#gemrb@irc.freenode.net logs for 6 Feb 2011 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[01:08:41] <-- barra_home has left IRC (Quit: Verlassend)
[04:08:02] <-- Maighstir_ has left IRC (Ping timeout: 240 seconds)
[05:19:20] --> Maighstir has joined #GemRb
[05:45:03] --> Elvashi has joined #GemRb
[06:10:53] <-- Elvashi has left IRC (Ping timeout: 255 seconds)
[06:25:53] <-- _pickle has left IRC (Remote host closed the connection)
[06:27:11] --> Elvashi has joined #GemRb
[06:31:06] --> raevol has joined #GemRb
[07:00:56] <-- raevol has left IRC (Ping timeout: 240 seconds)
[07:05:45] --> raevol has joined #GemRb
[07:14:13] --- Elvashi is now known as Textmode
[07:47:24] <-- raevol has left IRC (Quit: Leaving.)
[08:37:12] --> Avenger has joined #GemRb
[08:37:12] --- ChanServ gives channel operator status to Avenger
[08:37:20] <Avenger> hi
[08:38:31] * Textmode huggles Avenger
[08:38:43] <Avenger> today's first laugh for me: I pondered why some (but not all) points in the maze structure look like (-1,0) when (-1,-1) would be the totally correct form. Is there a secret meaning, or they just shot for the minimal good version.
[08:39:16] <Avenger> well, the answer is (as always) it's a bug: .text:00548A2D mov [edx+maze_struct.nordom_pos.x], 0FFFFFFFFh
[08:39:18] <Avenger> .text:00548A37 mov eax, [ebp+var_14]
[08:39:19] <Avenger> .text:00548A3A mov [eax+maze_struct.nordom_pos.x], 0FFFFFFFFh
[08:39:21] <Avenger> .text:00548A44 mov ecx, [ebp+var_14]
[08:40:14] <Textmode> yay! bugs!
[08:40:41] <Avenger> can you spot it? :)
[08:41:09] <Textmode> my ASM is weak...
[08:41:32] <Avenger> ah come on, you should be better than our average BIS programmer
[08:43:09] <Textmode> by "weak" I mean, "I always meant to learn ASM, but never did"
[08:43:18] <Textmode> I don't know what those registers are, for one.
[08:52:20] --> jschall has joined #GemRb
[08:53:35] <fuzzie> hehe
[08:55:29] <Avenger> the registers are there only to confuse you, the actual important fields are in text form :) the solution is, they are the same due to cut&paste
[08:55:58] <fuzzie> they're always the same due to cut&paste :P
[08:56:16] <fuzzie> they need more imaginative bugs
[08:56:29] <Avenger> i know it is, because 2 of 5 of these point initialisations are buggy :)
[08:57:58] <Textmode> ah right, pos.x, pos.x...
[08:58:19] <Textmode> so...what happens if you fix that? I'm guessing somethign breaks :P
[08:58:36] <Avenger> no, their code later checks if either x or y is -1
[08:58:47] <Avenger> it doesn't affect anything
[08:58:54] <fuzzie> it's just horrible
[08:59:02] <Textmode> heh
[08:59:10] <Avenger> except me pondering a lot after i looked at the .gam listing first
[09:00:41] <Avenger> you cannot be sure until you see the code, like above. Until that, it could have been some special stuff we need to imitate
[09:01:23] <Avenger> i'm sure there are already bugs we need to imitate, because the scripts of one or two engines rely on them
[09:01:48] <fuzzie> there's that stupid repeating-effect bug :(
[09:01:54] <fuzzie> hate that one
[09:02:06] <Avenger> rapid fire in the same gametime tick?
[09:02:10] <fuzzie> no
[09:02:21] <fuzzie> the one where the original engine wipes the lists when a recurring effect fires the first time
[09:02:34] <Avenger> hmmm
[09:02:39] <fuzzie> so only the first-running one ever runs
[09:02:56] <Avenger> what?
[09:03:22] <Avenger> i don't really understand, don't all of the recurring effects execute?
[09:03:33] <fuzzie> the fx_apply_effect_repeat ones
[09:04:00] <fuzzie> You *only* get the shortest-duration one.
