#gemrb@irc.freenode.net logs for 24 Jun 2009 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:48:08] --- barraTV is now known as barraAway
[01:05:25] --> |Cable| has joined #gemrb
[02:23:52] <-- dawid has left IRC ("Leaving.")
[02:31:37] --> xrogaan has joined #gemrb
[03:28:41] --> pupnik__ has joined #gemrb
[03:39:29] <-- pupnik has left IRC (Read error: 101 (Network is unreachable))
[03:41:32] <-- pupnik_ has left IRC (Read error: 110 (Connection timed out))
[06:51:30] --> lynxlynxlynx has joined #gemrb
[06:51:30] --- ChanServ gives channel operator status to lynxlynxlynx
[07:33:12] <-- |Cable| has left IRC ("Leaving")
[07:34:38] --> ltleitkie has joined #gemrb
[07:35:07] <ltleitkie> exultbot?
[07:35:56] <ltleitkie> anyway, tried to compile it and it froze the whole system reproducibel in GameScript.o
[07:38:00] <lynxlynxlynx> gcc, glibc, kernel?
[07:42:04] <ltleitkie> are installed, why?
[07:42:17] <lynxlynxlynx> haha
[07:42:22] <ltleitkie> waitamin
[07:51:23] <ltleitkie> 2.6.27.1 mips64el, 4.3.3-10, 2.9-10
[07:55:55] <lynxlynxlynx> mips uh
[07:56:58] <lynxlynxlynx> maybe it's a gcc bug
[07:57:11] <lynxlynxlynx> did you start from clean or just continued from the last time you built?
[07:58:09] <ltleitkie> I freshly checked out and that build hang, did not clean before the retries
[07:58:23] <lynxlynxlynx> ok
[07:58:36] <lynxlynxlynx> is there any interesting output before it freezes?
[07:59:08] <lynxlynxlynx> also, did it really freeze your system or just exhaust all your memory?
[07:59:36] <lynxlynxlynx> you can try building with --enable-debug which will disable all optimisations
[08:03:19] <ltleitkie> no output
[08:18:10] <fuzzie> how much RAM do you have available?
[08:18:33] <fuzzie> but that just leads to lynxlynxlynx's suggestion :)
[08:23:02] <ltleitkie> 1 GB
[08:24:13] <fuzzie> That is strange, then, gcc on mipsel's pretty good.
[08:26:54] <ltleitkie> brb
[08:27:04] <lynxlynxlynx> huh, the VB_DAMAGE is displayed even if i comment out the only caller
[08:27:17] <-- ltleitkie has left IRC ("Page closed")
[08:33:22] <lynxlynxlynx> ah, generic action
[08:39:37] <fuzzie> I think I have an actual plan for fixing all of this combat/spell timing once and for all, now.
[08:40:10] <lynxlynxlynx> :D
[08:55:31] <lynxlynxlynx> i see why we display the xp award before the damage that caused it
[08:56:09] <lynxlynxlynx> Damage first does a NewBase to find the damage and if that is enough, the pcf kills the actor
[08:57:01] <lynxlynxlynx> this would also mean we don't display any of the damage animations and probably explain many of those "stance overriden by death" messages
[08:57:24] <lynxlynxlynx> making a strong monk now to test chunking ;)
[08:57:33] <fuzzie> Our chunking code is bad too.
[08:57:45] <fuzzie> It's all kinda misorganised.
[08:58:36] <fuzzie> The Die() call should probably happen somewhere else, but I don't want to mess with it even before the release, I already caused one segfault.
[09:00:02] <fuzzie> I think devSin said that our CheckOnDeath() stuff is crazy anyway, we shouldn't be destroying corpses even when chunking.
[09:03:24] <CIA-20> gemrb: 03lynxlupodian * r6609 10/gemrb/trunk/gemrb/GUIScripts/ie_stats.py: sync of guiscript IE_FISTTHAC0 vs IE_FISTHIT
[09:04:58] <lynxlynxlynx> yeah, no chunking happens
[09:05:30] <lynxlynxlynx> fuzzie: i was thinking of just changing those NewBase calls in Damage to not call pcf and call it later manually
[09:05:53] <fuzzie> i think that probably breaks a whole bunch of other code
[09:06:44] <lynxlynxlynx> i'll try anyway, just to see how it works
[09:07:20] <fuzzie> I mean, if you change it in NewBase.
[09:07:34] <fuzzie> Maybe you're intending to manipulate it manually, that'd work, although be kind of horrid.
[09:08:44] <lynxlynxlynx> i meant to just call pcf_hitpoint in Damage later
[09:08:57] <lynxlynxlynx> but i see there is no NewBaseNoPCF
[09:09:59] <fuzzie> I think the NoPCF functions will kill you anyway.
[09:11:43] <lynxlynxlynx> why?
[09:11:50] <lynxlynxlynx> Die is called from the pcf
[09:13:00] <fuzzie> the NoPCF functions will call the pcf to check for dying
[09:13:16] <fuzzie> //might need to run the post change function (stat change can kill actor)
[09:14:58] <lynxlynxlynx> isn't that in the Stat one?
[09:15:41] <fuzzie> aren't they all Stat ones?
[09:15:52] <lynxlynxlynx> no
[09:16:10] <lynxlynxlynx> let's see how this turns out
[09:20:53] <lynxlynxlynx> no chunking :(
[09:21:29] <lynxlynxlynx> but it does fix the output order for xp
[09:22:06] <lynxlynxlynx> http://pastebin.ca/1472277 <
[09:24:05] <fuzzie> well, it's not going to do any harm that way, but that is truly a hack :)
[09:25:18] <lynxlynxlynx> do you have a better idea?
[09:25:25] <fuzzie> You tested how it works for MINHITPOINTS?
[09:25:44] <lynxlynxlynx> let me try the albion
[09:27:02] <fuzzie> The code at the top presumably shouldn't be using NewBase at all, and/or Die() probably shouldn't be called from the pcf.
[09:27:43] <lynxlynxlynx> i'll leave it be then
[09:28:00] <lynxlynxlynx> i couldn't kill the albion while he was shielded
[09:29:16] <fuzzie> Well, I just wondered whether it continued showing damage or not.
[09:29:42] <lynxlynxlynx> i think it did
[09:34:46] <fuzzie> I'll test original engine behaviour later.
[09:40:25] --> pupnik has joined #gemrb
[09:46:55] <fuzzie> lynxlynxlynx: did you ever check illy with immediate combat?
[09:47:04] <lynxlynxlynx> i can do it now
[09:47:20] <lynxlynxlynx> was just about to test something else there
[09:54:05] <lynxlynxlynx> if i attack her, only her mirror image fires
[09:54:13] <lynxlynxlynx> she starts casting if i stop
[09:54:30] <fuzzie> ok, as-expected
[09:56:59] <CIA-20> gemrb: 03lynxlupodian * r6610 10/gemrb/trunk/gemrb/GUIScripts/bg2/GUICommonWindows.py:
[09:56:59] <CIA-20> gemrb: fixed SetupThac0 to not assume the class table has the same order as the thac0 one
[09:56:59] <CIA-20> gemrb: fixes slow monk thac0 progression
[10:06:31] <fuzzie> Creature AI works quite beautifully when I fill it with quick hacks :)
[10:08:16] <lynxlynxlynx> pc ai is already cool
[10:08:48] <lynxlynxlynx> i had viconia memorise all the spells she could and she did a good job at supporting the fight
[10:09:04] <lynxlynxlynx> summoned some undead and kept casting other stuff
[10:09:53] <fuzzie> The spellcaster stuff is pretty good in svn, but the melee/ranged stuff not so much.
[10:11:11] <fuzzie> I didn't look in detail at the spellcasting stuff yet; aren't the casting times wrong?
