January 17, 2009 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
January 17, 2009 [00:01:03] *** groton_ has joined ##OpenGL[00:01:14] *** groton has quit IRC[00:09:23] <emzic> hi i got a question about the GLSL funciton shadow2D[00:09:43] <emzic> i assume the first paramter is the shadowmap sampler but what is the second one?[00:10:22] <MatthiasM> emzic: read the GLSL spec[00:10:35] <RTFM_FTW> the coordinate(s) to use during the sample operation[00:10:36] <emzic> or.... you just give me the short answer? :D[00:10:40] <RTFM_FTW> I just did[00:10:45] <emzic> cool[00:10:52] <emzic> but it takes a vec3[00:10:59] <emzic> and my coordinates are only vec2[00:12:36] <RTFM_FTW> read through the GL_ARB_fragment_program_shadow specification[00:13:06] <RTFM_FTW> it will help to clarify why shadow lookups use 3 components (for the non-perspective cases) versus 2 for 2D[00:13:55] <emzic> ok[00:14:33] <emzic> another question, is it theoretically possible to do shadowmapping with "regular" textures and lookups?[00:14:44] <MatthiasM> yes[00:14:45] *** vampirefrog has quit IRC[00:14:52] <MatthiasM> but more work[00:14:52] <RTFM_FTW> the third component is used for the R compare[00:14:57] <emzic> ok thanks![00:15:26] <RTFM_FTW> and yes you can do the R compare manually but I wouldn't encourage this[00:15:43] <RTFM_FTW> at least on HW which offers dedicated shadow sampling support[00:15:53] <emzic> i did it once in HLSL[00:16:15] <RTFM_FTW> since the underlying HW can do it is a much more optimal (and higher-quality) manner[00:16:20] <emzic> ok, where can i find more information on this "R compare" ?[00:16:27] <RTFM_FTW> read ARB_shadow[00:16:50] <RTFM_FTW> ARB_shadow, EXT_shadow_funcs, ARB_shadow_ambient and ARB_fragment_program_shadow are the extensions to note[00:16:59] <RTFM_FTW> ARB_depth_texture as well[00:17:01] <emzic> ok thanks[00:18:08] *** Bollinger has quit IRC[00:22:31] *** nailora has quit IRC[00:22:50] *** pietia has quit IRC[00:24:56] *** scai has left ##opengl[00:25:33] *** Marin_ has joined ##OpenGL[00:26:13] *** mmmmmmm has joined ##opengl[00:26:44] *** Marin has quit IRC[00:29:29] *** Yustme has quit IRC[00:32:00] *** groton_ has quit IRC[00:32:01] *** mmmmmmm is now known as mondi[00:32:30] *** Amorphous has quit IRC[00:38:26] *** Eforen has joined ##opengl[00:41:20] *** vick has quit IRC[00:45:40] *** Eforen has quit IRC[00:45:49] *** Eforen has joined ##opengl[00:46:38] *** k3t has quit IRC[00:47:01] *** Amorphous has joined ##opengl[00:55:09] *** eXtronuS_ has quit IRC[00:55:37] *** amz has joined ##opengl[00:58:28] *** mondi has quit IRC[00:59:54] *** itewsh has quit IRC[01:03:39] *** calav3ra has joined ##opengl[01:08:31] *** KU0N has quit IRC[01:12:02] *** GX_ has quit IRC[01:13:01] *** X-Scale has left ##OpenGL[01:14:35] *** djork has joined ##opengl[01:14:46] *** elite01 has quit IRC[01:15:55] *** dvoid has quit IRC[01:21:13] *** Suprano has quit IRC[01:24:01] *** AndyCR has joined ##OpenGL[01:24:14] <AndyCR> Quick question[01:24:40] <AndyCR> Let's say I wanted to do something like the iPhone does - where you have something in orthographic mode (IE a menu), but then want it to fly off into space[01:25:06] <AndyCR> The problem is, how would I know what the right coordinates for it are in perspective so there isn't a funny jump as it goes away?[01:25:27] <AndyCR> I have it rendering in ortho correctly, I can render in perspective, I just don't know how to smoothely blend between them[01:25:32] <MatthiasM> AndyCR: you do the math[01:26:02] <AndyCR> Do you have any pointers whatsoever for even what I would search for?[01:26:16] *** mm765 is now known as mm765^sleep[01:26:17] <MatthiasM> http://www.opengl.org/sdk/[01:26:22] <AndyCR> Projection matrices aren't my area of expertise[01:26:32] <Weiss> AndyCR: study the matrix maths involved with the perspective and orthographic projections[01:26:51] <Weiss> set up a set of equations that give the same (algebraic) value[01:27:30] <AndyCR> well, it's a start. i'll see what I can study.[01:27:31] <AndyCR> Thanks.[01:27:37] <Weiss> or... just keep the perspective projection all the time and figure out the right distance away for it to be to fill the screen[01:27:58] <AndyCR> The problem with me is I always try to get away with learning as little as possible to accomplish the goal. I need to stop that.[01:28:18] <AndyCR> I didn't care how projection matrices worked, as long as they did what I wanted them to.[01:28:45] <AndyCR> I guess that's a bit lazy[01:29:09] <Weiss> yep, i think for this you have no choice[01:30:08] <AndyCR> I'm used to 3D graphics engines, and the transition to pure OpenGL has been easier than I thought, but there are these sharp little roadbumps like this once in awhile.[01:31:59] <Weiss> how's your algebra? it might not be as bad as you think[01:32:15] <AndyCR> pretty good[01:32:22] <AndyCR> I never had any real trouble understanding math[01:32:27] <AndyCR> I should be fine, I guess[01:32:30] <Weiss> yeah[01:32:37] <AndyCR> My biggest issue with most things is that I'm inherently lazy :)[01:32:38] <Weiss> the manuals have the exact forms of the matrices which you need[01:33:08] <AndyCR> thanks so much for everything[01:33:38] <Weiss> no problem[01:33:44] * Weiss didn't do very much :)[01:36:32] *** montana has joined ##OpenGL[01:48:04] *** AndyCR has quit IRC[01:48:43] *** aib42 has left ##OpenGL[01:49:08] *** iammisc has joined ##OpenGL[01:50:24] *** pwned has quit IRC[01:51:46] <iammisc> Is it faster to use my own matrix class to do rotations, etc. and load them using glLoadMatrix or to use opengl functions and then use a glGet?[01:53:46] <MatthiasM> iammisc: depends on how good your math classes are[01:54:33] *** emzic has quit IRC[01:56:14] <RTFM_FTW> honestly neither approach is likely to be performance limiting[01:56:45] <RTFM_FTW> OTOH doing the math yourself (via a dedicated matrix class or otherwise) is far more flexible (and far cleaner IMHO)[01:57:16] <iammisc> so RTFM_FTW, you're saying not to use the opengl functions like glRotate, glTranslate, etc. at all?[01:57:40] <iammisc> Because I too, think that that is cleaner than using glGet, but I wasn't sure if that was a viable approach[01:57:42] <montana> hi guys is anyone willing to help me complete an opengl task, i have to parralax map a cube[01:57:43] <RTFM_FTW> I wouldn't recommend it[01:58:01] <iammisc> RTFM_FTW: that was kinda ambiguous wouldn't recommend what?[01:58:31] <RTFM_FTW> "... you're saying not to use the opengl functions like glRotate, glTranslate, etc. at all?" -> "I wouldn't recommend it"[01:59:13] <RTFM_FTW> in any case building a matrix class is a trivial exercise so do it[01:59:21] <iammisc> So you're saying that using my own matrices is something you do not recommend? or are you saying that I should use matrix classes AND opengl functions?[01:59:59] *** LordMetroid has quit IRC[02:00:59] <montana> anyone?[02:01:03] <montana> :)[02:03:14] *** linxuz3r has joined ##OpenGL[02:03:15] <linxuz3r> hi[02:03:21] <linxuz3r> can someone help please[02:09:28] *** iammisc has quit IRC[02:14:49] <montana> hi guys is anyone willing to help me complete an opengl task, i have to parralax map a cube[02:14:52] *** AlastairLynn has quit IRC[02:16:14] <MatthiasM> http://mikeash.com/getting_answers.html[02:23:39] *** vade has quit IRC[02:25:08] <linxuz3r> im using freeglut for my glut when i compile a simple program to display a polygon and run the program i see the program window with the polygon immediatly followed by a blank window as a result a blank window "overlaps" the window with the polygon how can i fix this[02:25:20] <linxuz3r> this happens everytime i run the program[02:25:22] <linxuz3r> please[02:25:58] *** Marin_ has quit IRC[02:38:45] <rnx> post the code to a pastebin[02:40:16] <linxuz3r> http://gamedev.pastebin.com/m399612b2[02:40:33] *** cs_student has quit IRC[02:43:46] <rnx> glutSwapBuffers at the end of drawglscene[02:46:49] *** korff_home has quit IRC[02:49:03] <RTFM_FTW> actually he doesn't need (nor does he want) to use glutSwapBuffers there[02:49:13] <RTFM_FTW> being that his context is single buffered[02:49:48] <RTFM_FTW> if you want glutSwapBuffers then you need to replace GLUT_SINGLE with GLUT_DOUBLE during context creation[02:53:45] <linxuz3r> i got it[02:54:01] <linxuz3r> its my compiz fusion that is getting in the way[02:56:53] <rnx> ah ... i only checked if it works with my first idea[03:00:40] <RTFM_FTW> well either route (single *or* double buffered) should work[03:02:04] *** montana has quit IRC[03:05:43] *** Vaydorg has joined ##OpenGL[03:06:26] <Vaydorg> Hello, i am facing a bit of a problem with skybox. It appears that the 'dimensions' of the box is apparent, and it doesn't give the illusion of an enviroment. http://img127.imageshack.us/my.php?image=hostve7.jpg , Thanks in advance.