[09:04:38] <Avenger> omg
[09:04:44] <Avenger> that's in the original?
[09:04:46] <fuzzie> yes
[09:04:57] <fuzzie> and so, of course, stuff depends on it :P
[09:05:21] <fuzzie> it is the Most Hated IE Bug Ever, i'm pretty sure you must already know about it
[09:05:26] <Avenger> no
[09:05:37] <Avenger> actually, i missed this totally, i cannot even believe it :)
[09:05:58] <Textmode> ew, thats sounds awful...
[09:06:00] <Avenger> this means, only one recurring effect can happen in an ai cycle?
[09:06:14] <Textmode> I hope thats in a compat flag you can turn off if the game being run doesn't need it :/
[09:06:20] <Avenger> or, what's worse, only the fastest one ever works?
[09:06:46] <Avenger> that is totally unacceptable, that would mean, poison and regenerate couldn't work together
[09:07:09] <Avenger> or disease+poison
[09:07:13] <Avenger> or whatever
[09:07:17] <fuzzie> http://forums.gibberlings3.net/index.php?showtopic=17833&st=75&p=157648&#entry157648
[09:07:47] <Avenger> yeah, totally crap
[09:08:14] <Avenger> heh, you would think, the separate list (the added complexity) only helps them to do this properly
[09:08:19] <fuzzie> Taimon's conclusion is "Totally broken, probably beyond repair."
[09:08:36] <Avenger> same conclusion here, i still cannot believe it
[09:09:17] <Avenger> we talk about bioware programmers now, not bis ones :)
[09:09:45] <Avenger> and what stuff relies on this bug?
[09:09:59] <fuzzie> there are effects which shouldn't get applied, in spells and stuff
[09:10:19] <Avenger> why? when their time has come, they should happen
[09:10:22] <fuzzie> there's some posts in the fixpack forums about them, with many attempts to fix them
[09:10:29] <fuzzie> because, cut&paste :P
[09:10:50] <fuzzie> you don't notice useless stuff in spells in the original, because the effects never happen
[09:10:52] <Avenger> well, this is core engine bug, a deep one, i doubt it can be fixed by modding
[09:11:02] --> barra_home has joined #GemRb
[09:11:13] <fuzzie> yeah, but you can try and remove all the uses of the repeating EFF stuff entirely
[09:11:18] <Avenger> ahh, so some spells would become more deadly?
[09:11:40] <fuzzie> i think there's a fix in ToBEx
[09:12:48] <Avenger> poor taimon got saturated by the requests
[09:13:00] <fuzzie> well, Taimon wrote one too
[09:13:09] <fuzzie> but people are incapable of talking about what they are working on, so now we have two :P
[09:14:46] <Avenger> ah you mean ascension64 and taimon, yeah, i know. I first thought they work together
[09:15:00] <fuzzie> they just share code, once finished
[09:15:06] <fuzzie> i think
[09:15:40] <Avenger> i should have noticed them doing this faster, then i would have used ida earlier, sparing us a year
[09:16:16] <Avenger> without it, i would still only look at opcode disassemblies :)
[09:18:25] <fuzzie> yes
[09:21:22] <fuzzie> i can't find where ToBEx documents which fixes are crazy broken
[09:23:15] <Avenger> btw, anyone here, can test gemrb with the pst maze?
[09:24:10] <fuzzie> not here :/
[09:26:30] <Avenger> it won't work with just moving to ar13fy, you need the modron cube (cube.itm), also a variable set (know_cube=1)
[09:28:25] <-- Jamer has left IRC (Remote host closed the connection)
[09:33:09] <Avenger> btw, we got something wrong with triggers, the constructs don't help each other
[09:33:23] <Avenger> this is the Help() trigger
[09:34:07] <Avenger> i hope you will fix the triggers sometime
[09:35:12] <fuzzie> hm
[09:36:29] <Textmode> Avenger: "bis"?
[09:36:42] <Avenger> black isle studios
[09:38:29] <fuzzie> the nice people who wrote iwd2
[09:38:55] <Avenger> and pst :)
[09:39:09] <fuzzie> but i think bioware's efforts belong firmly in the middle there
[09:39:51] <Textmode> fwiw, I've noticerd that obsidian doesn't seem to be much better, from what I saw in their nwn efforts...