[10:11:24] <lynxlynxlynx> the ai part is ok, i just set their scripts to melee and enjoy the masacre
[10:11:39] <lynxlynxlynx> yes, many are wrong
[10:15:57] --- barraAway is now known as barra_library
[11:26:36] --> ltleitkie has joined #gemrb
[11:30:25] <CIA-20> gemrb: 03fuzzie * r6611 10/gemrb/trunk/gemrb/plugins/Core/ActorBlock.cpp: WalkTo shouldn't repath if we're already at the destination
[11:30:31] <fuzzie> lynxlynxlynx: that should fix the orientation problems
[11:39:31] <lynxlynxlynx> cool
[11:51:48] <fuzzie> I'm very happy about the new movement stuff. Round and round and round!
[11:52:22] <fuzzie> I'm pretty sure we should be stopping movement when you try and initiate dialog with a neutral actor, but that would spoil it. :)
[12:04:24] <lynxlynxlynx> that was a bit annoying in the original actually
[12:04:56] <lynxlynxlynx> i think if you moved your party and someone talked to the first pc, everyone stopped instead of at least coming close
[12:06:43] <fuzzie> I only tested in bg1, where neutral actors stop moving if you click to talk to them, but party members don't stop when a script tries initiating dialogue with them.
[12:07:37] <fuzzie> Maybe it is different for some scripting actions, or different in bg2, and maybe only for some neutrals, and so on.
[12:08:36] <fuzzie> I still didn't work out why this "slide on hit" thing happens, maybe it's a buggy action.
[12:16:39] <lynxlynxlynx> i haven't noticed it anymore, since i test with an archer recently
[12:16:53] <lynxlynxlynx> let's see if my monk has this problem too
[12:17:37] <lynxlynxlynx> i can confirm the orientation is fixed
[12:20:21] <lynxlynxlynx> at least now illich doesn't attack you when you encounter him
[12:23:05] <lynxlynxlynx> hmpf
[12:23:20] <lynxlynxlynx> a poison trap applied to my monk with poison immunity
[12:23:37] <fuzzie> At least the poison doesn't apply every frame any more :)
[12:24:02] <fuzzie> I still need to fix the PERCENT version.
[12:25:29] <fuzzie> It's a really pity mattinm disappeared, I could do with someone to double-check this combat code.
[12:26:31] <lynxlynxlynx> yeah
[12:26:52] <fuzzie> And he really understood how it worked :)
[12:29:08] <pupnik> email
[12:30:15] <fuzzie> Last time he was around he said he was swamped, so I don't really want to bother him.
[12:35:17] <lynxlynxlynx> orientation during combat is now wierd sometimes
[12:36:33] <lynxlynxlynx> heh, maybe it is just the small folk - the assassin in the trap room also made the druergar slide on hit
[12:37:15] <fuzzie> Pathfinding/positioning/distance/orientation during combat is firmly in the "after release" category, because fixing it properly is a lot of work.
[12:37:40] <fuzzie> I need to fix that var.var thing though, otherwise PS:T's first areas are broken.
[12:38:04] <lynxlynxlynx> seggy with ulvaryl
[12:38:24] <fuzzie> :(
[12:41:46] <lynxlynxlynx> the trap assassin doesn't attack the druergar until the druergar is completely seen by the player :/
[12:42:58] <lynxlynxlynx> hah
[12:43:12] <lynxlynxlynx> the summon trap spawned a bunch of gnolls
[12:43:20] <lynxlynxlynx> that my party has control of :D
[12:43:25] <fuzzie> hah :)
[12:43:30] <lynxlynxlynx> Allegiance: 4 current allegiance:4
[12:44:47] <lynxlynxlynx> ah, the crash is reproducible
[12:44:55] <fuzzie> do you have a backtrace?
[12:45:11] <lynxlynxlynx> fx_replace_creature ends up calling Interface::SummonCreature with a null target
[12:45:33] <lynxlynxlynx> http://pastebin.ca/1472489
[12:47:05] <lynxlynxlynx> the effect doesn't do any sanity checking
[12:47:19] <fuzzie> well, the effect just passes NULL always, it seems?
[12:47:22] <lynxlynxlynx> yeah
[12:47:35] <fuzzie> SummonCreature already has one (target) check.
[12:48:37] <fuzzie> I guess it should probably just be using Owner's area if there's no target?
[12:48:57] <lynxlynxlynx> sounds fine
[12:49:53] <fuzzie> I'm just going through the commits to try and work out how it happened.
[12:52:02] <fuzzie> Maybe fx_replace_creature should be preserving the target? I would add the (target) check and use Owner's area for now. PRobably best for you to do it, if you can reproduce?
[12:53:11] <lynxlynxlynx> seems like i can and i also have save near
[12:58:03] <lynxlynxlynx> i see why my monk is getting all soft about magic - we don't handle resistance yet
[13:05:17] <lynxlynxlynx> ulvy turns into a bat too fast
[13:05:57] <lynxlynxlynx> in the original she slaughtered everyone first; i guess she gets damaged too fast
[13:07:06] <lynxlynxlynx> heh, you can move while asleep, but not attack
[13:07:09] <fuzzie> It happens on HPLT(Myself,10) or if she can't see any more enemies and a PC is present.
[13:07:47] <fuzzie> The damage is more likely, but it could be a See() bug.
[13:08:34] <lynxlynxlynx> i doubt it was below 10
[13:08:41] <lynxlynxlynx> she was injured at most
[13:09:27] <lynxlynxlynx> damn, cloudkill got me
[13:10:28] <CIA-20> gemrb: 03lynxlupodian * r6612 10/gemrb/trunk/gemrb/plugins/ (Core/Interface.cpp FXOpcodes/FXOpc.cpp): fixed a crash in SummonCreature
[13:23:32] <lynxlynxlynx> huh, the cutscene outside is pretty fast
[13:24:19] --> dawid has joined #GemRb
[13:24:21] <fuzzie> Too fast?
[13:24:38] <lynxlynxlynx> everyone is headtexting over each other
[13:24:54] <lynxlynxlynx> looks more like concurrent monologues than nialogs
[13:24:55] <fuzzie> Oh, you're probably using NullSound.
[13:25:02] <lynxlynxlynx> yes
[13:25:04] <fuzzie> The headtext waits for the sound to finish playing.
[13:25:22] <fuzzie> If you disable the sound then gemrb has no way of knowing that. :)
[13:25:48] <lynxlynxlynx> ah
[13:26:08] <wjp> hm, that sounds like a bug in NullSound
[13:26:17] <fuzzie> Well, NullSound doesn't load the WAV files.
[13:26:41] <fuzzie> Maybe it should do so, the sound in loadSound seems sufficient?
[13:26:46] <wjp> s/bug/missing feature/, then :-)
[13:27:40] <fuzzie> The core just waits for the time returned by AudioDrv()->Play().
[13:28:33] <fuzzie> Right now, that's always 1 second for NullSound.
[13:29:57] <dawid> hi there
[13:30:15] <dawid> got that thing working at last yesterday :)
[13:30:52] <dawid> it even looks better than when i played it under windoze but
[13:31:17] <dawid> all npcs are hasteing all the time
[13:31:52] <fuzzie> Heh, that's a new one. Which game, bg2?
[13:32:08] <dawid> yes
[13:32:11] <lynxlynxlynx> another two bad animations ><
[13:32:45] <dawid> i also have noticed it likes to play the beginning animations all time again when i do sth in the menu
[13:33:25] <dawid> oh,and the always look at me when the finish doing sth, that was not the case under windoze earlier
[13:33:30] <dawid> they*
[13:33:48] <fuzzie> That last one should be fixed in latest svn, but only today.
[13:34:09] <dawid> ill pull that in a minute
[13:34:44] <dawid> then i wanted to ask whether that script creating configure works for U
[13:34:47] <fuzzie> The movie stuff in the start menu needs some fixing, even with the intro stuff fixed to play once it still goes back to the SoA/ToB screen and causes more movies there.
[13:34:50] <lynxlynxlynx> Actionoverride failed for Gaelan and now i'm stuck in cutscene mode in his house
[13:34:51] <dawid> i had to modify it
[13:34:58] <fuzzie> Yes, the script should work fine with an up-to-date autoconf/automake/etc.
