#gemrb@irc.freenode.net logs for 25 Dec 2012 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:06:21] <-- rocket_hamster has left IRC (Quit: BuhBye!)
[00:06:40] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[01:21:20] <-- edheldil_ has left IRC (Ping timeout: 255 seconds)
[01:43:58] <Gekz> holy shity
[01:44:02] <Gekz> shit*
[01:44:09] <Gekz> DrMcCoy: that's a pretty well done engine reimplementation
[01:44:13] <Gekz> just watching the video from 2011.
[02:05:04] <-- CamDawg has left IRC (Quit: Leaving.)
[04:41:05] --> brada has joined #gemrb
[04:43:29] <-- kida_laptop has left IRC (Remote host closed the connection)
[04:59:58] <-- brada has left IRC (Quit: brada)
[06:30:27] --> brada has joined #gemrb
[07:19:27] <-- brada has left IRC (Quit: brada)
[08:35:50] --> avenger has joined #gemrb
[08:35:50] --- ChanServ gives channel operator status to avenger
[08:39:15] <avenger> morning
[08:40:53] <fireglow> morning, avenger
[08:41:06] <DrMcCoy> Møøning
[08:41:57] <DrMcCoy> Gekz: All just smoke and mirrors, there's no real gameplay yet :P
[09:14:28] --> edheldil_ has joined #gemrb
[09:14:53] <-- edheldil_ has left IRC (Client Quit)
[09:15:00] --> edheldil_ has joined #gemrb
[09:16:08] <edheldil_> DrMcCoy: sure, but still in a more advanced state than I was expecting aswell
[09:20:22] <avenger> yeah, it renders something in 3d :)
[09:21:15] <avenger> it doesn't render objects, right?
[09:21:24] <avenger> that place seemed empty
[09:21:44] <avenger> hah, now there are objects
[09:33:06] <DrMcCoy> No, objects (placeables and doors, at least) work fully. Creatures work
[09:33:10] <DrMcCoy> mostly too
[09:33:33] <DrMcCoy> Even animated a bit (default breathing animations)
[09:35:06] <DrMcCoy> The video doesn't show animations yet, though. Or "p"-type creatures (with each body part in a separate model), only their heads on the ground, I think
[09:35:31] <edheldil_> the lack of shadows iswhat makes it look empty, I think
[09:36:25] <DrMcCoy> Well, there isn't even a proper light system yet. It's all just flatly shaded
[09:36:49] <fuzzie> did you still not attract an opengler?
[09:38:13] <DrMcCoy> Well, there was a guy who said he's interested, but currently has no time. That was back in October, and he never said another peep
[09:39:14] <fuzzie> well, that is consistent with having no time :-p
[09:39:41] <DrMcCoy> Yes :/
[09:40:57] <DrMcCoy> avenger: p-Type creatures: http://drmccoy.de/zeugs/NWN_20120422T113139.png , http://drmccoy.de/zeugs/NWN_20120422T113352.png
[09:42:57] <DrMcCoy> I also experimented with lighting (can't find the screenshot right now). Looked good, but ran seriously slow, because I have no idea how to best do that. In the end, I basically tested each polygon against the nearest 8 lights
[09:43:44] <fuzzie> what does nwn do? :p
[09:44:07] <DrMcCoy> It also stopped working in certain outdoor sets, like a desert, with each tile having one light above, so that 9 lights are in the vincinity of each polygon. You could see the tile boundaries then :/
[09:44:30] <DrMcCoy> fuzzie: I have no clue. I didn't actually look that deep in the disasm :P
[09:44:41] <avenger> i'm pretty sure i saw tile boundaries in nwn :D
[09:44:50] <DrMcCoy> But not that extreme
[09:45:35] <fuzzie> ( http://drmccoy.de/eos/nwn/20110610T230251.png btw)
[09:46:20] <DrMcCoy> Ah, yes
[09:46:34] <fuzzie> there's a comment on the relevant post which implies that they were just using opengl's fixed-function lighting with a bunch of hacks
[09:47:25] <DrMcCoy> Also, this one shows the tile boundaries a bit: http://drmccoy.de/eos/nwn/20110610T220908.png , it was worse in the desert, though
[09:48:04] <DrMcCoy> Well, I am also doing fixed-function lighting with a bunch of hacks :P
[09:48:22] <fuzzie> you make me wish I had the time :/
[09:49:02] <fuzzie> lighting stuff not committed?