[09:40:03] <Textmode> seems to be the entire family... :/
[09:40:05] <Avenger> yep, exactly
[09:40:09] <fuzzie> they're all terrible, terrible people
[09:40:12] <Avenger> :P
[09:40:14] <fuzzie> all game developers, ever
[09:41:06] <Avenger> well, the iwd2 code is much cleaner, but the obsidian stuff was also famous about their bugs, and dragon age had its weird crap too
[09:42:31] <Avenger> anyway, there is difference between engine and dataset bugs. I hate engine bugs much more
[09:42:57] <fuzzie> yes. you can trivially fixpack the data bugs :)
[09:43:21] <Avenger> yep, and they will likely not affect gemrb
[09:43:48] <Avenger> engine bugs are sometimes relied on or fixed in another engine, so they need much more love
[09:45:05] <Avenger> hmm, our pst savegame is still not binary compatible
[09:46:49] <Avenger> i hoped it is, it would have helped with testing. I need to reach the end of the maze, and no ctrl-y in pst :(
[09:51:00] <Avenger> fuzzie, do you know if there is a breakpoints window in ida?
[09:51:13] <Avenger> i put one somewhere, but i don't find it :)
[09:51:21] <fuzzie> sure, ctrl-alt-b?
[09:51:23] <Avenger> ah nwm foud it
[09:51:25] <fuzzie> heh
[09:51:31] <Avenger> lol actually, i got two
[09:52:32] <Avenger> good news, the crash happens close to the game loader, not too deep
[09:52:35] <Avenger> i'll find it
[09:58:59] <Avenger> hmm, pst has this very odd variable handling. Probably we create a variable not on their var.var list
[09:59:13] <Avenger> it looks like they cannot create a new variable on the fly
[09:59:23] <fuzzie> stupid list :)
[09:59:34] <Avenger> yep, it is another of the top stupid stuff in pst
[09:59:49] <fuzzie> bg1 can handle any variables?
[09:59:56] <Avenger> yes
[10:00:03] <Avenger> i don't know what is this crap
[10:00:09] <fuzzie> probably what bioware had before bg1
[10:00:32] <Avenger> hehe
[10:00:48] <Avenger> it could be, this is definitely an evolutionary step back
[10:01:19] <Avenger> now, to achieve binary compatibility, we should restrict variables to var.var :(
[10:01:32] <Avenger> that would be a very bad thing
[10:01:47] <fuzzie> you could make it an option, and enforce it during save
[10:02:37] <Avenger> yes
[10:02:51] <fuzzie> just something like Variables::LoadInitialValues except saving instead of loading
[10:02:58] <Avenger> i still hope it is just some stupid format bug, or a diplicate variable or whatever
[10:03:03] <Avenger> hmm, oh
[10:03:07] <Avenger> we already got the list?
[10:03:16] <fuzzie> var.var has default values in it :)
[10:03:18] <Avenger> then it is not so terrible
[10:03:22] <fuzzie> important!
[10:03:28] <Avenger> i mean, we already have var.var loaded
[10:03:50] <fuzzie> no, i don't bother keeping it around
[10:04:01] <fuzzie> but we could do that, if it was needed
[10:04:43] <Avenger> so, we need a DropInvalidValues thing?
[10:05:05] <fuzzie> well, i thought we could just put it in the saving loop
[10:05:22] <Avenger> i would put it there, yes
[10:05:30] <fuzzie> if GF_ONLY_KNOWN_VARIABLES is set, then only write the valid variable names
[10:06:14] <Avenger> well, this means we need to load that damn list
[10:06:42] <Avenger> not a big deal, 100k more
[10:07:00] <fuzzie> only if you insist on keeping original pst compat though :P
[10:07:13] <fuzzie> and i'm pretty sure we still corrupt all kinds of stuff
[10:07:21] <Avenger> well, not too much
[10:07:25] <Avenger> i fix what i notice
[10:07:33] <Avenger> i will probably not fix this
[10:07:47] <Avenger> i just need to find the offending variable and delete it with dltcep
[10:08:01] <Avenger> and find out why it was set in the first place :)
[10:08:13] <fuzzie> we set loads of fairly random vars
[10:08:23] <Avenger> we?
