#gemrb@irc.freenode.net logs for 3 Feb 2013 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[01:50:26] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)
[01:54:46] <Seniorita> [bug] 3603146 - BG2: illithid brain items cannot be picked up http://sourceforge.net/tracker/?func=detail&atid=110122&aid=3603146&group_id=10122
[02:14:12] <-- i30817 has left IRC (Quit: ChatZilla 0.9.89 [Firefox 18.0.1/20130117041608])
[03:53:16] --- ermo is now known as ermo^
[06:00:58] <-- WingedHussar has left IRC (Ping timeout: 245 seconds)
[06:10:45] --> WingedHussar has joined #gemrb
[09:28:48] <-- traveler has left IRC (Ping timeout: 245 seconds)
[09:34:25] --> traveler has joined #gemrb
[09:45:28] --> fizzle has joined #gemrb
[10:07:56] --> lynxlynxlynx has joined #gemrb
[10:07:56] --- ChanServ gives channel operator status to lynxlynxlynx
[10:13:25] --> Yoshimo has joined #gemrb
[10:33:39] <Seniorita> [commit] Jens Granseuer: add new function for safely copying resrefs http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=65a066137067f79ae1361aa2a41f267703908c2d
[10:33:40] <Seniorita> [commit] Jens Granseuer: remove obsolete and broken autotools build system http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=3ac2b3a4e871b48f0379a842381afe0d00b08043
[10:33:58] <fizzle> somebody (who's not me) should update Release.txt
[10:36:15] <lynxlynxlynx> ugh
[10:37:08] <lynxlynxlynx> the autotools one is bad, people still use it
[10:37:34] <lynxlynxlynx> we don't maintain it much, but some platforms don't have cmake
[10:38:36] <DrMcCoy> Hmm, to start a flamewar or not...
[10:39:21] <wjp> if we want to undo that commit, there's a small window in time in which we can just force-push it away without too much hassle
[10:41:45] <fuzzie> hasn't it been broken for quite some time?
[10:41:57] <lynxlynxlynx> just a bunch of deletes, it should always work
[10:42:24] <lynxlynxlynx> i don't know, but for example even freebsd was still using (and patching) it until i pointed out cmake
[10:42:25] <fuzzie> so I have no opinion either way any more
[10:42:57] <fuzzie> although cmake is really pretty bad. :/
[10:42:58] <wjp> but it's been broken at least since the override->unhardcoded split, it seems
[10:44:13] <fizzle> lynxlynxlynx: it doesn't work
[10:46:30] <lynxlynxlynx> probably not much effort to fix it
[10:46:39] <lynxlynxlynx> so if you ask me it is still viable
[10:46:43] <fuzzie> it's often broken, because we don't use it, and porters only fix it at releases
[10:46:53] <fuzzie> but is anyone using it any more?
[10:47:05] <fizzle> well, maintaining two systems doesn't make much sense, does it?
[10:47:34] <fizzle> I actually prefer autotools to cmake but we should really decide on one
[10:48:09] <fizzle> the other will always be (near-)broken
[10:48:26] <fuzzie> well, as I understand it, the situation is that only autotools works for everyone, but no-one is able to maintain autotools beyond the very basics
[10:48:30] <lynxlynxlynx> we don't maintain the other one
[10:51:58] <fizzle> and you rely on random people to figure out what's changed since the last release?
[10:52:13] <fizzle> and update the scripts for their specific builds?
[10:54:31] <fizzle> sounds like going autotools all the way would be a better approach
[10:54:39] <wjp> FWIW, it seems only very superficially broken currently
[10:54:52] <wjp> (just some broken filespecs in override and unhardcoded and a few missing files in core)
[10:54:57] <lynxlynxlynx> we just left it as an option
[10:55:18] <lynxlynxlynx> the switch to cmake was years ago by now
[10:55:18] <fizzle> wjp: and iconv and strlcpy...