[09:49:47] <DrMcCoy> I put the lights in a list sortable by distance to arbitrary points, then I resorted them several times during rendering and switched the nearest lights on. I tried that for each tile, each plane in a model and each polygon, neither on them looked right completely
[09:49:59] <DrMcCoy> Yeah, I didn't push that to github
[09:50:13] <fuzzie> it looks pretty good anyway
[09:50:19] <fuzzie> from the screenshots
[09:50:42] <DrMcCoy> For some values of good
[09:56:25] --> lynxlynxlynx has joined #gemrb
[09:56:25] <-- lynxlynxlynx has left IRC (Changing host)
[09:56:25] --> lynxlynxlynx has joined #gemrb
[09:56:25] --- ChanServ gives channel operator status to lynxlynxlynx
[09:56:57] <fuzzie> if you expect perfection then you will, of course, never be happy :)
[09:58:19] <DrMcCoy> Story of my life
[09:59:56] <fuzzie> my copy of kotor is still sitting around waiting for some engine love in any case. :p
[10:00:03] <fuzzie> maybe I should someday try installing it
[10:01:07] <DrMcCoy> :)
[10:01:36] <fuzzie> our projectiles are really kinda too fast (in gemrb)
[10:01:59] <DrMcCoy> Okay, rebased my old lighting branch ontop the new master. Let's see if it still compiles and works :P
[10:04:44] <DrMcCoy> Of course, I have no idea what I fiddled with last. Might just be I completely broke it then :P
[10:04:52] <lynxlynxlynx> fuzzie: yeah, in the original you could move and even arrows clearly arced chasing you
[10:05:57] <fuzzie> and gosh we leak anims a lot more than we used to. :)
[10:12:11] <fuzzie> and I can still talk to the barrels in iwd2 :)
[10:15:34] <fuzzie> it doesn't work that badly though
[10:16:27] <lynxlynxlynx> i reinstalled my wine copy yesterday, but it unfortunately doesn't want to read the older saves
[10:16:28] <avenger> meh fuzzie, you promised one day you will fix the bouncy lightning bolt :D
[10:16:44] <lynxlynxlynx> avenger: question for you
[10:16:53] <avenger> ouch, you got me :D
[10:17:12] <lynxlynxlynx> can you check what those few unknowns at the end of the itmv2 header are for? they weren't there in previous versions
[10:17:25] <avenger> ?
[10:17:28] <lynxlynxlynx> iesdp and ielister still have them as unknown
[10:17:39] <lynxlynxlynx> 0070h Equip count 0001
[10:17:39] <lynxlynxlynx> 0072h UNKNOWN 00000000
[10:17:39] <lynxlynxlynx> 0076h UNKNOWN 00000000
[10:17:39] <lynxlynxlynx> 007ah UNKNOWN 00000000
[10:17:39] <lynxlynxlynx> 007eh UNKNOWN 00000000
[10:17:55] <avenger> is this iwd2?
[10:17:57] <lynxlynxlynx> v1 ended at 70
[10:17:58] <lynxlynxlynx> yes
[10:18:11] <avenger> iwd2 has 16 empty bytes i think
[10:18:37] <lynxlynxlynx> 16 is right
[10:18:42] <lynxlynxlynx> you sure they're always empty?
[10:19:15] <avenger> 99.9% sure
[10:19:36] <avenger> i'm 100% sure i didn't see other than 0 there
[10:20:23] <avenger> ielister has this: if(iwd2) { for(i=0;i<4;i++) { os.AddDword("UNKNOWN"); } }
[10:21:32] <avenger> there is a similar empty block in the area header, btw
[10:21:51] <avenger> and the store header also has an empty block, but i think one dword is used there
[10:22:08] <avenger> totally useless, though, it is for capacity
[10:22:41] <fuzzie> heh :)
[10:22:49] <lynxlynxlynx> too bad
[10:22:59] <lynxlynxlynx> i guess critical ranges were hardcoded aswell then
[10:23:23] <lynxlynxlynx> we already have a separate table for some of the multipliers
[10:23:34] <avenger> yea, critical ranges are hardcoded
[10:23:39] <avenger> by itemtype
[10:23:55] <avenger> do you want a list from ida?
