#gemrb@irc.freenode.net logs for 11 Oct 2009 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:05:32] --> raevol has joined #GemRB
[00:36:42] --- barra_away is now known as barra_home
[00:42:51] <-- Maighstir has left IRC (Read error: 145 (Connection timed out))
[00:45:08] <Gekz> umm
[00:45:17] <Gekz> fuzzie: someone highlighted me about 10 hours ago
[00:45:21] <Gekz> and my backlog is gone
[00:45:22] <Gekz> lol
[00:58:01] <-- barra_home has left IRC ("Verlassend")
[03:24:11] <-- Forgetful_Lion has left IRC (Read error: 110 (Connection timed out))
[03:32:10] --> Forgetful_Lion has joined #GemRB
[04:07:32] --> [1]Forgetful_Lio has joined #GemRB
[04:09:33] <-- Forgetful_Lion has left IRC (Read error: 145 (Connection timed out))
[04:09:34] --- [1]Forgetful_Lio is now known as Forgetful_Lion
[06:44:54] <-- Gekz has left IRC (Read error: 110 (Connection timed out))
[06:45:47] --> Gekz has joined #GemRB
[07:21:45] --> lynxlynxlynx has joined #gemrb
[07:21:45] --- ChanServ gives channel operator status to lynxlynxlynx
[08:05:20] --> spike411 has joined #GemRB
[08:05:27] <-- spike411 has left #GemRB ()
[08:18:55] --> xrogaan has joined #gemrb
[08:53:00] --> Avenger has joined #gemrb
[08:53:02] --- ChanServ gives channel operator status to Avenger
[09:10:56] <CIA-66> gemrb: 03avenger_teambg * r7404 10/gemrb/trunk/gemrb/plugins/ (6 files in 2 dirs):
[09:10:56] <CIA-66> gemrb: fixed weird cases of projectile types while hopefully didn't break anything that worked before
[09:10:56] <CIA-66> gemrb: removed redundant field (ProjectileType)
[09:15:52] <CIA-66> gemrb: 03avenger_teambg * r7405 10/gemrb/trunk/gemrb/plugins/Core/ (Actor.cpp CharAnimations.cpp): RangedType could be only: 0,1,2 (use 2-sling when in doubt)
[09:22:33] <Avenger> lol, someone wrote this in a gemrb review: works like a charm, runs the game as smoothly as the original Black Isle engine...
[09:26:11] <Gekz> lol
[09:26:17] <Gekz> are they insane
[09:33:02] <Avenger> yep, i guess we still a year away from 'black isle smoothness'
[09:39:58] <Avenger> wjp, if you are here: how r7398 changed anything? size-1>0 is the same as size>=2, no?
[09:45:24] <fuzzie> no
[09:45:48] <fuzzie> size-1>0 breaks if size is 0
[09:45:59] <fuzzie> because 0xffffffff>0
[09:48:15] <fuzzie> and i guess that DLTCEP problem was the TIS in the wrong place, hoorah :)
[09:53:30] <Avenger> oh hehe, size is size_t
[09:53:41] <Avenger> ok
[09:53:49] <fuzzie> you fixed erik?
[09:54:03] <Avenger> yes, he is doing fine
[09:54:13] <Avenger> shooting harmless arrows nicely
[09:54:21] <fuzzie> great :)
[09:54:30] <Avenger> nordom is a bit weird, though
[09:54:41] <Avenger> but i think it was always like that
[09:55:13] <Avenger> i changed a bit more than hacking ProjectileQualifier, though. I removed all traces of ProjectileType
[09:56:11] <Avenger> ProjectileQualifier is: 0,1,2,4. Projectiletype is 0,1,2,3,40,100. (the first 3 values are the same, 4=3, and 40,100 are for throwing axe/spear)
[09:56:26] <Avenger> so those could be the same as 3
[09:56:53] <Avenger> RangedType could be 0,1,2 which is ProjectileType-1, (0 is not projectile)
[09:57:39] <Avenger> RangedType was crashy before, i guess, with axes/spears, because ProjectileType-1 was well over 2
[09:59:13] <Avenger> Nordom's xbow lacks any ProjectileQualifier/Type, but his arrow has the ProjectileType set. Same for Erik's arrows
[09:59:45] <Avenger> And testing in bg1/bg2 showed, that projectiles themselves use the ProjectileType, and ignore the qualifier. Even if it is set in almost all of them
[10:00:08] <fuzzie> ah
[10:01:04] <Avenger> The only truly messy thing is nordom xbow, but in that game the only projectile is the xbow
[10:01:10] <Avenger> so they didn't need any qualifier
[10:01:35] <Avenger> we already had that hack, if there are no qualifiers set, it is assumed an xbow
[10:11:57] --> barra_library has joined #gemrb
[10:12:24] --- barra_library is now known as barra_hpme
[10:12:34] --- barra_hpme is now known as barra_home
[10:26:29] <-- Avenger has left IRC ("bye!")
[11:34:35] <fuzzie> hmm, Erik still not working properly
[11:44:27] <-- xrogaan has left IRC ("Why ?")
[11:45:00] <fuzzie> come baaack, Avenger
[11:47:44] <fuzzie> or anyone else who knows how the slot code works
[12:00:54] <fuzzie> SLOT_MELEE is 35, SLOT_MAGIC is 34, SLOT_MELEE-SLOT_MAGIC=1, so you can never equip anything else once Equipped==1
[12:02:16] <fuzzie> maybe Inventory.cpp:1330 should be SLOT_MAGIC-SLOT_MELEE? but then i am stuck waiting for Avenger to ask :)
[12:08:10] --> Avenger has joined #gemrb
[12:08:17] --- ChanServ gives channel operator status to Avenger
[12:08:19] <CIA-66> gemrb: 03fuzzie * r7406 10/gemrb/trunk/gemrb/plugins/Core/Inventory.cpp: use the right slot (i think?) when checking for magic weapons
[12:08:36] <Avenger> i heard someone is calling for me?
[12:08:47] <fuzzie> well, is that commit correct? :-)
[12:09:19] <Avenger> yes
[12:09:40] <Avenger> SLOT_MELEE should be subtracted from the real slot number to get the Equipped value
[12:09:58] <fuzzie> ok :)
[12:10:14] <fuzzie> Erik correctly re-equips the bow when I run away from melee, with that change
[12:10:15] <Avenger> i really hate that hack, it is causing lots of problems. Bioware stuff :)
[12:11:11] <Avenger> i really don't understand why they had to subtract SLOT_MELEE from it
[12:11:21] <fuzzie> i am confused by everything in the inventory code
[12:11:44] <Avenger> maybe the best would have been to convert that value when importing/exporting
[12:12:21] <Avenger> still confused?
[12:12:32] <fuzzie> yes, but by other things :)
[12:12:43] <fuzzie> did you find anything on how pst implem,ents stances?
[12:12:54] <Avenger> yes, and i sent it to you too :)
[12:13:05] <fuzzie> oh, heh, i didn't look for it in there :)
[12:13:09] <fuzzie> i will do so after lunch
[12:13:38] <Avenger> it is in pst_code\charanimation\objanimation_pst.txt
[12:13:57] <Avenger> not everything is annotated, or even listed. But it gives some insight
[12:14:19] <Avenger> start from the end of that file
[12:56:58] --> Maighstir has joined #gemrb
[13:13:11] <-- Forgetful_Lion has left IRC (" HydraIRC -> http://www.hydrairc.com <-")
[13:20:50] --> barra_away has joined #gemrb
[13:25:22] <-- barra_home has left IRC (Read error: 60 (Operation timed out))
[13:28:25] <fuzzie> what's with this "Couldn't create animationfactory: mor6de (e850)" stuff in iwd?
