JogAmp IRC Chat Logs

#jogamp @ irc.freenode.net - 20150718 05:05:24 (UTC)

20150718 05:05:24 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20150716145123.html
20150718 05:05:24 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20150718050524.html
20150718 05:40:26 * elect (~elect@anon) has joined #jogamp
20150718 05:49:57 * monsieur_max (~maxime@anon) has joined #jogamp
20150718 06:24:24 * elect (~elect@anon) Quit (Ping timeout: 264 seconds)
20150718 11:49:53 <xranby_> sgothel: i will take a look and try understand the root cause of the javadoc font issues... since we the jogamp css is based on a "standard" java 7 css things should work
20150718 11:53:05 <xranby_> aha! https://validator.w3.org/check?uri=http%3A%2F%2Fjogamp.org%2Fdeployment%2Fjogamp-current%2Fjavadoc%2Fjogl%2Fjavadoc%2Fcom%2Fjogamp%2Fopengl%2Futil%2Fglsl%2FShaderCode.html%23defaultShaderCustomization%2528com.jogamp.opengl.GL2ES2%2C%2520boolean%2C%2520boolean%2529%3F&charset=%28detect+automatically%29&doctype=Inline&group=0
20150718 11:53:24 <xranby_> we have errors like this in the code <code>count</count>
20150718 11:53:40 <xranby_> the end tag should be </code> instead of </count>
20150718 11:54:11 <sgothel> well . that is indeed in our source code .. (those tags) :)
20150718 11:54:33 <sgothel> usually injected via apidoc {@code count} .. etc
20150718 11:54:46 <sgothel> lets see ..
20150718 11:55:12 <xranby_> yes its better that we fix the code than to hack around it in the css
20150718 11:55:56 <sgothel> <code>count</count> :)
20150718 11:56:06 <sgothel> multiple places .. in the file
20150718 11:56:17 <sgothel> sure the css still needs a bit love ..
20150718 11:59:49 <xranby_> maybe update to a new version of java doc and use the new javadoc css + tweak for jogamp.org colors?
20150718 12:00:12 <xranby_> imho, the current css is ok as long as the fonts are readable
20150718 12:00:16 <sgothel> that is the current used css .. (tweaking some colors)
20150718 12:00:37 <sgothel> but those fonts sometimes become tiny - here it was the javadoc source code ..
20150718 12:00:38 <xranby_> yes i lookaed at the history and you changed nothing that would cause font issues
20150718 12:00:41 <sgothel> otherwise .. I dunno
20150718 12:00:54 <sgothel> then - we will also produce doxygen soon ..
20150718 12:00:58 <xranby_> i think they become tiny due to the html validation errors
20150718 12:01:11 <xranby_> and become bid due to the same reason
20150718 12:01:13 <sgothel> then - all good, so we need to find those bugs anyway
20150718 12:01:14 <xranby_> big
20150718 12:01:33 <sgothel> yeah .. tiresome stuff to validate our javadoc .. broken links .. etc
20150718 12:02:23 <xranby_> we currently have documentation at three places.. the html pages, the javadoc and the wiki
20150718 12:02:34 <xranby_> maybe we should move the content from the html pages into the javadoc ?
20150718 12:03:25 <sgothel> hu ?
20150718 12:03:31 <sgothel> which html pages ?
20150718 12:03:48 <xranby_> https://jogamp.org/gluegen/doc/manual/
20150718 12:03:57 <xranby_> pages such as that one
20150718 12:04:03 <sgothel> well, there is a difference between API doc and manual, etc .. hmm
20150718 12:04:24 <sgothel> if there would be a doc system, which would validate those links in between ..
20150718 12:04:31 <sgothel> and an editor ..
20150718 12:04:38 <sgothel> I would agree
20150718 12:05:17 <sgothel> hmm .. they still could be alive in their own repo though ..
20150718 12:05:58 <sgothel> but a document editor (like eclipse java code) .. which then helps w/ referencing code etc ..
20150718 12:06:22 <sgothel> we could start w/: offline css validation and dead-link checks
20150718 12:06:30 <sgothel> css/html validation - I mean
20150718 12:06:40 <sgothel> and slowly fix the docs ..
20150718 12:07:03 <sgothel> so we need an EDITOR (human being in this role)
20150718 12:08:04 <sgothel> either free or we need to have a cost/benefit eval .. and pay for it - since I doubt I will do this properly (time/energy)
20150718 12:08:05 <xranby_> https://marketplace.eclipse.org/content/jdoceditor
20150718 12:08:14 <xranby_> i wonder if this editor will do?