[13:35:08] <fuzzie> If you had to remove the -W from aclocal then your automake is too old.
[13:35:27] <dawid> the second option as well
[13:35:47] <fuzzie> We should probably just complain about an ancient copy of automake.
[13:35:48] <dawid> and yes i have slotted installs of automake and only newer ones than 1.7 have it
[13:36:01] <dawid> i have 1.10
[13:36:12] <dawid> 1.7 and 1.5 as well
[13:36:14] <fuzzie> automake 1.8 came out in *2004*, your default one should really be 1.10 by now :P
[13:36:21] <dawid> it is
[13:36:33] <lynxlynxlynx> it first tries plain automake
[13:36:40] <lynxlynxlynx> only then looks for versioned binaries
[13:36:55] <lynxlynxlynx> so if yours is pointing to something old, there's gonna be trouble
[13:36:59] <dawid> yes i figured that out but still
[13:37:10] <dawid> automake --version gives me 1.10.2
[13:37:20] <fuzzie> 'aclocal --version', though?
[13:37:37] <lynxlynxlynx> bbiaf
[13:37:39] <dawid> gnu automake 1.10.2
[13:37:54] <fuzzie> Huh.
[13:38:05] <dawid> i think the code at the beginning of the script more or less wants 1.7
[13:38:28] <dawid> it only executed that piese of code corresponding to this version
[13:38:55] <dawid> other than that it worked all like a charm :)
[13:39:30] <fuzzie> It looks like we should be requiring 1.10, but it should really work anyway. :/
[13:40:12] <dawid> ok, ill try ones more after pulling with a clean one
[13:40:20] <dawid> once -.-
[13:40:33] <fuzzie> I honestly have no idea about the hasted problem, though. Are you loading a savegame?
[13:41:54] <fuzzie> lynxlynxlynx: cut42a?
[13:42:05] <dawid> no i starded a new one
[13:51:42] <lynxlynxlynx> let me check
[13:52:13] <lynxlynxlynx> maybe, it is the last bcs shown, but there was plenty of other output later
[13:52:30] <lynxlynxlynx> this is after you enter slums and he teleports you to his home
[13:52:44] <fuzzie> Did he fail to teleport with you?
[13:52:52] <fuzzie> It should really work okay otherwise.
[13:52:58] <lynxlynxlynx> yep, i'm alone in there
[13:53:47] <fuzzie> The console should tell you if he tried the MoveBetweenAreas or not.
[13:54:01] <lynxlynxlynx> dawid: "automake (GNU automake) 1.10.2" or "gnu automake 1.10.2" ?
[13:54:13] <dawid> the first one
[13:54:36] <lynxlynxlynx> fuzzie: i guess that's the failing action override
[13:54:43] <fuzzie> I expect he didn't, because our FadeToColor/FadeFromColor are probably broken.
[13:54:56] <fuzzie> lynxlynxlynx: The failing action override is the 'start dialog', because he's not there.
[13:55:11] <lynxlynxlynx> ah
[13:55:36] <fuzzie> The script teleports everyone else away before he gets teleported, which probably stops his script from running.
[13:56:11] <lynxlynxlynx> dawid: i'll just remove / upgrade the WANT_X values then
[13:56:20] <lynxlynxlynx> they're old, but ere there specifically
[13:56:21] <fuzzie> That could be our actions waiting for too long, or it could be that his script is meant to keep running.
[13:56:32] <lynxlynxlynx> are there specifically for gentoo and mandriva
[13:56:55] <dawid> lynxlynxlynx: no clue really, just testing it line by line
[13:57:18] <lynxlynxlynx> dawid: stop and just change export WANT_AUTOMAKE="1.10" at the top of autogen.sh
[13:57:27] <dawid> lynxlynxlynx: ok
[13:58:06] <lynxlynxlynx> then retry the build of course
[13:58:11] <dawid> lynxlynxlynx: seems to go on
[13:58:27] <lynxlynxlynx> that translates to succes?
[13:59:03] <dawid> lynxlynxlynx: yes,building right now
[13:59:40] <lynxlynxlynx> ping me when it is done
[14:00:11] <dawid> lynxlynxlynx: o.O meaning?
[14:00:31] <lynxlynxlynx> tell me when the build finishes
[14:00:40] <dawid> lynxlynxlynx: ok xD
[14:01:47] <dawid> thatll take a while im gonna munch sth
[14:06:13] <dawid> lynxlynxlynx: finished
[14:06:20] <lynxlynxlynx> yeah, setting of a trap doesn't check any immunity
[14:06:27] <dawid> lynxlynxlynx: i mean, PING xD
[14:06:43] <lynxlynxlynx> i'll take that as 0
[14:07:37] <CIA-20> gemrb: 03lynxlupodian * r6613 10/gemrb/trunk/autogen.sh: autogen.sh: fixed the custom automake export to not use an ancient version
[14:09:11] <dawid> oh there one more thing that really hurts
[14:09:32] <dawid> the game scrambles my console
[14:09:49] <dawid> what i mean it got black with white text
[14:10:08] <lynxlynxlynx> and you use a different color scheme
[14:10:23] <lynxlynxlynx> we changed it to this so it also works on windows iirc
[14:10:24] <fuzzie> Someone should probably update the wiki installation page to give correct instructions and list the requirements.
[14:10:29] <dawid> yep i use the standard one for this terminal
[14:10:38] <fuzzie> A 'reset' might fix your console colours.
[14:10:41] <lynxlynxlynx> fuzzie: any new requirements?
[14:10:57] <dawid> they''ll be allright when the game quits
[14:11:34] <dawid> in the wiki this phrase bout configure was really misleading
[14:11:53] <fuzzie> lynxlynxlynx: It doesn't mention any of the autoconf/automake/etc.
[14:12:22] <fuzzie> And while it might not be necessary for the tarball, I think everyone who's come in here and asked in the last few months has been building svn.
[14:12:32] <lynxlynxlynx> SVN
[14:12:33] <lynxlynxlynx> cd gemrb/gemrb/trunk
[14:12:35] <lynxlynxlynx> Run ./autogen.sh2), it will also run ./configure
[14:12:36] <lynxlynxlynx> make
[14:12:41] <fuzzie> it should be 'cd gemrb', for a start.
[14:12:52] <lynxlynxlynx> you want specific versions?
[14:13:00] <lynxlynxlynx> oh yes, we changed the clone
[14:13:02] <fuzzie> Just mentioning that you need them at all would be nice :)
[14:13:15] <lynxlynxlynx> at first we instructed to just dl everything, so they'd get dltcep and iesh too
[14:13:40] <lynxlynxlynx> i'll add them, although it is pretty selfevident
[14:14:47] <fuzzie> Well, if you just try running autogen.sh then it'll complain about each missing thing individually and then quit, you end up running/installing several times.
[14:15:33] <lynxlynxlynx> two times
[14:16:28] <fuzzie> I think I ended up running for automake and libtool.
[14:16:57] <lynxlynxlynx> plus autoconf and you have everything
[14:18:38] <fuzzie> Can you remove the sf nomination thingy too?
[14:18:51] <lynxlynxlynx> nah, the vote has just begun
[14:19:13] <fuzzie> yes, but the nominations are long-gone
[14:19:36] <lynxlynxlynx> ah, wrong pic
[14:19:40] <fuzzie> so it's just linking to a "nominations are over" page
[14:20:39] <lynxlynxlynx> that says "vote here"
[14:20:41] <dawid> have sth new here on myscreen :/ during the beginning animation for a few secs the screen is becoming green
[14:20:47] <fuzzie> lynxlynxlynx: yes, but since you can't vote for us :)
[14:21:10] <lynxlynxlynx> we should move back to sf someday, now that it is wiki friendly
[14:21:33] <-- ltleitkie has left IRC (Ping timeout: 180 seconds)
[14:21:44] <fuzzie> It would be nice to have a more attractive front page with screenshots and things..
