#gemrb@irc.freenode.net logs for 30 Aug 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[01:03:45] <-- edheldil_ has left IRC (Ping timeout: 260 seconds)
[01:24:11] --> edheldil_ has joined #GemRb
[01:31:42] <-- Mataniko has left IRC (Remote host closed the connection)
[01:47:44] <-- edheldil_ has left IRC (Ping timeout: 276 seconds)
[05:48:24] --> raevol has joined #GemRb
[06:55:37] <-- raevol has left IRC (Quit: Leaving.)
[07:16:16] --> lynxlynxlynx has joined #GemRb
[07:16:16] --- ChanServ gives channel operator status to lynxlynxlynx
[07:19:06] --> spike411 has joined #GemRb
[07:19:10] <spike411> Morning
[07:21:05] <lynxlynxlynx> moo
[07:53:56] <fuzzie> hi
[07:54:25] <lynxlynxlynx> i was just about to say that the putactor asserts don't get triggered
[07:54:38] <lynxlynxlynx> i managed a corruption behind their back
[07:55:00] <fuzzie> your log had it complaining while writing a spell resref
[07:56:44] <lynxlynxlynx> yes, that syscall thing
[07:59:48] <fuzzie> which is .. strange
[08:01:48] <lynxlynxlynx> i'll step through it
[08:01:56] <lynxlynxlynx> i have a reliable reproduction path now
[08:02:11] <lynxlynxlynx> and i still think it involves that projected image
[08:02:56] <fuzzie> the spellbook does have a tendency to get corrupted
[08:03:27] <lynxlynxlynx> 41403 insertions(+), 41398 deletions(-) <-- source update! hah
[08:03:53] <fuzzie> dltcep?
[08:04:00] <lynxlynxlynx> yes
[08:04:20] <fuzzie> git is often not so clever with generated files, but that looks like the whole codebase :)
[08:05:04] <lynxlynxlynx> aha
[08:08:46] <fuzzie> i wasn't so clever when messing with spellbook here
[08:13:15] <lynxlynxlynx> so far so good
[08:14:20] <fuzzie> and those asserts are firing
[08:14:59] <fuzzie> grr
[08:20:16] <edheldil> that surely looks like THE update, but rest assured the message could be so much terser :)
[08:21:34] <CIA-26> GemRB: 03fuzzie * r1549a0cbc545 10gemrb/gemrb/ (core/Spellbook.cpp plugins/CREImporter/CREImporter.cpp): add/fix more random asserts/checks
[08:21:48] <fuzzie> those also make no difference to your thing
[08:21:53] <edheldil> hehe, Avenger changed newline chars :)
[08:23:27] <fuzzie> ah. needs more -b!
[08:27:21] <lynxlynxlynx> i've stepped through all the CREImporter::PutKnownSpells except for the nonparty npc that we always load/save
[08:27:40] <lynxlynxlynx> nothing interesting came uo
[08:27:41] <lynxlynxlynx> up
[08:28:36] <lynxlynxlynx> (i was checking the loop bounds)
[08:31:06] <lynxlynxlynx> now the s->size() == level assert triggers on load
[08:33:42] <fuzzie> oh, well, that is positive
[08:34:35] <fuzzie> i added that because i noticed all the callers were happily ignoring the "did this work?" return value
[08:38:50] <fuzzie> any idea what the caller is?
[08:41:06] <lynxlynxlynx> CREImporter::ReadInventory
[08:41:52] <fuzzie> right
[08:41:55] <fuzzie> bad caller
[08:42:43] <fuzzie> this is really stupid code
[08:43:40] <fuzzie> it did work before my early 2009 changes, but without my early 2009 changes, everything crashes.
[08:46:42] <lynxlynxlynx> i'm glad you know what's wrong
[08:48:08] <fuzzie> are you going to be able to test?
[08:48:20] <lynxlynxlynx> sure
[08:53:22] <fuzzie> grr, cmake is really dumb sometimes
[08:54:13] <lynxlynxlynx> it's a build system, it'd be odd otherwise
[08:57:28] <fuzzie> ok, try that. mostly untested.
[08:57:46] <CIA-26> GemRB: 03fuzzie * rcaf8a6af6bf6 10gemrb/gemrb/ (3 files in 2 dirs): try fixing the spellbook memorization API
[08:58:14] <fuzzie> it is still awful, really
[08:58:30] <fuzzie> but i hope it's the minimal-effort change which works..
