January 8, 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 8, 2009 [00:01:09] *** ol1veira has quit IRC[00:03:18] *** mikolalysenko has joined ##OpenGL[00:03:39] *** scai has left ##opengl[00:04:08] *** phrosty has quit IRC[00:05:57] *** Bollinger has quit IRC[00:06:18] *** oliveira has joined ##OpenGL[00:11:52] *** Walt has joined ##opengl[00:23:54] *** Amorphous has quit IRC[00:26:18] *** amz has joined ##opengl[00:30:43] *** vade has joined ##OpenGL[00:34:55] *** rnx has left ##opengl[00:38:30] *** phrosty has joined ##opengl[00:41:12] *** Amorphous has joined ##opengl[00:41:40] *** neoneye has quit IRC[00:46:47] *** Aintaer has quit IRC[00:51:32] *** pietia has quit IRC[00:52:05] *** AlastairLynn has quit IRC[00:57:19] *** x0rx0r has quit IRC[01:07:44] *** x0rx0r has joined ##OpenGL[01:09:07] *** asdfa has quit IRC[01:10:52] *** Suprano has quit IRC[01:11:18] *** bijoo_appdev has quit IRC[01:11:18] *** bijoo_osdev has quit IRC[01:12:44] *** bijoo_osdev has joined ##opengl[01:12:44] *** bijoo_appdev has joined ##OpenGL[01:14:17] *** bijoo_appdev_ has joined ##OpenGL[01:14:17] *** bijoo_osdev_ has joined ##opengl[01:16:14] *** bijoo_osdev has quit IRC[01:16:39] *** m4ggus has quit IRC[01:16:59] *** bijoo_appdev has quit IRC[01:19:17] *** m4ggus has joined ##opengl[01:21:53] *** korff_home has quit IRC[01:25:35] *** mikolalysenko has quit IRC[01:28:02] *** NinZine has quit IRC[01:28:47] *** jfroy|work has joined ##OpenGL[01:30:53] *** sparky_ has joined ##OpenGL[01:31:27] *** sparky has quit IRC[01:40:25] *** sparky has joined ##OpenGL[01:40:39] *** sparky_ has quit IRC[01:44:14] *** iion has quit IRC[01:48:08] *** emzic has quit IRC[02:05:16] *** tokemonstah has quit IRC[02:17:31] *** Dagyde_ has left ##opengl[02:18:11] *** Dagyde has joined ##OpenGL[02:21:47] *** kbotnen has joined ##OpenGL[02:22:59] *** amz has quit IRC[02:25:06] *** Xmas| has quit IRC[02:29:07] *** Entelin has quit IRC[02:32:53] *** dv_ has quit IRC[02:38:50] *** dvoid_ has quit IRC[02:43:45] *** vade has quit IRC[02:48:56] *** kbotnen_ has joined ##OpenGL[02:49:18] *** greg_ has joined ##OpenGL[02:49:47] *** HuntsMan has joined ##opengl[02:50:52] *** zstars has quit IRC[02:52:00] *** Dagyde has left ##OpenGL[02:56:01] *** iion has joined ##OpenGL[02:56:52] *** cplusplus2 has quit IRC[02:58:26] *** acetoxy has quit IRC[02:58:59] *** acetoxy has joined ##opengl[03:01:46] *** x0rx0r has quit IRC[03:03:24] *** Kasu- has quit IRC[03:08:14] *** kbotnen has quit IRC[03:09:46] *** twist has quit IRC[03:14:21] *** johndoe has quit IRC[03:15:01] *** Spkka has quit IRC[03:29:17] *** vade has joined ##OpenGL[03:37:30] *** MatthiasM has quit IRC[03:37:36] *** MatthiasM has joined ##opengl[03:41:03] *** jfroy|work has quit IRC[03:54:52] *** forrestv has quit IRC[03:55:33] *** b0000 has quit IRC[03:56:58] *** grindhold has joined ##OpenGL[04:02:51] *** jfroy has joined ##OpenGL[04:04:36] *** reprore_ has joined ##OpenGL[04:06:47] *** forrestv has joined ##OpenGL[04:09:49] *** GuShH_ has joined ##OpenGL[04:10:18] *** reprore_ has quit IRC[04:10:45] *** essoe- has joined ##opengl[04:12:42] *** calav3ra_ has quit IRC[04:13:47] *** Shel2476 has joined ##OpenGL[04:13:55] *** Shel2476 has left ##OpenGL[04:15:34] *** GuShH has quit IRC[04:25:55] *** mattn2|home_ has quit IRC[04:26:37] *** mattn2|home_ has joined ##OpenGL[04:33:45] *** amz has joined ##opengl[04:34:26] *** bijoo_appdev_ has quit IRC[04:34:26] *** bijoo_osdev_ has quit IRC[04:40:00] *** iion has quit IRC[04:44:17] *** _Rangar_ has quit IRC[04:51:04] *** ridley_ has joined ##OpenGL[04:52:09] *** ridley_ has left ##OpenGL[04:54:52] *** signalsea has joined ##OpenGL[05:05:23] *** kbotnen_ has quit IRC[05:05:42] *** kbotnen_ has joined ##OpenGL[05:13:38] *** qeed has quit IRC[05:21:06] <signalsea> Can anyone tell me why I am always getting hits in this section of code?: http://pastebin.com/d1261517a (see function pressMouse)[05:23:13] <signalsea> That is, when I click anywhere in the window, not just on the squares drawn.