#gemrb@irc.freenode.net logs for 21 Aug 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:03:11] <-- tomprince_loki has left #GemRb
[00:46:56] --> raevol has joined #GemRb
[01:10:57] <pupnik> time to lie down with some Econtalk.org lectures
[03:08:31] --> |Cable| has joined #GemRb
[03:09:57] <-- pupnik has left IRC (Read error: Connection reset by peer)
[03:10:18] --> pupnik has joined #GemRb
[07:02:40] --> lynxlynxlynx has joined #GemRb
[07:02:40] --- ChanServ gives channel operator status to lynxlynxlynx
[07:21:32] --> avenger has joined #GemRb
[07:21:32] --- ChanServ gives channel operator status to avenger
[07:21:52] <avenger> hi
[07:40:17] <-- raevol has left IRC (Quit: Leaving.)
[07:50:31] <pupnik> nhuhu
[08:26:08] <lynxlynxlynx> http://lynxlynx.info/ie/GemRB-v0.6.2.zip <-- here's tom's binary
[08:26:24] <lynxlynxlynx> we need volunteers for testing
[08:29:35] <fuzzie> the Windows one?
[08:30:08] <lynxlynxlynx> yes
[08:35:16] <lynxlynxlynx> heh, the current summon disable timing is too long
[08:36:33] <lynxlynxlynx> but back to the release
[08:39:54] <pupnik> how about adding a gemrb.cfg and echo config instructions after the 'make install'
[08:40:43] <fuzzie> you shouldn't need any of that
[08:40:47] <fuzzie> i mean, assuming you're using cmake
[08:40:58] <pupnik> used cmake, make, make install
[08:40:58] <pupnik> [Config]: Trying to open /usr/local/etc/gemrb/gemrb.cfg [NOT FOUND]
[08:41:17] <pupnik> just trying to emulate a first-time build here
[08:42:12] <pupnik> however i guess distro packagers should provide those
[08:43:08] <lynxlynxlynx> http://pastebin.ca/1921907 <-- comments on the release announcement?
[08:43:47] <lynxlynxlynx> we could echo the line to run the minimal dataset
[08:43:50] <fuzzie> more concerned about this Windows example build
[08:43:53] <fuzzie> which has no working example config
[08:44:46] <fuzzie> and wants about 2gb of RAM to find that out
[08:44:53] <lynxlynxlynx> huh
[08:45:21] <fuzzie> someone did report that their mingw build was doing that.
[08:47:31] <fuzzie> the openal32.dll shipped doesn't work
[08:48:29] <fuzzie> because it's the openal router, not openal.
[08:48:51] <fuzzie> there is a soft_oal.dll i can just copy over it, though.
[08:50:02] <fuzzie> but i wouldn't ship this build while it has the RAM thing.
[08:51:46] <fuzzie> if i replace openal32.dll and fix the config to point at the example data then it works, though.
[08:51:54] <fuzzie> python complains about being able to 'import site' but continues onward.
[08:52:16] <fuzzie> but then it does this RAM-eating thing when it quits.
[08:52:34] <lynxlynxlynx> meh
[08:53:12] <avenger> you use the tagged source? i mean, my last day commit isn't in this, i hope ;)
[08:53:25] <lynxlynxlynx> it isn't
[08:53:37] <fuzzie> avenger: could you avoid adding guiscript hacks like the selection thing?
[08:53:57] <fuzzie> it means i have to break it at some point
[08:53:59] <lynxlynxlynx> i guess i'll just decry (the) windows (build) in the announcement
[08:54:14] <avenger> you mean passing 0 as partyID would take the selected actor?
[08:54:28] <avenger> actually, for the buttons, i would completely remove the party ID
[08:54:29] <fuzzie> avenger: well, either don't do that, or make that work for everything relevant
[08:54:51] <avenger> hmm, something needs it ?
[08:54:55] <fuzzie> but i would prefer to do it properly using globalids, so that modders can use it
[08:55:22] <fuzzie> a whole bunch of the button stuff needs moving to python anyway, we already have mods hard-coded into the C++
[08:55:47] <avenger> huh, really?
[08:56:04] <fuzzie> some Avenger doesn't like zero-charge items disappearing, apparently :P
[08:56:34] <fuzzie> and a few other things which differ from the original but apparently i shouldn't just change back
[08:57:04] <avenger> well, i'm not sure about this zero charge thing
[08:57:18] <avenger> the engine has a bug where it wraps around to 65535 on charges
[08:57:22] <lynxlynxlynx> that's for one item that indeed didn't disappear iirc
[08:58:08] <avenger> it is definitely a bug in the original, if i 'hacked' anything, i intended to remove that
[08:58:16] <avenger> all else is probably just programming error
[08:59:06] <fuzzie> there's some other differences hard-coded deep in python which i forget right now
[08:59:07] <avenger> ahh, i remember something else,
[08:59:31] <avenger> ok, i didn't like when wands disappear on zero charge without able to recharge them :)
[08:59:53] <fuzzie> but it should be possible to do this either way
[09:00:03] <lynxlynxlynx> game flag time
[09:00:06] <fuzzie> but it isn't, because it's hard-coded into the C++ where it can never be modded without a game flag :)
[09:00:31] <avenger> there are so many bits in the recharge field, you can always make it so
[09:00:37] <avenger> no need for game flag
[09:00:44] <lynxlynxlynx> avenger: so how to make the action bar work for summons? Now only the hardcoded buttons are there, no class specifics
[09:01:02] <fuzzie> this is why i am whining about it being a hack :P
[09:01:06] <avenger> if you don't like my 'mod' then assign an unused bit for it in the recharge field
[09:01:30] <avenger> just make sure you pick an unused bit :P
[09:01:42] <avenger> well, we can move it later
[09:02:01] <avenger> lynx?
[09:02:29] <lynxlynxlynx> eg., i summon a deva and get only the talk, 2xweapons and the innates button
[09:02:33] <fuzzie> GetActionButtonRow doesn't work for summons because we store in pcstats
[09:02:39] <lynxlynxlynx> not the full row - they can cast spells
[09:03:06] <avenger> it should be the qslots default row
[09:03:13] <avenger> qslots.2da, row 0
[09:03:13] <fuzzie> default row for the class?
[09:03:33] <avenger> yes, that row should be presented for any classes not specifically mentioned
[09:03:36] <fuzzie> i mean, different summons get different action bars
[09:03:36] <lynxlynxlynx> so just all the extra classes need to be added there?
[09:04:10] <avenger> i didn't find any class based difference
[09:04:10] <lynxlynxlynx> i'll retry
[09:04:21] <avenger> other than already written
[09:04:26] <fuzzie> avenger: well, simple question: how do summons with spells work?
