#jogamp @ irc.freenode.net - 20130921 05:05:28 (UTC)
20130921 05:05:28 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20130919025727.html
20130921 05:05:28 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20130921050528.html
20130921 07:07:27 * monsieur_max (~maxime@anon) has joined #jogamp
20130921 09:04:23 * xranby-64 (~familjen@anon) has joined #jogamp
20130921 12:04:10 * xranby (~xranby@anon) Quit (Ping timeout: 245 seconds)
20130921 12:04:55 * xranby (~xranby@anon) has joined #jogamp
20130921 12:13:06 * Eclesia (~eclesia@anon) has joined #jogamp
20130921 12:13:21 <Eclesia> hi
20130921 12:14:04 * xranby (~xranby@anon) Quit (Ping timeout: 268 seconds)
20130921 12:17:32 <Eclesia> I am searching for the equivalent of java.awt.Toolkit.getDefaultToolkit().getScreenSize();
20130921 12:17:43 <Eclesia> using Newt only
20130921 12:18:13 * xranby (~xranby@anon) has joined #jogamp
20130921 12:18:13 <Eclesia> I have seens the Display and Screen classes, it looks like what I'm searching for.
20130921 12:18:19 <Eclesia> But can't find the entry point
20130921 12:18:38 <Eclesia> hello xranby
20130921 12:19:54 * xranby-64 (~familjen@anon) Quit (Quit: Leaving.)
20130921 12:25:17 <Eclesia> found it : Screen.getAllScreens();
20130921 12:30:31 <Eclesia> hm nope...
20130921 12:30:36 <Eclesia> empty list
20130921 12:43:53 * Eclesia (~eclesia@anon) Quit (Quit: Leaving.)
20130921 16:12:06 * xranby (~xranby@anon) Quit (Ping timeout: 268 seconds)
20130921 16:13:41 * xranby (~xranby@anon) has joined #jogamp
20130921 18:56:12 * void256 (~chatzilla@anon) has joined #jogamp
20130921 21:21:01 * xranby (~xranby@anon) Quit (Read error: Operation timed out)
20130921 21:37:05 * xranby (~xranby@anon) has joined #jogamp
20130921 21:53:46 * monsieur_max (~maxime@anon) has left #jogamp
20130921 22:01:00 * void256 (~chatzilla@anon) Quit (Remote host closed the connection)
20130921 22:01:03 * void256_ (~void@anon) has joined #jogamp
20130921 22:01:07 * void256_ is now known as void256
20130922 00:27:08 * void256 (~void@anon) Quit (Remote host closed the connection)
20130922 00:27:35 * void256 (~void@anon) has joined #jogamp
20130922 00:48:18 <void256> http://i.imgur.com/RbFVwqM.png
20130922 00:48:27 <void256> round caps and round line joins :>
20130922 00:48:59 <sgothel> looks smooth
20130922 00:49:02 <sgothel> nice
20130922 00:49:31 <sgothel> since you use OSX .. would you expect that we use HiDPI by default ?
20130922 00:49:43 <sgothel> i.e. best resolution possible .. or ..
20130922 00:52:06 <void256> I still use regular resolution so I can't really comment on that :/
20130922 00:52:48 <void256> and regular means 2560x1600 @ 30" :D
20130922 00:53:05 <void256> but yeah, no retina thing
20130922 00:53:26 <sgothel> thats pretty close though .. regarding fill-rate
20130922 00:53:35 <sgothel> same here ..
20130922 00:53:58 <sgothel> however, was thinking how can we possibly inject hat 'flag' .. and is using a flag, what is the default ?
20130922 00:54:06 <sgothel> IMHO native resolution should be it ..
20130922 00:54:26 <void256> inject into where?
20130922 00:54:35 <sgothel> good question :)
20130922 00:54:37 <void256> xD
20130922 00:54:43 <sgothel> NEWT .. caps ..
20130922 00:54:49 <void256> ah
20130922 00:54:52 <void256> ok
20130922 00:55:15 <void256> but why would you care?
20130922 00:55:27 <void256> I mean the user can request the resolution
20130922 00:55:31 <sgothel> well .. for OSX users ( not me personally )
20130922 00:55:31 <void256> isn't that enough?
20130922 00:55:36 <sgothel> can they ?
20130922 00:55:48 <void256> I don't know - I hope so? =)
20130922 00:56:06 <sgothel> i.e. you do a window .. if you don't set HiDPI .. you get some lower res AFAIK
20130922 00:56:29 <sgothel> it's quite retarded .. but well ..