[05:31:57] *** djork has quit IRC[05:33:12] *** djork has joined ##opengl[05:39:31] *** HuntsMan has quit IRC[05:43:06] *** HuntsMan has joined ##opengl[05:45:04] *** djork has quit IRC[05:59:29] <sparky> evenin[06:01:10] *** mikolalysenko has joined ##OpenGL[06:01:41] <kbotnen_> evening.[06:01:58] <boghog> hello[06:19:16] <Madsy> morning[06:22:14] *** tamlin has joined ##OpenGL[06:23:51] <kbotnen_> how can I draw a border around my window? using ortho mode? Want a 2D border, or a bitmap of a frame around my 3Dperspective. I have managed to set and leave ortho / perspective in correct manner, so I only need to figure out what coordinates I should use for the ortho draw.[06:24:45] <kbotnen_> I am currently drawing a quad with texture in orthomode. I only need to make it fit (right coords for the vertex2f() calls).[06:25:24] <HuntsMan> you want a border in the native window?[06:25:34] <tamlin> If the border is to be pixel-precise, use 0.5 offsets for x/y.[06:26:09] <kbotnen_> I want a border surrounding my "app". My display is 640x480, and thats also the size of my viewport.[06:26:49] *** LiraNuna_ has joined ##OpenGL[06:26:51] <HuntsMan> kbotnen_: well you can't draw with OpenGL outside of your window, if that's what you mean[06:26:52] <kbotnen_> The border is just a "test". if it works I can use same methods to draw onscreen 2D menus later if needed.[06:27:01] <kbotnen_> inside my window.[06:27:05] <kbotnen_> ofc :)[06:30:32] <kbotnen_> 0,0 is lower left right?[06:30:57] <kbotnen_> and the first that has to be drawn?[06:31:07] <tamlin> Yes, and no.[06:31:43] *** jedir0x has joined ##OpenGL[06:32:05] <jedir0x> Is there a VerticalSync extension to OpenGL similar to wglSwapInterval for GLX ?[06:32:42] <jedir0x> if so, i'd really appreciate some pointing in the right direction, i'ved grepped glext.h and found nothing[06:32:59] <jedir0x> i did find this though: http://www-f9.ijs.si/~matevz/docs/007-2392-003/sgi_html/ch11.html[06:33:15] <jedir0x> but that doesn't look right for some reason[06:33:36] <HuntsMan> yes, but i can't recall the name[06:34:47] *** kbotnen_ has quit IRC[06:34:48] <HuntsMan> you should be looking into glxext.h BTW[06:35:17] *** LiraNuna_ has quit IRC[06:35:20] <HuntsMan> GLX_SGI_swap_control[06:35:27] <jedir0x> sweet, thanks[06:35:28] <jedir0x> yeah[06:36:01] <tamlin> Hint: http://www.opengl.org/registry/[06:36:33] *** LiraNuna_ has joined ##OpenGL[06:36:42] <signalsea> Can anyone help me with my code? Not sure if no one was around earlier[06:36:49] *** m4ggus has quit IRC[06:37:17] *** Tenac has quit IRC[06:37:19] <tamlin> signalsea: What's the problem?[06:37:22] <signalsea> Trying to get picking to work (the code is at http://pastebin.com/d1261517a)[06:37:52] <signalsea> I always get a hit, no matter where I click[06:38:03] <signalsea> that's not the full code, just the drawing and mouse portions[06:38:42] <signalsea> and a bunch of global variables[06:39:37] <tamlin> Is 'colorstate' a vector of pointers?[06:39:54] <signalsea> that shouldn't matter it's just part of a color switch[06:40:52] <signalsea> though yes, it is[06:41:16] <tamlin> OK. What do you get in hitnums? Always 1?[06:41:21] <signalsea> yes[06:42:18] *** LiraNuna has quit IRC[06:42:55] <jedir0x> thanks for the help guys[06:42:59] <jedir0x> later[06:43:00] *** jedir0x has quit IRC[06:43:46] <signalsea> i feel like maybe I am doing something wrong with gluPickMatrix[06:44:08] <signalsea> but I am trying to get by without totally understanding projection matrices[06:45:15] <signalsea> oh[06:45:25] <signalsea> that's probably exactly my problem.