#gemrb@irc.freenode.net logs for 7 Apr 2012 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:01:29] <-- edheldil_ has left IRC (Ping timeout: 265 seconds)
[00:02:51] --> edheldil_ has joined #gemrb
[00:10:42] <-- SiENcE has left IRC (Quit: cya)
[00:31:53] <-- edheldil_ has left IRC (Ping timeout: 260 seconds)
[03:02:42] <nutron> DrMcCoy: :o The Witcher! =)
[03:02:58] <nutron> though the renderer is apparently... not the same.... so maybe not.
[03:19:24] --> joneirik has joined #gemrb
[04:51:48] <-- joneirik has left IRC (Remote host closed the connection)
[05:49:17] <-- brad_a has left IRC (Quit: brad_a)
[06:27:39] --> brad_a has joined #gemrb
[06:33:08] --> edheldil_ has joined #gemrb
[06:39:05] <-- brad_a has left IRC (Quit: brad_a)
[06:59:42] --> Yoshimo has joined #gemrb
[07:09:56] --> Avenger has joined #gemrb
[07:11:04] <Avenger> edheldil: strange, i could use the door now
[07:35:32] <-- edheldil_ has left IRC (Ping timeout: 246 seconds)
[07:44:16] <DrMcCoy> nutron: Well, models are partially working already for The Witcher. Nothing else, though :P
[08:13:09] --> lynxlynxlynx has joined #gemrb
[08:13:10] <-- lynxlynxlynx has left IRC (Changing host)
[08:13:10] --> lynxlynxlynx has joined #gemrb
[08:13:10] --- ChanServ gives channel operator status to lynxlynxlynx
[08:35:38] <lynxlynxlynx> Avenger: did you see that tot thread?
[08:36:12] <Avenger> tot thread?
[08:36:40] <lynxlynxlynx> http://forums.gibberlings3.net/index.php?showtopic=24282
[08:36:58] <Avenger> not yet
[08:37:31] <lynxlynxlynx> if i understand correctly, there is no signature involved and we check and fail that test needlessly
[08:38:25] <Avenger> i thought you wanted to show me the segment stuff (which we have)
[08:39:32] <lynxlynxlynx> no, i saw that
[08:41:55] <Avenger> so all you say, that checking for the 0xffff stuff is wrong because they are offsets that could be different? I agree with that.
[08:42:54] <Avenger> i didn't see any example where the 0 offset was different, but it could be
[08:47:32] <Avenger> drmccoy: i didn't play ME3 for long. I doubt i will play any future bioware game. Their direction is totally wrong for me.
[08:48:47] --- ChanServ gives channel operator status to Avenger
[08:49:29] <lynxlynxlynx> well, maybe we just don't write it properly, since i know the tlk override is not complete yet
[08:49:45] <lynxlynxlynx> but i see that sig error all the time
[08:53:59] <Avenger> try to remove the check maybe it will work then
[08:58:25] <lynxlynxlynx> it's not fatal, just some spam
[09:01:14] <Avenger> it returns false on init
[09:01:17] <lynxlynxlynx> seems the default biography strings are hardcoded somewhere
[09:02:10] <Avenger> we talk about iwd/how ?
[09:02:57] <lynxlynxlynx> iwd2
[09:04:17] <Avenger> i already wrote a bios.2da for iwd2
[09:04:30] <Avenger> it is hardcoded in the engine, but already recovered
[09:04:52] <Avenger> is that what you needed?
[09:05:05] <fuzzie> it's no fun looking for hardcoded stuff in the original engine, when you already un-hardcoded everything :P
[09:05:17] <Avenger> like anyone else would do it :P
[09:05:33] <Avenger> but, there are lots of things to recover
[09:05:45] <Avenger> proper avatars.2da for every game :D
[09:06:15] <fuzzie> well, 'fix avatars' and 'fix scripting' are on my TODO list to get around to someday
[09:06:30] <Avenger> also, i don't know how spontaneous casting in iwd2 works, if someone wants a smaller task
[09:06:48] <fuzzie> right now I am buried in spectral decompositions though :/
[09:07:09] <Avenger> you work during the holidays?