[10:24:33] <avenger> i thought i already wrote them down somewhere, i'm sure i saw the code for it
[10:25:15] <lynxlynxlynx> sure
[10:25:33] <lynxlynxlynx> we have the armor penalties and crit multipliers in itemdata.2da
[10:25:54] <avenger> ah crit multipliers
[10:25:58] <avenger> yeah, that is in itemdata
[10:26:04] <lynxlynxlynx> i hope there aren't so few item types, since the list is pretty short
[10:26:11] <avenger> in my unique undocumented compressed format :D
[10:26:35] <avenger> crit ranges are not there
[10:26:52] <avenger> i think that was the point where i couldn't squeeze everything in a single column
[10:27:07] <DrMcCoy> fuzzie: https://github.com/DrMcCoy/xoreos/tree/light (beware,
[10:27:12] <DrMcCoy> might be ugly)
[10:27:33] <lynxlynxlynx> well the max value is always 20, so only the minimum is needed
[10:27:59] <lynxlynxlynx> the table has double data already, so please don't make it any uglier
[10:28:50] <avenger> hehe
[10:29:20] <fuzzie> DrMcCoy: switchOnOff with a boolean parameter is a somewhat confusing name :-p
[10:30:01] <-- avenger has left IRC (Remote host closed the connection)
[10:31:19] <DrMcCoy> Yes, Yes :P
[10:31:37] <fuzzie> also, ok, yes, if you re-do the lights for every vertex then I can see why this would be slow :-p
[10:33:28] <DrMcCoy> :P
[10:34:28] <fuzzie> the idea seems pretty sound though
[10:34:48] <fuzzie> I mean, if you're trying to replicate what those crazy Bioware folks did.
[10:35:25] <DrMcCoy> Well, I want it to look good
[10:35:42] <DrMcCoy> In which case, I'm screwed :P
[10:36:52] <fuzzie> I like how even the configure script is really really slow for the boost bits.
[10:37:43] <fuzzie> also, foxpro.lo?!
[10:39:01] <DrMcCoy> Yes! For reading FoxPro databases!
[10:39:48] <DrMcCoy> NWN stores its campaign databases (variables that carry between .mod files that a grouped in a campaign) in such a database
[10:40:28] <DrMcCoy> .cdx, .dbf and .fpt
[10:41:10] <fuzzie> hm, I wonder if I can download the nwn data from bioware's site somehow
[10:42:16] <fuzzie> I assume not, since EA ate it.
[10:42:21] <DrMcCoy> You could download update packs in the past, dunno if they're still up (they pulled the whole NWN stuff a year? two years? ago because of "omfg hackers!" and never put them up again)
[10:42:34] <fuzzie> originally you could download the whole thing, with a valid cd key
[10:42:52] <DrMcCoy> Ah, right, yes, the linux base package
[10:42:53] <fuzzie> since they couldn't get the data out of the windows installer or something, despite having promised a linux binary years in advance :-p
[10:43:23] <fuzzie> unfortunately I don't seem to have NWN CD 2 here, which is a bit inconvenient.
[10:44:17] <DrMcCoy> :/
[10:48:08] <edheldil_> I remember the promised linux version :(
[10:48:31] <edheldil_> and months of "We are working on how mouse wheel works"
[10:49:00] <DrMcCoy> :/
[10:49:35] <edheldil_> in the end, it worked great, but they never bothered to release the construction kit
[10:50:25] <DrMcCoy> When NWN first came out, I still had a Windows partition. And when I kicked that out a year or so later, the Linux version was working okay
[10:50:54] <DrMcCoy> Or at least I don't remember any problems, beside the fiddlyness of it all
[10:51:39] <edheldil_> What was working okay? Linux version of the game or the construction kit?
[10:52:03] <DrMcCoy> Well, and no movies playing. There was a hack a bit later that hooked into the binary and called mplayer, and you had to reencode the Bink videos yourself using wine, the RADGameTools program and mencoder :P
[10:52:16] <DrMcCoy> The Linux game. Yeah, the editor never worked great with wine
[10:52:24] <DrMcCoy> Lots of access violations all the time
[10:52:42] <DrMcCoy> No native binary either, because the editor uses MFC shit
[10:53:03] <edheldil_> hmm, but it worked ... you just had to have a special version of wine. I haven't constructed anything big with it, though
[10:53:18] <DrMcCoy> Yes, but even that crashed all the time
[10:53:38] <DrMcCoy> You had to save constantly to avoid losing progress
[10:54:15] <edheldil_> I used to play on Arkhalia server here, with my friend. Some of the outdoorlocations were breathtaking
[10:54:51] <DrMcCoy> And I remember cursing out neveredit, a reimplementation effort of the toolset in Python. Never worked great and Python packaging problems in Debian didn't help :P
[10:55:52] <DrMcCoy> afk, need to take a shower :P
[11:01:13] <edheldil_> e.g.: http://www.eowyn.cz/gallery/games/nwn/maps/stromovy_haj/NWN0002.jpg . Good luck lighting this :)
[11:09:43] --> Avenger has joined #gemrb
[11:09:43] --- ChanServ gives channel operator status to Avenger
[11:11:01] <Avenger> lynx, maybe combatinfo class could be dropped into the scriptable subdir?
