#gemrb@irc.freenode.net logs for 10 Jun 2009 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:03:00] --> barra_library has joined #gemrb
[00:10:16] --- mattinm1987 is now known as mattinm
[00:18:09] <-- barra_away has left IRC (Read error: 110 (Connection timed out))
[00:28:12] <CIA-2> gemrb: 03mattinm * r6377 10/gemrb/trunk/gemrb/ (GUIScripts/bg2/GUIREC.py plugins/GUIScript/GUIScript.cpp):
[00:28:12] <CIA-2> gemrb: added IsDualWielding to GUIScripts
[00:28:12] <CIA-2> gemrb: GUIREC now shows the independent THAC0s if dual wielding
[00:28:30] --- ChanServ gives channel operator status to mattinm
[00:52:50] --> |Cable| has joined #gemrb
[02:56:43] <-- mattinm has left IRC (Read error: 110 (Connection timed out))
[04:05:59] --> mattinm has joined #GemRB
[04:06:18] --- ChanServ gives channel operator status to mattinm
[04:31:14] --> jdm- has joined #gemrb
[05:44:02] <mattinm> some interesting info:
[05:44:03] <mattinm> It should be noted that Baldur's gate does not store the actual rolls; when you level up, it merely adds the roll to the base HP value, if you get level drained, it takes a fraction depending on the amount drained (i.e., 2 level of an 8th level character = 25% of base hit points)
[06:30:47] --> lynxlynxlynx has joined #gemrb
[06:30:47] --- ChanServ gives channel operator status to lynxlynxlynx
[06:34:49] <CIA-2> gemrb: 03mattinm * r6378 10/gemrb/trunk/gemrb/GUIScripts/bg2/GUIREC.py: use Main/Off Hand THAC0 string instead of /
[06:43:19] <CIA-2> gemrb: 03mattinm * r6379 10/gemrb/trunk/gemrb/ (3 files in 3 dirs):
[06:43:19] <CIA-2> gemrb: penalties for two-weapons in BG2 (needs updating for IWD2)
[06:43:19] <CIA-2> gemrb: added wstwowpn.2da (and doc) to hold two-weapon penalties for BG2
[06:55:01] <lynxlynxlynx> :)
[06:55:32] <-- Gekz has left IRC (Read error: 104 (Connection reset by peer))
[06:55:44] --> Gekz has joined #GemRB
[06:55:56] <mattinm> hi lynx :)
[07:27:32] <-- Gekz has left IRC (Remote closed the connection)
[07:27:35] --> Gekz_ has joined #GemRB
[07:36:26] <CIA-2> gemrb: 03mattinm * r6380 10/gemrb/trunk/gemrb/ (3 files in 3 dirs):
[07:36:26] <CIA-2> gemrb: added two-handed weapon style to wstwohnd.2da (and docs)
[07:36:26] <CIA-2> gemrb: moved the style code to the less intrusive GetCombatDetail
[07:56:35] <lynxlynxlynx> why 3 levels?
[07:56:55] <mattinm> it's just easier for compatibility with 2 weapons
[07:57:15] <lynxlynxlynx> ah
[07:57:23] <mattinm> easier to make them all get the 3rd level (even though it can't be assigned), then store a check for it
[07:57:43] <mattinm> sword+shield is the worst :)
[07:57:50] <mattinm> it just has a +ac vs missile mod
[08:00:10] <lynxlynxlynx> single-handed is not too good either
[08:00:35] <lynxlynxlynx> monks are too good to use weapons and their ac is low anyway
[08:00:47] <mattinm> yeah; the *only* use i can see for single-handed is if you want to increase your chances of a critical on a backstab
[08:01:52] <mattinm> two-handed might be the craziest of all though; with the -4 speed mod, +1 critical range, and +1dmg with Carsomyr, you can do serious damage
[08:08:35] <lynxlynxlynx> i prefer the worpal sword ;)
[08:09:03] <lynxlynxlynx> vorpal
[08:10:04] <mattinm> lilacor ftw :)
[08:10:20] <lynxlynxlynx> of course
[08:10:33] <lynxlynxlynx> i can't understand why some people want to shut him up
[08:11:55] <mattinm> i don't either, he's pretty hilarious
[08:20:58] <CIA-2> gemrb: 03mattinm * r6381 10/gemrb/trunk/gemrb/ (5 files in 3 dirs):
[08:20:58] <CIA-2> gemrb: added single-weapon and sword/shield tables and docs
[08:20:58] <CIA-2> gemrb: masked off the proficiencies
[08:20:58] <CIA-2> gemrb: added defensive modifiers for the new styles
[08:31:38] <-- |Cable| has left IRC ("Leaving")
[08:33:26] <CIA-2> gemrb: 03mattinm * r6382 10/gemrb/trunk/gemrb/plugins/Core/Actor.cpp: changed DamageType to weapon_damagetype[DamageType]
[08:33:47] <mattinm> ok, time to go do some more research
[08:33:59] <mattinm> i can say with certainty that spells in BG2 are on their own round timer
[08:34:30] <mattinm> and auto-pause: end of round pauses when a spell is cast and when the combat round is up, which can get annoying fast
[08:34:42] <mattinm> i'm gonna go check in bg1 now
[08:34:45] <-- Gekz_ has left IRC (Read error: 104 (Connection reset by peer))
[08:36:52] --> Gekz has joined #GemRB
[08:49:05] <mattinm> http://www.sorcerers.net/Games/BG2/SpellsReference/GeneralSpellInfo.htm#CastAndAttack
[08:49:11] <mattinm> alrighty, i'm off to bed :)
[09:11:54] <fuzzie> interesting :)
[09:13:39] <fuzzie> the rest of that page is interesting too!