[10:55:39] <-- Yoshimo has left IRC (Quit: Yoshimo)
[10:57:54] <fizzle> well, anyway, whichever way you like; no strong feelings either way
[10:58:03] <fizzle> even though having to build systems still doesn't sound smart to me
[10:58:09] <fizzle> *two
[10:58:34] <lynxlynxlynx> we have a large disclaimer in autogen
[11:00:11] <lynxlynxlynx> i suggest we remove it after the next release
[11:00:24] <lynxlynxlynx> i'll add a warning/cfv to the notes
[11:00:30] <fizzle> I know I missed that disclaimer when I first tried to build :)
[11:00:38] <wjp> so... force-push it back? Although half an hour is starting to feel slightly uncomfortable
[11:02:43] <lynxlynxlynx> why not do it cleanly with a revert?
[11:04:22] <wjp> that's the alternative. But it's a matter of perspective which is cleaner :-)
[11:04:32] <wjp> the transition is cleaner with a revert, but the end result isn't
[11:05:21] <fuzzie> pretty sure gemrb is quiet enough that you can take a while for it
[11:26:27] <traveler> when freebsd was using autotools?
[11:27:05] <traveler> i mean, it must have been when i was looking in the other way
[11:30:26] <traveler> ah, ok
[11:30:34] <traveler> 2,5 years ago
[11:33:05] --> mijas has joined #gemrb
[11:35:06] <lynxlynxlynx> maybe netbsd then
[11:35:23] <lynxlynxlynx> a thomas is the packager
[11:36:35] <lynxlynxlynx> yeah, netbsd & klausner
[11:38:42] --> kida has joined #gemrb
[11:40:13] <-- mijas has left IRC (Quit: Page closed)
[12:19:16] <lynxlynxlynx> so
[12:19:40] <lynxlynxlynx> i've got a fix for this pile issue to push
[12:53:34] --> edheldil_ has joined #gemrb
[12:57:18] <Seniorita> [commit] Jaka Kranjc: PerformAttack: moved the immobility check to AttackCore http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=e43fe71409434e42c67d8659665671c0239648ee
[12:57:20] <Seniorita> [commit] Jaka Kranjc: in the original (bg2) only party members suffer fatigue with time http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=1497a4a7e0ff8aacc7be2f15d7e2b81f11abca68
[12:57:21] <Seniorita> [commit] Jaka Kranjc: UseContainer: if the pile is too far, try with the normal range after 10 attempts http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=baa98f8b12536bcab14929dfcb76e5187675581a
[12:57:22] <Seniorita> [commit] Jaka Kranjc: redone fatigue handling to be more like in the original http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=ef16ed22444fbe236811eb7f8f6154eb7a073ea4
[12:57:23] <Seniorita> [commit] Jaka Kranjc: Revert "remove obsolete and broken autotools build system" http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=61ec1a9f1e765c6535082e358f24bc96f8838577
[12:57:36] <edheldil_> there goes my build system, sniff ...
[12:57:49] <lynxlynxlynx> not yet
[12:58:11] <lynxlynxlynx> but soon (queue creepy eyes)
[13:04:26] <wjp> (cue)
[13:04:43] <wjp> I'll commit these automake fixes
[13:05:02] <wjp> haven't completed the iconv detection yet, though
[13:12:27] --> avenger has joined #gemrb
[13:12:28] --- ChanServ gives channel operator status to avenger
[13:12:46] <avenger> lol what are you doing with automake? Remove or not, that is the question.
[13:13:49] <Seniorita> [commit] Willem Jan Palenstijn: Fix autotools build http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=124a9d8dc60c8b37deeea8664fd5bc3593faee42
[13:14:01] <wjp> we're not quite sure yet :-)
[13:14:51] <avenger> you got bluffed into fixing it?