[06:45:26] <tamlin> Well, PICK_TOL must be width and height of the picking region. Is it?[06:45:41] <signalsea> yeah, it's 3[06:45:45] <signalsea> that's in pixels, right?[06:46:20] <tamlin> Close enough.[06:47:06] *** essoe- has quit IRC[06:47:11] <signalsea> Would i be correct in thinking that if the projection matrix is set up to render the whole thing, that it will always give a hit?[06:47:22] *** LiraNuna_ has quit IRC[06:47:23] <signalsea> the whole thing being not just the part around the cursor[06:47:31] *** LiraNuna has joined ##OpenGL[06:48:06] <tamlin> What do you think? Does that make sense or not?[06:48:15] <tamlin> ;-)[06:49:02] <tamlin> If you have something filling the whole viewport, would it make sense for any x/y coord picking to always return something?[06:51:06] <signalsea> I will tinker some more :)[06:51:37] <signalsea> I didn't quite understand how gluPickMatrix was deciding whether or not to elicit a hit[06:51:48] <signalsea> until i read the doc just now[06:51:52] <signalsea> "To effectively use the generated pick matrix for picking, first call glLoadIdentity to load an identity matrix onto the perspective matrix stack. Then call gluPickMatrix, and finally, call a command (such as gluPerspective) to multiply the perspective matrix bythe pick matrix."[06:51:56] <signalsea> :)[06:52:15] *** mikolalysenko has quit IRC[06:52:27] <signalsea> thanks regardless[06:52:47] <tamlin> Read doc's first. Ask questions later. You're welcome.[06:53:29] <signalsea> yes yes, i know... I had overlooked that one[06:54:46] <tamlin> No worries. Happy hacking.[07:03:07] *** UUncia has joined ##OpenGL[07:27:22] *** amz has quit IRC[07:41:23] *** servus_ has joined ##opengl[07:41:23] *** servus has quit IRC[07:42:08] *** rnx has joined ##opengl[08:07:36] *** korff_home has joined ##OpenGL[08:09:01] *** phrosty has quit IRC[08:12:44] *** fargiola` has joined ##OpenGL[08:13:51] *** tamlin has left ##OpenGL[08:13:52] *** godling has joined ##opengl[08:14:51] *** fargiolas has quit IRC[08:17:32] *** kenws has joined ##OpenGL[08:18:28] *** Tenac has joined ##opengl[08:20:26] <Tenac> Is the only way to get perfect per-pixel object selection by rendering the entire scene twice, once with each mesh having a unique color, checking what color the mouse is over, and then re-rendering everything with textures?[08:22:38] <mlucassmith> no[08:23:01] <Tenac> Or I guess there is ray/triangle collision from shooting a ray from the camera out to the scene, but then I'd need to limit what triangles to test for collision or else that would be slow[08:23:23] <mlucassmith> instead, you can read from the z-buffer, then apply the window x, y with the z from the z-buffer at that x, y, then put that 3d coordinate through your various matrix transforms and do a bounding volume check to see if the point is in your object[08:23:43] <mlucassmith> that way you're only testing on your geometry, not on triangles or ray casting or rendering twice[08:24:50] *** groton has joined ##OpenGL[08:25:09] <Tenac> a "bounding volume check" would only be a bounding sphere/box, right?[08:25:20] <mlucassmith> depends on what geometry objects you've got[08:25:32] <Tenac> Weird shaped triangle meshes[08:25:42] <mlucassmith> something defined those right?[08:25:54] <mlucassmith> like, a set of bezier patches?