[09:04:50] <avenger> i don't know, i guess they got mage class?
[09:04:52] <fuzzie> they should just have an existing class?
[09:05:00] <lynxlynxlynx> devas have 255
[09:05:06] <avenger> hmm
[09:05:08] <lynxlynxlynx> that's nothing or something like that
[09:05:41] <avenger> which deva?
[09:05:57] <lynxlynxlynx> good
[09:06:15] <avenger> devagood has cleric class
[09:06:18] <fuzzie> but it won't work anyway because all the calls like SetupSpellIcons and SpellCast and loads of other functions take an actor slot too :)
[09:06:24] <lynxlynxlynx> Class: 255 current class:255
[09:06:33] <avenger> resref?
[09:06:35] <fuzzie> so someone has to fix a lot of code
[09:06:37] <avenger> is it summoned?
[09:06:45] <avenger> not all devas are summoned
[09:06:52] <avenger> 'devagood' has cleric class
[09:06:57] <lynxlynxlynx> it is summoned
[09:07:05] <avenger> what is its resref?
[09:07:09] <lynxlynxlynx> devagood is the script name, i don't have the resref yet
[09:07:29] <avenger> it should be in the summoning spell !
[09:07:50] <lynxlynxlynx> it was from the paladin hla
[09:07:56] <avenger> spcl923?
[09:08:02] <fuzzie> we don't seem to sabotage it in the effect code
[09:08:09] <lynxlynxlynx> yes
[09:08:22] <avenger> it summons devagood
[09:08:27] <avenger> via external eff
[09:08:34] <avenger> devagood should have cleric class
[09:08:41] <fuzzie> does it then apply 0x48?
[09:08:53] <avenger> what is that fuzzie?
[09:08:57] <fuzzie> change id
[09:09:45] <avenger> i don't know, who would do that
[09:10:16] <avenger> my devagood.cre has 3 in the cleric field, isn't that so for you?
[09:10:23] <avenger> err, class field ;)
[09:11:49] <lynxlynxlynx> no
[09:12:03] <avenger> hmm
[09:12:16] <lynxlynxlynx> patch/fixpack
[09:12:47] <avenger> ok, i change it to ff, and see how my original game works
[09:12:48] <fuzzie> devagood gets set to cleric in the fixpack
[09:13:54] <fuzzie> i copied the original file, and a copy from a only Bioware official patched version, both say 'Class: None' in ielister
[09:14:59] <fuzzie> i see no relevant effects on it
[09:15:15] <lynxlynxlynx> planevil/plangood are also ff, but planet01 is tanarri
[09:15:47] <lynxlynxlynx> my game is unpatched and unmodded
[09:15:51] <avenger> well, i have to see this first, since all the cre's i checked had their class 'corrected' i didn't suspect any extra
[09:15:59] <fuzzie> i guess you have fixpack :)
[09:16:03] <avenger> sure i have
[09:16:15] <fuzzie> it works fine here, but i don't see what on earth field you'd check
[09:16:56] <avenger> well, play with wizard eye :)
[09:17:05] <avenger> at least that is done
[09:17:22] <fuzzie> meh, i leave all this stuff in the hands of you two
[09:17:36] <fuzzie> i want to fix scripting and the search map and etc
[09:17:40] <avenger> ok
[09:19:26] <lynxlynxlynx> the invisible stalker gets the full row :)
[09:20:31] <Lightkey> are you talking about me?
[09:20:48] <lynxlynxlynx> ups, he's on to us
[09:20:54] * lynxlynxlynx casts non-detection
[09:21:57] <fuzzie> i imagine spells should work if you just copy-and-paste Avenger's slot/selection thing and add some hack to the guiscript to pass 0 when GemRB.GameGetFirstSelectedPC doesn't give you a pc?
[09:22:19] <fuzzie> but there's too many calls for me to bother doing it
[09:23:03] <avenger> no, the button row is hardwired to class
[09:23:14] <avenger> in the original and ours too
[09:23:25] <avenger> so if someone has 0xff as class, i don't know what happens in the engine
[09:23:31] <fuzzie> you didn't try it?
[09:23:34] <avenger> probably there is some other layer
[09:23:39] <fuzzie> just remove the devagood.cre from your override
[09:23:49] <avenger> i modified it to 0xff
[09:23:52] <fuzzie> oh, ok
[09:24:04] <avenger> and i see it can still cast
[09:24:23] <fuzzie> it's odd, i don't see a hardcode
[09:24:36] <avenger> where?
[09:24:54] <avenger> i dont even know where is the wizard eye part, atm :)
[09:25:29] <avenger> ahh it is: 0068E453
[09:26:45] <avenger> i will have to go over this with ida :)
[09:27:36] <fuzzie> i mean, there's no hardcoded resref :)
[09:27:37] <avenger> meanwhile, see if you got 'init_action_buttons' in the latest bunch of commented files, i got one from aug. 02
[09:27:53] <avenger> it is damn close to what is already in our code
[09:27:59] <fuzzie> ohh no. i go eat breakfast :)
[09:28:11] <fuzzie> it is cool that PCStats get created for summons though
[09:28:14] <avenger> the first selected thing is there :)
[09:28:41] <fuzzie> sure, obviously we should do the first selected thing
[09:28:46] <avenger> they do it this way: none selected: empty row, more than one selected: group row, one selected: draw based on class
[09:28:49] <avenger> that's what i see
[09:29:01] <fuzzie> but either you have to do that in the python and use global ids, or you have to fix about 50 functions
[09:29:25] <avenger> well when you are back we talk about this :)
[09:29:36] <fuzzie> and it makes more sense to me to use global ids, since then modders can do magical multi-PC action bars
[09:29:46] <fuzzie> or whatever crazy things they want
[09:29:53] <fuzzie> but i don't really care, i just want to know which way to do it
[09:30:11] <fuzzie> (we can just start global ids at 0x80 or something like the original engine..)
[09:30:18] <fuzzie> but, yes, breakfast
[09:30:50] <avenger> there seems to be a special mark '0x70' for default class
[09:32:25] <fuzzie> oh, thanks for the charm code by the way
[09:33:39] <lynxlynxlynx> so, first things first - any comments on the release notes? I just changed the windows binary thing
[09:34:04] <avenger> you could mention the 10 years?
[09:34:17] <lynxlynxlynx> nx> http://pastebin.ca/1921907 <-- comments on the release
[09:34:29] <avenger> ah ok
[09:34:58] <lynxlynxlynx> i'll start with the uploading
[09:36:49] <avenger> i'm trying to figure out what gives the deva's spellcasting :)
[09:37:25] <avenger> hmm, one guess, maybe the default has the casting in it
[09:37:30] <avenger> never seen any other summon
[09:37:50] <avenger> if there are no spells, the button is grayed
[09:37:52] <fuzzie> could be. easy to check?
