#gemrb@irc.freenode.net logs for 25 Oct 2012 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:14:21] --> Canageek has joined #gemrb
[02:36:59] <-- Canageek has left IRC (Quit: KVIrc 4.0.4 Insomnia The future is here. It's just not widely distributed yet. - William Gibson)
[03:23:51] <-- |Cable| has left IRC (Ping timeout: 260 seconds)
[03:37:38] --> |Cable| has joined #gemrb
[04:42:44] --> lynxlynxlynx has joined #gemrb
[04:42:44] <-- lynxlynxlynx has left IRC (Changing host)
[04:42:44] --> lynxlynxlynx has joined #gemrb
[04:42:44] --- ChanServ gives channel operator status to lynxlynxlynx
[05:44:25] --> edheldil_ has joined #gemrb
[06:59:00] <-- edheldil_ has left IRC (Read error: Operation timed out)
[07:38:14] <-- Drakkar has left IRC (Ping timeout: 244 seconds)
[07:47:22] --> Drakkar has joined #gemrb
[08:01:45] --> rocket_hamster has joined #gemrb
[10:04:34] <-- rocket_hamster has left IRC ()
[10:23:36] --> rocket_hamster has joined #gemrb
[10:29:40] <lynxlynxlynx> rocket_hamster: today i'll have time again
[10:34:59] <rocket_hamster> yeah i should be back home at 19 00
[10:35:10] <rocket_hamster> have to go to school to listen to some crap :/
[10:35:32] <rocket_hamster> so is 1900 acceptable?
[10:37:00] <lynxlynxlynx> sure
[10:37:13] <lynxlynxlynx> just ping me once you start your cycle
[10:43:12] <rocket_hamster> sure thing, cya later
[10:43:14] <-- rocket_hamster has left IRC ()
[13:51:37] --> WingedHussar has joined #gemrb
[14:00:32] <-- WingedHussar has left IRC (Ping timeout: 252 seconds)
[14:01:35] --> WingedHussar has joined #gemrb
[15:36:08] --> brada has joined #gemrb
[15:36:13] <brada> check it out http://hg.libsdl.org/SDL/rev/533131e24aeb
[15:36:42] <brada> may now more than mac and ios can give people gui alerts :)
[15:36:46] <fuzzie> useful
[15:46:02] <lynxlynxlynx> cool
[15:58:03] --> Beholder has joined #gemrb
[15:58:29] <-- xrogaan has left IRC (Ping timeout: 252 seconds)
[16:00:16] <Beholder> hi
[16:02:05] --> Guest13362 has joined #gemrb
[16:03:31] <-- WingedHussar has left IRC (Quit: WingedHussar)
[16:05:28] <-- Guest13362 has left IRC (Remote host closed the connection)
[16:05:50] <-- brada has left IRC (Quit: brada)
[16:16:20] --> xrogaan_ has joined #gemrb
[16:16:20] <-- xrogaan_ has left IRC (Changing host)
[16:16:20] --> xrogaan_ has joined #gemrb
[16:45:42] --> brada has joined #gemrb
[16:51:40] <Beholder> hi Brad
[16:51:56] <brada> hey. been awhile
[16:52:13] <Beholder> some questions to you
[16:52:40] <brada> ok lets hear
[16:53:40] <Beholder> SDL20VideoDriver::ProcessFirstTouch
[16:53:40] <Beholder> if (!(MouseFlags & MOUSE_DISABLED) && firstFingerDown.fingerId) - always false in android. This is reason that touch screen not work
[16:53:55] <brada> i see
[16:54:00] <brada> let me think for a min
[16:54:08] <fuzzie> zero fingerId?
[16:54:14] <brada> do you know which parts are false?
[16:54:19] <brada> there are 2 :)
[16:55:01] <brada> glad to see you are again working on getting multitouch working
[16:55:35] <brada> wich we could get true SDL 2 rendering working so performance wouldnt be abysmal
[16:56:36] <Beholder> !(MouseFlags & MOUSE_DISABLED) is always true, !(0&2)
[16:56:42] <Beholder> in menu
[16:56:54] <Beholder> yes fingerid=0 for first finger
[16:57:14] <Beholder> true & false = false
[16:58:41] <brada> well i guess you have to fake finger id somehow?