[13:28:33] <fuzzie> it's kind of game-breaking right from the start
[13:28:47] <fuzzie> "mor3a4e (e820)" too
[13:29:09] <fuzzie> and "mor3a4 (e820)"
[13:33:28] <lynxlynxlynx> why is it game breaking?
[13:33:38] <fuzzie> because i can't see any enemies there at all
[13:34:02] <lynxlynxlynx> don't worry, your archers can :P
[13:34:04] <fuzzie> i guess if you have high-level chars that's no problem, but my low levels kind of die
[13:34:26] <lynxlynxlynx> let me refresh my clone
[13:34:57] <fuzzie> the cause seems to have been Avenger removing a bunch of the lines?
[13:35:10] <fuzzie> if i revert that then they appear
[13:36:04] <fuzzie> don't know if it's coincidence
[13:36:07] <Avenger> what lines
[13:36:17] <fuzzie> Avenger: in how/avatars.2da
[13:36:24] <lynxlynxlynx> well, how and iwd have different entries for mor6
[13:36:37] <lynxlynxlynx> gemrb/override/how/avatars.2da:337:0xE850 MOR6 MOR6 MOR6 MOR6 9 2
[13:36:39] <lynxlynxlynx> gemrb/override/iwd/avatars.2da:331:0xE808 MOR6 MOR6 MOR6 MOR6 9 2
[13:36:58] <lynxlynxlynx> iirc the falling is done backwards, so 20 doesn't go to 50
[13:36:58] <Avenger> i can tell you which is the correct
[13:37:10] <lynxlynxlynx> oh wait, that's mor3
[13:37:23] <fuzzie> the E808 line was in how/avatars.2da, but got removed by r7380, same for a bunch of other lines
[13:37:34] <fuzzie> i don't know anything about how this works, just noticing..
[13:37:55] <fuzzie> works fine with iwd/avatars.2da
[13:38:05] <Avenger> i removed, or intended to remove lines that are in iwd2
[13:39:12] <Avenger> because, after i changed to the interval based animation lookup, the extra animation entries based on iwd2 would 'poison' the interval
[13:39:57] <Avenger> and the animation ids rarely use the first, 'round' entry. They usually have an ignored last nibble
[13:40:41] <Avenger> so all extra entries coming from iwd2 or bg2 should be removed
[13:40:48] <Avenger> hmm, i think it is more like bg2
[13:42:10] <Avenger> hm is there MOR6?
[13:42:56] <lynxlynxlynx> no
[13:42:56] <Avenger> it seems to me that it is e810 - mor1, e820 - mor2, e830-mor3, e840-mor4, and e850-mor5
[13:43:01] <lynxlynxlynx> only up to mor5
[13:43:08] <lynxlynxlynx> :)
[13:43:10] <Avenger> this is in the code
[13:43:22] <Avenger> the last digit is ignored in the match
[13:43:47] <Avenger> so e808 was intentionally removed, because it would poison the e800-e80f interval
[13:44:08] <Avenger> e850 should be mor5
[13:44:18] <fuzzie> so what should e800 be?
[13:44:38] <Avenger> hmm no entry for that
[13:44:59] <fuzzie> well, it would be nice if one of you who know how it work could fix the 2da, anyway :)
[13:45:17] <Avenger> why what animation you got anyway?
[13:45:24] <Avenger> how shouldn't have e808
[13:45:45] <fuzzie> well, it works when i revert that entire commit, i don't know why
[13:45:47] <Avenger> ahh, actually it could be put back since it is an unused interval
[13:46:03] <lynxlynxlynx> how doesn't have it, it's iwd that still does
[13:46:42] <fuzzie> if all of those E8xx lines are bad, then that might explain all the problems?
[13:46:44] <Avenger> i didn't change iwd, mostly because i read only how code
[13:47:22] <Avenger> oh the lines are bad
[13:47:46] <Avenger> this should be the correct way-->it seems to me that it is e810 - mor1, e820 - mor2, e830-mor3, e840-mor4, and e850-mor5
[13:48:19] <fuzzie> the other columns need changing too?
[13:48:28] <Avenger> yes, apparently
[13:48:39] <Avenger> there is no e800-e80f
[13:49:03] <Avenger> i wonder how it worked before
[13:49:44] <Avenger> this was the right entry-->0xE818 MOR1 MOR1 MOR1 MOR1 9 2 1 *
[13:49:53] <Avenger> but it should lose its last digit to 0
[13:49:59] <Avenger> instead i removed the line, hehe
[13:50:08] <Avenger> because there was already an e810 line
[13:50:13] <Avenger> but that was the wrong line
[13:51:02] <Avenger> so if you change all with a matching e8x0 = morx, it will work
[13:51:04] <fuzzie> bg1's avatars seem to work really well, apart from the xvart thing
[13:52:55] <Avenger> btw, the lookup can only work from the highest entry counting down. As long as the entry is > than the searched value, i go lower. As soon as it is <= i know i found it.
[13:53:52] <Avenger> that's why i need to order the entries now
[13:54:37] <Avenger> do you understand how it works?
[13:55:27] <fuzzie> i guess bg1's MXVT entry is type 2 (IE_ANI_FOUR_FILES) but the bam is type IE_AI_FOUR_FILES_2?
[13:55:48] <Avenger> i don't know
[13:56:11] <fuzzie> the mxvtg2 bams have 16 cycles..
[13:56:41] <Avenger> mxvt has 4 files, 2 normal, 2 east
[13:56:58] <fuzzie> and with type 2, gemrb tries using too-high cycles
[13:57:23] <Avenger> one cycle has 8 frames (3 are in the east files)
[13:57:41] <Avenger> so it is 5+3 style
[13:58:58] <fuzzie> 5+3 style + 4 files + 16 cycles per file = IE_ANI_FOUR_FILES_2?
[13:59:08] <Avenger> i don't know
[13:59:12] <Avenger> i didn't look at it yet
[13:59:15] <fuzzie> ok :)
[13:59:37] <fuzzie> i don't dare change how's avatars.2da, i don't understand all the columns, i poke through the code now
[13:59:42] <Avenger> IE_ANI_FOUR_FILES_2: Like IE_ANI_FOUR_FILES but with only 16 cycles per frame.
[13:59:48] <Avenger> we have 8 cycles per frame
[14:00:05] <Avenger> err, 8 frames per cycle...
[14:00:10] <Avenger> hmm
[14:00:25] <Avenger> i don't know why the comment says cycles/frame :)
[14:00:37] <Avenger> when cycles are larger units than frames
[14:00:49] <fuzzie> i think it just means "but with only 16 cycles"
[14:00:53] <fuzzie> but i guess you wrote it?
[14:01:12] <Avenger> i think it is more like frames/cycle, see the multiplier ?
[14:02:15] <fuzzie> well, all i know is: IE_ANI_FOUR_FILES tries using cycles above 16 in 'g2' bams, IE_ANI_FOUR_FILES_2 stays below 16 cycles in 'g2' bams
[14:02:17] <Avenger> heh odd
[14:02:30] <fuzzie> and i want the second one, because the xvart only has 16 cycles :)
[14:03:02] <Avenger> both has orient/2
[14:03:13] <fuzzie> IE_ANI_FOUR_FILES has '16 + Orient/2'
[14:03:15] <Avenger> that means 8 directions per stance. (or 8 frames per cycle) :)
[14:03:25] <fuzzie> so that doesn't work at all
[14:03:34] <fuzzie> same for METT..