[03:07:07] *** qeed_ has quit IRC[03:10:19] *** reprore_ has joined ##OpenGL[03:10:24] *** NotWil has joined ##OpenGL[03:16:17] *** zacs7 has joined ##opengl[03:17:22] *** rabbit- has joined ##OpenGL[03:19:19] <Vaydorg> It sure is quite in here.[03:20:48] *** a-stray-cat` has quit IRC[03:25:03] *** Wil has quit IRC[03:25:32] *** cs_student has joined ##OpenGL[03:27:32] <Vaydorg> anybody fammiliar with the concept of 'skybox' ?[03:29:30] *** jm has joined ##OpenGL[03:29:52] *** ectropy_ has joined ##OpenGL[03:31:35] <Spark> render a box first[03:31:36] <Spark> done[03:32:49] <Vaydorg> Spark: What do you mean ?[03:32:59] <Vaydorg> hold on let me give you a screen shot[03:33:39] <zacs7> Vaydorg, er... render a box and put textures on it then...? :-)[03:34:10] <Vaydorg> http://img207.imageshack.us/my.php?image=hostda5.jpg[03:34:20] <Vaydorg> you see the lines on the box is apparent.[03:35:04] <Vaydorg> Spark: How do i solve that ?[03:35:45] <Spark> what am i looking at exawctly[03:36:12] <Spark> problem with mipmaps and clamping?[03:36:38] <Vaydorg> I'm not sure :([03:37:06] *** Orphis has joined ##opengl[03:37:12] <zacs7> Vaydorg, clamp & you may have to turn off interpolation[03:37:37] *** MatthiasM has quit IRC[03:37:43] *** MatthiasM has joined ##opengl[03:37:55] *** calav3ra has quit IRC[03:37:58] *** Orphis_ has quit IRC[03:38:16] *** djork has quit IRC[03:38:25] *** amz has quit IRC[03:38:26] <Vaydorg> zacs7: I am using http://rafb.net/p/wixvII22.html as parameters for texturing[03:39:09] <linxuz3r> what is the commonly used windowing system with opengl?[03:39:21] <zacs7> glfw or glut or anything really[03:39:51] <zacs7> Vaydorg, try turn off linear interpolation ie (GL_NEAREST)[03:40:04] *** vade has joined ##OpenGL[03:40:08] <linxuz3r> what about sdl[03:40:12] <zacs7> What about it?[03:40:14] <linxuz3r> is sdl much better than glut[03:40:24] <zacs7> They do different things[03:40:27] <zacs7> really..[03:40:34] <zacs7> But yes, some may say it is[03:40:43] <zacs7> I'd go for glfw if I was you[03:41:21] <Vaydorg> zacs7: Other than those parameters, what else do i need ?[03:41:31] <zacs7> Vaydorg, nothing[03:41:49] <Spark> there is also cegui for example[03:42:15] <Spark> and similar toolkits that are designed for rendering with meshes/textures[03:42:16] <Vaydorg> zacs7: But there still seems to be a line, can't i do anything about it ?[03:42:28] <zacs7> Call ze police[03:42:36] <Vaydorg> lol[03:42:38] <Vaydorg> really[03:42:39] <zacs7> Ooh line not lion :D[03:43:13] <Spark> turn off mipmaps[03:43:19] <Spark> it's probably bleeding from the other side[03:45:19] <Vaydorg> How do i do that ?[03:45:49] <zacs7> glPleaseTurnOffMipmapping()[03:46:01] <linxuz3r> is it easy to learn NeHe and redbook with SDL or with GLUT?[03:46:13] <Vaydorg> heh[03:46:14] <zacs7> linxuz3r, makes no difference[03:46:41] <linxuz3r> why do you say makes no difference?[03:47:36] <zacs7> linxuz3r, because the redbook and Nehe teach opengl[03:47:45] <linxuz3r> lol i know[03:47:54] <linxuz3r> but it they use glut[03:48:21] <zacs7> linxuz3r, it makes no difference what they use.[03:49:16] <linxuz3r> ok ill learn both[03:49:27] *** vade has quit IRC[04:04:52] *** vade has joined ##OpenGL[04:05:32] *** AndyCR has joined ##OpenGL[04:15:56] *** Wil has joined ##OpenGL[04:16:16] *** vade has quit IRC[04:18:03] *** AndyCR has left ##OpenGL[04:20:29] *** amz has joined ##opengl[04:23:15] *** HuntsMan has joined ##opengl[04:27:41] *** NotWil has quit IRC[04:53:49] *** dusted has quit IRC[05:01:51] *** cplusplus has quit IRC[05:03:56] *** kbotnen has joined ##OpenGL[05:10:00] *** bijoo_osdev has quit IRC[05:10:00] *** bijoo_appdev has quit IRC[05:11:57] *** bijoo_appdev has joined ##OpenGL[05:11:57] *** bijoo_osdev has joined ##opengl[05:17:52] *** zacs7 has quit IRC[05:21:22] *** cs_student has quit IRC[05:30:50] *** jm has quit IRC[05:31:43] *** ol1veira_ has joined ##OpenGL[05:44:46] *** jfroy has joined ##OpenGL[05:49:22] *** Eforen has quit IRC[05:58:45] *** zacs7 has joined ##opengl[06:06:41] *** kbotnen has quit IRC[06:07:38] *** sbwilson has quit IRC[06:16:22] *** rabbit- has quit IRC[06:21:58] *** reprore_ has quit IRC[06:30:38] *** ectropy_ has quit IRC[06:34:23] *** TheLorax has joined ##opengl[06:40:02] *** Cultz has quit IRC[06:40:10] *** TheLorax has quit IRC[06:40:15] *** TheLorax has joined ##opengl[06:43:03] *** zacs7 has quit IRC[06:44:20] *** stringfellow has joined ##opengl[06:49:47] *** Quentarez has quit IRC[07:05:39] *** rutski_ has quit IRC[07:06:20] *** rabbit- has joined ##OpenGL[07:14:07] *** TheLorax has quit IRC[07:34:24] *** hdon has joined ##OpenGL[07:35:38] *** kbotnen has joined ##OpenGL[07:39:43] *** rabbit- has quit IRC[07:41:36] *** vade has joined ##OpenGL[07:47:47] *** bijoo_appdev has quit IRC[07:47:47] *** bijoo_osdev has quit IRC[07:48:20] *** bijoo_osdev has joined ##opengl[07:48:20] *** bijoo_appdev has joined ##OpenGL[07:48:58] <Vaydorg> Hello.[07:51:03] *** zacs7 has joined ##opengl[07:53:56] <hdon> hola[07:56:09] *** UUncia has joined ##OpenGL[08:04:54] *** epifanio has quit IRC[08:20:02] *** servus_ has quit IRC[08:23:41] *** servus_ has joined ##opengl[08:27:05] *** korff_home has joined ##OpenGL[08:27:57] *** amz has quit IRC[08:28:37] *** kbotnen has quit IRC[08:28:55] *** vade has quit IRC[08:40:20] *** neoneye has joined ##OpenGL[08:44:18] <hdon> should glViewport() clear my color buffer?[08:48:13] <mlucassmith> no[08:53:06] <hdon> interesting.. it seems to be doing just that...[08:56:47] <mlucassmith> only Clear will clear the active buffer[08:56:58] <mlucassmith> and it'll only do it for the layers you specify[08:58:12] *** GinoMan has joined ##OpenGL[08:59:14] <hdon> is there a transformation matrix that affects the values incoming fragments are tested against? (for instance adding an offset to the stencil buffer)[09:00:32] *** bijoo_osdev has quit IRC[09:00:32] *** bijoo_appdev has quit IRC[09:06:08] <hdon> mlucassmith: i am only working with the color buffer and the stencil buffer. calling glViewport() between back-buffer flips results invariably in a black screen on my machine[09:09:54] <mlucassmith> viewport only changes the final device coordinates that your scene renders in to AFAIK[09:11:24] <hdon> mlucassmith: yeah that's what the docs say, and that's what makes *sense*[09:11:46] <hdon> i'm going to try something..[09:15:20] <hdon> i am passing my arguments in the wrong order[09:18:31] <hdon> actually it seems that glViewport() does what i was hoping it would do, and defines a rasterization offset and boundaries for rasterization[09:19:04] <hdon> for instance i am a points at the same location four times and successfully using glViewport() to arrange them so they don't end up in the same spot[09:19:46] <hdon> mlucassmith: if i give you a test program can you try it and see if it has the same effect?[09:20:17] <hdon> mlucassmith: http://gist.github.com/48290[09:20:26] <hdon> it uses SDL but you can change it a bit to suit different needs[09:22:48] *** linxuz3r has quit IRC[09:23:15] <mlucassmith> uhm, no, sorry, it's past midnight and i'm playing with ring buffers[09:24:45] <hdon> ok[09:24:47] *** vade has joined ##OpenGL[09:24:53] <hdon> mlucassmith: what are ring buffers?