[13:15:23] <wjp> nah, I was just curious how broken it was
[13:15:55] <avenger> actually, i think it is a good idea to have 2 build systems, if they really work
[13:16:04] <avenger> options are not bad
[13:20:40] <traveler> explain
[13:23:04] --> Yoshimo has joined #gemrb
[13:24:29] <traveler> ok, that was flamebait. but it obviously goes against all plausible reasonable practice, bbl.
[13:30:44] <-- Yoshimo has left IRC (Quit: Yoshimo)
[13:44:31] <avenger> traveler: if you are multiplatform, with really obscure targets, it is inevitable that one build system is broken. If we have two, it is like a backup system.
[13:54:30] <lynxlynxlynx> wb
[14:10:41] <-- avenger has left IRC (Remote host closed the connection)
[14:16:59] --> Avenger has joined #gemrb
[14:16:59] --- ChanServ gives channel operator status to Avenger
[14:20:04] <Avenger> lynx, just liked at iwd2 Unhide action we still miss: http://forums.gibberlings3.net/index.php?showtopic=25032
[14:20:06] <Seniorita> Unhide - The Gibberlings Three Forums
[14:20:34] <Avenger> is there any other action you want to know about atm?
[14:21:44] <lynxlynxlynx> not really
[14:22:08] <lynxlynxlynx> we don't fully support the modal combat feats yet, but i don't think that goes through actions
[14:33:57] <lynxlynxlynx> what would be interesting to know is if RemoveTraps really does halve and reroll skill like we have it now
[14:34:04] <lynxlynxlynx> you can check that with bg2
[14:34:56] <Avenger> halve and reroll?
[14:36:25] <Avenger> this --> roll = core->Roll(1, skill/2, 0); ?
[14:36:45] <Avenger> i think the original rolls d10 and adds it to the stat
[14:37:00] <Avenger> or something like that, it is not that random like ours
[14:39:45] <Avenger> the d10 is adjusted by luck (if luck is 2 then d10 is 2-10) or so
[14:40:47] <Avenger> so roll = skill + LuckyRoll(10) or something like that
[14:47:20] <lynxlynxlynx> no, i mean for the other games
[14:47:29] <lynxlynxlynx> for iwd2 we already have the right calc
[14:47:38] <lynxlynxlynx> the others use this skill/2 roll + skill/2
[14:47:48] <lynxlynxlynx> a bit too harsh if you ask me
[15:11:22] <lynxlynxlynx> stuck in a cutscene in totl currently
[15:11:30] <lynxlynxlynx> almost halfway there
[15:19:55] <lynxlynxlynx> nothing obvious goes wrong http://paste.debian.net/231412/
[15:19:57] <Seniorita> debian Pastezone
[15:20:14] <lynxlynxlynx> the spell is cast, vars get set, but the cutscene isn't terminated
[15:20:26] <lynxlynxlynx> it was started from dialogue just as usual
[15:33:37] <fuzzie> what does INNATE_HOBART_TO_RAKSHASA do?
[15:42:16] <Avenger> fuzzie: which game and resref?
[15:44:47] <fuzzie> that's iwd, right? so i assume spin180 (spell.ids has it as 3180)
[15:44:55] <Avenger> fuzzie: spin180 does: animation removal, summon rakshasa to self, destroys self
[15:45:20] <Avenger> no idea why didn't they use replace self
[15:45:22] <fuzzie> right, that's what I thought :P
[15:45:29] <Avenger> that would be 1 opcode instead of 3
[15:45:40] <fuzzie> so.. if you force that on yourself then you stop running actions, surely :)
[15:45:57] <fuzzie> and thus that EndCutSceneMode will never be reached
[15:46:13] <Avenger> well, it calls an instant death opcode
[15:46:25] <Avenger> iwd2 lets you endcutscenemode while dead?
[15:46:38] <fuzzie> I guess that would be pretty harmless.
