#jogamp @ irc.freenode.net - 20130512 05:05:53 (UTC)
20130512 05:05:53 -CatOut- Previous @ http://jogamp.org/log/irc/jogamp_20130511050639.html
20130512 05:05:53 -CatOut- This channel is logged @ http://jogamp.org/log/irc/jogamp_20130512050553.html
20130512 10:04:52 * xranby_ac100 (~xranby@anon) has joined #jogamp
20130512 10:12:17 <xranby_ac100> good day
20130512 10:46:55 * Guest41141 is now known as rmk0
20130512 10:46:58 * rmk0 (~rmk0@anon) Quit (Changing host)
20130512 10:46:58 * rmk0 (~rmk0@anon) has joined #jogamp
20130512 10:47:04 <rmk0> 'lo
20130512 10:47:26 <rmk0> ... no sgothel?
20130512 10:48:16 <xranby_ac100> rmk0: sgothel is taking the time off for some days
20130512 10:50:10 <xranby_ac100> 20130506 19:45:25 <sgothel> btw: out of home/office: May 9th - May 15th
20130512 10:51:25 <rmk0> ah, right
20130512 10:51:28 <rmk0> he deserves it!
20130512 10:51:47 <xranby_ac100> btw.. if someone enjoy working on a/v sync code.. i have a FFMPEGMediaPlayer implementation with sound under X11/Linux but the a/v sync code can be improved quite a lot
20130512 10:51:47 <rmk0> never seen someone work so intensely on one single project for so long
20130512 10:53:23 <xranby_ac100> https://github.com/xranby/jogl/commits/FFMPEGMediaPlayer-audio
20130512 10:54:39 <xranby_ac100> i am working on https://jogamp.org/bugzilla/show_bug.cgi?id=725
20130512 10:55:46 <rmk0> nice
20130512 10:58:16 <xranby_ac100> rmk0: i basically need to have a video frame cache as well, right now video frames is handed over directly to OpenGL which is fast but prevents good a/v sync
20130512 10:59:55 <rmk0> any licensing issues with ffmpeg?
20130512 11:00:00 <rmk0> random parts of it are GPLd
20130512 11:01:28 <xranby_ac100> rmk0: right now we avoid the licensing issue by dynamicaly linking to the system installed libav/ffmpeg if jogamp want to bundle libav/ffmpeg in the future then we need to make sure we only include lgpl codecs and codecs with patent issues
20130512 11:01:29 * [Mike] (~Mike]@anon) Quit (Quit: Changing server...)
20130512 11:01:48 <rmk0> hm, right
20130512 11:02:14 * [Mike] (~Mike]@anon) has joined #jogamp
20130512 11:02:21 * [Mike] (~Mike]@anon) Quit (Client Quit)
20130512 11:02:53 * [Mike] (~Mike]@anon) has joined #jogamp
20130512 11:03:03 <xranby_ac100> rmk0: https://jogamp.org/bugzilla/show_bug.cgi?id=686 - Add libav native Libraries for GLMediaPlayer fallback w/ patent free Codecs like WebM
20130512 11:09:24 * [Mike] (~Mike]@anon) Quit ()
20130512 11:46:08 * DemoscenePassiv (~Lutsche@anon) has joined #jogamp
20130512 12:25:38 * mjacob_ (~foobar@anon) Quit (Read error: Operation timed out)
20130512 12:28:53 * mjacob (~foobar@anon) has joined #jogamp
20130512 14:23:08 * mjacob (~foobar@anon) Quit (Ping timeout: 256 seconds)
20130512 14:24:14 * xranby_ac100 (~xranby@anon) Quit (Remote host closed the connection)
20130512 15:16:04 * Lusche (~Lutsche@anon) has joined #jogamp
20130512 15:24:46 * DemoscenePassiv (~Lutsche@anon) Quit (*.net *.split)
20130512 19:39:26 * void256 (~void@anon) has joined #jogamp
20130512 20:48:39 * gouessej (527946de@anon) has joined #jogamp
20130512 20:48:48 <gouessej> Hi
20130512 20:49:06 <void256> Hi
20130512 20:49:25 <gouessej> When I use a MonitorMode whose rotation is not zero, it doesn't work as expected, only the behaviour of the mouse changes
20130512 20:49:52 <gouessej> this was working just before the major changes in the API, with the addition of MonitorMode and MonitorModeUtil
20130512 20:52:08 <gouessej> Ok I'll speak to Sven when he's back
20130512 20:52:21 * gouessej (527946de@anon) Quit (Client Quit)
20130512 21:02:26 * xranby_ac100 (~xranby@anon) has joined #jogamp
20130512 21:03:06 * xranby_ac100 is reading http://www.ibm.com/developerworks/library/j-jni/ Best practices for using the Java Native Interface
20130512 21:07:25 <void256> JNA?