[09:07:22] <fuzzie> I have an exam on Tuesday morning.. :-)
[09:07:49] <lynxlynxlynx> there's a sponcast.2da that has all the needed info
[09:08:00] <lynxlynxlynx> evil alignment -> hurt, rest -> heal
[09:08:24] <Avenger> i never used it
[09:08:27] <lynxlynxlynx> i'll go check the tot thing now
[09:08:49] <Avenger> is that 'right click' --> replace spell with alignment specific one?
[09:08:50] <lynxlynxlynx> you shift-pressed when spellcasting and it was converted before the targeting phase
[09:09:00] <Avenger> i see, sounds like easy
[09:09:05] <lynxlynxlynx> yep
[09:09:19] <Avenger> we could even add that as a feature to other games :)
[09:11:58] <lynxlynxlynx> :)
[09:33:59] --> alx3apps has joined #gemrb
[10:18:08] <-- Avenger has left IRC (Quit: bye!)
[10:22:04] --> Gekz_ has joined #gemrb
[10:22:05] <-- Gekz_ has left IRC (Changing host)
[10:22:05] --> Gekz_ has joined #gemrb
[10:23:08] <-- Gekz has left IRC (Read error: Connection reset by peer)
[11:03:33] <CIA-28> GemRB: 03avenger_teambg * r9a46c990e66e 10gemrb/gemrb/core/GameScript/Actions.cpp: fixed bashing doors (now it works!)
[11:03:43] <CIA-28> GemRB: 03avenger_teambg * r68c3441ec8b2 10gemrb/gemrb/GUIScripts/GUIRECCommon.py: Merge branch 'master' of ssh://gemrb.git.sourceforge.net/gitroot/gemrb/gemrb
[11:05:38] <CIA-28> GemRB: 03avenger_teambg * r77b06e286a6f 10gemrb/gemrb/core/GameScript/ (GSUtils.cpp GSUtils.h): use the movement effect for delayed move to area
[11:05:46] --> Avenger has joined #gemrb
[11:05:46] --- ChanServ gives channel operator status to Avenger
[11:06:30] <Avenger> i hope the move to area effect works, it was still under testing, i just committed part of it by mistake :)
[11:06:47] <lynxlynxlynx> yep, was just about to complain
[11:07:23] <lynxlynxlynx> on the previous topic, is it possible my default.tot is already corrupted?
[11:07:27] <Avenger> the bash door fix involves only one function
[11:07:31] <lynxlynxlynx> i get invalid seeks when trying to write to it
[11:07:51] <Avenger> it could be that the offset is not calculated properly
[11:09:28] <lynxlynxlynx> that's what it looks like if i compare it with iesdp
[11:09:48] <lynxlynxlynx> trying to write the backref in the char array
[11:10:28] <Avenger> so it writes the offset at a bad position?
[11:11:24] <lynxlynxlynx> it looks like there is +/- error
[11:11:45] <lynxlynxlynx> line 176 of tlkoverride.cpp
[11:12:44] <Avenger> 176 is ieDword backp = 0xffffffff; for m
[11:12:45] <Avenger> me
[11:14:40] <lynxlynxlynx> oh
[11:14:55] <lynxlynxlynx> seek offset+4
[11:15:30] <lynxlynxlynx> but i see now it didn't give me a good new strref too, it's STRREF_START
[11:17:35] <lynxlynxlynx> AuxCount is 0
[11:17:55] <Avenger> it should have given a new one then
[11:18:38] <lynxlynxlynx> it's one of the bio strrefs, so maybe they're at different numbers in iwd2?
[11:18:55] <lynxlynxlynx> i thought these were reserved for this purpose
[11:21:35] <lynxlynxlynx> ah, they are out of the range, so we do need new ones
[11:22:28] <lynxlynxlynx> we should increase AuxCount beforehand then or check differently
[11:23:17] <Avenger> i don't understand
[11:23:32] <lynxlynxlynx> me neither yet, just thinking aloud
[11:23:39] <lynxlynxlynx> why is STRREF_START so big?