[09:00:49] <lynxlynxlynx> now it segfaults on the broken save
[09:01:02] <fuzzie> o.O
[09:01:15] <lynxlynxlynx> actually make that any save
[09:01:56] <fuzzie> that should only happen when you exceed NUM_BOOK_TYPES or MAX_SPELL_LEVEL
[09:01:56] <lynxlynxlynx> assert(spl->Number == 0 && spl->Number2 == 0); // unused
[09:02:04] <lynxlynxlynx> spl->Number is inaccessible
[09:02:11] <fuzzie> well, presumably spl is NULL
[09:02:16] <lynxlynxlynx> no
[09:02:27] <fuzzie> but it segfaults?
[09:02:38] <lynxlynxlynx> yes
[09:03:04] <fuzzie> what's the type?
[09:03:12] <lynxlynxlynx> i'll retry
[09:03:41] <lynxlynxlynx> type is also inaccessible
[09:03:57] <fuzzie> what happens when you try?
[09:04:01] <lynxlynxlynx> all the public members are
[09:04:09] <lynxlynxlynx> (gdb) p spl->Type
[09:04:10] <lynxlynxlynx> Cannot access memory at address 0x45544954
[09:04:14] <lynxlynxlynx> (gdb) p spl
[09:04:15] <fuzzie> so 'spl' is corrupt, i guess
[09:04:16] <lynxlynxlynx> $2 = (CRESpellMemorization *) 0x4554494e
[09:04:20] <lynxlynxlynx> yeah
[09:04:21] <fuzzie> just print 'Type'
[09:04:27] <lynxlynxlynx> 0
[09:04:56] <fuzzie> and 'Level'?
[09:04:59] <lynxlynxlynx> 3
[09:06:31] <fuzzie> Spellbook.cpp:470 should probably be size()-1
[09:06:49] <fuzzie> afaict not relevant though
[09:07:04] <fuzzie> oh, no, it should be size().
[09:07:09] <fuzzie> that is very strange.
[09:07:16] <fuzzie> but i have to go to a meeting now :(
[09:07:33] <lynxlynxlynx> ok, have fun :)
[09:42:49] <Lightkey> uh.. ich muss weg! :-D
[09:47:34] --> SiENcE has joined #GemRb
[10:00:44] <-- mattinm has left IRC (Ping timeout: 258 seconds)
[10:13:16] <fuzzie> hi
[10:13:29] --> mattinm has joined #GemRb
[10:23:04] <fuzzie> no more insights?
[10:30:42] <lynxlynxlynx> no, i stopped looking
[10:38:09] <fuzzie> grr
[10:38:19] <lynxlynxlynx> type is 3
[10:39:16] <fuzzie> that is invalid, no?
[10:39:22] <lynxlynxlynx> yeah
[10:39:38] <fuzzie> oh, i forgot to add that assert :(
[10:39:40] <lynxlynxlynx> maybe GetSpellLevelCount doesn't return a nice value
[10:40:14] <lynxlynxlynx> rechecking, i've got a very sweet debugging route
[10:40:34] <lynxlynxlynx> yep
[10:40:41] <lynxlynxlynx> GetSpellLevelCount(type) = 4260014834
[10:40:49] <fuzzie> yes, because type is invalid :P
[10:40:53] <lynxlynxlynx> yes
[10:40:57] <CIA-26> GemRB: 03fuzzie * rf99d5c50863c 10gemrb/gemrb/core/Spellbook.cpp: add a forgotten assert
[10:40:59] <fuzzie> i forgot to add the assert there, my 'git add -i' skills are not so good
[10:41:31] <lynxlynxlynx> well shouldn't it return 0 for invalid types?
[10:42:44] <fuzzie> this is GetMemorizedSpellsCount calling it, i guess?
[10:43:08] <lynxlynxlynx> GetSpellMemorization
[10:43:08] <fuzzie> i'm just adding assert()s in an attempt to find the bad caller
[10:44:40] <fuzzie> i think returning 0 there would be v.dangerous
[10:46:36] <fuzzie> but i don't know what we're meant to do with type 3.
[10:50:37] <CIA-26> GemRB: 03fuzzie * r6f87794c27d7 10gemrb/gemrb/ (core/Spellbook.cpp plugins/CREImporter/CREImporter.cpp): return NULL on invalid spell type, check for this in CREImporter asserts
[10:50:41] <fuzzie> oh, i suck.
[10:53:02] <CIA-26> GemRB: 03fuzzie * r9b91615c0475 10gemrb/gemrb/plugins/CREImporter/CREImporter.cpp: .. and fix the GetSpellMemorization parameter order
[10:53:28] <fuzzie> try that.