20150718 12:08:14 <sgothel> ah
20150718 12:08:55 <sgothel> commit 6e68e6a9d2472d556c7cab69b2ee9e55aeb60f36 :)
20150718 12:10:28 <xranby_> nice!
20150718 12:11:39 <xranby_> so if someone finds a javadoc page with odd looking fonts.. the routine will be .. run it through the validator and fix hte code
20150718 12:12:07 <xranby_> having offline automatic validation surely helps as well, but I do not know of any suitable tool
20150718 12:12:11 <sgothel> yup - same for dead links (i.e. those {@link Class#lala(..)} things
20150718 12:12:42 <sgothel> javadoc itself does it .. (turned off by now) - just would need the energy to walk this HUGE list!
20150718 12:13:01 <sgothel> volunteers ? :)
20150718 12:13:19 <sgothel> this will probably increase the EDITORs commit stats by a huge margin :)
20150718 12:13:51 <xranby_> :)
20150718 12:14:00 <xranby_> jogamp hall of fame highscore list
20150718 12:14:03 <sgothel> maybe I can throw in some hardware or other candy for motivation :)
20150718 12:15:32 <xranby_> we do accept pull request for typo's and spelling errors!
20150718 12:15:58 <sgothel> indeed!
20150718 12:18:38 <xranby_> i sat and experimented last night trying to use the internal newt api's to initialize x11 windows on my raspberry pi system, when the system is by default uses the bcm.vc.iv driver, trying to fix 1178
20150718 12:19:10 <xranby_> https://jogamp.org/bugzilla/show_bug.cgi?id=1178
20150718 12:20:07 <sgothel> hmm .. your use case is more like: NEWT/BCM overlay over AWT .. isn't it?
20150718 12:20:13 <sgothel> or it could be seen that way ..
20150718 12:20:21 * doev (~doev@anon) has joined #jogamp
20150718 12:20:27 <xranby_> i don tkn ow if i want to use awt :)
20150718 12:20:40 <xranby_> but sure, bcm overlay over awt would work as well
20150718 12:21:06 <xranby_> the alternative is to do the full X11/EGL-blit https://jogamp.org/bugzilla/show_bug.cgi?id=626#c3
20150718 12:21:16 <sgothel> so if xorg is available - problem is: creating a dependency to X11 is a bit cracy .. isn't it?
20150718 12:21:36 <sgothel> awt: one can use GLJPanel .. quite the same thing
20150718 12:22:14 <sgothel> so an optional X11MouseTracker .. w/ X11 dependencies ?
20150718 12:22:42 <xranby_> my motivation to work on 1178 is that its such a common usecase for raspberry pi users to try run jogl applications from inside x11
20150718 12:22:44 <sgothel> here one needs to capture the X11 events from the root window
20150718 12:22:53 <sgothel> yup .. understood
20150718 12:23:29 <sgothel> then using the X11MouseTracker shall also disable our BCM mouse pointer
20150718 12:23:57 <xranby_> we may want to use our BCM mousepointer when the mousepointer is inside our newt window overlay
20150718 12:24:02 <sgothel> then you go ahead and write such a tracker .. done, maybe have a common interface w/ LinuxMouseTracker
20150718 12:24:03 <xranby_> else you cant see the mousepointer
20150718 12:24:22 <sgothel> and add a method: 'boolean needsPointerRendering()'
20150718 12:25:03 <sgothel> -> interface PointerTracker
20150718 12:25:45 <sgothel> NEWTs X11 code may help here alot to impl. the X11 event dispatcher ..
20150718 12:26:10 <sgothel> then .. it shall probe whether X11 is even available .. etc etc
20150718 12:26:53 <xranby_> indeed, its easy to think about it,, but when writing the code you have to think of all the details
20150718 12:27:11 <xranby_> and then its not an easy problem at all
20150718 12:27:45 <sgothel> ofc .. but hey, its not that complicated - but a bit tedious. i.e. its only a small part of NEWT X11 DisplayDriver
20150718 12:28:28 <sgothel> sure .. one would need to learn how the whole event dispatching currently works etc
20150718 12:29:12 <sgothel> Java_jogamp_newt_driver_x11_DisplayDriver_DispatchMessages0 <- reuse that maybe? hmm
20150718 12:29:47 <xranby_> *bookmarking for evening reading and hacking*
20150718 12:29:51 <xranby_> the sun is up right now
20150718 12:30:00 <xranby_> i will go out and enjoy the sun
20150718 12:30:18 <sgothel> hehe .. yesterday we finally got the blinders installed .. what a relieve
20150718 12:30:45 <xranby_> thank you for healty pointers, directions and hands on help to resolve all small details!