20130922 00:56:51 <sgothel> so this would affect window size and all the little math for it .. etc
20130922 00:56:57 <void256> really? odd! well, as I said, haven't really looked into that because of lack of hardware ;)
20130922 00:57:27 <sgothel> yeah .. I got some request in the pipe, hence a bit brainstorming .. what others think
20130922 00:59:59 <void256> glView = [[NSOpenGLView alloc] initWithFrame:newFrame pixelFormat:[NSOpenGLView defaultPixelFormat]];
20130922 01:00:01 <void256> [glView setWantsBestResolutionOpenGLSurface:YES];
20130922 01:00:47 <sgothel> what is best ? :)
20130922 01:00:55 <sgothel> thx .. have to look into it ..
20130922 01:00:57 <void256> https://developer.apple.com/library/mac/documentation/AppKit/Reference/NSViewOpenGLAdditions/Reference/Reference.html
20130922 01:01:17 <void256> "Apps that want to take advantage of the higher pixel density on high-resolution displays need to opt-in to get a high-resolution backing surface. "
20130922 01:01:37 <void256> that's the flag you'll probably use
20130922 01:02:02 <sgothel> ah .. yup, like the HiDPI flag they have somewhere ..
20130922 01:02:22 <void256> to enable it and that would be the flag one would want to set somewhere and yes, I'd expect that to be set by default when this is supported by the hardware
20130922 01:02:47 <sgothel> my thinking as well .. best resolution ..
20130922 01:02:52 <void256> aye
20130922 01:03:18 <sgothel> so the NativeWindow Capabilities could have an option like 'fluffyRes' :)
20130922 01:03:45 <void256> ^^
20130922 01:04:22 <sgothel> so w/ HiDPI .. we have to transform the NativeWindow size / Pointer coords .. etc to match
20130922 01:04:41 <sgothel> assuming Viewport uses the hiDPI values .. ofc
20130922 01:04:55 <void256> yep
20130922 01:05:04 <sgothel> great .. thank you Jens!
20130922 01:05:09 <void256> o_O
20130922 01:05:14 <void256> I've not done anything ^^
20130922 01:05:26 <void256> I was just a rubber duck I guess ;)
20130922 01:05:39 <sgothel> of course .. chatting w/ me .. telling me that me not insane - yes :)
20130922 01:05:51 <void256> sven?
20130922 01:05:54 <sgothel> now the CALayer positioning bug ..
20130922 01:05:56 <void256> you're not insane
20130922 01:05:57 <void256> ^^
20130922 01:06:07 <void256> (as far as I can tell ... hehe)
20130922 01:06:19 <sgothel> sometimes chatting w/ Xerxes and getting all the new ideas for project .. me not so sure :)
20130922 01:06:57 <void256> quantum diamond shape shader plugin extension for jogamp? ;>
20130922 01:07:07 <sgothel> yeah .. something like that :)
20130922 01:20:39 <void256> hmm
20130922 01:21:12 <void256> my line shader has some overdraw issues when used with a color that has an alpha value < 1 :/
20130922 01:21:36 <void256> the solution for that case would be to use the shader to create an alpha mask
20130922 01:21:46 <void256> I think
20130922 01:22:10 <sgothel> some culling / alpha-test ?
20130922 01:22:42 <void256> trying to figure something out the whole evening without much success yet :/
20130922 01:23:41 <sgothel> mgl_alphatest.fp ?
20130922 01:24:03 <sgothel> in jogamp.opengl.util.glsl.fixedfunc.shaders
20130922 01:24:05 <void256> http://i.imgur.com/d8RxqSs.png
20130922 01:24:26 <void256> ok, letme take a look ...
20130922 01:24:35 <sgothel> which one is not desired ?
20130922 01:24:41 <sgothel> the darker gray ?
20130922 01:24:48 <void256> yep
20130922 01:25:17 <sgothel> how to read the current pixel color ? AFAIK avail in ES3 ?
20130922 01:25:31 <sgothel> (so no FBO would be required)
20130922 01:26:33 <sgothel> interesting
20130922 01:27:19 <sgothel> tessellating before .. creating 2 objects to render, i.e. cut off before overdrawing ?
20130922 01:27:32 <void256> not really :>
20130922 01:28:18 <sgothel> I am sure graph does something like that .. at some level, triangulation .. new ctrl points
20130922 01:28:38 <void256> yes but no :)
20130922 01:28:51 <sgothel> so if you would know the distance to the overdraw .. hmm
20130922 01:29:23 <sgothel> would make things complicated .. yup
20130922 01:29:29 <void256> that really should be solved at the pixel/fragment level I would think
20130922 01:29:34 <sgothel> GLSL read current pixel ?
