#jogamp @ irc.freenode.net - 20160426 05:06:19 (UTC)
20160426 05:06:19 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20160425050619.html
20160426 05:06:19 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20160426050619.html
20160426 06:25:15 * bigpet (uid25664@anon) has joined #jogamp
20160426 06:29:53 * elect (~GBarbieri@anon) has joined #jogamp
20160426 06:30:20 <elect> hi
20160426 07:34:02 * Eclesia (~husky@anon) has joined #jogamp
20160426 07:34:11 <Eclesia> good morning
20160426 07:49:16 <elect> hey#
20160426 08:07:03 * jvanek (jvanek@anon) has joined #jogamp
20160426 08:54:27 <zubzub> you know what's really hard?
20160426 08:54:39 <monsieur_max> mmmh not sure i want to know
20160426 08:54:47 <zubzub> have people collaborate on a single project
20160426 08:55:32 <zubzub> (I'm trying to get a robovm fork going)
20160426 08:56:01 <zubzub> but it seems people are mostly interested in continueing working on their own one man pet projects
20160426 08:56:53 <zubzub> 'all I want to do is be productive so send me pull request if you want'
20160426 08:57:07 <zubzub> problem is that if they all say that, that won't work out
20160426 08:57:22 <zubzub> and they do all seem to say that
20160426 08:58:36 <monsieur_max> shame, but understandable
20160426 09:44:42 <elect> did you try to say exactly this to them?
20160426 09:46:39 <zubzub> elect: I just did
20160426 09:47:15 <elect> link
20160426 09:47:51 <zubzub> https://github.com/MobiDevelop/robovm/issues/1
20160426 09:47:57 <zubzub> last comment
20160426 09:48:32 <zubzub> https://github.com/bugvm/bugvm/issues/47#issuecomment-214657345
20160426 09:48:34 <zubzub> again last comment
20160426 09:49:02 <zubzub> I think I was rather diplomatic but firm
20160426 09:49:12 <zubzub> I don't expec them to change their decision behavior though
20160426 09:49:24 <zubzub> shame :(
20160426 09:51:50 <elect> it's funny that that mobidevelop mention politics
20160426 09:51:56 <elect> what has politics to do?
20160426 09:51:59 <elect> I cant get it
20160426 09:52:08 <zubzub> neither do I...
20160426 09:52:44 <zubzub> I guess he thinks finding a good starting base/project == politics
20160426 09:52:49 <elect> :D
20160426 09:53:17 <zubzub> while neither of the people involved in that thread have their own project or are in any way involved in the mentioned projects
20160426 09:53:32 <elect> I guess some relative long battle and diplomacy are waiting for you, zubzub ^^
20160426 09:53:43 <zubzub> I'm afraid so
20160426 09:54:03 <zubzub> lots of wasted energy
20160426 09:54:16 <elect> at least priand and dexdev seems to agree
20160426 09:54:21 <zubzub> idd
20160426 09:54:22 <elect> that's a nice starting point
20160426 09:55:02 <zubzub> I guess I'll try to setup something tomorrow
20160426 09:55:14 <zubzub> to get started
20160426 09:55:31 <elect> in the worst case scenario, is that you 3 focus your energies together and sooner or later also the two einstein ibinti and politics-guy jump on board
20160426 09:55:41 <zubzub> yeah idd
20160426 09:56:04 <elect> anyway, I look forward to it
20160426 09:56:06 <zubzub> they won't jump on board but will probably c/p fixes & code into their own project
20160426 09:56:17 <zubzub> and hopefully submit bug reports
20160426 09:56:24 <elect> if your get famous/grows, they might to
20160426 09:56:35 <elect> anyway, to sum robovm pro
20160426 09:56:45 <elect> java on ios, compiled java, and?
20160426 09:56:51 <elect> obscured sources?
20160426 09:57:04 <elect> anything left?
20160426 09:57:34 <zubzub> s/java/bytecode
20160426 09:57:51 <zubzub> also optimized bytecode by soot
20160426 09:58:21 <zubzub> language->bytecode->soot (optimize + ast) -> llvm IR -> (optimize) llvm native
20160426 09:58:42 <elect> yep, but you don't have any profiling numbers so far, right?
20160426 09:58:54 <zubzub> only for iOS iirc
20160426 09:58:55 <elect> only nice concept on papares
20160426 09:59:07 <zubzub> on iOS iirc it was on par with objc
20160426 09:59:08 <elect> can you share some?