[09:18:24] <fuzzie> gives even more evidence to pst rounds being 2s. bla. i'll go have a play myself.
[09:21:25] <CIA-2> gemrb: 03lynxlupodian * r6383 10/gemrb/trunk/NEWS: /me picks up the lyre and sings a ballad of NEWS
[09:22:42] <fuzzie> you still didn't mention your 'actually being able to save' fixes :)
[09:24:01] <fuzzie> the combat work really was been quite amazing
[09:26:13] <lynxlynxlynx> it's under "bugfixes"
[09:26:32] <lynxlynxlynx> i generalise a lot to keep the blob short
[09:27:05] <fuzzie> lynxlynxlynx: Do you know if gemrb saving works in iwd2?
[09:27:24] <lynxlynxlynx> i think so, let me check
[09:27:44] <fuzzie> I tried saving a new game and it wasn't loadable, but maybe gemrb's initial state is just broken.
[09:28:50] <lynxlynxlynx> i can save and load just fine, but i started with an original save
[09:29:11] <lynxlynxlynx> we know for sure something is wrong with our initial state
[09:29:22] <lynxlynxlynx> you can't use any of the common windows
[09:29:32] <fuzzie> Yes, I thought it might just be a GUI bug :/
[09:29:52] <lynxlynxlynx> it works for original saves, so i find that unlikely
[09:30:08] <fuzzie> Well, I mean, a GUI setup problem. Weird.
[09:30:11] <lynxlynxlynx> better chances of chargen not being complete
[09:30:59] <fuzzie> I'll try looking at Edheldil's save stack sometime, I guess.
[09:31:50] <fuzzie> Combat/etc seems fairly safe right now in mattinm's hands, which is nice :)
[09:57:05] <-- Gekz has left IRC (Read error: 104 (Connection reset by peer))
[09:58:40] --> Gekz has joined #GemRB
[10:12:51] --> pupnik_ has joined #gemrb
[10:31:02] <-- pupnik has left IRC (Read error: 101 (Network is unreachable))
[10:59:13] <-- Gekz has left IRC (Read error: 113 (No route to host))
[11:02:54] --> Gekz has joined #GemRB
[11:35:31] <fuzzie> Okay, good afternoon!
[12:14:30] <-- pupnik_ has left IRC ("Leaving")
[12:15:12] --> pupnik has joined #gemrb
[12:31:11] <fuzzie> gemrb really uuses too much RAM :)
[12:35:09] <CIA-2> gemrb: 03fuzzie * r6384 10/gemrb/trunk/gemrb/plugins/Core/ (EventMgr.cpp EventMgr.h Window.cpp Window.h): seperate keyboard and mouse event focus
[12:35:10] <fuzzie> No more having to click on the game control before pressing keys!
[12:36:45] <fuzzie> same for focus in dialogs and other screens, so I'm hoping I didn't break anything.
[12:44:05] <pupnik> excellent
[12:44:16] <pupnik> how much ram?
[12:44:32] <fuzzie> After moving through a few areas it's using 300mb.
[12:45:13] <fuzzie> Not so good if we want it to run well on devices with less RAM than that total. :) Probably just memory-leaking bugs, though.
[12:45:19] <pupnik> do you think it will be p.. ok
[12:46:18] <pupnik> next gen portables will typically have 256MB. problem is ram consumes power
[12:46:48] <fuzzie> I'm sure we can make it much lower than that.
[12:47:02] <pupnik> ok my disc is in place, can walk again, must clean house
[12:47:05] <pupnik> cheers
[12:47:07] <fuzzie> good luck :)
[13:03:24] <CIA-2> gemrb: 03fuzzie * r6385 10/gemrb/trunk/gemrb/plugins/Core/ (GameControl.cpp GlobalTimer.cpp): allow scrolling while paused
[13:35:44] <fuzzie> the fade stuff should perhaps be done while dialogs have scripts frozen too
[13:40:14] <Gekz> I see development
[13:40:14] <Gekz> xD
[13:41:49] <fuzzie> the real development has been people working on combat, it actually works now
[13:42:13] <Gekz> :o
[13:44:27] <fuzzie> gemrb is suddenly *playable*, it's scary.
[13:44:35] <CIA-2> gemrb: 03fuzzie * r6386 10/gemrb/trunk/gemrb/plugins/Core/ (GameControl.cpp GameControl.h): keep track of original dialog targets so we can link back to the right one
[13:54:34] <Gekz> fuzzie: wow
[13:54:37] <Gekz> very scary
[13:56:22] <CIA-2> gemrb: 03fuzzie * r6387 10/gemrb/trunk/gemrb/plugins/Core/Triggers.cpp: fix some item triggers to look inside bags
[13:56:35] <Gekz> fuzzie: you're rather awesome.
[14:03:23] <pupnik> gekz, were you looking at graphics for nonstandard resolutions last year?
[14:03:30] <pupnik> forgot who...
[14:04:09] <pupnik> fuzzie, what is ram usage when starting a game?
[14:05:09] <Gekz> pupnik: me
[14:06:42] <pupnik> think youre interested in starting on it again?
[14:06:51] <pupnik> maybe in the fall
[14:06:52] <Gekz> no
[14:06:54] <Gekz> because
[14:06:56] <Gekz> I have work to do
[14:06:58] <Gekz> shitloads of it
[14:07:00] <pupnik> k
[14:07:04] <Gekz> right now I'm meant to be writing an english essay
[14:07:06] <Gekz> but guess what
[14:07:11] <Gekz> I hate Hendrik Ibsen
[14:18:16] <fuzzie> Can't work out why item equipping is completely broken in ToB. I figured some trigger was unequipping everything, but I don't see it.
[14:19:27] <fuzzie> pupnik: 40mb or so. Really it shouldn't need to be much higher than that.