[15:46:54] <Avenger> wait, death is delayed by 1
[15:46:59] <Avenger> so this should work
[15:47:18] <fuzzie> lynx's pasted script has Wait(2) and Wait(1) between the spell cast and the EndCutSceneMode
[15:47:18] <Avenger> all opcode in this is delayed by 1
[15:47:21] <fuzzie> so that's 3*15 ai ticks
[15:47:46] <Avenger> uh, they wait in script and wait in spell opcodes, why
[15:48:03] <lynxlynxlynx> fuzzie: a polymorph or maybe even a char replacement (i don't know of such an opcode though)
[15:48:30] <lynxlynxlynx> ah, you got it covered
[15:58:39] <fizzle> I'm running into some effects trouble
[15:59:03] <fizzle> a creature with an item that applies "slow" (instant_limited)
[15:59:17] <fizzle> if you save and reload, the effects gets applied again
[15:59:27] <fizzle> after the fourth time the creature is immobile...
[16:00:13] <fizzle> and then a script tells it to MoveToPoint which obviously doesn't work and never finishes
[16:02:27] <fizzle> I guess the problem is the timing mode isn't instant_while_equipped here
[16:03:55] <lynxlynxlynx> that's one of the permanent timing modes?
[16:05:36] <fizzle> instant_limited isn't "permanent" as I understand it but the effect has a duration of 999 which sounds fairly long in any case
[16:06:32] <lynxlynxlynx> you could make fx_set_slowed_state exit if it is called with the same params
[16:06:46] <lynxlynxlynx> but why do the effects get stacked?
[16:07:07] <fizzle> you mean why slow allows stacking
[16:07:15] <lynxlynxlynx> ah no, you can't be double slowed anyway, you can just quit on the state being set already
[16:07:40] <fizzle> oh, okay, if double-slow is not valid then that's the simple fix
[16:10:44] <fuzzie> well, Persistent() returns true for the instant_limited case
[16:10:56] <fuzzie> so the effect will get saved and the item will re-apply the effect I guess?
[16:11:20] <fuzzie> so them stacking makes sense
[16:12:32] <fizzle> yes, that's where it comes from
[16:12:39] <fizzle> but it shouldn't happen in this case
[16:12:46] <fuzzie> well, it probably *should* happen
[16:12:55] <fuzzie> so probably what lynx said :)
[16:13:59] <fizzle> you're making me dizzy ;)
[16:14:21] <lynxlynxlynx> add the state check
[16:14:45] <fizzle> yes, on it
[16:17:29] <lynxlynxlynx> uhh
[16:17:40] <lynxlynxlynx> stuck now for real
[16:17:44] <lynxlynxlynx> there's a talking door
[16:18:09] <lynxlynxlynx> no dialog, there seems to be a hidden actor nearby, while the door has a talking animation
[16:18:29] <fuzzie> right, thankfully it is not pst :)
[16:18:39] <fuzzie> los issues?
[16:19:51] <lynxlynxlynx> no, i suspect the actor is disabled
[16:20:15] <lynxlynxlynx> the door has a script, but only relevant with the dialog (which is not part of the door)
[16:20:30] <lynxlynxlynx> just watched the lp to see how it looked
[16:22:08] <Seniorita> [commit] Jens Granseuer: don't allow "slowed" effects to stack http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=10499e967edde5870c1333671789dbd3be509908
[16:28:02] <lynxlynxlynx> there is some kind of active region on the floor
[16:28:37] <lynxlynxlynx> and on part of the door itself
[16:28:46] <lynxlynxlynx> dltcep time
[16:33:27] <lynxlynxlynx> the actor has some interesting bits set
[16:34:02] <lynxlynxlynx> for Fields dltcep's tooltip only knows about 1 and 8, but this guy has 7, so there's two more
[16:34:47] <-- fizzle has left #gemrb
[16:36:47] <lynxlynxlynx> gemrb only knows the same two
[16:37:37] <lynxlynxlynx> Avenger: any ideas about creature area flags' bits 2 and 3? we only cover AF_CRE_NOT_LOADED and AF_NAME_OVERRIDE
[16:38:07] <Avenger> yes some
[16:38:15] <Avenger> i defined the names
[16:39:07] <Avenger> somewhere..
