#gemrb@irc.freenode.net logs for 3 Dec 2011 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:12:33] <-- lynxlynxlynx_ has left IRC (Remote host closed the connection)
[00:41:19] <-- SiENcE has left IRC (Ping timeout: 248 seconds)
[01:06:58] <-- Maighstir has left IRC (Read error: Connection reset by peer)
[01:16:25] <-- edheldil_ has left IRC (Ping timeout: 244 seconds)
[01:25:24] <-- Yoshimo has left IRC (Ping timeout: 240 seconds)
[01:27:56] --> duckpunch has joined #gemrb
[01:40:50] --> edheldil_ has joined #gemrb
[01:46:55] <-- edheldil_ has left IRC (Ping timeout: 248 seconds)
[03:47:28] --> joneirik has joined #gemrb
[04:11:30] <-- joneirik has left IRC (Remote host closed the connection)
[04:48:25] <-- demitar has left IRC (Ping timeout: 260 seconds)
[07:24:28] --> Beholder has joined #gemrb
[07:24:28] <-- Beholder has left #gemrb
[09:17:44] --> edheldil_ has joined #gemrb
[09:29:35] <-- edheldil_ has left IRC (Ping timeout: 244 seconds)
[10:13:53] --> lynxlynxlynx has joined #gemrb
[10:13:53] --- ChanServ gives channel operator status to lynxlynxlynx
[10:41:44] --> Yoshimo has joined #gemrb
[10:56:54] --> lloyd has joined #gemrb
[10:56:59] <lloyd> hey guys
[10:59:10] <lynxlynxlynx> oj
[10:59:28] <lloyd> hey I finally found the source of my 2da errors
[10:59:42] <lloyd> it was the random number
[10:59:49] <lloyd> DisplayStringCore(actor, VB_SELECT+core->Roll(1,3,-1), DS_CONST|DS_CONSOLE);
[11:00:22] <lloyd> so it chooses a number from 1-3 or whatever
[11:00:36] <lloyd> problem is that not all things have three different select sounds
[11:01:00] <lloyd> some only have two -- so that there is a 1/3 chance of it failing
[11:02:15] <lloyd> and that's what generates the missing .2da error...because after the it returns -1 it tries asking resource manager for something
[11:02:31] <lynxlynxlynx> so i was right :)
[11:02:35] <lloyd> yep!
[11:02:52] <lynxlynxlynx> you can use GetStringReference to check it, i think
[11:03:43] <lynxlynxlynx> nah, that's for strings.2da
[11:04:01] <lloyd> ya I checked it with GDB and dltcep
[11:04:01] <lynxlynxlynx> core->GetString
[11:04:52] <lynxlynxlynx> actually GetVerbalConstant should be enough
[11:05:15] <lloyd> how can we check how many select sounds an actor has?
[11:06:48] <lloyd> ya GetVerbalConstant is where it happens -- if the index strref passed to GetVerbalConstant is 28 and the actor only has two select sounds (27-28) you get the problem
[11:07:20] <lloyd> I'm thinking that something in VCMap could be done
[11:07:34] <lynxlynxlynx> you could use Actor::VerbalConstant
[11:07:49] <lynxlynxlynx> it's odd how it counts, but it looks appropriate otherwise
[11:15:40] <lloyd> lol I'm tired what is that function doing?
[11:18:14] <lynxlynxlynx> using a random vb at given offset and their count
[11:18:36] <lynxlynxlynx> it looks biased a bit, but it does check if they resolve to valid strrefs
[11:18:56] <lloyd> I don't see it checking for -1?
[11:19:07] <lynxlynxlynx> it's in the end
[11:19:51] <lloyd> oh right yeah
[11:19:52] <lloyd> cool
[11:20:14] <wjp> lynxlynxlynx: "a bit"? :-)
[11:21:05] <lynxlynxlynx> i didn't compute it
[11:22:27] <lloyd> ya it would be a bit...
[11:55:06] --> SiENcE has joined #gemrb
[11:57:27] --> Maighstir has joined #gemrb
[12:13:43] <-- Maighstir has left IRC (Quit: .)
[12:19:06] --> Maighstir has joined #gemrb
[12:37:30] <Yoshimo> how about skipping alignment during bg character creation if there is just one choice anyway?
