JogAmp IRC Chat Logs

#jogamp @ irc.freenode.net - 20130131 05:05:55 (UTC)

20130131 05:05:55 -CatOut- Previous @ http://jogamp.org/log/irc/jogamp_20130130050542.html
20130131 05:05:55 -CatOut- This channel is logged @ http://jogamp.org/log/irc/jogamp_20130131050555.html
20130131 05:09:48 * hharrison (~chatzilla@anon) has joined #jogamp
20130131 06:20:34 * hharrison (~chatzilla@anon) Quit (Ping timeout: 256 seconds)
20130131 06:58:42 <sgothel> good morning (cont. w/ JOAL - got sleepy)
20130131 07:28:48 <sgothel> git clone --recursive git://jogamp.org/srv/scm/joal.git
20130131 07:29:08 <sgothel> openal-soft is now a git-submodule of joal
20130131 07:57:52 * ebrayet (~Adium@anon) has joined #jogamp
20130131 09:49:33 <xranby> sgothel: ok go ahead
20130131 09:50:28 <xranby> sgothel: it will be the first time i get to see git submodules in use
20130131 10:08:57 <sgothel> fixes some joal/gluegen issues (duplicates in ext ..) .. and testing now manually, then on nodes (compile/test)
20130131 10:20:07 <xranby> sgothel: i completed some unit of work on the raspberry pi keyboard driver
20130131 10:20:10 <xranby> last night
20130131 10:20:16 <xranby> its part of the pull request
20130131 10:20:41 <xranby> https://github.com/sgothel/jogl/pull/56
20130131 10:24:16 <sgothel> now we are getting sophisticated 'eh ? :) - nice
20130131 10:26:22 <sgothel> AL lib: alcOpenDevice: Option 'format' is deprecated, please use 'channels' and 'sample-type'
20130131 10:26:22 <sgothel> devices:
20130131 10:26:22 <sgothel> OpenAL Soft
20130131 10:26:22 <sgothel> capture devices:
20130131 10:26:22 <sgothel> Built-in Audio Analog Stereo
20130131 10:26:22 <sgothel> EAX present:false, EAX retrieved: false
20130131 10:26:22 <sgothel> Looping 1: false
20130131 10:26:37 <sgothel> com.jogamp.openal.test.manual.OpenALTest
20130131 10:26:41 <xranby> sgothel: nice!
20130131 10:26:51 <sgothel> shouldn't EAX be there ?
20130131 10:27:08 <xranby> EAX only exist on some pre vista drivers
20130131 10:27:11 <xranby> EAX != EFX
20130131 10:27:19 <sgothel> ah ok, thx
20130131 10:27:21 <xranby> its a creative specific thing
20130131 10:27:53 <xranby> EAX is depricated in favour of EFX
20130131 10:28:44 <xranby> creative dont even include EAX in their latest drivers
20130131 10:52:06 <sgothel> joal linux-arm6hf openal-soft compiled w/o change .. just picked up toolchain ..
20130131 10:52:28 <sgothel> same for build-linux-armv6
20130131 10:52:42 <sgothel> no checking for android
20130131 10:52:58 <sgothel> i.e. we need to squeeze in the android patch from Brice as well now ..
20130131 10:56:47 <xranby> nice, happy that cmake is sane
20130131 11:01:32 <sgothel> ok - need to update to android-ndk-r8d-linux-x86 ..
20130131 12:39:43 * rmk0 (~rmk0@anon) has joined #jogamp
20130131 12:40:02 * rmk0 kicks irssi autojoin
20130131 13:47:19 <xranby> sgothel: i have now mapped up all Linux evdev keys to NEWT VK keys. the hard part i have left is to return a printable character for each VK when pressed and depending on keyboard modifiers in short the keymap in use.
20130131 13:47:52 <xranby> do we have some magic function in place that do all this based on VK keys?