20150718 12:31:38 <sgothel> thank you!
20150718 12:33:02 <xranby_> btw, with the javadoc fix you did jogl/build/javadoc/jogl/javadoc/com/jogamp/opengl/util/glsl/ShaderCode.html looks stellar on my machine
20150718 12:33:22 <xranby_> correct font sizes everywhere
20150718 12:40:59 * doev (~doev@anon) Quit (Quit: Verlassend)
20150718 12:41:06 * rmk0 (~rmk0@anon) Quit (Remote host closed the connection)
20150718 13:12:33 * rmk0 (~rmk0@anon) has joined #jogamp
20150718 13:13:52 * rmk0 (~rmk0@anon) Quit (Client Quit)
20150718 13:14:16 * rmk0 (~rmk0@anon) has joined #jogamp
20150718 13:14:16 * rmk0 (~rmk0@anon) Quit (Changing host)
20150718 13:14:16 * rmk0 (~rmk0@anon) has joined #jogamp
20150718 14:13:37 * juank_prada (~juank_pra@anon) has joined #jogamp
20150718 17:14:16 * juank_prada (~juank_pra@anon) Quit (Ping timeout: 246 seconds)
20150718 17:32:47 * juank_prada (~juank_pra@anon) has joined #jogamp
20150718 18:58:31 * elect (~elect@anon) has joined #jogamp
20150718 19:10:21 <xranby_> https://github.com/xranby/gluegen/commit/708a163e333e9e7786c1828d028c8ca10b584f47
20150718 19:10:56 <xranby_> fixes miniature font size of the nested bullet lists at: http://jogamp.org/deployment/jogamp-current/javadoc/jogl/javadoc/
20150718 19:47:34 * juank_prada (~juank_pra@anon) Quit (Ping timeout: 246 seconds)
20150718 19:56:06 <elect> sgothel, you there?
20150718 20:21:48 * xranby_ grabs an Xlib tutorial https://tronche.com/gui/x/xlib-tutorial/
20150718 20:23:42 <elect> xranby_, I get ont ShaderCore Caused by: java.lang.IllegalArgumentException: path is empty <>
20150718 20:28:02 <xranby_> elect:which ShaderCode create constructor did you use?
20150718 20:28:46 <xranby_> most of the constructors assume you store your shaders as resources
20150718 20:30:06 <xranby_> relative to the class
20150718 20:38:45 <elect> it works only if shaders are in the same path as the class
20150718 20:39:20 <elect> if I move the shader one directory above
20150718 20:39:25 <elect> and pass ../
20150718 20:39:29 <elect> it doesnt work
20150718 20:42:27 <xranby_> i do not thin java supports .. when resolving resources
20150718 20:42:41 <xranby_> that only work for unic filesystems
20150718 20:42:43 <xranby_> unix
20150718 20:43:44 <elect> im on ubuntu
20150718 20:44:15 <xranby_> but java on ubuntu cant use .. to when resolving resources
20150718 20:44:52 <xranby_> because it is not part of the java specification
20150718 20:46:54 <elect> ok
20150718 20:47:04 <elect> I remember I saw it in some constructor in JOGL
20150718 20:51:36 <elect> https://github.com/sgothel/jogl/blob/master/src/jogl/classes/jogamp/opengl/macosx/cgl/MacOSXCGLContext.java#L164-L167
20150718 20:52:08 <xranby_> interesting
20150718 20:54:06 <xranby_> so if i understand correctly, using such syntax do not work for you using ubuntu, as you have verified
20150718 20:55:21 <xranby_> elect: do the src/test/com/jogamp/opengl/test/junit/jogl/demos/es2/av/MovieSimple.java work for you?