20130512 21:09:42 <void256> I mean I have no idea what you're doing :) but JNA is pretty cool too ...
20130512 21:09:59 <void256> or http://www.swig.org/
20130512 21:12:13 <xranby_ac100> void256: i am reviewing how we currently pass decoded audio and video frames from libav/ffmpeg up to jogl in the FFMPEGMediaPlayer backend
20130512 21:14:05 <xranby_ac100> ideally we dont want to copy the arrays of decoded data in order to build an efficient cache and a/v frame sync
20130512 21:15:53 <xranby_ac100> most of jogl code is generated by gluegen that do this efficiently
20130512 21:16:22 <xranby_ac100> but in the ffmpegmediaplayer case we use some handwritten code
20130512 21:18:27 <xranby_ac100> http://jogamp.org/gluegen/www/
20130512 22:08:39 * Lusche (~Lutsche@anon) Quit (Ping timeout: 256 seconds)
20130512 22:12:03 * xranby_ac1001 (~xranby@anon) has joined #jogamp
20130512 22:12:03 * xranby_ac100 (~xranby@anon) Quit (Read error: Connection reset by peer)
20130512 22:15:06 * void256_ (~void@anon) has joined #jogamp
20130512 22:16:22 * void256 (~void@anon) Quit (Ping timeout: 246 seconds)
20130512 22:16:30 * void256_ is now known as void256
20130512 22:18:47 <xranby_ac1001> void256: btw. of you want to disable vsync use gl.setSwapInterval(0);
20130512 22:19:21 <xranby_ac1001> you asked about it some days ago
20130512 22:21:32 <xranby_ac1001> http://jogamp.org/deployment/jogamp-current/javadoc/jogl/javadoc/javax/media/opengl/GLBase.html#setSwapInterval(int)
20130512 22:26:21 <void256> xranby_ac1001: thanks!
20130512 22:26:27 <void256> doesn't change anything
20130512 22:26:30 <void256> still 60 fps
20130512 22:26:31 <void256> >_<
20130512 22:26:55 <void256> nifty looks pretty smooth with it tho :)
20130512 22:27:20 <void256> do I have to combine setSwapInterval(int) with some other configuration on the Animator or so?
20130512 22:30:36 <xranby_ac1001> void256: are you using NEWT or some AWT component?