20160426 09:59:37 <zubzub> https://medium.com/@harrycheung/mobile-app-performance-redux-e512be94f976#.3sbdm0cv3
20160426 10:00:26 <elect> is there anything that may represent standard java perfs?
20160426 10:01:07 <elect> because I see also Java Android, but on Nexus 9 and Moto X
20160426 10:01:14 <elect> that's different platform
20160426 10:01:18 <zubzub> http://dacapobench.org/
20160426 10:01:30 <zubzub> but I don't think robovm will be able to run that
20160426 10:01:41 <zubzub> as it only supports android runtime
20160426 10:01:48 <elect> because matching objC perfs is nice
20160426 10:01:51 <zubzub> idd
20160426 10:02:00 <elect> but actually I am already able to do that now, with plain java
20160426 10:02:11 <elect> simple scenario ofcourse
20160426 10:02:15 <elect> but still
20160426 10:02:23 <elect> I'll share them asap
20160426 10:02:33 <elect> especially in the opengl channel :trollface:
20160426 10:03:45 <elect> how much do you code in java, btw, zubzub ?
20160426 10:04:00 <monsieur_max> 5
20160426 10:04:15 <elect> cm?
20160426 10:04:24 <elect> cmon, it could have been worst ^^
20160426 10:04:24 <monsieur_max> cold water
20160426 10:05:09 <zubzub> what do you mean with how much?
20160426 10:05:18 <zubzub> it's my full time job (embbeded osgi on arm)
20160426 10:05:18 <elect> I mean, are you a 100% java dev
20160426 10:05:29 <zubzub> but I can write C pretty well too
20160426 10:05:29 <elect> or do you use also other languages?
20160426 10:05:47 <elect> ok
20160426 10:06:30 <zubzub> I also did python full time for half a year
20160426 10:06:53 <elect> apart from your job, what is your general vision on java?
20160426 10:07:02 <elect> I mean, do you like as language?
20160426 10:07:21 <zubzub> oracle is killing it, java as a language is really nice but could be improved upon
20160426 10:07:38 <zubzub> I'm thinking of looking at kotlin
20160426 10:12:38 <zubzub> modular jdk9 looks nice
20160426 10:12:44 <elect> link
20160426 10:12:48 <zubzub> jigsaw
20160426 10:12:59 <elect> anyway, still no opengl kotlin binding?
20160426 10:13:08 <zubzub> http://openjdk.java.net/projects/jigsaw/spec/sotms/
20160426 10:13:10 <elect> but given is 100% java interoperable..
20160426 10:13:15 <elect> there is no need, or?
20160426 10:13:26 <zubzub> afaik you should be able to use jogl
20160426 11:10:01 <Eclesia> nobody interested in a vp8 (webm/webp) decoder ?
20160426 11:10:27 <monsieur_max> ( don't answer yes, he'll told you to code it )
20160426 11:11:43 <Eclesia> alredy made 40% of the work, but I'm a bit stuck, so another pov would help
20160426 11:18:24 <elect> sorry Eclesia
20160426 11:22:02 <Eclesia> :'(
20160426 12:14:55 * BAZtm (~BAZtm@anon) has joined #jogamp
20160426 12:15:20 <BAZtm> Has anyone ever had any issues with OSX El Capitan and TileRenderer?
20160426 12:43:11 <elect> what kind of issue do you have? (I am not on mac btw)
20160426 13:14:05 <BAZtm> elect: I have a TileRenderer which is used to create a High resolution screenshot
20160426 13:14:37 <BAZtm> Works fine on Ubuntu + Window, however on El-capitan it seems to render the first tile repeatedly
20160426 13:15:07 <BAZtm> I'm pretty sure I tested this a while back on OSX and it worked fine
20160426 13:15:28 <BAZtm> but that might have been before el-capitan
20160426 13:24:46 <elect> don't you have the chance to test on anything before than el-capitan, do you?
20160426 14:08:56 <BAZtm> elect: Not in the short term
20160426 14:09:37 <BAZtm> I've found that Ubuntu and Windows read GL_BACK however OSX seems to read GL_COLOR_ATTRIBUTES0
20160426 14:10:25 <BAZtm> GL_COLOR_ATTACHMENT0**
20160426 14:11:04 <elect> where?