[09:25:06] <mlucassmith> circular buffers, buffers that wrap around[09:25:20] <hdon> mlucassmith: that's what i would have thought, but... sounds unrelated to opengl[09:25:57] <mlucassmith> it is[09:28:15] <zacs7> Not totally unrelated ;0[09:29:14] *** SEAL_45 has joined ##OpenGL[09:30:26] *** SEAL_45 has quit IRC[09:32:23] *** zacs7 has quit IRC[09:37:51] *** Yustme has joined ##OpenGL[09:37:52] *** pwned has joined ##opengl[09:39:10] *** tmccrary has quit IRC[09:41:31] *** scai has joined ##opengl[09:41:51] *** tmccrary has joined ##OpenGL[09:49:38] *** fargiolas has joined ##OpenGL[09:50:38] *** vade has quit IRC[09:51:33] *** groton has joined ##OpenGL[09:58:54] *** ata2 has joined ##OpenGL[10:00:30] *** zacs7 has joined ##opengl[10:02:21] *** fargiol`` has quit IRC[10:11:34] *** ata2 has quit IRC[10:15:13] *** pietia has joined ##OpenGL[10:16:04] *** Marin has joined ##OpenGL[10:17:04] <Marin> I set that when I press key, zPos+=10; but it doesn't seems like it zooms in, just when I increase it over the nearZ it dissapears, nothing else[10:17:38] <Marin> Does anyone know what am I doing wrong=?[10:19:27] <zacs7> Marin: what are your near and far culling values?[10:19:39] <Marin> f[10:19:59] <Marin> so after I press the key 10 times it dissapears[10:20:05] <Marin> nothing else changes, like zooming.[10:26:27] *** pietia has quit IRC[10:31:05] *** pietia has joined ##OpenGL[10:31:51] *** geocalc has quit IRC[10:32:34] *** pietia has quit IRC[10:34:09] *** nailora has joined ##OpenGL[10:40:51] *** dvoid has joined ##OpenGL[10:41:05] *** Ingenu has joined ##OpenGL[10:41:18] *** geocalc has joined ##opengl[10:43:10] *** pietia has joined ##OpenGL[10:48:25] *** Gorgoroth has quit IRC[10:49:45] <Marin> Why it seems like this GL_LINE render seems like lines aren't totally white, http://imagebin.ca/view/zV_cMg.html[10:49:59] <Marin> It doesn't look smooth[10:50:14] <zacs7> Marin: seriously a bitmap!?!?[10:50:24] <hdon> yeah seriously[10:50:25] <Marin> That?[10:50:28] <MatthiasM> Marin: use PNGs[10:50:31] <hdon> Marin: use PNG[10:50:33] <hdon> ha[10:50:37] <hdon> Marin: next time for your screen shot[10:50:57] <hdon> Marin: did you want the lines to be aliased?[10:50:58] <Marin> Stupid windows only give .bmp's when I take screenshot :([10:51:06] <zacs7> Marin: then convert it[10:51:07] <MatthiasM> Marin: and make sure you render lines on pixel locations - otherwise you get this kind of anti-aliasing[10:51:19] <MatthiasM> Marin: windows doesn't write a file at all[10:51:24] <Marin> Pleaase, take a look at lines of this game http://www.frontiernet.net/~imaging/lunar_lander_game.html[10:51:28] <Marin> It is really smooth[10:51:33] <Marin> Not smooth[10:51:42] <Marin> I mean they are whiter[10:52:02] <Marin> And only one line, when you look close at my ship you will see that there are actually 2 rows of pixels to represent one line[10:52:10] <hdon> MatthiasM: will rounding line vertex coordinates to pixel values really work without drawing orthogonal lines?[10:52:22] <hdon> Marin: that's antialiasing[10:52:30] <hdon> Marin: you'll have to disable it, but i'm no expert on that[10:52:34] <Marin> How to turn it off?[10:52:41] <Marin> glDisable(GL_ANTIALIASING) ?[10:52:47] <zacs7> Marin: NO[10:52:51] <MatthiasM> read the red book on pixel exact rendering[10:53:07] <zacs7> Marin: http://www.opengl.org/resources/code/samples/sig99/advanced99/notes/node120.html[10:53:14] <Marin> Man, I don't have enough time now to read any book, can you give me quick solution ,thx[10:53:20] <hdon> Marin: if you're only doing 2D graphics and you don't need openGL, SDL 1.3 has line drawing primitives[10:53:43] <Marin> I am practicing OpenGL,that's why I use it :)[10:53:47] <hdon> ok[10:53:50] <hdon> go read then[10:54:13] <Rangar> Does anyone know of a simple sample that uses CgFX to display a quad with a texture?... no matter what I do i cannot get the texture to display :([10:54:38] <hdon> is that an #opengl appropriate question?[10:55:03] <Rangar> yes because it's through opengl i'm trying to use Cg[10:55:18] <Rangar> it's the opengl interface side i'm hacving issues with[10:55:26] *** povman has joined ##OpenGL[10:55:34] <hdon> ah[10:55:43] <hdon> how do you know? maybe your shader is broken ;)[10:55:53] *** maxton has quit IRC[10:55:59] <Rangar> it works fine in the nVidia browser[10:57:23] *** maxton has joined ##OpenGL[10:57:29] *** Bollinger has joined ##OpenGL[10:58:22] *** servus_ has quit IRC[10:59:40] *** eXtronuS_ has joined ##OpenGL[11:12:14] *** Suprano has joined ##OpenGL[11:19:47] *** servus_ has joined ##opengl[11:22:36] *** rnx has left ##opengl[11:33:22] *** zacs7 has quit IRC[11:35:01] *** predaeus has joined ##opengl[11:54:27] *** zacs7 has joined ##opengl[11:56:37] *** Vaydorg has quit IRC[11:58:24] *** sohail has quit IRC[12:01:08] *** Renderwahn has joined ##OpenGL[12:10:48] *** elite01 has joined ##opengl[12:11:47] <Marin> When I use glRotatef(yAngle,0,0,1); glTranslatef(0,yPos,zPos); it seems like object is rotating about some other invisible object, instead of himself, how to fix it?[12:13:12] <zacs7> Marin: did you translate to the point you want to rotate around first?[12:13:27] <Marin> no, I rotate first[12:13:28] <predaeus> Marin, read http://www.glprogramming.com/red/chapter03.html#name2[12:14:16] <predaeus> Marin, and study the other chapters of the book aswell, it's a good introduction.[12:19:45] <hdon> I'm trying to load GL_RED GL_UNSIGNED_BYTE pixel data into a texture using glTexImage2D(), but my texture is always solid red even though my data is just random()%256 at every pixel: http://gist.github.com/48314[12:19:48] *** b0000 has joined ##opengl[12:21:22] <hdon> i am rendering it with glColor3d(1,1,1) and glOrtho(0,0,screenw,screenh) with my rectangle aligned orthogonally 256 pixels wide[12:21:54] <hdon> but the color is solid every time, like random() was giving me the same values every time it was called[12:22:36] <pietia> anyone familiar with lwjgl ?[12:22:42] <zacs7> Well, glColor3d(1, 1, 1) is wrong for one ;)[12:22:52] <hdon> zacs7: why?[12:22:59] <hdon> pietia: i am a little[12:23:15] <hdon> pietia: i evaluated writing games with it in javascript using rhino -- way too slow[12:23:58] <hdon> (it seems rhino makes no optimizations of the code it compiles when it compiles them to java classes)[12:24:20] <pietia> hdon, were native libs should be placed ?[12:24:43] <hdon> pietia: let me check where i have them[12:24:48] <hdon> pietia: i remember having that difficulty too[12:24:50] <pietia> hdon, thx[12:24:53] <pietia> Exception in thread "main" java.lang.UnsatisfiedLinkError: no lwjgl in java.library.path[12:24:55] <pietia> :([12:25:03] <pietia> placed it _everywhere_ in JAVA_HOME[12:25:15] <pietia> and in my libs , current path, bin folder...[12:25:17] *** servus__ has joined ##opengl[12:25:22] <Marin> I am making simple lunar lander game, and I got what I like(lander moves in direction I want when I use glRotate first), but it looks like he is rotating about some other object not itself, and when I use gltranslate first ofcourse, it rotates okay, around himself, but movement is always down, does anyone know how to fix that!?!?[12:25:23] <hdon> pietia: have you tried setting java.library.path on the commandline when you invoke the vm?[12:25:40] <pietia> hdon, it's java's parameter?[12:26:01] <hdon> Marin: what do you mean by "movement is always down?"[12:26:09] *** servus__ is now known as servus[12:26:17] *** servus_ has quit IRC[12:26:46] <Marin> well, It is lunar lander right, so with SPACE key I add some velocity and with LEFT, RIGHT keys I rotate the Lunar Lander[12:26:59] <Marin> And when I use glRotatef first[12:27:12] <Marin> It is facing the right angle direction when I press SPACE key[12:27:29] <zacs7> Marin: You can rotate and translate more than once you know...[12:27:33] <hdon> pietia: yes[12:27:44] <hdon> pietia: use -Djava.library.path="path"[12:27:57] <pietia> hdon, path is dir , not list of iles ?[12:28:04] <pietia> files[12:29:17] <hdon> pietia: path usually means this: http://support.netricks.com/index.php?