[11:12:21] <lynxlynxlynx> perhaps
[11:12:28] <lynxlynxlynx> think it's a better place?
[11:12:42] <Avenger> well, it is used only by actor.cpp like pcstats
[11:13:06] <lynxlynxlynx> true
[11:13:21] <lynxlynxlynx> i'll move it later, have to finish the tohit stuff first
[11:13:36] <DrMcCoy> edheldil_: :P
[11:13:38] <Avenger> no problem :)
[11:13:44] <lynxlynxlynx> judging from the manual, all minimum critical ranges are easy to add
[11:13:50] <lynxlynxlynx> all except for large swords
[11:13:52] <edheldil_> :D
[11:14:13] <lynxlynxlynx> they have 19, while scimitars - also large swords, have 18 without any extra effects attached
[11:14:50] <lynxlynxlynx> http://paste.debian.net/218983/ <-- feat wise
[11:16:29] <lynxlynxlynx> and scimitars have it at 18 for sure, since people raved about that gleefully (you could get them to a 16-20 range)
[11:24:38] <Avenger> why is it difficult for large swords??
[11:24:51] <Avenger> it is based on itemtype
[11:25:19] <lynxlynxlynx> scimitars are also large swords
[11:25:20] <Avenger> i'm pretty sure there is only itemtype difference
[11:25:40] <lynxlynxlynx> i've even checked a few items of each and they do map to the same type
[11:26:29] <Avenger> hmm
[11:27:34] <Avenger> and when you equipped a plain scimitar and a plain sword?
[11:27:50] <Avenger> did you see a difference?
[11:28:24] <Avenger> the only difference i see is that the longsword has the breakable flag set :D
[11:28:35] <Avenger> but i doubt iwd2 uses that field
[11:28:54] <lynxlynxlynx> no, i was checking with dltcept
[11:29:08] <Avenger> me too
[11:29:29] <lynxlynxlynx> the web is full of scimitars, so i'm a bit suspicious everybody would get it wrong
[11:29:43] <Avenger> the scimitar of acid is 'keen'
[11:29:45] <lynxlynxlynx> i didn't see any debug prints that would show the range
[11:29:52] <Avenger> what does 'keen' do?
[11:30:00] <lynxlynxlynx> exactly what needed
[11:30:03] <Avenger> isn't it to increase the critical range
[11:30:23] <Avenger> scimitar of acid +2 is keen, normal scimitar isn't
[11:30:42] <Avenger> so... the test would be to see if the plain scimitar is also increasing range or not, i doubt it does
[11:31:06] <lynxlynxlynx> it does, it also has the keen bit set
[11:31:26] <Avenger> really?
[11:31:27] <edheldil_> then it's clear
[11:31:28] <lynxlynxlynx> i thought you could have a keenx2 scimitar
[11:31:44] <lynxlynxlynx> there's still two points to get to 16, but i'll investigate
[11:31:51] <Avenger> i looked at a plain one, it has no keen flag here
[11:32:00] <Avenger> 00CWSWDN.itm
[11:32:01] <edheldil_> lynxlynxlynx: with a feat
[11:32:20] <lynxlynxlynx> 00swdc01
[11:32:32] <lynxlynxlynx> improved criticals?
[11:32:40] <Avenger> sure that is keen here as well
[11:33:17] <Avenger> so, there is no item specific range increase, just the keen bit and the feat?
[11:33:20] <lynxlynxlynx> yeah, improved critical
[11:33:30] <lynxlynxlynx> 1 bonus left
[11:35:33] <lynxlynxlynx> A keen scimitar used with improved crit would give the wielder an impressibe crit range from 16-20. <-- meh
[11:37:11] <edheldil_> hmm, was that keen scimitar in IE or in DnD?