20130512 22:31:29 <xranby_ac1001> to create the window for nifty
20130512 22:31:36 <void256> AWT as far as I can tell
20130512 22:31:53 <void256> java.awt.Frame
20130512 22:35:14 <xranby_ac1001> void256: try cd jogamp-all-platforms and run
20130512 22:35:26 <xranby_ac1001> xranby@ac100-armhf:~/jogamp-all-platforms$ java -cp jar/jogl-test.jar:jar/jogl-all.jar:jar/gluegen-rt.jar:/usr/share/java/junit4.jar com.jogamp.opengl.test.junit.jogl.demos.es2.newt.TestGearsES2NewtCanvasAWT -time 40000 -vsync 0
20130512 22:35:55 <xranby_ac1001> if this work then look at that junit tests sourcecode
20130512 22:36:55 <xranby_ac1001> you may place a NEWT GLWindow inside AWT
20130512 22:37:55 <xranby_ac1001> the problem with using only awt is that rendering gets blocked by the awt event dispatch thread during input repaint etc
20130512 22:38:30 <xranby_ac1001> http://jogamp.org/jogl/doc/NEWT-Overview.html
20130512 22:42:02 <void256> well
20130512 22:42:04 <void256> thanks
20130512 22:42:09 <void256> it's not that I'm complaining
20130512 22:42:15 <void256> it looks pretty smooth :)
20130512 22:42:24 <xranby_ac1001> sounds good
20130512 22:42:31 <void256> I just wanted to compare the regular with the batched renderer in terms of FPS
20130512 22:42:45 <void256> but since this is acutally stupid
20130512 22:42:54 <void256> I'll simply go with the ms differences
20130512 22:42:58 <void256> :)
20130512 22:43:23 <void256> it looks smoother in jogl as it does with lwjgl (with more fps)
20130512 22:43:27 <void256> :>
20130512 22:43:59 <xranby_ac1001> i want to try this on opengl es 2
20130512 22:44:49 <void256> what do I have to change then? ^^
20130512 22:45:36 <void256> ah, wait ... I don't have a batched renderer with core profile support yet
20130512 22:45:41 <void256> for jogl
20130512 22:45:59 <void256> but since porting the regular one took something like an hour
20130512 22:46:12 <void256> adding core profile should be easy as well
20130512 22:46:13 <xranby_ac1001> void256: assume you use shaders, first request a GL2ES2 profile that is compatible with both OpenGL ES 2 and desktop GL 2
20130512 22:46:16 <void256> next weekend eventually
20130512 22:46:25 <void256> aye
20130512 22:47:35 <void256> it's on my list of things to do
20130512 22:47:49 <void256> but now sleep
20130512 22:47:52 <void256> thanks for the help :)
20130512 22:48:07 <xranby_ac1001> inside the GLEventListener use GL2ES2 gl = drawable.getGL().getGL2ES2();
20130512 22:48:18 <xranby_ac1001> and then fix all compile errors
20130512 22:48:23 <xranby_ac1001> if you have any
20130512 22:49:15 <xranby_ac1001> you may need to adjust the sharders slightly
20130512 22:49:45 <void256> nifty shaders are simple
20130512 22:49:59 <void256> and I think the jme3 stuff runs on android already
20130512 22:50:01 <void256> using ES2
20130512 22:50:05 <void256> so that should be fine
20130512 22:50:06 <xranby_ac1001> add a header at the start of the shader like
20130512 22:50:33 <xranby_ac1001> #ifdef GL_ES
20130512 22:50:34 <xranby_ac1001> precision mediump float;
20130512 22:50:34 <xranby_ac1001> precision mediump int;
20130512 22:50:34 <xranby_ac1001> #endif
20130512 22:50:42 <void256> aye
20130512 22:51:13 <xranby_ac1001> and thats all
20130512 22:52:41 <xranby_ac1001> ok good night
20130512 22:53:11 <void256> thanks again
20130512 22:53:20 <void256> I'm pretty sure this will work
20130512 22:53:29 <void256> nighty night
20130512 22:53:38 <xranby_ac1001> you are welcome, nice looking forward to test it :D
20130512 22:54:19 <void256> eventually you'll be the first to know ;)
20130512 22:54:39 * void256 (~void@anon) Quit (Quit: ChatZilla 0.9.90 [Firefox 20.0/20130326150557])
20130512 23:12:21 * xranby_ac1001 (~xranby@anon) has left #jogamp
20130513 00:16:54 * mjacob (~foobar@anon) has joined #jogamp
20130513 02:23:45 * mjacob_ (~foobar@anon) has joined #jogamp
20130513 02:27:04 * mjacob (~foobar@anon) Quit (Ping timeout: 255 seconds)
20130513 05:06:06 -CatOut- Continue @ http://jogamp.org/log/irc/jogamp_20130513050606.html