[11:24:37] <Avenger> updatestring returns the next string ref (newly allocated)
[11:24:43] <Avenger> if it gets an invalid strref
[11:25:42] <Avenger> strref start is so high because the lesser strrefs are not created by tlk override, they are reserved for the normal tlk table
[11:26:02] <Avenger> this value is from ToB, but it might be different in other games
[11:27:06] <lynxlynxlynx> 300k is good enough for everything with all the mods
[11:27:18] <lynxlynxlynx> default tob is like 60k
[11:27:40] <Avenger> there is also BIO_START and BIO_END to make things even more shitty
[11:28:06] <Avenger> those are overridden, despite they are <300k
[11:29:07] <lynxlynxlynx> exactly
[11:29:23] <lynxlynxlynx> anyway, that +/- thing didn't help
[11:29:37] <lynxlynxlynx> Invalid seek position 4 in file default.tot(limit: 0)
[11:29:40] <lynxlynxlynx> wierd limit
[11:30:30] <Avenger> there is a bug i think
[11:30:56] <Avenger> GetNewStrRef should honor the strref value of BIO_START - BIO_END
[11:31:12] <Avenger> it shouldn't create a new strref >300k
[11:31:17] <Avenger> it should use the given value
[11:32:59] <lynxlynxlynx> the update tries that, but since the strref doesn't exist, it goes for a new one
[11:39:14] <Avenger> see you later!
[11:39:16] <-- Avenger has left IRC (Quit: bye!)
[11:42:08] <lynxlynxlynx> doh
[11:42:22] <lynxlynxlynx> it's failing because the file was just created, so it has 0 length
[12:04:53] <-- edsrzf has left IRC (Ping timeout: 246 seconds)
[12:22:15] <CIA-28> GemRB: 03lynxlupodian * r34bc7926733a 10gemrb/gemrb/GUIScripts/iwd2/CharGen9.py:
[12:22:15] <CIA-28> GemRB: iwd2::cg: save the character bio
[12:22:15] <CIA-28> GemRB: (custom bio still doesn't work though)
[12:22:25] <CIA-28> GemRB: 03lynxlupodian * r8a86df767252 10gemrb/gemrb/plugins/TLKImporter/TlkOverride.cpp:
[12:22:25] <CIA-28> GemRB: don't check the tot "signature" or the whole tlk override is disabled if it
[12:22:25] <CIA-28> GemRB: doesn't exist already
[12:22:25] <CIA-28> GemRB: 03lynxlupodian * r112ac22bf880 10gemrb/gemrb/ (5 files in 4 dirs): fixed compilation with gcc 4.7
[13:36:35] --> SiENcE has joined #gemrb
[14:49:41] <-- alx3apps has left IRC (Quit: Leaving.)
[15:04:23] <-- Yoshimo has left IRC (Quit: Yoshimo)
[15:13:34] --> edheldil_ has joined #gemrb
[15:20:03] --> Yoshimo has joined #gemrb
[15:49:49] <-- wrotek__ has left IRC (Read error: Connection reset by peer)
[15:50:36] --> lowb has joined #gemrb
[16:03:01] <-- lowb has left IRC (Quit: Page closed)
[16:13:53] --> Avenger has joined #gemrb
[16:13:53] --- ChanServ gives channel operator status to Avenger
[16:14:13] <Avenger> lynx, why don't you simply use if EditControl: ?
[16:15:31] <Avenger> that type thing is probably not working well, if the control referenced by editcontrol is destructed, its type won't be int, but it will be correct with if EditControl:
[16:28:25] <-- edheldil_ has left IRC (Ping timeout: 246 seconds)
[16:31:14] --> edheldil_ has joined #gemrb
[16:31:34] <lynxlynxlynx> i did the easy thing
[16:31:43] <lynxlynxlynx> if editcontrol is valid, it will be an object
[16:32:29] <lynxlynxlynx> it was already set up, but i didn't bother checking if it should be delayed or nulled or something else to make it work for all cases
[16:33:25] <lynxlynxlynx> a bit safer would be to return if it is an int
[16:43:53] <Avenger> lol, in HoW, i didn't find the string 'to backpack' in dialog tlk. It turned out it is written as 'Backback'
[16:47:47] <CIA-28> GemRB: 03avenger_teambg * rf3d9bb57c2a1 10gemrb/gemrb/GUIScripts/GUISTORE.py: correct label for 'to backpack'
[17:07:26] --> brad_a has joined #gemrb
[17:19:23] <Avenger> testing Attack vs AttackReevaluate in tob engine, they really seem to be all the same. even a single Attack( [pc]) lets you switch targets switches to closer pc
[17:20:06] <Avenger> and the switching is unconditional, it doesn't count if i use 1 or 100 or 10000 in attackreevaluate
[17:27:24] <lynxlynxlynx> the disassambly isn't actually the same?
