#gemrb@irc.freenode.net logs for 4 Aug 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:34:33] --> raevol has joined #GemRb
[06:50:32] <-- raevol has left IRC (Quit: Leaving.)
[06:55:57] --> lynxlynxlynx has joined #GemRb
[06:55:57] --- ChanServ gives channel operator status to lynxlynxlynx
[07:18:53] <-- |Cable| has left IRC (Remote host closed the connection)
[08:54:42] <fuzzie> morning
[09:00:16] <lynxlynxlynx> oj
[09:10:04] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)
[09:27:58] --> SiENcE has joined #GemRb
[09:28:49] --> budlust has joined #GemRb
[09:39:30] <fuzzie> lynxlynxlynx: do you know of any effects other than the recurrent damage ones where applying them multiple times a frame would be a problem?
[09:41:45] <lynxlynxlynx> no
[09:46:55] <fuzzie> ok :) thanks
[09:48:01] <lynxlynxlynx> is it intentional that speakerID is always the id of the person who started the dialog?
[09:49:04] <fuzzie> hmm
[09:49:29] <fuzzie> i think so, yes
[09:49:55] <fuzzie> because targetID is the current speaker
[09:51:32] <fuzzie> but intentional doesn't necessarily mean correct
[09:51:43] <fuzzie> i know Avenger found that SetGabber doesn't work how we want it to
[09:52:19] <lynxlynxlynx> confusing name
[09:53:24] <fuzzie> trying to fix the centering thing?
[09:59:28] <fuzzie> the stupid scroll learning still doesn't work
[10:01:12] <lynxlynxlynx> scroll learning has worked for me since it was implemented
[10:01:31] <fuzzie> well, it worked fine originally, as long as the game wasn't paused
[10:01:31] <lynxlynxlynx> i'm trying to add the white speaker circle
[10:01:47] <fuzzie> and then you implemented feedback via manual LearnSpell call from the python, and that doesn't work
[10:01:57] <lynxlynxlynx> centering would be nice too, but I don't know enough about how it was done
[10:02:20] <fuzzie> so Avenger implemented a timed event to get the feedback after the item use is done, but that breaks even moreso when paused
[10:02:37] <fuzzie> we really just have to instant-use the scroll, i guess
[10:02:54] <lynxlynxlynx> aha
[10:04:02] <fuzzie> beware of my "ground circles are not drawn in cutscenes" comment in Actor::Draw
[10:06:20] <fuzzie> it sounds annoying to implement actually
[10:06:43] <fuzzie> since the first example i can think of doesn't even involve dialog at all
[10:07:03] <lynxlynxlynx> i saw that, but didn't get the verbal constant thing
[10:07:04] <fuzzie> but, oh well, brave you :)
[10:07:15] <lynxlynxlynx> what i have now uses speaker id, so it colors my pc instead
[10:07:45] <fuzzie> well, the lack of white speaker circle is most annoying when you're not in a dialog at all
[10:07:46] <lynxlynxlynx> will see how much needs to change for the switch to targetid
[10:08:15] <fuzzie> e.g. the leaving-Candlekeep cutscene in bg1
[10:08:34] <lynxlynxlynx> i know it is used elsewhere, but i'm just trying to fix that spellhold bug of indistinguishable speakers
[10:08:57] <lynxlynxlynx> also the random babbling the party npcs do highlights them
[10:09:00] <fuzzie> so i wrote 'VerbalConstant' there without really knowing what i meant
[10:09:19] <lynxlynxlynx> which is verbal constant stuff
[10:09:28] <fuzzie> ok, well, that is good :-)
[10:10:17] <fuzzie> i'm not sure targetID will help you either, though
[10:12:46] <fuzzie> but i guess if you can make it work at all, then it can be fixed
[10:20:30] <fuzzie> in other stupid UI bugs, we need to disable the UI in all dialogs, not sure why we don't
[10:21:20] <lynxlynxlynx> oversight
[10:25:28] <fuzzie> "Unless something major or cool turns up, this is hopefully the end of my involvement with engine/IESDP stuff" -- devSin, June 2006
[10:26:06] <lynxlynxlynx> ;)
[10:28:54] --> Maighstir_laptop has joined #GemRb
[10:35:33] <fuzzie> got to remember to ask Avenger about SPMINDAT..
[10:58:29] <Maighstir_laptop> The chargen in GemRB/BG1 doesn't seem to fully use the various values for a loaded character. Backing up to earlier steps makes some choices (class/alignment) disabled depending on the first loaded character (an all disabled if the current loaded is the first).
[10:58:29] <Maighstir_laptop> In the original it's possible to change the class and alignment (of course needing to redo all the steps after) on a loaded character, in GemRB as of now I need to go back another step to choose the race.
[10:59:12] <fuzzie> loaded meaning imported?
[10:59:17] <Maighstir_laptop> yes
[10:59:32] <fuzzie> meh, i thought that was working for bg1
[10:59:48] <fuzzie> it is not
[10:59:55] <fuzzie> guh
[11:00:21] <lynxlynxlynx> patches welcome :)
[11:00:47] <lynxlynxlynx> i maybe broke your demo btw
[11:00:59] <fuzzie> the GUICGxx.py files use GetVar() still
[11:01:38] <fuzzie> they need replacing with "MyChar = GemRB.GetVar ("Slot")" and then "GemRB.GetPlayerStat (MyChar, IE_RACE)" or IE_CLASS or equivalent
[11:03:06] <fuzzie> although it is a little more complex than that i suppose :|
[11:03:43] <lynxlynxlynx> only the vars that are used to get the values from the gui should stay
[11:04:31] <lynxlynxlynx> (the ones that are used with setvarassoc)
[11:04:45] <lynxlynxlynx> the rest should be all GetPlayerStat/SetPlayerStat
[11:04:57] <fuzzie> the trick, i suppose, is to look at where each stat is *set*, and it seems obvious how to modify the stat users from there
[11:05:28] <fuzzie> but e.g. the "Class" var is turned into IE_CLASS via a table lookup, it seems? so can't directly substitute
[11:05:37] <fuzzie> i shall leave it alone, want to finish other things
[11:06:02] <-- budlust has left IRC (Quit: Lost terminal)
[11:06:32] <Maighstir_laptop> Do more important stuff you, I'll see if I can figure it out and help out
[11:08:11] <fuzzie> any objections to changing gemrb's scrollbars so clicks work as in the original engine?
[11:08:37] <fuzzie> (clicking above/below the slider scrolls a page up/down, rather than the current gemrb behaviour, which is to move the slider to the position you clicked)
[11:10:02] <lynxlynxlynx> Class is one of those gui values, so it is used as an index into the table
[11:10:43] <lynxlynxlynx> i object, yes
[11:11:08] <fuzzie> ok
[11:11:12] <lynxlynxlynx> we have mouse scrolling and pg/pgdown could be made to work like that (if they don't already)
[11:11:15] <fuzzie> so, next question: have an idea how to do it?
[11:11:17] <fuzzie> well
[11:11:21] <fuzzie> this is for the message window in-game
[11:11:57] <fuzzie> atm it's impossible to scroll through it reliably without a middle mouse button
[11:12:08] <fuzzie> and obviously pgup/pgdn are bound to something different there
[11:12:34] <lynxlynxlynx> you can't just click on the top/bottom buttons?
[11:12:42] <fuzzie> they scroll line-by-line
[11:12:47] <lynxlynxlynx> yes
[11:12:55] <fuzzie> i mean, i'm asking this because it's getting really annoying to debug anything
[11:13:09] <lynxlynxlynx> you don't have a scroll mouse?
[11:13:11] <fuzzie> often you get a few pages of output at once
[11:13:15] <fuzzie> i'm using a laptop :)
[11:13:33] <lynxlynxlynx> it doesn't have the scrolling touchpad?
