#jogamp @ irc.freenode.net - 20150924 05:05:12 (UTC)
20150924 05:05:12 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20150923050512.html
20150924 05:05:12 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20150924050512.html
20150924 05:25:53 * monsieur_max (~maxime@anon) has joined #jogamp
20150924 06:02:16 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150924 06:12:23 * jvanek (jvanek@anon) has joined #jogamp
20150924 07:16:58 * monsieur_max (~maxime@anon) has joined #jogamp
20150924 07:44:50 <monsieur_max> xranby : had a bit of time to investigate my video playback issues, you'll be glad to learn that MovieSimple is doing the job perfectly, so again, it's my fault. But i still have to understand why.
20150924 08:02:45 <xranby_f22> monsieur_max: thank you for that, so the playboack work fine now when you use the last frame in the buffer instead of the first one in the buffer?
20150924 08:03:53 <xranby_f22> (jogamp have a buffer of decoded video frames usually 3) depending on which call you make to fetch the frame you get the buffered frame (as you do now and should work great)
20150924 08:04:30 <monsieur_max> well yes, but I made other changes, so i can't really say that it was the faulty part of the code. However, removing the vsync made things smoother. We'll stick to 60Fps
20150924 08:04:36 <xranby_f22> i recall you mentioned that before you used the wrong api call and got the last decoded frame hence you did not use the buffer at all
20150924 08:05:08 <xranby_f22> ok.. good to hear, so there may be something to investigate why it is not smooth with vsync
20150924 08:05:35 <monsieur_max> well, it's still smooth, but not as smooth :)
20150924 08:06:15 <xranby_f22> monsieur_max: hmm,, ok well then we need to earmark it and look into why it is not as smooth as it should be
20150924 08:06:17 <monsieur_max> my remaining issue is the sound jumps, I still have to understand what are the differences from the MovieSimple test and my code
20150924 08:06:46 <xranby_f22> the eye can detect subtile speed changes hence... we must make better sure that the time between two frames stays constant
20150924 08:08:43 <xranby_f22> and is the same time as the consecutive frames
20150924 08:09:19 <xranby_f22> if you say you have sound jumps then hmm... need to look into that
20150924 08:09:32 <zubzub> xranby_f22: do you know if avian support embedded minimal stripped build using openjdk?
20150924 08:09:42 <zubzub> https://github.com/ReadyTalk/avian/issues/455#issuecomment-142800201
20150924 08:10:28 <xranby_f22> zubzub: ask dicej!
20150924 08:10:55 <xranby_f22> zubzub: i have never tested it
20150924 08:12:21 <xranby_f22> kudos, i look forward to the answer
20150924 08:13:27 <xranby_f22> zubzub: at least i know you will get an avian binary
20150924 08:13:32 <xranby_f22> if you use openjdk
20150924 08:14:03 <xranby_f22> when you build against openjdk you get one avian binary and one libjvm.so hotspot replacement lib
20150924 08:14:15 <xranby_f22> thus you have two ways to use it
20150924 08:14:18 <xranby_f22> that I know of
20150924 08:14:33 * xranby_f22 (~familjen@anon) Quit (Quit: Leaving.)
20150924 08:14:36 <zubzub> ok
20150924 08:14:40 <zubzub> I asked in the ticket
20150924 08:15:40 <monsieur_max> xranby: No sound jumps with MovieSimple, so it's my fault :)
20150924 09:15:49 * xranby (~xranby@anon) Quit (Remote host closed the connection)
20150924 09:17:04 * eclesia (~husky@anon) has joined #jogamp
20150924 09:17:10 <eclesia> good morning
20150924 09:29:40 * xranby (~xranby@anon) has joined #jogamp
20150924 09:30:51 <monsieur_max> hello eclesia
20150924 09:30:55 <monsieur_max> xranby: No sound jumps with MovieSimple, so it's my fault :)
20150924 09:33:35 <xranby> monsieur_max: thank you for reporting... did you test MovieSimple and MovieCube with your video?
20150924 09:34:26 <xranby> you video is quite unique being a 60fps video
20150924 09:35:12 <xranby> hence disable vsync may be a must..
20150924 09:35:42 <monsieur_max> xranby: yes, that's exactly what i told you, i tested it with MovieSimple, just to be sure if it was my fault or not
20150924 09:36:00 <monsieur_max> and MovieSimple is doing perfectly
20150924 09:36:17 <monsieur_max> the issues only happens in my game
20150924 09:42:25 <monsieur_max> xranby: i tested with a webm video, 60FPS, MAx quality, and MovieSimple handled it perfectly :)
20150924 10:02:54 <xranby> monsieur_max: do MovieCube handle it superb as well?
