#jogamp @ irc.freenode.net - 20140119 05:05:39 (UTC)
20140119 05:05:39 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20140118050539.html
20140119 05:05:39 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20140119050539.html
20140119 07:22:31 * D3zmodos1 (~D3zmodos@anon) has joined #jogamp
20140119 07:25:53 * D3zmodos (~D3zmodos@anon) Quit (Ping timeout: 264 seconds)
20140119 09:07:24 * monsieur_max (~maxime@anon) has joined #jogamp
20140119 10:02:27 * hija (~hija@anon) has joined #jogamp
20140119 10:14:52 <monsieur_max> hi there, got a little issue, not quite sure it's on my side but the jogl-test.jar for 2.1.3 seems corrupted ( http://jogamp.org/deployment/archive/rc/v2.1.3/jar/jogl-test.jar ), however jogl-test.jar.gz is fine.
20140119 11:32:12 * monsieur_max (~maxime@anon) has left #jogamp
20140119 11:33:24 * monsieur_max (~maxime@anon) has joined #jogamp
20140119 11:34:40 * monsieur_max (~maxime@anon) Quit (Client Quit)
20140119 11:35:38 * monsieur_max (~maxime@anon) has joined #jogamp
20140119 11:52:23 * monsieur_max (~maxime@anon) has left #jogamp
20140119 11:59:46 * monsieur_max (~maxime@anon) has joined #jogamp
20140119 14:51:47 <sgothel> @Max .. the usual 'don't download anything via firefox issue ..' .. but use 'wget' .. b/c of our pack.gz mime-type mapping .. etc
20140119 16:21:17 <monsieur_max> sgothel: thanks for your answer
20140119 17:31:37 * hija (~hija@anon) Quit (Quit: hija)
20140119 18:34:20 * kermyt (~kermyt@anon) Quit (Read error: Operation timed out)
20140119 18:38:26 * kermyt (~kermyt@anon) has joined #jogamp
20140119 18:51:50 * hija (~hija@anon) has joined #jogamp
20140119 18:53:03 * Eclesia (~eclesia@anon) has joined #jogamp
20140119 18:53:42 <Eclesia> hi
20140119 18:56:27 <Eclesia> I have something strange when using float type textures
20140119 18:56:57 <rmk0> lo
20140119 18:57:10 <Eclesia> I has set a out layout of type vec3
20140119 18:57:26 <Eclesia> I set the value to f_out1 = vec3(1,2,3)
20140119 18:57:52 <Eclesia> and when I read the values using : gl.glGetTexImage(GL4.GL_TEXTURE_2D, 0, glFormat, glSampleType, buffer);
20140119 18:57:59 <Eclesia> the buffer contains : 2,3,4
20140119 18:58:36 <Eclesia> as if a +1 was made on all components
20140119 18:59:10 <rmk0> is blending enabled?
20140119 18:59:21 <Eclesia> yes
20140119 18:59:39 <rmk0> what values were in the framebuffer before you wrote to it?
20140119 19:00:01 <Eclesia> clean new buffer. so normaly nothing
20140119 19:00:22 <rmk0> a buffer that's been cleared with glClear?
20140119 19:00:32 <rmk0> typically the contents of a freshly allocated texture are undefined
20140119 19:00:44 <rmk0> framebuffer, whatever
20140119 19:02:11 <Eclesia> hm so matbe there was something at this location on memory that's what you say ?
20140119 19:02:17 <Eclesia> maybe*
20140119 19:02:50 <rmk0> is the only thing i can think of, off-hand... if there was (1,1,1) in the framebuffer and you wrote (2,3,4) with additive blending, you'd obviously get (2,3,4)
20140119 19:03:07 <rmk0> would try clearing the framebuffer with (0,0,0,1)
20140119 19:03:25 <rmk0> er, that came out wrong
20140119 19:03:34 <rmk0> (1,1,1) + (1,2,3) = (2,3,4), obviously
20140119 19:04:47 <rmk0> the contents of freshly allocated framebuffers quite often have interesting contents
20140119 19:05:05 <Eclesia> found it
20140119 19:05:09 <rmk0> it'll pull the memory from anywhere, so it can contain the now-deallocated memory of other processes
20140119 19:05:15 <Eclesia> my clear color was set on 1,1,1,1
20140119 19:05:26 <rmk0> \o/
20140119 19:05:37 <Eclesia> rmk0 +1
20140119 19:05:42 <Eclesia> good memory ^^
20140119 19:05:49 <rmk0> hehe, was a good guess
20140119 19:05:57 <rmk0> helps that i've just been doing additive blending work
20140119 19:07:08 <Eclesia> is there a way to disable blending on some color attachement ? I want somekind 'always override' mode
20140119 19:07:33 <Eclesia> I'm writing position,normals,uv, stuffs like that for deferred rendering
20140119 19:07:41 <rmk0> yes, know the feeling
20140119 19:07:47 <rmk0> feel like there is a way to do that on newer GLs...
20140119 19:10:01 <rmk0> hm, no
20140119 19:10:06 <rmk0> seems like blend state is still global
20140119 19:10:22 <rmk0> there's a way to set specific blending modes per draw buffer, but there's no way to specify that blending is on or off for a given framebuffer
20140119 19:10:34 <rmk0> it's either on or off globally
20140119 19:21:04 <Eclesia> rmk0 : you mean glBlendFunc ?