[11:13:38] <fuzzie> no
[11:13:54] <fuzzie> but, i guess i'll make this an option
[11:14:08] <lynxlynxlynx> that's fine
[11:14:12] <fuzzie> and maybe we can bind shift-pgdn/pgup or something, once i have keybindings working
[11:17:01] --> budlust has joined #GemRb
[11:17:50] <fuzzie> should really finish that, but i keep being distracted
[11:18:22] <Maighstir_laptop> Did you have a key to keep the cursor inside the window? It's slightly annoying when playing and the cursor needs to be at the extreme edge of the game window to scroll the map. A moot point if cursor trapping works, but the original games in window mode scrolls when the cursor is at the edge of the screen, not the window (equally annoying on multidisplay setups).
[11:18:44] <fuzzie> Maighstir_laptop: scroll lock, i think
[11:19:15] <Maighstir_laptop> Great
[11:20:21] <lynxlynxlynx> yep
[11:20:29] <fuzzie> thank lynx, who added it :)
[11:33:56] <fuzzie> meh, iwd+pst have more of these stupid effects than fxopcodes does :P
[12:11:29] <SiENcE> hey
[12:12:05] <SiENcE> if you want todo some optimizations...gemrb need really much cpu power, even if nothing is used
[12:12:44] <SiENcE> so in mainmenu, 80-90% cpu power but you don't have to draw the screen or something, if no mouse is moved
[12:13:08] <lynxlynxlynx> we need a contributor to join us and work only on performance
[12:13:29] <SiENcE> maybe i can do some patches for you
[12:13:34] <lynxlynxlynx> yes please :)
[12:13:46] <SiENcE> because cpu power eats my battries ;)
[12:13:53] <lynxlynxlynx> exactly
[12:14:01] <SiENcE> downscaler is working now and mouse also
[12:14:07] <lynxlynxlynx> it's a major porting obstacle too
[12:14:28] <SiENcE> what is a major porting obstacle?
[12:14:40] <fuzzie> but dirty-rectangling is not so simple
[12:14:46] <lynxlynxlynx> and wouldn't it be great if gemrb was consistently faster than the original? That would bring so many new users ;)
[12:14:54] <lynxlynxlynx> SiENcE: performance
[12:15:02] <SiENcE> yes
[12:15:14] <fuzzie> put a hundred or so saves in your save directory and time the 'load' screen in the original and gemrb :P
[12:15:23] <SiENcE> caching should be optimized and screen redraw only when needed
[12:15:26] <fuzzie> gemrb is about half a second, original is about 30 seconds :(
[12:16:26] <lynxlynxlynx> the fog of war updating is slow too
[12:17:03] <SiENcE> should i use #if xx for patches or a complete integration?
[12:17:30] <lynxlynxlynx> if they're not dingux specific then there's no need for ifdefs
[12:17:46] <SiENcE> ok
[12:18:02] <SiENcE> but i dont think people will play in 320x240 ;)
[12:19:40] <lynxlynxlynx> why not? that'd be four gemrbs on my screen
[12:19:53] <lynxlynxlynx> all my personalities could join in multiplayer someday
[12:20:27] <fuzzie> hehe
[12:20:43] <fuzzie> but i guess really, you want to fix the GUI properly for 320x240
[12:21:35] <lynxlynxlynx> in this situation, i could have one gemrb only for the message window :)
[12:24:31] <fuzzie> centering stuff is not too hard, it seems
[12:24:32] <SiENcE> ;)
[12:24:39] <SiENcE> i hope
[12:24:50] <fuzzie> it even follows moving targets :)
[12:25:07] <SiENcE> later i want to port it too the new handhelds, wiz and caanoo
[12:25:27] <SiENcE> they have touchscreen...which is better than a virtual mouse (d-pad)
[12:28:46] <fuzzie> but there are two sets of obvious optimisations
[12:28:59] <fuzzie> one is that we re-blit the entire screen every frame
[12:29:06] <fuzzie> and the other is that we use way more RAM than we need
[12:30:15] <fuzzie> and the Pandora/N900/etc people have 256MB RAM to play with, so i think the RAM limit is only important to your port
[12:34:27] <fuzzie> while the drawing stuff is important even for modern desktop machines
[12:36:25] <lynxlynxlynx> especially if you're debugging
[12:37:09] <wjp> You'd mainly save only on background tile rendering, right?
[12:37:28] <wjp> and the interface
[12:37:29] <fuzzie> wjp: and the UI
[12:37:30] <fuzzie> yes
[12:38:08] <wjp> for profiling purposes it would be interesting to separate the two and see how much performance could be gained
[12:38:10] <fuzzie> you can also implement the options available in the original engine (simpler transparency, no area animations, etc)
[12:38:52] <wjp> those could help a lot
[12:39:13] <fuzzie> but i'm not sure how many of the performance issues reported by the Pandora/N900 people are openal :P
[12:40:04] <fuzzie> i guess i should make accounts and ask on the relevant forums
[12:40:09] <fuzzie> or someone should
[12:40:22] <lynxlynxlynx> i don't understand why they don't come to us
[12:40:47] <lynxlynxlynx> sure it is mostly about getting the build environment, but when it comes to improvements ...
[12:41:05] <fuzzie> i think everyone assumes that the people building the ports will report back to us, but no-one does
[12:56:43] <SiENcE> i do ;)
[12:57:14] <SiENcE> loading is so slow from sd-card
[12:58:36] <SiENcE> what you think, how much ram is needed to go ingame? 32mb?
[12:58:54] <SiENcE> menues are around ~12mb
[13:00:57] <fuzzie> depends
[13:01:38] <fuzzie> my local copy is using ridiculous amounts of ram to go ingame
[13:01:39] <fuzzie> what on earth
[13:08:20] <fuzzie> 36mb of tiles. awesome.
[13:08:59] <SiENcE> what parts need most ram?
[13:09:08] <SiENcE> the background tiles?
[13:09:12] <SiENcE> of an area?
[13:09:17] <fuzzie> plus 9mb of tiles in SDL land, 7.7mb of BAMs, 5.2mb of MOS files, 2.5mb of audio, etc.
[13:10:03] <SiENcE> ~25mb?
[13:10:20] <SiENcE> +~10mb for binaries and plugins
[13:10:46] <SiENcE> 35mb is too much for the dingoo a320
[13:10:59] <SiENcE> but i will try my a330...which has 64mb ram
[13:11:09] <fuzzie> it didn't use to be this bad
[13:11:27] <fuzzie> something is wrong in the WEDImporter, i think
[13:14:34] <Maighstir_laptop> the scripts for exporting a char are pretty much incomplete, right? trying to export from in-game brings up the window, but no file is saved, and the pregen saves with a pre-set name (gembak)
[13:14:44] <fuzzie> umm
[13:14:46] <fuzzie> that sounds like a bug
[13:15:03] <fuzzie> gembak is meant to be for chargen only
[13:15:28] <Maighstir_laptop> chargen/pregen
[13:15:42] <wjp> fuzzie: I semi-recently started caching tiles converted to display format
[13:15:50] <fuzzie> no, only chargen
[13:16:17] <wjp> fuzzie: which significantly speeds up tile rendering, but at the cost of RAM
[13:16:18] <Maighstir_laptop> eh, wait..no, that's my wrong
[13:16:21] <fuzzie> wjp: there is 9mb of SDL_CreateRGBSurface allocs from GetTile, but 36mb for its own allocations, i thoguht it used to be far better
[13:16:40] <wjp> 'its own' = internal to SDL?
