#jogamp @ irc.freenode.net - 20140305 05:05:51 (UTC)
20140305 05:05:51 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20140304050551.html
20140305 05:05:51 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20140305050551.html
20140305 05:21:09 * phao (~phao@anon) has joined #jogamp
20140305 07:30:25 * hija (~hija@anon) Quit (Ping timeout: 240 seconds)
20140305 07:35:27 * hija (~hija@anon) has joined #jogamp
20140305 08:11:07 * kermyt (~kermyt@anon) Quit (*.net *.split)
20140305 08:11:08 * xranby (~xranby@anon) Quit (*.net *.split)
20140305 08:18:38 * monsieur_max (~maxime@anon) has joined #jogamp
20140305 08:42:41 * hija (~hija@anon) Quit (Ping timeout: 244 seconds)
20140305 09:00:57 * xranby (~xranby@anon) has joined #jogamp
20140305 09:00:57 * kermyt (~kermyt@anon) has joined #jogamp
20140305 09:30:11 <zubzub> is there some kind of requirements/test suite for javax.media.nativewindow or newt?
20140305 09:30:28 <zubzub> is nativewindow btw an official jsr?
20140305 09:41:18 <zubzub> (I want to start investigating port nativewindow to wayland and additionally newt)
20140305 10:17:22 <xranby> zubzub: there is several tests for newt javax.media.nativewindow use inside jogl/src/test/com/jogamp/opengl/test/junit/newt/
20140305 10:18:44 <xranby> zubzub: jogamp do not plan to be part of the jsr/jep/... processes
20140305 10:19:09 <zubzub> but afaik jogl is a jsr (implementation?)
20140305 10:19:37 <xranby> the old jsr only cover the depricated fixed function opengl pipelines
20140305 10:19:52 <zubzub> ic
20140305 10:20:30 <xranby> there was some progress of sun to create a new jsr for programmable opengl pipelines but it newer became reality as i understood it.
20140305 10:23:27 <xranby> http://jogamp.org/jogl/doc/Overview-OpenGL-Evolution-And-JOGL.html this page give some introduction to how JOGL has evolved to map the current OpenGL with profiles
20140305 10:24:00 <xranby> you probably want to have a look at the NEWT overview as well: http://jogamp.org/jogl/doc/NEWT-Overview.html
20140305 10:25:05 <zubzub> ok will do!
20140305 10:41:24 <sgothel1> @zuzub: Simply implement all the *Driver impl. classes of NEWT for e.g. Wayland - pass some OpenGL-less tests. Then we need to see how well it plays together w/ our JOGL/EGL backend. You may orient your NEWT impl. to jogamp.newt.driver.bcm.vc.iv, since this uses EGL as well.
20140305 10:41:55 <sgothel1> the spec is written inside the javadoc API, if something is missing - we need to add it.
20140305 10:42:21 <sgothel1> *javadoc of interfaces* that is
20140305 10:43:49 <zubzub> k
20140305 10:49:29 <xranby> zubzub: take a look at http://jogamp.org/git/?p=jogl.git;a=commit;h=0ca481381b51b4082ac2b3bbae80cfaf5b60c3b8 - this commit added the possibility to use NEWT on raspberry pi by passing -Dnativewindow.ws.name=jogamp.newt.driver.bcm.vc.iv
20140305 10:51:10 <xranby> zubzub: autodetection that removed the need to pass the -Dnativewindow.ws.name=jogamp.newt.driver.bcm.vc.iv got then added in a later commit: http://jogamp.org/git/?p=jogl.git;a=commit;h=8a1e50dc33e14e9423e2f4bc066b32a835bb33fb
20140305 10:53:41 <xranby> i assume wayland got some way to handle input that we need to map into the newt input events
20140305 10:54:39 <zubzub> http://wayland.freedesktop.org/docs/html/sect-Protocol-Input.html
20140305 10:55:02 <zubzub> I'd have to investigate how it really works
20140305 10:55:11 <zubzub> I know how it works in X but not really in wayland
20140305 11:24:44 <xranby> i belive jolla uses the maliit input methods
20140305 11:24:50 <xranby> https://wiki.maliit.org/images/c/c1/Fosdem2013-wayland-input-methods.pdf
20140305 11:29:45 <xranby> maliit provides the jolla on screen keyboard thing
20140305 12:04:59 <zubzub> the slides are not really clear
20140305 12:05:03 <zubzub> https://wiki.maliit.org/Wayland_Input_Method_System_Proposal
20140305 12:11:49 <zubzub> http://www.freedesktop.org/wiki/Software/libinput/
20140305 12:12:54 <zubzub> I guess if you put libinput into a seperate process, you almost have what the maliit proposes (?)