[17:29:49] <Avenger> no it isn't
[17:34:14] <fuzzie> um
[17:34:33] <fuzzie> i'm pretty sure they're not the same
[17:36:27] <fuzzie> but why would you use [PC] to test..?
[17:37:01] <fuzzie> that's ids filtering, that should be instantly switched if it changes, for any action
[17:37:02] <Avenger> because it is easy to build the test environment
[17:37:12] <fuzzie> do you not see it instantly switching?
[17:37:24] <Avenger> i see it instantly switching yes.
[17:37:28] --> wrotek has joined #gemrb
[17:37:36] <fuzzie> the target storage is for filters
[17:37:40] <fuzzie> like LastAttackerOf(Myself)
[17:37:48] <Avenger> But i thought AttackReevaluate would switch only the given period
[17:37:56] <Avenger> only after
[17:37:58] <fuzzie> only if it's a stored target like a filtered one
[17:38:18] <fuzzie> [PC]-style ids filtering is reevaluated every time
[17:38:36] <Avenger> then i wonder why the original game is full with scripts like that
[17:38:44] <fuzzie> why not?
[17:38:54] <Avenger> IF
[17:38:56] <Avenger> See([0.0.0.TANARI])
[17:38:58] <Avenger> THEN
[17:38:59] <Avenger> RESPONSE #100
[17:39:01] <Avenger> AttackReevaluate([0.0.0.TANARI],30)
[17:39:03] <Avenger> END
[17:39:16] <fuzzie> heh :)
[17:39:20] <Avenger> this is the same as plain attack, then
[17:39:52] <fuzzie> yes
[17:40:17] <fuzzie> to test, put a See([PC]) as trigger, and then AttackReevaluate the LastSeenBy(Myself)
[17:40:47] <Avenger> btw, noaction really doesnt' stop modal, but displaystringhead does
[17:40:52] <fuzzie> *nod*
[17:40:59] <Avenger> i always use displaystringhead, that's why i got confused
[17:41:26] <fuzzie> but yes, bioware's scripting is weird ..
[17:54:41] <Avenger> ok, this is entirely different
[17:54:57] <Avenger> but i couldn't even make it switch targets with attackreevaluate
[17:55:21] <fuzzie> I did test this all as working fine in ToB, I promise.. :)
[17:55:27] <fuzzie> It is a bit tricky though.
[17:55:36] <Avenger> IF
[17:55:38] <Avenger> Global("attacktype","global",1)
[17:55:39] <Avenger> See([PC])
[17:55:41] <Avenger> THEN
[17:55:43] <Avenger> RESPONSE #100
[17:55:44] <Avenger> AttackReevaluate(LastSeenBy(Myself),30)
[17:55:46] <Avenger> END
[17:55:54] <Avenger> this never seemed to switch once i set attacktype
[17:56:06] <fuzzie> once you set it to 1?
[17:56:09] <Avenger> yes
[17:56:13] <fuzzie> that looks like it should work
[18:00:02] <Avenger> lol, never switches
[18:01:29] <fuzzie> weird
[18:01:46] <fuzzie> as long as the See trigger gets re-evaluated, it should switch every 30 ticks
[18:06:39] <Avenger> ok, attackreevaluate with 1 works like before when i used [pc]
[18:06:46] <Avenger> that seems to be correct
[18:07:22] <Avenger> yep, it lets me ditch my attacker very easily
[18:18:34] <-- Avenger has left IRC (Quit: bye!)
[18:24:09] <edheldil_> Avenger: the travel through that door still does not work
[18:30:46] --> Avenger has joined #gemrb
[18:30:47] --- ChanServ gives channel operator status to Avenger
[18:31:02] <Avenger> ed: how come, i tried it just now.