[16:40:12] <Avenger> and implemented some?
[16:40:46] <Avenger> AF_SEEN_PARTY and AF_INVULNERABLE
[16:40:58] <Avenger> they are in AREImporter
[16:42:30] <Avenger> it is likely tob doesn't use more flags
[16:43:57] <lynxlynxlynx> oh, just bad ordering, a few lines below
[16:44:51] <lynxlynxlynx> i wonder why this is important
[16:46:00] --> Yoshimo has joined #gemrb
[17:29:17] <lynxlynxlynx> ok, great
[17:29:32] <lynxlynxlynx> this door is the only blocker, the rest of totl is fine
[17:41:26] <lynxlynxlynx> i don't see what could start a dialog here
[17:42:01] <lynxlynxlynx> the "spy" has no script, the area one is boring, the door has only open/close stuff
[17:50:26] <Avenger> which area?
[17:53:53] <lynxlynxlynx> doesn't look hardcoded either
[17:54:01] <lynxlynxlynx> sec
[17:54:37] <lynxlynxlynx> ar9717
[17:55:01] <-- kida has left IRC (Ping timeout: 245 seconds)
[17:55:28] <Avenger> magic mouth talk creature?
[17:55:32] <lynxlynxlynx> yep
[17:56:02] <lynxlynxlynx> oh, he has an area override for the class script
[17:56:06] <Avenger> yes
[17:56:11] <Avenger> dummtalk
[17:56:52] <Avenger> contains: !Dead("yself)") // yself)
[17:56:54] <lynxlynxlynx> !Dead("yself)") // yself) <--
[17:56:54] <Avenger> hahahah
[17:57:05] <lynxlynxlynx> isn't this just a decompiler bug?
[17:57:05] <Avenger> echo off
[17:57:09] <Avenger> no
[17:57:20] <lynxlynxlynx> i've seen it before
[17:57:27] <Avenger> it is a scripter failure
[17:57:41] <Avenger> they changed Dead() but their scripters used to the original
[17:58:33] <lynxlynxlynx> you think they added hack to parse this anyway?
[17:58:53] <Avenger> it is always true
[17:59:08] <Avenger> dead("whatever") is false
[17:59:15] <Avenger> !dead("whatever") is true
[17:59:38] <lynxlynxlynx> bg2 is full of these errors
[17:59:47] <Avenger> bg2?
[17:59:48] <lynxlynxlynx> 279
[18:00:01] <Avenger> heh, i thought this is only in iwd
[18:01:14] <lynxlynxlynx> anyway, the way you describe it this should work normally
[18:01:35] <lynxlynxlynx> so maybe it's a See thing again
[18:07:26] <-- Avenger has left IRC (Quit: bye!)
[18:43:05] <-- Yoshimo has left IRC (Quit: Yoshimo)
[18:57:07] <lynxlynxlynx> ok, found the problem
[18:57:51] <lynxlynxlynx> GameScript::GlobalTimerExpired always returns 0, since we also check the value (which is already zero, so we skip the equality check)
[18:58:02] <lynxlynxlynx> i could swear we changed this before
[18:58:03] --> rocket_hamster has joined #gemrb
[18:59:09] <lynxlynxlynx> indeed
[18:59:37] <lynxlynxlynx> experiment in the other direction failed d18a3cce
[19:04:13] <lynxlynxlynx> i guess it wasn't set at all
[19:24:46] --- ermo^ is now known as ermo
[20:08:32] <lynxlynxlynx> i don't see how this can work in the original
[20:09:08] <lynxlynxlynx> all the timer setter are behind variable walls that rely on the timer not being expired yet to be true
[20:17:40] <fuzzie> so did you ever fix the bug mentioned in d18a3cce?
[20:17:48] <-- rocket_hamster has left IRC (Quit: BuhBye!)