[14:21:55] <fuzzie> I don't know what would be necessary for that.
[14:22:41] <lynxlynxlynx> someone who cares ;)
[14:23:04] <fuzzie> well, I'm sure I could persuade someone to do an html design, but on a wiki might be more of a pain :)
[14:23:17] <lynxlynxlynx> everyone is laughing at web developers, but in the end, most of the projects i work with have a bad online presence
[14:23:30] <lynxlynxlynx> the wiki is themable
[14:24:11] <fuzzie> But presumably that's a lot harder than just a design, so it might be more difficult to find some help.
[14:24:22] <fuzzie> Just some embedded screenshots would be nice, though.
[14:25:36] <lynxlynxlynx> you mean for the background?
[14:25:58] <lynxlynxlynx> ad removed
[14:26:55] <fuzzie> I mean, some on the main page.
[14:27:22] <lynxlynxlynx> not a bad idea an easy to do
[14:27:29] <fuzzie> Maybe just adding some better ones to the gallery would help, too.
[14:27:30] <lynxlynxlynx> do we have anything sufficiently sexy?
[14:27:38] <lynxlynxlynx> the stuff on sf and hp is also bad
[14:27:54] <fuzzie> There must be some impressive-looking scene full of projectiles. :)
[14:28:40] <lynxlynxlynx> with a dragon
[14:28:49] <lynxlynxlynx> and zombies, people love zombies :)
[14:29:05] <lynxlynxlynx> i don't think there are any resources for ponies though
[14:30:16] <fuzzie> The current screenshots don't even have any weapons in the action bar, so meh.
[14:30:31] <fuzzie> Anyway, I am just wondering how to make everything more interesting to possible newcomers..
[14:30:54] <lynxlynxlynx> hey idling community! Get us some sexy screenshots
[14:31:24] <dawid> from the original heh? ;P
[14:31:25] <fuzzie> dawid: Some screenshots of the green might help, I'm not sure what you're seeing.
[14:32:10] <dawid> fuzzie: ill try to, about hasting npcs i dont see em now,the pseed is ok,but the clock is still irregularly ticking
[14:32:59] --- lynxlynxlynx has changed the topic to: GemRB 0.4.0 | Gekz, it's at http://gemrb.sf.net | Be wary of your words for there are Modron sensors in this channel: http://log.usecode.org/gemrblog.php | Hey you, we need some awesome screenshots!
[14:34:26] <fuzzie> dawid: How is the fps?
[14:34:48] <dawid> fuzzie: ultra low, even falling under 20
[14:35:25] <fuzzie> That's probably the problem; anything less than 28fps or so will cause timing problems.
[14:35:53] <fuzzie> If you're running at a high resolution you might try making it a bit smaller.
[14:35:59] <dawid> fuzzie: o.O not good i rarely get such fps xd
[14:36:15] <dawid> fuzzie: 800x600
[14:36:32] <fuzzie> But my poor 1ghz laptop can manage 30fps at 800x600, so as long as you compiled it optimised (the default), it really should be okay on anything modern.
[14:36:53] <dawid> -0s
[14:37:12] <fuzzie> I know that our fog-of-war code is kind of slow, though.
[14:37:13] <dawid> sry,shouldnt apply here -.-
[14:37:24] <lynxlynxlynx> ?
[14:37:40] <lynxlynxlynx> Os is like O2 with some size saving
[14:37:48] <pupnik> just try to keep performance in mind when implementing new things
[14:37:57] <fuzzie> The original engine is a bit cleverer about that, lagging the fog-of-war by a few frames or something.
[14:38:52] <pupnik> but leave FOW alone for now
[14:39:16] <lynxlynxlynx> as you command, sire
[14:39:26] <pupnik> that can be fixed by non-gemrb experts
[14:39:43] <dawid> i love its working fully polish from the beginning :)
[14:40:21] <dawid> 30+ fps timing again
[14:40:29] <lynxlynxlynx> polish or Polish?
[14:40:40] <dawid> the second one
[14:40:45] <lynxlynxlynx> yeah :/
[14:40:48] <fuzzie> If you have a slow terminal application then gemrb can get pretty slow when it's outputting a lot of text onto the console.
[14:41:12] <lynxlynxlynx> yeah, just redirect everything into the void
[14:41:26] <lynxlynxlynx> add &> /dev/null at the end
[14:41:29] <fuzzie> Someday we'll only make it output errors, but not yet. :)
[14:43:02] <lynxlynxlynx> the traps use ForceSpell which doesn't take magic resistance into account (only ApplyEffect does it), that's why my monk is always hurt
[14:44:03] <dawid> wasnt screenshot one of the last function keys?
[14:44:18] <lynxlynxlynx> we don't have that yet
[14:44:26] <lynxlynxlynx> oh, maybe i'm wrong
[14:44:55] <lynxlynxlynx> i just always use my native scrotter
[14:44:55] <fuzzie> lynxlynxlynx: The code seems to be applying effects when the casting is started. That's broken too, right?
[14:45:24] <lynxlynxlynx> it doesn't happen always, but it is wrong, yes
[14:45:52] <fuzzie> All of the spellcasting seems to call SpellCast which does AddAllEffects() right away, it seems that that'd always happen.
[14:46:20] <fuzzie> (Then it treats CastingTime as in seconds, which is why casting times are too long, I guess.)
[14:46:39] <lynxlynxlynx> maybe the animations confuse me
[14:46:58] <lynxlynxlynx> some of the aoe spells take long to draw
[14:47:23] <fuzzie> Anyway, yet more stuff I don't want to mess with before the release. :/
[14:48:29] <lynxlynxlynx> maybe the casting time is an easy win - shouldn't it just be divided?
[14:48:55] <fuzzie> It shouldn't be done with SetWait() at all.
[14:50:46] <fuzzie> But .. ROUND_SIZE*header->CastingTime/(GetStat(IE_MENTALSPEED)*10).
[14:51:04] <fuzzie> except we don't initialise IE_MENTALSPEED, because that would be helpful :)
[14:52:15] <lynxlynxlynx> ;)
[14:52:49] <fuzzie> Anyway, I probably misunderstand castingtime so I wouldn't commit that.
[14:53:32] <fuzzie> And actually, IE_MENTALSPEED is probably wrong.
[14:53:56] <lynxlynxlynx> it is the correct stat
[14:54:12] <fuzzie> Because it's used by Improved Alacrity and that description only talks about eliminating the pause, not reducing the casting time.
[14:54:13] <lynxlynxlynx> i checked what the robe of vecna applies
[14:55:07] <fuzzie> But I don't have DLTCEP/NI here so you can probably check this better than I can.
[14:56:05] <fuzzie> Oh, no, Improved Alacrity is fiddling with the auracleansing stat.
[14:56:11] <fuzzie> Never mind, then, IE_MENTALSPEED seems ok.
[14:56:54] <lynxlynxlynx> the robe sets the stat to 4 (says iesdp and gemrb), while the description says it decreases the spellcasting time by 4
[14:57:10] <lynxlynxlynx> so i guess it is used as a modifier when checking the casting time
[14:57:12] <fuzzie> By a multiple of 4?
[14:57:15] <fuzzie> Or just a modifier?
[14:57:37] <lynxlynxlynx> Alters the casting time of spells by the value specified by the 'Speed Modifier' field. This effect sets stat #77 to param #1
[14:58:03] <lynxlynxlynx> gemrb does the same, I think: STAT_MOD( IE_MENTALSPEED );
[14:58:10] <fuzzie> That seems not very conclusive. :)
[14:59:10] <fuzzie> If it's just subtracting from casting time then I would guess the minimum is still 1?
[14:59:22] <fuzzie> I haven't looked at this in the original game recently enough to remember.
[15:00:52] <lynxlynxlynx> probably
[15:05:24] <fuzzie> http://fuzzie.org/nfs/gemrb/untested_casttime_hack.txt <- something like this?
[15:05:51] <fuzzie> oh, no, that's broken of course.