20140305 12:54:58 <phao> Hey, is it just a bug in my program (hehe) or the defaults for glClearDepth and glDepthFunc vary accross versions of opengl? Or maybe jogl sets them differently?
20140305 12:56:54 <phao> For example, this code ( http://pastie.org/8870720 ) without the lines 189 and 190 works strangely. THis is a simple cube with a simple modelview matrix.
20140305 12:57:18 <phao> The idea is to specify a rotation/scale modelview matrix and a cube [0..1]^3
20140305 12:58:06 <phao> the front face (z=1, right?) should obscure the back face (z=0) since the that modelview matrix still keeps the camera pointing at z negative.
20140305 12:58:49 <phao> unless I specify that depth test configuration, the defaults make the back face obscure the front face.
20140305 12:59:15 <phao> Orientation and all seems all correct. I'm sorry the program is kind of messy.
20140305 13:40:32 * bbbruce (~bx@anon) Quit (Ping timeout: 252 seconds)
20140305 14:08:49 * bbbruce (~bx@anon) has joined #jogamp
20140305 14:35:46 * phao (~phao@anon) Quit (Quit: Fui embora)
20140305 14:46:39 <zubzub> hmm
20140305 14:46:57 <zubzub> there seems to be a dependency in javax.media.nativewindow.util.InsetsImmutable
20140305 14:47:06 <zubzub> to com.jogamp.common.type.WriteCloneable
20140305 14:47:30 <zubzub> am not sure if that is supposed to be the case...
20140305 14:48:04 <zubzub> the jogl-all-noawt doesnt have it at least
20140305 14:48:09 <zubzub> (jar)
20140305 14:48:33 <zubzub> com.jogamp.common is not in the jar
20140305 14:50:59 <zubzub> but imho javax.media.nativewindow should not have deps to anythong jogamp/jogl
20140305 14:56:36 <xranby> jogamp.common is found in gluegen
20140305 14:56:45 <xranby> jogl depend on gluegen
20140305 14:56:59 <xranby> javax.media.nativewindow is implemented in jogl
20140305 14:57:10 <xranby> zubzub: include gluegen-rt.jar
20140305 15:01:42 <zubzub> why is it an implicit dep?
20140305 15:13:54 <xranby> jogl, joal, jocl .. jovg... all re-use functionality in gluegen for native library symbol lookup and utility classes
20140305 15:14:22 <xranby> the jogamp documentation is quite clear that gluegen-rt is allways needed: http://jogamp.org/jogl/doc/deployment/JOGL-DEPLOYMENT.html
20140305 15:14:43 <zubzub> then why is it not a maven dep (if it is always needed)?
20140305 15:16:44 <xranby> i pass that question to rmk0
20140305 15:16:54 <xranby> the current maven docs are: http://jogamp.org/wiki/index.php/Maven
20140305 15:17:46 <zubzub> the docs btw idd say that gluegen is needed for the jogl jar
20140305 15:18:00 <zubzub> but what if I were to use the nativewindow jar?
20140305 15:18:25 <zubzub> it does not depend on jogl as far as I can see
20140305 15:19:12 <zubzub> and the only deps on gluegen are some small util interface
20140305 15:19:17 <zubzub> (not that I mind that much)
20140305 15:33:19 <sgothel1> dependency: gluegen << nativewindow [ << jogl ] << newt
20140305 15:33:48 <sgothel1> newt doesn't need jogl, sure not, only that GLWindow does
20140305 15:35:26 <sgothel1> off-topic: puppet mastery now in Europe .. proxy wars, etc .. the usual - http://www.zerohedge.com/news/2014-03-05/behind-kiev-snipers-it-was-somebody-new-coaltion-stunning-new-leak-reveals-truth
20140305 15:38:41 <sgothel1> http://en.wikipedia.org/wiki/First_they_came...
20140305 15:40:41 <sgothel1> .. the older you get, the more boring all of this sort of becomes - like, 'the same procedure as last year .. AGAIN?' .. oh dear.