[14:20:11] <Gekz> fuzzie: not even for caching?
[14:20:47] <pupnik> fuzzie, that is GREAT news
[14:21:21] <pupnik> anything under 170MB is doable for pandora and nokia N900
[14:21:26] <fuzzie> great :)
[14:21:31] <pupnik> the rest can swap
[14:21:40] <pupnik> maybe 200
[14:23:29] <fuzzie> A fight and a move to a second area uses 110mb. I expect the tilemap is the main culprit, and we can easily optimise that.
[14:26:20] <fuzzie> Huh, it's only the main character who can no longer equip anything.
[14:26:30] --- barra_library is now known as barraAway
[14:27:28] <pupnik> :/
[14:28:34] <fuzzie> Oh, there's a "remove all effects" modifier attached, heh.
[14:29:38] <Gekz> awesome
[14:46:41] <fuzzie> something does applyspell() with resetchr on the player, and the effect never goes away (indeed, it is set as permanent until dead).
[14:46:49] <fuzzie> can't work out what, though. dinnertime!
[14:50:27] <pupnik> bon appetit!
[16:02:47] <lynxlynxlynx> fuzzie: re: main char: yes, but that has maybe to do with the fact that he was the only traveller or did you try with a party?
[16:03:15] <lynxlynxlynx> nice work on the focus stuff :)
[16:10:49] <fuzzie> well, judging by the applyspell debug, it's probably applied to anyone in the party when you first talk to sarevok or osmething
[16:11:03] <fuzzie> but i can't see which script calls it :(
[16:11:34] <fuzzie> So I'll hope that someone else can work it out.
[16:13:09] <pupnik> can you leave a note in todo?
[16:15:11] <lynxlynxlynx> i'll ask edheldil if ie_sh supports decompiling
[16:15:31] <lynxlynxlynx> it would be useful to have a dump of all dialogs and scripts somewhere to grep
[16:16:19] <fuzzie> i think edheldil was working on it.
[16:18:53] <fuzzie> I'm sure weidu can do it, but I can't work out how to use weidu.
[16:22:36] <fuzzie> --list-files gives me BCS files, but how do you get it to extract those?
[16:23:29] <lynxlynxlynx> i've never used it, at least not directly
[16:28:16] <fuzzie> "./weidu.asm.exe --game ~/src/gemrb/bg2 --list-files | grep BCS | xargs ./weidu.asm.exe --game ~/src/gemrb/bg2" worked!
[16:29:53] <fuzzie> And it's in cut205a. But not in the copy on my Windows machine. Perhaps removed by the first patch?
[16:31:20] <fuzzie> In any case, it's just a ApplySpellRES("ResetCHR", Myself) on every member of the party.
[16:33:31] <fuzzie> And that just applies fx_current_hp_modifier as a 'permanent after death' effect with flag 0x20000, which is RemoveAllNonPermanentEffects.
[16:37:11] <lynxlynxlynx> wierd
[16:42:38] <fuzzie> Avenger added this code just to make ctrl-r work, so maybe it's not too well tested.
[16:42:57] <fuzzie> But it looks like everything is working as it should be, to me.
[16:45:28] <lynxlynxlynx> yes, the input is problematic
[16:50:19] <lynxlynxlynx> http://www.bioware.com/games/shadows_amn/support/patches/ ‹-- doesn't appear to be fixed by the patch / it is undocumented
[16:54:19] <fuzzie> well, it's a ToB bug, so :)
[16:57:47] <lynxlynxlynx> doh
[16:57:56] <fuzzie> ok, it is in the ToB patch
[16:58:15] <fuzzie> cabextract it, unshield the resulting cab file, new cut205a.bcs :)
[17:00:23] <fuzzie> there's nothing so serious in the log, though, so we're probably doing something wrong
[17:00:47] <lynxlynxlynx> Fixes bug where drow scimitars would kill bards instantly.
[17:02:03] <lynxlynxlynx> is the applyspell the only difference?
[17:03:03] <fuzzie> yes
[17:03:45] <fuzzie> the most suspicious patch log entry is the multi-class one
[17:04:41] <lynxlynxlynx> hlas?
[17:05:33] <fuzzie> hm, i guess the resetchr is only used in cut205a..
[17:08:12] <fuzzie> Maybe this is also deserving of a forum thread.
[17:08:12] <lynxlynxlynx> maybe we could just dump a dud "ResetCHR" into our override
[17:11:03] <fuzzie> well, I wonder if that might just cover up a bug :)
[17:11:28] <lynxlynxlynx> yeah, i'm just being practical
[17:11:42] <lynxlynxlynx> they had to have good reason to have put that in
[17:12:06] <fuzzie> I imagine they were trying to stop the Blindness effect from the battle from interfering.
[17:12:41] <lynxlynxlynx> or any other effect, since it reset everything
[17:13:12] <lynxlynxlynx> the question is why it has a permanent-after-death duration
[17:13:35] <lynxlynxlynx> it would be enough to be a normal use-once effect
[17:13:49] <fuzzie> it is strange :)
[17:18:06] <fuzzie> is "cutscenes should switch to the main game view" solved now?
[17:20:48] <lynxlynxlynx> i'll try that later
[17:20:58] <lynxlynxlynx> it was easy to reproduce
[17:21:07] <fuzzie> I'm just glancing through todo, wondering what might be fixed.
[17:30:05] <-- pupnik has left IRC (Read error: 110 (Connection timed out))
[17:31:46] <lynxlynxlynx> grr
[17:32:01] <lynxlynxlynx> i made a perfect blade and on the end of chargen, a crash
[17:36:00] <lynxlynxlynx> well, at least it is reproducible
[17:36:58] <lynxlynxlynx> nothing obvious
[17:37:34] <lynxlynxlynx> http://pastebin.ca/1455595
[17:38:58] <lynxlynxlynx> pArgs is null
[17:39:12] <fuzzie> everything is built up-to-date, not just Core?