[15:06:52] <fuzzie> Updated to something which makes a little more sense.
[15:07:08] <fuzzie> But I can't test it at the moment myself.
[15:09:45] <lynxlynxlynx> trying it out
[15:12:18] <lynxlynxlynx> still not perfect, but it seems to look better
[15:14:43] <fuzzie> If it's still bad then it's probably not worth the commit.
[15:21:40] <lynxlynxlynx> unconclusive
[15:21:53] <lynxlynxlynx> i think it is a bit faster now, since you can see some animation lag
[15:26:11] --> ltleitkie has joined #gemrb
[15:39:58] <-- ltleitkie has left IRC (Ping timeout: 180 seconds)
[15:41:23] <CIA-20> gemrb: 03lynxlupodian * r6614 10/gemrb/trunk/gemrb/override/bg2/avatars.2da: MGHL also has a IE_ANI_FOUR_FILES_2 animation type
[16:03:57] --> tombhadAC has joined #gemrb
[16:47:48] --> Avenger has joined #gemrb
[16:47:51] --- ChanServ gives channel operator status to Avenger
[16:47:54] <Avenger> hello
[16:50:41] <lynxlynxlynx> oj
[16:52:01] <lynxlynxlynx> i found out damage types *are* a bitfield
[16:52:15] <lynxlynxlynx> the fireball does fire/missile damage
[16:53:20] <lynxlynxlynx> or maybe it was the arrow of flame
[16:53:34] <Avenger> they never mix damagetypes
[16:53:42] <lynxlynxlynx> retesting, i can't hurt anyone with fireballs
[16:53:51] <Avenger> huh?
[16:53:58] <Avenger> the projectile isn't working?
[16:54:09] <lynxlynxlynx> it is shown but it does not damage
[16:54:18] <Avenger> then it isn't working
[16:54:19] <Avenger> meh
[16:54:27] <Avenger> i thought area projectiles work
[16:54:39] <Avenger> last time i tested, web and cloud worked
[16:54:49] <lynxlynxlynx> web and cloud
[16:56:07] <Avenger> btw i saw some chat about improved alacricity
[16:56:22] <Avenger> there is a stat: auracleansing
[16:56:29] <lynxlynxlynx> yeah
[16:56:39] <lynxlynxlynx> it was a dead end, we were talking about casting times
[16:57:03] <lynxlynxlynx> which you probably read about too
[16:57:42] <Avenger> ok, i just go and create some more projectiles
[17:06:52] <lynxlynxlynx> [CharAnimations]: Couldn't create animationfactory: usar1a2e <-- missing usar1a2e.bam for one of the final three in ch1 (sarevok?)
[17:08:29] <Avenger> yes that's sarevok
[17:08:52] <lynxlynxlynx> can something be done about it?
[17:09:28] <Avenger> there is no usar1a2
[17:09:47] <Avenger> a2 is probably one handed attack form
[17:09:53] <Avenger> sarevok has only 2 handed weapons
[17:09:56] <Avenger> i guess
[17:10:36] <Avenger> or the other way around
[17:10:54] <Avenger> it looks like these are one handed attacks
[17:11:40] <lynxlynxlynx> so maybe we're screwing up somewhere and he doesn't equip his weapon
[17:11:54] <Avenger> yes, i noticed that elsewhere too
[17:12:13] <lynxlynxlynx> sometimes actors just disappear
[17:12:28] <Avenger> isn't it because their animation couldn't be rendered?
[17:12:58] <lynxlynxlynx> don't know
[17:13:14] <lynxlynxlynx> sarevok is the only big spammer
[17:14:53] <lynxlynxlynx> his sword is twohanded
[17:15:48] <lynxlynxlynx> the animation is set to 2h too
[17:17:24] <lynxlynxlynx> can you tell from dltcep if the item is equipped or not?
[17:25:18] --> |Cable| has joined #gemrb
[17:44:49] <lynxlynxlynx> bah, now a crash
[17:49:49] <Avenger> yes
[17:50:29] <Avenger> in the items&spells tab, 'selected slot'
[17:50:44] <Avenger> the actual slot is selected slot+first weapon slot
[17:50:54] <Avenger> so, quivers are negative, weapon slots are 0-3
[17:50:58] <Avenger> fist is 1000
[17:51:20] <lynxlynxlynx> all the items show weapons slot 0 there
[17:51:29] <Avenger> that means, first weapon
[17:51:40] <Avenger> that's quite normal :)
[17:51:50] <fuzzie> hi
[17:51:55] <Avenger> hello fuzzie
[17:52:06] <fuzzie> sorry for confusion, Avenger, i found the auracleansing after a bit of looking :)
[17:53:39] <fuzzie> do you know what the stat means?
[17:53:43] <fuzzie> is it just zero/non-zero?
[17:55:08] <-- dawid has left IRC (Read error: 110 (Connection timed out))
[17:56:39] --> dawid has joined #GemRb
[17:56:48] <fuzzie> agh, ActionOverride is kind of horrid
[17:57:48] <Avenger> auracleansing is a boolean stat
[17:58:05] <Avenger> if it is 0, then there must be some delay between spells (one round)
[17:58:21] <Avenger> if it is 1, then you can start casting as soon as the previous spell was cast
[17:58:25] <fuzzie> yes, that is fine
[17:58:32] <fuzzie> don't suppose you know spell timings too? :)
[17:58:39] <CIA-20> gemrb: 03lynxlupodian * r6615 10/gemrb/trunk/gemrb/plugins/Core/CharAnimations.cpp: added sleep animation handling for ogre mages
[17:58:54] <fuzzie> i wrote a quick hack using IE_MENTALSPEED and CastingTime earlier, but it seems not perfect
[17:59:28] <Avenger> sometimes i think it is 7 ai updates, sometimes 15
[17:59:50] <fuzzie> It's probably 9.
[17:59:59] <Avenger> probably depends on the ai update frequency, or gametypes too
[18:00:24] <Avenger> how you came up with 9?
[18:00:33] <fuzzie> ROUND_TIME/10
[18:00:37] <fuzzie> erm, ROUND_SIZE
[18:00:53] <fuzzie> that is what seems suggested on the web, but i didn't have time to check it properly
[18:01:42] <Avenger> watch this piece:
[18:01:45] <Avenger> 00515AB9 8B 88 D0 1D 00 00 mov ecx,dword ptr [eax+1DD0h] ;;gametime ticks
[18:01:47] <Avenger> 00515ABF 89 4D D4 mov dword ptr [ebp-2Ch],ecx
[18:01:48] <Avenger> 00515AC2 8B 55 BC mov edx,dword ptr [ebp-44h]
[18:01:50] <Avenger> 00515AC5 8B 42 24 mov eax,dword ptr [edx+24h]
[18:01:51] <Avenger> 00515AC8 2B 45 D4 sub eax,dword ptr [ebp-2Ch]
[18:01:53] <Avenger> 00515ACB 33 D2 xor edx,edx
[18:01:54] <Avenger> 00515ACD B9 0F 00 00 00 mov ecx,0Fh
[18:01:56] <Avenger> 00515AD2 F7 F1 div eax,ecx
[18:01:58] <Avenger> there is a 15 :)
[18:02:20] <Avenger> edx+24 is the duration in the .spl file
[18:02:29] <Avenger> 1dd0 is the ticks counter in cgameobject
[18:03:35] <fuzzie> yes, that kind of code is all over the place
[18:04:00] <Avenger> yes, it is to calculate the finish time of a spell given a duration
[18:04:08] <Avenger> or something like that
[18:04:46] <Avenger> in iwd2 i see 7's
[18:04:48] <fuzzie> but what's -44h, is the question
[18:05:01] <fuzzie> oh, it's zeroed out there anyway
[18:05:53] <fuzzie> isn't that dividing the spell duration by the AI ticks? if so, that doesn't make much sense unless it already modified the spell duration at load.