[10:53:31] <lynxlynxlynx> hah
[10:54:14] <lynxlynxlynx> loading works again
[10:54:35] <lynxlynxlynx> borked save normally asserts
[10:54:40] <lynxlynxlynx> now retrying to break it
[10:55:38] <lynxlynxlynx> that's successful too :(
[10:58:16] --> barra_library has joined #GemRb
[11:02:26] <fuzzie> v.strange
[11:04:26] <fuzzie> but, well, CopySelf copies all the effects, you get for example the fx_bonus_wizard_spells stuff, i guess?
[11:05:46] <lynxlynxlynx> it's not my image
[11:06:44] <fuzzie> how do you break it?
[11:07:16] <lynxlynxlynx> start the dialog -> contingency including mislead -> dragon breath -> save
[11:07:52] <lynxlynxlynx> one thing i noticed now is that one of the dead actors - not sure if it is the copy, has memorised spells that it doesn't know
[11:08:03] <lynxlynxlynx> Known spells: 0: spwi219 L: 1 T: 1
[11:08:08] <lynxlynxlynx> Memorized spells:
[11:08:10] <lynxlynxlynx> 0: spwi217 1
[11:08:11] <lynxlynxlynx> 1: spwi212 1
[11:08:16] <lynxlynxlynx> 2: spwi219 1 x2
[11:08:55] <lynxlynxlynx> i think it is the original caster, since the loot is there
[11:09:44] <lynxlynxlynx> he's a walking dead man now, probably since the breath made him unconscious after death (at least as far as messages can be trusted)
[11:30:38] <lynxlynxlynx> hmm, it worked this time
[11:30:43] <lynxlynxlynx> i let it run for a long while
[11:31:06] <lynxlynxlynx> the clone probably unsummoned if it wasn't killed properly before
[11:32:07] <fuzzie> no spell-modifying effects on the original caster?
[11:33:11] <lynxlynxlynx> will check
[11:55:48] <lynxlynxlynx> nothing like that
[11:56:50] <lynxlynxlynx> http://pastebin.ca/1928965 <-- in either of them
[11:58:57] <fuzzie> the CastSpellOnCondition is there afterwards?
[12:02:19] <lynxlynxlynx> sure, the fireshield is still active
[12:02:34] <lynxlynxlynx> it can't deliver anything though, i sabotaged the condition to be always false
[12:02:52] <fuzzie> you'd think it would get copied too, but i guess it gets applied after the mislead
[12:03:08] <lynxlynxlynx> yep
[12:08:11] <lynxlynxlynx> well, this is pretty conclusive now
[12:08:42] <lynxlynxlynx> saving before the mislead expires breaks the save, while afterwards it doesn't
[12:09:54] <lynxlynxlynx> and what i forgot to mention (ups) is that also two dangling item warnings are printed, so the inventory is probably screwed up too
[12:12:36] <lynxlynxlynx> yeah, those are ne
[12:12:41] <lynxlynxlynx> w
[12:18:42] <lynxlynxlynx> i guess i'll repeat the putactor step-by-step, but later
[13:15:34] <lynxlynxlynx> hmm
[13:15:46] <lynxlynxlynx> ItemSlotsOffset is a bit big
[13:16:01] <lynxlynxlynx> (gdb) p MemorizedSpellsOffset
[13:16:02] <lynxlynxlynx> $12 = 1200
[13:16:04] <lynxlynxlynx> (gdb) p MemorizedSpellsCount
[13:16:04] <fuzzie> i guess i probably don't check that one
[13:16:05] <lynxlynxlynx> $13 = 26
[13:16:07] <lynxlynxlynx> (gdb) p ItemSlotsOffset
[13:16:09] <lynxlynxlynx> $14 = 10488
[13:16:17] <fuzzie> but ItemSlotsOFfset is after the inventory, right?
[13:16:45] <lynxlynxlynx> let's see
[13:16:49] <fuzzie> yes, so you have to check ItemsOffset and ItemsCount
[13:17:06] <fuzzie> oh, and the effects are in the middle too
[13:18:07] <lynxlynxlynx> ItemsOffset is the same as ItemSlotsOffset, but that makes sense, since there are no slot items
[13:27:02] <-- barra_library has left IRC (Read error: Connection reset by peer)
[13:30:18] <lynxlynxlynx> do we special-case fists also in saving/loading?
[13:30:37] <lynxlynxlynx> ItemsCount is 0, but the fists are there
[13:30:49] <fuzzie> they shouldn't be saved, if that's what you mean
[13:30:55] <lynxlynxlynx> ok
[13:32:39] <fuzzie> so what is the visible issue with the save?
[13:33:56] <lynxlynxlynx> Assertion `spl && spl->Number == 0 && spl->Number2 == 0' failed.