[17:39:56] <lynxlynxlynx> heh, false lead, pArgs is set to null the line before
[17:40:07] <lynxlynxlynx> i didn't do a make clean if that's what you mean
[17:40:27] <lynxlynxlynx> let's see if a plain fighter works
[17:40:42] <lynxlynxlynx> yep
[17:40:47] <fuzzie> i did test chargen after my focus patches, but only with a plain fighter
[17:40:58] <lynxlynxlynx> i'll rebuild
[17:42:49] <fuzzie> it looks like corruption, though, at a glance
[17:45:49] <lynxlynxlynx> didn't help
[17:45:56] <fuzzie> can you give me reproduction instructions?
[17:46:05] <fuzzie> just making a random blade doesn't seem to do it.
[17:46:10] <lynxlynxlynx> i just created a human lg fighter
[17:46:24] <lynxlynxlynx> as fast as chargen can get
[17:46:50] <fuzzie> using enter, or another key?
[17:46:54] <lynxlynxlynx> enter
[17:47:01] <fuzzie> huh. i can't reproduce. :/
[17:47:01] <lynxlynxlynx> but hah, this time it worked
[17:48:23] <lynxlynxlynx> next time broke again
[17:48:41] <lynxlynxlynx> maybe i'm too fast
[17:48:49] <lynxlynxlynx> this time the bt is useless
[17:49:09] <fuzzie> are you doing anything over than clicking entries and then hitting enter to move on?
[17:49:19] <lynxlynxlynx> nope
[17:49:25] <fuzzie> and which chargen, soa?
[17:49:31] <lynxlynxlynx> tob
[17:49:36] <fuzzie> unpatched?
[17:49:45] <lynxlynxlynx> yes
[17:50:08] <fuzzie> do you have .chu files in your override/?
[17:50:39] <lynxlynxlynx> yes
[17:50:45] <lynxlynxlynx> guichap.chu guiconn.chu guimg.chu guirec.chu start.chu
[17:51:55] <fuzzie> and it segfaults where, at the end?
[17:51:58] <lynxlynxlynx> and the only code difference from HEAD is a printf
[17:52:23] <lynxlynxlynx> yes, when it should exit and enter the game
[17:55:09] <fuzzie> wow, gemrb loads an incredible number of useless files before entering chargen
[17:55:24] <lynxlynxlynx> like all the npcs
[17:55:28] <fuzzie> indeed
[17:58:53] <fuzzie> okay, valgrind reports nothing unusual, although it does always complain a bit about the python.
[17:59:06] <fuzzie> but if it's something in your override/ then I wouldn't be able to see it anyway.
[17:59:54] <fuzzie> hm
[18:01:19] <fuzzie> my svn is *not* equal to head, let me fix that
[18:01:43] <fuzzie> this might be the fix, it couuld lead to corruption
[18:03:04] <CIA-2> gemrb: 03fuzzie * r6388 10/gemrb/trunk/gemrb/plugins/Core/DataStream.cpp: forgot to commit this, oops: always terminate resrefs
[18:06:00] <fuzzie> if that doesn't work then i'll have a look through the commits, see if i made any other stupid mistakes recently :)
[18:08:02] <lynxlynxlynx> let's see
[18:08:11] --> Edheldil has joined #gemrb
[18:08:11] --- ChanServ gives channel operator status to Edheldil
[18:08:44] <lynxlynxlynx> first try: succes
[18:09:03] <lynxlynxlynx> second too
[18:09:06] <lynxlynxlynx> time for a blade
[18:10:13] <lynxlynxlynx> huh
[18:10:25] <lynxlynxlynx> half-orcs have minimum int 11 and cha 15?
[18:10:48] <lynxlynxlynx> looks like we're reading the wrong row
[18:10:48] --- barraAway is now known as barra_library
[18:14:49] <lynxlynxlynx> oh, must've been the class
[18:18:30] <lynxlynxlynx> fuzzie: your fix for cutscene vs windows improves things, but i still see the pick-it-up inventory instead of the action bar on some redraws during dialogs (and when choosing answers)
[18:19:08] <lynxlynxlynx> meh, another crash
[18:19:27] <lynxlynxlynx> pretty horrible last three days
[18:24:43] <fuzzie> if you had a container/pile open on area change, that's a long-standing crash
[18:25:55] <lynxlynxlynx> can't we close those too?
[18:26:04] <fuzzie> yes, but i don't know where to do it
[18:27:19] <fuzzie> there are related bugs, like a dialog start should force-switch back to the main game view
[18:27:22] <fuzzie> otherwise that can cause crashes
[18:31:36] <lynxlynxlynx> hmpf, everyone is invulnerable in this iteration of illasera an co
[18:32:56] <lynxlynxlynx> wierd
[18:33:19] <lynxlynxlynx> i cured myself with ^r and then everyone was back to normal pokyness
[18:36:45] <fuzzie> that's because ctrl-r does the resetchr trick :p
[18:36:54] <fuzzie> or i would imagine so, anyway
[18:37:06] <lynxlynxlynx> but probably only on the target
[18:37:32] <fuzzie> one of illasera's spells negatively affects your rolls quite badly, i think
[18:37:43] <fuzzie> i mean, i'm just guessing at a cause :)
[18:38:10] <lynxlynxlynx> -4 for saves - the usual spell
[18:38:21] <lynxlynxlynx> console suggested i was landing hits just fine
[18:39:16] <fuzzie> the console output is before the to-hit rolls, i think
[18:39:31] <fuzzie> maybe that changed recently with the fixes
[18:39:47] <lynxlynxlynx> with 3 or 4 apr and 10min of playing, i'm pretty sure the odds are on my side
[18:40:47] <lynxlynxlynx> i was invulnerable too
[18:41:56] <lynxlynxlynx> looks like the flail of ages is problematic
[18:42:12] <lynxlynxlynx> when i had something else equipped, there was no problem
[18:42:42] <fuzzie> maybe get mattinm to have a look
[18:43:36] <fuzzie> http://fuzzie.org/nfs/gemrb/tooltip.jpg <- getting there..