[12:38:17] <lynxlynxlynx> just press enter
[12:38:53] <lynxlynxlynx> if that doesn't work, we just need to add autofocus
[12:39:26] <Yoshimo> there is nothing like "enter" on my cell ;)
[12:41:45] <lynxlynxlynx> none of the buttons are bound as default?
[12:43:33] --> edheldil_ has joined #gemrb
[12:44:49] <Yoshimo> i only have the possibility to go back, go to my homescreen , lock the cell and fiddle with volume
[12:56:18] <lynxlynxlynx> sucks
[12:59:56] --> Beholder has joined #gemrb
[13:00:03] <Beholder> hi all
[13:02:04] --> demitar has joined #gemrb
[13:02:13] <lynxlynxlynx> oj
[13:02:43] <CIA-44> GemRB: 03lynxlupodian * r1fbae665d10f 10gemrb/gemrb/core/ (Game.cpp Map.cpp Map.h): made partial rest actually rest the actors
[13:02:43] <CIA-44> GemRB: 03lynxlupodian * ra0edaa9c0090 10gemrb/gemrb/GUIScripts/GUICommon.py:
[13:02:43] <CIA-44> GemRB: SetGamedaysAndHourToken: also set the bg1 token for days
[13:02:43] <CIA-44> GemRB: fixes the game day not being displayed in the pause button tooltip
[13:02:44] <CIA-44> GemRB: 03lynxlupodian * ra372bd972990 10gemrb/gemrb/core/ (Game.cpp Map.cpp):
[13:02:44] <CIA-44> GemRB: made Map::Rest advance the game time, so interruptions don't ignore it
[13:02:45] <CIA-44> GemRB: and partial rest becomes possible in that window
[13:26:36] <Yoshimo> Beholder: how many hardware buttons on android should be mapped by default?
[13:35:21] <Beholder> I use this scheme: 'Search' as containers highlighting (default), 'Menu' as mouse right (defined in SDL configuration), 'Back' as Esc (default), volume keys are not overridden
[13:47:02] <Beholder> I added logging to SDLVideoDriver::PollEvents(), touch events not handling
[13:47:23] <Beholder> mouse events too
[14:48:23] <Beholder> hmm. all events not handled
[14:48:27] <Beholder> keys too
[14:51:27] <Beholder> I added 'printMessage( "SDLVideo", "Event handler...", WHITE );' after 'while ( SDL_PollEvent(&event) ) {' no one message been handled and logged
[14:51:34] <Beholder> odd
[14:53:57] <-- edheldil_ has left IRC (Ping timeout: 244 seconds)
[15:05:36] --> SiENcE_ has joined #gemrb
[15:08:12] <-- SiENcE has left IRC (Ping timeout: 240 seconds)
[15:10:10] --> jeremyagost has joined #gemrb
[15:13:55] <Beholder> where menu events handling?
[15:34:50] <Beholder> ok, i understand
[15:35:02] <Beholder> all touch events works
[15:35:09] <Beholder> mouse events not
[15:35:31] <Beholder> it not be pushed too
[15:36:17] <Beholder> new code for handling touch events required
[15:36:55] <Beholder> without simulating mouse activity
[15:53:54] --> brad_a has joined #gemrb
[15:59:42] <brad_a> Beholder: I will rewrite the touch code to stay within the touch events and just put and #if IOS in the mouse events to have them immediately return. that way it will work with both platforms. unfortunately i wont have time to do it for a couple weeks at least, but Ill put it at my highest gemrb priority after getting the TTF plugin merged, however.
[16:00:39] <brad_a> a simple interim solution may be to add a mose event at the end of the touch handlers
[16:00:49] <brad_a> but that may be the wrong order i cannot recall
[16:02:15] <brad_a> or in the android SDL code you can put an SDLSendMouse
[16:05:55] <Beholder> maybe i'm rewrite this code?
[16:06:21] <Beholder> i have some free time now
[16:07:14] <Beholder> ios gets mouse events with finger events, so?
[16:07:51] <Beholder> fingerdown with mousedown and fingerup with mouseup
[16:15:42] <brad_a> well i wont stop you, but im not sure it will be clear what everything is doing
[16:16:07] <brad_a> but yes ios sdl sends both, but i dont remember the order
[16:18:32] <brad_a> if you really want to hack on it yourself refer to this for what the touches currently do: http://gemrb.sourceforge.net/wiki/doku.php?id=install:ios#touch_input
[16:20:50] <-- duckpunch has left IRC (Ping timeout: 258 seconds)
[16:22:05] <Beholder> thanks
[16:22:21] <Beholder> beautiful keyboard!