[08:26:20] <Tenac> No, just imported meshes and such[08:26:34] <mlucassmith> oof[08:26:40] <mlucassmith> that's kind of low level :)[08:26:48] <Tenac> =\[08:27:11] <mlucassmith> the approach i described means you can do forces too, since you know -where- on the object they clicked[08:28:00] *** fargiola` is now known as fargiolas[08:28:21] <Tenac> I have the code to find the 3D location of where the mouse is over the mesh[08:28:45] <mlucassmith> Anyway, the first approach you described works well (from experience)[08:28:45] <Tenac> But I don't have any matrix math set up and don't know too much about it[08:29:02] <mlucassmith> well.. you can use gluUnproject if you really need to[08:29:48] <Tenac> I'm thinking of going with this, but I think it means 2x the rendering time: http://gpwiki.org/index.php/OpenGL_Selection_Using_Unique_Color_IDs[08:29:58] <mlucassmith> Yeah I could tell that from your description :)[08:30:32] <mlucassmith> you can do it in one pass btw, just build your own framebuffer and render on to the screen and your separate buffer in your fragment shader[08:30:34] *** pwned has quit IRC[08:30:50] *** pwned has joined ##opengl[08:31:18] <mlucassmith> Then you can copy only a bit of the renderbuffer back that is near where the mouse cursor is to see what's there, should be very fast[08:34:47] <Tenac> I don't know much about framebuffers, so I think I'm just going to attempt ray/bounding box collision[08:54:23] *** groton has quit IRC[08:55:20] *** tokemonstah has joined ##OpenGL[08:56:27] *** Bollinger has joined ##OpenGL[08:56:51] *** sparky_ has joined ##OpenGL[08:57:00] *** jfroy has quit IRC[09:00:10] *** phrosty has joined ##opengl[09:00:51] *** phrosty has quit IRC[09:00:57] *** phrosty has joined ##opengl[09:05:14] *** neoneye has joined ##OpenGL[09:08:18] *** sparky has quit IRC[09:08:19] *** geocalc has quit IRC[09:11:45] *** [AD]Turbo has joined ##OpenGL[09:18:12] <[AD]Turbo> yo[09:20:26] *** grindhold has quit IRC[09:25:50] *** tokemonstah has quit IRC[09:34:53] *** maxton has joined ##OpenGL[09:42:29] *** sohail has quit IRC[09:52:23] *** mm765^away is now known as mm765[09:53:36] *** phrosty has quit IRC[09:53:58] *** LtJax has joined ##OpenGL[09:56:50] *** scai has joined ##opengl[09:59:02] *** godling has quit IRC[10:00:29] *** GuShH_ has quit IRC[10:02:21] *** phrosty has joined ##opengl[10:02:34] *** eXtronuS has quit IRC[10:03:13] *** eXtronuS has joined ##OpenGL[10:04:41] *** predaeus has joined ##opengl[10:08:04] *** groton has joined ##OpenGL[10:12:52] *** groton_ has joined ##OpenGL[10:15:23] *** Nescafe has joined ##OpenGL[10:16:28] *** sparky_ has quit IRC[10:16:53] *** Ingenu has joined ##OpenGL[10:18:19] *** groton has quit IRC[10:18:22] *** sparky has joined ##OpenGL[10:18:51] *** groton has joined ##OpenGL[10:19:22] *** Yustme has joined ##OpenGL[10:21:14] *** dvoid_ has joined ##OpenGL[10:23:37] *** eXtronuS has quit IRC[10:24:42] *** groton_ has quit IRC[10:38:43] *** ewn has joined ##OpenGL[10:42:38] *** doub has joined ##opengl[10:45:25] *** dusted has joined ##OpenGL[10:48:26] *** electrostatic has joined ##OpenGL[10:48:43] <pa> mh.. suppose i have a rotation of angle n around axis a. how could i extract a sequence of 3 rotations around x, y and z which reflects in the initial rotation?[10:49:26] <electrostatic> I've asked this a couple of times before, but here it goes again... Does anyone know a definitive, standardized way of ensure that an asynchronous PBO read has completed?[10:50:20] <MatthiasM> electrostatic: what do you mean ?[10:50:49] <MatthiasM> pa: search for "matrix and quaternion FAQ"[10:51:38] <LtJax> electrostatic: I think that's one of the extensions in GL3, but before that, you are out of luck[10:52:45] <electrostatic> MatthiasM: When you start a readPixels in a PBO, it starts reading the pixels in a different thread, and apparently has no way of blocking or even checking for completion...[10:52:58] <electrostatic> LtJax: That is seriously messed up[10:53:06] <electrostatic> How could the not notice that?