[18:43:50] <fuzzie> annoyingly mispositioned though
[18:44:30] <lynxlynxlynx> doesn't look far off
[18:54:13] <fuzzie> do you know how i add strrefs to gemrb?
[18:54:22] <fuzzie> just put them in all the 2da files and the .h file?
[18:54:49] <lynxlynxlynx> yeah
[18:55:11] <lynxlynxlynx> you can look at 294633a4ef50be93d for an example
[18:55:52] <fuzzie> git-svn revisions are random, it seems
[18:56:21] <lynxlynxlynx> 6361
[18:56:30] <fuzzie> thanks
[19:00:33] <wjp> not random, but they do depend on how much history you have imported
[19:01:04] <wjp> and the svn username <-> name/email map you use
[19:01:07] <fuzzie> ah
[19:02:05] <-- barra_library has left IRC ("Verlassend")
[19:14:34] --> barra_library has joined #gemrb
[19:26:18] <fuzzie> bg1's really improved from the new combat code, too
[19:31:34] <CIA-2> gemrb: 03fuzzie * r6389 10/gemrb/trunk/gemrb/ (9 files in 8 dirs): implement actor tooltips
[19:31:58] <lynxlynxlynx> :D
[19:32:00] <fuzzie> tested in bg1, bg2, pst, iwd2, and all seems ok
[19:33:15] <lynxlynxlynx> only bg2 looks ok from the commit
[19:33:30] <lynxlynxlynx> i can find the missing strings though, except for bg1
[19:33:41] <fuzzie> there are no strings in bg1/pst afaik
[19:33:53] <fuzzie> and i couldn't find any in iwd2
[19:34:23] <lynxlynxlynx> oh, what's the dialogf.tlk in pst?
[19:34:23] <fuzzie> pst is special-cased and the other games fall back to bg1 behaviour
[19:34:45] <fuzzie> lynxlynxlynx: identical to dialog.tlk
[19:34:57] <lynxlynxlynx> ok
[19:36:37] <fuzzie> but feel free to fiddle with it, i'm done for now :)
[19:37:26] <lynxlynxlynx> i've got a nice oneliner for string search, but it takes long to extract
[19:37:34] <lynxlynxlynx> i should just dump them somewhere
[19:59:45] <lynxlynxlynx> they're around 2944 elsewhere
[20:02:32] <fuzzie> in which game?
[20:02:46] <lynxlynxlynx> all iwds
[20:02:58] <lynxlynxlynx> so also likely in bg1
[20:03:14] <fuzzie> well, i tried real bg1 and i didn't get the injured text at all
[20:03:38] <lynxlynxlynx> ok, makes it even easier
[20:05:03] <fuzzie> String #2943 is ~Uninjured~
[20:05:05] <fuzzie> String #2944 is ~Barely injured~
[20:05:06] <fuzzie> ^- iwd2
[20:05:11] <lynxlynxlynx> yep
[20:05:15] <lynxlynxlynx> same in all three
[20:05:55] <fuzzie> fuzzie@alyssa:~/src/weidu/weidu$ ./weidu.asm.exe --game ~/src/gemrb/bg1/ --string 2943
[20:05:58] <fuzzie> String #2943 is ~Right then! Thank you for the information, but you'd best be off home. I'll take care of this lot.~
[20:06:19] <fuzzie> looks fine for the iwds though, as you say
[20:08:27] <fuzzie> '--traify-tlk' gives a beautiful list for grepping
[20:09:11] <CIA-2> gemrb: 03lynxlupodian * r6390 10/gemrb/trunk/gemrb/override/ (how/strings.2da iwd/strings.2da iwd2/strings.2da): added injury levels to icewind dales
[20:10:51] <lynxlynxlynx> now to try this
[20:11:56] <lynxlynxlynx> feels unfamiliar, i'm not sure they were displayed over the head
[20:13:16] <lynxlynxlynx> nice, it updates during combat :)
[20:13:41] <fuzzie> works fine in iwd2
[20:14:35] <lynxlynxlynx> iwd too
[20:15:05] <fuzzie> they're probably not displayed overhead
[20:15:30] <fuzzie> i imagine the -size thing should be removed, i just can't check atm
[20:16:31] <fuzzie> and there's still varied update bugs but for another day :)
[20:16:50] <lynxlynxlynx> the tooltip update nicely for me
[20:19:33] <lynxlynxlynx> bg2 displays them at the cursor position
[20:19:44] <lynxlynxlynx> the cursor is hidden for that time
[20:20:13] <lynxlynxlynx> it also shows the injury level for neutrals
[20:20:39] <fuzzie> hm
[20:20:57] <fuzzie> which neutrals?
[20:21:06] <fuzzie> i tried some and it didn't give me anything except the name
[20:21:12] <lynxlynxlynx> blue circled folk
[20:21:20] <fuzzie> well, there are different kinds :)
[20:21:24] <lynxlynxlynx> neutral in the affinity sense, not alignment
[20:21:28] <fuzzie> i tried just checking EA_NEUTRAL
[20:21:34] <fuzzie> maybe it is more subtle.
[20:21:49] <lynxlynxlynx> getting another sample
[20:23:17] <lynxlynxlynx> not for a peasant in docks
[20:23:42] <lynxlynxlynx> the ones that had it were somehow proper npcs, the allies in the bodhi fight
[20:24:17] <fuzzie> Yes, those should show injured strings. Summoned creatures, too.