[18:07:12] <Avenger> bp-whatever are local variables
[18:07:25] <Avenger> bp+whatever are parameters
[18:07:35] <fuzzie> yes, i'm used to reading this in IDA Pro because i am lazy :)
[18:07:38] <Avenger> cx+whatever are struct elements
[18:07:58] <Avenger> oh well, i should try it one day
[18:08:08] <Avenger> it would probably improve my productivity
[18:08:29] <fuzzie> the free version on their website is very nice
[18:08:51] <Avenger> yes fuzzie, it modifies the spell duration at 'first apply' somewhere before that
[18:08:59] <fuzzie> so that is the magic bit :(
[18:09:18] <Avenger> it always calculates the absolute end of the spell from duration
[18:09:22] <Avenger> you can see that in savegames
[18:09:32] <Avenger> the duration is replaced on first apply
[18:09:40] <Avenger> we do that too, btw :D
[18:09:47] <fuzzie> where is it done?
[18:09:51] <Avenger> in our code?
[18:09:56] <fuzzie> yes
[18:10:27] <Avenger> PrepareDuration
[18:10:29] <fuzzie> right now the spell stuff is some miserable broken hack
[18:10:30] <Avenger> in effectqueue
[18:10:34] <fuzzie> that is just for effects though :)
[18:11:17] <Avenger> don't break more than you fix, though
[18:11:30] <fuzzie> i'm not going to change spellcasting for no
[18:11:31] <fuzzie> w
[18:11:36] <Avenger> phew :P
[18:11:43] <fuzzie> i don't understand it, as you can see
[18:11:58] <fuzzie> actions are much easier :)
[18:12:13] <Avenger> there are some bad parts, this isn't easy stuff
[18:12:27] <Avenger> actions are not easy either
[18:12:33] <fuzzie> right now the spell casting applies the effects on cast start
[18:12:47] <Avenger> no
[18:12:49] <fuzzie> and i think lynx said that it is not even checking resistances
[18:12:56] <Avenger> casting applies only the cfb
[18:12:59] <Avenger> which it should
[18:13:11] <Avenger> the real effects simply go to the projectile
[18:13:20] <Avenger> the projectile effects apply on impact
[18:13:27] <fuzzie> strange :) but i didn't check any of it
[18:13:37] <Avenger> well, don't kill the cfb :D
[18:13:49] <Avenger> it is the 'equipping effects' of spells
[18:14:06] <fuzzie> but we should be keeping the spell around and just adjusting the duration inside it, you think?
[18:14:16] <Avenger> ?
[18:14:19] <fuzzie> at the moment we FreeSpell() immediately after the cfb
[18:14:28] <fuzzie> and then create it again when we finish casting
[18:14:39] <fuzzie> i was just going to keep that code
[18:14:44] <Avenger> well it is cached
[18:14:50] <Avenger> it is the same as we 'create' items
[18:14:56] <Avenger> we don't really create it
[18:14:57] <fuzzie> and then have the action do the casting time
[18:15:03] <Avenger> just borrow the cache pointer
[18:15:21] <fuzzie> but you say that we should keep the duration in the spell? or you mean just the effects once they're applied?
[18:15:31] <Avenger> no i never said that
[18:15:42] <fuzzie> the effect durations seem specified in AI updates (15), definitely
[18:15:44] <Avenger> part 1: each effect may have a different duration
[18:16:01] <fuzzie> but the casting time is the time before the projectile is even created
[18:16:02] <Avenger> so it is not spell based, it is separate for each effect
[18:16:07] <Avenger> yes
[18:16:14] <fuzzie> and that is what i am interested in :)
[18:16:14] <Avenger> that's how long you wiggle your hand
[18:17:27] <fuzzie> but you don't know anything about that particularly?
[18:17:43] <fuzzie> that is all i want to change, i think, i don't think the effect timing needs any more changes.
[18:18:06] <Avenger> castspellend is the end of the casting
[18:18:20] <Avenger> that's where the casting time wound down to 0
[18:18:25] <fuzzie> http://fuzzie.org/nfs/gemrb/untested_casttime_hack.txt
[18:18:28] <fuzzie> ^- this is what i am talking about :)
[18:18:47] <fuzzie> well, SetWait is wrong, it should not be doing SetWait, but you get the idea
[18:18:58] <Avenger> yes
[18:19:05] <fuzzie> i don't mean to change anything else! :)
[18:19:07] <Avenger> and after setwait you call castspellend
[18:19:21] <Avenger> i mean, after the wait time was over
[18:19:23] <fuzzie> yes, that works already, it is just very buggy
[18:20:04] <Avenger> ok, but how casting time is done now
[18:20:16] <Avenger> i notice casting time, it is quite long
[18:20:26] <fuzzie> yes, that is the problem
[18:20:28] <Avenger> there is no wait, and it is still there?
[18:20:41] <Avenger> i doubt that
[18:20:43] <fuzzie> there is a SetWait right now, you can see it at the top there
[18:20:53] <fuzzie> but it does *AI_UPDATE_TIME and that is too long :)
[18:21:38] <Avenger> well, try it without it
[18:21:40] <fuzzie> but i don't think the new code there is quite right either
[18:22:13] <fuzzie> anyway, it is something to change tomorrow..
[18:22:39] <Avenger> ahh it is in spellcast
[18:23:09] <Avenger> why would you want to remove it from spellcast and move it to the action?
[18:23:21] <fuzzie> because SetWait is not interruptable
[18:23:26] <fuzzie> and spells are interruptable
[18:24:04] <fuzzie> and some actions (the Force ones) cast spells instantly, so they must not set a wait
[18:24:30] <Avenger> ise
[18:24:38] <Avenger> well
[18:24:57] <fuzzie> but i didn't test it enough in the original engine yet.
[18:25:55] <fuzzie> anyway it is really really easy to fix once i am sure how it should work :)
[18:26:02] <fuzzie> i just wanted to know if you knew how it should work already
[18:29:17] <fuzzie> all the actions are kind of complicated because often things change in the middle of their execution, i don't know how much that applies to spellcasting yet
[18:29:25] <Avenger> i know, but of course, it was a very first implementation :) reallyforcespell was added on its top
[18:29:51] <Avenger> spellcasting is very complicated
[18:30:12] <Avenger> your target can vanish, or leave the area
[18:30:23] <Avenger> you yourself can die during casting
[18:30:38] <Avenger> moving near target is required too
[18:30:39] <fuzzie> for most actions, you target can change in the middle of the action
[18:31:17] <Avenger> well, for open door, i don't expect it to be that tough :)
[18:31:33] <fuzzie> hehe, yes, those ones are easy :)
[18:31:59] <fuzzie> but i spent a few days checking everything and talking to devSin, i just didn't get to spellcasting yet
[18:32:02] <Avenger> there is also the immunity to creature type opcode
[18:32:56] <fuzzie> that is not all handled in effects?
[18:32:57] <Avenger> if an actors would be target is affected by it, the actor cannot attack (target) the target.
[18:33:03] <fuzzie> oh, interesting :)
[18:33:12] <fuzzie> i'll have to add that to the list
[18:33:19] <Avenger> no, this should be handled in the scripts targeting, but i don't think it is done yet
[18:33:27] <lynxlynxlynx> protection from undead huh
[18:33:32] <Avenger> i don't know if the target should be completely invisible or not
[18:33:43] <Avenger> if it is a complete invisibility by script, then it is easier
[18:34:17] <Avenger> you simply check if the target is affected by 'immunetocreature', caster->Race
[18:34:41] <fuzzie> ok. that is added by spells and things?
[18:34:41] <lynxlynxlynx> 35: chevil10 - (0 0 0) 68e8 Wt: 1 x 9Lb
[18:34:43] <lynxlynxlynx> Equipped: 0
[18:34:53] <lynxlynxlynx> so sarevok is spawned with the sword in the inventory
[18:35:03] <lynxlynxlynx> 35 SLOT_INV13
[18:35:13] <Avenger> equipped:0 means first weapon slot
[18:35:25] <Avenger> fuzzie: yes
[18:35:38] <Avenger> fuzzie: there is an effect queue call to check on that, i think.