[18:31:31] <Avenger> You need to use the lever after you compiled the update
[18:32:39] <Avenger> if you just reload the saved game from after you used the lever, it won't work of course
[18:33:40] <-- Avenger has left IRC (Client Quit)
[19:16:17] --> Abbedied has joined #gemrb
[19:21:52] <lynxlynxlynx> thomas_sch: ping :)
[19:32:56] <-- Abbedied has left IRC (Ping timeout: 245 seconds)
[19:42:18] <edheldil_> Avenger: strange - on the second try it works (and I am not dumb to go through closed door, thank you)
[19:43:50] <edheldil_> but who did sabotage guiscripts? selecting PCs now does not display green rectangles around portraits and damage shows as a solid black rectangle :((
[19:45:36] <edheldil_> ah selection uses black rectangles
[19:48:12] --> Avenger has joined #gemrb
[19:48:20] <Avenger> yeah, someone sabotaged the portrait icon borders
[19:49:05] <-- SiENcE has left IRC (Ping timeout: 260 seconds)
[19:49:41] <Avenger> must have been lynx ;)
[19:49:52] <edheldil_> anyway, I am loathe to deem the bug solved, because it did not work on the first try...
[19:50:35] <Avenger> this was lynx changing rgb http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=112ac22bf88019e515b133bebfeb2fce434cb643
[19:51:13] <Avenger> you probably loaded an area where you already used the lever
[19:51:22] <fuzzie> um, yes, that is not a good commit
[19:51:39] <Avenger> huh, it is a stack overwrite!!!!
[19:51:40] <fuzzie> you can't change ints to bytes like that! :P
[19:51:47] <fuzzie> python will indeed stomp right over your stack there
[19:51:53] <fuzzie> since it was told to
[19:51:57] <Avenger> lynxlynxlynxlynxlynx :D
[19:51:59] <edheldil_> Avenger: no, don't use gemrb saves
[19:52:15] <lynxlynxlynx> oj :)
[19:52:33] <lynxlynxlynx> i thought it was doing the same as before
[19:52:45] <Avenger> ed: you can try it a third time
[19:52:59] <lynxlynxlynx> but i see i forgot the cast
[19:53:02] <Avenger> 2 vs 1 would be decisive :D
[19:53:04] <fuzzie> you can't cast them
[19:53:14] <fuzzie> if you pass 'i' to python, you have to give it an int*
[19:53:18] <Avenger> use ints for python, then cast them to byte
[19:54:11] <edheldil_> Avenger: already tried 6 times :)
[19:54:29] <Avenger> and? 5 vs 1?
[19:55:34] <edheldil_> 6 vs. 1
[19:55:58] <Avenger> that is better than 0 vs 7 that would have been before my change
[19:56:20] <Avenger> if it is still buggy, it could be some script trigger timing bug
[19:56:51] <Avenger> but that is different from sending a generic trigger/disabling the trap/then not sending the real trigger bcause trap was already triggered
[19:58:05] <-- brad_a has left IRC (Quit: brad_a)
[20:01:56] <edheldil_> ok, another two attempts... Seems to be fixed
[20:02:51] <edheldil_> btw, gemrb's fights are extremely static and rather boring
[20:03:17] --> SiENcE has joined #gemrb
[20:04:30] <edheldil_> another good gsoc project would be pathfinding with bumping. Those narrow corridors are annoying
[20:04:47] <Avenger> yep
[20:05:01] <Avenger> what makes the fights 'static and boring' ?
[20:05:09] <Avenger> the opponents don't use spells?
[20:05:18] <Avenger> or they don't switch target?
[20:05:54] <edheldil_> that too, but I would have to compare
[20:06:55] <Avenger> well, it is most likely scripting problem, lets hope fuzzie will have time eventually
[20:07:10] <edheldil_> one annoying problem is that the foes are not highlighted unless you pause, so it's hard to see whom do you fight. Moreover, if they are behind an obstacle, you do not see them at all
[20:07:33] <Avenger> isn't that an option?