[20:25:05] <fuzzie> I think it's probably a gemrb bug if allies are EA_NEUTRAL..
[20:27:28] <lynxlynxlynx> i was checking in the original
[20:27:41] <lynxlynxlynx> nobody in docs shows the level
[20:28:00] <fuzzie> I mean, my tooltip code shows injury for everyone except EA_NEUTRAL chars.
[20:28:59] <lynxlynxlynx> yes, but i was only testing in ie, so ie shows it for some neutrals if IE_NEUTRAL is what dictates the ground circle
[20:29:11] <fuzzie> it isn't
[20:29:39] <fuzzie> oh, nm, maybe it is :)
[20:30:09] <fuzzie> gemrb's Actor::SetCircleSize does green circles for EA_ALLY
[20:38:48] <fuzzie> So if you save in IE and load in gemrb, it's broken?
[20:39:21] <lynxlynxlynx> haven't even tried
[20:44:03] <fuzzie> ok, was just wondering if i should be debugging
[20:46:13] <lynxlynxlynx> we don't understand what triggers the display for some of the neutrals, so there's not much to debug
[20:46:29] <lynxlynxlynx> you could remove the height addition though
[20:47:07] <fuzzie> ok, let me try and see if just removing it gets it right
[20:48:43] --> Avenger has joined #gemrb
[20:48:47] --- ChanServ gives channel operator status to Avenger
[20:49:03] <Avenger> fuzzie, i think your resref change caused this bug:
[20:49:08] <fuzzie> hm, it's pretty much right
[20:49:22] <Avenger> Traceback (most recent call last):
[20:49:23] <Avenger> File "./GUIScripts/bg2/CharGen9.py", line 40, in NextPress
[20:49:25] <Avenger> FinishCharGen()
[20:49:27] <Avenger> File "./GUIScripts/bg2/CharGen9.py", line 186, in FinishCharGen
[20:49:29] <Avenger> if inventory_exclusion & item['Exclusion']:
[20:49:34] <fuzzie> Avenger: you're using latest svn?
[20:49:48] <Avenger> depends what's latest, some 10 minutes old ;)
[20:50:03] <Avenger> lemme see again
[20:50:12] <lynxlynxlynx> the original centers the tooltip at the cursor position, so it should be easy to place it
[20:50:21] <fuzzie> lynxlynxlynx: yes, turns out you can just remove the height
[20:50:39] <fuzzie> Avenger: r6388 is the important one to have
[20:51:02] <Avenger> i got it, but i might forgot to actually perform make install
[20:51:25] <Avenger> i tried to create a fighter cleric dwarf in ToB
[20:51:34] <fuzzie> yes, it was crashing before 6388 :)
[20:51:35] <CIA-2> gemrb: 03fuzzie * r6391 10/gemrb/trunk/gemrb/plugins/Core/GameControl.cpp: don't position tooltips above actors
[20:51:49] <fuzzie> really stupid mistake on my part, i just forgot to commit the fix
[20:52:00] <Avenger> hmm, if you just put a zero in the end, this isn't correct
[20:52:25] <fuzzie> ieResRef is 9 bytes long, and code does strlen() on it, so it must be null-terminated
[20:52:29] <Avenger> you should also remove all trailing spaces
[20:52:40] <fuzzie> the code already removes trailing spaces
[20:53:14] <Avenger> ok i'm just trying it again
[20:53:22] <fuzzie> i think that is a bug, the original engine does not like nulls at all, but i didn't change it for now.
[20:53:52] <Avenger> [KEYImporter]: Searching for CHAN09.itm...[FOUND]
[20:53:54] <Avenger> [KEYImporter]: Searching for * .itm...[ERROR]
[20:53:56] <Avenger> Traceback (most recent call last):
[20:53:57] <Avenger> File "./GUIScripts/bg2/CharGen9.py", line 40, in NextPress
[20:53:59] <Avenger> FinishCharGen()
[20:54:00] <Avenger> File "./GUIScripts/bg2/CharGen9.py", line 186, in FinishCharGen
[20:54:02] <Avenger> if inventory_exclusion & item['Exclusion']:
[20:54:03] <Avenger> TypeError: 'NoneType' object is unsubscriptable
[20:54:05] <Avenger> so, no, this is still buggy
[20:54:37] <fuzzie> that is strange. your 'svn diff' is empty?
[20:54:53] <Avenger> yes
[20:54:54] <lynxlynxlynx> it worked for me too
[20:55:06] <Avenger> hmm, wait, i got gamecontrol
[20:55:12] <lynxlynxlynx> but let's try that dwarf f/c
[20:55:14] <Avenger> but that was r6391 :)
[20:55:43] <fuzzie> yes, that is definitely broken, it should strip all spaces
[20:55:52] <Avenger> what did i said :D
[20:56:08] <lynxlynxlynx> creation was fine here
[20:56:11] <fuzzie> but it works fine here
[20:56:13] <Avenger> heh
[20:56:15] <fuzzie> so i'm not really sure what to suggest
[20:56:16] <Avenger> odd
[20:56:37] <fuzzie> maybe that comes from a 2da and not via resref?
[20:57:21] <Avenger> that comes from a 2da
[20:57:37] <Avenger> the starting item 2da, forgot its name
[20:58:16] <fuzzie> abstart maybe
[20:58:21] <Avenger> no
[20:58:22] <fuzzie> or 25stweap.2da or similar
[20:58:26] <Avenger> yes
[20:59:12] <fuzzie> in any case i guess the bug is that the python should not be trying to make items from * fields :)
[20:59:20] <lynxlynxlynx> no
[20:59:23] <Avenger> it worked before
[20:59:24] <lynxlynxlynx> it avoids that
[20:59:33] <lynxlynxlynx> but he gets " *" for some reason
[20:59:39] <lynxlynxlynx> so the != * check fails
[20:59:45] <Avenger> "* "
[20:59:52] <lynxlynxlynx> whatever :P
[20:59:53] <Avenger> but it stripped those spaces before
[21:00:24] <Avenger> i don't know when
[21:00:38] <fuzzie> how do you reproduce it, which class/kit?