20130922 01:29:44 <sgothel> starting from which GLSL version ?
20130922 01:30:13 <void256> probably 4.x
20130922 01:30:21 <void256> like all the neat stuff ;)
20130922 01:30:22 <sgothel> autch
20130922 01:30:37 <void256> haven't seen it in core 3.2 what I'm using right now
20130922 01:32:00 <void256> actually I haven't seen a "read current pixel" at all to be honest o_O
20130922 01:32:00 <sgothel> yeah .. so creating a mask in a texture (a 1st rendering pass) seems to be most compatible :/
20130922 01:32:06 <void256> aye
20130922 01:32:13 <void256> at least for alpha != 1
20130922 01:32:27 <sgothel> or leave it in for art :)
20130922 01:32:33 <void256> and with the benefit to have gradients in lines and texture fills and so on ^^
20130922 01:32:51 <void256> yeah ;)
20130922 01:33:07 <void256> it gets even worse with other line joins tho
20130922 01:33:15 <void256> like the bevel line join
20130922 01:33:21 <void256> it gets really ugly ;)
20130922 01:34:05 <void256> but yes I'll leave it in
20130922 01:34:21 <void256> it should be okay for alpha = 1 and small lines anyway
20130922 01:34:43 <sgothel> yup, sounds ok .. i.e. can be optimized later on - great job!
20130922 01:35:05 <void256> aye
20130922 01:35:18 <sgothel> ha!
20130922 01:35:49 <void256> if you don't mind the evil word in the url ;)
20130922 01:35:50 <void256> https://github.com/void256/lwjgl-coregl/tree/master/lwjgl-coregl-utils-examples/src/main/resources/line
20130922 01:35:59 <void256> here are the shaders in the current version
20130922 01:36:25 <sgothel> your fp could write the alpha value also to an FBO, which can be read (concurrently - ugly - undefined ?) as a texture ..
20130922 01:36:36 <sgothel> the multi-target thing ..
20130922 01:36:42 <void256> yes that might work
20130922 01:36:59 <sgothel> but keep in mind that concurrent write/read is undef .. hmm
20130922 01:37:05 <void256> aye
20130922 01:37:14 <sgothel> as long you render them sequential .. it should work
20130922 01:37:39 <sgothel> why lwjgl ?
20130922 01:37:51 <void256> only because of historical purpose
20130922 01:37:55 <void256> ;>
20130922 01:38:11 <sgothel> hmm .. guess I have to make the suffix in ShaderCode flexible ..
20130922 01:38:27 <void256> suffix?
20130922 01:38:38 <sgothel> .fs .fp ...
20130922 01:38:53 <void256> ah
20130922 01:38:54 <sgothel> our tool ShaderCode uses .fp
20130922 01:39:26 <void256> http://www.opengl.org/wiki/Fragment_Shader
20130922 01:39:35 <void256> and so on
20130922 01:39:45 <sgothel> to my knowledge it was a fragment program first :)
20130922 01:39:51 <void256> I think it was called program with the ARB extension
20130922 01:39:53 <void256> aye
20130922 01:40:30 <void256> after reading the opengl super bible I'm more in a "I don't care there is only core profile now for me" state of mind ;)
20130922 01:40:32 <sgothel> bookmarked your URI
20130922 01:41:00 <sgothel> pretty much so ..
20130922 01:41:03 <void256> nice
20130922 01:41:42 <sgothel> so back to dirty tedious business .. hope I will have time soon to work w/ these nice things again
20130922 01:42:09 <void256> yes, have fun ... I have to sleep anyway ><
20130922 01:42:16 <void256> and I have some bug reports for Nifty as well
20130922 01:42:23 <void256> but I kinda ignore them ;)
20130922 01:42:30 <void256> lines are more fun at the moment ^^
20130922 01:42:38 <void256> cu later :)
20130922 01:43:27 <sgothel> MRT - multiple render target (for reference .. above)
20130922 01:43:31 <sgothel> yup .. laters
20130922 01:47:27 * void256 (~void@anon) Quit (Quit: ChatZilla 0.9.90.1 [Firefox 23.0.1/20130814063812])
20130922 03:53:51 * juank_prada (~juank_pra@anon) has joined #jogamp
20130922 04:13:46 * juank_prada (~juank_pra@anon) Quit (Quit: Leaving)
20130922 05:05:28 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20130922050528.html