[20:08:20] <edheldil_> if it is, GemRB does not use it, works for me with the orig
[20:08:44] <Avenger> set marker feedback to maximum
[20:09:00] <Avenger> maybe this is a value problem with the marker feedback
[20:09:14] <Avenger> hmm, no
[20:09:21] <Avenger> we don't honor it i guess
[20:13:21] <Avenger> well, only the 2 highest marker feedbacks show red circle when unpaused
[20:14:02] <Avenger> you also need to check 'dither always'
[20:19:20] <edheldil_> ok... I can't check the orig now. For some reason when I exit it, I am unable to start it again w/o restart
[20:20:46] <CIA-28> GemRB: 03avenger_teambg * r78d38085f760 10gemrb/gemrb/plugins/GUIScript/GUIScript.cpp: fixed Color byte vs int problems
[20:21:34] <Avenger> ok, border should be back
[20:24:31] <edheldil_> yes, damage is better as well
[20:40:06] --> edsrzf has joined #gemrb
[20:46:01] <-- gembot has left IRC (Ping timeout: 246 seconds)
[20:46:12] --> gembot has joined #gemrb
[20:52:08] <lynxlynxlynx> still playing iwd?
[21:04:29] <edheldil_> me? Not at the moment
[21:04:50] <edheldil_> but I can test with gemrb, if you need something
[21:06:50] <-- SiENcE has left IRC (Quit: cya)
[21:10:35] <-- edheldil_ has left IRC (Quit: Really?)
[21:12:01] --> edheldil_ has joined #gemrb
[21:12:02] <lynxlynxlynx> no, it was about the boring combat
[21:12:40] <lynxlynxlynx> that's pretty much the default in iwd up until the very late stages
[21:12:41] <edheldil_> I have just restarted so I can start the orig
[21:13:18] <lynxlynxlynx> but it is also true iwd opcodes are buggier than the rest in gemrb, since they get less testing
[21:32:24] <edheldil_> I checked the combat in orig. The PCs seem to be less lazy and more doing 'something'.
[21:33:22] <edheldil_> the thief is easily finding all traps in gemrb, unlike orig
[21:33:34] <fuzzie> we don't do the charbase animations yet I think.
[21:33:38] <fuzzie> so no fake animations filling the gaps.
[21:34:05] <edheldil_> yeah, quite possibly
[21:39:10] <edheldil_> loading game in gemrb, only thief is selected. In orig, whole party is
[21:42:04] <edheldil_> failed disarm report 'pickpocket failed'
[21:42:13] <edheldil_> reports
[21:44:19] <edheldil_> we don't honor 'dither always' or whatever should cause the foes not to be hidden behind obstacles
[21:44:43] <edheldil_> we do not highlight targets ...
[21:44:47] <edheldil_> etc
[21:52:46] <-- Yoshimo has left IRC (Quit: Yoshimo)
[21:55:42] <gembot> build #358 of nmake-msvc++10 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B10/builds/358 blamelist: Avenger
[21:56:20] <fuzzie> well disarm is presumably string error :)
[22:20:38] <-- wrotek has left IRC (Ping timeout: 240 seconds)
[22:33:20] <Avenger> fuzzie do you know why gembot blames me?
[22:34:22] <fuzzie> not your fault
[22:44:06] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:00:19] <CIA-28> GemRB: 03avenger_teambg * r0125e6e00980 10gemrb/gemrb/core/Scriptable/ (Actor.cpp Actor.h Scriptable.cpp Scriptable.h): implemented 'Always Dither' option (FIXME: drop all spritecovers when changing this option)
[23:00:19] <-- edheldil_ has left IRC (Read error: Connection reset by peer)
[23:00:22] <Avenger> ok edheldil Always Dither is implemented
[23:01:25] <-- Avenger has left IRC (Quit: bye!)
[23:34:53] <xrogaan> > https://twitter.com/#!/TrentOster/status/188769402706071552
[23:35:01] <xrogaan> they want LINUX experts
[23:35:07] <xrogaan> I say, just give them a hand
[23:35:21] <xrogaan> s/want/need/
[23:51:04] <gembot> build #359 of nmake-msvc++10 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B10/builds/359
[23:54:55] <gembot> build #635 of mingw32 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/mingw32/builds/635 blamelist: Avenger