[09:37:53] <avenger> but it is there
[09:38:13] <avenger> yep, just tell me some summonable creature? invisible stalker?
[09:38:19] <avenger> that's ok, i guess
[09:39:00] <lynxlynxlynx> invisible stalker gets the spellcasting icon in gemrb
[09:39:08] <lynxlynxlynx> has no spells, so it is grayed out
[09:39:39] <lynxlynxlynx> more examples that should be able to cast and do have stuff memorised: djini, efreet, hakeashar, nishruu
[09:40:26] <fuzzie> just cast monster summoning I
[09:40:34] <avenger> wait so invisible stalker has a class?
[09:40:51] <avenger> hmm it is 0xff for me
[09:41:09] <avenger> so it cannot get spellcasting icon in gemrb, or the deva would too
[09:41:42] <avenger> at the moment i think the default row give the spellcasting icon
[09:42:13] <avenger> so i need to know if there is any summon in the original that has no such button at all
[09:42:32] <lynxlynxlynx> Class: 1 current class:1 here
[09:42:59] <lynxlynxlynx> 8: mage01 - (0 0 0) 68a9 Wt: 1 x 0Lb <-- also supportive
[09:43:38] <lynxlynxlynx> so try monster summoning, those are just grunts
[09:44:06] <pupnik> ls cd2 // Data Movies
[09:44:32] <pupnik> is it normal for bg2 to have caps for the Moves/Data dirs? and normal for gemrb to look for lowercase?
[09:45:53] <lynxlynxlynx> yes
[09:47:30] <pupnik> so this is just a bad gamedata dir?
[09:48:05] <pupnik> ohh my symlink is broken
[09:48:41] <lynxlynxlynx> do you have CaseInsensitive enabled?
[09:48:47] <avenger> ok we are good, invis stalker has the same button row as the deva
[09:48:55] <pupnik> boink - there it is ty
[09:49:08] <avenger> ogrillon too
[09:49:16] <avenger> ogrillon could cast, if got spells ;)
[09:50:02] <avenger> skeleton too, ok, so lynx, you just have to set up the default row in qslot.2da
[09:50:45] <lynxlynxlynx> cool
[09:51:02] <avenger> there is only a small thing we cannot do, the default row has an attack icon and an empty icon instead of weapon1/2
[09:51:23] <avenger> we currently have those slots hardcoded
[09:52:07] <avenger> qslots2 already has them unhardcoded, btw
[09:52:34] <avenger> so qslot.2da could be converted to the same format by adding the missing 3 columns
[09:52:52] <lynxlynxlynx> sound like a good idea
[09:53:16] <avenger> originally i didn't have those because they ARE hardcoded in iwd2 too
[09:53:30] <avenger> iwd2 has only 9 customizable slots
[09:53:36] <avenger> the first 3 are hardcoded
[09:54:08] <avenger> and iwd2 is the least hardcoded about slots :)
[09:55:27] <fuzzie> sounds good
[09:55:46] --- lynxlynxlynx has changed the topic to: GemRB 0.6.2 | http://gemrb.sf.net | Be wary of your words for there are Modron sensors in this channel: http://log.usecode.org/gemrblog.php | Hey <CHARNAME>, we need some awesome screenshots!
[09:56:13] <fuzzie> avenger: hey, do you know how the original engine does ticks?
[09:56:13] <lynxlynxlynx> only g3 to go :)
[09:56:18] <fuzzie> or anyone, for that matter
[09:56:50] <fuzzie> i'm wondering if i can get away with limiting gemrb to a single tick per frame
[09:57:06] <avenger> i don't know
[09:57:34] <fuzzie> at the moment we can increase GameTime by loads of ticks at once
[09:57:38] <fuzzie> it breaks scripts, etc
[09:57:56] <avenger> well, it seems it is called once per tick
[09:58:08] <avenger> each increase of 0x1dd0
[09:58:31] <avenger> but that is mostly effect based
[09:58:37] <avenger> i don't know how actions work
[09:58:52] <fuzzie> they get run every tick
[09:58:56] <avenger> opcodes seem to be called once for each pGame->0x1dd0 increase
[09:59:19] <avenger> well, if they are recurring, the IE has some opcode call optimisation
[09:59:20] <fuzzie> i think scripts get checked once per second, the tick at 15%globalid
[09:59:49] <avenger> that could be
[10:00:13] <fuzzie> erm, globalid%15
[10:00:13] <avenger> you meant globalid%15 ?
[10:00:20] <fuzzie> they seemed to quite like the idea of doing globalid%x
[10:00:31] <avenger> because it spreads the load
[10:00:39] <fuzzie> that thing i was talking about the other day was Taimon saying that Turn() apparently does it to work out the modifier
[10:00:57] <fuzzie> which is, well, not so smart, since the globalid there is the turner :)
[10:01:22] <avenger> maybe he just mistook that part, there is a similar thing in spawns, too
[10:01:39] <fuzzie> the spawns thing makes sense, i think
[10:01:47] <fuzzie> anyway, i mustn't get so distracted
[10:02:11] <avenger> i think it was a good idea to do that
[10:02:41] <avenger> i mean, doing the globalid%15 thing :) that might help with timing problems
[10:04:14] <fuzzie> it makes them worse :P
[10:04:31] <fuzzie> much worse, actually
[10:04:38] <fuzzie> but i will try it after i fixed the relevant bugs
[10:04:46] <avenger> then global id is not generated correctly ?
[10:05:30] <fuzzie> no
[10:05:36] <fuzzie> but we have all kinds of code in the wrong place
[10:05:36] <pupnik> is there an override directory for bg2 cd3/Data? Keyimporter is erroring out now on all cd3 stuff
[10:05:48] <fuzzie> i think it'll work fine, once i fix a bunch of other bugs
[10:07:55] <lynxlynxlynx> pupnik: you have to do a full install
[10:10:46] <pupnik> lynxlynxlynx: sorry, is that something different than "make install"?
[10:11:15] <lynxlynxlynx> of the game :)
[10:11:40] <lynxlynxlynx> avenger: please post http://pastebin.ca/1921931 to the webring
[10:13:53] <pupnik> lynxlynxlynx: the game was full-installed, i.e. all directories are full of data - i can't see why cd3 fails when the rest work
[10:14:05] <pupnik> but this is just something stupid, so i'll be quiet
[10:15:35] <lynxlynxlynx> did you set the CD3 path correctly?
[10:18:19] <fuzzie> if we can get this logging stuff which tomprince started in, hopefully we can actually start providing useful errors
[10:19:12] <pupnik> yeah it's all right
[10:19:31] <pupnik> http://pastebin.com/X2Z76aC7 see how this guy gets all the cd3 errors, but gemrb is ok with it?