[21:00:48] <Avenger> just a fighter/cleric dwarf
[21:00:51] <lynxlynxlynx> [KEYImporter]: Searching for CHAN09.itm...[FOUND]
[21:00:53] <lynxlynxlynx> [KEYImporter]: Searching for HELM07.itm...[FOUND]
[21:00:54] <Avenger> but i had some fixpack
[21:01:00] <lynxlynxlynx> from my fighter/cleric dwarf
[21:01:07] <lynxlynxlynx> hmm
[21:01:10] <fuzzie> maybe the fixpack added spaces to the end of the 2da :)
[21:01:16] <lynxlynxlynx> do you have a 25weaps override?
[21:01:20] <Avenger> yes
[21:01:45] <Avenger> no spaces on end of rows
[21:02:08] <Avenger> no tabs in the file, just spaces
[21:02:13] <Avenger> this should work
[21:04:50] <Avenger> i don't know what happens here
[21:05:02] <fuzzie> could you upload/email the 2da?
[21:06:14] <fuzzie> the code there is pretty simple, it just reads the 2DA directly
[21:06:41] <Avenger> this fixpack sucks, it creates lots of files in the override
[21:07:02] <fuzzie> well, the official patch does too
[21:07:08] <Avenger> they could have repackaged it to bifs
[21:07:17] <Avenger> there is a weidu command for that
[21:07:33] <fuzzie> it turns out that original ToB has a cutscene which casts 'resetchr.spl' on all party memebers, and resetchr has one effect, which is a permanent-until-dead "remove all effects" hp modifier
[21:08:05] <fuzzie> and it is quietly removed from the cutscene in the patch without any real comment :)
[21:08:29] <fuzzie> and for finding things like that it is nice to have things obvious in override
[21:12:00] <Avenger> it slows down everything, dltcep, the game, or just browsing the file for the mail :)
[21:16:12] <Avenger> got the file?
[21:17:41] <fuzzie> yes
[21:18:22] <fuzzie> still can't reproduce
[21:18:28] <fuzzie> [KEYImporter]: Searching for 25stweap.2da...[Found in Override]
[21:18:28] <fuzzie> [KEYImporter]: Searching for CHAN09.itm...[FOUND]
[21:20:30] <lynxlynxlynx> grr
[21:21:46] <fuzzie> Avenger's copy has the LEATH14 fix, otherwise it is identical to mine
[21:22:06] <Avenger> heh
[21:22:17] <fuzzie> is this on Windows?
[21:22:34] <Avenger> linux 64 bit
[21:22:59] <Avenger> i have lots of regressions
[21:23:01] <fuzzie> Huh. Then I have no ideas. :/ The code definitely goes nowhere near the resref code I changed, though.
[21:23:35] <Avenger> 1. cannot load any tob games, 2. imoen and sarevok say their death string, and go yellow circled, but they are not hurt
[21:23:39] <Avenger> 3. this start new game bug
[21:24:03] <Avenger> i wanted to test the tob changes, but i cannot continue either way :)
[21:25:34] <Avenger> i wonder why my experience is different from yours :( i have only some window related changes withhold from svn
[21:25:36] <fuzzie> have you tried 'make clean'? maybe a stupid suggestion, but all i can think of is that your binaries are maybe corrupt somehow
[21:25:57] <lynxlynxlynx> if i just remove the resetchr calls from cut205a, of course my character remains blind, but the effect looks permanent :/ (nature's beauty?)
[21:25:59] <Avenger> i recompile
[21:26:08] <fuzzie> I *did* change window-related code today.
[21:26:22] <Avenger> that shouldn't cause any of these bugs
[21:26:30] <fuzzie> If your withheld code changes focus at all then it might be a problem, but I can't see how it would cause corruptin.
[21:27:05] --> pupnik has joined #gemrb
[21:27:14] <fuzzie> and I ran valgrind a few times today. I'll try it again.
[21:27:26] <Avenger> my changes were targeted at the gui flickering. bottom side of the portrait bar
[21:27:41] <Avenger> did you fix that?
[21:28:18] <lynxlynxlynx> nope
[21:28:21] <fuzzie> no, didn't change anything to do with that.
[21:28:25] <Avenger> i didnt manage to fix it, though i threw a lot of code on that :)
[21:28:35] <fuzzie> my local changes are just some built-in projectiles work
[21:28:43] <fuzzie> and i tried a build without it and it's still fine
[21:28:44] <pupnik> .
[21:32:35] <-- Avenger has left IRC ("bye!")
[21:32:49] <fuzzie> hm, i got a new valgrind complaint
[21:36:29] <lynxlynxlynx> i guess the patch also modified illasera's spell / effect to not last so long
[21:42:30] <fuzzie> i added some abort()s all over and none trigger, it is strange
[21:43:53] --> Avenger has joined #gemrb
[21:44:58] <Avenger> fuzzie: i had some corrupted 25stweap file in my gemrb override too. it had a default value * and a lot of spaces
[21:45:15] <fuzzie> aha :) but what about the other problems?
[21:45:42] <Avenger> well, i know lynx screwed the tob savegame dir :)
[21:46:07] <Avenger> the third bug is very odd, the yellow circle
[21:46:16] <Avenger> i don't know what causing that
[21:46:31] <fuzzie> is that in a new game, or a gemrb savegame?