[13:16:47] <fuzzie> sorry
[13:16:54] <fuzzie> 'its own' = GetTile itself, allocating memory for the pixels
[13:18:58] <Maighstir_laptop> does exporting work at all, then? at the end of the pregen the "accept" button should change to "export" and act accordingly
[13:19:30] <SiENcE> maybe you should make an config option, for low ram usage
[13:20:01] <SiENcE> like disabling music, sounds, smaller tilecache etc.
[13:21:08] <fuzzie> yes
[13:21:37] <fuzzie> Maighstir_laptop: you're using bg1?
[13:21:44] <Maighstir_laptop> yes
[13:21:51] <fuzzie> how are you doing pregen?
[13:22:06] <Maighstir_laptop> multiplayer->pre-generate
[13:22:18] <fuzzie> but that should just take you back to the multiplayer party screen, no?
[13:22:42] <fuzzie> or you mean we misimplemented that earlier?
[13:23:31] <fuzzie> if i click pregen, i get only the party arbitration screen
[13:23:45] <SiENcE> does bg1 or iwd1 uses less ram with gemrb?
[13:24:00] <Maighstir_laptop> not in the original, you get the export screen, then get back to the multiplayer start screen (with pre-gen, connect, and back to start)
[13:24:06] <fuzzie> SiENcE: yes, bg1 is more lightweight
[13:24:14] <SiENcE> thx :) good to hear
[13:24:20] <fuzzie> Maighstir_laptop: so the bug is that we're displaying the party arbitration here, instead?
[13:24:26] <Maighstir_laptop> yes
[13:24:53] <fuzzie> the bg1 chargen only has "accept and enter game, saving a backup to gembak if you have GUIEnhancements enabled" and "go back to party arbitration" paths
[13:25:00] <fuzzie> so i think it's just that nothing else is implemented
[13:25:27] <fuzzie> in fact, bg1 doesn't even have an ExportFile.py
[13:26:36] <fuzzie> so bg1 would need an ExportFile.py (probably modified to match bg1's ui), and then you have to copy the ExportFile-opening code from the very end of bg2/CharGenEnd.py to the same place at the end of bg2/CharGenGui.py
[13:28:34] <fuzzie> and you need to compare the PregenPress function from bg2/Start2.py into bg1/Start.py - I think you just replace GUIMP with CharGen
[13:28:49] <fuzzie> should maybe have a wiki page for these things
[13:29:16] <fuzzie> erm, it should be 'to the same place at the end of bg1/CharGenGui.py' above, ofc
[13:29:52] <Maighstir_laptop> that's what I did, changed GUIMP to CharGen
[13:31:27] <fuzzie> well, if you only change the menu button, then it is predictable it won't work, i guess :)
[13:32:11] <fuzzie> the first 'playmode = GemRB.GetVar ("PlayMode")' in bg2/CharGenCommon.py is responsible for the accept/export text.
[13:32:22] <fuzzie> but i guess that is a minor detail.
[13:36:13] <fuzzie> wjp: so i guess SDL_CreateRGBSurfaceFrom is recursively calling SDL_CreateRGBSurface for some tiles..
[13:36:15] <Maighstir_laptop> playmode>=0: "accept", else: "export", I assume
[13:36:20] <fuzzie> yes
[13:37:06] <fuzzie> gah
[13:37:18] <fuzzie> so we load all area tiles at least twice?!
[13:37:45] <fuzzie> i have too much of a headache to look at this now, i think
[13:39:34] <fuzzie> but the WEDImporter tries loading tiles twice, once for without rain and one for with rain
[13:42:16] <fuzzie> so, well, :-(
[13:43:19] <fuzzie> it doesn't even ever use them for the main tiles! DrawOverlays always uses the non-rain version :(
[13:45:05] <fuzzie> dammit
[13:48:14] <CIA-26> GemRB: 03fuzzie * rdd7737f1b732 10gemrb/gemrb/plugins/WEDImporter/WEDImporter.cpp: don't try loading a rain version of the main overlay
[13:49:23] <fuzzie> so there's 25mb saved in the first area of bg2
[13:50:01] <fuzzie> and a huge speed increase
[13:50:10] <fuzzie> i will go despair in the direction of a sandwich
[13:50:12] <-- budlust has left IRC (Remote host closed the connection)
[13:50:43] --> budlust has joined #GemRb
[13:53:48] <SiENcE> which git repository should i take to build an current version of gemrb?
[13:53:58] <fuzzie> the sourceforge one
[13:54:46] <SiENcE> git://gemrb.git.sourceforge.net/gitroot/gemrb/gemrb ?
[13:54:54] <fuzzie> yes
[13:54:56] <SiENcE> ok
[14:06:16] <lynxlynxlynx> cool, it works
[14:09:46] <fuzzie> :)
[14:17:37] <-- Maighstir_laptop has left IRC (Ping timeout: 276 seconds)
[14:20:03] <fuzzie> lynxlynxlynx: the colours for the message window are hard-coded into the source right now?
[14:20:13] <lynxlynxlynx> yes
[14:20:16] <fuzzie> lynxlynxlynx: if so, can i add a 2da?
[14:20:22] <lynxlynxlynx> yes, please
[14:20:42] <lynxlynxlynx> currently only the names are colored specially, everything else is in hex
[14:21:20] <lynxlynxlynx> the original seems to have used palleted fonts though, because if you look carefully, for example the xp text isn't all of the same color
[14:22:17] <fuzzie> within one game?
[14:26:11] <lynxlynxlynx> yes
[14:28:25] <fuzzie> hmm
[14:28:31] <fuzzie> i have been colour-sampling from ToB
[14:28:51] <fuzzie> and they're the same for quest XP and monster-kill XP across the whole party
[14:29:03] <fuzzie> but perhaps i misunderstand, or i'm just not managing to reproduce
[14:31:44] <lynxlynxlynx> i mean in the character sense
[14:32:02] <lynxlynxlynx> i should check an original screenshot
[14:33:03] <fuzzie> there's quest xp like http://fuzzie.org/nfs2/bg2/one.png
[14:42:43] <lynxlynxlynx> yep
[14:43:14] <-- budlust has left IRC (Read error: Operation timed out)
[14:43:15] <lynxlynxlynx> using a color picker it is clear that there are different colors or some sort of blending
[14:44:03] <lynxlynxlynx> resizing probably isn't as accurate, but this is nicely visible at 1000%
[14:44:10] <fuzzie> oh, right
[14:44:15] <fuzzie> you mean, their fonts aren't ugly :)
[14:44:38] <fuzzie> i think they AA their fonts and we don't
[14:44:58] <fuzzie> i mean, maybe it is paletted, i don't know
[14:45:11] <fuzzie> i'll look later
[14:45:23] <lynxlynxlynx> looks like we already do something like that
[14:45:36] <lynxlynxlynx> (looking at your promenade screenshot)
[14:45:48] <lynxlynxlynx> so single color looks fine to me
[14:48:02] --> |Cable| has joined #GemRb
[14:49:20] --> budlust has joined #GemRb
[14:54:40] <-- budlust has left IRC (Ping timeout: 240 seconds)
[14:56:10] --> budlust has joined #GemRb
[14:58:06] <fuzzie> ok, i just got that 'you cannot rest because you do not have control of all your party members' message in the original
[14:58:10] <fuzzie> so i guess i do seei t :)
[15:00:42] <-- budlust has left IRC (Ping timeout: 265 seconds)
[15:28:02] <CIA-26> GemRB: 03lynxlupodian * r40728bf18089 10gemrb/gemrb/core/ (GUI/GameControl.cpp Scriptable/Actor.cpp):
[15:28:02] <CIA-26> GemRB: highlight the speaker's feet circle by whitening
[15:28:02] <CIA-26> GemRB: currently only works for real dialogs
[15:37:39] --> Maighstir_laptop has joined #GemRb
[15:38:16] <Maighstir_laptop> I'll let you in on a secret: GemRB doesn't like GemRB.SetNextScript (GemRB.GetToken("NextScript")) when the token NextScript isn't set :-P
[15:39:10] <lynxlynxlynx> gasp! :P
[15:39:48] <Maighstir_laptop> had to force the computer to power off through the power button
[15:42:03] <lynxlynxlynx> i'll go fix that
[15:43:58] <lynxlynxlynx> http://pastebin.com/QHMExPDK <-- try with this
[15:47:49] --> budlust has joined #GemRb
[15:52:51] <-- budlust has left IRC (Ping timeout: 264 seconds)
[15:55:15] <Maighstir_laptop> much better
[15:56:36] <lynxlynxlynx> :)
[15:56:50] <CIA-26> GemRB: 03lynxlupodian * rec2655540eb8 10gemrb/gemrb/plugins/GUIScript/GUIScript.cpp: bail out of GemRB_SetNextScript if an empty parameter is passed
[16:00:51] --> budlust has joined #GemRb
[16:05:21] <CIA-26> GemRB: 03lynxlupodian * rf328cba101e1 10gemrb/gemrb/GUIScripts/iwd/GUICommonWindows.py: iwd: applied 6d2e6c022e (try to preserve selection across screens)
[16:05:37] --> Avenger has joined #GemRb
[16:05:37] --- ChanServ gives channel operator status to Avenger
[16:05:52] <Avenger> hi
[16:05:55] <lynxlynxlynx> oj
[16:07:50] <CIA-26> GemRB: 03lynxlupodian * re7137b72b4a1 10gemrb/gemrb/GUIScripts/iwd2/GUICommonWindows.py: iwd2: applied 6d2e6c022e (try to preserve selection across screens)
[16:08:16] <fuzzie> hi Avenger
[16:09:14] <Avenger> i learned a little more about the queue :)
[16:09:16] <fuzzie> Avenger: you have some idea why Hold's SPMINDAT plays underneath the player and not above?