[10:53:07] <LtJax> why? it's automatically blocking[10:53:15] <MatthiasM> electrostatic: when you map the buffer it is done[10:53:24] <LtJax> exactly.[10:53:39] <MatthiasM> glMapBuffer will block until all gl operations on that buffer are finished[10:53:56] <MatthiasM> but this is stated clearly in the spec ![10:54:00] <electrostatic> MatthiasM: I must have a bug in my drive r then, because it isn't[10:54:14] <MatthiasM> paste code[10:55:07] <pa> thanks![10:55:40] <MatthiasM> :)[10:56:09] <electrostatic> one sec...[10:59:09] <electrostatic> http://pastebin.com/m19a8ce19 <- code[10:59:29] <electrostatic> http://incp.us/img/lines.png <- result[10:59:59] <electrostatic> And if I uncomment the for loop, so it just "spins" for a bit, all of those ugly black lines go away[11:00:19] *** Suprano has joined ##OpenGL[11:00:36] <MatthiasM> electrostatic: why do you map/unmap the buffer directly ?[11:00:53] <electrostatic> I do not have to[11:00:56] *** geocalc has joined ##opengl[11:00:59] <MatthiasM> and why do you use PBO to transfer data inside OGL ?[11:01:10] <electrostatic> I was only trying to find some way to get it to block until it was finished[11:01:18] <electrostatic> I don't need to map any of them at all[11:01:20] <LtJax> you don't check for valid mapping return values either[11:03:25] <electrostatic> All the returns are valis, just checked[11:03:29] <electrostatic> *valid[11:03:31] <MatthiasM> electrostatic: make a small test case which shows this bug - and use (if possible) only 1 texture, 1 PBO etc[11:03:39] <electrostatic> k[11:04:01] <MatthiasM> and then send a bug report[11:04:07] <MatthiasM> or ask RTFM_FTW[11:04:30] <MatthiasM> but you use PBO in a way for which it was not intended to be used[11:04:53] <MatthiasM> and can you rty to write the buffer after the mapping into a file and check that file ?[11:05:16] <MatthiasM> if the lines are also there after the glMapBuffer in line 11[11:05:43] <electrostatic> MatthiasM: The spec is very precise in that you can use it for anything that is leagal for a buffer[11:06:09] <electrostatic> You can even use it to store vertex coords[11:06:12] <LtJax> electrostatic: I agree[11:06:27] <MatthiasM> well - maybe they missed this usecase in the synchronisation ?[11:06:35] <LtJax> everything that the spec allows is perfectly valid[11:06:52] <electrostatic> And for the record, I was testing out unpacking YUYV image data on the GPU[11:06:56] <MatthiasM> because you could use glCopyTex...[11:06:58] <LtJax> then again, stuff like this sometimes uses the ostrich algorithm for error handling ;)[11:07:11] <electrostatic> and this was the least GPU intensive way I could find to do some of the operations[11:08:14] <LtJax> but you'd still get some nice error values from the mapping commands - you did not just check them with a debugger, did you?[11:10:05] <electrostatic> 0x793000 - ponter, 0x7b9000 - ponter, 0x793000 - ponter, 0x7b9000 - ponter, and GL_ERROR 0x0[11:10:17] <MatthiasM> ponter?[11:10:26] <electrostatic> Those were the 4 returns I got and the resulting GL_ERROR code[11:10:30] <electrostatic> lol[11:10:32] <electrostatic> lointer[11:10:36] <electrostatic> gah[11:10:40] <electrostatic> *POINTER[11:10:41] <MatthiasM> heh[11:11:30] <MatthiasM> and if you remove all glMapBuffer calls it also happens ?