[17:00:24] <Beholder> you check only first finger in this function?
[17:00:41] <brada> I thought i had to use it to get SDL_Touch, but that is actually touchId
[17:01:09] <brada> so looks like fingerid is only being used to make sure there is a touch
[17:01:26] <brada> so thats pretty easy to code around
[17:02:13] <Beholder> If I touch to a screen this function calls many times, why?
[17:02:23] <brada> i could probably either check something else or you could just make sure to set it to 1 whenever i assign to firstFinger
[17:02:36] <brada> which function?
[17:03:11] <Beholder> SDL20VideoDriver::ProcessFirstTouch. First call is valid, fingerid=0 and coordinates are proper
[17:03:38] <Beholder> all another call get a garbage in fingerid and coordinates =0,0
[17:04:47] <Beholder> I do not studied you code very well
[17:04:52] <brada> wherever the code is assigning a valid touch to firstFinger try forcing fingerId to 1
[17:05:37] <brada> looks like line 410 is the only place that happens
[17:06:22] <brada> or better yet the code should probably be using touchId
[17:06:54] <brada> you can try replacing fingerId with touchId
[17:07:48] <brada> touchid HAS to be valid or the code wont work for other reasons
[17:08:06] <brada> i use the touchId to get SDL_Touch structs
[17:08:33] <lynxlynxlynx> the fact that it gets run multiple times sounds interesting too
[17:08:40] <Beholder> one min, i'll try
[17:11:09] <brada> i dont get why the android implementation isnt assigning a fingerId
[17:11:21] <brada> but hopefully it doesnt matter
[17:11:39] <lynxlynxlynx> or maybe it enumerates them from 0 :)
[17:11:49] <brada> beholder: i assume you are using the latest sdl
[17:12:15] <brada> lynx: possibly, but it seems queer that it would be diffrent between ios and android
[17:12:33] <lynxlynxlynx> yeah
[17:12:49] <lynxlynxlynx> not that stuff like that doesn't happen though -.-
[17:12:53] <brada> true
[17:13:14] <Beholder> replaced fingerid by touchid in ProcessFirstTouch. Buttons in menu pressed well, but nothing happens. Buttons stay pressed))) Of course latest SDL is used
[17:13:32] <brada> you replaced fingerId EVERYWHERE it is used?
[17:13:46] <Beholder> no
[17:13:53] <Beholder> replace it?
[17:13:55] <brada> do replace them all
[17:13:57] <brada> yes
[17:13:58] <Beholder> ok
[17:15:07] <brada> its a good sign that the buttons pressed
[17:15:19] <brada> that means this is not hopeless :)
[17:19:13] <Beholder> hm works, i see that a little finger motion disables action
[17:19:48] <brada> interesting
[17:19:56] <brada> this works well on ios
[17:20:19] <Beholder> or not .... If I do my touches very fast it works good
[17:20:20] <brada> so it would be hard for me to say what is wrong
[17:20:51] <Beholder> but if i hold my finger some times not work...
[17:21:13] <brada> you may have to turn on the on screen log and add some log statements to the code
[17:21:19] <Beholder> may be i can increase delay?
[17:21:30] <brada> delay?
[17:21:57] <brada> you mean GetRKDelay?
[17:22:25] <brada> did you remember to replace fingerId on line 298?
[17:23:41] <Beholder> of course i've replaced it. what is GetRKDelay do?
[17:23:57] <brada> i believe it is for repeating events
[17:24:12] <brada> like when you pres and hold the thif skill + or - buttons
[17:24:19] <brada> thief
[17:24:21] <lynxlynxlynx> only used for that kind of stuff though
[17:24:43] <lynxlynxlynx> not on game control
[17:24:48] <brada> right, but im wondering if that code is getting triggered on android
[17:25:07] <lynxlynxlynx> it's our extension, so we have to set it on the control explicitly
[17:26:20] <brada> i cant think of what would cause the behavior he describes
[17:27:43] <lynxlynxlynx> emulated rightclick maybe
[17:27:53] <brada> yeah
[17:27:57] <brada> that sounds likely
[17:27:57] <lynxlynxlynx> movement would be obvious with feedback on
[17:28:45] <brada> right click delay is currently set by the TOUCH_RC_NUM_TICKS define
[17:29:12] <brada> so maybe on android firstFingerDownTime is not set
[17:29:31] <Beholder> control works properly only if i knocking screen, any delay results action not works
[17:29:32] <brada> no that is our code...