[16:09:40] <Avenger> in gemrb?
[16:09:42] <fuzzie> the y is +1, the z is -32, no weird flags, but maybe there's some flag in the effect, i didn't look yet
[16:09:49] <fuzzie> erm, the z is 32
[16:10:00] <fuzzie> and do tell about the queue :)
[16:10:12] <fuzzie> (i guess this is still bg2?)
[16:10:25] <fuzzie> oh, and Edheldil would like a copy of your RE stuff
[16:11:27] <Avenger> no idea
[16:12:07] <Avenger> i will send the mail once i'm back to linux ;)
[16:12:58] <Avenger> well, last time i was a bit confused about which actions clear the queue before they add an action to it
[16:13:20] <fuzzie> ah
[16:13:51] <fuzzie> i know that ActionOverride only clears any running blocking action
[16:13:58] <fuzzie> i don't know if that helps
[16:14:09] <fuzzie> probably not, i guess, since ActionOverride is not an action :|
[16:14:25] <Avenger> i know what breaks currently running actions, i found that earlier
[16:14:34] <Avenger> setting 2b8h does it
[16:14:41] <fuzzie> hehe. not too helpful to us :P
[16:15:02] <-- budlust has left IRC (Ping timeout: 252 seconds)
[16:15:59] <Avenger> i guess, it isn't helpful if i tell you how groupattack works either :)
[16:16:09] <Avenger> because it is probably never used
[16:17:04] <Avenger> basically, it is an unlimited AttackNoSound. Normal attack stops when the target becomes invalid, GroupAttack simply reevaluates the target and starts again
[16:17:29] <Avenger> it does this by pushing itself on top of the queue, then pushing an attacknosound too
[16:17:32] <fuzzie> it is not used by scripts, but maybe internally?
[16:17:57] <Avenger> not sure, the only way to break from it (and continue executing) is to set 0x2b8 to 1 :0
[16:19:43] <Avenger> the worst news: as far as i see clearactions just pushes a noaction on top of the queue, and sets 0x2b8 to 1
[16:19:45] <fuzzie> beserk-related?
[16:19:57] <Avenger> clearactions doesn't seem to purge the queue
[16:20:40] <Avenger> i don't quite see why it needs that push
[16:24:10] <Avenger> noaction clears the modal state only
[16:24:35] <Avenger> so clearactions just breaks any long running actions (blocking actions), and brings the actor out of any modal stuff
[16:27:31] <Avenger> oh i forgot, the nointerrupt flag can override this 2b8 thing
[16:32:48] <Avenger> fuzzie: does the berserk action work in normal scripting ?
[16:34:11] <Avenger> because from what i see, it cannot :)
[16:34:27] <Avenger> it has an int0parameter, a counter for berserk time
[16:34:38] <Avenger> heh
[16:34:51] <SiENcE> i got gemrb working ingame on my a320 dingoo with 32mb ram using zram (20mb) and swaptoSDCard(30mb)
[16:34:59] <SiENcE> currently i play BG2
[16:35:08] <SiENcE> but without music
[16:35:16] <Avenger> grats
[16:35:19] <SiENcE> slows down and uses too much ram
[16:35:27] <Avenger> an what caused the crash
[16:35:39] <SiENcE> thx
[16:35:40] <SiENcE> ?
[16:36:01] <Avenger> i heard you had some problems earlier
[16:36:08] <SiENcE> all problems solved
[16:36:16] <Avenger> nothing we should know about?
[16:36:25] <fuzzie> gemrb was making two copies of every TIS tile
[16:36:26] <SiENcE> nothing
[16:36:36] <SiENcE> only porting issues and some python problems
[16:36:55] <fuzzie> but otherwise boring things :)
[16:37:03] <fuzzie> you make a video, SiENcE?
[16:37:06] <SiENcE> gemrb must be optimized, screen redraw is done every frame...thats really bad
[16:37:14] <SiENcE> fuzzie, of couse later
[16:37:23] <SiENcE> and sound using openal is damn slow
[16:37:26] <Maighstir_laptop> what's the flag for an editable editbox? IE_GUI_EDIT_EDITABLE?
[16:37:29] <SiENcE> better use sdlaudio
[16:37:38] <SiENcE> and caching must be optimized
[16:37:58] <SiENcE> and please don't load the whole area tileset into ram
[16:38:06] <SiENcE> or make an low memory option in config
[16:38:10] <SiENcE> thats all
[16:38:12] <Avenger> yeah, tileset caching
[16:38:12] <SiENcE> for now
[16:38:18] <fuzzie> sure, we'll do it all :)
[16:38:28] <SiENcE> and i help and send patched
[16:38:34] <Avenger> i don't promise sdl audio ;)
[16:38:34] <SiENcE> for downscaling and so on
[16:38:42] <fuzzie> i think i'll do sdl audio
[16:38:45] <Avenger> oh good
[16:38:51] <SiENcE> or direkt sound through oss or alsa
[16:38:55] <fuzzie> it doesn't have to be good :P
[16:39:07] <Avenger> will that know all the positional stuff or it will be just a cheap option?
[16:39:13] <fuzzie> a cheap option
[16:39:27] <SiENcE> yes
[16:39:29] <fuzzie> modern openal always mixes floating-point in surround sound
[16:39:30] <SiENcE> for handhelds
[16:39:37] <fuzzie> and handhelds have to emulate fp
[16:39:41] <Avenger> yeah, that sounds good
[16:39:54] <Avenger> you don't really need positional sound there anyway :)
[16:40:11] <SiENcE> where can i configure the keysets?