[16:24:44] <brad_a> thank you :)
[16:25:25] <fuzzie> the modifiers there are really rather nicely integrated
[16:26:02] <Beholder> can i use you keyboard for Android? can it possible?
[16:26:17] <Beholder> *your
[16:26:32] <fuzzie> that's just the iOS soft keyboard with a few overrides, right?
[16:26:51] <brad_a> yes its a"hacked up" ios keyboard
[16:27:04] <brad_a> but ive seen android apps with similar keyboard mods
[16:27:43] <fuzzie> apps, or devices?
[16:27:47] <Beholder> Android keyboard fills full screen with no transparency. ugly
[16:28:03] <fuzzie> well, there are lots of different android keyboards
[16:28:09] <brad_a> well the ios keyboard isnt transparent either. that was a hack
[16:28:17] <fuzzie> and that means that you can't mess with their views reliably
[16:28:42] <brad_a> because the keyboard often covers what you are typing in in gem rb the easiest thing to let you see was to make it semi transparent
[16:29:38] <Beholder> looks great
[16:30:22] <brad_a> its kinda funny how much time and effort i put into it considering i cant play it :)
[16:30:27] <fuzzie> yes, you are mad
[16:30:33] <fuzzie> is the text entry needed much for gemrb?
[16:30:42] <brad_a> just want to share the joy of gemrb
[16:30:57] <brad_a> do journal entries not work?
[16:31:01] <brad_a> i mean custom ones
[16:31:16] <brad_a> or adding map markers?
[16:31:31] <fuzzie> well, i mean, i'm just wondering if the easiest thing for android wouldn't just be to pop up a native text field
[16:32:03] <brad_a> sure. i was thinking of addign a text field to the keyboard accessory view for ios too
[16:32:27] <brad_a> there is certainly room
[16:34:12] <brad_a> i did move the console to the top tho so you can see that clearly. if anybody was wondering why its at the top now.
[16:40:36] <Beholder> i think we need to ignore mouse events for ios (add temporary hack) until SDL developers fix it
[16:43:49] <brad_a> thats what i said :-P
[16:45:27] <brad_a> in retrospect that probably what i should have done in the first place. the code would be a lot less hacky if i had
[16:45:48] <-- SiENcE_ has left IRC (Quit: cya)
[16:56:38] <Beholder> ok :)
[17:02:27] <brad_a> Beholder: i tried implementing the audio pause in the sdl window events for android, but im guessing it is crashing on minimize because i did something wrong?
[17:02:47] <brad_a> i though i just copied what the handlers in gemrb.cpp were doing
[17:03:39] <Beholder> i look backgrounding after fixing touch events
[17:03:50] <Beholder> what does this code:
[17:03:59] <Beholder> if (Evnt && numFingers != core->NumFingInfo) {
[17:03:59] <Beholder> Evnt->KeyRelease( GEM_ALT, 0 );
[17:04:16] <Beholder> what is core->NumFingInfo
[17:05:04] <fuzzie> number of fingers which have to be held down for simulating 'alt'
[17:05:31] <brad_a> if the number of fingers in contact with the screen == numfinginfo then send alt+tab
[17:05:32] <fuzzie> (since alt displays info like containers/HP/etc)
[17:05:33] <Beholder> ah, ok
[17:05:36] <brad_a> that
[17:05:55] <fuzzie> this is where the touch stuff gets really tricky
[17:05:59] <brad_a> yes
[17:06:25] <brad_a> i do apologize for any aneurisms :-P
[17:06:47] <Beholder> NumFingInfo is a config variable?
[17:06:52] <brad_a> yes
[17:06:57] <brad_a> as is numfingkboard
[17:07:04] <brad_a> and one other
[17:07:08] <brad_a> i think numfingscroll
[17:07:55] <fuzzie> you think that's enough input to play with?
[17:08:09] <brad_a> did you look at the touch input guide?
[17:08:15] <brad_a> i think i covered everything
[17:08:42] <brad_a> looking at that will probably help navigate the coe too
[17:08:49] <brad_a> code
[17:09:04] <fuzzie> yes, i am really asking if you think it's enough
[17:09:31] <brad_a> i may just be myopic but yes...
[17:09:36] <brad_a> what are you thinking?