[18:35:53] <fuzzie> so it is not quite so easy, i'll have to check what happens when a marked object gains the effect in the middle of an action
[18:36:01] <lynxlynxlynx> yes, but if the prefixes in the inventory part are the slots, he has only the fist weapon
[18:37:17] <lynxlynxlynx> http://pastebin.ca/1472973
[18:37:30] <lynxlynxlynx> barehanded
[18:43:02] <fuzzie> ok, i must go for a while. good luck with projectiles :)
[19:05:23] --> barra_away has joined #gemrb
[19:06:06] <pupnik> wow
[19:21:31] <-- barra_library has left IRC (Read error: 110 (Connection timed out))
[19:23:22] --- barra_away is now known as barra_library
[19:45:06] <lynxlynxlynx> the area just spawns them with CreateCreature
[19:47:42] <lynxlynxlynx> you don't notice that sarevok doesn't have the sword equipped since a sword is baked into his animations
[19:47:46] <fuzzie> ok, hi again
[19:48:25] <lynxlynxlynx> o hi
[19:49:45] <lynxlynxlynx> nothing in the official patch notes or the fixpack
[19:50:18] <fuzzie> you tried saving there in the original engine and looking at the actor?
[19:50:52] <lynxlynxlynx> i can't run tob :/
[19:51:26] <fuzzie> which bit in particular should i look at?
[19:51:53] <lynxlynxlynx> what it has equipped
[19:52:03] <lynxlynxlynx> but you won't be able to save in the original
[19:52:24] <lynxlynxlynx> this is the sarevok apparition, not the npc
[19:53:24] <lynxlynxlynx> so i don't think you can check any more than i do
[19:59:52] <fuzzie> oh?
[20:00:01] <fuzzie> i mean, i can just sabotage the script after the CreateCreature
[20:00:35] <lynxlynxlynx> and change his ea and kill everyone else and I still am not sure if you would be able to save
[20:00:54] <fuzzie> there are ways to trick the engine into saving anyway
[20:00:56] <lynxlynxlynx> but you can try it, of course
[20:01:07] <lynxlynxlynx> really? good
[20:03:18] <fuzzie> hm, there are some really inaccurate things in the g3 forums :)
[20:05:35] <fuzzie> one post talks about ActionOverride() blocking on the action, but it is not true
[20:07:37] <fuzzie> i was worried, that would be painful to do
[20:10:35] <fuzzie> it does wipe the action queue, but i don't know how to do that properly yet
[20:13:07] --> Edheldil has joined #gemrb
[20:13:07] --- ChanServ gives channel operator status to Edheldil
[20:14:22] <fuzzie> lynxlynxlynx: the cutscene bug is in our Fade actions, btw
[20:14:36] <lynxlynxlynx> gealan?
[20:14:39] <fuzzie> yes
[20:15:06] <fuzzie> our FadeToColor and FadeFromColor actions block, and they shouldn't
[20:15:47] <fuzzie> you can try commenting out the SetWait calls in FadeToColor and FadeFromColor in Actions.cpp, if you have the time, and see if the cutscene works then
[20:16:11] <lynxlynxlynx> sure, I can try
[20:18:02] <fuzzie> there are some other bugs there, so it's something to fix after the release, but i hope that's the bug breaking the cutscene
[20:19:35] <lynxlynxlynx> yes, it works now
[20:19:46] <lynxlynxlynx> i've commented out all three SetWaits
[20:20:33] <lynxlynxlynx> the viewport didn't center on the scene in slums though, but this is most likely unrelated (i can check if this was ok before)
[20:22:19] <lynxlynxlynx> the next cutscene was perfect :)
[20:22:29] <fuzzie> :)
[20:22:35] <lynxlynxlynx> of course it was just ending when i said that and there is a problem at the very end
[20:22:42] <fuzzie> hehe, what's it?
[20:22:50] <lynxlynxlynx> i'm now in the building where it took place
[20:23:05] <fuzzie> i'll try this all myself of course, but i already have a list of bugs. do you know the .bcs?
[20:23:16] <lynxlynxlynx> movie01c.bcs
[20:23:43] <fuzzie> and the one before?
[20:24:10] <lynxlynxlynx> b
[20:25:01] <fuzzie> a calls StorePartyLocations() and c calls RestorePartyLocations()
[20:26:02] <lynxlynxlynx> one of which doesn't work
[20:26:07] <fuzzie> those are implemented in gemrb, but they store the location in a local variable and without the area name
[20:26:30] <fuzzie> i don't know why gemrb's implementation is like it is, because i don't see those strings anywhere in the exe
[20:27:17] <lynxlynxlynx> do you want me to commit the Fade comment-outs or do you have other plans?
[20:27:19] <fuzzie> yes, it is meant to be stored in the GAM, not in a local variable
[20:27:30] <fuzzie> go ahead and commit them for now
[20:27:48] <fuzzie> i'll fix it properly after the release
[20:28:58] <CIA-20> gemrb: 03lynxlupodian * r6616 10/gemrb/trunk/gemrb/plugins/Core/Actions.cpp:
[20:28:58] <CIA-20> gemrb: hack: GameScript::Fade*Color should not be blocking
[20:28:58] <CIA-20> gemrb: enabled you to reach chapter2 in soa
[20:30:13] <fuzzie> ok, gemrb doesn't implement saved locations yet. it is easy though.
[20:31:04] <Edheldil> lynxlynxlynx, I saw that the wiki debate did not develop well :/
[20:31:08] <lynxlynxlynx> don't feel pressured to do this before this release, I'm sure you have plenty of stuff on your plate already
[20:31:30] <lynxlynxlynx> Edheldil: it's not that bad; could've been much worse
[20:32:20] <fuzzie> I just want to fix that var.var thing, then I am done before the release.
[20:34:09] <Edheldil> of course it could, and I had suspected it would. But I really miss the 'diff to some older version' feature :(
[20:34:35] <lynxlynxlynx> yep
[20:34:50] <lynxlynxlynx> i miss transparency
[20:35:16] <Edheldil> like in "no SPOF" sense?
[20:35:27] <lynxlynxlynx> spoof?
[20:35:35] <fuzzie> Well, I think I'm just going to write my own actions reference.
[20:35:39] <Edheldil> spof=single point of failure
[20:35:51] <lynxlynxlynx> kinda, i guess
[20:36:08] <fuzzie> The IESDP one isn't useful enough for me and Igi doesn't want it to *be* useful enough, which is fine.
[20:36:13] <Edheldil> i.e. less of (benevolent) dictatorship and more of community
[20:36:24] <lynxlynxlynx> yes
[20:36:42] <lynxlynxlynx> fuzzie: bloat comes just with bad design and we already have a likeable design
[20:37:05] <lynxlynxlynx> it's all just about separating any elaborate info into subpages or categories
[20:37:17] <fuzzie> Well, I want things like "this is definitely confirmed as working in all ways in all engine versions".
[20:37:31] <fuzzie> Not just "we copied this from the BG2 actions and gosh we hope it's correct".
[20:37:39] <Edheldil> hehe
[20:37:48] <fuzzie> And this doesn't work very well with the current layout.
[20:37:52] <lynxlynxlynx> some kind of tagging would be simple to do
[20:37:53] <Edheldil> that could be even useful to the modders
[20:37:58] <fuzzie> But please don't bring this up in the thread :)
[20:38:11] <lynxlynxlynx> keeping the current layout is a must at the start
[20:38:22] <Edheldil> sure. Anything more and it will catch fire
[20:38:23] <fuzzie> I'll just quietly write something new for my own use.
[20:38:45] <lynxlynxlynx> the cool thing about wiki content is that you can write as many interfaces as you want
[20:39:00] <fuzzie> I'm not sure anyone would, though :)
[20:42:03] <lynxlynxlynx> Avenger: i see the fireball from the spell uses an already present projectile (pfire2) and sets a global var in the cfb (mgArea1)
[20:42:18] <Avenger> so?