[16:40:24] <SiENcE> i need to disable the scrolling via cursor keys
[16:40:32] <Avenger> i don't think you can configure it yet
[16:40:38] <SiENcE> mh
[16:40:40] <Avenger> you need to hack the code
[16:40:41] <SiENcE> ok
[16:40:58] <SiENcE> anyway...good work on gemrb so far
[16:41:20] <Avenger> thanks :) it will be a 10 years old project this month
[16:41:40] <SiENcE> a lot better than before...i think i made the 3rd party windows version 4years ago
[16:41:57] <Avenger> yeah it starts to work
[16:41:59] <SiENcE> our Iris2 project is also 10years old now
[16:43:01] <SiENcE> ok cya
[16:43:05] <Avenger> see you
[16:43:08] <SiENcE> i make a video laters
[16:43:20] <-- SiENcE has left IRC (Quit: @all: cya)
[16:44:53] <Maighstir_laptop> hmm, why can't I edit the editbox?
[16:44:54] <Maighstir_laptop> FileNameEditBox = ExportWindow.GetControl (7)
[16:44:54] <Maighstir_laptop> FileNameEditBox.SetEvent (IE_GUI_EDIT_ON_CHANGE, FileNameChange)
[16:47:57] --> budlust has joined #GemRb
[16:48:25] <Avenger> FileNameChange is there?
[16:48:39] <Maighstir_laptop> yes
[16:49:14] <Avenger> hmm, the cursor doesn't appear in the editbox, or it appears, but you see bugs on the console
[16:49:23] <Avenger> if it doesn't appear, then it is a focus problem
[16:49:45] <Maighstir_laptop> cursor doesn't appear when I click on the box
[16:49:58] <Avenger> hmm, not even when you click... that's odd
[16:50:15] <Avenger> what happens if you hit enter
[16:51:34] <Avenger> make sure you added this: FileNameEditBox.SetStatus(IE_GUI_CONTROL_FOCUSED)
[16:51:52] <Avenger> though i think clicking on it can still work without that
[16:53:00] <-- budlust has left IRC (Remote host closed the connection)
[16:53:32] --> budlust has joined #GemRb
[16:54:52] <Maighstir_laptop> The character name window obviously works, so I tried to copy that code, but it doesn't work for the export window
[16:55:05] <Maighstir_laptop> might be something completely different
[16:55:40] <fuzzie> the control id is right?
[16:55:59] <Maighstir_laptop> according to DLTCEP, yes
[16:56:54] <Maighstir_laptop> http://pastebin.com/zZBPQgxi current code
[16:58:09] <-- budlust has left IRC (Ping timeout: 265 seconds)
[17:00:07] <Avenger> fuzzie: i'm positive: Berserk() used in bg2 is a nop :)
[17:00:28] <Avenger> You gotta change it to Berserk(I:time*) to actually have any use
[17:00:50] <Avenger> then Berserk(20) will keep the guy in berserk state for 20 ai cycles
[17:01:53] <fuzzie> hmm
[17:02:29] <fuzzie> yes, it seems not used
[17:02:44] <Avenger> it is used in bg1: kiel.dlg
[17:02:58] <fuzzie> d/bg2/kiel.d:Berserk()~ JOURNAL #23540 /* ~~ */ EXIT
[17:03:09] <Avenger> that is inactive in bg2
[17:03:15] <Avenger> just residue from bg1
[17:03:29] <fuzzie> hm, my bg1 has no kiel.d
[17:03:34] <Avenger> hmm
[17:03:38] <Avenger> odd
[17:03:57] <Avenger> i do
[17:04:19] <Avenger> it is kiel.dlg, d is a weidu output
[17:04:24] <fuzzie> yes
[17:04:28] * Avenger ducks. :)
[17:04:53] <Avenger> kiel dlg is in dialog.bif
[17:05:08] <Avenger> of bg1
[17:05:13] <Avenger> well, you got full totsc?
[17:05:24] <fuzzie> there is a good question
[17:05:28] <fuzzie> trying to work out how to tell which version
[17:06:09] <fuzzie> yes, it is full totsc
[17:07:19] <fuzzie> it is weidu being stupid
[17:07:30] <fuzzie> grr
[17:07:33] <Avenger> ok
[17:07:49] <Avenger> well, nevermind
[17:08:03] <fuzzie> well, it means i have to go fix this :)
[17:08:06] <Avenger> the point is, Berserk is not working in bg2
[17:08:10] <fuzzie> but yes, i see it is left over
[17:08:22] <Avenger> and i'm unsure about bg1
[17:08:25] <Avenger> since it turns enemy anyway
[17:08:50] <Avenger> so, you wanted to know about berserk, the action is surely not too useful
[17:08:57] <Avenger> so you want to see the opcode now?
[17:09:13] <fuzzie> it is not important :) only if you are bored
[17:09:15] <Avenger> i guess it is used internally btw
[17:09:27] <Avenger> somewhere in get hit/morale check
[17:09:30] <Avenger> i have to find that
[17:09:41] <fuzzie> i have bigger and bigger list of things i am working on..
[17:09:56] --> budlust has joined #GemRb
[17:09:59] <Avenger> i won't stop reversing for some time now
[17:10:33] <Avenger> now the code is not threatening any more :)
[17:11:26] <fuzzie> ok
[17:11:28] <lynxlynxlynx> can you fix the applykit stuff first, please?
[17:11:42] <fuzzie> where does it break?
[17:11:46] <fuzzie> i think there was some confusion
[17:12:06] <Avenger> what uses applykit?
[17:12:10] <lynxlynxlynx> dual classed chars don't get their first class/kit stuff
[17:12:32] <Avenger> ah you meant the level up part, not the action
[17:12:36] <Avenger> i guess
[17:12:51] <lynxlynxlynx> the action still boggles me, so i left it alone
[17:13:13] <lynxlynxlynx> a pcf_kit with this stuff would be good too, since some guiscript code could be removed
[17:13:16] <Avenger> don't bother with the action, i guess it is not used except in some obscure places, and maybe some kit mods
[17:14:07] <Avenger> yes, if pcf_kit would be there, you could probably get rid of the script side kit code
[17:14:42] <Avenger> what is set first, level or kit/class?
[17:14:58] <Avenger> i think i need both/all three
[17:15:47] <lynxlynxlynx> first class, then level based on it
[17:15:52] <lynxlynxlynx> then kit if one is chosen
[17:16:16] <lynxlynxlynx> atleast in cg
[17:17:44] <-- budlust has left IRC (Ping timeout: 252 seconds)
[17:33:44] --> edheldil_ has joined #GemRb
[17:56:28] <Avenger> fuzzie, i'm sure cutscenes usually don't add setinterrupt actions when queueing the cutscene block. There is a check, if the cutscene object is an actor, no interrupt trick, the extra actions are added only if it is not an actor
[17:57:00] <Avenger> so our cutscene code is ~good about the queueing
[17:57:24] <Avenger> there is one thing, before queueing, 0x2b8 is set to 1 :)
[17:58:15] <Avenger> so any current blocking action is broken, then the actions are simply added to the end of the queue
[18:00:36] <fuzzie> sure
[18:00:38] <fuzzie> don't we do that?
[18:00:56] <fuzzie> (ReleaseCurrentAction is our equivalent)
[18:01:41] <fuzzie> we sure do
[18:01:42] <Avenger> yes
[18:02:18] <Avenger> i just tell you the current standing of my scientific research :) it is good if theory and practice support each other
[18:02:21] <fuzzie> are you sure you read this right?
[18:02:25] <Avenger> yes
[18:02:29] <fuzzie> non-actors are non-interruptible by default
[18:02:34] <Avenger> no
[18:02:45] <Avenger> non-actors have an interruptible flag
[18:02:47] <fuzzie> meh, this is some game difference?