[10:08:42] <Avenger> game variables?
[10:08:58] <fuzzie> yes
[10:09:00] <Avenger> i don't think so, we set only some that are used in an engine variant
[10:09:16] <Avenger> like chapter
[10:09:28] <Avenger> i don't know of any other, from the top of my head
[10:09:54] <fuzzie> all the death flags
[10:10:03] <Avenger> hmm
[10:10:30] <Avenger> actually, i ctrl-y'd a few constructs :)
[10:11:25] <Avenger> ok, you know what, i will add a game feature: no new variables
[10:11:28] <Avenger> that will fix pst
[10:11:49] <Avenger> and that's closer to what they do
[10:12:06] <Avenger> i will just print some warning on the console
[10:12:16] <Avenger> so it is possible to catch the new variables
[10:12:19] <fuzzie> oh, i guess that works
[10:12:27] <Avenger> hmm, actually, i will just add the warning first
[10:13:01] <fuzzie> if pst really doesn't like new variables, though, you have to add the feature
[10:13:27] <fuzzie> but i guess you'll find out
[10:13:53] <Avenger> yes
[10:14:04] <Avenger> i mean, i will have to add the feature
[10:27:22] --> Maighstir__ has joined #GemRb
[10:30:37] <-- Maighstir has left IRC (Ping timeout: 255 seconds)
[10:35:14] <Avenger> meh, i fixed the setvariable actions, but of course, not the numerous instances where the core sets a variable :(
[10:38:17] <fuzzie> and they are probably the issue
[10:38:30] <fuzzie> which is why i was suggesting the death vars as an example
[10:38:47] <fuzzie> it is tricky because you can't just sabotage SetAt, since that would break the core..
[10:39:08] <Avenger> i did, somewhat
[10:39:32] <Avenger> added a nocreate boolean to the SetAt which handles dwords
[10:39:56] <Avenger> it is a default param with false as default
[10:40:15] <fuzzie> i would add a 'force' boolean
[10:40:21] <Avenger> so, it won't do much if i don't use true
[10:40:56] <Avenger> i use the true for the special handling because if(nocreate) { print warning; return; }
[10:41:06] <fuzzie> hm
[10:41:07] <fuzzie> ok
[10:41:11] <fuzzie> which is the special handling?
[10:41:20] <fuzzie> i mean, you want the special case to be the core functions which force creation
[10:41:35] <fuzzie> because you don't want people to accidentally add SetAt calls which make new variables
[10:41:51] <Avenger> well, actually i thought about the opposite
[10:42:10] <fuzzie> and it will be obvious if someone does it for non-game variables, since, it won't work :)
[10:42:20] <Avenger> i would rather add the extra param where the variable creation is needed
[10:42:34] <Avenger> i mean, where it is not always needed
[10:42:49] <Avenger> it is the new behaviour
[10:42:51] <fuzzie> well, i know i will accidentally cause bugs, if i have to remember to add an extra param to avoid invisible bugs.
[10:43:18] <Avenger> or you have to add it to force a variable setting :)
[10:43:32] <Avenger> either way it is a new source of bugs
[10:43:33] <fuzzie> well, if i have to add it to force a variable setting, i will notice when it doesn't work :)
[10:43:49] <fuzzie> if i have to add it to force some strange behaviour which only matters when loading saves in original pst, i will never notice :P
[10:44:20] <Avenger> if no one notices it, it isn't a bug :D
[10:44:32] <fuzzie> well, ok, then you don't need to add any of this code at all :P
[10:44:36] <-- jschall has left IRC (Remote host closed the connection)
[10:45:29] <fuzzie> i guess it really sucks either way
[10:45:35] <Avenger> yep
[10:46:04] <fuzzie> meh
[10:47:35] <Avenger> you don't know what you miss by not playing with the modron cube :)
[10:47:51] <fuzzie> bugs? :P
[10:48:10] <Avenger> Hordes of imaginary creatures cheer the cube's victory. You can almost see a tiny oily tear brimming on one of its eyes... it is a hero, the greatest cube ever to roam the Planes, and everyone loves it.