[11:37:47] --> rocket_hamster has joined #gemrb
[11:38:02] <lynxlynxlynx> iwd2 fixpack: All scimitars in the game benefit from keen effect (x3 critical).
[11:38:19] <lynxlynxlynx> man are people confused
[11:40:50] <Avenger> keen scimitar is: 18-20 (x2)
[11:41:08] <Avenger> without improved crits
[11:41:49] <lynxlynxlynx> i remember, i think keen doubled the gap
[11:41:53] <lynxlynxlynx> that would explain it
[11:42:15] <lynxlynxlynx> how did you determine that?
[11:42:22] <Avenger> i can test that
[11:42:26] <Avenger> i load the game :D
[11:42:31] <Avenger> and looked at stats
[11:42:55] <Avenger> now turned off the keen flag
[11:42:58] <Avenger> on the same weapon
[11:43:10] <Avenger> 19-20 (x2)
[11:43:26] <Avenger> a staff is only 20 (x2)
[11:43:56] <Avenger> conclusion: itemtype specific range
[11:44:10] <edheldil_> keen scimitar +improved crit should have 12-20
[11:44:23] <Avenger> no way :)
[11:44:38] <Avenger> i don't even see how it would be 16-20
[11:44:39] <edheldil_> yes
[11:44:56] <lynxlynxlynx> oh, in the guirec display?
[11:44:59] <Avenger> yes
[11:45:19] <lynxlynxlynx> edheldil_: that's only in true 2ed
[11:45:20] <edheldil_> scimitar : 18-20, keen x2, but keen +improved crit=x3
[11:45:21] <lynxlynxlynx> *3ed
[11:45:35] <edheldil_> sure
[11:45:37] <Avenger> ed x3 is for damage, not threat range
[11:45:38] <lynxlynxlynx> Avenger: what happens if you add the feat?
[11:46:31] <edheldil_> no, threat range, see here: http://www.enworld.org/forum/showthread.php?7696-Keen-scimitar
[11:46:56] <edheldil_> keen + feat result only in x3, not x4
[11:47:54] <Avenger> 18-20 (x2)
[11:48:08] <Avenger> but now i have not keen scimitar
[11:48:20] <Avenger> making it keen will result in 17-20
[11:49:28] <Avenger> yes 17-20
[11:52:07] <lynxlynxlynx> ok
[11:52:28] <lynxlynxlynx> odd reports then
[11:52:57] <lynxlynxlynx> thanks
[12:00:01] <Avenger> --> Increased Threat Range: Sometimes a threat range is greater than 20. That is, a threat can be scored on alower number. Longswords, for instance, give threat on a natural attack roll of 19 or 20.
[12:00:14] <wjp> which d&d edition are we talking about?
[12:00:23] <Avenger> icewind dale II ;P
[12:00:36] <wjp> is that some bastardization of 3rd ed, then? :-)
[12:00:42] <Avenger> yes
[12:00:46] <lynxlynxlynx> yep
[12:01:04] <wjp> in D&D3 18-20 x2 would become 15-20 x2 when keen
[12:01:37] <Avenger> yeah, apparently iwd only increase it by one when keen
[12:04:44] <Avenger> ok looking for improved critical usages in the engine
[12:05:46] <lynxlynxlynx> Boring beetle - Attacks Melliana
[12:05:52] <lynxlynxlynx> Pick pocket successful
[12:05:54] <lynxlynxlynx> :D
[12:06:31] <lynxlynxlynx> Avenger: in gemrb we only need to plug in one place, everything else is already set (ok, except exposure on the gui side)
[12:33:16] <Avenger> you know there is a stat for critical range?