[18:02:54] <Avenger> bg2 i am in
[18:03:02] <fuzzie> i forget where we tested this, maybe bg1
[18:03:18] <fuzzie> but the 'interruptible' flag of non-actors is 'false' by default, according to all our tests
[18:03:18] <Avenger> we already talked about the interruptible flag, ages ago
[18:04:01] <Avenger> well, i don't know what blocking actions are there that work in a non-actor
[18:04:12] <fuzzie> oh
[18:04:12] <Avenger> i mean, real blocking actions
[18:04:16] <fuzzie> ok, we talk about different flags?
[18:04:30] <fuzzie> i don't talk about non-interruptible actions
[18:05:01] <fuzzie> just the SetInterrupt stuff
[18:05:06] <Avenger> to me a blocking action is any action that is not executed immediately
[18:05:14] <fuzzie> ok
[18:05:21] <fuzzie> well, think of Wait as your example :)
[18:05:23] <Avenger> that is what we mark by AF_BLOCKING
[18:05:30] <Avenger> ok, wait is special
[18:05:49] <Avenger> for example: actionlistempty will return true, if the current action is wait, LOL
[18:05:56] <Avenger> even if there is something AFTER wait
[18:06:03] <Avenger> at least that's what i read here
[18:06:07] <fuzzie> not unlikely
[18:06:13] <fuzzie> actionlistempty is only used by combat scripts i think
[18:06:18] <Avenger> it is a major bug to me
[18:06:28] <fuzzie> so wait is a stupid thing to be running :P
[18:06:37] <fuzzie> but i mean, it is a fine example for this discussion :)
[18:07:12] <Avenger> i would change it like: if the queue is 0 long-empty, if it is >1 long: it is not empty, if it is 1 long and the action is wait/noaction: it is empty
[18:07:26] <Avenger> they just check if it is empty, or the top action is noaction/wait
[18:07:28] <fuzzie> sure, we can add some parameter check
[18:07:46] <fuzzie> int0Param is 1 means "don't be stupid" :p
[18:07:58] <Avenger> uhm
[18:08:27] <Avenger> so, the biggest conflict of research here is this cutscene queueing
[18:08:49] <Avenger> taimon or some other guy said it adds the setinterrupt to everyone
[18:08:51] <fuzzie> it doesn't make much sense for it to queue SetInterrupt actions only on non-actors
[18:08:57] <fuzzie> it would make sense if it only queued them on actors
[18:09:09] <Avenger> when i found it first, i thought the same!
[18:11:31] <Avenger> you will see it when i send it to you
[18:12:01] <fuzzie> i wonder if this is a leftover from bg1
[18:13:27] <Avenger> i don't even know any cutscene scriptable that is non-actor
[18:15:34] <fuzzie> yes
[18:15:35] <fuzzie> so it is odd
[18:17:10] <fuzzie> i see none
[18:17:31] --> budlust has joined #GemRb
[18:24:58] <-- edheldil_ has left IRC (Ping timeout: 260 seconds)
[18:26:33] <-- budlust has left IRC (Remote host closed the connection)
[18:27:22] --> budlust has joined #GemRb
[18:37:18] <-- budlust has left IRC (Remote host closed the connection)
[18:37:55] --> budlust has joined #GemRb
[18:39:54] <lynxlynxlynx> Performattack for Black Bear, target is: Squirrel :)
[18:41:26] <fuzzie> hehe
[18:42:39] <-- budlust has left IRC (Ping timeout: 240 seconds)
[18:46:47] <lynxlynxlynx> they keep following me through areas in tob
[18:46:55] <lynxlynxlynx> don't know where they are though
[18:47:16] <fuzzie> ah, we keep areas running for a while
[18:47:23] <fuzzie> i don't know the exact rules for that
[18:47:40] <fuzzie> there's some stuff from devSin/Avenger/etc on the forums about it
[18:47:59] <fuzzie> in particular some weird stuff about how actors are 'carried' along with the party when they run an action to move to a new area
[18:48:18] <fuzzie> and you can see in ToB that the pocket plane area keeps being loaded..
[18:51:16] --> edheldil_ has joined #GemRb
[19:01:08] --> budlust has joined #GemRb
[19:09:51] <Avenger> heh, randomwalk seems to check class
[19:10:30] <Avenger> i think it is to keep group of random critters together
[19:10:32] <fuzzie> seperately from the standard id matching checks?
[19:10:51] <fuzzie> i thought it matched a whole bunch of fields
[19:10:51] <Avenger> it is flock mentality :)
[19:11:02] <fuzzie> horribly overcomplicated
[19:11:20] <Avenger> it could
[19:11:36] <Avenger> the class check stands out because it is a call
[19:11:44] <Avenger> the rest are direct field access
[19:17:27] <Avenger> yes it checks all ids targeting field
[19:19:04] <fuzzie> meh :P
[19:19:07] <fuzzie> it sounds overcomplicated
[19:19:19] <fuzzie> but if you can make some notes, would be good to implement, i guess
[19:23:42] <Avenger> strongestof object: it handles fighter/ranger/paladin specially (probably gives them priority)
[19:24:28] <fuzzie> not barbarian?
[19:25:01] <fuzzie> we check fighter/barbarian/ranger/paladin for some stuff, but i guess in bg2 barbarian is just a kit
[19:25:19] <Avenger> barbarian is a fighter kit
[19:25:56] <fuzzie> i think it's a whole class in iwd2
[19:26:12] <lynxlynxlynx> yes
[19:26:23] <lynxlynxlynx> but iwd2 is sane about this stuff
[19:26:27] <fuzzie> but you don't look at iwd2 so this is irrelevant, sorry :)
[19:35:36] <edheldil_> Avenger: could you send me your disasm notes? I ran pst through ida and use the result as a light bedtime reading
[19:36:24] <Avenger> yes i will mail you once i'm in linux
[19:36:37] <Avenger> i don't have much of pst, though
[19:36:41] <Avenger> opcodes mostly
[19:37:15] <Avenger> and projectiles
[19:37:54] <edheldil_> it will be helpful to get some sense of meaning :)
[19:38:39] <fuzzie> Scient (pst fixpack exe hacker) probably has best set of pst notes (IDA file too, i think)
[19:38:49] <fuzzie> don't know if he's still around though
[19:44:41] <lynxlynxlynx> hmpf, got stuck in tob
[19:45:45] <fuzzie> oh?
[19:45:56] <lynxlynxlynx> all of a sudden the pp exit cutscene gets stuck somewhere or with more careful party placement, the old locations are supposedly bad (top of watcher's keep)
[19:46:26] <lynxlynxlynx> argh, couldn't restore party member 1!
[19:47:05] <fuzzie> did you add a party member since you left?
[19:47:12] <lynxlynxlynx> no
[19:47:33] <fuzzie> and you have >1 member?
[19:47:42] <lynxlynxlynx> 6
[19:47:48] <lynxlynxlynx> all of them error
[19:47:56] <fuzzie> including 0?
[19:48:43] <lynxlynxlynx> 1-6
[19:49:04] <fuzzie> oh right, it's printing i+1
[19:49:57] <fuzzie> it means that it doesn't have any old locations, as i guess you summarised
[19:50:05] <lynxlynxlynx> the cutscene gets stuck before the visuals kick in and if the ordering is honored, that means only CutSceneId(Player1) gets run
[19:50:21] <lynxlynxlynx> cut221a.baf
[19:51:09] <fuzzie> means it doesn't get run :P
[19:51:22] <lynxlynxlynx> the gui gets hidden though
[19:51:51] <fuzzie> that's done beforehand, i think
[19:52:04] <lynxlynxlynx> ok, now i've succeeded again
[19:52:24] <lynxlynxlynx> game->GetPlaneLocationCount() is 0
[19:52:30] <fuzzie> yes
[19:53:24] <fuzzie> so i guess fx_cutscene2 didn't get run after the last ExitPocketPlane
[19:53:35] <lynxlynxlynx> the only thing that clears them is the exit pp action
[19:53:51] <fuzzie> i'm not sure i'd trust Avenger's comment on the clearing action
[19:54:00] <fuzzie> if you can run original ToB, it should be trivial to check
[19:54:07] <fuzzie> i can't run it right now though
[19:54:28] <lynxlynxlynx> well, it should have exited before clearing it
[19:54:29] <Maighstir_laptop> are thief proficiencies randomised in chargen?