[10:48:34] <Avenger> pst writers are the best
[10:50:16] <Avenger> cool, i think i found the offending variable
[10:50:32] <Avenger> kaputz/drone_light
[10:51:25] <Avenger> but, there is: kaputz/kill_drone_light
[10:51:38] <Avenger> so, it could be a simple bug
[10:52:36] <fuzzie> if (KillVar[0]) {
[10:52:36] <fuzzie> if (core->HasFeature(GF_HAS_KAPUTZ) ) {
[10:52:41] <fuzzie> ^- this bit is .. weird?
[10:53:14] <fuzzie> it makes no sense
[10:53:52] <fuzzie> oh, it's from 6 years ago :(
[10:54:39] <fuzzie> the KILL_ bit from below should be there, yo'd think. although i am looking at old code, maybe it's already fixed.
[10:54:51] <fuzzie> anyway, i ahve to go work.
[10:54:57] <Avenger> yes
[10:55:07] <Avenger> the kill bit below also works
[10:55:14] <Avenger> so, maybe this branch should be entirely cut
[10:55:54] <fuzzie> the kill bit below only runs if scriptname is set
[10:56:11] <fuzzie> so move that to here, maybe?
[10:57:32] <Avenger> dunno
[10:57:53] <Avenger> scriptname is there even if it is 'none' ;)
[10:57:57] <Avenger> and it works now
[10:58:04] <Avenger> kill_drone_light is set
[11:07:14] <Avenger> huh, apparently gemrb sets a variable even by loading an original game and resaving it (also causing a crash)
[11:07:37] <Avenger> so, while the death vars were wrong, they didn't fix the problem
[11:08:29] <Avenger> good thing, dltcep orders the variables :)
[11:12:35] <Avenger> haha, it is the chapter variable
[11:12:45] <Avenger> so, hmm, pst has no chapter?
[11:14:43] <Avenger> odd, the code still contains it, but i guess, it doesn't work because it isn't listed in var.var
[11:15:20] <Avenger> there is even incrementchapter action
[11:16:02] <Textmode> weird.
[11:16:31] <Avenger> yes, i guess, if you add it to var.var, it would even work
[11:17:53] <Avenger> pst is weird
[11:20:30] <barra_home> but still a great game :-)
[11:26:07] <Textmode> aye
[11:31:47] <Avenger> meh, there are still 20 extra bytes in the gemrb version
[11:35:06] <Avenger> ok, no extra bytes, but it still crashes
[11:35:23] <barra_home> on a related note: how complete is PS:T support with GemRB these days?
[11:35:35] <barra_home> is there any kind of list of unsupported features?
[11:36:05] <Avenger> projectiles and overlays are the major missing features, and of course the maze isn't done yet
[11:36:22] <Avenger> other than that, i don't think so
[11:36:55] <barra_home> coolio, thanks for filling me in
[11:37:02] <barra_home> modron maze Avenger ?
[11:37:07] <Avenger> yes
[11:37:14] <Avenger> it is half working already :)
[11:37:23] <barra_home> I've read that it's the only randomly generated maze in all IE games?
[11:37:32] <Avenger> yes
[11:37:54] <Avenger> though, you could do a maze like this in bg2
[11:38:24] <barra_home> never played BG2 in detail, hated the beginning and gave up :-/
[11:39:26] <Avenger> well, any ie version which has doors or can somehow deactivate travel regions
[11:39:42] <Avenger> i think bg1 doesn't qualify
[11:40:22] <barra_home> does projectile support affect magic as well?
[11:40:47] <barra_home> or does it just matter if you have nordom in your party as he has a crossbow?
[11:41:34] <Avenger> we did a random maze in glory of istar
[11:42:15] <barra_home> is that a custom game based on GemRB?