20150718 20:56:18 <xranby_> src/test/com/jogamp/opengl/test/junit/jogl/demos/es2/av/MovieSBSStereo.java src/jogl/classes/jogamp/opengl/util/glsl/GLSLTextureRaster.java and src/jogl/classes/jogamp/opengl/macosx/cgl/MacOSXCGLContext.java uses ../ syntax
20150718 20:56:31 <elect> I never succeeded to build jogl
20150718 20:57:35 <zubzub> when refering to classpath resources you're actually constructing the path part of a resource url
20150718 20:58:04 <zubzub> the formt of the classpath url is undefined and implementation specific, but it should comply with the official url spec
20150718 20:58:52 <zubzub> https://en.wikipedia.org/wiki/Uniform_resource_locator
20150718 20:59:25 <zubzub> afaik .. is not allowed as a path fragment from a url
20150718 21:01:10 <xranby_> elect: try cd jogamp-all-platforms/jar and run java -cp jogl-test.jar:jogl-all.jar:gluegen-rt.jar com.jogamp.opengl.test.junit.jogl.demos.es2.av.MovieSimple
20150718 21:03:30 <xranby_> if that work.. then ../ syntax did work
20150718 21:10:23 <xranby_> will it work if you use .. instead of ../ ?
20150718 21:40:30 <sgothel> good evening ..
20150718 21:40:42 <sgothel> yes, '../' should work w/ ShaderCode
20150718 21:40:59 <sgothel> @elect .. cont. w/ the suffix stuff :)
20150718 21:41:32 <sgothel> ShaderCode uses our IOUtils to resolve resources using ClassLoader and the like .. usually works nice
20150718 21:41:57 <sgothel> @xerxes: tronche's X11 lib guide is gold - I use it as well
20150718 21:42:10 <zubzub> except it's X11
20150718 21:42:15 <zubzub> Xlib
20150718 21:42:17 <zubzub> and not xcb :)
20150718 21:42:53 <sgothel> further, GlueGen has its own Uri impl., which solves the URI limitations (UTF .. etc)
20150718 21:43:13 <sgothel> (born our of frustration when we moved to URI first :)
20150718 21:43:28 <sgothel> *out*
20150718 21:43:57 <sgothel> well .. wayland will be next though
20150718 21:44:31 <zubzub> :)
20150718 21:44:54 <zubzub> supporting wayland should be really simple
20150718 21:45:09 <zubzub> as it's mostly setting up an egl context
20150718 21:45:20 <zubzub> maybe input handling might be a bit more tricky
20150718 21:45:46 <sgothel> well .. the details .. the details - yes: JOGL's NativeWindow part and NEWT
20150718 21:46:00 <sgothel> but sure .. it will be a step by step thingy
20150718 21:52:46 <elect> xranby_, porn movie is running
20150718 21:53:18 <elect> zubzub, thanks for killing all my hopes
20150718 21:53:30 <elect> sgothel, I never get you when you speak
20150718 21:53:56 <elect> you pay the same also if you write down all the letters ^^
20150718 21:54:05 <sgothel> hu?
20150718 21:54:28 <elect> just a joke
20150718 21:54:46 <elect> you meant "continuing with suffix stuff"?
20150718 21:54:47 <sgothel> whats your issue trying to compile jogl?
20150718 21:55:00 <sgothel> yes .. editing it now
20150718 21:55:03 <elect> I had a lot
20150718 21:55:11 <elect> that I gave up
20150718 21:55:16 <sgothel> oh .. you read the howto ofc ..
20150718 21:55:21 <elect> ofc
20150718 21:55:28 <elect> ok, not on the first play
20150718 21:55:31 <elect> but on the second
20150718 21:55:32 <sgothel> best if using a vanilla ant installation - not from distri ..
20150718 21:55:50 * juank_prada (~juank_pra@anon) has joined #jogamp
20150718 21:55:52 <elect> + it was compiling for Itanium x64
20150718 21:55:56 <elect> dont ask why plz
20150718 21:55:57 <sgothel> I had issues w/ the latter as well - but that was .. err .. month/years ago
20150718 21:56:04 <sgothel> hmm
20150718 21:56:09 <elect> anyway
20150718 21:56:11 <elect> never mind
20150718 21:56:19 <elect> we can continue our lives anyway
20150718 21:56:26 <elect> I will keep this sin inside
20150718 21:56:50 <sgothel> omg .. deep philosophy tonight? - usually its just me talking stuff :)
20150718 21:57:05 <elect> xranby_, anyway I am watching the video, but no sound, normal?
20150718 21:57:05 <sgothel> or you try your new 'elisa' bot ? :)
20150718 21:57:16 <elect> not yet
20150718 21:57:27 <elect> btw, if the movie is running
20150718 21:57:33 <sgothel> so xerxes wants you to watch a porn movie? :)
20150718 21:57:33 <elect> ../ should work
20150718 21:57:38 <elect> YES
20150718 21:57:40 <elect> HE MEAN
20150718 21:57:50 <elect> HIS FAULT, DONT LOOK AT ME
20150718 21:57:52 <elect> :D
20150718 21:57:55 <sgothel> good red red wine?