[17:30:01] <brada> beholder: are you just touching and holding still?
[17:30:07] <brada> because that is a right click
[17:30:37] <Beholder> only small time holding, one second
[17:30:49] <brada> right so you are right clicking
[17:31:18] <brada> touch + hold still for 500 ticks is a right click
[17:31:38] <brada> you can increase the number of ticks by the afore mentioned define
[17:31:48] <Beholder> where?
[17:32:26] <brada> line 31
[17:33:01] <brada> i need to find time to work on this code to fix the keyboard gesture
[17:33:10] <Beholder> buttons on windows works very strange, it causes actions 1 time from 3. any times it stay pressed and do no action
[17:33:53] <brada> i dont know what to tell you since this works fine on the only touch system i have
[17:34:00] <Beholder> *any other times
[17:35:55] <Beholder> can you try start gemrb on iphone?
[17:36:01] <brada> if the controls stay pressed that probably means EvntManager->MouseUp is not getting called
[17:36:14] <brada> beholder: i have run it in simulator yes and it works
[17:36:32] <brada> i dont have an actuall device, but others have confirmed to me it works fine
[17:38:48] <brada> the only thing that i know is broken is the keyboard/console gesture
[17:38:50] <Beholder> hm, with MIN_GESTURE_DELTA_PIXELS 10, not really. In iphone this should be 40-50 minimum, I think.
[17:38:56] <brada> and i know why and what to do to fix that
[17:39:25] <brada> beholder: i plan on getting rid of that define
[17:39:33] <brada> but yes that is roughly why it is broken
[17:40:11] <brada> i need to not pop the console until it is a new swipe event
[17:40:22] <brada> currently the same even can do both
[17:40:36] --> rocket_hamster has joined #gemrb
[17:41:09] <brada> the TOUCH_RC_NUM_TICKS should be turned into a config option too
[17:41:42] <rocket_hamster> yo
[17:42:09] <rocket_hamster> lynx ill have to set up home server and try NAT on it instead of router
[17:42:19] <Beholder> hm, you're right mouse up not works as it should
[17:43:03] <brada> right, but the question remains why
[17:43:09] <brada> since it does work on iOS
[17:43:45] <brada> it will be painfull for you to debug most liekly :p
[17:44:00] <brada> but we have an on screen logger now so that should help
[17:44:33] <Beholder> if i try to select party members by selection shape, shape does not disappears from screen(
[17:44:56] <brada> that is a known issue that occurs sometimes
[17:45:11] <brada> but it doesnt effect gameplay so i havent bothered fixing it
[17:45:42] <lynxlynxlynx> rocket_hamster: ok, tell me when/if you're ready
[17:50:58] <rocket_hamster> ok probably not so soon :/
[17:52:54] <brada> beholder: what did you change line 350 to?
[17:53:24] <brada> it probably has noting to do with your current issue tho
[17:53:35] <brada> since that only applies to multi-finger events
[17:56:33] <-- rocket_hamster has left IRC (Ping timeout: 252 seconds)
[17:59:22] <brada> anyway for you line 350 shoud be firstFingerDown.touchId = event.tfinger.touchId;
[18:05:16] --> raevol has joined #gemrb
[18:05:32] <raevol> so that direwolf on the gorion death map is an issue outside of the save game i had
[18:05:49] <brada> what difficulty are you on?
[18:05:57] <raevol> i think the original BG1 has some sort of exception when it first loads that map and loads a special set of mobs
[18:06:02] <raevol> easiest i believe
[18:08:01] <brada> what exactly is the problem?
[18:21:28] <lynxlynxlynx> too tough random spawns
[18:21:48] <lynxlynxlynx> we just don't understand it enough
[18:21:53] <brada> i see
[18:22:30] <Beholder> Brad, why you don't check deltas to MIN_GESTURE_DELTA_PIXELS before setting ignoreNextFingerUp to true in SDL_FINGERMOTION. This is a reason of my touch problems.