[14:03:59] <Avenger> well, try it, if it works, then use it, hehe
[14:04:12] <Avenger> i cannot say anything against a working animatoin
[14:05:21] <Avenger> i agree that IE_ANI_FOUR_FILES_2 stays below 16 :) it has only 2 stances in g2
[14:05:30] <fuzzie> maybe i should write an automated tool which checks all of these
[14:05:57] <Avenger> mxvt has 2 attack like stances in g2
[14:06:12] <Avenger> yep, could be attack / shoot
[14:06:44] <Avenger> they are more like attack1/attack2
[14:07:12] <Avenger> dltcep has some automation built in
[14:07:17] <Avenger> 'generate avatars.2da'
[14:07:38] <Avenger> it also checks an existing 2da, and suggests some values
[14:07:42] <fuzzie> it says 'anisnd not found, sorry'?
[14:07:43] <Avenger> but it is not perfect
[14:08:00] <Avenger> anisnd, hehe, so it works in bg2 only
[14:08:14] <Avenger> you could steal one from bg2?
[14:09:14] --> Gekz_ has joined #GemRB
[14:10:03] <Avenger> if you got an existing avatars.2da. copy it to bg1, and 'check/avatars.2da'
[14:10:14] <Avenger> that would not require anisnd
[14:10:20] <Avenger> and would perform a check
[14:10:27] <Avenger> but it is really not finished :)
[14:11:21] <fuzzie> it produces '0x7C00 MXVT MXVT MXVT MXVT 3 2 0', so IE_ANI_TWO_FILES
[14:12:47] <-- Gekz has left IRC (Read error: 60 (Operation timed out))
[14:14:13] <Avenger> odd
[14:14:21] <Avenger> maybe it is not correct :)
[14:14:43] <Avenger> there are definitely 4 files
[14:15:27] <Avenger> two files is only g1/g1e
[14:15:30] <Avenger> a pity
[14:15:32] <Avenger> so much to fix
[14:18:33] <Avenger> interesting, my code intends to look for g2/g2e and if those exist it should choose ie_ani_four_files
[14:19:20] <fuzzie> but it doesn't check the cycles?
[14:19:53] <fuzzie> Avenger: my pst_code/charanimation is empty?
[14:20:07] <Avenger> it doesn't look for ie_ani_four_files_2
[14:20:12] <Avenger> empty?
[14:20:17] <fuzzie> yes, that zip only has the empty directory in it
[14:20:37] <Avenger> meh, i must have compressed the stuff before i copied it in
[14:21:01] <Avenger> at the time i sent it i had it there so i assumed it is in :)
[14:21:42] <Avenger> but you can still look at the HoW code
[14:21:59] <Avenger> if you need support to fix the 0xe8../MORx stuff
[14:23:12] <fuzzie> yes, i see the '0xe810-0xe850' in there
[14:24:48] <Avenger> ok, do you know how many cycles are in g1/g2 in a genuine four_files/four_files_2 scheme?
[14:25:06] <Avenger> that means i need 2x2 numbers
[14:25:40] <Avenger> if you don't know that, do you know 2 animations that are correct?
[14:26:02] <fuzzie> well, mxvt has 48 in both g1 files and 16 in both g2 files
[14:26:23] <Avenger> but mxvt is not a genuine one :) it is the one under scrutiny :)
[14:26:39] <Avenger> i could make the check matching it, but it would ruin the experiment
[14:26:41] <fuzzie> well, i have no idea which are genuine :)
[14:27:17] <Avenger> one that works and already set as their representative type
[14:27:46] <fuzzie> my notes say i see also 48 in g1 and 24 in g2, for a real four_files one, but i don't know which one i found for that
[14:28:10] <Avenger> looks like mdog is fine?
[14:28:19] <Avenger> but i never seen one
[14:28:43] <fuzzie> well, i worry that maybe they just don't use the broken animation in-game when i saw them
[14:28:56] <Avenger> mdog is 48/16
[14:29:15] <Avenger> mdog is set as type = 3
[14:29:22] <Avenger> that is four_files...
[14:29:28] <Avenger> hmm not four_files_2
[14:29:37] <Avenger> so mdog is buggy too?
[14:29:43] <fuzzie> "#define IE_ANI_TWO_FILES 3", no?
[14:29:57] <Avenger> that sucks
[14:30:00] <Avenger> mdog has 4 files as well
[14:30:02] <Avenger> like mxvt
[14:30:05] <Avenger> MEEH
[14:31:22] <Avenger> we really need someone who is more orderly than I, with a bit chart around, matching every anims
[14:31:31] <Avenger> bit->big
[14:32:20] <Avenger> i try to fix dltcep to actually KNOW about the animation schemes
[14:32:33] <Avenger> it doesn't know 13/14
[14:32:45] <Avenger> so no wonder it couldn't diagnose mxvt
[14:33:47] <Avenger> well, you said 48/16 for four_files_2, lets assume that's correct
[14:33:56] <Avenger> now i just need the numbers for the other scheme
[14:37:28] <Avenger> four_files is 48/24, MSIR is my reference
[14:39:04] <-- Gekz_ has left IRC (Connection timed out)
[14:50:54] <Avenger> ok, i taught dltcep to know animation type 14 (four_files_2)
[14:52:16] <Avenger> and it lists: mogh, mogn, mber, mdog, mett, mghl, mgib, mspi, mzom, mlic???, mumb???
[14:53:02] <Avenger> something isn't right yet, mlic has no east animations
[14:53:36] <Avenger> heh and mxvt isn't on the list
[14:53:59] <Avenger> ah it is there, and suggested as 14 too
[14:54:37] <Avenger> mlic and mumb were suggested as 4, not 14. hehe
[14:54:43] <Avenger> that is probably correct
[15:01:53] * fuzzie peers at MGCL
[15:09:01] <Avenger> i think it was hasty to add animation type 15
[15:09:07] <Avenger> i guess 4 is simply buggy
[15:09:09] <Avenger> no?
[15:09:51] <fuzzie> that is my thought
[15:13:20] <Avenger> what is the difference? dltcep knows only the resource names, and that seem to be the same for 4 and 15
[15:13:38] <Avenger> going simply by resource names, it is type 4
[15:14:26] <fuzzie> is there a difference?
[15:15:01] <Avenger> IE_ANI_CODE_MIRROR_3: Almost identical to IE_ANI_CODE_MIRROR_2, but with fewer cycles in g26
[15:15:02] <fuzzie> sorry, i look at my own things, but it seems that type 4 is trying to use cycles that don't exist anywhere
[15:15:09] <Avenger> this is the code
[15:15:20] <fuzzie> maybe i just look at bad examples
[15:15:33] <Avenger> well, i guess, i will add the cycle checker just as i did before with four_files
[15:15:38] <Avenger> then i could find them easily
[15:15:44] <Avenger> i just need the cycle counts for all files :)
[15:15:50] <Avenger> and it is a lot, here ;)
[15:16:05] <fuzzie> but i look at type 4 and gemrb tries non-existant cycles for all the examples i tried..
[15:16:30] <Avenger> well, lets see one example, and list its cycles
[15:16:37] <Avenger> i just need 13 numbers :)
[15:16:37] <fuzzie> and it seems that this IE_ANI_CODE_MIRROR_3 gets it right
[15:17:08] <fuzzie> MGCL and friends, for example
[15:17:24] <Avenger> ok i look at mgcl
[15:17:38] <Avenger> 54 in g1
[15:19:25] <Avenger> 54 in g1*, and 63 in g2* ?
