrb
[20:06:01] --- ChanServ gives channel operator status to lynxlynxlynx
[20:37:14] <pupnik> o/ fuzzie hero of mine
[20:54:52] <Maighstir> Any idea why GemRB would break when I set a window to visible? Other windows work fine, and if I comment out the line BodyWindow.SetVisible(WINDOW_VISIBLE) it doesn't crash. I see nothing of value in the log.
[20:55:26] <fuzzie> well, if it's not visible, it doesn't draw
[20:55:35] <fuzzie> so a buggy image, maybe?
[20:55:53] <tomprince> Is it crashing? Do you have a backtrace?
[20:56:23] <fuzzie> as usual .. what tomprince said :)
[20:56:51] <Maighstir> The python file in question is this: http://pastebin.com/EmZLnCmf - it crashes as it reaches line 364
[20:57:24] <Maighstir> no, no backtrace
[20:57:49] <Maighstir> the same data has worked before
[20:57:59] <fuzzie> what do you mean by 'crash', though?
[20:58:59] <Maighstir> "gemrb.exe has stopped working, Windows can check online for a solution to the problem"
[20:59:30] <fuzzie> i see you have some prints
[20:59:48] <fuzzie> but i think the most interesting thing to print is probably 'Avatar', in UpdateBody()
[21:00:06] <fuzzie> and which gemrb version is this?
[21:00:34] <Maighstir> git, as of a day ago at most
[21:01:17] <fuzzie> i ask because 2DA files were broken for a few days, but it was fixed on Wednesday night
[21:03:50] <Maighstir> It says it's loading CMC1.bam or CFC1.bam depending on whether I chose male or female two screens earlier (as it should, that's what the files are called, and that's what it says in the AT_1 column)
[21:04:44] <Maighstir> CMC1.plt is, of course not found, CMC1.bam is though
[21:04:59] <Maighstir> and that's the last line in the log
[21:06:23] <fuzzie> but did the paperdoll ever work?
[21:07:05] <Maighstir> it has worked, yes, a few weeks ago (I'm not making progress too fast)
[21:07:43] <Maighstir> it's not animating though, but then IWD2 doesn't either in chargen (though the original does)
[21:08:57] <fuzzie> the iwd2 one still works fine for you?
[21:09:02] <Maighstir> yes
[21:09:15] <fuzzie> i'm not really familiar enough with all these flags off the top of my head
[21:09:15] <Maighstir> wait, um
[21:09:49] <fuzzie> so stuff like "IE_GUI_BUTTON_NO_IMAGE | IE_GUI_BUTTON_PICTURE | IE_GUI_BUTTON_ANIMATED" sounds like a really weird combination :)
[21:10:37] <fuzzie> (i do see it in GUIINV)
[21:10:38] <Maighstir> commenting out that line makes no difference though
[21:11:00] <Maighstir> and yes, iwd2 shows its paperdoll
[21:11:45] <fuzzie> well, you add the IE_GUI_BUTTON_ANIMATED back in UpdateBody
[21:11:52] <Maighstir> I pretty much copied stuff from IWD2 as I wanted the same system of having the bam as the paperdoll
[21:12:39] <fuzzie> and honestly i have no idea what that's meant to do
[21:14:04] <Maighstir> And the curious thing is that it's only breaking as soon as I display the window, not before when stuff is being done
[21:14:24] <fuzzie> i imagine it is crashing in Button::Draw
[21:14:33] <fuzzie> but that is difficult to tell, if you can't get a backtrace from a debugger
[21:16:10] <Maighstir> At least, the log doesn't show a backtrace of python functions, I could get a disassembly out of visual studio, but nothing more comprehensive as I built it with MinGW
[21:17:03] <fuzzie> the IE_GUI_BUTTON_PICTURE is the tricky paperdoll code
[21:17:47] <Maighstir> that's what makes it accept a palette change?