[12:34:34] <Avenger> executioner eyes spell improves critical range
[12:39:06] <Avenger> two handed flag affects critical range/multiplier
[12:43:58] --> Yoshimo has joined #gemrb
[12:44:10] <Avenger> haha, i see iwd2 had some weapon styles. But now the place where it determines if you have a style always returns 0
[12:45:09] <edheldil_> least effort hack? :)
[12:45:24] <edheldil_> who would guess there are such things in ie :)
[12:49:24] <Avenger> it's fun because i can use that stump to find combat specific code :)
[12:50:00] <fuzzie> funny, you'd think with the amount of optimization they did, that would have been optimized out :p
[12:50:22] <lynxlynxlynx> that's just the usual critical bonus stat, same as in the other games
[12:53:19] <Avenger> i still don't see what makes quarterstaff 20 and greatsword 19-20
[12:53:46] --> CamDawg has joined #gemrb
[13:06:40] <lynxlynxlynx> 20 is the default
[13:06:48] <lynxlynxlynx> greatswords are hardcoded
[13:32:23] <-- Yoshimo has left IRC (Quit: Yoshimo)
[14:06:16] <-- edheldil_ has left IRC (Ping timeout: 250 seconds)
[14:48:16] <Avenger> lynx: i meant, i don't see the 'hard code' :D
[14:48:54] <lynxlynxlynx> ideally it would be in some other unknown field
[14:49:42] <lynxlynxlynx> in the end we found that all non-20's are 19, so only one bit somewhere akin to keen is needed
[14:49:59] <Avenger> i doubt that
[14:56:08] <Avenger> cool i found it
[14:56:18] <Avenger> actually, i already named that function
[14:56:39] <Avenger> it is as i thought, itemtype specific
[14:59:53] <lynxlynxlynx> http://paste.debian.net/219008/ <-- here are our named itemtypes vs min crit range
[15:00:13] <lynxlynxlynx> that's just judging from the manual
[15:03:50] <Avenger> I found this: 16 DAGGER 19 SMALLSWORD 20 BIGSWORD 23 FLAIL 27 CROSSBOW 31 BOLT 57 GREATSWORD 69 BASTARDSWORD
[15:04:34] <Avenger> this is where the iwd2 engine increases the threat range by 1
[15:13:08] <lynxlynxlynx> cool, it matches
[15:13:18] <lynxlynxlynx> daggers are small blades prof-wise
[15:13:55] <lynxlynxlynx> don't know why bolts are there though
[15:14:27] <lynxlynxlynx> maybe they realised they weren't looking at the launcher later on
[15:19:55] <Avenger> lol
[15:21:54] --> kida_laptop has joined #gemrb
[15:47:34] --> brada has joined #gemrb
[15:49:18] <lynxlynxlynx> Avenger: are you doing anything with those types or should i?
[15:49:38] <Avenger> just do what you can
[15:51:51] <lynxlynxlynx> ok
[15:51:57] <lynxlynxlynx> i'll add another column to that table
[16:17:13] <-- CamDawg has left IRC (Quit: Leaving.)
[16:18:18] <lynxlynxlynx> i hate table loading
[16:18:28] <lynxlynxlynx> no matter how we do it, it looks ugly
[17:21:03] --> edheldil_ has joined #gemrb
[17:29:35] --> CamDawg has joined #gemrb
[17:45:46] <-- CamDawg has left IRC (Read error: Connection reset by peer)
[17:46:25] --> CamDawg has joined #gemrb
[17:46:44] <-- CamDawg has left #gemrb
[19:26:04] <-- rocket_hamster has left IRC (Ping timeout: 246 seconds)
[20:53:50] <lynxlynxlynx> AreaAnimation is the same as an ambient?
[20:55:34] <lynxlynxlynx> nah, can't be
[21:01:45] <fuzzie> ambient is just sound, right?
[21:03:35] <lynxlynxlynx> yeah
[21:03:58] <fuzzie> so, no :)
[21:04:02] <lynxlynxlynx> we don't handle (de)activation properly in iwd2
[21:04:13] <lynxlynxlynx> still looking around
[21:04:31] <lynxlynxlynx> Deactivate("sound_portal") <-- plain call like this
[21:04:43] <lynxlynxlynx> and we end up doing nothing, since it doesn't have an animation attached
[21:08:27] <lynxlynxlynx> yeah, clearly separate from the portal anim
[21:08:45] <lynxlynxlynx> i'll delay that error
[21:30:18] <lynxlynxlynx> http://sprunge.us/FAGC?diff <-- ack?