[10:20:12] <pupnik> my problem seems to be with the overridepath... right after those errors he loads gemrb.ini, but I get error [ResourceManager]: Searching for gemrb.ini...[ERROR]
[10:20:28] <fuzzie> right up to the point where you try using some of the cd3 data
[10:20:39] <fuzzie> is your game type set correctly?
[10:21:02] <fuzzie> and did you make sure not to change anything under 'HERE BE DRAGONS'?
[10:21:34] <pupnik> game type is bg2
[10:21:44] <pupnik> so it should pull from /usr/local/share/gemrb/override/bg2
[10:22:42] <pupnik> i have no 'HERE BE DRAGONS'- which gemrb.cfg should i be sourcing from?
[10:23:15] <pupnik> nm found it - recopyingsettingup
[10:23:17] <pupnik> gemrb-0.6.2$ grep -r DRAGONS *
[10:24:04] <fuzzie> it's in the gemrb subdir
[10:24:20] <fuzzie> there should only be one example gemrb.cfg which isn't marked 'noinstall'..
[10:27:01] <pupnik> ok that one works :/
[10:27:05] <pupnik> ty
[10:31:34] <pupnik> loading a bg2 game @ 23hours gives me a long slew of 'romantic' dialogues
[10:31:44] <pupnik> as if there was a backlog
[10:32:02] <pupnik> and i lost the green/red circles marking the destinations
[10:32:09] <pupnik> maybe i should start from beginning
[10:33:11] <pupnik> "Now... I wonder where my Calimshite darling went off to...always time for one last quickie..." lol @ Uncle Lester
[10:35:52] <fuzzie> the circles are set in your baldur.ini
[10:36:00] <fuzzie> i think we don't set the defaults correctly
[10:38:35] <pupnik> how do i disable the intro scene with irenicus and dungeon?
[10:39:59] <pupnik> they seem to work initially, then after clicking on a unwalktoable space, the green circlies disappear
[10:40:04] <avenger> lynxlynxlynx: http://forums.gibberlings3.net/index.php?showtopic=20838
[10:40:23] <pupnik> ty
[10:40:34] <lynxlynxlynx> yay
[10:40:47] <pupnik> oh heh
[10:40:50] <pupnik> congrats!
[10:41:54] <avenger> btw, with igi gone, teambg was reincarnated again :)
[10:43:16] <lynxlynxlynx> what, was he preventing it?
[10:43:39] <avenger> no
[10:43:48] <avenger> he just had another reincarnation
[10:44:24] <avenger> it is coming back like a good zombie :)
[10:45:02] <Lightkey> like a paradoxon?
[10:46:25] <avenger> so what's up with the windows binary?
[10:47:00] <avenger> isn't it some memory allocation with -1 size? or the like
[10:47:18] <pupnik> so happy the party-selection keys work - thanks so much
[10:47:37] <avenger> my windows gemrb runs perfectly, though
[10:48:01] <lynxlynxlynx> maybe it's the new mingw toolchain
[10:48:18] <lynxlynxlynx> the older releases were with older gcc and co
[10:48:19] <fuzzie> it is just mingw
[10:48:30] <fuzzie> destruction order issues, i think
[10:49:11] <avenger> gemrb takes up 13M on my machine !
[10:49:17] <avenger> even svchost is bigger
[10:49:53] <avenger> takes up 30-40 cpu, that's a lot though :)
[10:49:55] <fuzzie> doing what, though? :)
[10:50:05] <avenger> standing around in irenicus dungeon
[10:50:08] <fuzzie> huh
[10:50:13] <fuzzie> well, that is just something lying to you :P
[10:50:32] <fuzzie> just the MOS for the world map is 5mb
[10:50:58] <avenger> it is now 17M
[10:51:00] <fuzzie> and the dungeon tiles are at least 10mb, so that's more than 13mb already
[10:51:45] <avenger> strange, i quit to main menu, and it is now 18
[10:52:11] <fuzzie> sure
[10:52:18] <fuzzie> i mean, how are you reading these numbers?
[10:52:24] <fuzzie> just out of task manager?
[10:52:37] <avenger> huh, now loaded game again, and it is 60
[10:52:41] <avenger> yes, out of task manager
[10:52:52] <fuzzie> the first rule of memory management is: the only numbers you can trust are ones you keep track of yourself
[10:52:54] <avenger> i guess it doesn't add the pieces up?
[10:53:04] <fuzzie> task manager just tells you how many pages are in RAM
[10:53:07] <avenger> 60m is more believable
[10:53:09] <fuzzie> that is pretty meaningless
[10:53:49] <avenger> well, that still doesn't explain why the first load showed only 13 :)
[10:53:52] <avenger> it was a fresh load
[10:54:00] <fuzzie> you can free() hundreds of mb, and Task Manager will show no difference
[10:54:13] <fuzzie> but yes, 13mb is just a lie, i don't know why that would happen :)
[10:54:52] <pupnik> 14MB here
[10:55:08] <pupnik> well 6.7% mem @ 2GB
[10:55:19] <fuzzie> again, you're just reading that from top? :)
[10:55:21] <CIA-26> GemRB: 03lynxlupodian * r6c2d44c7f321 10gemrb/admin/announcement.template: announcement.template: intro for other sites
[10:55:31] <pupnik> heh oops 140
[10:55:34] <avenger> ok, now when i quit/reload the same area it goes back from 60 to 24
[10:55:54] <avenger> i guess that's 36 is reliably the memory requirement for the first area
[10:56:03] <avenger> mostly the tiles?
[10:56:17] <pupnik> yes fuzzie
[10:56:22] <fuzzie> avenger: i think so
[10:56:26] <fuzzie> i have an exact list somewhere
[10:56:39] <fuzzie> 5mb for the worldmap MOS, 20mb or so for the tiles, most of the rest is animation BAMs
[10:56:48] <avenger> anyway, i just wanted to say, there is no memory hog in my version
[10:56:49] <fuzzie> and the rest is <1%
[10:56:56] <avenger> so, any hogging is some 3rd party dll
[10:56:57] <fuzzie> but yes, the memory issue is definitely mingw
[10:57:04] <avenger> or that
[10:57:32] <lynxlynxlynx> ok, so everything is back to normal
[10:57:55] <lynxlynxlynx> free your latent magics :)
[10:58:04] <avenger> i can give someone a binary build, but i'm not up to create the whole package
[10:58:34] <avenger> hmm, also, this is a debug build, shall i recompile with optimisations?
