**** BEGIN LOGGING AT Sun Jan 15 02:59:57 2012
Jan 15 15:25:14 <PuffTheMagic> Brybry, i need code that goes from scancode to unicode
Jan 15 15:25:25 <PuffTheMagic> i found some code in SDL_fbevents.c which does it
Jan 15 15:25:34 <PuffTheMagic> but the necessary ioctls fail
Jan 15 15:25:41 <PuffTheMagic> so I cant use that approach
Jan 15 15:28:55 <Brybry> I didn't even think SDL gave valid scancodes?
Jan 15 15:32:32 <PuffTheMagic> SDL doesnt give scancodes
Jan 15 15:32:38 <Brybry> like SDL_keysym.scancode is always 0x00 or something
Jan 15 15:32:42 <PuffTheMagic> ya
Jan 15 15:32:45 <PuffTheMagic> well
Jan 15 15:32:52 <PuffTheMagic> the keyboard provides the scancode
Jan 15 15:32:53 <PuffTheMagic> not SDL
Jan 15 15:33:07 <Brybry> yeah, I mean even from a real keyboard
Jan 15 15:33:24 <PuffTheMagic> you are talking about just SDL on webos?
Jan 15 15:33:33 <PuffTheMagic> or evern sdl on your desktop
Jan 15 15:33:38 <Brybry> maybe I've seen 0x40 but it always the same value from a source, not like a key value
Jan 15 15:33:40 <Brybry> just webos
Jan 15 15:33:47 <PuffTheMagic> well here is the thing
Jan 15 15:33:54 <PuffTheMagic> if unicode processing is enabled
Jan 15 15:34:05 <PuffTheMagic> then it will try and use this method i found
Jan 15 15:34:16 <PuffTheMagic> assuming webos sdl uses the fbcon backend
Jan 15 15:34:24 <PuffTheMagic> but since the ioctl doesnt work
Jan 15 15:34:31 <Brybry> yeah, iirc it uses fbcon
Jan 15 15:34:32 <PuffTheMagic> the scancodes get changed to 0
Jan 15 15:36:24 <PuffTheMagic> the reason this is an issue is that I changed the vkb code to use all sym code
Jan 15 15:36:28 <PuffTheMagic> and no unicode
Jan 15 15:36:40 <PuffTheMagic> i was hoping sdl would do the unicode processing
Jan 15 15:36:43 <PuffTheMagic> but it doesnt
Jan 15 15:38:47 <Brybry> hmm
Jan 15 15:39:41 <Brybry> I wonder if you can just make javascript KeyEvents and send them to the object and it'll magically all work
Jan 15 15:40:15 <PuffTheMagic> well i can go back to sending the unicode from js
Jan 15 15:40:22 <PuffTheMagic> and manually supply the unicode for each key
Jan 15 15:52:30 <Brybry> is there like a list of scancodes somewhere that you're going off of?
Jan 15 16:11:27 <PuffTheMagic> in the kernel code there is
Jan 15 16:12:17 <PuffTheMagic> and SDL_fbevents.c has some code that takes a sdl key event and uses the scancode/modifiers to look up in this map
Jan 15 16:19:58 <PuffTheMagic> Brybry, checkout the branch i pushed
Jan 15 16:34:47 <Brybry> I see what you're trying to do but I'll honestly have to take a look at it tomorrow as I'm going to pass out :)
Jan 15 20:34:25 <PuffTheMagic> dwc-, ping
Jan 15 20:37:00 <PuffTheMagic> another vkb layout for wTerm -> http://ompldr.org/vY2E3Nw/wterm_2012-15-01_153353.png
Jan 15 21:39:03 <sconix> I would rather get rid of right shift and place arrow keys as they should be and then the dedicated del / ins keys around up arrow, at least I am not fond of breaking the arrow key formation
Jan 15 21:39:41 <sconix> just my opinion
Jan 15 21:42:40 <sconix> since at least how the right shift is placed it seems anyway out of place so using it would be bit odd, one expects to find the modifier keys where they should be
Jan 15 21:49:25 <dwc-> agree with sconix about the arrows
Jan 15 21:49:56 <dwc-> http://www.digibarn.com/collections/parts/swyft-card/apple-swyft.jpg
Jan 15 21:50:08 <dwc-> I'd argue those arrows would be better
Jan 15 21:51:01 <dwc-> but it's all customizable, so we should be ok with whatever you pick for your default
Jan 15 21:51:01 <sconix> but why is right shift needed?
Jan 15 21:51:10 <sconix> its a console, does anything separate left/right shift
Jan 15 21:51:36 <dwc-> depends on how people type on their tablet
Jan 15 21:52:21 <sconix> well yes but the keyboard aint anyway 1:1 to the tablet keyboard so at least I would rather loose right shift/ctrl and have properly formatted other extra keys, again just my opinion
Jan 15 21:52:44 <dwc-> and if the scankey sent is the same from left ctrl/alt/shift, as right, then I'd argue alt/ctrl should be dropped before shift
Jan 15 21:53:18 <dwc-> everyone should just use hjkl ;)
Jan 15 21:54:48 <dwc-> what if the cursor keys were made half-height
Jan 15 21:54:59 <dwc-> or even just up/down
Jan 15 21:57:50 <dwc-> or, move shift to the right of up and reorder left/down/right? (shrink right ctrl?)