20160426 14:12:43 <BAZtm> Within the TileRenderer.endTile() function during "glReadPixels"
20160426 14:16:12 <BAZtm> https://github.com/sgothel/jogl/blob/master/src/jogl/classes/com/jogamp/opengl/util/TileRenderer.java#L464
20160426 14:17:09 <BAZtm> I'm confused on why OSX is refererring to an FBO but on Windows/Ubuntu just refers to the back buffer
20160426 14:23:08 <rmk0> BAZtm: i don't know the details, but this sort of thing is almost always down to it being easier to use an FBO to work around platform-specific breakages
20160426 14:23:32 <rmk0> i seem to remember sven having a good six months of near constant OSX-related issues and doing a lot of work to paper over the cracks
20160426 14:25:53 <BAZtm> Ahh yeah I've just found a forum post where someone is discussing the exact same issue and Svens responded
20160426 14:25:56 <BAZtm> Going to give it a read
20160426 14:26:30 * jvanek (jvanek@anon) Quit (Quit: Leaving)
20160426 14:34:04 * jvanek (jvanek@anon) has joined #jogamp
20160426 14:38:26 <BAZtm> rmk0: Yeah it looks like OSX uses FBO's to patch over some issues
20160426 14:39:11 <BAZtm> However I'm rendering to the back buffer, and would like to read the back buffer, but TileRenderer with FBO's only reads from the Front buffer
20160426 14:39:48 <BAZtm> Which explains why I'm probably getting the repeating tile as it's just repeatedly rendering the last thing in that position
20160426 14:41:27 <BAZtm> Thing is I don't really care about swapping buffers as I dont want to display what I'm doing (which is just composing a large image)
20160426 14:45:11 * jvanek (jvanek@anon) Quit (Quit: Leaving)
20160426 14:46:25 <elect> link?
20160426 14:46:49 <BAZtm> http://forum.jogamp.org/glReadBuffer-behavior-on-MAC-td4028039.html
20160426 14:47:10 <elect> thanks
20160426 14:47:37 <elect> anyway, both OSX and Win/Linux refer to an fbo
20160426 14:48:05 <elect> because the back buffer is the color attachment of the default fbo
20160426 14:48:24 <BAZtm> aha yeah
20160426 15:36:12 <monsieur_max> mmmh great, i have a sound issue on macosX, the fix is in joal 2.4... any idea when it'll be released ?
20160426 15:36:49 * monsieur_max is afraid that the answer'll be "never"
20160426 15:59:16 <bruce-> when the moon is in the seventh house
20160426 15:59:26 <bruce-> and jupiter aligns with mars
20160426 16:01:33 <monsieur_max> ok, it checks "never"
20160426 16:02:20 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20160426 16:13:45 * Eclesia (~husky@anon) has left #jogamp
20160426 17:10:12 * elect (~GBarbieri@anon) Quit (Ping timeout: 276 seconds)
20160426 17:22:02 * Eclesia (~eclesia@anon) has joined #jogamp
20160426 17:34:16 * monsieur_max (~maxime@anon) has joined #jogamp
20160426 17:41:01 <Eclesia> time to get motivated, VP8 decoding evening
20160426 17:48:00 * elect (~elect@anon) has joined #jogamp
20160426 17:55:23 * dwbrite (4474ac1d@anon) has joined #jogamp
20160426 17:55:47 <dwbrite> Hey, just wanted to drop in and say "I'm still alive"
20160426 17:56:23 <dwbrite> and Eclesia: http://dev.dwbrite.com/post/143271862389/this-blog-post-is-mostly-just-a-test-for-code :)
20160426 17:56:48 <dwbrite> I'll be back after May 7th, when I take the SATs.
20160426 17:57:10 <dwbrite> Ciao!
20160426 17:57:22 * dwbrite (4474ac1d@anon) Quit (Client Quit)
20160426 17:57:30 <Eclesia> \o/
20160426 18:00:15 <Eclesia> I won't ask you anything else, you can stay ^^
20160426 18:00:15 * BAZtm (~BAZtm@anon) Quit (Ping timeout: 276 seconds)
20160426 18:11:24 * elect (~elect@anon) Quit (Ping timeout: 250 seconds)
20160426 19:25:23 <Eclesia> I love google developers. the worse specifications ever : "for every x value where x ≡ 0 (mod 2), a green value at x is positioned into the 4 least-significant bits of the green value at x / 2, a green value at x + 1 is positioned into the 4 most-significant bits of the green value at x / 2."
20160426 19:49:06 * elect (~elect@anon) has joined #jogamp
20160426 19:58:17 * Eclesia (~eclesia@anon) Quit (Quit: Leaving.)
20160426 19:58:44 * elect (~elect@anon) Quit (Ping timeout: 276 seconds)
20160426 20:47:03 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20160426 22:12:44 * bigpet (uid25664@anon) Quit (Quit: Connection closed for inactivity)
20160427 05:06:19 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20160427050619.html