[10:58:37] <fuzzie> i have visual studio 2010 around here
[10:58:43] <fuzzie> which produces much faster builds
[10:58:47] <fuzzie> but i can't do it right now
[10:59:21] <Lightkey> seems to be a recurring theme ;-)
[10:59:35] <avenger> well, i can build the binary now :P
[10:59:37] <fuzzie> yep
[10:59:59] <fuzzie> and, well, if you can build a binary which doesn't depend on anything else..
[11:00:19] <avenger> i have problems with python
[11:00:37] <avenger> that's why i'm not up to create the whole package
[11:00:57] <fuzzie> with guiscript.dll, or with python?
[11:01:18] <avenger> the python dll has different name for each release
[11:01:42] <fuzzie> oh
[11:04:19] <avenger> hehe, our core is still only near 1M
[11:05:03] <avenger> i wonder if it is possible to do a partially static build?
[11:05:16] <avenger> like compile in plugins that got no alternative
[11:05:44] <lynxlynxlynx> our cmake supports fully static builds
[11:06:07] <avenger> fully static sucks, because you cannot switch to nullsound
[11:06:21] <lynxlynxlynx> but since mingw is broken and we don't have cc set up, it is pretty irrelevant here
[11:06:48] <avenger> why is it important to have the windows binary with mingw?
[11:06:57] <lynxlynxlynx> i doubt people that need binaries want no sound
[11:07:13] <lynxlynxlynx> it is not important, it was just the easiest way
[11:07:39] <avenger> i got a compile with python26
[11:07:48] <fuzzie> you can static link with multiple audio/video plugins
[11:08:13] <fuzzie> but then you require that the relevant DLLs be present+working
[11:08:14] <avenger> fuzzie: well, ok, then it is good, that is not much overhead anyway
[11:08:47] <avenger> ah well, with sound it is openal and nothing else? or sdl mixer is a special stuff?
[11:09:01] <fuzzie> you get to pick via a AudioDriver= line in config
[11:09:16] <fuzzie> the plugin will only initialise if it matches
[11:09:27] <fuzzie> even if statically linked
[11:09:57] <avenger> python is a requirement anyway
[11:10:00] <fuzzie> yes
[11:10:02] <avenger> there is no alternative
[11:10:13] <avenger> and what else, i wouldn't link in the ogg part anyway
[11:10:26] <fuzzie> but why link in anything?
[11:10:29] <lynxlynxlynx> is that 90/100 multiplier in the summoning from the original? to get a sane duration, it needs to be much lower in current gemrb (1/10 is pretty close)
[11:10:33] <avenger> to have a single exe?
[11:10:35] <fuzzie> you don't gain if you still have some things in plugins
[11:10:48] <avenger> well, to have less dlls
[11:11:19] <fuzzie> lynxlynxlynx: you have numbers?
[11:11:32] <lynxlynxlynx> no, this is just by trying
[11:11:33] <avenger> if i compile in everything, it might be smaller than the pieces. Even if i compile in the multiple plugins
[11:11:53] <fuzzie> well, yes, i guess it saves some space
[11:12:11] <fuzzie> i think compared to a 4gb+ game install an extra mb doesn't matter, but :)
[11:12:18] <avenger> and before you ask, no i don't really care about that space, i just thought about comparing it to the original engine size
[11:12:41] <lynxlynxlynx> 1/10 is just a tiny bit too slow, 9/10 is horrendously late
[11:12:42] <fuzzie> and i think the duration calculation is just bad, for SummonCreature
[11:13:21] <fuzzie> because it is a non-prepared duration, right?
[11:13:27] <avenger> hehe
[11:13:30] <avenger> that could be
[11:13:42] <fuzzie> so it is 15x too high or something
[11:14:03] <avenger> gotta divide by ai update time?
[11:14:25] <fuzzie> why not just create it with the native timing mode?
[11:14:58] <fuzzie> or is it already?
[11:14:58] <avenger> if you want to divide by 15, simply change the AI_UPDATE_CYCLE to 1 :P
[11:15:15] <avenger> newfx->Duration = vvc->GetSequenceDuration(1)*90/100;
[11:15:28] <fuzzie> yeah, but then you get really awful times
[11:15:39] <lynxlynxlynx> 15 looks good
[11:16:16] <fuzzie> the original engine really uses FX_DURATION_INSTANT_LIMITED here?
[11:16:58] <fuzzie> i would use FX_DURATION_ABSOLUTE and use the existing code, adding GameTime to Duration
[11:17:23] <avenger> yes fuzzie, that would work too
[11:17:27] <fuzzie> but if the original engine really does things in seconds, then ok
[11:17:48] <avenger> i didn't really check how this particular timing works, just quickly typed in something
[11:17:53] <fuzzie> ah, ok
[11:18:14] <fuzzie> lynxlynxlynx: well, i guess that, then: make SummonCreature use FX_DURATION_ABSOLUTE and add the GameTime to the Duration
[11:18:46] <avenger> the engine does most durations that way anyway :)
[11:18:57] <fuzzie> and now i am getting out of here
[11:19:04] <fuzzie> be back later-ish
[11:19:09] <avenger> see you
[11:24:54] <CIA-26> GemRB: 03lynxlupodian * rb94eb4adfde5 10gemrb/gemrb/core/Interface.cpp: fixed the summondisable timing, yay for teamwork
[11:31:48] <avenger> hmm, i got one layer closer to the action button row setup
[11:32:38] <avenger> and here it loads the bam, so i will see actual cycle numbers
[11:45:29] <avenger> cool i see them now
[11:45:41] <avenger> bg also has 100 for empty, just like iwd2 ;)
[12:33:00] <fuzzie> huh, msys git comes with vim o.O
[12:33:38] <fuzzie> i accidentally typed 'vim' into this shell without thinking and it worked
[12:33:57] <fuzzie> tortoisegit, on the other hand, is the most useless thing
[13:23:28] --> xrogaan_ has joined #GemRb
[13:23:31] <-- xrogaan_ has left IRC (Remote host closed the connection)
[13:32:34] <pupnik> havent found any critically missing things playing from start yet
[13:32:55] <pupnik> divvying up potions with double click not working
[14:23:21] <fuzzie> mm, that is annoying :)
[14:26:52] <pupnik> lightning bolt in irenicus dungeon 'strange machine' room only did 2 damage to minsk
[14:27:02] <pupnik> i wish i recalled the rules
[14:27:32] <fuzzie> the bug there is that it doesn't bounce, i think
[14:27:41] <fuzzie> it's not really meant to hurt you too much
[14:29:04] <pupnik> ahh
[14:32:04] <pupnik> user-interface (game screen border) is dissappearing when one character is swinging
[14:32:14] <fuzzie> that means the gui code broke :(
[14:32:24] <fuzzie> save/quit/reload
[14:32:50] <fuzzie> we still haven't tracked that bug down, but i'm pretty sure it's because of all the guiscript code which hacks variables for HideGUI/ShowGUI
[14:33:04] <fuzzie> so it gets into an inconsistent state
[14:33:13] <pupnik> ahhh
[14:33:48] <fuzzie> tomprince suggested replacing that a while ago, but i thought it would just cause more bugs
[14:33:54] <fuzzie> but no, it is just broken and needs replacing
[14:40:09] <pupnik> imoen and minsc attacking animations not running vs lightning mephit (console output shows Performattack and Initround)
[14:40:20] <pupnik> jaheira is animating
[15:06:48] <pupnik> Ogre Mage from 1st lev only dropped a Bastard sword
[15:08:56] <lynxlynxlynx> what do you miss?