[15:20:21] <Avenger> yep seems like that. there is g1, g11-15, g2, g21-g26
[15:20:23] <fuzzie> it seems to be
[15:20:41] <Avenger> i will use this as reference for 15
[15:20:53] <Avenger> lets see if there is any remaining
[15:20:56] <Avenger> that will be 4
[15:21:03] <Avenger> if none remains, then we can move 15 to 4
[15:21:44] <fuzzie> the thing is
[15:22:02] <fuzzie> gemrb's IE_ANI_CODE_MIRROR_2 picks cycles which exist
[15:22:06] <fuzzie> but which are simply blank in the data files
[15:22:36] <fuzzie> for instance, for IE_ANI_ATTACK_JAB it tries cycle 45 in g26, and this is blank in everything i look at.
[15:23:04] <fuzzie> so i think if there is a difference, it is probably not detectable via cycle counts?
[15:23:24] <Avenger> 4 is simply wrong
[15:23:31] <Avenger> i found meae and meas only
[15:23:34] <Avenger> try those?
[15:24:07] <fuzzie> meaeg26 has forward-facing cycles only?
[15:24:50] <Avenger> that seems utterly bullshit :)
[15:25:08] <fuzzie> i guess it does have one at 45..
[15:25:09] <Avenger> is there any valuable cycle in g26 in other anims?
[15:25:13] <fuzzie> but what on earth?
[15:25:39] <fuzzie> sure, type 15 uses valuable anims in g26
[15:26:35] <Avenger> they are valuable only if they don't exist elsewhere
[15:26:50] <fuzzie> hm, well, i don't see well enough to see if they're different :)
[15:28:17] <Avenger> hmm, ok, move every 4's to 15 then, except meae and meas ?
[15:28:35] <Avenger> Invalid animation MGLC (0x7F07) for type 4.
[15:28:37] <Avenger> Invalid animation MSNK (0x7F13) for type 4.
[15:28:39] <Avenger> Invalid animation AMOO (0x7F16) for type 4.
[15:28:43] <Avenger> these are also not fitting
[15:28:57] <Avenger> (i used type 4 as our type 15 check)
[15:29:02] <fuzzie> i don't think it works even if i move to 15
[15:29:23] <Avenger> well, meae won't work
[15:29:27] <Avenger> but most others will
[15:29:52] <fuzzie> MGCL has the first 8 cycles empty, then the next 9 cycles good to mirror
[15:30:16] <lynxlynxlynx> ooh, nice job
[15:30:56] <fuzzie> and the SixteenToNine conversion doesn't help there at all
[15:31:21] <Avenger> amoo has 54/55 cycles, instead of 54/63
[15:31:22] <fuzzie> eg, it adds 18 for IE_ANI_ATTACK_JAB, but the cycle is actually at 27
[15:31:53] <Avenger> well, there are different schemes :)
[15:32:00] <fuzzie> so i don't see where all these numbers come from
[15:32:01] <Avenger> maybe it worked for one, exactly one
[15:32:11] <fuzzie> neither AddVHR3Suffix nor AddVHR2Suffix make sense for anything i look at
[15:32:12] <Avenger> we first need to map all schemes
[15:32:16] <Avenger> big chart :)
[15:32:34] <Avenger> and it seems it isn't enough to have resource names, we need at least cycle counts
[15:32:41] <Avenger> and lets hope it is enough
[15:32:53] <fuzzie> but at least we can fix MXVT and etc to 14, if you made dltcep check that?
[15:33:02] <Avenger> yes i did
[15:33:20] <Avenger> but i don't want to stop here
[15:33:38] <Avenger> most of these exist in bg1 too
[15:34:25] <lynxlynxlynx> cycle counts sounds enough to me, all of the issues i've hit were either missing cycles, missing files (suffixes) or missing avatar table rows
[15:34:55] <Avenger> i don't know how amoo exists
[15:35:01] <Avenger> it has resources with 55 cycles
[15:35:10] <Avenger> that doesn't compute :(
[15:36:34] <fuzzie> it's 6 cycles + 36 is wasted
[15:37:06] <Avenger> no
[15:37:30] <fuzzie> cycle #36 is identical to #37, and otherwise it's 6 sets of 9 orientations
[15:37:32] <Avenger> it has 9 cycles per stance
[15:37:34] <fuzzie> sorry, i don't mean 'cycles'
[15:37:52] <fuzzie> but you see that it makes sense if you ignore #36?
[15:38:27] <Avenger> BUT WHY!!!! this is so insane
[15:38:37] <fuzzie> yes
[15:38:41] <Avenger> i see what you say, but why on earth they made this???
[15:39:05] <Avenger> i really feel like getting a big hammer and smash the computer
[15:39:42] <lynxlynxlynx> di^Wbovine intervention?
[15:40:01] <Avenger> this sucks
[15:40:08] <Avenger> and g2 is surely used
[15:40:35] <Avenger> besides all of g2* are the same
[15:40:44] <Avenger> this is royally screwed up
[15:41:36] <Avenger> we probably will need a scheme just for the moose?
[15:42:47] <Avenger> look at msnk
[15:42:49] <fuzzie> there's nothing else with 55 cycles?
[15:43:05] <Avenger> msnkg2 has the first stance in g2, then all other cycles are empty
[15:43:18] <lynxlynxlynx> maybe the pst ghost type could be used
[15:43:28] <lynxlynxlynx> the one we have for mmist in bg2
[15:43:50] <Avenger> i think they don't use lots of these cycles
[15:43:57] <Avenger> so we could have the same schemes
[15:44:08] <Avenger> we just have to find which cycles they actually use
[15:45:05] <Avenger> msnkg26 has a single useful cycle :)
[15:45:37] <fuzzie> 'useful' is pushing it for any of those msnk cycles, i think
[15:46:01] <fuzzie> i guess "pushing it" is terribly colloquial, i mean: i think not really 'useful' :)
[15:46:19] <Avenger> yep, the whole resource is crap
[15:46:37] <Avenger> if we are using strong terms :P
[15:47:09] <Avenger> i would really be happy if we don't need g26 then
[15:47:16] <fuzzie> so i guess lynx added this VHR3 one
[15:47:37] <fuzzie> lynxlynxlynx: any idea where you got the numbers from? some specific animatio?
[15:48:32] <lynxlynxlynx> what's the name of the type?
[15:48:42] <fuzzie> the MIRROR_3 one
[15:49:19] <lynxlynxlynx> which avatars use it?
[15:49:43] <lynxlynxlynx> ah
[15:49:45] <fuzzie> METT for bg2?
[15:49:51] <Avenger> IE_ANI_CODE_MIRROR_3: Almost identical to IE_ANI_CODE_MIRROR_2, but with fewer cycles in g26
[15:49:53] <lynxlynxlynx> msps apparently
[15:49:59] <lynxlynxlynx> sword spider i think
[15:49:59] <fuzzie> but not for bg1, which confuses me
[15:50:15] <Avenger> so, if it is about g26, then to the hell with it, we should simply eliminate all g26 references :)
[15:50:23] <fuzzie> hm, no, i look up wrong number
[15:50:29] <Avenger> msnk has a single cycle in it, which makes it useless completely
[15:50:39] <fuzzie> Avenger: but msnk seems useless for all cycles
[15:50:40] <lynxlynxlynx> only bg2 uses that type in our table
[15:50:52] <Avenger> g26, at least, no?