[19:54:38] <lynxlynxlynx> no
[19:54:47] <Maighstir_laptop> hmm... weird
[19:55:03] <fuzzie> lynxlynxlynx: but maybe that already got executed a previous time, and it didn't get restored?
[19:55:10] <Maighstir_laptop> doing a few tests and getting strange results
[19:55:38] <lynxlynxlynx> there was some extra teleporting going on, but the last pp entry was done by me
[19:57:03] <lynxlynxlynx> you kill yagashura and the solar teleports you to the pp, then there's more of the talk and then you open challenge2 and then you get teleported back to the siege camp where m shows up and there's more talking and finally you're free to do what you will again
[19:57:54] <lynxlynxlynx> i went to wk to sell some stuff and then teleported back to the pp, finished ch2 and this is where i'm stuck
[20:00:38] <fuzzie> well, the tricky bit seems to be whether fx_cutscene2 got run or not
[20:00:44] <fuzzie> i have no real idea how this works outside the load/save
[20:01:14] <lynxlynxlynx> i don't know why this time would be special
[20:01:22] <lynxlynxlynx> the pp worked before
[20:01:47] <fuzzie> i am puzzling over why VMware Player Setup is giving me IESDP :(
[20:02:22] <lynxlynxlynx> huh
[20:03:14] <fuzzie> argh, it does it after a reboot
[20:03:15] <fuzzie> stop that!
[20:05:06] <-- budlust has left IRC (Quit: Changing server)
[20:05:56] <fuzzie> seems that vmware setup puts some html files in a temp directory and then displays index.html, and that index.html is really the IESDP script actions page. huh. :|
[20:06:17] --> budlust has joined #GemRb
[20:08:08] <Maighstir_laptop> Take a look at these images http://maighstir.imgur.com/import_character , from my understanding they should be the same (barring the portrait to distinguish them), but the stats differ between all of them. One is created through the GemRB pregen, the other through original BG1's pregen, then both are loaded in both engines to compare differences (see descriptions as for which is which)
[20:08:52] <Maighstir_laptop> s/stats/proficiences/
[20:09:10] <lynxlynxlynx> iirc the original does the dex bonus dynamically and doesn't show it in cg
[20:09:33] <lynxlynxlynx> so that would only be the racial bonus there plus whatever you distributed
[20:09:33] <fuzzie> the BG->gemrb one looks bad though
[20:09:36] <fuzzie> 0/0/30/0?
[20:09:57] <fuzzie> oh, we create the dex bonus at chargen time and never apply it?
[20:10:37] <lynxlynxlynx> we apply it
[20:10:51] <lynxlynxlynx> well, i'm not sure about bg1
[20:11:02] <fuzzie> i mean, outside of chargen
[20:11:15] <lynxlynxlynx> i don't think so
[20:11:27] <lynxlynxlynx> it's a todo
[20:13:14] <fuzzie> hum
[20:13:17] <fuzzie> difficult to fix :|
[20:13:38] <lynxlynxlynx> not really
[20:14:19] <lynxlynxlynx> it's like any other stat based bonus
[20:14:34] <fuzzie> do we handle any of those?
[20:14:41] <lynxlynxlynx> plenty
[20:15:18] <fuzzie> right, RefreshPCStats
[20:15:21] <fuzzie> i completely forgot
[20:15:40] <fuzzie> more broken stuff in here, meh
[20:35:04] --> SiENcE has joined #GemRb
[20:36:06] <lynxlynxlynx> cut217c.baf is also problematic, the spimpt bam keeps on playing over and over again (the fr one plays once as expected)
[20:37:18] <edheldil_> SiENcE: I have heard that you have ps:t ida files?
[20:37:37] <lynxlynxlynx> no, that's scient
[20:37:51] <lynxlynxlynx> iirc he's from ppg
[20:38:56] <edheldil_> ah
[20:39:05] <edheldil_> good night
[20:39:42] <fuzzie> lynxlynxlynx: any idea which flags spimpt.vvc has?
[20:40:20] <lynxlynxlynx> spimpt.vvc doesn't exist
[20:40:31] <lynxlynxlynx> the fr one does though, so maybe that's why it works properly
[20:41:03] <-- edheldil_ has left IRC (Read error: Operation timed out)
[20:41:31] <fuzzie> oh, ugh
[20:42:00] <fuzzie> GameData::GetScriptedAnimation just makes some new VVC and doesn't configure it at all?
[20:42:59] <fuzzie> it should probably be setting PlayOnce() itself, or maybe we should get that if iterations==0
[20:43:05] <fuzzie> horrible :)
[20:44:14] <fuzzie> something to note on the todo page (CreateVisualEffect with only a BAM leads to infinitely-looping play)?
[20:45:37] <lynxlynxlynx> you don't think it is a quick fix?
[20:46:28] <lynxlynxlynx> i wonder, maybe the sparkles are also bams and this is why they are persistent
[20:48:01] <Maighstir_laptop> Changes to BG1 chargen/pregen: http://pastebin.com/72budg0F , the cancel button in the export window doesn't work (I didn't manage to figure out how), and you're thrown back to the start menu, not the multiplayer menu as in the original (small difference, not really worth the trouble, by my reckoning)
[20:50:44] <lynxlynxlynx> all those stats are saved in the player, not just in those variables?
[20:51:32] <fuzzie> the sparkles are just very broken C++ code
[20:52:23] <fuzzie> and replacing GetVar("Class") with GetPlayerStat directly is wrong, i thought
[20:52:37] <fuzzie> it's done right below there
[20:53:02] <fuzzie> maybe we should just unset all the gui vars once the stats are set
[20:53:06] <fuzzie> make it clear where chargen is broken
[20:55:15] <fuzzie> but it looks ok other than the pdoll stuff not looking too careful
[20:58:40] <Maighstir_laptop> the change there was GemRB.GetPlayerStat (MyChar, IE_SEX) instead of GemRB.GetVar("Gender") (same for the other two), because it didn't know the right values so it couldn't show a doll
[20:58:53] <Maighstir_laptop> when importing
[20:58:56] <SiENcE> hey
[20:59:18] <lynxlynxlynx> fuzzie: actually it does help with sparkles, they don't restart
[20:59:33] <SiENcE> question, why gemrb decompresses everytime it loads the same *.wed file?
[20:59:50] <SiENcE> i think its already decompressed in my cache
[21:00:10] <fuzzie> lynxlynxlynx: are we talking about different sparkles?
[21:00:31] <fuzzie> SiENcE: do you restart it in the middle?
[21:01:16] <lynxlynxlynx> they are still buggy, eg. not like in the original
[21:01:28] <lynxlynxlynx> sparkles, the stuff for example dispel magic leaves behind
[21:01:35] <fuzzie> the white dots?