[15:09:36] <pupnik> i need to run through this with wine as well - thought there was more
[15:09:53] <pupnik> just saw that goblin animation when near death is the 'dying' animation over and over
[15:12:29] <lynxlynxlynx> nothing is missing :)
[15:12:39] <pupnik> ok ty
[15:18:04] <pupnik> Imoen cast stinking cloud on goblin archers held by web-spell, and was teleported to inside the radius of the stinking cloud! :/
[15:18:26] <pupnik> oops, no, jaheira was
[15:18:36] <pupnik> hmm
[15:18:55] <avenger> maybe she just walked there by AI?
[15:19:24] <pupnik> yep i'm thinkin maybe i missed observing a placement
[15:20:24] <avenger> heh, there is an action button with the picture of a sheep
[15:20:31] <avenger> i wonder when it is shown
[15:20:54] <lynxlynxlynx> for the summon cow spell? :)
[15:21:05] <avenger> not a spell icon
[15:21:19] <avenger> just like an attack/turn/etc button
[15:21:23] <lynxlynxlynx> i don't know the spell, so i thought it summoned a cow
[15:21:34] <avenger> it summons a cow, and it falls on the target
[15:21:43] <avenger> the falling cow is a projectile :)
[15:21:45] <lynxlynxlynx> hehe
[15:22:05] <avenger> this is something different
[15:22:10] <fuzzie> debug code? :)
[15:23:08] <avenger> well, i just found 'holes' in the button types, which never got assigned by the normal 'button row by class' thing
[15:23:27] <avenger> and when i went deeper, on button level, i found that one of them is this sheep :)
[15:23:40] <fuzzie> or maybe you found code for an easter egg
[15:23:48] <avenger> hehe
[15:24:25] <avenger> i didn't find how to trigger it, i could modify the exe to show this for bard's singing or such
[15:24:30] <avenger> then i could press it ;)
[15:28:41] <pupnik> picking stuff up and inventory is nice and reliable it seems - yaay
[15:30:45] <pupnik> nice how i can still target monsters standing behind doors +++
[15:31:31] <fuzzie> really must fix the stupid searchmap code :p
[15:32:01] <pupnik> sorry i mean select them as targets - the door is open but the monster icon is behind it
[15:32:12] <pupnik> so giving monster priority is a great way to do things
[15:39:10] <avenger> there is one thing we don't do correctly: the special abilities button row for cleric/thief starts with thieving (in the original)
[15:44:41] <pupnik> another is the sparkly 'snow' in the steam mephit room runs down the screen forever
[15:44:52] <pupnik> one of those difficult ones from long ago, iir
[15:44:53] <pupnik> c
[15:51:42] <lynxlynxlynx> sparkle puff
[15:52:29] <pupnik> mephits are a great introduction to the game - shows off some of the status spells nicely
[15:53:33] <lynxlynxlynx> they're pretty annoying
[16:15:52] <avenger> i looked at sparkles today
[16:17:29] <pupnik> could you rule some causes out?
[16:17:48] <fuzzie> the 'runs down the screen forever' thing is just because our sparkles never expire and never get removed
[16:18:15] <pupnik> reminds me of some websites :D
[16:18:18] <fuzzie> but someone obviously wrote all of this code for a reason, so i don't want to touch it
[16:18:21] <avenger> there is one problem, the sparkles in the ie got a duration field, it stops generating new particles after that time is off
[16:18:52] <fuzzie> well, our particles also don't move randomly
[16:18:59] <fuzzie> despite there being code for this
[16:19:18] <pupnik> now i got machine-gun sound effects repitions
[16:19:38] <fuzzie> pupnik: something with Fireshield?
[16:19:58] <pupnik> looks like the stone shield that imoen has
[16:20:02] <avenger> yeah, that machine gun stuff is annoying
[16:20:04] <fuzzie> yes, that
[16:20:07] <fuzzie> it is avenger's fault :P
[16:20:15] <avenger> sure blame it on me :P
[16:20:23] <fuzzie> the effect apply stuff doesn't use triggers
[16:20:42] <fuzzie> it just pokes around at the internals of objects
[16:21:08] <fuzzie> i think i said i would fix that :) but not right this very moment
[16:25:07] <pupnik> gosh that 'web' radius seems huge @ 640x480
[16:25:50] <pupnik> should i test at 800x480?
[16:32:38] <fuzzie> don't think it's necessary
[16:35:09] <lynxlynxlynx> looks fine to me
[16:49:42] <pupnik> quitting and reloading seems to be a practical workaround to a number of things :)
[16:59:02] <pupnik> hehe the neighbor's horse looked a little unsettled when i turned-up the orchestral background just now
[17:16:13] <-- deepinthewoods has left IRC ()
[17:29:34] --> deepinthewoods has joined #GemRb
[18:09:28] <avenger> hahaha
[18:09:41] <avenger> if i target an immobile opponent, i get hurt myself
[18:10:21] <avenger> 1. use chromatic orb on a goblin (imoen is good), 2. cast magic missile on it
[18:10:36] <fuzzie> your projectile code is doing something silly?
[18:10:51] <avenger> no it is not the projectile code
[18:10:56] <lynxlynxlynx> chromatic orb is a bad example
[18:11:08] <lynxlynxlynx> iirc the stun effect is always applied on the caster there
[18:11:15] <avenger> i have fixpack
[18:11:19] <lynxlynxlynx> oh
[18:11:19] <avenger> so it works here
[18:11:37] <avenger> you can try to freeze your opponent with something else
[18:11:42] <avenger> maybe web or entangle
[18:11:44] <fuzzie> what do you think it is, if not projectile code?
[18:11:50] <avenger> targeting
[18:11:59] <fuzzie> what kind of targeting?