20140305 16:58:12 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20140305 17:58:20 * monsieur_max (~maxime@anon) has joined #jogamp
20140305 19:07:58 * rallan (~rallan@anon) has joined #jogamp
20140305 19:40:26 * void256 (~void@anon) has joined #jogamp
20140305 20:01:48 * void256_ (~void@anon) has joined #jogamp
20140305 20:05:00 * void256 (~void@anon) Quit (Ping timeout: 264 seconds)
20140305 20:05:14 * void256_ is now known as void256
20140305 20:06:54 * rallan (~rallan@anon) Quit (Remote host closed the connection)
20140305 20:51:49 <void256> The field GL2.GL_STACK_OVERFLOW is ambiguous JoglGL.java /nifty-jogl-renderer/src/main/java/de/lessvoid/nifty/renderer/jogl/render line 208 Java Problem
20140305 20:51:59 <void256> with: <jogamp.version>2.1.4</jogamp.version>
20140305 20:52:08 <void256> doesn't happen with: <jogamp.version>2.0.2-rc12</jogamp.version>
20140305 20:52:13 <sgothel1> oh
20140305 20:52:16 <void256> how do I fix that?
20140305 20:52:18 <void256> :)
20140305 20:52:35 <void256> someone asked me to up the version to 2.1.4
20140305 20:52:53 <sgothel1> ../build-x86_64/jogl/gensrc/classes/javax/media/opengl/GL2ES1.java: public static final int GL_STACK_OVERFLOW = 0x0503;
20140305 20:52:53 <sgothel1> ../build-x86_64/jogl/gensrc/classes/javax/media/opengl/GL2ES2.java: public static final int GL_STACK_OVERFLOW = 0x0503;
20140305 20:53:35 <sgothel1> quick fix: say 'GL2ES2.GL_STACK_OVERFLOW'
20140305 20:53:55 <void256> excellent!
20140305 20:53:58 <void256> that fixed it :)
20140305 20:54:03 <void256> danke sven :)
20140305 20:54:12 <void256> how are you?
20140305 20:55:10 <sgothel1> ^^ introduced w/ GL_KHR_debug .. hmm, have to check whether we shall drop this in GL for next version ..
20140305 20:55:45 <sgothel1> thank you Jens .. hacking along w/ graph, text rendering/AA .. and better not look to hard on world matters :)
20140305 20:55:57 <sgothel1> spring is here .. hurray 'eh ? :)
20140305 20:56:07 <sgothel1> and you ? all fine I hope?
20140305 21:02:27 <void256> nice! it's still a bit grey here :| looking forward to spring too! I've been quite lazy this year :(
20140305 21:08:03 * monsieur_max sends sun and heat all over the place
20140305 21:35:52 * yradostev (~yradostev@anon) has joined #jogamp
20140305 21:42:51 * yradostev (~yradostev@anon) Quit (Ping timeout: 244 seconds)
20140305 22:03:40 * void256 (~void@anon) Quit (Quit: ChatZilla 0.9.90.1 [Firefox 27.0.1/20140212131424])
20140305 22:28:47 * monsieur_max (~maxime@anon) has left #jogamp
20140306 00:21:14 * phao (~phao@anon) has joined #jogamp
20140306 00:22:49 <phao> Hi. Earlier I was having some issues. This code ( http://pastie.org/8870720 vertex shader = http://pastie.org/8871912 ), without the lines 189 and 190, it works strangely. The program simply draws a cube, but makes the vertex (1,1,1) white while the others are black so to stand it out. Without lines 189 and 190, this vertex gets hidden by some other vertex ([1,1,0] afaik). That's puzzling me because since by default I'm looking at -z, the face fo
20140306 00:22:49 <phao> r z=0 shouldn't hide the face for z=1. RIght?
20140306 01:04:36 <phao> it turns out
20140306 01:04:44 <phao> identity projection matrix is ortho 3d for near=1 and far=-1
20140306 01:04:52 <phao> I was assuming it was for near=-1 and far=1
20140306 01:05:03 <phao> that needs row=3,col=3 of projection matrix to be -1 and rest be like an identity.
20140306 01:05:19 <phao> changing the vertex shader with an ad-hoc projection matrix did the trick
20140306 01:49:08 * phao (~phao@anon) Quit (Read error: Connection reset by peer)
20140306 01:49:33 * phao (~phao@anon) has joined #jogamp
20140306 05:05:51 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20140306050551.html