[15:51:32] <Avenger> g1 has 9-17 filled
[15:52:06] <fuzzie> MSPS is another of these files which only has forward-facing anims in g26
[15:52:09] <Avenger> g11: 0-8
[15:52:53] <Avenger> well, lets find an animation which uses something from g26 ;)
[15:53:08] <Avenger> and then lets see if it could be replaced
[15:53:19] <Avenger> if g26 is not used, then happiness
[15:54:08] <Avenger> lynx: only bg2 uses it because you didn't change the other tables :)
[15:54:23] <Avenger> these anims are shared among engine versions
[15:54:25] <lynxlynxlynx> yes, i only change things when i find problems
[15:56:29] <Avenger> msnk is really nicely cleaned up, almost all cycles are empty
[15:56:52] <Avenger> the only problem is i cannot tell which one is a casting animation ;)
[16:00:31] <Avenger> lynx, i think mirror_3 was not needed, i don't find any animation which needs g26
[16:01:14] --> kettuz has joined #gemrb
[16:01:46] <Avenger> the moose is still bad, though, because all of its g2* resources got fewer cycles than expected
[16:04:06] <Avenger> mglc has only 62 cycles in its g2* files
[16:04:28] <Avenger> i beleive they just stuffed crap cycles in it :)
[16:08:06] <Avenger> i really wonder how these animations work at all
[16:09:02] <lynxlynxlynx> so all mirror_2 avatars just have a crappy/random g26?
[16:09:12] <Avenger> i don't know
[16:09:42] <Avenger> if g26 contains casting, or such, it could be that it is unused for something like a snake
[16:10:02] <Avenger> we just need to find the real cycles used for walking
[16:10:34] <Avenger> i look at mglc and it is really ugly too
[16:11:02] <Avenger> even if an animation has a good starting frame, it could be damaged
[16:12:33] <Avenger> i found only 2 stances for mgcl, one is a kind of twitching, the other is an attack, i didn't even find a walking anim yet
[16:12:53] <Avenger> found it: g11
[16:14:30] <Avenger> i don't think we need any g2*
[16:14:54] <Avenger> maybe g2: 0-8 for attack
[16:15:57] <Avenger> g21: 9-17: another attack form
[16:27:47] <Avenger> ok, i simply remove g26 from the needed resources, it is crap everywhere
[16:29:08] <fuzzie> even in iwd2+pst?
[16:34:09] <Avenger> lynx added the new resource because of the attack jab animation was empty, i think
[16:34:26] <Avenger> case IE_ANI_ATTACK_JAB:
[16:34:28] <Avenger> strcat( ResRef, "g26" );
[16:34:30] <Avenger> Cycle+=18;
[16:34:31] <Avenger> break;
[16:35:05] <Avenger> but it is empty even for animations that seem to have a well developed g26, like mkuo
[16:35:40] <Avenger> g26 seems to contain only casting graphics
[16:36:51] <fuzzie> well, the only meaningful thing in g26 seems to be that :)
[16:37:19] <Avenger> they seem to use equal cycle numbers in all gn* resources, but use only selected cycles
[16:37:28] <Avenger> i guess this is to spare with memory
[16:42:00] <Avenger> if there is no casting graphic, all anims over 45 or maybe 36, are crap
[16:42:02] <Avenger> just filler
[16:42:12] <Avenger> i guess the moose would crash if you try to cast with it
[16:42:57] <Avenger> this is why some g2* files have crippled cycle counts
[16:43:10] <Avenger> but we shouldn't have a new animation scheme just to support that
[16:43:18] <Avenger> they won't cast in the game anyway :)
[16:44:11] <Avenger> i just need to find case IE_ANI_ATTACK_JAB:
[16:44:48] <Avenger> but it could be this animation scheme has no 3 different attacks
[16:45:07] * fuzzie peers at pst
[16:45:15] <Avenger> pst is different :)
[16:45:24] <Avenger> you always flutter from game to game, heh
[16:45:33] <Avenger> i cannot follow
[16:45:37] <fuzzie> 0x40 in our avatars.2da is DHUMB but i guess the HUMB is the real magic there?
[16:46:06] <fuzzie> the 'D' bit is just "stand and look stupid animation"
[16:46:46] <fuzzie> ah yes you document this :) cool
[16:46:57] <Avenger> no
[16:47:04] <Avenger> it is d<stance>humb
[16:47:19] <Avenger> like dat1humb
[16:47:46] <Avenger> well i guess you found the docs
[16:50:31] <fuzzie> i was hoping some of this might apply to pst, is all :(
[16:50:37] <fuzzie> i didn't realise it was all different
[16:51:32] <fuzzie> looks like it'd be really easy to fix up pst though, just needs some lookup table for weird pst stances
[16:51:37] <fuzzie> ok, dinnertime
[16:52:17] <Avenger> oh well, i will mail you the missing file then :)
[16:52:20] <-- Avenger has left IRC ("ChatZilla 0.9.85 [Firefox 3.5.3/20090824101458]")
[16:59:47] --> D_T_G has joined #gemrb
[17:06:29] <D_T_G> wow, i didn't know there are so many graphics from bg1 left in iwd1 resources
[18:08:58] <-- kettuz has left IRC ("Leaving")
[18:32:03] <CIA-66> gemrb: 03avenger_teambg * r7407 10/gemrb/trunk/gemrb/override/ (how/avatars.2da iwd/avatars.2da): fixed iwd orc animations (mor6/7 exist in iwd2)
[18:52:35] --> Avenger has joined #gemrb
[18:52:37] --- ChanServ gives channel operator status to Avenger
[18:55:23] <D_T_G> i have a guiscript question
[18:55:59] <Avenger> yep?
[18:56:08] <D_T_G> if window: window.unload() creates an error can't unload window if i click some back buttons
[18:56:31] <D_T_G> i see lots of such blocks in code
[18:56:53] <D_T_G> and it all works, but i don't see what is done to work all the time
[18:57:36] <Avenger> was that window opened normally?
[18:57:57] <D_T_G> yeah, normally
[18:58:06] <D_T_G> i think so
[18:58:26] <Avenger> after you unload it, do this: window = None
[18:58:53] <D_T_G> i work on implementing this: http://img98.imageshack.us/img98/817/howu.jpg
[18:59:11] <Avenger> window.unload() doesn't clear the variable itself
[18:59:24] <D_T_G> i have this so far: http://wklej.org/id/172079/txt
[18:59:42] <Avenger> what's that, network setup?
[18:59:49] <fuzzie> Avenger: full game vs expansion only
[19:00:03] <fuzzie> appears when you click 'create game' in HoW
[19:00:08] <Avenger> ahh, that's important
[19:00:32] <D_T_G> in iwd1 it is like SoA vs ToB starting
[19:00:45] <D_T_G> but chargen is no difference
[19:00:46] <fuzzie> D_T_G: i think Avenger has the trick, you probably need some setting to None..
[19:01:34] <lynxlynxlynx> maybe we should just do it in the wrapper
[19:01:35] <fuzzie> why on earth is there both 'no' and 'cancel' in that warning dialog?
[19:01:56] <D_T_G> bioware did it so
[19:01:58] <fuzzie> d_t_g makes no mistake there, they both do the same thing
[19:02:09] <D_T_G> the do the same in exe
[19:02:14] <fuzzie> yes, i tried in exe too :)
[19:02:18] <fuzzie> i dare not ask
[19:02:28] <D_T_G> no and cancel both just return to the main screen
[19:02:33] <fuzzie> yep
[19:02:55] <D_T_G> i think maybe this window behaves diffrenet when final_save is present
[19:03:32] <fuzzie> that is simple to test..