[18:12:02] <avenger> clicking on an immobile returns 0, maybe
[18:12:17] <avenger> if you supply 0 as target, the caster will be the target
[18:12:34] <avenger> well, actually, it could be the projectile code
[18:13:00] <fuzzie> i don't see it in the projectile code, it just uses global ids
[18:14:03] <fuzzie> oh dear, TryToCast never sets a target
[18:14:13] <fuzzie> oh, it does
[18:15:04] <fuzzie> //hidden actors are not selectable by clicking on them
[18:15:04] <fuzzie> Actor* actor = area->GetActor( p, GA_DEFAULT | GA_SELECT | GA_NO_DEAD | GA_NO_HIDDEN);
[18:15:16] <fuzzie> ^- so, yes, you can't cast at an immobile target if you pass GA_SELECT!
[18:16:07] <fuzzie> that is just wrong, right?
[18:16:18] <fuzzie> PerformActionOn should just check GA_SELECT itself before selecting something?
[18:20:01] <fuzzie> ok, no reply, you get a commit :P
[18:20:31] <CIA-26> GemRB: 03fuzzie * r24baaa66a635 10gemrb/gemrb/core/GUI/GameControl.cpp: you don't have to be able to select an actor to target it
[18:21:33] <fuzzie> i see other bugs with this too
[18:24:42] <fuzzie> this GameControlSetTargetMode code everywhere defaults to GA_SELECT too, also not my fault
[18:25:44] <CIA-26> GemRB: 03fuzzie * r6a21eb3a6c57 10gemrb/gemrb/core/Map.cpp: fix GetActorInRect to only check GA_SELECT for party mode
[18:25:46] <fuzzie> with that commit, GameControlSetTargetMode is the last issue, though
[18:25:51] <fuzzie> everything else uses it correctly
[18:26:32] <avenger> i'm looking at the spark code
[18:27:50] <fuzzie> cool
[18:28:00] <fuzzie> i think our code is just buggy..
[18:28:27] <fuzzie> i just don't understand how it's meant to work :)
[18:43:25] <CIA-26> GemRB: 03fuzzie * rfabc9be28632 10gemrb/gemrb/ (4 files in 3 dirs): always check target types, always reset target types
[18:48:35] <CIA-26> GemRB: 03fuzzie * r26a0d10bddae 10gemrb/gemrb/core/GUI/GameControl.cpp: don't default to GA_SELECT in target type
[18:49:00] <fuzzie> so, lots of changes there, someone should check them
[18:49:50] <fuzzie> but i hope those fix the stupid targetting bugs
[18:57:11] <CIA-26> GemRB: 03fuzzie * re9d0908609de 10gemrb/gemrb/core/GUI/GameControl.cpp: don't allow casting at the ground if a target is required
[18:58:19] <fuzzie> it still breaks if you try casting at doors or infopoints etc
[18:59:29] <-- deepinthewoods has left IRC (Ping timeout: 265 seconds)
[19:00:12] <fuzzie> but i don't see how to detect those from spelldata.Target
[19:00:53] <lynxlynxlynx> they can't just be excluded either or knock would stop working
[19:04:46] <fuzzie> knock has target type TARGET_CREA..
[19:06:51] <fuzzie> so i guess this is a question for avenger
[19:11:53] --> deepinthewoods has joined #GemRb
[19:13:38] <avenger> ok, now i have a fountain sparkle, it is a bit slow yet
[19:14:06] <fuzzie> are the fountain sparkles really used?
[19:14:20] <fuzzie> i mean, apart from caspenar, which i know is the one you like :p
[19:19:52] <avenger> ok, i call cespenar's sparkle type as shower, the fountain one is a tiny explosion
[19:20:15] <fuzzie> ah, okay, those are the real problem :)
[19:20:21] <fuzzie> if you fix those then it would be really great
[19:22:45] <fuzzie> lynx/avenger: the original engine seems to handle travel regions by forcing the target to a certain point and then everyone just trying to get 'near enough' to it, is that a good summary of what i should fix?
[19:23:39] <avenger> how would you summary our engine's work?
[19:24:15] <fuzzie> gemrb does "don't keep track of the travel region, just try and form a formation somewhere vaguely near where the user clicked"
[19:24:20] <avenger> if someone gets near the travel region, then the rest of the party should be 'not scattered'
[19:24:40] <lynxlynxlynx> it also gives up after some time and just proceeds if at least one selectee is close enough
[19:24:42] <avenger> that not scattered check is the same as for resting
[19:24:45] <fuzzie> this means you can't get to travel regions if they're behind another one (because everyone travels through the first), and some travel regions you can't use at all, because the actors try doing a formation
[19:25:06] <fuzzie> the check whether you are allowed to travel is a different thing, which i don't intend to change here :)
[19:25:08] <avenger> well yes, that's bad
[19:27:05] <fuzzie> the scattered check is in InfoPoint::Entered, i guess
[19:27:13] <fuzzie> which checks for individual actors
[19:27:45] <fuzzie> and then CheckTravel, which uses EveryoneNearPoint
[19:32:18] <fuzzie> so, hm, i wonder why that fails
[19:33:20] <fuzzie> it checks Entered for every actor, and if any actor passes, it calls UseExit
[19:34:15] <fuzzie> and then it only fails if the distance is above MAX_TRAVELING_DISTANCE, which is 400, 25 search squares
[19:37:29] <fuzzie> well, i guess the distance to the actor is checked, and the lead actor is perhaps sometimes quite far from the last one
[19:48:26] --> pupnik_ has joined #GemRb
[19:49:32] <-- pupnik has left IRC (Read error: Operation timed out)
[19:55:07] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[20:01:22] <-- avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
[20:24:19] --> Maighstir has joined #GemRb
[20:36:21] --> tomprince_loki has joined #GemRb
[20:39:12] <fuzzie> tomprince_loki: i think if you're going to have a function which aborts, it should probably be called LogAndAbort or something
[20:41:11] <tomprince_loki> Sensible.
[20:42:50] <fuzzie> i was going to comment more on some silly bits, but it looks like this is just automated
[20:45:57] <tomprince_loki> A lot of it is, but not all.
[20:46:29] <fuzzie> sorry, i mean, i thought you'd done really silly things, but then realised you haven't actually manually inspected all of this :)
[20:47:00] <fuzzie> but e.g. the DebugDump() functions shouldn't really have a new 'Actor' in front of each line
[20:47:16] <fuzzie> there might well be a cleverer way to do DebugDump in general, though
[20:48:01] <tomprince_loki> I pondered that.
[20:49:16] <tomprince_loki> But it doesn't seem entirely unreasonable to prefix everything with '[Actor]: '. That way, everything lines up as you would expect.
[20:49:46] <fuzzie> but mostly: this patch just changes every single printf
[20:49:46] <tomprince_loki> And it makes grepping for it easier.
[20:50:04] <fuzzie> including the "this is too broken to call any functions in core, aborting" one. :P
[20:50:11] <tomprince_loki> Yes.
[20:50:39] <tomprince_loki> Which one is that?