20140119 19:23:30 <rmk0> yeah, there's a variant that allows it for specific draw buffers
20140119 19:23:59 <rmk0> glBlendFuncSeparatei and friends
20140119 19:24:28 <Eclesia> found : https://www.opengl.org/sdk/docs/man4/xhtml/glBlendFunc.xml
20140119 19:24:40 <Eclesia> glBlendFunci
20140119 19:26:54 <rmk0> yeah, is just that the blending enabled/disabled state is global
20140119 19:27:03 <rmk0> even for glEnablei(GL_BLEND, n);
20140119 19:27:39 <rmk0> you could maybe do:
20140119 19:27:48 <rmk0> glDisablei(GL_BLEND, 1);
20140119 19:28:06 <rmk0> and make sure that you always write the data that you don't want blended to draw buffer 1
20140119 19:28:18 <rmk0> is unpleasant, and error-prone
20140119 19:28:35 <Eclesia> I can also make a SRC porter duff rule
20140119 19:28:41 <Eclesia> to always replace the value
20140119 19:29:07 <rmk0> hm, yes
20140119 19:37:40 <Eclesia> rmk0: seems like the docs say we can use : glBlendFunci(attIndex,...) and glEnable(GL_Blend,attIndex)
20140119 19:37:56 <Eclesia> so we can control blending for each attachment
20140119 19:39:18 <rmk0> you can, yes, but glEnable(GL_Blend,attIndex) will enable blending on attIndex for *all* framebuffers
20140119 19:39:30 <rmk0> it's not per-framebuffer state in the state tables
20140119 19:40:46 <rmk0> see table 23.21 in the 4.3 spec
20140119 19:40:55 <Eclesia> ho I see.
20140119 19:41:36 <Eclesia> but there is only 1 FBO in writing mode at the same time in a GL context. so we can change the blend modes when changing fbo
20140119 19:41:44 <rmk0> yep
20140119 19:42:18 <rmk0> in my renderer, i'm drawing all opaque objects and then all translucent objects, so it's actually only set twice per frame
20140119 19:42:31 <rmk0> tend to think of it as global state
20140119 19:42:57 <Eclesia> like all other glEnable/Diable properties
20140119 19:43:03 <rmk0> right
20140119 19:44:48 <Eclesia> ha finaly have correct values in my output image :)
20140119 19:45:06 <Eclesia> rmk0: thanks for the help
20140119 19:46:06 <rmk0> no problem
20140119 20:56:49 * hharrison (~chatzilla@anon) has joined #jogamp
20140119 20:57:41 <hharrison> has anyone been following Valve's efforts to create an OpenGL debugger?
20140119 21:00:40 <rmk0> someone just mentioned it to me an hour ago
20140119 21:01:16 <rmk0> am hoping they do things the way apitrace does... that log of calls that can be replayed in the debugger without needing to run the original program
20140119 21:01:24 <rmk0> amazingly useful
20140119 21:03:08 <rmk0> is the valve thing actually... usable, yet?
20140119 21:12:58 <hharrison> Not from what I've seen
20140119 21:13:12 <hharrison> Not really released yet
20140119 21:13:19 <hharrison> But shouldn't be too long
20140119 21:14:25 <rmk0> apitrace doesn't reall work for anything other than 2.1 here
20140119 21:14:28 <rmk0> *really
20140119 21:14:36 <rmk0> the valve thing claims to support more than that already
20140119 21:14:42 <rmk0> some issue with opening a context, i think
20140119 21:14:43 <hharrison> vogl is essentially 3.3 and above
20140119 21:14:57 <hharrison> 4.x coming Real Soon Now
20140119 21:15:15 <rmk0> i'll have a 4.x context one of these years!
20140119 21:15:34 <hharrison> Oh, I totally read this slide wrong
20140119 21:15:44 <hharrison> Currently supports 1 -> 3.3
20140119 21:15:48 <rmk0> right
20140119 21:16:01 <hharrison> core or compat profiles
20140119 21:16:20 <hharrison> Partial 4.x currently, full 4.x on the way
20140119 21:16:34 <hharrison> I like that they have intel, amd and nvidia all providing patches here
20140119 21:17:27 <rmk0> i'm kind of amused that people are ganging up on microsoft
20140119 21:17:49 <rmk0> suddenly everyone wants to be valve's friend
20140119 21:18:22 <hharrison> Attacking windows on one side...but I find it interesting that directX is back on the defensive
20140119 21:19:23 <hharrison> have to step away for a bit, desparate dog
20140119 22:13:05 <Eclesia> good night all ++
20140119 22:13:12 * Eclesia (~eclesia@anon) Quit (Quit: Leaving.)
20140119 22:15:30 * monsieur_max (~maxime@anon) has left #jogamp
20140120 00:17:56 * D3zmodos1 (~D3zmodos@anon) Quit (Read error: Connection reset by peer)
20140120 00:18:17 * sgothel (~sgothel@anon) Quit (Ping timeout: 252 seconds)
20140120 00:18:46 * sgothel (~sgothel@anon) has joined #jogamp
20140120 00:18:56 * sgothel (~sgothel@anon) Quit (Changing host)
20140120 00:18:56 * sgothel (~sgothel@anon) has joined #jogamp
20140120 00:18:56 * ChanServ sets mode +v sgothel
20140120 00:18:57 * D3zmodos (~D3zmodos@anon) has joined #jogamp
20140120 03:43:21 * torgashov (~torgashov@anon) has joined #jogamp
20140120 04:14:04 * torgashov (~torgashov@anon) Quit (Remote host closed the connection)
20140120 04:42:03 * hharrison (~chatzilla@anon) Quit (Ping timeout: 260 seconds)
20140120 05:05:39 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20140120050539.html