[19:04:07] <D_T_G> maybe to find final_save on the net :)
[19:04:55] <D_T_G> and the script should also be simpler if HoW is not installed
[19:06:37] <fuzzie> my final save seems to make no difference
[19:07:13] <fuzzie> i cheated to produce it, maybe it doesn't have the right bits
[19:07:23] <D_T_G> but there must be a usage of string 26317 and 26586 from tlk...
[19:08:30] <D_T_G> oh, 26586 is for exporting chars to a file
[19:08:50] <D_T_G> but 26317 clearly says to import a team from finish main game
[19:08:58] <D_T_G> *finished
[19:09:00] <fuzzie> yes, it does
[19:09:51] <D_T_G> hmm
[19:10:20] <D_T_G> what about starting already in lonelywood, would it be hard to implement?
[19:10:25] <fuzzie> nope
[19:10:41] <fuzzie> did you try just setting the expansion, like in bg1/bg2?
[19:11:09] <D_T_G> i don't remember totsc only game in bg1
[19:11:12] <fuzzie> gemrb just checks startare.2da and startpos.2da
[19:11:32] <fuzzie> and the expansion details in those are fine for how
[19:12:25] <D_T_G> but the dialog with hjollder is different from when you transfer from kuldahar and start in lonelywood
[19:12:28] <fuzzie> you just want SetVar("PlayMode", 2) i guess?
[19:12:54] <fuzzie> i would try just running that in your guiscript and seeing what happens
[19:13:02] <fuzzie> set it to 0 for normal non-expansion mode
[19:13:29] <Maighstir> bg1 doesn't have an option to only play the expansion, as it incorporates fully into the original game, instead of adding on after it
[19:14:08] <D_T_G> yeah, there was a very special totsc save after install, but that's all
[19:14:46] <D_T_G> re:HoW - well, i'll try maybe to finish the gui first
[19:14:56] <fuzzie> ok.
[19:15:20] <D_T_G> if i will understand how unloading windows work, maybe i could fix that things in bg2 :)
[19:17:22] <fuzzie> some strings in iwd exe are a bit mad
[19:17:39] <fuzzie> i see "That's no master area, that's my wife!" here in action stuff :)
[19:18:05] <D_T_G> sounds like easter egg :)
[19:19:34] <fuzzie> i see some RETURN_TO_LONELYWOOD, GEM_ACTIVE and ACT_06 strings
[19:20:08] <D_T_G> hmm, probably the first unset 9101_HOBART_INIT and SPAWN
[19:20:08] <fuzzie> and this weird AR2109/HOWSEER stuff i noticed the other day
[19:21:07] <D_T_G> after you finish totl quest Hobart dissapears
[19:21:22] <D_T_G> from whistling gallows i think so
[19:22:33] <fuzzie> and hardcoded 'door2616a'
[19:24:29] <D_T_G> where is 2109?
[19:25:06] <D_T_G> ah, it's the house in kuldahar where hjollder is
[19:25:27] <D_T_G> hm, maybe also it gets set when you return to kuldahar
[19:25:44] <Avenger> that wife stuff is surely an assertion, with some buggy areas
[19:26:06] <fuzzie> Avenger: yes, it just funny :)
[19:26:12] <fuzzie> HOWSEER is some movie, i guess
[19:26:24] <Avenger> try to remove all entries from mastarea.2da, you might get it :)
[19:26:55] <Avenger> howseer is a movie, it also has a 2da
[19:27:14] <Avenger> i guess it is the expansion starting movie
[19:27:24] <Avenger> it is scriptable, probably
[19:27:30] <Avenger> before entergame, you can play it
[19:27:39] <Avenger> if you enter the expansion
[19:27:49] <D_T_G> expansion_only game
[19:28:10] <D_T_G> yeah and normally in kuldahar it's triggered from script
[19:28:21] <D_T_G> because this movie is played in gemrb now
[19:28:34] <fuzzie> well, maybe some other thing to bear in mind, if you write gui
[19:28:46] <Avenger> yep, i just mentioned, if it doesn't activate somehow
[19:29:31] <fuzzie> Avenger: it would be nice if hitting enter in the 'setup name' combobox of DLTCEP closed the dialog
[19:29:54] <fuzzie> or maybe i miss some other quick way to switch between games
[19:29:54] <Avenger> fuzzie, i committed a the avatar update, if you didn't see it
[19:30:03] <D_T_G> must go now, bye!
[19:30:07] <-- D_T_G has left IRC ()
[19:31:00] <Avenger> hmm, you can set up multiple games, then select the wanted setup in the combo box and press the save&back button
[19:31:05] <fuzzie> ok, 'door2616a' is surely some leftover from bg1
[19:31:17] <Avenger> that's fast enough, you want faster?
[19:31:39] <fuzzie> Avenger: it's just instinctive for me to hit enter after i select the wanted setup, i guess :) it surely doesn't matter
[19:31:49] <Avenger> door2616a is in bg1 exe :)
[19:31:54] <fuzzie> yes
[19:32:03] <fuzzie> it is the door where you start in Candlekeep..
[19:32:14] <fuzzie> or, well, the region for that door
[19:32:18] <Avenger> and in bg2 exe too
[19:32:27] <Avenger> hehe, they really kept all crap around
[19:32:38] <fuzzie> i wonder, what on earth is so special about that door?
[19:32:59] <Avenger> maybe we'll never know :D
[19:33:42] <Avenger> so, got your mail about pst stances?
[19:33:47] <fuzzie> yes
[19:34:26] <Avenger> it seems they keep lots of slots around
[19:34:35] <Avenger> even after the translation
[19:35:23] <Avenger> like this: 007B8E91 83 BA 22 05 00 00 00 cmp dword ptr [edx+522h],0
[19:35:44] <Avenger> i didn't map all slots but they are somewhere around 60?
[19:35:51] <Avenger> compared to our ~18
[19:36:38] <fuzzie> i thought animstat.ids was complete, at 78
[19:37:01] <Avenger> yeah but some stances are not using a new slot
[19:37:22] <fuzzie> the jumptable /looks/ completely mapped?
[19:37:24] <Avenger> 7,8,9 for example: assertion
[19:38:06] <Avenger> the jumptable has 48 valuable entries, but there is only one slot for misc1-20
[19:38:30] <Avenger> that is handled by: 007B8C01 83 BC 90 E6 03 00 00 cmp dword ptr [eax+edx*4+3E6h],0 ;;has miscX?