[21:46:54] <Avenger> that's a savegame from the original engine
[21:47:56] <Avenger> even more strangely, it summons sarevok, and sarevok has some odd blur effect on hm
[21:48:04] <fuzzie> could be something corrupting the IE_EA stat, perhaps
[21:48:23] <Avenger> did you fix sparks?
[21:48:35] <Avenger> because cespenar now has nice shower type sparks on him
[21:48:46] <fuzzie> no, that is still your code :)
[21:48:59] <Avenger> odd, because it works :)
[21:49:55] <fuzzie> apart from the input/tooltips today, all i've done recently is bugfixes
[21:49:59] <Avenger> in windows version sarevok doesn't get yellow circle
[21:50:06] <Avenger> so this is 64 bit problem
[21:50:12] <Avenger> or uninitialized variable
[21:51:28] <Avenger> oh wait, this was the original game
[21:51:30] <Avenger> hehe
[21:51:40] <Avenger> i start to mix up the two
[21:51:45] <fuzzie> valgrind seems happy on 32-bit, except for all the usual python complaints
[21:52:44] <Avenger> ok, of course cespenar had the nice sparks in the original engine :(
[21:52:50] <-- lynxlynxlynx has left IRC (Remote closed the connection)
[21:53:24] <fuzzie> oh, and also mattinm's maxhpconbon is not so good.
[21:53:42] <fuzzie> a not-very-clever memset there, but it would not cause corruption.
[21:53:51] <Avenger> ok, the yellow circle happens here too
[21:54:24] <Avenger> when i quit from gemrb, i get corrupted heap warning too
[21:54:37] <fuzzie> are you able to run valgrind?
[21:54:44] <Avenger> not today
[21:54:57] <Avenger> i would need to reboot to linux
[21:55:11] <fuzzie> if you could email the savegame i could try reproducing some more
[21:55:21] <Avenger> it is in gamedata->ClearCaches
[21:55:45] <Avenger> free(wspattack[i]);
[21:55:54] <Avenger> this is some new stuff
[21:55:59] <fuzzie> heh
[21:56:01] <fuzzie> i'm looking at that now
[21:56:14] <fuzzie> it is maybe indeed broken
[21:56:41] <fuzzie> but it should be in your override for bg2..
[21:56:47] <Avenger> for (i=0; i<=wspattack_rows
[21:56:56] <Avenger> it should be only i<wspattack_rows
[21:56:58] <Avenger> i'm sure
[21:57:03] <fuzzie> yes
[21:57:24] <Avenger> check all those <= relations, i think they are all wrong
[21:57:38] <fuzzie> they are all wrong, and there are other bugs :)
[21:57:48] <fuzzie> so i guess this is the problem
[21:58:25] <fuzzie> oh, maybe not wrong. hm
[21:58:47] <Avenger> not ALL wrong
[21:58:54] <Avenger> wspecial_max = tm->GetRowCount()-1;
[21:58:57] <Avenger> this is fine
[21:59:19] <Avenger> because he then uses calloc(wspecial_max+1,
[21:59:35] <Avenger> but wspattack_rows is wrong
[21:59:39] <fuzzie> it is just wspattack which is broken
[22:00:05] <Avenger> ok, i gotta sleep
[22:00:09] <fuzzie> night!
[22:00:10] <Avenger> see you!
[22:00:12] <-- Avenger has left IRC ("ChatZilla 0.9.84 [Firefox 3.0.10/2009042316]")
[22:12:26] <fuzzie> hum, not so simple.
[22:37:16] <CIA-2> gemrb: 03fuzzie * r6392 10/gemrb/trunk/gemrb/plugins/Core/ (Actor.cpp Actor.h): don't refresh PC stats on non-PC chars (plus some alloc/free fixes)
[22:37:17] <fuzzie> mattinm: please review r6392 if you have a moment
[22:47:15] <CIA-2> gemrb: 03fuzzie * r6393 10/gemrb/trunk/gemrb/plugins/Core/GameControl.cpp: add actor tooltip positioning hack for PS:T
[22:56:53] <fuzzie> now i am really left with only python complaints in valgrind.
[22:56:58] <fuzzie> so i hope that fixes any corruption issues.
[23:10:16] * mattinm gets dinner and reads log
[23:14:17] <fuzzie> summary: flail of ages casues problems when equipped, I committed some new stuff, r6392 fixes some bugs in your code but perhaps introduces more, hence the review request.
[23:15:08] <mattinm> yep, just finished :)... it's odd, i have a character with the FoA, and it caused slow in the off-hand and main-hand; that's actually how i tested the dual-wield code :/
[23:18:07] <mattinm> r6392 look fine :)
[23:20:49] <fuzzie> oh, and combat in iwd2 is awfully bad :)
[23:20:56] <fuzzie> but i'm sure this is no surprise to anyone
[23:26:18] <mattinm> no surprise to me :)
[23:27:00] <mattinm> i *try* to keep my code iwd2 compatible, but when i can't test, it's somewhat difficult
[23:28:15] <fuzzie> I think the problem is more that it just didn't work before and it needs new code, to be honest.
[23:29:00] <mattinm> yeah; i'll do some more reading up on iwd2 tonight and try to see if i can see the problem
[23:29:19] <mattinm> or problems ;)
[23:34:48] <fuzzie> i wouldn't worry about it - i think your time might be better spent on what you're familiar with, you certainly seem to know bg2 well :)
[23:48:43] <jdm-> hee hee, I try out iwd2 every few days now to see how any changes have affected it
[23:48:48] <jdm-> and yeah, combat's pretty gross
[23:49:05] <jdm-> if you have anything you want me to test, let me know
[23:53:37] <-- jdm- has left IRC ("http://www.mibbit.com ajax IRC Client")