20150924 10:03:13 <xranby> i am cross checking if our junit MovieSimple and MovieCube code is identical
20150924 10:04:00 <monsieur_max> xranby: i could make the test tonight with movie cube
20150924 11:30:56 <xranby> one of the researchers at the univerity of Lund/Sweden have created a Java to C translator with realtime garbage collector: http://www.robot.lth.se/java/
20150924 11:31:10 <xranby> LJRT - Lund Java based Real-Time
20150924 11:32:18 <xranby> LGPL 2.1
20150924 11:35:06 <xranby> http://www.robot.lth.se/java/ReferenceManual.pdf
20150924 11:35:34 <xranby> The set of target systems considered include small (350 MHz PPC G3 with32 MB ram) to very small (AVRμcontroller at 8MHz/32 kB RAM) embeddedcomputers.
20150924 11:41:53 <xranby> Lund Java-based Real-Time (LJRT) library http://cs.lth.se/eda040/ljrt/ The LJRT Java2C translation system is now hosted at https://launchpad.net/ljrt
20150924 11:42:14 <xranby> http://cs.lth.se/eda040/ljrt/
20150924 11:43:08 <xranby> LJRT includes a Java-to-C translator (and libraries) that enables the use of Java source code for real-time execution on machines without a JVM. This project is hosted at http://git.cs.lth.se/ljrt/ljrtcompiler
20150924 11:43:31 <xranby> hence the latest source is available at -> http://git.cs.lth.se/ljrt/ljrtcompiler
20150924 11:46:24 <zubzub> java 2 C ?!
20150924 11:46:31 <zubzub> what's the catch?
20150924 11:46:55 <xranby> guillaum1: ^
20150924 11:47:31 <xranby> zubzub: the catch usually is that it only work for the usecases that the researchers use
20150924 11:47:57 <xranby> i think you have to limit youself to java 1.5
20150924 11:48:05 <zubzub> System.out.println("Hello world!");?
20150924 11:48:10 <zubzub> gah
20150924 11:48:13 <xranby> they use it for robotics
20150924 11:48:16 <zubzub> always the same
20150924 11:48:24 <zubzub> hey we have this super cool thing!
20150924 11:48:43 <zubzub> but it's not usuable for 99% of interested people
20150924 11:49:04 <xranby> well.. if you write java without the invoke dynamic bytecode it should work
20150924 11:49:24 <xranby> i mean most of the java 1.6 -> 1.9 is only syntactic shugar
20150924 11:49:37 <zubzub> 1.5 -> 1.9 :p[
20150924 11:49:37 <xranby> the classfiles are basically the same
20150924 11:49:55 <zubzub> does it compile bytecode or source file?
20150924 11:51:39 <xranby> i am reading the reference manual... looks like this operate on sourcecode hmm
20150924 11:51:52 <xranby> unlike the gcc ECJ that can operate on class files
20150924 11:51:59 <xranby> robovm operate on class files as well
20150924 11:52:21 <xranby> gcc GCJ can operate on class files
20150924 11:57:05 <xranby> zubzub: thus robovm is likely a better java to c translator for you
20150924 11:57:15 <xranby> or simply use gcj
20150924 11:57:22 <zubzub> java 1.8
20150924 11:57:24 <zubzub> that's the catch
20150924 11:57:27 <zubzub> (for me)
20150924 11:57:43 <xranby> yes compile using java 8 -> class then compile this class using robovm or gcj
20150924 11:57:51 <zubzub> gcj does 1.8?
20150924 11:58:15 <xranby> javac does 1.8 but can output classes that are 1.6 compatible
20150924 11:58:33 <xranby> and gcj can use the classes produced by javac
20150924 12:01:50 <zubzub> ic
20150924 12:02:12 <zubzub> what about 1.8 classes like Optional?
20150924 12:02:29 <zubzub> I assume gcj does not know those
20150924 12:02:33 <zubzub> or can it us ea 1.8 jdk?
20150924 13:40:23 * xranby (~xranby@anon) Quit (Ping timeout: 264 seconds)
20150924 13:53:10 * xranby (~xranby@anon) has joined #jogamp
20150924 14:44:54 * jvanek (jvanek@anon) Quit (Quit: Leaving)
20150924 15:23:06 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150924 16:03:35 * eclesia (~husky@anon) has left #jogamp
20150924 17:32:07 * Eclesia (~eclesia@anon) has joined #jogamp
20150924 17:55:42 * monsieur_max (~maxime@anon) has joined #jogamp
20150924 20:31:46 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150924 22:12:03 * Eclesia (~eclesia@anon) Quit (Ping timeout: 250 seconds)
20150925 05:05:12 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20150925050512.html