[11:42:19] <Avenger> projectile support affects all spells and special abilities
[11:42:31] <Avenger> no, it was just a total conversion for the tob engine
[11:42:55] <Avenger> not all projectiles are missing, only the ones with hardcoded crap
[11:43:34] <Avenger> there is an acid serpent, or morte's skills, or such, all the nice fancy stuff that makes pst unique
[11:43:50] <Avenger> sadly they were also hardcoded in an ugly way
[11:45:01] <barra_home> so there's no easy and clean way to implement support for them?
[11:45:44] <Avenger> yeah, not easy, mostly because i don't want to make stuff that is only useful for a single thing
[11:46:16] <Avenger> it is probably possible to do them, but they would just bloat our core
[12:28:47] <Avenger> oh damn
[12:29:29] <Avenger> this is the offending variable: Dictionary_Githzerai_ Hjacknir
[12:30:06] <fuzzie> Dictionary?
[12:30:32] <Avenger> that's just part of the variable name
[12:30:46] <Avenger> the cause of the bug is the space in it
[12:31:11] <Avenger> this variable would never work, but this is how it is in var.var
[12:31:32] <Avenger> i would break the fingers of the monkey who typed it in
[12:32:44] <Avenger> this is a really bad one
[12:33:04] <Avenger> since var.var contains it with space, it is how it is assigned in the .gam
[12:33:24] <Avenger> gemrb 'fixes' it
[12:33:37] <Avenger> but then the fixed .gam cannot be loaded
[12:34:09] <Avenger> and all this for a variable never used
[12:34:23] <Avenger> it wouldn't work anyway, since scripts get their spaces dropped
[12:34:34] <Avenger> so the scripts couldn't reference the bad variable
[12:36:11] <Avenger> i got no real solution for this
[12:36:46] <Avenger> var.var should be fixed, or the corrupted name should be added to .gam :(
[12:37:21] <Avenger> the variable is totally unused, btw
[12:38:31] <Avenger> i'm not going to fix this, i will fix my var.var locally
[12:38:33] <Avenger> damn bis
[12:47:49] <Avenger> ok, another crash, but i definitely got farther, done with variables :)
[13:55:06] <Avenger> hah, found the bug
[13:55:13] <Avenger> took some time :(
[13:55:39] <Avenger> it is our one byte anim id, we simply lose the high word
[13:56:05] <fuzzie> um
[13:56:12] <Avenger> so the animation becomes 0032 instead of 6032, but pst also supports the old style animations
[13:56:21] <fuzzie> yes indeed
[13:57:01] <Avenger> well, i hope i can simply preserve the high part
[13:58:22] --> edheldil_ has joined #GemRb
[14:03:35] --> _pickle has joined #GemRb
[14:04:54] <Avenger> meh, this is in some script
[14:05:44] <Avenger> anims.2da
[14:07:22] <Avenger> cool, the progress bar went to the end before the next crash :)
[14:07:35] <-- Avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
[14:11:16] <CIA-29> GemRB: 03avenger_teambg * r3963cfaa6b87 10gemrb/gemrb/override/pst/anims.2da: for binary compatibility with the original engine
[14:19:28] --> Avenger has joined #GemRb
[14:19:28] --- ChanServ gives channel operator status to Avenger
[14:19:38] <Avenger> fuzzie you were right about the killvar vs. scriptname thing in Actor.cpp
[14:27:12] <CIA-29> GemRB: 03avenger_teambg * r732a7bc0c197 10gemrb/gemrb/ (11 files in 6 dirs):
[14:27:12] <CIA-29> GemRB: implemented GF_NO_NEW_VARIABLE
[14:27:12] <CIA-29> GemRB: fixed death variables in pst
[14:31:08] <-- tomprince has left IRC (Ping timeout: 250 seconds)
[14:38:54] --> deepinthewoods has joined #GemRb
[16:06:16] <-- deepinthewoods has left IRC ()
[16:30:42] --> tomprince has joined #GemRb
[17:41:21] <-- Avenger has left IRC (Quit: bye!)
[18:05:09] --> lynxlynxlynx has joined #GemRb
[18:05:09] <-- lynxlynxlynx has left IRC (Changing host)
[18:05:09] --> lynxlynxlynx has joined #GemRb
[18:05:09] --- ChanServ gives channel operator status to lynxlynxlynx
[18:24:07] <lynxlynxlynx> fun backlog today
[18:35:28] <fuzzie> oh?