[19:38:37] <Avenger> it takes up 20 slots
[19:39:53] <Avenger> well, hmm 48+20 is 68, so yeah, maybe there are ~60 entries, the maximum so far is 75, i could only remove 7,8,9 with 100% surety
[19:40:52] <Avenger> there are 4 slots for 3 attack slots, i think
[19:41:48] <fuzzie> hm, you're obviously much better at reading this than me :)
[19:41:57] <Avenger> oh hmm no, more than 4
[19:42:42] <fuzzie> oh right, you include the stance conversion table at bottom
[19:42:59] <Avenger> 416,41a, 41e - these are used by jumptable #0, 13,14,15. but #0 is also referenced at least 4 times through the translation table
[19:43:02] <fuzzie> that's helpful
[19:43:03] <Avenger> yeah, i do
[19:43:17] <Avenger> that is a must to read :)
[19:44:11] <fuzzie> first thing i do is reverse that table
[19:44:24] <fuzzie> so, "jumptable 0: stances 0, 9, c, d, e"
[19:45:00] <fuzzie> so that's ATTACK, SHOOT, ATTACK_SLASH, ATTACK_BACKSLASH, ATTACK_JAB
[19:45:01] <Avenger> yeah this is it, i just didn't add any numerical indices: 007B91A0 007B81DC ;;0 attack (shoot, slash, backslash, jab)
[19:45:06] <fuzzie> which makes sense :)
[19:45:46] <Avenger> well, we have 6 slots for this
[19:45:54] <Avenger> 3 melee anims and 3 shooting anims
[19:46:13] <Avenger> pst has either melee chars or shooting chars
[19:46:17] <Avenger> no mix
[19:46:56] <Avenger> so, we have a little room
[19:47:02] <Avenger> but not enough
[19:47:11] <Avenger> probably we need to make that allocation dynamic
[19:47:13] <fuzzie> yes, it gives me a much better understanding as to why you didn't think just increasing MAX_STANCE was a good idea
[19:47:22] <Avenger> yep
[19:47:51] <Avenger> well, max_stance could be increased, but the internal animation slots are not the same
[19:48:03] <Avenger> a stance could resolve to different animations
[19:48:25] <Avenger> and they have a generic 'attack' stance that could be a random jab/slash/backslash
[19:48:37] <Avenger> that is actually in any game
[19:48:49] <Avenger> so that is 4 stances with 3 slots
[19:49:28] <Avenger> we should also check if ANY of those misc slots are ever used
[19:49:38] <Avenger> if not, then to the hell with it :) 20 slots are not needed
[19:50:17] <Avenger> we will still need a nice translation table
[19:51:00] <Avenger> i still don't know why awake/die are using the same slot. Yeah, it is probably just a reversed animation
[19:51:07] <Avenger> but how it is reversed, i don't know
[19:51:16] <fuzzie> the misc slots are used, yes
[19:51:26] <Avenger> oh, s...
[19:51:31] <Avenger> well
[19:51:32] <Avenger> where?
[19:51:33] <fuzzie> but maybe only misc1?
[19:51:41] <fuzzie> just check the scripts :)
[19:51:50] <Avenger> i don't have those super grep functions
[19:51:59] <fuzzie> well, the list of scripts is far too long to say
[19:52:06] <fuzzie> see 0100dab1?
[19:52:19] <fuzzie> baf/pst/1201ci13.baf: PlaySequence(ANIM_MISC2)
[19:52:27] <fuzzie> baf/pst/1201cv13.baf: PlaySequence(ANIM_MISC3)
[19:52:54] <Avenger> i see that is the hammering dabus :)
[19:53:08] <fuzzie> that's as high as is used in scripting, it seems
[19:53:23] <Avenger> ok, so misc3
[19:53:34] <Avenger> well, lets see if this has any trace in resdata
[19:54:07] <Avenger> Misc1=cat1dab
[19:54:10] <Avenger> lol
[19:54:14] <Avenger> it is an attack animation
[19:54:40] <Avenger> misc1=cms1dei (deionarra)
[19:54:46] <Avenger> so we cannot simply map it to attack
[19:55:08] <Avenger> Misc1=dat1god (godsman)
[19:55:26] <Avenger> damned deionarra, she doesn't attack, they could have squeezed it into that slot
[19:55:58] <fuzzie> no chance of her attacking in fortress?
[19:56:05] <Avenger> no, she is just standing around
[19:56:13] <Avenger> attack1=cstddei
[19:56:15] <Avenger> attack2=cstddei
[19:56:51] <Avenger> misc1=dms1nof
[19:56:53] <Avenger> misc2=dms2nof
[19:56:56] <Avenger> this is nameless one
[19:57:15] <fuzzie> http://fuzzie.org/nfs/gemrb/pst_bcs_used_stances.txt
[19:57:20] <Avenger> ok, transcendent one has 5
[19:57:36] <Avenger> checked dialogue too?
[19:57:45] <Avenger> check some transcendent one
[19:57:50] <Avenger> he got 5 misc anims
[19:58:02] <fuzzie> i can't find anything in dialogue
[19:58:27] <Avenger> well, grep for 'isc' in resdata.ini
[19:58:30] <fuzzie> no SetAnimState or PlaySequence calls at all
[19:58:32] <Avenger> fits a screen
[19:58:40] <fuzzie> oh!
[19:58:41] <fuzzie> damn
[19:58:51] <fuzzie> ok, my pst dialogue folder only has dmorte.d in it :(
[19:58:58] <Avenger> LOL
[20:00:02] <Avenger> ok, i think we can safely say we need to support only 5 misc slots, but if we go with dynamic animation slots/translation tables, we can support any number. This is only if we hardcode anything
[20:00:37] <Avenger> so my estimate of ~60 is close :)
[20:02:40] <fuzzie> ok, i set weidu extracting dialog, it will take a while
[20:04:47] <Avenger> doh, even the transcendent one has only 17 animations
[20:05:06] <Avenger> so our hardcoded limit wouldn't be a limit, just need some clever mapping
[20:05:38] <Avenger> nameless has 21
[20:10:35] <Avenger> fuzzie, what do you think: armor=x in resdata.ini is the base armor class bonus?
[20:10:41] <fuzzie> dialog uses MIMEDIE, MIMEGETHIT, MIMEATTACK1, MISC1, MIMESTAND, MIMESTANCETOSTAND, MIMESTANDTOSTANCE
[20:10:54] <fuzzie> Avenger: isn't it sound?
[20:11:16] <Avenger> sound? it is like armor=10 or armor=0
[20:11:24] <Avenger> or any number between 0 and 10
[20:15:38] <fuzzie> yes, it is sound
[20:15:57] <fuzzie> it's combined in some way with 'hitsound' to work out the on-hit sound to play
[20:17:03] <Avenger> oh,
[20:17:18] <Avenger> so it isn't some kind of armor class
[20:17:24] <Avenger> but the 'clink' sound of getting hit?
[20:17:28] <fuzzie> as far as i know
[20:17:37] <fuzzie> my notes say that some PST reverse-engineer agrees
[20:17:58] <fuzzie> ah, there is a whole gibberlings3 thread on this
[20:18:10] <fuzzie> http://forums.gibberlings3.net/index.php?showtopic=1818
[20:18:17] <Avenger> ignus has it as armor=10
[20:20:46] <Avenger> hitsound=IGN068a,IGN068b,IGN068c
[20:20:49] <Avenger> how these two add up
[20:22:24] <Avenger> even worse is there is no ign068* file
[20:25:20] <fuzzie> i think the armor value corresponds to some letter, A-K?
[20:25:29] <fuzzie> see the HIT_* sounds
[20:25:35] <fuzzie> but it has been a long time
[20:30:15] <Avenger> i see
[20:30:29] <Avenger> extremely complicated ;)
[20:33:52] <fuzzie> is it, really?
[20:34:05] <fuzzie> i think it's probably simple
[20:34:16] <Avenger> looks like it is hit_0<1-4><a-k>[1]
[20:34:26] <Avenger> i don't know what is the 1-4
[20:34:30] <Avenger> damage level?
[20:34:53] <Avenger> the final optional 1 is probably the miss sound?
[20:35:20] <fuzzie> the final optional one makes it quieter, certainly
[20:35:44] <fuzzie> maybe the 1-4 is damage type?