[13:37:52] <lynxlynxlynx> the clone is different in the way we save it - PutActorGemRB vs PutActorBG
[13:41:27] <fuzzie> oh
[13:41:35] <fuzzie> right, PutActorGemRB should never be used
[13:41:49] <fuzzie> so there's your problem, i guess
[13:41:55] <lynxlynxlynx> CREVersion is 0, so maybe it is just a default
[13:42:05] <fuzzie> change CopySelf() to copy 'version' as well
[13:45:15] <lynxlynxlynx> yep :D
[13:45:56] <fuzzie> you win! :)
[13:48:08] <CIA-26> GemRB: 03lynxlupodian * rf96d4ad0d278 10gemrb/gemrb/core/Scriptable/Actor.cpp:
[13:48:08] <CIA-26> GemRB: Actor::CopySelf: also copy the CREVersion, so the illusion/copy doesn't
[13:48:08] <CIA-26> GemRB: break any save games! wooohoo
[13:48:22] <lynxlynxlynx> thanks for your aid
[14:10:12] <lynxlynxlynx> would it be better to make a setter for that?
[14:10:39] <SiENcE> hey
[14:10:57] <SiENcE> what commit version should i take todo a release for dingux now?
[14:12:38] <lynxlynxlynx> latest == best
[14:15:56] <pupnik> how about a big splash page about how to contribute to gemrb development on startup
[14:18:44] <fuzzie> ok, i am gone for the evening
[14:19:00] <fuzzie> i'm sure life is conspiring against me getting anything done..
[14:19:11] <lynxlynxlynx> pupnik: clicking the openhatch link is a good way, bu obscure
[14:19:23] <lynxlynxlynx> clicking the development link is pretty obvious
[14:20:03] <pupnik> oh doh
[14:22:44] <pupnik> it might be fun to run bg2 in wine, and record video and mouse clicks and movements and keyboard input. load a savegame, write a savegame $name1. close. cp $name1 $name1.bg2 run gemrb and playback keyboard/mouse input and save/close - mv $name1 $name1.gemrb
[14:23:41] <pupnik> oh wait, bg2 doesn't run the random number generator deterministically does it
[14:23:45] <pupnik> sorry nm
[14:25:03] <lynxlynxlynx> we're not at a point where this would make sense either
[14:25:06] <pupnik> i was imagining automating the process of running a few seconds of game, from a variety of savegames, then putting the results (savegames and videos) on a webpage
[14:25:09] <pupnik> ok
[14:25:59] <pupnik> is it practically not yet useful, or theoretically impossible?
[14:29:43] <lynxlynxlynx> honestly i don't think it would be ever that useful
[14:30:03] <fuzzie> youtube has plenty of let's play things for bg2, anyhow
[14:30:17] <fuzzie> i just wish someone would do a serious one for pst
[14:30:37] <fuzzie> but i guess it would be sooo long
[14:34:59] <edheldil> what do you mean?
[14:35:12] <edheldil> what's "let's play" ?
[14:35:57] <wjp> gameplay videos
[14:39:40] <edheldil> ah, you mean walkthroughs or quickruns?
[14:43:25] <wjp> yeah
[14:43:32] <wjp> there's tons of them on youtube
[14:53:05] <edheldil> I saw some of ps:t, but only snippets. They consisted of clicking through dialogs
[14:54:49] <wjp> that seems very faithful to the game :-)
[15:17:48] <-- spike411 has left IRC (Quit: Manga & anime pokec na Jabberu: manga.cz@conf.netlab.cz)
[16:01:13] --> barra_home has joined #GemRb
[16:03:03] --> pupnik_ has joined #GemRb
[16:06:23] <-- pupnik has left IRC (Ping timeout: 276 seconds)
[16:22:32] --> spike411 has joined #GemRb
[16:37:41] --> Avenger has joined #GemRb
[16:37:41] --- ChanServ gives channel operator status to Avenger
[16:39:08] <Avenger> hello
[16:58:10] <-- SiENcE has left IRC (Quit: @all: cya)
[17:03:23] <pupnik_> huhu Avenger
[17:12:34] <Avenger> lynx if you are here: LoadScreen.py could easily get the loadhints from loadh25 if the game is tob.