20150718 21:57:57 <sgothel> hehe
20150718 21:58:04 <elect> I dont drink
20150718 21:58:11 <elect> wait, alcohol, water is needed
20150718 21:58:27 <elect> btw, if the video is running, ../ should work, shouldnt?
20150718 21:58:29 <sgothel> naturally high - I like that as well
20150718 21:58:37 <xranby_> elect: sounds is supposed to work and the move is supposed to be all about furries
20150718 21:59:29 <elect> I love furries
20150718 22:02:17 <xranby_> you can try MovieCube as well.. and tell if AudioSink is stuck java -cp jogl-test.jar:jogl-all.jar:gluegen-rt.jar com.jogamp.opengl.test.junit.jogl.demos.es2.av.MovieCube
20150718 22:02:44 <elect> sgothel, https://github.com/elect86/jogl-samples/blob/master/jogl-samples/src/tests/gl_300_fbo_multisample.java#L100-L103
20150718 22:03:03 <xranby_> If the AudioSink do not consume frames then it may be a bug decoding the audio frames
20150718 22:03:37 <xranby_> i do not hear a thing on my machine as well :/
20150718 22:03:46 <xranby_> AudioSink[frames [p 0, q 0, f 1, c 0], time 0, bytes 0]
20150718 22:03:55 <elect> lol
20150718 22:04:00 <elect> neither btw
20150718 22:04:18 <elect> anyway dont worry
20150718 22:04:24 <xranby_> we need to look into it
20150718 22:04:39 <elect> video is playing, so ../ seems ok
20150718 22:04:41 <xranby_> you can get inside the cube using the scroll wheel
20150718 22:05:53 <elect> anyway, another thing I really really really really really really really miss looking at C OpenGL is the DDS tool that decompressing offers also internalFormat/externalFormat and pixel type
20150718 22:06:15 <elect> on jogl this features is only offered by the Texture class
20150718 22:08:17 <elect> GLI seems to be more complete in this sense
20150718 22:11:23 <xranby_> ah solved the audio issue... need to add joal.jar to the classpath
20150718 22:11:38 <elect> ah
20150718 22:11:43 <elect> ok
20150718 22:13:15 <xranby_> java -cp joal.jar:jogl-test.jar:jogl-all.jar:gluegen-rt.jar com.jogamp.opengl.test.junit.jogl.demos.es2.av.MovieCube
20150718 22:14:47 <elect> sgothel, I get another error with shaders in the same path of the class
20150718 22:14:47 <elect> Program link failed: 1
20150718 22:14:48 <elect> Vertex info
20150718 22:14:48 <elect> -----------
20150718 22:14:48 <elect> 0(10) : error C1013: function "main" is already defined at 0(10)
20150718 22:15:08 <elect> defaultShaderCustomization(gl3, false, false);
20150718 22:17:08 <elect> xranby_, yep, now it is running
20150718 22:19:23 <sgothel> -Djogl.debug.GLSLCode
20150718 22:19:37 <sgothel> ^^ will enable debug .. dumps failed shader etc
20150718 22:19:56 <sgothel> looks like your are somehow concat 2 shader programs w/ main
20150718 22:22:53 <elect> wait wait
20150718 22:22:54 <elect> my fault
20150718 22:29:16 <elect> I was passing same vertex shader type also for fragment
20150718 22:29:30 <elect> I fixed but I still get path is empty
20150718 22:29:59 <elect> I moved the shaders in the class
20150718 22:30:04 <elect> and passing null
20150718 22:30:37 <elect> no joking
20150718 22:30:42 <elect> problem is on the texture
20150718 22:30:44 <elect> ok
20150718 22:31:19 <elect> night
20150718 22:31:21 <elect> ^^
20150718 22:31:32 <sgothel> good night
20150718 22:32:48 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150718 22:35:43 * elect (~elect@anon) Quit (Ping timeout: 252 seconds)
20150719 00:12:18 * jk4 (~jk4@anon) Quit (Quit: WeeChat 0.4.3)
20150719 01:21:05 * zubzub (~zubzub@anon) Quit (Ping timeout: 264 seconds)
20150719 01:21:13 * zubzub (~zubzub@anon) has joined #jogamp
20150719 03:44:33 * juank_prada (~juank_pra@anon) Quit (Read error: Connection reset by peer)
20150719 05:05:24 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20150719050524.html