20130131 13:48:38 <sgothel> no - and due to this one bug - it would be layout based - still questionable task
20130131 13:48:45 <sgothel> however, we now have isPrintable() method
20130131 13:49:08 <sgothel> you also would need to respect the SHIFT key and others .. it's quite a task
20130131 13:49:20 <sgothel> .. other modifiers
20130131 13:49:53 <xranby> yup i need a class that abstracts all this logic ... the logic that we usually let the os provide
20130131 13:50:13 <sgothel> if we have that all in Java .. that would be great ofc ..
20130131 13:51:06 <sgothel> https://jogamp.org/bugzilla/show_bug.cgi?id=641 -> discussing semantics of VK_ (virtual key codes) to chars .. incl. keyboard layout etc ..
20130131 13:51:14 <sgothel> but be WARNED: This is huge!
20130131 13:51:32 <xranby> i know.. i am well aware that this is a hairy problem to solve
20130131 13:51:37 <sgothel> i.e. native libs which do this for X11, or other platforms .. well, its huge :)
20130131 13:51:48 <sgothel> you could make a simple 1:1 mapper
20130131 13:52:13 <sgothel> VK_A -> 'a' | 'A' .. VK_1 -> '1' | '!' .. etc depending on SHIFT
20130131 13:52:47 <xranby> yes.. it would talke about 3h constant typing to get a fixed keymap in place
20130131 13:52:53 <sgothel> you could use the isPrintable* in KeyEvent .. as a template
20130131 13:53:07 <sgothel> me fighting w/ Android openal-soft ..
20130131 13:53:49 <xranby> thank you for the input
20130131 13:53:57 <xranby> i will try figure something out
20130131 13:54:01 <sgothel> always lovely how they sneak in undocumented gcc flags in their NDK .. grr
20130131 14:09:37 <ebrayet> @sgothel Hi Sven. I updated the pull request for bugs 677 and 668
20130131 14:10:00 <sgothel> would it be possible to have them separated ?
20130131 14:10:20 <sgothel> however, I will look into it after joal work
20130131 14:10:25 <sgothel> thank you!
20130131 14:10:28 <ebrayet> Sure
20130131 14:11:34 <ebrayet> Hum, I'm not sure how to separate a pull request in two, but I'll give it a try
20130131 14:31:52 <xranby> ebrayet: i would suggest to leave the current pull request and branch as is. then two new brances on top of sgothel master checkout one of the new branched and then cherry pick the changes needed checkout the other new branch and cherry pick the changes needed there as well
20130131 14:38:33 <ebrayet> Ok. And then create 2 new pull requests, one for each new branch, and let Sven refuse the one on master, right ?
20130131 14:44:38 <xranby> yes sounds right
20130131 15:19:13 <sgothel> sven@risa:/usr/local/projects/JOGL/joal/make$ file ../build-android-armv6/openal-soft/libopenal.so.1.15.1
20130131 15:19:14 <sgothel> ../build-android-armv6/openal-soft/libopenal.so.1.15.1: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), dynamically linked (uses shared libs), not stripped
20130131 15:20:52 <xranby> very nice!
20130131 15:22:57 <xranby> sgothel: can you check readelf --program-headers libopenal.so.1.15.1
20130131 15:22:58 <sgothel> now adding it to APK .. and create a test activity ..
20130131 15:23:23 <xranby> to see which dynamic linker the library will use
20130131 15:23:38 <xranby> linux usually want [Requesting program interpreter: /lib/ld-linux.so.3]
20130131 15:23:55 <xranby> while android want [Requesting program interpreter: /system/bin/linker]
20130131 15:24:20 <xranby> -Wl,--dynamic-linker=/system/bin/linker
20130131 15:24:47 <sgothel> yeah .. but don't see that entry in readelf dump
20130131 15:26:55 <sgothel> readelf --headers .. does
20130131 15:27:13 <sgothel> Requesting program interpreter: /system/bin/linker - good
20130131 15:28:23 * xranby realise that a standard linux console keymap is about 7384 lines
20130131 15:28:31 <xranby> for one language ;)
20130131 15:29:02 <xranby> thats 390kb data
20130131 15:29:08 <sgothel> can't you just use the existing native functionalities ? aren't there any ?