[18:38:38] <CIA-29> GemRB: 03avenger_teambg * r76c14357c2d2 10gemrb/gemrb/GUIScripts/pst/Start.py: play intro videos not only for the first time (SkipIntroVideos can still disable them)
[18:42:51] --> Avenger has joined #GemRb
[18:42:51] --- ChanServ gives channel operator status to Avenger
[18:43:02] <Avenger> hello
[18:47:36] <lynxlynxlynx> cool, another feature from beholder
[18:48:14] <lynxlynxlynx> http://pastebin.com/Km47YzRJ wider and visible gc scrolling regions
[18:50:33] <Avenger> semi transparent?
[18:51:42] <Avenger> hmm, that's just lines, right?
[18:51:54] <Avenger> no semi transparent rectangle
[18:52:48] <Avenger> anyway, it if helps the phones, it is cool
[18:53:12] <Avenger> the android downloads surpass all the other regular downloads
[18:53:37] <Avenger> i wonder how that happens, but even if 1% of those tries the PC, it is a good advertising
[18:55:38] <fuzzie> well, there are rather a lot of android users
[18:55:45] <Avenger> also Korea became the top country of downloaders, whoa
[18:56:14] <Avenger> i wonder if they will make tournaments like with other games :D
[18:56:56] <lynxlynxlynx> hehe
[18:57:10] <lynxlynxlynx> goblin rush
[19:00:11] <Avenger> 55 downloaders from Korea on a single day, and they made ~100 around this month. Sounds like a rush, yeah
[19:36:44] <Avenger> hmm i wonder why the dialog text is always in a different color
[19:42:26] --> Maighstir___ has joined #GemRb
[19:45:39] <-- Maighstir__ has left IRC (Ping timeout: 276 seconds)
[19:49:28] <Avenger> hmm doors became opaque?
[19:49:42] <Avenger> i mean, transparent ...
[19:53:35] <Avenger> i wonder who changed PATH_MAP_NO_SEE to PATH_MAP_SIDEWALL
[19:54:13] <fuzzie> in which code?
[19:54:44] <fuzzie> oh, in IsVisible
[19:55:06] <fuzzie> i changed it, ages ago
[19:55:26] <fuzzie> i don't know what all these searchmap changes recently did though
[19:55:45] <fuzzie> but i got tired of people seeing through walls
[19:55:58] <edheldil_> in theory, it should be clear from commit messages :-P
[19:56:14] <Avenger> i have difficulties finding them
[19:56:24] <fuzzie> "use PATH_MAP_SIDEWALL in Map::IsVisible, since PATH_MAP_NO_SEE is never present"
[19:56:38] <fuzzie> is the commit message
[19:58:50] <edheldil_> btw, WMPICKL/WMPICKR is used only in two functions in pst, and only one looks like distinguishing between them
[19:59:44] <edheldil_> although, there might be more references in wrongly parsed parts
[19:59:47] <CIA-29> GemRB: 03lynxlupodian * r53047aeff40a 10gemrb/ (3 files in 2 dirs):
[19:59:47] <CIA-29> GemRB: applied Beholder's patch for better touchpad scrolling support
[19:59:47] <CIA-29> GemRB: plus a switch to cmake (define TOUCHPAD to enable)
[20:02:53] <Avenger> path map no see is never present?
[20:02:55] <Avenger> whoa
[20:03:04] <Avenger> that sounds like a bug
[20:03:36] <fuzzie> well
[20:04:02] <fuzzie> at the time, we didn't bother with half of the searchmap info, we just dropped it
[20:04:05] <Avenger> that was not a good change
[20:04:19] <Avenger> sidewalls should be different from direct blocks
[20:04:21] <fuzzie> i don't know if anyone fixed this since, but it doesn't look like it
[20:04:29] <Avenger> sidewalls let you see 1-2 tiles farther
[20:04:55] <fuzzie> i tried it in the original engine, and IsVisible seemed to stop at sidewalls
[20:05:27] <fuzzie> but no-one would answer my questions at the time, so i made the change which stopped people from shooting through walls :P
[20:07:08] <Avenger> uh
[20:07:15] <Avenger> well
[20:07:21] <fuzzie> this is nothing to do with what you can see on-screen
[20:07:32] <fuzzie> not used in the exploration code
[20:08:04] <edheldil_> Avenger: btw what about ignoring baldur.ini? It seems that it only causes confusion
[20:08:20] <fuzzie> but NO_SEE has almost always been broken, you removed it from pathfind.2da 6 years ago
[20:08:52] <fuzzie> for some reason, SIDEWALL is always set instead.