[18:24:20] <Beholder> touchscreen in my device is not so precise, or my fingers too thick and simulate motions)))
[18:24:21] <brada> i wonder why ios has no problem
[18:24:44] <brada> so did you add that and it fixes everything?
[18:24:57] <brada> at least everything critical
[18:26:06] <brada> if thats the case then a better solution would probably be to check against that elsewhere
[18:26:29] <brada> before event handeling switch
[18:28:15] <raevol> you know, something that i noticed is that the autopause settings weren't really being handled as i set them in the options
[18:28:16] <brada> above the switch you can put something like: if event type is gesture but is < MIN_GESTURE_DELTA_PIXELS then return
[18:28:30] <raevol> it seems like all autopause settings are enabled by default, and don't really disable when you mess with them
[18:28:44] <raevol> i wonder if the difficulty setting as it is set in the menu is being ignored
[18:30:22] <brada> beholder: probably bbest to put that at the top of Process event
[18:30:27] <raevol> sorry i don't have hard facts on these issues, i chat with you guys at work in the morning, then go home at the end of the day and play, and haven't been writing down actual information
[18:30:36] <brada> otherwise we would do a bunch of useless processing
[18:32:27] <brada> beholder: what device do you test with?
[18:32:41] <brada> im curious if the fullscreen scaling works on android
[18:33:43] --> rocket_hamster has joined #gemrb
[18:33:53] <Beholder> HTC DesireHD
[18:34:02] <brada> what is the screen resolution?
[18:34:18] <Beholder> 800x480 pix 4.3 inch
[18:34:55] <rocket_hamster> sry lynx
[18:35:00] <rocket_hamster> i cant get NAT working properly
[18:35:01] <brada> so if you set resolution to 640x480 in config and set fullscreen=1 does it still work?
[18:35:30] <Beholder> checking for MIN_GESTURE_DELTA_PIXELS avoiding all problems with clicking by providing problems with gestures)
[18:35:55] <Beholder> I use 800x480 with widescreen
[18:37:19] <lynxlynxlynx> raevol: depends on your gemrb version, but both should work in git head; the settings don't get saved though
[18:38:02] <Beholder> I think buffering deltas would help me
[18:40:43] <rocket_hamster> lynx u still avaible?
[18:42:19] <brada> does widescreen disable the native resolutions?
[18:42:35] <brada> because id still like you to test it if you can
[18:48:27] <raevol> lynxlynxlynx: i'll try to remember testing resetting the difficulty before the gorion death map is loaded, see if it pulls a different mob table
[18:50:13] <brada> beholder: ill try to find time next week to mess with the touch input
[18:50:43] <rocket_hamster> who is working on bg2 gui?
[18:53:20] <Beholder> Not needed. Setting MIN_GESTURE_DELTA_PIXELS to 4 and checking it on SDL_FINGERMOTION avoid all my problems. But SDL still buggy((( App not closed properly, dying when phone sleeps and other. I ask to Pelya for new release of SDL2.
[18:53:52] <brada> well why is it dying?
[18:54:10] <lynxlynxlynx> rocket_hamster: sure
[18:54:11] <brada> becuae gemrb does try to react to those events
[18:54:36] <lynxlynxlynx> raevol: i doubt the difficulty affects that
[18:54:40] <brada> ie does it crash because of what we do in SDL_WINDOWEVENT_MINIMIZED?
[18:55:11] <rocket_hamster> lemme reboot lynx!
[18:57:29] --> rockethamster has joined #gemrb
[18:57:51] <-- rocket_hamster has left IRC (Read error: Connection reset by peer)
[18:58:14] <rockethamster> rdy
[19:00:24] <Beholder> I don't know, and if i trying to exit by pressing 'EXIT' menu button and confirm, game frozen, not exit.
[19:02:12] <brada> well im wondering if SDL_WINDOWEVENT_MINIMIZED event happens after our audo driver is destroyed
[19:02:34] <brada> try commenting out line 491 of sdl20video
[19:04:25] <Beholder> tomorrow, get some sleep)
[19:04:58] <Beholder> bye
[19:06:04] <-- Beholder has left #gemrb
[19:06:33] <brada> well at leats touch input finally works for android :)
[19:11:44] <raevol> lynxlynxlynx, brada, difficulty doesn't effect which monsters spawn?