[17:30:02] <lynxlynxlynx> ok
[17:32:23] <lynxlynxlynx> cool, the structure is the same
[17:35:06] <CIA-26> GemRB: 03lynxlupodian * r0ba9f05138fa 10gemrb/gemrb/GUIScripts/bg2/LoadScreen.py: bg2: if tob is installed, use the tob load hint table
[17:41:01] <Avenger> i think we do dialogs wrong, i have not much proof, i just see that dialogs are stored in the actors. Each actor has 2 dialog structures in them (taking up 200 bytes in every critter)
[17:41:30] <Avenger> this may be the reason why they can do interjections much better than us
[17:42:04] <Avenger> if not, then it may be simply crappy coding ;)
[17:50:16] <Avenger> hmm, getting attacked may remove the charm effect (as expected)
[17:51:04] <Avenger> i think i found the entry point of 'actor->sendTrigger'
[17:56:47] <lynxlynxlynx> yes, it depends on the ea
[17:57:02] <lynxlynxlynx> charmed revert, dominated don't
[17:58:29] <Avenger> morale penalty: when get hit below 25% of maxhp, morale-=2
[17:58:49] <Avenger> we have something about this already, but not correct
[18:02:27] <Avenger> if someone killed ANYONE else, the killer receives +3 morale
[18:02:40] <Avenger> not really sophisticated :)
[18:04:20] <-- barra_home has left IRC (Ping timeout: 272 seconds)
[18:06:56] <Avenger> it looks like the 'friends' of the killer receive +2, and the enemies of the killer receive -1 too
[18:07:43] <Avenger> the friend/enemy stuff is not too complicated either. Either goodcutoff/evilcutoff marks the two groups
[18:09:53] --> barra_home has joined #GemRb
[18:13:02] <pupnik_> heh
[18:16:57] <Avenger> killing of innocent/flaming fist (class) changes reputation too, the columns in reputati.2da: 0/3 respectively. As the column headers hint it. I guess we need some more unhardcoding here
[18:17:52] <lynxlynxlynx> we already a
[18:18:00] <lynxlynxlynx> we already handle this
[18:20:52] <Avenger> there is some weird shit about alignments too
[18:21:50] <lynxlynxlynx> yes, there's also a column for killing evil
[18:22:11] <lynxlynxlynx> but it wasn't used in the end, you'd get 20 rep in chateau irenicus already
[18:23:25] <Avenger> yes, true, it is not that
[18:23:40] <Avenger> this is more like: if you are evil, you don't get rep change?
[18:24:50] <lynxlynxlynx> for the same column it would actually make some sense
[18:25:42] <Avenger> oh wait, no, this is not alignment but EA :)
[18:25:55] <Avenger> it is checking if the killer has EA < 15
[18:26:01] <Avenger> this is a new ea value
[18:26:30] <Avenger> if an EA < 15, then the kill is the responsibility of the party
[18:26:56] <Avenger> if the ea > 15 (could still be in the party friendly side), the kill is not chalked up to the party
[18:27:00] <Avenger> this is good to know
[18:27:25] <Avenger> btw, you won't get xp either :)
[18:31:35] <Avenger> i think i can see all the cases where one can get xp: the 3 addxp actions, opcode 93 (learn spell),receiving the killed trigger, remove traps, pick lock
[18:34:45] <Avenger> oh, the notneutral EA value is used by projectiles too
[18:36:46] <Avenger> good you added it yesterday, looks like there is use of that
[18:38:30] <Avenger> omg, what a hack?
[18:38:36] <Avenger> i cannot believe this
[18:39:00] <Avenger> but i swear it is true: if your mage is called Edwin, they cannot be interrupted in their casting?
[18:39:38] <Avenger> maybe this is the scripting name, but i'm not entirely sure :)
[18:40:59] --> SiENcE has joined #GemRb
[18:44:42] <-- barra_home has left IRC (Ping timeout: 240 seconds)
[18:47:57] <lynxlynxlynx> sounds fishy
[18:50:15] <Avenger> totally
[18:52:45] <Avenger> btw, it looks like this is only in tob, not even bg2...
[18:56:42] <spike411> So… any idea why I cannot heal Reig in IWD2? http://paste.jabbim.cz/4833
[18:57:38] <spike411> I guess chmb1.2da is GemRB's override.
[18:58:42] <spike411> Hmm, cannot find it anywhere.
[18:59:27] <Avenger> no, that's supposed to be an avatar anim
[18:59:47] <Avenger> do you see your pc's casting animation?
[19:00:06] <lynxlynxlynx> i think i did
[19:00:20] <lynxlynxlynx> does iwd use different effects for healing?
[19:00:29] <Avenger> i'm not sure about ResistSpell2
[19:00:38] <Avenger> if anywhere, it should be in IWD2
[19:00:52] <spike411> Avenger: Yes
[19:01:00] <spike411> I saw the animation.
[19:01:06] <Avenger> oh, i see, chmb2ca
[19:01:30] <Avenger> chmb1.2da is suposed to be a monster sound 2da :)
[19:01:44] <Avenger> not sure why it tries to access that, when avatars got soundsets
[19:01:54] <Avenger> pc avatars
[19:01:56] <spike411> Meh, I still can't understand all the filenames… :D
[19:02:57] <Avenger> well, starting from chmb1.2da (erroneously accessed to get the casting sound), cgtable.2da (casting glow animation table)
[19:03:08] <Avenger> conjucg.vvc is fine to not be around, the bam is there.