_m=knowledgebase&_a=viewarticle&kbarticleid=64&nav=0,12[12:29:41] <hdon> pietia: but also in java they let you point to archive files which contain more directories and files inside of them (like JARs)[12:30:02] <hdon> pietia: i don't know if java will look inside archive files to find native libraries[12:30:11] <hdon> pietia: but in the case of lwjgl, it's just in the "native" directory[12:32:32] <hdon> pietia: also, in the case of programs like java or command shells, there are strings called "search paths" like the PATH environment variable or like a Java class path, which are lists of places to look for things[12:33:35] <predaeus> hdon, do you enable texturing?[12:33:58] <hdon> predaeus: when? before drawing? yes, but if i don't, the square is white, as my glColor3d(1,1,1) is white[12:34:08] <hdon> predaeus: if i change GL_RED to GL_BLUE i get a uniform blue square[12:34:13] <predaeus> hdon, do you specify texture coordinates?[12:34:25] *** wey has joined ##opengl[12:34:49] <hdon> predaeus: yes, i --- OH , i bet i forgot i have texture coords GL_CLAMPed[12:36:41] *** elite01 has quit IRC[12:36:50] *** groton has quit IRC[12:37:44] <hdon> predaeus: ugh...[12:37:53] <hdon> setting GL_TEXTURE_WRAP* to GL_REPEAT didn't help[12:38:22] *** groton has joined ##OpenGL[12:38:36] <hdon> predaeus: i'll paste all of the code[12:38:42] <predaeus> hdon, do you have different texture coordinates at each vertex?[12:39:13] <fusi> holding a clipboar and looking serious wiont get you through our building security[12:39:13] <hdon> predaeus: yes, my tex coord are identical to my vertex coords[12:39:17] <hdon> http://gist.github.com/48314[12:39:37] <hdon> just search for the word "texture" to find the relevant parts[12:39:39] <predaeus> fusi, :-D[12:40:00] <fusi> o/ :)[12:41:59] * fusi idly sets fire to their ati card[12:42:04] <predaeus> hdon, can it be that your texture coordinates lie outside [0, 1], they will be clamped to [0, 1] unless you use the texture_rectancle extension.[12:42:09] <predaeus> *rectangle[12:42:11] * fusi goes off in search of a functional nv card that has working drivers[12:42:29] <predaeus> hdon, use texture coordinates inside [0, 1].[12:42:50] <hdon> oh god.. i forgot i wasn't scaling my texture matrix in this app[12:42:53] <hdon> thanks predaeus[12:42:54] <fusi> do texture rectangles still do thr0~e 09..width thing[12:43:03] <fusi> the 00...width*[12:43:18] <predaeus> hdon, happens, np[12:43:19] <fusi> its been a while[12:45:53] <predaeus> fusi, what do you mean? texture_rectangle uses coordinates [0, {width, height}] AFAIK.[12:46:30] <fusi> yea thats what i mean[12:46:37] <fusi> just wondered if they might of changeed it :p[12:46:56] <fusi> i have onlyjjust gort up and am not thinking cvlearer[12:47:07] <predaeus> :-)[12:47:16] *** elite01 has joined ##opengl[12:47:35] <fusi> just after awakening my brain->mouth connection is on by default[12:47:49] <hdon> fusi: texture-mapped fragments are calculated by "mapping" texels (texture pixels) to raster positions (screen pixels) by multiplying all kinds of numbers and things. by default, the transformation matrix applied to the procedure casts the position (0,0) in one corner, and (1,1) in the opposite[12:48:08] <fusi> thanks hdon[12:48:25] <hdon> fusi: but you can change that by transforming the texture matrix[12:48:48] <fusi> to get repetition etc[12:52:59] <Madsy> [pedantic] hdon: Or use whatever transformations you want in the vertex shader [/pedantic][12:53:16] <Madsy> Why people still use the old matrix functions is beyond me :-)[12:53:18] *** XT95_ has joined ##OpenGL[12:53:46] <Madsy> glMult/LoadMatrix, glRotate and glTransform[12:54:04] <hdon> Madsy: i've only ever used fragment shaders, but this card doesn't have vertex shaders[12:55:02] <hdon> Madsy: one reason i can think of to use old matrix functions is for backward compatibility. if what you want to do can be done equally well without shaders, you might at least *consider* supporting cards like mine ;)[12:55:31] <hdon> err.. wait this card has vertex shaders, it's fragment shaders that it doesn't have[12:55:46] <hdon> what card has fragment shaders and no vertex shaders? unheard of[12:56:08] <zacs7> The card I made out of corn chips![12:57:42] <hdon> i believe in your corn chip gup zacs[12:57:44] <hdon> gpu*[12:57:49] <Madsy> Hm, I guess you can call glAttachObject only for a fragment shader and omit a vertex shader. I didn't realize that there are cards that only support fragment shaders.[12:57:58] <Madsy> Are you sure that's the case? It sounds rather weird.[12:58:19] <Madsy> Ah, I read upwards now :-P[12:58:23] <hdon> i guess i'm not positive, but the previous owner of this machine said it was. i took his word because he was the guy who started the Beryl project[12:58:26] <Madsy> That makes much more sense.[12:58:45] <hdon> oh, haha... i misread your message thinking you were already up to date[12:58:48] <hdon> on my mishaps[12:58:53] <hdon> but no one can be up to date on those[12:58:58] <hdon> they come too fast...[12:59:43] <Madsy> hdon: Vertex shaders and vertex programs have been around since.. err 2001 or so :)[12:59:52] <Madsy> GLSL fragment shaders since 2003-2004[13:00:03] <Madsy> fragment programs since around 2002[13:00:56] <Madsy> So, yeah. People without even support for fragment program and vertex program can go screw themselves in my opinion :-P[13:01:24] <hdon> Madsy: well it's not OEM, the video card is probably from around that era. it is also dual head, which was rare at the time[13:01:35] <Madsy> Matrox?[13:02:01] <hdon> yes[13:02:14] <hdon> it has some problems with the nv driver :([13:02:23] <hdon> i have to use the binary from nvidia[13:02:40] <Madsy> Matrox G400 was a cool card[13:02:53] <Madsy> The first one to support environmental bumpmapping[13:03:04] <Madsy> Back in 99 or so :-P[13:03:08] <zacs7> It still can't beat the best card in history... FX5200 or MX440 (I can't decide!)[13:03:09] <hdon> but for whatever reason, nvidia's binary driver seems to be TERRIBLY slow at non-opengl drawing under the x.org that ships with Debian 4.0[13:03:25] *** jm|earth has joined ##OpenGL[13:03:31] <Madsy> zacs7: The Geforce FX series? Are you serious?[13:03:39] <Madsy> That family suck balls![13:04:17] <zacs7> Madsy: Even the fantastic 5200?[13:04:26] <zacs7> For the money it costed at the time...[13:07:31] <Plagman> hdon: install 180.22[13:08:01] <hdon> Plagman: is that a version of the nvidia blob?[13:08:22] <Madsy> Heh.. blob. I love that word.[13:08:55] <Plagman> yes[13:09:22] <Plagman> it's the latest stable one, and probably not the one that's in debian 4.0[13:09:46] <Plagman> and it's supposed to help a lot with 2D acceleration[13:10:19] <Plagman> oh wait, your card is legacy?[13:10:23] <Plagman> you can't use that then[13:10:26] <Plagman> whatever[13:10:34] <hdon> :( :( :([13:11:31] <hdon> ugh... glCopyTexImage2D() seems really slow. am i imagining things?[13:11:43] <hdon> i don't remember it being this slow on my iBook G4[13:11:49] <zacs7> hdon: It is VERY slow[13:11:57] <hdon> zacs7: what should i do?[13:12:05] <zacs7> Find another way ;)[13:12:32] <hdon> can't you just tell me another way?[13:12:36] <hdon> :P[13:13:04] <hdon> i think i used the Sub variety of this function before, but i doubt that's the critical difference[13:13:51] <hdon> oh wait, maayybe it is[13:14:23] <hdon> i notice there's no internalformat argument[13:14:26] <hdon> and it is definitely faster![13:14:55] <Madsy> hdon: Do you want render-to-texture? Use Pixelbuffer Objects or Framebuffer Objects[13:15:24] <Madsy> It saves you from a slow copy.[13:15:33] <hdon> Madsy: ah, good point[13:16:49] <Madsy> I avoided pixel buffer objects for a long time, because it requires a lot of platform-specific handling.[13:16:59] <hdon> ugh[13:17:02] <hdon> lame[13:17:02] <Madsy> But the ARB extension is surprisinly clean.[13:17:11] <Madsy> s/requires/required[13:17:20] <Madsy> I did before, but they improved it.[13:17:22] <hdon> hmm, wait.. i think i heard something about SDL reconciling PBO API discrepancies between platforms[13:17:27] <Madsy> Gah. It*[13:18:08] *** LordMetroid has joined ##OpenGL[13:18:14] <Madsy> GL_ARB_pixelbuffer_object doesn't depend on an OS-specific memcopy or anything like that anymore.[13:18:32] <Madsy> You can use glMapBuffer for streaming, just like you do with vertex buffer objects.