[17:09:50] <fuzzie> i am thinking "instead of thinking for ages about what input is needed, i can just ask" :)
[17:10:06] <brad_a> oh :)
[17:10:20] <fuzzie> sorry, no wisdom, just curiousity
[17:10:45] <brad_a> i see. usually fuzzie asking if i think something is followed by an epiphany :)
[17:15:47] <brad_a> Beholder: i must leave you in fuzzie's capable hands for answering any other questions today. i must go.
[17:15:51] <brad_a> good luck
[17:15:55] <-- brad_a has left IRC (Quit: brad_a)
[17:22:46] <Beholder> Menu buttons processing handled on Mouse Up or Mouse Down event?
[18:02:45] <Yoshimo> can you create diffrent shortcuts on windows which all start a diffrent game with a diffrent gemrb.cfg?
[18:03:36] <jeremyagost> yeah, you should be able to
[18:03:58] <jeremyagost> in the shortcut you need to specify the command line parameter for a specific config file
[18:04:22] <jeremyagost> it's '-c' I believe
[18:04:24] <Yoshimo> which would be ?
[18:04:28] <Beholder> this action needs to write special launcher app. No other way.
[18:05:05] <jeremyagost> so make a shortcut and add "-c C:\my\config\path.cfg" to the executable path
[18:05:15] <Beholder> android apps does not support command line options
[18:05:39] <jeremyagost> Yoshimo asked about windows
[18:05:51] <Beholder> )
[18:06:00] <Yoshimo> a select game on startup might be handy for android too nevertheless^^
[18:06:04] <jeremyagost> yeah
[18:06:14] <jeremyagost> on a totally unrelated note, is there a way to slow down the framerate?
[18:06:28] <jeremyagost> the original games had a configuration option for manual framerate control
[18:06:43] <jeremyagost> I've gotten bouncing lightning working
[18:06:57] <jeremyagost> but it's spazzing out sometimes
[18:06:58] <Yoshimo> some of the original game ingame options actually do nothing in gemrb
[18:07:03] <Beholder> hm, i know it limited to 30fps in latest versions
[18:07:33] <jeremyagost> hold on I've gotta record a video of this...
[18:09:44] <Yoshimo> if your computer is too fast some old dos games run amok, same problem here?
[18:10:21] <fuzzie> well, original didn't *really* have framerate control, all kinds of things broke if you twiddled it
[18:10:29] <fuzzie> but you can change the delay in SDLVideo in our source
[18:11:40] <jeremyagost> ok
[18:11:45] <jeremyagost> nah it
[18:12:11] <jeremyagost> nah it's just it would be easier to debug the lightning pathing if it wasn't changing direction so fast
[18:13:48] <jeremyagost> http://www.youtube.com/watch?v=bGxCKyK2gJU
[18:14:07] <jeremyagost> ^^ work-in-progress :P ^^
[18:14:35] <fuzzie> nice
[18:14:49] <Yoshimo> yea it looks a bit wrong ;)
[18:22:04] <jeremyagost> alright well this is interesting
[18:22:27] <jeremyagost> in BG2 the lightning goes past the walls a bit before turning around
[18:23:32] <tomprince> It seems like it is bouncing of multiple sides of the same box.
[18:24:53] <tomprince> Would it be easier to figure out if you replaced it with something more point like? Say a single magic missle?
[18:24:57] <fuzzie> jeremyagost: are you hitting the right walls?
[18:25:50] <fuzzie> (i assume you worked out that you can see searchmap with ctrl-8, the relevant bits should be highlighted in different colours if i didn't mess it up since)
[18:25:53] <jeremyagost> well what I'm doing is calling GetBlocked
[18:26:04] <fuzzie> well, GetBlocked returns a mask
[18:26:06] <jeremyagost> yeah
[18:26:53] <jeremyagost> specifically, GetBlocked() & PATH_MAP_PASSABLE
[18:27:29] <jeremyagost> the blocked test occurs after the next point is calculated
[18:27:36] <fuzzie> well, did you look at ctrl-8?
[18:27:39] <jeremyagost> if it is blocked, then it tests a new point 90º CCW
[18:27:51] <jeremyagost> if THATS blocked then it tests a point 90º CW
[18:28:01] <jeremyagost> fuzzie: I'll try that
[18:28:27] <tomprince> I think you need to be a bit smarter. There are similiar issues with visibility, since the blocks are comparatively big.