[19:03:23] <Avenger> splprot.2da is the iwd style targeting ids stuff
[19:03:44] <Avenger> the 5 translation errors are missing feature, not harming you
[19:03:52] <Avenger> the next line is fishy
[19:04:10] <Avenger> then it is the casting animation, which you saw fine
[19:04:30] <spike411> ResistSpell2 line is the fishy one?
[19:04:33] <Avenger> then the normal standing around anim
[19:04:34] <Avenger> yes
[19:04:54] <Avenger> can you heal yourself?
[19:05:21] <Avenger> btw, in the past, healing worked
[19:05:26] <Avenger> so, this is weird
[19:09:17] --> Micru has joined #GemRb
[19:09:42] <pupnik_> :/
[19:10:21] <lynxlynxlynx> are you sure it worked in iwd2?
[19:10:47] <spike411> http://paste.jabbim.cz/4834 cure on myself didn't seem to have any effect
[19:11:04] <spike411> Also I cannot use the spell on my portrait, I guess that's a bug?
[19:11:43] <spike411> ‘Select All’ is supposed to select only all the party members, or summoned creatures as well?
[19:13:27] <lynxlynxlynx> the portrait thing is easy to fix if you want to
[19:13:41] <Avenger> hmm there should be something like ActOnPC in the the guicommonwindows.py script
[19:14:12] <Avenger> well, i see actonpc in there
[19:14:13] <Avenger> so, hmm
[19:15:23] <lynxlynxlynx> maybe the indices are wrong
[19:15:33] <Avenger> spike411, you left out the important part from the console log :)
[19:15:52] <spike411> I guess not.
[19:16:19] <spike411> There were no lines after that.
[19:16:38] <Avenger> no more after not finding the vvc? there should have been one about finding the bam
[19:17:04] <spike411> Nope
[19:17:06] <Avenger> or maybe it was cached and didn't looked for again ?
[19:17:14] <Avenger> but you saw casting glow?
[19:17:18] <spike411> Yes
[19:17:42] <Avenger> k
[19:18:09] <Avenger> lynx: what indices? the spell was correctly resolved as a cure light wounds one
[19:18:27] <lynxlynxlynx> portrait/pc
[19:18:39] <lynxlynxlynx> like does it work on the second party member?
[19:18:39] <Avenger> i see
[19:18:59] <lynxlynxlynx> we have an off-by-one in a game or two set up
[19:19:31] <Avenger> yes, but it is not the case, i see: Button.SetVarAssoc ("PressedPortrait", i+1)
[19:19:38] <Avenger> that's how it is in all other games too
[19:20:06] <Avenger> and i fixed dragging recently
[19:21:33] <Avenger> a pity fuzzie isn't here, i finally found the action queue stuff, with the instants and such :)
[19:23:27] <-- SiENcE has left IRC (Ping timeout: 265 seconds)
[19:25:00] <spike411> Well the dragging is still not perfect. :)
[19:25:54] <Avenger> what is the situation? is the portrait stuck on your cursor too much, or too little?
[19:26:16] <Avenger> if too much sticky, then you don't have the latest source :)
[19:28:30] <spike411> Hmm, another audio-thread crash http://paste.jabbim.cz/4835
[19:28:46] <spike411> Avenger: I'm always on the bleeding edge. ;)
[19:29:22] <spike411> BTW shouldn't Ctrl (or Cmd on OS X) + click multi-select creatures/party members?
[19:29:38] <Avenger> yes it should
[19:29:41] --> SiENcE has joined #GemRb
[19:29:46] <Avenger> maybe shift click?
[19:29:56] <lynxlynxlynx> we have shift, yes
[19:30:34] <spike411> OK
[19:31:57] <spike411> Fun fact: I'm using Czech keyboard layout where I can access numbers on alphanumeric keyboard with Shift modifier, so I usually go through dialogs with Shift+num, Shift+num, Enter, Enter… but when the last dialog screen comes, I have to press num-only…
[19:32:24] <spike411> Although sometimes I have to press numbers without Shift throughout the whole dialog.
[19:32:55] <spike411> I checked and menubar in OS X still shows Czech layout…
[19:35:01] <CIA-26> GemRB: 03lynxlupodian * rfa146c865cc4 10gemrb/gemrb/core/GameScript/Actions.cpp:
[19:35:01] <CIA-26> GemRB: fixed LeaveAreaLUA*Entry to call LeaveAreaLUA* correctly, since their
[19:35:01] <CIA-26> GemRB: parameters differ. An empty mos means we'll look it up again on load
[19:35:01] <CIA-26> GemRB: this too broke saving, since the wrong mos was looked up
[19:36:37] <spike411> Avenger: Ad dragging 1) I select a portrait. 2) Drag it. 3) Drop it. 4) Now another portrait seems highlighted and I cannot select the portrait from (1) until I select another portrait first.