[13:18:40] <hdon> hmm, it looks like it was a proposed API for 1.2.6 way back in 2004, but i don't see it in 1.2.13[13:18:48] <hdon> streaming?[13:18:54] <hdon> like between client memory and video memory?[13:19:00] <Madsy> Yeah.[13:19:04] <hdon> i don't need to do that[13:19:14] <hdon> although... i guess i will in the future[13:19:22] <hdon> right now i'm just trying to get the feedback loop working[13:19:28] <Madsy> The buffer contents aren't flagged for a specific purpose either. You can render to the buffer and use it as vertex data.[13:19:47] <hdon> oh.. i could do all kinds of cool things with that[13:19:55] <hdon> i'll haev to think about it[13:20:06] <Madsy> I don't know how efficient that is, though.[13:20:14] <hdon> probably varies[13:20:38] <Madsy> The geometry shaders are better for such a usage.[13:20:44] <hdon> oh, i guess doing water ripples is probably the first thing everyone does with this capability[13:21:14] <Madsy> Water ripples can be done trivially in the vertex shader.[13:22:04] <hdon> sure, but if you wanted to use arb buffers in a learning exercise..[13:22:15] <hdon> do you have a better idea for a learning exercise?[13:22:27] <Madsy> Learning how to use what?[13:22:53] <hdon> these buffers defined in this ARB extension[13:24:30] <Madsy> Which one of them? I've mentioned GL_ARB_pixel_buffer_object, GL_ARB_framebuffer_Object, and GL_ARB_vertex_buffer_object.[13:24:54] <hdon> i thought you said you could store any data in those buffers[13:25:13] <Madsy> Only in pixel_buffer_object, which extends on vertex_buffer_object[13:25:22] <hdon> ah ok[13:25:26] <hdon> well that's what i meant[13:25:55] *** emzic has joined ##OpenGL[13:26:22] <Madsy> I'd say, don't learn something just due to the "bling" alone. If you don't see a need, it's probably not needed.[13:26:52] <pietia> hdon, ok, now i'm getting java.lang.UnsatisfiedLinkError: org.lwjgl.DefaultSysImplementation.getNativeLibraryVersion()Ljava/lang/String;[13:27:02] <pietia> hdon, path is pointed correctly[13:27:03] <Madsy> If you want to render to textures, use GL_ARB_framebuffer object.[13:27:47] <hdon> pietia: i'm not an expert but i think you might have a version of lwjgl for a newer version of java than you have[13:28:08] <hdon> pietia: no nevermind i misread[13:28:20] <pietia> hdon, this an error from tutorial[13:28:29] <pietia> http://www.lwjgl.org/installation.php[13:28:50] <pietia> error when i try to run code from tutorial :)[13:29:33] <hdon> pietia: let me see the command you used to run your program[13:30:01] <pietia> java -cp .:res:jar/lwjgl.jar:jar/lwjgl_test.jar:jar/lwjgl_util.jar:jar/lwjgl_fmod3.jar:jar/lwjgl_devil.jar:jar/jinput.jar: -Djava.library.path=/dane/projekty/lib/lwjgl/lwjgl-2.0.1/native/linux/ org.lwjgl.test.WindowCreationTest[13:30:06] <hdon> pietia: also i recommend creating a shell script / batch file to run your program so that you don't have to retype the command[13:32:21] <pietia> hdon, ha ![13:32:37] <hdon> pietia: what?[13:32:54] <pietia> hdon, i use 1.6.0.11 , but it doesn't want to work. tried openjdk (1.7) and it works ;)[13:33:07] <hdon> great[13:33:16] <pietia> maybe some bug in 11[13:33:23] *** jimtendo has joined ##OpenGL[13:33:49] <jimtendo> Hey guys. Has anyone here ever worked with Collada?[13:33:50] <hdon> pietia: are you programming games mostly in java? or just using the JVM as a platform?[13:34:29] <hdon> man i'm having a hard time finding straightforward reference documentation for these ARB PBO APIs[13:34:35] <predaeus> jimtendo, I'm a total collada newb, but what do you need?[13:34:39] *** emzic has quit IRC[13:34:51] <hdon> all i'm finding is tutorials and stuff[13:35:28] <pietia> hdon, i learn opengl with qt (university course) but i like java and i would like to learn also lwjgl, jme etc[13:35:29] <jimtendo> predaeus, I'm trying to understand what the parameter "offset" means when used like this: <input offset="1" semantic="NORMAL" source="#Cube-Geometry-Normals"/>[13:35:55] <hdon> pietia: ok, good luck[13:36:02] <predaeus> hdon, look here http://www.opengl.org/registry/[13:36:11] <hdon> predaeus: there now ;)[13:36:44] <predaeus> jimtendo, gimme a sec, was confused about that too[13:36:49] <jimtendo> Cheers.[13:37:53] *** LtJax has joined ##opengl[13:44:10] <predaeus> jimtendo, I hope I don't mix up stuff here. The offset is the offset into the <p> indices array that is used for that input. To get the stride, you need to compute the maximum offset.[13:45:10] <predaeus> jimtendo, correction: the stride is the maximum offset + 1[13:45:38] <jimtendo> Ohhh, I think I'm starting to get it.[13:45:52] <hdon> what's collada?[13:46:31] <predaeus> jimtendo, I crosschecked with the Collada book and my very unfinished implementation. Unfortunately the book isn't that good in my opinion. But it's the only one and it helps.[13:46:46] <jimtendo> Is that the official documentation?[13:46:48] *** johndoe has joined ##opengl[13:47:04] <jimtendo> I think I understand what it's doing now.[13:47:08] <jimtendo> Cheers predaeus.[13:47:23] <predaeus> jimtendo, np[13:47:51] <hdon> now i remember why i didn't use PBOs back the last time i did this. these extension specs make references to all kinds of other extensions i don't know anything about either[13:48:14] <hdon> it's like i no frame of context[13:49:23] <predaeus> jimtendo, nah, it's Remi's book.[13:52:00] <Marin> If In coordinate system I have some point, let's say (-3,1) how can I calculate the angle of that point, if you understand me[13:52:31] *** povman has quit IRC[13:52:52] <predaeus> Marin, afaik something like 'angle of a point' doesn't exist in geometry.[13:52:56] <hdon> you mean the angle of that vector[13:53:01] <Marin> Yes[13:53:16] <hdon> cos(d\x/dy)[13:53:21] <hdon> remoev that \ in there[13:53:27] <hdon> err[13:53:28] <hdon> acos[13:53:45] <Marin> thx[13:53:47] <hdon> but, Marin, you might not need the angle[13:54:00] <hdon> Marin: you might just need the ratio of its components[13:54:03] <hdon> Marin: which is just dx/dy[13:54:09] <Marin> I need that, yes[13:54:24] <Marin> DO you know how to claculate it?[13:54:38] <hdon> Marin: i don't even know what it is you're calculating[13:54:45] <hdon> Marin: just telling you a mistake i made a long time ago ;)[13:54:47] *** predaeus has quit IRC[13:55:03] <hdon> it was a game with balls that bounced off each other and off the walls[13:55:12] <Marin> Ill draw :)[13:55:19] <hdon> and i didn't need angles, i really just needed ratios[13:58:28] <Marin> http://i310.photobucket.com/albums/kk431/systat_photo/lol.jpg[13:58:30] <Marin> hdon:[13:58:54] <Marin> That is lunar lander, and I want to know how I need to translate it, when It is on that angle[13:59:12] <hdon> yes, you don't need the angle. just think about what you would do with it once you knew the angle[13:59:21] <hdon> you would use trig functions to convert the angle into X and Y components[13:59:35] <hdon> which is exactly what you have to use to get from component values to an angle value[13:59:37] <Marin> I set angle, I mean I rotate my lunar lander, with LEFT RIGHT keys, and store the value[14:00:12] <hdon> pos_x += velocity * pos_x / pos_y[14:00:17] <hdon> OH[14:00:19] <hdon> to rotate it[14:00:28] <hdon> yeah i guess if you aren't using proper physics, you do need the angle[14:00:30] <Marin> NO[14:08:58] <hdon> lol[14:11:41] *** zacs7 has quit IRC[14:12:11] *** LordMetroid has quit IRC[14:14:52] *** Tempoe has quit IRC[14:14:52] *** GNU\caust1c has quit IRC[14:14:52] *** ville has quit IRC[14:14:52] *** dindinx has quit IRC[14:15:09] *** dindinx has joined ##OpenGL[14:15:09] *** ville has joined ##OpenGL[14:15:09] *** GNU\caust1c has joined ##OpenGL[14:15:09] *** Tempoe has joined ##OpenGL[14:15:23] <Marin> It works[14:15:45] <Marin> I used this formula yPos+=yVelocity*sin((yAngle/180)*PI);[14:16:08] <Marin> But It moves like thruster is on other part of lunar lander[14:18:12] <Marin> http://imagebin.ca/view/Z_zYV05.html[14:23:23] *** johndoe has quit IRC[14:26:14] *** sohail has joined ##OpenGL[14:27:06] *** johndoe has joined ##opengl[14:28:19] *** Wagoo has joined ##opengl[14:29:20] *** XT95_ has quit IRC[14:31:00] *** oogaw has quit IRC[14:37:58] *** elite01 has quit IRC[14:39:33] *** KU0N has joined ##opengl[14:43:34] *** pietia has quit IRC[14:48:45] *** pietia has joined ##OpenGL[14:51:34] *** elite01 has joined ##opengl[15:01:27] *** kbotnen has joined ##OpenGL[15:05:15] <Jupp3> Marin: So it's a simple matter of += -> -= or?