[21:01:56] <lynxlynxlynx> i'm sure we talk about the same thing, since it is loading sprklclr.2da
[21:01:58] <lynxlynxlynx> yes
[21:02:00] <fuzzie> maybe i am too tired for this :)
[21:02:23] <lynxlynxlynx> now it only does it once, so the whole thing is a lot better
[21:02:32] <fuzzie> but the sparkles stuff is in Particles.cpp and has nothing to do with the animation factory
[21:02:52] <fuzzie> well, that is not entirely true
[21:03:29] <fuzzie> but if it fixed that, i think i don't understand anything about that :)
[21:03:30] <lynxlynxlynx> i'm very tempted to commit this PlayOnce
[21:03:50] <fuzzie> but every time so far someone thought sparkles were fixed in some way, it was coincidence
[21:04:01] <fuzzie> i'm sure the funniest was when Avenger ran the original engine to test, by mistake
[21:04:54] <fuzzie> i think the trouble with putting a PlayOnce in the GetScriptedAnimation bit is that it is wrong
[21:05:00] <lynxlynxlynx> now the white fog just keeps dropping all over the map, but it never returns
[21:05:21] <fuzzie> because it'll break any BAM which needs to loo[
[21:05:22] <fuzzie> loop
[21:05:30] <fuzzie> and it seems those exist
[21:05:45] <fuzzie> but you could find the relevant callers, and put it there if there's obviously no need to loop
[21:06:08] <lynxlynxlynx> how to detect when you need to loop?
[21:06:11] <fuzzie> but then i don't know how to tell it was a BAM :(
[21:06:18] <lynxlynxlynx> do you know of a bam like that?
[21:06:30] <fuzzie> well, if you look at GetVVCEffect in GSUtils.cpp, any case with 'iterations' non-zero needs to loop
[21:07:08] <fuzzie> only examples i know by heart are from pst
[21:07:24] <lynxlynxlynx> ok
[21:07:26] <fuzzie> but i guess if you look at that function, you'll see how it works
[21:10:16] <fuzzie> hmm
[21:12:09] <fuzzie> so i guess projectiles make sparkles, maybe they are controlled by animations sometimes
[21:12:41] <fuzzie> but the projectile code seems to always manually call PlayOnce on anything it creates itself
[21:14:11] <fuzzie> so, confusing
[21:15:50] <lynxlynxlynx> isn't there a flag for playonce in projectiles?
[21:17:53] <fuzzie> i mean, when the projectile plays VVCs
[21:18:00] <fuzzie> i am just rambling, sorry, i don't understand how it works :)
[21:18:20] <fuzzie> i only know about the iterations thing because i fixed it at some point
[21:22:50] <-- Avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
[21:23:19] <lynxlynxlynx> yep, almost everything sets PlayOnce itself
[21:29:50] <fuzzie> wait
[21:29:57] <fuzzie> i was talking nonsense above
[21:30:54] <lynxlynxlynx> Maighstir_laptop: does it still work for normal cg though?
[21:30:55] <fuzzie> if GetVVCEffect gets a zero 'iterations', it's going to loop forever unless you call PlayOnce()
[21:31:17] <fuzzie> lynxlynxlynx: bg1 chargen is pretty careful about only setting/using stats everywhere, so it looks ok
[21:32:08] <fuzzie> i would have preferred to fix the VVC looping stuff at a higher level, but unfortunately the PCFs etc ruin that
[21:32:33] <lynxlynxlynx> in the interim it is better to add that in the else case in GetVVCEffect?
[21:32:45] <fuzzie> that sounds like it should always be correct
[21:32:51] <lynxlynxlynx> awesome :)
[21:37:18] --> Avenger has joined #GemRb
[21:37:19] --- ChanServ gives channel operator status to Avenger
[21:37:37] <Maighstir_laptop> lynxlynxlynx: as far as I can see, yes
[21:38:06] <fuzzie> Maighstir_laptop: about my comments: you lookup an index for IE_CLASS elsewhere, but not in the pdoll
[21:38:16] <fuzzie> that's correct? i don't know the tables at all
[21:38:43] <fuzzie> just seems that if an extra lookup is necessary in one place, using the value elsewhere without any lookup is a bit odd
[21:38:43] <Maighstir_laptop> I fixed the pdoll function as well, here
[21:39:43] <lynxlynxlynx> it's just a -1 offset, but it is more safe to always look it up
[21:40:15] <lynxlynxlynx> that way the row order can be changed without ill effect (just in this case though, since we rely on it in other places)
[21:41:34] <Maighstir_laptop> I'll go through it again before I submit it so I can catch any errors, that'll have to wait for tomorrow though
[21:41:36] <SiENcE> sorry fuzzie, i was away
[21:42:00] <-- Avenger has left IRC (Client Quit)
[21:42:02] <SiENcE> no i start from beginning..and everytime it decompresses this file
[21:42:36] <fuzzie> then yes, that is correct, by default the cache is wiped at startup
[21:42:52] <fuzzie> it is not safe to use across sessions because gemrb modifies the decompressed files
[21:43:33] <fuzzie> we should have a proper optional cache for the tile data
[21:43:37] <fuzzie> but we don't :(
[21:46:22] <-- Maighstir_laptop has left IRC (Quit: Maighstir_laptop)
[21:48:54] <lynxlynxlynx> hmm, this doesn't fix implosion though
[21:49:27] <lynxlynxlynx> too bad i lost the previous save
[21:49:54] <fuzzie> well, this is why i was looking for which GetScriptedAnimation call might be responsible
[21:49:56] <lynxlynxlynx> i can run the cutscene manually though, but not tonight
[21:50:31] <lynxlynxlynx> there's only 1-3 that don't either set playonce directly or conditionally
[21:50:43] <lynxlynxlynx> will need to step through it
[21:50:46] <lynxlynxlynx> good night
[21:50:50] <fuzzie> night
[21:51:12] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[22:06:03] <-- budlust has left IRC (Quit: leaving)
[22:15:13] <SiENcE> http://img193.imageshack.us/img193/9576/screenshot21v.png
[22:15:23] <SiENcE> IWD
[22:16:43] <SiENcE> Icewind Dale 1 with GemRb on Dingoo Handheld a320 (using Dingux), with 32mb ram
[22:17:11] <SiENcE> using the nosound plugin due the lack of a flootingpoint unit for openal
[22:20:02] <fuzzie> hehe, really got to mod the UI :)
[22:20:32] <fuzzie> but that is very cool
[22:24:46] <SiENcE> :)
[22:28:49] <SiENcE> but when i go outside
[22:29:02] <SiENcE> i cant see the background tile
[22:29:18] <SiENcE> i think the ram it not cleaned carefully
[22:34:37] <SiENcE> in my blog are more pics
[22:34:38] <SiENcE> http://crankgaming.blogspot.com/2010/08/icewind-dale-on-dingux-using-gemrb.html
[22:34:58] <SiENcE> screenshots taked directly from my handheld using a framegrabber
[22:44:42] <SiENcE> i hope you can help me for a sdlaudio plugin?
[22:44:49] <fuzzie> i wonder if some kind of zoom thing would work well
[22:44:57] <fuzzie> if you're using the dpad for mouse anyway
[22:45:36] <fuzzie> and yes, sdlaudio should be easy, but i have exams next week so might not be until after exams :)
[22:50:13] --> SiENcE_ has joined #GemRb
[22:50:26] <SiENcE_> sorry my inet was away
[22:50:35] <SiENcE_> i can wait for sdlaudio :)
[22:50:59] <SiENcE_> first i have to disable the cursor keys for scrolling, because i need them for mouse cursor movement
[22:51:08] <SiENcE_> ok...cya tomorrow
[22:51:28] <-- SiENcE has left IRC (Ping timeout: 265 seconds)
[22:51:33] <-- SiENcE_ has left IRC (Client Quit)
[22:57:28] --> nickdaly has joined #GemRb
[23:43:09] --> nickdaly` has joined #GemRb
[23:45:04] <-- nickdaly has left IRC (Ping timeout: 276 seconds)