[21:48:53] <lynxlynxlynx> great, now i hit a cutscene halt
[21:50:36] <lynxlynxlynx> hah, CanSee
[21:53:43] <lynxlynxlynx> phew, just the line of sight check fails
[22:36:47] --> CamDawg has joined #gemrb
[22:37:01] <fuzzie> diff seems fine
[22:37:17] <fuzzie> i would suggest comment w/testcase though
[22:38:43] <lynxlynxlynx> i see we have a soundactivate that does just that, but it's unused in iwd2
[22:39:14] <lynxlynxlynx> bg2 is full of it on the other hand
[22:39:39] <fuzzie> it seems harmless to change in both though
[22:39:52] <fuzzie> well, hopefully harmless :)
[22:40:19] <fuzzie> assuming iwd2 does use it for anims ofc
[22:40:42] <fuzzie> (i.e. we shoulsn't be mapping iwd2's activate to soundactivate)
[22:40:58] <lynxlynxlynx> i agree about that
[22:41:18] <lynxlynxlynx> anyway, i've looked further and it's odd how the cutscene works
[22:42:03] <fuzzie> unfortunately not at home now
[22:42:07] <lynxlynxlynx> but it all ends up breaking there when the dialog is supposed to start :s
[22:42:29] <lynxlynxlynx> i'm still digging, but unfortunately my save doesn't load in the original anymore
[22:42:32] <fuzzie> i've had issues before with such failures
[22:42:56] <fuzzie> but usually just due to pathfinding failure in crowded spaces I think
[22:44:37] <fuzzie> don't know if that could possibly be your problem. but i guess there are likely to be various possibilities in iwd2, they changed quite a lot.
[22:46:12] <lynxlynxlynx> no, it fails the LOS check
[22:46:34] <lynxlynxlynx> looking at the script though, i don't see how we manage to move from the portal to that place though
[22:46:43] <lynxlynxlynx> it's needed, but i don't see the code
[22:48:17] <lynxlynxlynx> the block increases both the dialog and visual range, but BeginDialog is called with BD_CHECKDIST, so they don't play a role
[22:49:33] <lynxlynxlynx> due to that flag we call CanSee explicitly with range disabled and that takes care of the visual one, while dialog range is handled later and doesn't get a chance
[22:53:44] <lynxlynxlynx> http://paste.debian.net/219113/ <-- this is the block
[23:07:59] <fuzzie> No idea what might be suspicious. :/
[23:08:22] <lynxlynxlynx> ok, disregard the movement bit, it was my own oversight
[23:08:45] <lynxlynxlynx> judging from the edge of the explored area, we're not far off
[23:08:57] <lynxlynxlynx> if i continue the cutscene later, all goes fine
[23:09:01] <fuzzie> Is the LOS just way too short?
[23:09:26] <lynxlynxlynx> not *way*, let's say 100px
[23:10:27] <lynxlynxlynx> or less
[23:11:42] <fuzzie> So the likely culprit is the visual range? I forget the units, honestly. but I assume larger than pixels.
[23:12:02] <lynxlynxlynx> like i said, it doesn't even get checked
[23:14:16] <fuzzie> Well, assuming that is a gemrb bug. I wouldn't be surprised if we do the checks wrong, at let for owd2/
[23:14:24] <fuzzie> least for iwd2.
[23:14:52] <fuzzie> Just wondering where might be worth a look if this is still a puzzle when I get home tomorroq. :)
[23:16:15] <lynxlynxlynx> disabling BD_CHECKDIST makes it work, but the actors are not visible
[23:16:26] <lynxlynxlynx> it gives a nice cue though
[23:16:40] <lynxlynxlynx> it looks like they move in too slowly
[23:17:35] <fuzzie> The movement speed is often hardcoded in anims..
[23:17:39] <lynxlynxlynx> but that could also be a thing from after the dialog, since they use JumpToPoint to move
[23:17:52] <lynxlynxlynx> yeah
[23:18:05] <lynxlynxlynx> they jump from the portal and that's fine
[23:18:15] <lynxlynxlynx> at that point they should be in range to eavesdrop
[23:18:34] <lynxlynxlynx> the dialog itself is not for the protagonist, they only realise you're there later
[23:19:17] <lynxlynxlynx> http://lparchive.org/Icewind-Dale-2/Update%2032/ <-- after RUN
[23:19:36] <fuzzie> Ah.
[23:20:24] <lynxlynxlynx> those screenshots look like they have the whole area without fog
[23:20:37] <lynxlynxlynx> not that it would matter in this case
[23:21:24] <lynxlynxlynx> no, there is some fog
[23:22:49] <lynxlynxlynx> odd stuff
[23:32:13] <lynxlynxlynx> the origin of that false is from 2007
[23:36:21] <lynxlynxlynx> toggling that, the visual range check fails, but significantly - 1619 > 50*15
[23:36:28] <lynxlynxlynx> double