[15:08:40] *** wey has quit IRC[15:09:10] <hdon> Marin: stop posting bitmaps![15:09:33] <hdon> Marin: and good for you making your thing move[15:13:44] *** pietia has quit IRC[15:15:34] *** pietia has joined ##OpenGL[15:21:12] *** itewsh has joined ##OpenGL[15:23:24] *** mm765^sleep is now known as mm765[15:24:59] *** qeed has joined ##opengl[15:25:55] *** gusnan has joined ##OpenGL[15:30:04] *** montana has joined ##OpenGL[15:31:42] *** Orphis has quit IRC[15:32:15] *** Orphis has joined ##opengl[15:32:39] *** AlastairLynn has joined ##opengl[15:34:17] *** Marin__ has joined ##OpenGL[15:35:29] *** Marin has quit IRC[15:40:35] *** KU0N has quit IRC[15:41:30] *** d3mn0id has joined ##OpenGL[15:45:04] <d3mn0id> What's the most standardized way to get an OpenGL window open on Linux?[15:45:31] <Jupp3> glx?[15:45:47] <Jupp3> many use sdl[15:46:03] <Jupp3> which can do way more than open the window, and it's way more portable too[15:46:10] <d3mn0id> GLX is outdated, apparently.[15:46:45] <AlastairLynn> the whole X window system is outdated imho but everybody still uses it and glx is the official way to do OpenGL with the X window system, SDL only runs on top of it[15:46:59] *** phenom9600 has joined ##OpenGL[15:47:07] <Jupp3> not really[15:47:13] <phenom9600> anyone who can give some hints on how to make a chatterbot for GNOME with 3D animations?[15:47:23] <Jupp3> it runs on top of whatever host system uses[15:47:32] <Jupp3> but yes, glx on linux[15:47:36] <AlastairLynn> which, on linux, is typically glx[15:47:45] <HuntsMan> d3mn0id: there's only one way, GLX[15:47:49] <AlastairLynn> yes, linux was all he asked about :)[15:48:08] <Jupp3> but that's not "the whole truth" on sdl[15:48:09] <HuntsMan> d3mn0id: and i which way would be GLX outdated? :)[15:48:26] <Jupp3> and you should NEVER count on havinh glx, ig you use sdl[15:48:46] <Jupp3> HuntsMan: It's ready, and doesn't need many updates anymore? :)[15:48:57] *** UUncia has quit IRC[15:49:07] *** UUncia has joined ##OpenGL[15:49:08] <d3mn0id> No, X still has life in it and I don't know HuntsMan.[15:49:14] *** UUncia has quit IRC[15:49:18] <HuntsMan> Jupp3: well it was updated to create GL3 context via extensions :P[15:49:30] <Jupp3> true[15:50:30] <d3mn0id> Jupp3, have you never heard of package managers like Synaptic (which is a wrapper for apt)[15:51:46] <Jupp3> d3mn0id: Sure, why?[15:52:47] <Jupp3> d3mn0id: Are you now trying to prove it's not "outdated" after all? :)[15:54:01] <d3mn0id> They can take care of dependencies and I have no interest in cross-platform compatibility.[15:54:09] <Jupp3> With updates I meant mostly "new features"[15:54:35] <Jupp3> And I guess glx is rather "stabilized" by now[15:54:45] <d3mn0id> So, GLX is actually fine[15:54:58] <Jupp3> Becouse it's your only option? :)[15:54:58] <d3mn0id> SDL is also stable.[15:55:06] <Jupp3> Depends[15:55:12] <Jupp3> There are big changes happening[15:55:15] <d3mn0id> mostly.[15:55:28] <Jupp3> Which will break quite a bit of backwards compatibility when the next big release is out[15:56:20] <d3mn0id> Exactly, GLX is safe for learning on.[15:56:38] <Jupp3> afaik, there are way more planned changes for SDL than for GLX[15:56:41] <d3mn0id> Then the new SDL can be done later.[15:56:47] <Jupp3> Well SDL does way more things so... :)[15:57:08] *** jimtendo has quit IRC[16:00:14] *** NorthStar has joined ##OpenGL[16:02:00] *** sparky has joined ##OpenGL[16:03:51] *** Orphis has quit IRC[16:03:58] *** Orphis has joined ##opengl[16:05:06] <sparky> mornin fools[16:06:20] <d3mn0id> hello[16:11:08] *** AlastairLynn has quit IRC[16:13:03] *** karsten_ has joined ##OpenGL[16:16:09] *** d3mn0id has left ##OpenGL[16:17:51] *** m4ggus has joined ##opengl[16:19:25] <sparky> :)[16:19:44] <sparky> karsten: din dansker![16:22:36] *** montana has quit IRC[16:22:46] *** itewsh has quit IRC[16:27:09] *** dv_ has joined ##opengl[16:32:29] *** emzic has joined ##OpenGL[16:34:19] <fusi> OMFG[16:34:23] <fusi> i found the problem[16:34:26] <fusi> jesus christ[16:34:36] <fusi> all the way[16:34:44] <fusi> right back into the guts of my math lib[16:35:09] <fusi> a bloody debug float member in the vec2 offsetting the size[16:35:10] <fusi> GAH![16:36:06] <hdon> fusi: you should always assert those things ;)[16:36:22] *** calav3ra has joined ##opengl[16:36:46] <fusi> indeed[16:36:49] <Rolenun> fusi: You should tell em hindsight is 20/20 and backseat programmers are many. :D[16:36:55] <fusi> heh[16:41:25] *** dv_ has quit IRC[16:44:05] <emzic> hi guys, i have a question about the GLSL shadow2D function[16:44:11] <Madsy> fusi: Hah. I told you about proper stride before.[16:44:17] <Madsy> ;)[16:44:20] <emzic> how do i get the texture coordinates that i put in the second parameter?[16:44:57] <Madsy> emzic: What?[16:45:10] <Madsy> How should we know what second parameter you're talking about?[16:45:31] <emzic> shadow2D takes two parameters[16:45:47] <emzic> the first is the texture (sampler2Dshadow), and the second is a vec4[16:45:59] <Madsy> Yep.[16:46:00] <emzic> but where do i get that vec4?[16:46:03] <Madsy> s,t,r,q[16:46:13] <Madsy> From your texture coordinates.[16:46:20] <Madsy> Pass them as varying to your fragment shader.[16:46:41] <emzic> first of all thanks for your help[16:46:58] <emzic> but i am afraid i dont fully understand. where do i get these texture coordinaes from in the first place?[16:47:02] <emzic> glTexGen ?[16:47:41] <Madsy> glTexGen or your own data passed via glTexCoordPointer or a custom vertex attribute.[16:48:11] *** itewsh has joined ##OpenGL[16:48:21] <Madsy> I have no idea what you use the shadow texture for, so it's hard to make any assuption what the "right" source would be.[16:48:33] <emzic> ok[16:48:45] <Madsy> We're not mind readers.[16:48:49] *** m4ggus has quit IRC[16:48:55] <Madsy> Some people have a hard time understanding that..[16:49:10] <Madsy> Not refering to you in that case.[16:49:20] <emzic> :)[16:52:28] *** johndoe has quit IRC[16:53:25] <emzic> what is actually the difference between shadow2D and shadow2DProj ?[16:53:43] <MatthiasM> emzic: did you read the description ?[16:54:20] <emzic> i have been reading the GLSL spec pdf from opengl.org[16:54:37] <emzic> but i didnt understand everything so was asking here :)[16:55:34] <MatthiasM> then read it again[16:55:43] <calav3ra> hi[16:56:21] <calav3ra> my question is not truely a opengl question - i want to create rectangular 3d planettextures without distortion, like here http://www.oera.net/How2/PlanetTexs/EarthMap_2500x1250.jpg[16:56:47] *** itewsh has quit IRC[16:56:52] <MatthiasM> calav3ra: well - this jpg is already distorted[16:57:18] <calav3ra> yes, but mine not :)[16:57:23] <calav3ra> i basically have such textures available, but without polarmapping already. now i ask myself what would be the best way to map thetexture[16:57:44] <calav3ra> either dynamically with a shader or prepare them before[16:57:47] <MatthiasM> as a rectangular texture doesn't fit onto a sphere ...[16:58:07] <calav3ra> it does, a sphere is twice as wide as high[16:58:20] <MatthiasM> not the pole caps[16:58:38] <MatthiasM> the area over the caps is much smaller then over the equator[16:58:40] <calav3ra> yes, thats the point[16:58:42] <calav3ra> i know[16:58:49] <calav3ra> thats what i want to solve[16:58:51] <calav3ra> :)[16:58:57] <MatthiasM> and your texture must be created to account for that[16:59:13] <MatthiasM> otherwise you gets artifacts[16:59:24] <emzic> it says that the sampler that is used in the shadow2D functions should be a depth-texture. is it ok if it is a framebufferobject with a depth component?[16:59:57] <MatthiasM> emzic: no - you can use a framebuffer as texture - but you can attachj a depth texture as depth attachment to an FBO[17:00:12] <MatthiasM> you can't use a framebuffer ...