[18:29:35] <fuzzie> you can see there PATH_MAP_SIDEWALL (blue) and PATH_MAP_IMPASSABLE (red) which might correspond to what you see in original
[18:30:04] <tomprince> In candlekeep, if you go to the outside northwestern corner of the inner keep, and stand just south of the tower, you can see into the keep. (If i remember correctly.
[18:30:17] <fuzzie> in gemrb?
[18:30:33] <tomprince> Yes.
[18:30:34] <fuzzie> there's a ridiculously stupid gemrb bug where we calculate the visibility from a point which is slightly too far south/east
[18:30:45] <fuzzie> but no-one seems to care about it whenever i bring it up
[18:31:12] <jeremyagost> yeah I watched it work with the searchmap
[18:31:14] <jeremyagost> hmm
[18:31:47] <jeremyagost> is there any other (smarter) way to find walls?
[18:32:01] <fuzzie> the original is definitely using the searchmap for bouncing
[18:32:19] <jeremyagost> yeah
[18:32:25] <jeremyagost> one sec
[18:32:32] <fuzzie> i think specifically checking for those red squares, so where GetBlocked() returns 0
[18:33:04] <fuzzie> but i don't know what it does with them :) and can't think of any obvious trivial idea
[18:37:00] <jeremyagost> even the original game can be pretty retarded it seems
[18:37:14] <jeremyagost> but it looks less stupid because the bolt isn't moving so fast
[18:38:47] <jeremyagost> http://www.youtube.com/watch?v=aJ08KrC3DmQ
[18:41:11] <Yoshimo> is it really that ugly cutting through the wall and overlapping ?
[18:41:34] <jeremyagost> yup
[18:43:34] <Yoshimo> is there something we can do about that?
[18:44:42] <jeremyagost> well lightning in BG2 was always kludgy
[18:47:33] <jeremyagost> what do we usually use for trig in gemrb?
[18:47:33] <jeremyagost> plain old math.h
[18:47:33] <jeremyagost> ?
[18:48:02] <fuzzie> yes, the Projectile code already uses some
[18:48:38] <jeremyagost> yeah I see it
[18:54:54] <jeremyagost> where was that huge comment about sprite angles?
[18:55:03] <fuzzie> CharAnimations.cpp
[18:55:04] <jeremyagost> actor?
[18:55:06] <jeremyagost> oh ok
[18:58:33] <lynxlynxlynx> ok, bg1 works
[19:01:49] <lynxlynxlynx> a few noncritical but annoying bugs on the way, but that's it
[19:03:24] <Beholder> how mouse events works? I need to equivalent coordinates of UP and DOWN events to handle action (such as processing a button in main menu)?
[19:06:10] <lynxlynxlynx> don't look at me
[19:07:32] --> CJS|2 has joined #gemrb
[19:10:26] <-- _CJS_ has left IRC (Ping timeout: 240 seconds)
[19:13:34] <-- Beholder has left #gemrb
[19:22:11] --> Beholder has joined #gemrb
[19:31:06] <-- jeremyagost has left IRC (Quit: jeremyagost)
[20:06:35] <Yoshimo> i think i found a bug with pickpocketing
[20:08:31] <Yoshimo> with gemrb, you can pick quite some load of stuff , in this case money from npcs, like ratava artsyms 17500 gold in the friendly arm inn, with a 220 pickpocketing skilled thief in the original bg2, you only get a scroll and a gem, and then "the target has no items", only way to get the gold in bg2 is to kill her and then you also get her armor
[20:09:49] <lynxlynxlynx> bg1/bg2
[20:10:01] <Yoshimo> bgt
[20:10:25] <Yoshimo> run with gemrb : problem, original engine, fine
[20:11:12] <Yoshimo> happens also with ribald from bg2
[21:21:56] --> jeremyagost has joined #gemrb
[21:24:54] <Beholder> I'll post experimental SDL1.3 build tomorrow, i think. Only for ARMv7.
[21:29:34] <-- Beholder has left #gemrb
[21:30:52] <Yoshimo> what could cause the pickpocket diffrence i saw?
[21:48:34] <lynxlynxlynx> the obvious, code differences
[21:54:08] <Yoshimo> ive been so far already^^
[23:34:47] <-- Yoshimo has left IRC (Quit: Yoshimo)
[23:52:19] <-- demitar has left IRC (Ping timeout: 260 seconds)
[23:57:10] <-- lynxlynxlynx has left IRC (Remote host closed the connection)