[19:38:34] <spike411> Also sometimes when I'm draggin a portrait it seems to be dropped randomly and automatically.
[19:39:13] <spike411> Nevermind, I guess that's because I leave the portrait area accidentaly.
[19:39:43] <spike411> (Still it's confusing.)
[19:42:31] <Avenger> yes, maybe thats not like the original, but i cannot fix that easily
[19:42:49] <spike411> OK
[19:44:45] <spike411> OK, Shift+click can select multiple characters via portraits. Not via avatars though.
[19:45:08] <spike411> (Again, not sure about the original behaviour though.)
[19:50:29] --> edheldil_ has joined #GemRb
[19:51:41] <spike411> Huh http://img185.imageshack.us/img185/2353/gemrb062iwd2glitch007.png
[19:51:55] <spike411> Is the goblin talking to me? :o
[19:52:37] <Avenger> haha
[19:52:50] <Avenger> that is some misplaced constant string, i guess
[19:52:57] <Avenger> a pity you don't have strref on
[19:53:11] <Avenger> do you know how to turn them on?
[19:53:49] <spike411> In the game's ini file or something…?
[19:54:01] <Avenger> edit your original game config, [program options] Strref On=1
[19:54:40] <spike411> Aaaand another audio crash.
[19:55:27] <Avenger> 2742 is the strref, btw
[19:56:22] <Avenger> i don't know what causes it
[19:56:29] <spike411> OK, turned on strrefs.
[19:56:38] <spike411> git pull && make…
[19:56:44] <spike411> Also. :)
[19:56:58] <spike411> Didn't have the last commit yet.
[19:57:29] <spike411> Oooh, funny.
[19:58:59] <spike411> Known? [GameScript]: Invalid scripting action: runtopoint([1869.957])
[20:02:17] <lynxlynxlynx> dudd
[20:02:46] <spike411> Are there any messages for which I should really look out?
[20:02:54] <spike411> (errors / warnings)
[20:05:52] <lynxlynxlynx> not really
[20:07:24] <spike411> Avenger: When I cast the cure spell on another party member (with full HP though): http://paste.jabbim.cz/4836
[20:08:11] <spike411> Meh, I get these audio-thread crashes fairly often. Maybe it's indirectly triggered by spellcasting.
[20:09:49] <lynxlynxlynx> make a debug build
[20:09:58] <lynxlynxlynx> run gemrb in gdb
[20:10:59] <spike411> Lovely idea. :)
[20:11:33] <spike411> Do you have a make target for debug? (I can see I'm building RelWithDebInfo or something.)
[20:11:59] <lynxlynxlynx> pass -DCMAKE_BUILD_TYPE=Debug
[20:12:27] <spike411> Ah, thanks.
[20:13:18] <Avenger> maybe that is the osx openal implementation being crappy
[20:13:31] <spike411> Avenger: No surprise there… :D
[20:18:38] <spike411> I guess I could try SDL mixer instead (?) of OpenAL.
[20:20:49] <Avenger> well, that was recently fixed too, and could still be somewhat buggy. if you want to take no risk, maybe use nullsound
[20:23:14] <spike411> Note: It's not a good idea to cast Cure on Reig before the dialog starts.
[20:23:47] <Avenger> bye!
[20:23:49] <-- Avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
[20:24:07] <spike411> No user interface, no cursor…
[20:29:55] <lynxlynxlynx> i guess you interrupted the cutscene
[20:30:03] <spike411> Yep
[20:31:58] <lynxlynxlynx> wow, abazigal is sure deadly
[20:37:52] <spike411> And now I cannot get it to crash. How typical.
[20:38:09] <spike411> Anyways, gotta go to bed. Getting up at 4 AM, yay.
[21:01:44] <-- Micru has left IRC (Remote host closed the connection)
[21:05:41] <edheldil_> heh, is it just me, or is MSDN down? :)
[21:32:08] <fuzzie> hi
[21:34:16] <lynxlynxlynx> oj
[21:34:48] <lynxlynxlynx> i had to hack HPPercentLT to not return true for dead actors or i'd get infinite dragons >>
[21:34:57] <fuzzie> hehe
[21:35:08] <lynxlynxlynx> hostile dragons
[21:35:58] <lynxlynxlynx> does that sound like a sane place to be doing that?
[21:36:21] <fuzzie> this is in the effect thing?