[17:00:28] <emzic> oh[17:00:39] <emzic> but i can use the depth component of the framebuffer?[17:00:56] <MatthiasM> a framebuffer has no depth component - only a depth attachment[17:01:47] <emzic> sorry, i thought so, because this is how i attach it: glRenderbufferStorageEXT(GL_RENDERBUFFER_EXT, GL_DEPTH_COMPONENT, width, height);[17:02:24] <MatthiasM> that's a renderbuffer - you can't use this as a texture[17:02:30] *** bijoo_appdev has joined ##OpenGL[17:02:30] *** bijoo_osdev has joined ##opengl[17:02:37] <MatthiasM> you need a texture[17:02:37] <emzic> but can i use it for the sampler2D function?[17:02:41] <MatthiasM> no[17:02:45] <emzic> ok[17:02:57] <MatthiasM> just read the spec or google for examples[17:03:02] <HuntsMan> sampler2D is a type, not a function[17:03:27] <emzic> sorry, i meant shadow2D[17:03:42] <HuntsMan> that's a type also :P[17:04:01] *** Jernej has quit IRC[17:04:03] <HuntsMan> bah, no, sampler2DShadow would be the type[17:04:10] *** JernejL has joined ##OpenGL[17:04:15] <emzic> really? cause it says in the spec: vec4 shadow2D (sampler2DShadow sampler, vec3 coord [, float bias] )[17:04:52] *** NightVisio has joined ##OpenGL[17:04:54] <HuntsMan> yes, my mistake :P[17:04:56] *** amz has joined ##opengl[17:04:59] <emzic> :)[17:05:28] <emzic> ok, so just to make sure i understood. i can not use a FBO as a sampler2DShadow ? right?[17:05:44] *** LordMetroid has joined ##OpenGL[17:05:52] *** AlastairLynn has joined ##opengl[17:06:09] <HuntsMan> not an FBO, a texture bound to the FBO[17:06:31] <MatthiasM> emzic: how hard is it to understand ?!? you can ONLY use textures for samplers ![17:06:52] <emzic> that would be something, that i call the glBindRenderbufferEXT with?[17:07:11] <MatthiasM> emzic: A TEXTURE !!!!!!![17:07:30] <emzic> MatthiasM: sorry, i am just learning this stuff... =)[17:07:31] <MatthiasM> a renderbuffer is not a texture[17:07:35] <emzic> ok[17:07:37] <MatthiasM> no myou don't[17:10:36] *** NightVisio has quit IRC[17:13:16] <hdon> can i texture map with a render buffer object?[17:13:34] *** NightVisio has joined ##OpenGL[17:13:45] <hdon> err, frame buffer object[17:13:58] <MatthiasM> hdon: NO[17:14:39] *** wey has joined ##opengl[17:14:42] <MatthiasM> why is it so hard to read the spec - it even has examples[17:15:11] <hdon> MatthiasM: they aren't always the first thing to show up on google :([17:15:20] <MatthiasM> then learn to google[17:15:52] <hdon> and generally, the docs for OpenGL extensions cite all kinds of other OpenGL extensions, so they're very hard to read if you haven't used any extensions before[17:16:13] <hdon> in my limited experience[17:16:35] <HuntsMan> then start using them[17:17:18] <hdon> as a criticism, that's a logical fallacy :([17:17:27] <hdon> but i am trying :)[17:17:48] <MatthiasM> hdon: start with an easy extension and read them[17:17:57] <MatthiasM> like ARB_vertex_buffer_object[17:20:11] <hdon> also not having man pages for all the extension' functions is pretty annoying[17:23:43] <emzic> can i attach a depth-texture to a FBO?[17:23:59] <MatthiasM> emzic: I already said yes[17:24:09] <emzic> ok thanks :)[17:24:16] <MatthiasM> emzic: you really need to learn to read[17:25:16] <hdon> MatthiasM: are buffer objects the most common things people ask about?[17:26:49] <MatthiasM> hdon: no - most people ask the thing that another one just gets answered[17:26:58] <hdon> heh[17:27:22] <hdon> i've been working for the past few hours trying to absorb this information[17:27:41] <hdon> i want to implement a cellular automata with graphics hardware acceleration sans fragment shaders[17:28:29] <HuntsMan> hdon: seems you need something like CUDA or OpenCL[17:28:40] <hdon> i had an inkling to do it with stencil tests and a glTexSubImage2D() feedback loop, but that function seems slow[17:29:00] <hdon> HuntsMan: care to elaborate?[17:29:15] * hdon googles[17:29:31] <HuntsMan> http://en.wikipedia.org/wiki/CUDA[17:29:34] <hdon> wow the cuda page is flash. lame.[17:29:49] <HuntsMan> CUDA is the compute engine in NVIDIA graphics processing units or GPUs, that is accessible to software developers through industry standard programming languages.[17:30:39] <hdon> contrary to what MatthiasM may tell you, i can read ;)[17:31:01] <HuntsMan> :B[17:31:37] <MatthiasM> hdon: well - that was not targeted at you[17:31:39] <MatthiasM> (yet)[17:32:54] <hdon> my GPU is significantly older than these technologies, i'm likely not compatible[17:33:14] <AlastairLynn> what kind of speed increase do you get from the use of VAOs?[17:33:26] <hdon> anyhow it sounds like those technologies wouldn't necessarily enable me to display my data as quickly as i could process it ;)[17:33:41] <HuntsMan> hdon: buy a new GPU :)[17:33:54] *** mor_ has joined ##OpenGL[17:33:56] <hdon> HuntsMan: looking for a job at the moment :O[17:35:08] <MatthiasM> AlastairLynn: depends on your app and how you use them and how the driver is written[17:35:25] <hdon> hm, my card doesn't seem to support framebuffer objects either[17:35:28] <MatthiasM> AlastairLynn: what you save is a few function calls - depending on how many VBOs and arrays you need[17:35:32] <HuntsMan> and if your app is CPU bound[17:35:38] <HuntsMan> hdon: what card and OS?[17:35:52] <RTFM_FTW> the difference between shadow2D and shadow2DProj would be that the latter does a projective divide[17:36:01] <RTFM_FTW> err prospective divide[17:36:08] <AlastairLynn> perspective[17:36:10] <AlastairLynn> ?[17:36:13] <hdon> HuntsMan: old dual-head matrox (GeForce4 MX 4000) debian 4.0[17:36:14] <MatthiasM> which is clearly stated in the spec[17:36:24] <AlastairLynn> or prospective?[17:36:26] <MatthiasM> hdon: forget it[17:36:34] *** ol1veira__ has joined ##OpenGL[17:36:37] <hdon> MatthiasM: which part?[17:36:48] <MatthiasM> hdon: a) matrix != geforce[17:36:59] <MatthiasM> b) with both you can't use Cuda or OpenCL[17:37:05] <MatthiasM> *matrox[17:37:14] <hdon> oh, i guess this is a real nvidia card[17:37:20] <hdon> maybe the matrox was the older card[17:37:29] <hdon> 01:00.0 VGA compatible controller: nVidia Corporation NV18 [GeForce4 MX 4000 AGP 8x] (rev c1)[17:37:47] <MatthiasM> if you get any kind of OpenGL acceleration out of a Matrox card then you are more then lucky[17:37:48] <RTFM_FTW> haha[17:38:19] <MatthiasM> hdon: and shader support on a Geforce4 is also very limited[17:39:10] <hdon> MatthiasM: that's why i'm not using shaders for this :)[17:39:19] <RTFM_FTW> in any case all of the *Proj lookups will do the 1 / Q (actually W) step in the shader...[17:42:05] *** NorthStar has quit IRC[17:46:34] <fusi> yay[17:46:42] <fusi> finally something onscreen[17:46:54] <fusi> http://91.135.13.31/yay.jpg \o/[17:48:23] <hdon> fusi: how many triangles?[17:48:40] <fusi> dunno not counting ^^[17:48:44] *** ol1veira_ has quit IRC[17:48:48] * hdon counts[17:48:53] <fusi> i will[17:49:04] <fusi> i just havent written it yet[17:52:26] <MatthiasM> fusi: cool - what approch do you use ?[17:52:44] <MatthiasM> and how do you texture it ?[17:52:58] *** emzic has quit IRC[17:53:07] *** sedeki has joined ##OpenGL[17:54:07] *** XT95_ has joined ##OpenGL[17:55:42] <fusi> http://uk.youtube.com/watch?v=Jd3-eiid-Uw[17:55:56] <fusi> MatthiasM: just gmm with splatting[17:56:21] <fusi> im building it to work with huge terrain[17:56:28] <fusi> havent got that far yet[17:57:13] <MatthiasM> and which algorithm do you use for gmm ?[17:59:05] <fusi> i used to stitch but now i skirt[17:59:44] <fusi> faster and simpler[17:59:56] <fusi> and the actual pixel error is miniscule[18:00:09] <MatthiasM> hmm - I use BTT :)[18:00:16] <fusi> btt?[18:00:23] <fusi> im not hot on acronyms[18:00:32] <MatthiasM> that doesn't require stitching or skirts[18:00:38] <MatthiasM> binary triangle trees[18:00:40] *** GX has joined ##OpenGL[18:00:41] *** NorthStar has joined ##OpenGL[18:00:51] <fusi> orly[18:00:58] <fusi> is that like roam[18:01:10] <MatthiasM> the only requirement is that neighbour tiles need to be +-1 LOD away[18:01:16] <MatthiasM> no - tile based[18:01:27] <fusi> got a link to a paper?