[19:12:10] <brada> i thought it did, but that is not my realm of gemrb knowledge
[19:12:28] <brada> or even original implementation knowledge
[20:08:05] <raevol> well, i know enough to know that a dire wolf spawning in that area is a bug, in the original game, every time i played i cleared that area, and the only thing that ever may have killed me was a black bear if i aggroed it
[20:08:21] <brada> i recall regular wolves
[20:09:23] <raevol> yea
[20:12:44] <lynxlynxlynx> sure
[20:15:50] <brada> could it be cuased by dire wolf spawns in later chapters?
[20:16:02] <brada> doesnt dire wolves spawn there in chapter 6?
[20:26:54] <raevol> that's what i was thinking, that there's two different mob tables (or whatever they are) and it's picking the higher level one where it should be picking the lower one
[20:27:25] <fuzzie> is this bg1?
[20:27:38] <fuzzie> because, bg1 spawns are quite weird
[20:28:01] <raevol> yea
[20:28:02] <fuzzie> ah yes, gorion spawn map
[20:28:09] <fuzzie> it might just be our algorithm being buggy still, honestly
[20:28:14] <raevol> bg1, the unloved stepchild of gemrb ;) hehe
[20:28:17] <fuzzie> I am really truly buried in house moving stuff though. :/
[20:28:36] <raevol> no worries
[20:29:25] <raevol> do you guys have a roadmap or a bug tracker?
[20:30:23] <fuzzie> yes
[20:30:27] <fuzzie> in theory :p
[20:30:39] <fuzzie> todo wiki page is still most-used I think
[20:30:42] <raevol> is it the wiki? haha
[20:31:05] <fuzzie> yeah :
[20:31:06] <fuzzie> :/
[20:31:56] <lynxlynxlynx> :)
[20:32:39] <raevol> i'm learning a lot about open source projects helping out with openmw
[20:32:48] <fuzzie> ah cool
[20:32:53] <raevol> wish i was a coder though :(
[20:33:01] <fuzzie> presumably they have quite similar challenges with regards to keeping track of bugs etc
[20:33:46] <raevol> https://bugs.openmw.org/projects/openmw/issues
[20:34:05] <fuzzie> conspicious lack of issues there though :P
[20:34:24] <raevol> well, it's in severe pre-alpha
[20:34:37] <fuzzie> yeah, it gets really nightmareish later on
[20:34:40] <raevol> a ton of "this doesn't work" bugs aren't helpful when you've only implemented a tiny portion of the functionality hehe
[20:34:47] <fuzzie> right, but you get them anyway usually :P
[20:35:09] <fuzzie> but of course even the original developers have *ridiculous* numbers of bugs to track
[20:35:16] <raevol> we don't have many lolligagger users posting bugs, for some reason most of our users understand it's very much a work in progress
[20:35:28] <fuzzie> you can see the crazy amounts of bugs fixed in the patches to games like bg and mw, even after a huge complex QA process
[20:35:48] <fuzzie> and then the fan patches fixing the huge amounts of stuff they missed..
[20:36:15] <raevol> from my experience they usually didn't miss it, usually the producer wanted the thing out the door without fixing them, hehe
[20:36:39] <fuzzie> well I know that at least for bg, there was a huge amount of QA
[20:36:50] <raevol> yea
[20:36:53] <fuzzie> there's just too many fiddly interactions to ever do it perfectly
[20:37:00] <raevol> closed QA and open QA are very different beasts
[20:37:05] <fuzzie> but yeah you have to ship at some point
[21:21:48] <-- rockethamster has left IRC (Ping timeout: 260 seconds)
[21:21:57] --> rockethamster has joined #gemrb
[21:32:11] <-- rockethamster has left IRC (Quit: Leaving)
[21:45:41] --> rocket_hamster has joined #gemrb
[22:42:39] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[22:51:22] <-- brada has left IRC (Quit: brada)
[23:18:06] <-- raevol has left #gemrb
[23:33:49] <-- rocket_hamster has left IRC (Read error: No route to host)
[23:34:12] --> rocket_hamster has joined #gemrb