[20:42:41] <lynxlynxlynx> so it is not a problem with an unhardcoded projectile
[20:42:46] <Avenger> fireball is not hardcoded, yes
[20:43:01] <lynxlynxlynx> what's the global var for?
[20:43:12] <Avenger> i think they wanted it to signal an incoming fireball
[20:43:20] <Avenger> so monster AI can scatter and run :)
[20:43:22] <lynxlynxlynx> it is reset after it hits
[20:43:27] <lynxlynxlynx> ooh, nice
[20:43:30] <Avenger> yes, of course
[20:43:44] <Avenger> i don't know if they use it anywhere
[20:45:45] <fuzzie> yes
[20:45:49] <fuzzie> see areawarn.bcs
[20:46:02] <fuzzie> i think that is used
[20:46:24] <fuzzie> it doesn't do anything useful, mind :)
[20:47:31] <fuzzie> it's also used in mage14t, maybe someone can look using ie_sh
[20:48:01] <lynxlynxlynx> it's not really important
[20:48:15] <fuzzie> yes, just for curiousity :)
[20:51:11] <lynxlynxlynx> casting the fireball doesn't even call fx_damage, no wonder nothing happens
[20:55:29] <fuzzie> Avenger: do you work on anything?
[20:55:39] <Avenger> projectiles
[20:55:49] <fuzzie> ok :)
[21:02:52] <fuzzie> Avenger: you added StorePartyLocations in r2174 but you stored them in a local variable instead of the GAM, do you think that was just a guess?
[21:03:28] <Avenger> no, party locations are stored in the gam in bg2
[21:03:44] <fuzzie> ok, so i can fix it :) thanks
[21:04:22] <Avenger> i think some engines store them in variables, though
[21:04:54] <fuzzie> well, the action is only in the bg2 list on iesdp, and the variable used in gemrb is not in any of the exe files
[21:05:07] <Avenger> there are some features which were first developed by blackisle. The bioware guys didn't take the code, but reimplemented the feature in their own way
[21:05:56] <Avenger> that's the main reason for these weirdly similar features
[21:06:06] <fuzzie> yes, this is slightly annoying when gemrb handles everything by name :)
[21:07:38] <Avenger> about fireballs, if you cannot live without them, look for a call for SecondaryProjectile() in Projectile.cpp
[21:07:46] <Avenger> remove the condition from around it
[21:08:16] <Avenger> err SecondaryTarget
[21:09:14] <lynxlynxlynx> sounds like a hack
[21:09:29] <lynxlynxlynx> interestingly the delayed blast fireball works fine
[21:09:52] <Avenger> because it has secondary projectile
[21:10:03] <Avenger> it is not a hack
[21:10:36] <lynxlynxlynx> so the two if tests should be swapped? now the PAF_SECONDARY is commented out
[21:10:38] <Avenger> the default projectile is 0 :)
[21:10:53] <Avenger> none of those conditions are needed
[21:11:04] <lynxlynxlynx> so i can just remove everything, ok
[21:11:24] <Avenger> yes
[21:11:36] <Avenger> though PAF_SECONDARY does something, i don't know what is it :)
[21:11:55] <Avenger> there are about 5-6 flags i still don't know
[21:12:10] <Avenger> oddly enough, even bg1 has some of those
[21:12:21] <lynxlynxlynx> it works :D
[21:12:53] <Avenger> release?
[21:13:04] <lynxlynxlynx> tommorow
[21:13:05] <Avenger> tomorrow?
[21:13:07] <Avenger> ok
[21:13:57] <CIA-20> gemrb: 03lynxlupodian * r6617 10/gemrb/trunk/gemrb/plugins/Core/Projectile.cpp: make fireballs hurt, curtesy of Avenger
[21:14:35] <Avenger> i still don't know where are those damn fireball flames in bg1
[21:16:57] <lynxlynxlynx> i don't have it installed to check
[21:33:08] <fuzzie> Avenger: does http://fuzzie.org/nfs/gemrb/read_initial_variables.txt seem ok?
[21:33:28] <fuzzie> it is not great, but it works for now, and i'd like it in the release because otherwise PS:T is quite broken
[21:34:30] <Avenger> yes, you can remove those commented out lines from map too, //map->vars=new Variables(); and the next one
[21:34:57] <fuzzie> ok
[21:35:38] <Avenger> it was a good catch you found these default values :)
[21:36:01] <fuzzie> It is only thanks to DLTCEP's variable list that I noticed that var.var existed at all :)
[21:36:12] <Avenger> hehe
[21:36:39] <Avenger> at one point i must knew it has values, but it wasn't even working correctly in dltcep ;)
[21:37:04] <Avenger> pst wasn't my favourite anyway, especially this feature
[21:37:21] <fuzzie> PS:T is my favourite game and my least favourite engine :)
[21:42:51] <CIA-20> gemrb: 03fuzzie * r6618 10/gemrb/trunk/gemrb/plugins/ (4 files in 3 dirs): add Variables::LoadInitialValues hack to read initial values from pst's var.var file
[21:43:09] <-- Avenger has left IRC ("ChatZilla 0.9.85 [Firefox 3.0.11/2009060215]")
[21:54:10] <CIA-20> gemrb: 03avenger_teambg * r6619 10/gemrb/trunk/gemrb/override/ (16 files in 2 dirs): more projectiles
[21:56:34] <CIA-20> gemrb: 03avenger_teambg * r6620 10/gemrb/trunk/gemrb/plugins/Core/Projectile.cpp: fix 0 explosion count
[22:10:34] <-- lynxlynxlynx has left IRC (Remote closed the connection)
[22:15:29] <Edheldil> good night
[22:15:31] <-- Edheldil has left IRC ("Really?")
[22:21:34] <fuzzie> i'm getting a python error while saving a bg2 game: GUISAVE's ConfirmedSaveGame calls StartLoadScreen does Middle.SetMOS (LoadPic) which fails due to 'scene01.mos' not existing
[22:25:30] <dawid> U lucky :D i dont get errors
[22:26:04] <fuzzie> I guess because LoadMos contains the TextScreen details at that point....
[22:28:27] <fuzzie> hehe, when I reload the save, the dead actors from the dungeon cutscene come back to life!
[22:28:43] <fuzzie> I guess that is a bug with cutscenes never executing death scripts.
[22:29:32] <dawid> are U able to talk to minsc? in the cage i mean
[22:29:40] <fuzzie> yes, but you have to move close enough first
[22:29:51] <fuzzie> stupid bug :(
[22:30:01] <dawid> standing at the cage should suffice?
[22:30:03] <fuzzie> we don't try and find paths through doors
[22:30:05] <fuzzie> yes
[22:30:18] <dawid> ill try ince more to get closer then
[22:31:55] <fuzzie> you should be able to get all the way through the dungeon, but battles are very buggy right now, so it's too annoying :)
[22:37:48] <dawid> yep,was standing even further away from him,but this time i managed to talk to him
[22:39:27] <fuzzie> http://fuzzie.org/nfs/gemrb/dungeon.html is my complaints from when I last tried it.
[22:40:41] <fuzzie> I have a fix for the attack stuff, but it's waiting for after the 0.5 release.
[22:42:36] <dawid> SPWI311 produces a "Color:SetRGB (-1023376384, 5)" for me. Endian bug? U have big endian machine? o.O
[22:42:46] <fuzzie> yes
[22:43:01] <fuzzie> it isn't an endian bug, though, it happens on little-endian too :)
[22:46:46] <dawid> i noticed w/o fog of war the game is unplayable at all, but thats prolly not so important
[22:47:40] <dawid> oh bout that green things on my screen,i wont make shots of em, after i changed resolution the disappered,so i guess its my local thingy
[22:49:23] <fuzzie> Heh, I guess you really can't get past the start of Chapter 2 in gemrb. :(
[22:51:36] <-- xrogaan has left IRC ("Why ?")
[22:53:28] <dawid> didnt get that far yet
[23:51:19] --> barra_desktop has joined #gemrb