[18:01:33] <MatthiasM> I use 32x32 tiles[18:01:54] <fusi> ah nvm google provides[18:02:04] <MatthiasM> :)[18:02:46] <MatthiasM> the generation takes some time - esp if the vertex cache optimisations are applied[18:02:46] <fusi> how much runtime processing is involved[18:02:53] <MatthiasM> at rendering none[18:03:00] <fusi> cool[18:03:36] <MatthiasM> www.matthiasmann.de/worldscape[18:03:57] <fusi> nice[18:04:20] *** XT95__ has joined ##OpenGL[18:04:49] <MatthiasM> http://www.matthiasmann.de/java/worldscape/ufo.jnlp <-- uses a 16k x 16k terrain[18:04:57] <tmccrary1> 16x16![18:05:11] <MatthiasM> 16385 x 16385[18:05:28] <fusi> :)[18:05:34] <fusi> you need some grass[18:05:37] <fusi> heh[18:06:03] <MatthiasM> this game needs a lot of thinks - but I'm working on aother projects currently[18:06:07] <MatthiasM> *things[18:07:08] <fusi> :) looks good[18:07:08] <tmccrary1> this london fog latte is kind of strange tasting[18:07:31] <fusi> london fog latte?[18:07:39] <fusi> sounds carcanogenic[18:07:55] <tmccrary1> I hope so[18:08:01] <tmccrary1> I like challenges[18:08:22] <fusi> lul[18:08:36] <tmccrary1> zhul[18:15:07] *** GuShH has quit IRC[18:19:30] *** XT95_ has quit IRC[18:32:19] *** qeed has quit IRC[18:33:59] *** elite01 has quit IRC[18:36:28] *** montana has joined ##OpenGL[18:47:38] *** predaeus has joined ##opengl[18:50:48] *** ljungk has joined ##OpenGL[19:00:43] *** cplusplus has joined ##OpenGL[19:00:48] *** Ademan has quit IRC[19:03:03] *** ol1veira__ has quit IRC[19:10:16] *** Quentarez has joined ##OpenGL[19:32:32] *** jfroy has quit IRC[19:34:20] *** chai has quit IRC[19:38:09] *** l3dx has joined ##OpenGL[19:38:30] *** Ademan has joined ##OpenGL[19:40:17] *** lewymati has joined ##OpenGL[19:40:44] *** mor_ has quit IRC[19:49:08] <montana> anyone want to help with some openGL work, ill pay... a little bit :)[19:55:24] <Ingenu> just ask[19:58:13] <montana> well im tottally confused[19:58:17] <montana> so prob need a lot of hands on work[19:58:22] <montana> anyone chat on msn?[19:58:24] <montana> or pm[19:58:37] *** jfroy has joined ##OpenGL[20:03:20] *** druggy_ has joined ##opengl[20:07:48] *** wey has quit IRC[20:10:38] *** NightVisio has quit IRC[20:13:43] <pietia> anyone familiar with blender ?[20:13:53] <pietia> why blender's exporter generates only 2 values in obj file ? f 6//12 7//12 3//12 ?[20:16:10] <tmccrary1> your vertices don't have texture coordinates[20:16:26] <tmccrary1> select the geometry, hit tab and then hit u[20:16:41] <tmccrary1> this will be you options for unwrapping the geometry[20:16:46] *** rnx has joined ##opengl[20:16:56] <tmccrary1> then you can change one of your views into UV mode and fine tune everything[20:18:46] <sparky> can someone here confirm that cosine squared (of the vertical angle) when integrated over the hemisphere is (2 * pi) / 3[20:19:43] *** druggy has quit IRC[20:20:25] *** gusnan has quit IRC[20:26:39] *** montana has quit IRC[20:27:43] *** montana has joined ##OpenGL[20:32:45] *** Ademan has quit IRC[20:40:35] <montana> hi[20:42:44] *** KU0N has joined ##opengl[20:43:28] <montana> anyone want to help with some openGL work, ill pay... a little bit :) pm me[20:43:35] <sparky> never mind, I got it[20:46:10] <Madsy> montana: Nagging doesn't help. And if you want to pay for work, go to rentacoder.com or something else.[20:47:03] <Madsy> We help people for free, given that they show the will to learn.[20:49:37] <sparky> madsy: I have a feeling he means develop som work[20:49:45] <sparky> not help him with his understanding of it[20:49:47] <sparky> but I may be wrong[20:50:22] <Madsy> sparky: ..[20:51:57] *** vade has joined ##OpenGL[20:54:57] *** predaeus has quit IRC[20:55:00] *** emzic has joined ##OpenGL[20:57:17] <Madsy> RTFM_FTW :)[20:59:38] <RTFM_FTW> hey Madsy[21:00:32] <RTFM_FTW> having fun beating down the unwashed masses? :P[21:00:39] <RTFM_FTW> heheh[21:00:40] <Madsy> As always.[21:03:46] <Madsy> RTFM_FTW: So, what's up?[21:04:33] <Madsy> Man I'm full. Probably ate a kg chicken and rice for dinner.[21:04:52] <Madsy> Hopefully drinking more beer will help[21:04:56] <montana> lol[21:04:57] <tmccrary1> I read that as a keg of chicken[21:05:00] <tmccrary1> I was like[21:05:02] <tmccrary1> damn[21:05:24] <montana> liquidised that chicken and suck it outta keg[21:05:28] <Madsy> tmccrary1: haha[21:05:32] <RTFM_FTW> heh not much... wondering what (if anything) I should do today :P[21:06:32] <Madsy> RTFM_FTW: You could waste some time and give me comments and suggestions for improvements on my fixedpoint class: http://www.mechcore.net/files/code/fixedpoint/fixed.hpp[21:06:37] <Madsy> Or you could take a wank.[21:06:41] <Madsy> :)[21:06:50] <RTFM_FTW> hahah[21:07:07] <Madsy> Here's how it is used: http://www.mechcore.net/files/code/fixedpoint/testcase.cpp[21:07:30] <Madsy> If you remove the runtime-dependancy (rand in this example), the objects are optimized away into constants.[21:07:40] <Madsy> That also applies for full expressions.[21:11:31] *** Tibor__ has joined ##OpenGL[21:21:48] *** johndoe has joined ##opengl[21:23:50] *** Spkka has joined ##OpenGL[21:27:35] *** Gaia has joined ##OpenGL[21:27:48] *** Gaia is now known as LoufoQ[21:37:51] *** kbotnen has quit IRC[21:41:49] *** Ingenu is now known as Sir_dor[21:41:57] *** Sir_dor is now known as Sir_d[21:45:05] *** Sir_d is now known as Ingenu[21:49:03] *** JoshDreamland has joined ##OpenGL[21:50:32] <JoshDreamland> I don't see anything about fonts in your topic[21:50:38] <JoshDreamland> which lib do you recommend?[21:50:39] <hdon> what is the behavior of glCopyTexImage2D() if you haven't called glBindTexture()?[21:51:04] *** iion has joined ##OpenGL[21:52:43] <hdon> i'm looking at a program that makes calls to functions like glTexParameteri() and glCopyTexImage2D() without ever generating texture handles or binding a texture -- is there some default texture handle?[21:53:02] *** XT95__ has quit IRC[22:04:58] *** plasticpool has joined ##OpenGL[22:05:25] *** plasticpool has quit IRC[22:05:36] *** plasticpool has joined ##OpenGL[22:06:49] *** Entelin has quit IRC[22:07:05] <RTFM_FTW> glCopyTex[Sub]Image will use whatever texture unit (and texture ID) which is currently bound and active[22:07:05] *** Quentarez has quit IRC[22:07:22] <RTFM_FTW> in the *default* case that would be TID = 0 and TEXTURE0[22:08:23] <RTFM_FTW> texture / vertex / buffer ID zero is discussed within the GL API specification[22:08:24] *** XT95_ has joined ##OpenGL[22:12:06] *** dv_ has joined ##opengl[22:15:28] *** JoshDreamland has quit IRC[22:24:39] <hdon> RTFM_FTW: i see![22:25:52] *** Eforen has joined ##opengl[22:34:07] *** pietia has quit IRC[22:34:38] *** pietia has joined ##OpenGL[22:42:53] *** wisey has joined ##OpenGL[22:52:23] *** groton_ has joined ##OpenGL[22:56:49] *** tmccrary has quit IRC[22:57:39] *** ljungk has quit IRC[22:57:54] *** iion has quit IRC[22:58:55] *** tmccrary has joined ##OpenGL[22:59:42] *** iion has joined ##OpenGL[23:01:55] *** groton has quit IRC[23:02:24] *** XT95_ has quit IRC[23:03:09] *** iion has quit IRC[23:06:53] *** XT95_ has joined ##OpenGL[23:10:54] *** tmccrary has quit IRC[23:12:32] *** NevroPus has quit IRC[23:12:54] *** NevroPus has joined ##OpenGL[23:18:49] *** Stevethe1irate has quit IRC[23:19:52] *** Stevethepirate has joined ##OpenGL[23:21:28] *** vade has quit IRC[23:24:24] *** groton_ has quit IRC[23:27:19] *** Renderwahn has quit IRC[23:31:35] *** GuShH has joined ##OpenGL[23:32:46] *** jcazevedo has quit IRC[23:35:23] *** e_roder has joined ##OpenGL[23:35:34] *** Madsy has quit IRC[23:35:34] <e_roder> hey[23:37:46] <GuShH> hoy[23:38:57] <e_roder> i'm having trouble with some matrix/perspective/something math[23:40:19] <e_roder> i'm working with a shader, trying to recreate the ffp i had in place before[23:40:56] <e_roder> when i translate a triangle away from the origin and then rotate it, it seems to be rotating in a circle, but not around the origin[23:41:36] *** phenom9600 is now known as kaputt[23:41:54] <e_roder> switching the order of the multiplication seems to have no effect either[23:43:03] *** XT95_ has quit IRC