20130131 15:29:18 <xranby> i have not seen any
20130131 15:29:56 <xranby> the data i get is what the keyboard manufacturer decided
20130131 15:30:41 <xranby> raw keycodes
20130131 15:30:49 <sgothel> b/c you hook to a raw input device ?
20130131 15:30:54 <xranby> yes
20130131 15:30:57 <sgothel> isn't there a translated one ?
20130131 15:31:19 <sgothel> however .. we support less than 255 keys .. hmm
20130131 15:34:16 <xranby> the linux console also use less than 255 keycodes.
20130131 15:34:39 <xranby> a mapping for a single key in this example the nr 4 key looks like
20130131 15:34:44 <sgothel> i.e. just use the keys as mentioned in isPrintable .. IMHO
20130131 15:34:44 <xranby> keycode 5 = four
20130131 15:34:44 <xranby> altgr keycode 5 = dollar
20130131 15:34:44 <xranby> control keycode 5 = Control_backslash
20130131 15:34:44 <xranby> shift control keycode 5 = Control_backslash
20130131 15:34:44 <xranby> altgr control keycode 5 = Control_backslash
20130131 15:34:44 * xranby (~xranby@anon) Quit (Excess Flood)
20130131 15:35:27 * xranby (~xranby@anon) has joined #jogamp
20130131 15:35:35 <sgothel> don't flood :)
20130131 15:35:39 <xranby> :)
20130131 15:41:20 <xranby> ok that is nice it will solve the issue for most keys quickly
20130131 15:53:25 <xranby> https://github.com/xranby/jogl/commit/ae89ca7342b8673e38b7d76bff386175e84a9173
20130131 15:57:07 <xranby> i will now depart, pack and head to fosdem, i will try to be available this evening
20130131 15:57:25 <sgothel> ay .. laters - me trying to get joal in .. then we will see ..
20130131 15:58:36 <xranby> ok! if possible try get brayet's key fixes in and similar pull requests
20130131 15:58:51 <sgothel> ay
20130131 15:58:54 <xranby> big kudos
20130131 16:01:19 * xranby (~xranby@anon) Quit (Quit: Leaving.)
20130131 17:08:03 * ebrayet (~Adium@anon) Quit (Quit: Leaving.)
20130131 19:43:18 * hharrison (~chatzilla@anon) has joined #jogamp
20130131 19:43:23 * xranby (~familjen@anon) has joined #jogamp
20130131 21:59:33 * gouessej (5279c53d@anon) has joined #jogamp
20130131 21:59:39 <gouessej> hi
20130131 22:02:22 <sgothel> hi Julien
20130131 22:02:28 <rmk0> lo
20130131 22:03:08 <sgothel> already 11pm .. uh .. joal needs a few replacements for Android I guess .. trying to circumvent javax.audio* stuff ..
20130131 22:05:26 <gouessej> I will copy the content of my hard drive for the demo
20130131 22:05:35 <gouessej> LibGDX works fine
20130131 22:05:42 <sgothel> awesome!
20130131 22:05:53 <gouessej> JMonkeyEngine 3 will need some more work for ES2
20130131 22:08:02 <xranby> sgothel: i ran some tests of the new joal builds, the built libopenal.so work on the ac100 using a cortex-a9 but fail on armv6
20130131 22:08:13 <gouessej> Sven, do you still have my phone number?
20130131 22:08:34 <xranby> sgothel: i can workaround it for fosdem using -Djoal.SystemOpenAL
20130131 22:08:38 <sgothel> I don't know - maybe re-sent via email - thx!
20130131 22:08:43 <gouessej> ok
20130131 22:09:07 <sgothel> oh .. so some cmake settings are too harsh .. for arm
20130131 22:09:24 <xranby> sgothel: possibly using the cross compilers default
20130131 22:09:28 <sgothel> yeah .. on Android it already loads .. inits .. fails loading Wav file .. finishing demo then I push
20130131 22:10:02 <sgothel> if you know how - go ahead and make such a cmake file like for android ..
20130131 22:10:49 <xranby> gouessej: hi, will you arrive toporrow?
20130131 22:10:55 <xranby> to morron in brussels?