[20:09:27] <Avenger> i remember something, yes, but i forgot why
[20:09:49] <fuzzie> well, 2005 is a long time before i started coding on this :P
[20:11:17] <Avenger> yep, i start to forget more of the code than i write :)
[20:11:43] <fuzzie> anyway, i guess it is obvious from the commit message why i changed it.
[20:12:27] <fuzzie> it seems like pathfind.2da destroys a lot of information, still
[20:13:03] <fuzzie> but i never understood it
[20:13:54] <Avenger> i'm sure it was me who removed one of the door flags from cval = PATH_MAP_DOOR_OPAQUE|PATH_MAP_DOOR_TRANSPARENT;
[20:14:00] <Avenger> that was wrong too
[20:14:36] <Avenger> those names don't cover the true meaning of the flags
[20:14:50] <Avenger> OPAQUE is about visibility, TRANSPARENT is about passability
[20:16:25] <CIA-29> GemRB: 03lynxlupodian * radeabc38ca85 10gemrb/gemrb/core/GUI/GameControl.cpp: fixed compilation when TOUCHSCREEN is set
[20:22:19] <lynxlynxlynx> hmpf, indeed, soa intro doesn't work anymore
[20:22:37] <lynxlynxlynx> when i looked back, my char is dead in the cage :/
[20:24:04] <lynxlynxlynx> [Variables]: Cannot create new variable: SPRITE_IS_DEADishaddud <-- i thought this should only block in pst
[20:25:14] <lynxlynxlynx> trying again
[20:25:24] <lynxlynxlynx> nothing else was odd in the output
[20:26:44] <Avenger> meeh
[20:26:52] <lynxlynxlynx> worked this time, maybe it was a bad char i imported (created new this time round)
[20:27:10] <Avenger> i messed it up
[20:27:28] <DrMcCoy> Btw: How's PST support? :P
[20:27:39] <Avenger> i use direct booleans, instead of core->HasFeature(GF_NO_NEW_VARIABLE)
[20:27:42] <lynxlynxlynx> 3 damage suffered, but i doubt the other guy has so little hp :)
[20:28:13] <Avenger> we are working on it
[20:28:37] <lynxlynxlynx> hah, he does! 2/2 hp :)
[20:28:44] <lynxlynxlynx> must've been a debug char of mine
[20:29:17] <lynxlynxlynx> ah, level 1 thief
[20:41:36] <CIA-29> GemRB: 03avenger_teambg * rc4fe576656fa 10gemrb/gemrb/ (5 files in 4 dirs): added door hacks to the maze setup (closer to the original)
[20:50:06] <Avenger> ok, i fixed that variable hack
[20:50:26] <CIA-29> GemRB: 03avenger_teambg * r3a7f9bdc9b6b 10gemrb/gemrb/core/ (Game.cpp Scriptable/Actor.cpp): fixed the recent variable hack
[20:51:04] <-- Avenger has left IRC (Quit: bye!)
[21:59:39] <CIA-29> GemRB: 03avenger_teambg * rddefcba69744 10gemrb/gemrb/plugins/PSTOpcodes/PSTOpcodes.cpp: pst: fixed play bam opcode
[22:05:20] <CIA-29> GemRB: 03avenger_teambg * rdf2d84965605 10gemrb/gemrb/ (GUIScripts/pst/Maze.py plugins/GUIScript/GUIScript.cpp): maze: adding trap script to the enabled traps too
[23:26:57] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:46:30] --> pieplay has joined #GemRb
[23:46:56] <pieplay> is their a launchpad for GemRB?
[23:55:30] <tomprince> launchpad?