[20:19:41] <lynxlynxlynx> yeah, it was that activation thing
[20:20:03] <lynxlynxlynx> actors start active, except in bg1
[20:47:52] --> Yoshimo has joined #gemrb
[21:32:36] <-- |Cable| has left IRC (Ping timeout: 248 seconds)
[21:40:21] <-- Yoshimo has left IRC (Quit: Yoshimo)
[21:45:23] --> |Cable| has joined #gemrb
[22:08:12] <lynxlynxlynx> lynxlynx.info/ie/how.baf.tar.bz2 <-- fuzzie, your dump was missing totl, this one is complete
[22:52:58] <-- traveler has left IRC (Ping timeout: 245 seconds)
[23:01:12] --> traveler has joined #gemrb
[23:07:21] --> rocket_hamster has joined #gemrb
[23:18:14] <rocket_hamster> heya
[23:20:19] <lynxlynxlynx> oj
[23:21:42] <rocket_hamster> hows things? :D
[23:22:11] <lynxlynxlynx> awesome
[23:22:44] <rocket_hamster> owww
[23:22:48] <Seniorita> [commit] Jaka Kranjc: guirec: use a proper strref for electricity resistance in iwd http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=abef6ff20deebf556c44418d165c2b3bccf7d70d
[23:22:49] <Seniorita> [commit] Jaka Kranjc: fx_tenser_transformation: also give a tohit bonus; cleanup http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=449a34f34f86d98723b46a51c51095271679ada6
[23:22:50] <Seniorita> [commit] Jaka Kranjc: SetupDamageInfo: partially disable the iwd2 hpbar, so the full-hp max isn't visible http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=83b41c65592f480983fbbbcd1a8f6a74c2f7f89e
[23:22:52] <Seniorita> [commit] Jaka Kranjc: fx_smite_evil: give the actual boni http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=870cdba9ae19d9db1750ace958039ee5756eefeb
[23:23:15] <rocket_hamster> [GameScript]: Displaying string on: imoen >>> on every game tick? :(
[23:23:30] <lynxlynxlynx> i fixed a resting bug
[23:23:44] <lynxlynxlynx> old saves are affected, of course, but resting will reset everything
[23:24:56] <traveler> \o/
[23:25:23] <traveler> not that i have experienced this bug, but this is bg2 stuff?
[23:26:06] <rocket_hamster> correct
[23:31:02] <lynxlynxlynx> all the games
[23:31:24] <lynxlynxlynx> luck would max out in certain circumstances
[23:31:49] <lynxlynxlynx> you'd always hit and it would do max damage
[23:32:00] <lynxlynxlynx> or they would :)
[23:32:15] <lynxlynxlynx> pretty obvious when it happened
[23:33:07] <rocket_hamster> are you still working on iwd mostly?
[23:33:47] <lynxlynxlynx> today was iwd1 day
[23:33:53] <lynxlynxlynx> iwd2 i try not to touch anymore
[23:34:46] <lynxlynxlynx> two totl bugs to fix, then at least the two pst crashers, then back to the rare inventory problems
[23:34:47] <rocket_hamster> :)
[23:35:11] <traveler> well i did not notice luck bug
[23:35:37] <traveler> because i would anyway always hit probably
[23:35:56] <rocket_hamster> i noticed damage dealt was constant for long time
[23:36:01] <rocket_hamster> havent tested recently
[23:36:05] <traveler> hmm
[23:36:20] <lynxlynxlynx> once you hit the bug, it was permanent, since we were updating the base stat
[23:36:20] <traveler> or i wouldn't be fatigued anyway too
[23:36:34] <traveler> ah
[23:37:05] <lynxlynxlynx> i don't really know what the remaining corner case was
[23:37:26] <lynxlynxlynx> the current way is more robust plus matches the original more closely
[23:37:53] <traveler> cool :)
[23:48:44] --> joneirik has joined #gemrb