20130131 22:10:58 <xranby> morrow
20130131 22:11:15 <gouessej> I arrive Saturday morning in Brussels
20130131 22:11:44 <xranby> ok we meet then!
20130131 22:14:40 <sgothel> @Xerxes: when I hear sound on Android - I push joal and test builds
20130131 22:16:30 <xranby> sgothel: good!
20130131 22:16:55 <sgothel> I hear a little piezo sound .. from speaker - not exactly the wav file :)
20130131 22:16:56 <xranby> sgothel: you have really done wonders to joal today
20130131 22:17:31 <xranby> i assume you hear something better on desktop?
20130131 22:17:39 <sgothel> sure
20130131 22:21:29 <sgothel> pushed current state
20130131 22:21:59 <gouessej> I've just sent you my phone number ;)
20130131 22:22:20 <sgothel> thx
20130131 22:22:28 <xranby> sgothel: gouessej: i got the same phone number i used in LA
20130131 22:22:30 <gouessej> I hope I will have something better to show for Anaheim
20130131 22:23:13 <gouessej> Xerxes, maybe we will have some time to make JMonkeyEngine 3 work on your Raspberry Pi
20130131 22:23:14 <xranby> gouessej: can you demo java3d?
20130131 22:23:47 <gouessej> I could demo Java3D but I have nothing really interesting to show
20130131 22:23:53 <gouessej> with it
20130131 22:24:23 <gouessej> I can demo LibGDX and JMonkeyEngine 3.0 for sure
20130131 22:25:01 <sgothel> hi Harvey: Java3D from Assenco (sic!) ?
20130131 22:26:44 <xranby> gouessej: if we can mention that java3d is back then it would be great.
20130131 22:28:05 <xranby> gouessej: and of course demo libgdx and jme3, its excellent work!
20130131 22:29:37 <xranby> the list of things that have been done during the last months is really huge, good to see continuation and revival in community run spirit
20130131 22:30:26 <gouessej> the game based on Java3D crashed the last time I wanted to use it
20130131 22:31:07 <gouessej> I don't see the interest of showing something very poor
20130131 22:31:29 <gouessej> Should I show SweetHome 3D once again?
20130131 22:31:43 <xranby> that would be fine
20130131 22:31:47 <sgothel> now really w/ new Java3D ? sure
20130131 22:36:33 <gouessej> I'll speak about that to Emmanuel Puybaret
20130131 22:42:19 <hharrison> Yeah, probably best to show sweethome, here at Ausenco, we've mostly moved to a direct JOGL2 backend anyway
20130131 22:42:31 <hharrison> The Java3d one has become more and more 'legacy'
20130131 22:43:27 <hharrison> I am _extremely_ jealous of the fun weekend you gentlemen have in store
20130131 22:47:53 <gouessej> Harvey, if you need another scenegraph, let me know
20130131 22:50:52 <hharrison> We don't we literally wrote our own direct-to-opengl framework
20130131 22:51:11 <hharrison> That was what I pushed out to github the other day, in case anyone has time to poke it
20130131 22:51:28 <hharrison> Still lots of legacy from the old j3d way of working, but way more performant
20130131 22:51:47 <hharrison> ...basically, excuse the mess, work in progress
20130131 23:06:17 <gouessej> Harvey, I'll give it a look soon ;) Good night
20130131 23:06:24 * gouessej (5279c53d@anon) Quit (Quit: Page closed)
20130201 01:32:52 * xranby (~familjen@anon) has left #jogamp
20130201 02:28:55 * hharrison (~chatzilla@anon) Quit (Quit: ChatZilla 0.9.89 [Firefox 19.0/20130123083802])
20130201 03:04:12 * sgothel (~sven@anon) Quit (Quit: Leaving.)
20130201 04:18:55 * hharrison (~chatzilla@anon) has joined #jogamp
20130201 04:19:20 * masterzen (~masterzen@anon) Quit (Ping timeout: 276 seconds)
20130201 04:23:18 * masterzen (~masterzen@anon) has joined #jogamp
20130201 05:06:08 -CatOut- Continue @ http://jogamp.org/log/irc/jogamp_20130201050608.html