Jan 15 21:58:36 <sconix> at least placing all arrow keys on the bottom row would be better
Jan 15 21:59:12 <sconix> than the current one, to get the shift in its place
Jan 15 22:00:36 <dwc-> actually, since I hold the tablet in one hand, I might even prefer one of those funky split-layout ones with the tab/caps/ctrl/alt/enter  in the middle
Jan 15 22:00:49 <dwc-> s/since/when/
Jan 15 22:01:17 <dwc-> since the keys I'd hit more (a-z,0-9) are less of a stretch from the edge
Jan 15 22:02:54 <dwc-> oh, no, I guess some of them do retain them, nevermind (http://icorecomputing.com/wp-content/uploads/2011/02/1298444887-24.jpg)
Jan 16 00:02:56 <dtzWill> PuffTheMagic: are you wanting SDL to populate the unicode field for an SDL event?
Jan 16 00:03:07 <dtzWill> i so, there's something like SDL_EanbelUNICODE
Jan 16 00:03:12 <dtzWill> err EnableUNICDOE
Jan 16 02:22:18 <PuffTheMagic> i know about that
Jan 16 02:22:35 <PuffTheMagic> but it doesnt work for pdk apps
Jan 16 02:23:03 <PuffTheMagic> dtzWill, the unicode translations are specific to each of the display mechanisms
Jan 16 02:23:08 <PuffTheMagic> directfb has its own
Jan 16 02:23:13 <PuffTheMagic> fbcon has its own
Jan 16 02:23:19 <PuffTheMagic> x11 has its own
Jan 16 02:23:20 <PuffTheMagic> etc
Jan 16 02:23:34 <PuffTheMagic> and the code which does that for fbcon does not work
Jan 16 02:23:38 <dtzWill> it works fine for pdk apps
Jan 16 02:23:55 <dtzWill> the SDL_EnableUNICODE is used in x11 i was pretty sure
Jan 16 02:24:07 <dtzWill> now if it helps you with your goals, i don't know. i barely remember why i wanted it.
Jan 16 02:24:17 <PuffTheMagic> it might work for x11 apps
Jan 16 02:24:19 <dtzWill> so if you're saying it's worthles to you, you're quite possibly right and sorry :)
Jan 16 02:24:34 <dtzWill> i mean
Jan 16 02:24:41 <dtzWill> x11 is a pdk app
Jan 16 02:24:41 <PuffTheMagic> SDL_EnableUNICODE is enabled in wterm
Jan 16 02:24:44 <PuffTheMagic> it does nothing
Jan 16 02:24:51 <dtzWill> 1 sec
Jan 16 02:25:25 <dtzWill> event.key.keysym.uniode
Jan 16 02:25:34 <dtzWill> should be populate d then
Jan 16 02:25:47 <PuffTheMagic> it doesnt get populated
Jan 16 02:25:47 <dtzWill> *unicode
Jan 16 02:25:51 <dtzWill> dammit wtf is with my internet ?!
Jan 16 02:25:57 <dtzWill> sorry *ahem*
Jan 16 02:26:17 <dtzWill> i actively use it in x11 for touchpad keyboard support
Jan 16 02:26:23 <PuffTheMagic> first, it think the translate to unicode code only gets called for real keyboard events
Jan 16 02:26:39 <dtzWill> oh that's quite possible.
Jan 16 02:26:46 <PuffTheMagic> second, i tried ripping that code out of SDL_fbevent.c
Jan 16 02:26:49 <dtzWill> like i said it might still be worthless, i was just objecting didn't work for pdk apps.
Jan 16 02:26:51 <PuffTheMagic> and i could not get it to work
Jan 16 02:26:52 <dtzWill> aww :(. and no dice?
Jan 16 02:26:53 <dtzWill> x.x
Jan 16 02:27:02 <PuffTheMagic> so if u want to try go ahead, i could have done something wrong
Jan 16 02:27:21 <PuffTheMagic> all the vkb keys are still sending proper sym codes
Jan 16 02:27:37 <PuffTheMagic> so it should be possible to set the scancode to the symcode and then go from there
Jan 16 02:27:38 <PuffTheMagic> idk
Jan 16 02:27:58 <dtzWill> no i'm not saying you're wrong, i was just saying if your only problem is that it wasn't running for pdk apps then that was wrong and so maybe that's helpful
Jan 16 02:28:17 <dtzWill> but sounds like that's not the case and you've done all kinds of other things too
Jan 16 02:28:46 <PuffTheMagic> the code in SDL_fbevents relied on some ioctls that I couldnt get to work
Jan 16 02:28:54 <PuffTheMagic> like i said, i might have been doing it wrong
Jan 16 02:29:41 <PuffTheMagic> you need to open a tty
Jan 16 02:29:51 <PuffTheMagic> and then get some keyboard info from it
Jan 16 02:30:00 <PuffTheMagic> and populate this lookup table
Jan 16 02:40:04 <PuffTheMagic> dtzWill, Brybry dwc- stbuehler will you guys test HEAD of my repo and let me know if there are any issues that need to get fixed before i tag and push a new release
Jan 16 02:40:08 <PuffTheMagic> and also FYI
Jan 16 02:40:16 <PuffTheMagic> when building you have 2 new options to set
Jan 16 02:40:21 <PuffTheMagic> export DEBUG
Jan 16 02:40:25 <PuffTheMagic> and export LOGLEVEL
Jan 16 02:40:37 <PuffTheMagic> debug 1 is -g
Jan 16 02:40:42 <PuffTheMagic> 2 -g -Wall
Jan 16 02:40:50 <PuffTheMagic> 3 -g -Wall -pendantic
Jan 16 02:41:32 <PuffTheMagic> for LOGLEVEL it filters up to a certain syslog severity
Jan 16 02:41:35 <PuffTheMagic> so 0 is nothing
Jan 16 02:41:48 <PuffTheMagic> 7 is everything down to LOG_DEBUG
**** ENDING LOGGING AT Mon Jan 16 02:59:57 2012