[20:36:04] <fuzzie> it sounds like it
[20:36:24] <Avenger> hmm 1-4 is more like damage type
[20:36:38] <Avenger> some sound like hammered, some sound like slashed, with the same number
[20:38:04] <Avenger> hit_02 sounds are probably the hit by blunt sounds?
[20:38:56] <Avenger> well i guess if there is no documentation for this, i should reverse it
[20:39:29] <-- Avenger has left IRC ("bye!")
[20:45:09] --> Avenger has joined #gemrb
[20:45:13] --- ChanServ gives channel operator status to Avenger
[20:45:24] <Avenger> yeah, it is easy, it is in the damage opcode, you got the file too
[20:45:30] <Avenger> just no annotation yet
[20:46:05] <Avenger> the pst damage opcode uses lots of hardcoded spell resrefs
[20:46:09] <Avenger> smells like trouble
[20:46:38] <Avenger> it is especially obsessed with SPWI803
[20:46:53] <fuzzie> Deathbolt?
[20:47:58] <fuzzie> i wonder why
[20:48:11] <fuzzie> maybe to make sure the damage doesn't happen if the fx_death was successful?
[20:48:22] <fuzzie> still weird though
[20:51:41] <Avenger> the armor code is in the subroutine starting with ;;damage animation?
[20:51:45] <Avenger> if you want to take a look
[20:52:03] <Avenger> this is the armor string-->9381F8h
[20:52:36] <Avenger> so far i see it taking the target's animation ID, subtracts 0xe000
[20:52:51] <Avenger> i don't know how that works with 0x6000+x anims
[20:53:28] <Avenger> i would have used and 0x0fff
[20:55:39] <Avenger> then there is an ugly jumptable handling the hit_... resrefs one by one, lol
[20:56:12] <Avenger> no wonder this code is so huge
[20:56:53] <Avenger> but now i see, this is some damage type thingie
[21:02:55] <-- Avenger has left IRC ("ChatZilla 0.9.85 [Firefox 3.5.3/20090824101458]")
[21:13:49] --> D_T_G has joined #gemrb
[21:21:42] <D_T_G> http://wklej.org/id/172244/txt
[21:22:10] <D_T_G> expansion game in start.py
[21:24:19] --> Avenger has joined #gemrb
[21:24:22] --- ChanServ gives channel operator status to Avenger
[21:27:00] <D_T_G> now about starting in lonelywood i should do that in PartyFormation.py:EnterGamePress right?
[21:27:33] <Avenger> yes, i think so
[21:28:25] <Avenger> fuzzie: pst errors out for me: File "./GUIScripts/pst/GUICommonWindows.py", line 604, in CheckLevelUp
[21:28:34] <Avenger> does it work for you?
[21:28:50] <Avenger> it happens when i get xp, probably able to level up, dunno
[21:29:14] <lynxlynxlynx> what's the error?
[21:29:36] <Avenger> probably i should just include CanLevelUp
[21:29:48] <Avenger> GemRB.SetVar ("CheckLevelUp"+str(pc), CanLevelUp (pc))
[21:29:50] <Avenger> NameError: global name 'CanLevelUp' is not defined
[21:30:14] <Avenger> import CanLevelUp from LuCommon ?
[21:30:36] <lynxlynxlynx> sure
[21:31:22] <Avenger> from GUICommon import *
[21:31:24] <Avenger> import GUICommon
[21:31:27] <Avenger> both of these are needed?
[21:31:40] <Avenger> i never learned this stuff :)
[21:32:14] <Avenger> but i guess the second line is > than the first
[21:32:28] <D_T_G> GameSetExpansion (1) does not work for HoW
[21:33:53] <Avenger> Traceback (most recent call last):
[21:33:55] <Avenger> File "./GUIScripts/pst/GUICommonWindows.py", line 605, in CheckLevelUp
[21:33:57] <Avenger> GemRB.SetVar ("CheckLevelUp"+str(pc), CanLevelUp (pc))
[21:33:58] <Avenger> File "./GUIScripts/LUCommon.py", line 40, in CanLevelUp
[21:34:00] <Avenger> Class = ClassTable.FindValue (5, Class)
[21:34:01] <Avenger> AttributeError: 'NoneType' object has no attribute 'FindValue'
[21:34:03] <Avenger> lynx: now i get this :)
[21:34:14] <Avenger> well, definitely better than before, but not perfect
[21:34:31] <Avenger> classtable is not loaded? i wonder why
[21:35:47] <lynxlynxlynx> maybe it's the wrong order
[21:36:05] <lynxlynxlynx> try also importing ClassTable from GUICommon
[21:37:18] <D_T_G> the starting of tob is in CharGen9.py? i don't see anything like checking startpos.2da there...
[21:38:51] <lynxlynxlynx> it is
[21:39:32] <CIA-66> gemrb: 03avenger_teambg * r7408 10/gemrb/trunk/gemrb/GUIScripts/pst/GUICommonWindows.py: import CanLevelUp
[21:40:49] <Avenger> ok, this is my contribution to fuzzie's pst improvement project :)
[21:40:50] <CIA-66> gemrb: 03avenger_teambg * r7409 10/gemrb/trunk/gemrb/plugins/Core/ (Actor.cpp Actor.h): play PST specific weapon hit sounds
[21:40:57] <D_T_G> GemRB.SetVar("oldgame",1) was of the same purpose as my GameType?
[21:41:58] <Avenger> bye!
[21:42:01] <-- Avenger has left IRC ("bye!")
[21:45:47] <lynxlynxlynx> D_T_G: GameType is read-only for guiscripts
[21:47:45] <D_T_G> i think when i click a button with SetVarAssoc ("GameType", x) its get set to x, isn't it?
[21:49:06] <lynxlynxlynx> sure, but don't do that
[21:49:19] <D_T_G> ?
[21:49:38] <lynxlynxlynx> don't touch GameType
[21:50:04] <D_T_G> where?
[21:50:16] <D_T_G> it must be set somewhere
[21:50:25] <lynxlynxlynx> it is set by the core
[21:50:52] <D_T_G> oh, should i change the variable name?
[21:51:15] <lynxlynxlynx> yes
[21:52:00] <D_T_G> ExpansionGame?
[21:52:24] <lynxlynxlynx> why don't you use oldgame?
[21:52:53] <D_T_G> yeah, why 'old'? :)
[21:53:29] <lynxlynxlynx> i didn't name it
[21:54:40] <D_T_G> let me rename it sth more meaningful, 'old' is opposite to 'new' for me, and here it's not about starting new game or loading old game, but about starting new game in easthaven or lonelywood :)
[21:55:07] <lynxlynxlynx> that's how it is used in bg2 too
[21:55:22] <lynxlynxlynx> arguing that a boolean should have the opposite name is silly :P
[21:56:12] <lynxlynxlynx> but it is not really a boolean, i think it is also used for mp, but this is simpler in iwd, so it is probably not needed
[21:56:25] <D_T_G> for me old game is a saved game
[22:03:34] <D_T_G> http://wklej.org/id/172268/txt better?
[22:04:06] <D_T_G> well, without EnterGamePress - none works here
[22:04:08] <lynxlynxlynx> i can look at it later /today/
[22:04:30] <D_T_G> thx, i go sleep, nnight
[22:04:31] <-- lynxlynxlynx has left IRC (Read error: 104 (Connection reset by peer))
[22:04:37] <-- D_T_G has left IRC ()
[22:05:36] <fuzzie> what d_t_g needed was 'PlayType', i assume
[22:05:50] <fuzzie> i'm sure i was clear about this in backscroll