[21:38:55] <lynxlynxlynx> the effect is replace creature, but luckily the level is already bad, so only one is summoned as a replacement
[21:39:14] <fuzzie> i mean, where is this done?
[21:39:22] <lynxlynxlynx> script
[21:39:28] <lynxlynxlynx> bg2/abazigal.baf
[21:39:45] <lynxlynxlynx> the second spell has that effect
[21:40:02] <lynxlynxlynx> even with my hack, i got two dragons, but that's maybe just a timing bug
[21:40:18] <lynxlynxlynx> or death being sucky
[21:41:50] --> Maighstir has joined #GemRb
[21:41:50] <fuzzie> i think the trigger should return true
[21:42:53] <fuzzie> and the action should error out if the target is dead
[21:43:28] <fuzzie> based on, well, the disasm of the action checking the dead state and exiting :)
[21:43:51] <lynxlynxlynx> ReallyForceSpellRES?
[21:44:09] <fuzzie> well, there is only one function for ForceSpell, and i glance at the ReallyForceSpell bit
[21:47:19] <fuzzie> but i mean, i wouldn't implement anything based on this, i would ask Avenger
[21:50:15] <lynxlynxlynx> afterwards there were some other bugs, but i can't tell if they weren't related yet
[21:51:08] <lynxlynxlynx> and i had a wierd dialog bug happen a few times too
[21:52:03] <lynxlynxlynx> it ended abrupty instead of continuing
[21:52:30] <lynxlynxlynx> bg2/bazeye01.d:93 <-- break out instead of 12
[21:53:01] <lynxlynxlynx> err no
[21:53:16] <lynxlynxlynx> 12 did start, i just usually didn't get the kobolds option
[21:53:56] <fuzzie> it just exits?
[21:54:12] <lynxlynxlynx> yes
[21:54:13] <fuzzie> if so, you should get a console error
[21:55:34] <lynxlynxlynx> [Dialog]: Can't redirect dialog
[21:56:14] <fuzzie> mph
[21:57:54] <lynxlynxlynx> i can step through that tommorow to see where it became null
[21:58:41] <fuzzie> no
[21:58:42] <fuzzie> i mean
[21:58:47] <lynxlynxlynx> interdia was loaded, so not in the last block
[21:59:01] <fuzzie> you probably have Anomen, Imoen or Jan in your party?
[21:59:23] <fuzzie> and they probably have the SoA dialog assigned
[21:59:40] <lynxlynxlynx> none of those
[22:00:13] <fuzzie> ok, then it is weird :)
[22:00:23] <lynxlynxlynx> they do all have the old dialogs
[22:00:31] <lynxlynxlynx> there was supposed to be an interjection?
[22:00:44] <fuzzie> if one of those were present, yes
[22:01:04] <fuzzie> but obviously an interjection will fail and abort the dialog..
[22:01:52] <lynxlynxlynx> oh, i do have keldorn
[22:02:14] <lynxlynxlynx> in my file it checks for korgan, minsc and keldorn
[22:02:31] <fuzzie> huh :)
[22:02:34] <fuzzie> well, that is your answer, anyway
[22:02:38] <lynxlynxlynx> IF ~~ THEN REPLY #72328 /* ~Kobolds?~ */ GOTO 13
[22:02:39] <lynxlynxlynx> IF ~IsValidForPartyDialog("Minsc")~ THEN EXTERN ~MINSC25J~ 13
[22:02:41] <lynxlynxlynx> IF ~IsValidForPartyDialog("Korgan")~ THEN EXTERN ~KORGA25J~ 15
[22:02:42] <lynxlynxlynx> IF ~IsValidForPartyDialog("Keldorn")~ THEN EXTERN ~KELDO25J~ 18
[22:02:51] <lynxlynxlynx> yeah, but how come it works some times?
[22:03:05] <fuzzie> because Keldorn isn't in sight, or Keldorn is the speaker
[22:03:11] <lynxlynxlynx> ah
[22:03:17] <lynxlynxlynx> ok
[22:03:22] <fuzzie> (this is why the tweak pack has a "get rid of IsValidForPartyDialog from every dialog" option)
[22:04:36] <fuzzie> so i hope that makes sense :)
[22:05:22] <fuzzie> we should really print more information in that error..
[22:21:15] <-- lynxlynxlynx has left IRC (Read error: Connection reset by peer)
[22:32:43] <-- mattinm has left IRC (Ping timeout: 272 seconds)
[22:44:07] --> mattinm has joined #GemRb
[22:56:34] <-- Maighstir has left IRC (Quit: ~ Trillian Astra - www.trillian.im ~)
[23:08:58] <-- SiENcE has left IRC (Quit: cya @all)