[21:19:21] <fuzzie> anyway
[21:19:32] <fuzzie> if you want an iwd2-style image, you shouldn't be using SetPLT
[21:20:47] <fuzzie> i really can't tell why it's crashing, at a glance, but my guess would be some stupid interaction between SetBAM and SetPLT there
[21:20:52] <Maighstir> SetAnimation instead?
[21:20:59] <fuzzie> yes, and SetAnimationPalette
[21:21:59] <fuzzie> SetBAM *does* seem to do the right thing and remove the results of SetPLT..
[21:24:00] <Maighstir> changing the lines didn't make a difference
[21:24:01] <Maighstir> BodyDisplay.SetAnimation(Avatar, 1) BodyDisplay.SetAnimationPalette(0, AvatarMinorColour, AvatarMajorColour, AvatarSkinColour, 0, 0, AvatarHairColour, 0)
[21:24:18] <Maighstir> (as two lines, of course)
[21:24:40] <fuzzie> and your BAM has a cycle 1?
[21:24:45] <tomprince> anyway to get a backtrace on win32?
[21:24:46] <Maighstir> yes
[21:24:55] <fuzzie> tomprince: fighting with gdb..
[21:24:56] <Maighstir> about backtrace, don't know
[21:26:55] <fuzzie> we lost the errors in BAMImporter too, when tomprince refactored
[21:27:24] <fuzzie> just noting as i poke through
[21:29:44] <wjp> isn't doing a SetPLT followed by a SetBAM rather weird?
[21:29:55] <wjp> oh, fuzzie already mentioned that I see
[21:29:56] <fuzzie> yes, it doesn't work
[21:29:57] <fuzzie> hehe
[21:30:14] <wjp> for added confusion SetPLT can also load a BAM :-)
[21:30:21] <fuzzie> and it was, here..
[21:30:28] <fuzzie> i can only think it's some fallout from the file handling code if it stopped working
[21:31:11] <fuzzie> but it's weird if only your code is broken.
[21:31:27] <Maighstir> I think I tried SetAnimation and SetAnimationPalette first, then checked the IWD2 scripts and saw that curiosity as well, so I just copied that... but it was a few weeks ago, so I don't really remember
[21:33:15] <fuzzie> tomprince: how much did you test this on win32?
[21:33:17] <wjp> ah, right, our iwd2 RefreshPDoll also does SetPLT followed by SetBAM
[21:34:27] <tomprince> Not much.
[21:35:06] <Maighstir> I'm building on Linux now, to check there as well
[21:35:13] <fuzzie> great
[21:35:36] <fuzzie> tomprince: ah well :)
[21:35:50] <fuzzie> difficult to tell what kind of other thing might've broken it, but lots of things we miss
[21:36:16] <tomprince> I try to avoid having log on to my windows vm...
[21:46:50] <tomprince> one of the reasons I'd like better tests.
[21:55:16] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[22:07:59] <Maighstir> Phew, finally (why does samba mess with me? had to use a USB stick to move the files). Ubuntu says "segmentation fault". Not very informative.
[22:08:16] <fuzzie> hmph
[22:08:20] <fuzzie> well, run it under gdb, i guess?
[22:11:11] <Maighstir> I don't completely understand gdb's man page, how do I run gemrb with arguments through gdb?
[22:11:20] <wjp> gdb ./gemrb
[22:11:23] <wjp> set args (args)
[22:11:24] <wjp> run
[22:11:46] <tomprince> or gdb --args ./gemrb (args)
[22:14:56] <Maighstir> pastebin.com/qRL2Wi36 -- the culprit seems to be ScrollBar::draw
[22:15:39] <Maighstir> hmm, wonder how I've set my scrollbars...
[22:29:46] --> edheldil_ has joined #gemrb
[22:33:35] <Maighstir> Ah, there's the error, I had apparently managed to change the name of the bam slightly, but not updated the chu
[22:35:51] <Maighstir> And it did show in the log, I just didn't see it.
[22:51:26] <-- Bo_Thomsen has left IRC (Quit: Leaving.)