[20:50:49] <fuzzie> well, i guess that one becomes a LogAndAbort
[20:50:55] <fuzzie> and, Interface::SanityCheck
[20:51:32] <fuzzie> it would be nice to have a smarter way to do that, but i couldn't think of one which would actually work
[20:51:45] <fuzzie> although checking things like sizeof(Interface) would make sense too
[20:53:04] <fuzzie> i had an argument for the DebugDump thing but it has somewhat evaporated in the light of further thinking
[20:53:14] <fuzzie> this "GameScript/Actions" stuff should go, though
[20:53:21] <fuzzie> but that can just be re-sedded, i guess
[20:54:05] <fuzzie> but other things i noticed: the ResourceManager etc and other things which don't have a '\n' don't work out well
[20:54:26] <fuzzie> the same for things which do have a '\n' but are meant to come later, i guess
[20:55:10] <fuzzie> and i think it *is* important that things are still able to print debug info as they go without it spread over 10+ lines.
[20:58:31] <fuzzie> a lot of the things i'm noticing are broken in the original code, though..
[20:58:42] <fuzzie> lots of missing newlines in errors
[21:00:08] <fuzzie> but i am guessing that fixing those in trunk is going to be really annoying.
[21:02:31] <fuzzie> i think 'LogAbort' still isn't clear enough :P but no big issue, i guess
[21:04:10] <fuzzie> if i were doing this then i'd rewrite this so it does the Logging.h first and then gradually moves everything over file-by-file, so everyone can patch things up as they go
[21:08:23] <fuzzie> (not sure if that's actually easier or anything, just commenting)
[21:09:14] <fuzzie> but mostly it looks fine, other than the above my only issue is that it 'formalises' a lot of the errant printf()s so that it's difficult to tell that they should really be removed rather than being intended to be permanent
[21:09:30] <fuzzie> hence the thought about bit-by-bit.
[21:23:45] <fuzzie> tomprince_loki: i've had that "PyDict_Merge" commit for a while and encountered absolutely no issues, btw
[21:43:11] <Lightkey> o.o
[21:45:55] <fuzzie> reviewing by means of walls of text :p
[21:50:04] <CIA-26> GemRB: 03avenger_teambg * rd49a39de9147 10gemrb/gemrb/ (6 files in 2 dirs): better sparkles
[21:52:53] <-- deepinthewoods has left IRC (Remote host closed the connection)
[21:56:11] --> Avenger has joined #GemRb
[21:56:12] --- ChanServ gives channel operator status to Avenger
[21:56:26] <Avenger> so, are the sparkles worse or better :)
[22:12:17] <pupnik_> ooo
[22:17:34] <pupnik_> ty for those fixes fuzzie - hope it works out good :)
[22:38:51] <fuzzie> Avenger: cool
[22:40:38] <fuzzie> doesn't build though
[22:41:29] <Lightkey> and what's Fonzie like?
[22:42:05] <fuzzie> hi, spamkey :)
[22:42:12] <Lightkey> shush
[22:44:41] <CIA-26> GemRB: 03fuzzie * rea962e633c3a 10gemrb/gemrb/core/Particles.cpp: add brackets
[22:53:50] <Avenger> whoa doesn't build? it did for me O_o
[22:54:02] <fuzzie> not enough brackets :)
[22:55:31] <fuzzie> Avenger: you know how targetting for doors/containers/etc works?
[22:55:50] <Avenger> global id?
[22:55:53] <fuzzie> i mean
[22:56:02] <fuzzie> Knock has TARGET_CREA set, for example
[22:56:19] <fuzzie> so we just let all spells be targetted at all scriptables?
[22:56:21] <Avenger> when it detects you clicked on a door, it constructs an action
[22:56:39] <Avenger> and copies the whole object into the action's object
[22:56:53] <fuzzie> yes. i didn't explain well. spells! :)
[22:56:59] <Avenger> spells too
[22:57:10] <fuzzie> TARGET_CREA just means "all scriptables"?
[22:57:16] <Avenger> yes
[22:57:39] <fuzzie> i didn't commit global ids yet, so all spells cast at scriptables are broken
[22:58:12] <fuzzie> i just wondered if it was a targeting bug, or just the lack of global ids
[22:58:54] <fuzzie> i hope i fixed all other GUI targetting bugs now
[23:00:18] <fuzzie> wow, sometimes projectiles in gemrb are really pretty :)
[23:00:53] <fuzzie> playing with the Radiant Mephit, i don't remember it being this pretty in the original
[23:01:42] --> deepinthewoods has joined #GemRb
[23:02:08] <fuzzie> the sparkles disappear nicely though - they're still not random though, and they're far too dense (but maybe that is an effect bug, and there are just a lot on top of each other?)
[23:04:06] <Avenger> the sparkles are not pixel exact yet ;)
[23:04:14] <fuzzie> they are MUCH better
[23:04:28] <fuzzie> now they just look funny, no big problem :)
[23:04:32] <pupnik_> nice
[23:04:37] <fuzzie> so it is great
[23:04:39] <Avenger> yes, but they need lots of adjustment to be pixel exact
[23:08:01] <fuzzie> well, it is great anyway
[23:08:32] <fuzzie> bg2 really starts coming together
[23:11:31] <Avenger> you didn't believe it, did you :)
[23:12:16] <Avenger> there are still some nasty bugs, but far less
[23:12:44] <Avenger> btw, knock worked earlier
[23:13:08] <Avenger> though, not exactly like in the original
[23:13:10] <fuzzie> yes, knock has some hack :)
[23:13:28] <fuzzie> it still breaks sometimes though
[23:14:10] <fuzzie> i just wanted to know whether knock was hard-coded
[23:14:16] <fuzzie> or whether you could really target anything in the original
[23:14:51] <fuzzie> but i tried it, and you can really target anything, so it is all projectile bugs
[23:15:14] <fuzzie> and it turns out that, yes, Projectile::GetTarget() happily just returns caster if it has no target :)
[23:16:41] <fuzzie> so i think we just add global ids for containers, it won't find a target actor and will give it, it'll all magically work
[23:17:10] <fuzzie> and now i go sleep
[23:17:42] <Avenger> yes, that will work
[23:54:24] <Maighstir> Fixed some stuff in IWD chargen: http://pastebin.com/hyHYCRps - when I chose Dwarf I got Gnome in-game and vice versa, paperdoll default colours were wrong, and unlike the original, GemRB forced you to use all attribute points
[23:54:54] <Maighstir> if something looks strange, feel free to yell at me
[23:55:34] <-- Avenger has left IRC (Quit: bye!)
[23:55:52] <-- deepinthewoods has left IRC (Quit: Leaving)
[23:57:29] <pupnik_> even works pretty good at 1024x768! wooohoo!