[11:11:46] <electrostatic> yes, and a bit worse[11:11:58] <electrostatic> If I uncomment the spin loop it doesn't happen at all[11:12:18] <electrostatic> For the life of me, it just seems like it hasn't finished yet and doesn't care[11:12:18] <MatthiasM> ok, then make a test where you map one buffer after the glReadPixels and write that into a image file[11:12:44] <electrostatic> That probably wouldn't show it[11:12:50] <MatthiasM> try it[11:12:58] <electrostatic> becasue writing would be slower than the transfer[11:13:11] <MatthiasM> ten copy it into another buffer and write later[11:13:18] <MatthiasM> just use memcpy[11:13:28] <electrostatic> I might later, but It's almost 2:30 here now[11:13:52] <electrostatic> The memcopy might be a good idea though[11:14:51] <electrostatic> One more thing, can glCopyTex be used to change the color depth vs size of the image like I am doing?[11:15:01] <MatthiasM> no[11:15:26] <electrostatic> ANd actually, that part doesn't do that, but the Y plane does[11:15:27] <MatthiasM> but why do you need it ?[11:16:07] <electrostatic> Trying for very fast packing / unpacking of YUYV to YUVP[11:16:27] <electrostatic> So you read a YUYV frame in as RGBA[11:16:51] <MatthiasM> why not just use a shader with MRT ?[11:16:51] <electrostatic> Nad B=Y1 and B=Y2[11:18:02] <electrostatic> Well, I might be able to do that, acually[11:18:04] *** KU0N has joined ##opengl[11:18:23] <electrostatic> If I read the data as LUMA_ALPHA into the texture[11:18:49] <MatthiasM> you could upload the image as 1 channel texture instead of RGBA[11:19:08] <MatthiasM> and use ARB_texture_rectangle[11:19:23] <electrostatic> But then you still would have to take out the UV "gaps" to get the Y plane[11:19:33] <MatthiasM> and then use +0 for Y1, +1 for U, +2 for Y2 and +3 for V[11:19:36] <electrostatic> I have figered out how to do that, using GL_NEAREST[11:19:41] <MatthiasM> yep[11:20:10] <electrostatic> But you need a min tex size of 4096 to handle HD that way[11:20:29] <MatthiasM> you could address the texture unsing ((int)texCoordS)*4 + channel[11:21:51] <electrostatic> Somehow I do not think that would be as optimized as using the vector conversions[11:22:47] <electrostatic> BAnd it still doesn't solve the texture size problem[11:23:22] <electrostatic> But If I read it as RGBA and used CopyTex to get a LUMA_ALPHA copy also I think I could do it pretty easy[11:24:34] <electrostatic> But I couldn't use copytex for that[11:24:37] <electrostatic> meh[11:25:13] <electrostatic> I could, however, do the U and V palnes without a PBO[11:25:30] <electrostatic> Which means I could limit my PBO to just the Y plane[11:25:48] <electrostatic> and do the U and V between the PACK and UNPACK[11:26:31] <electrostatic> Or better yet, PACK, do U plane, UNPACK, do V palne[11:26:39] <electrostatic> *plane[11:26:51] <electrostatic> oh well, time for sleep tonight anyway[11:29:53] *** electrostatic has quit IRC[11:42:15] *** kenws has quit IRC[11:46:56] *** XT95_ has joined ##OpenGL[11:47:20] *** Eforen-AtSchool is now known as Eforen-Sleeping[11:52:07] *** kenws has joined ##OpenGL[11:56:25] *** Knightingale has joined ##OpenGL[12:14:50] *** XT95__ has joined ##OpenGL[12:16:17] *** cplusplus has joined ##OpenGL[12:22:43] *** XT95_ has quit IRC[12:24:59] *** sparky has quit IRC[12:27:51] *** groton___ has quit IRC[12:28:08] *** sparky has joined ##OpenGL[12:28:15] *** Knightingale has quit IRC[12:28:18] *** GuShH has joined ##OpenGL[12:47:55] *** TiLex has joined ##opengl[12:52:10] *** pietia has joined ##OpenGL[12:55:49] *** LordHavoc has quit IRC[12:58:02] *** Madsy has quit IRC[12:58:08] *** Adrinael has quit IRC[13:03:32] *** TiLex has quit IRC[13:03:41] *** Killari_ has joined ##OpenGL[13:03:54] *** TiLex has joined ##opengl[13:04:20] *** Adrinael has joined ##opengl[13:05:13] *** cplusplus2 has joined ##OpenGL[13:06:33] *** bijoo_appdev has joined ##OpenGL[13:06:33] *** bijoo_osdev has joined ##opengl[13:07:31] *** bijoo_appdev has quit IRC[13:07:31] *** bijoo_osdev has quit IRC[13:07:51] *** dvoid_ has quit IRC[13:10:07] *** rnx has left ##opengl[13:10:28] *** cplusplus has quit IRC[13:13:52] *** Killari has quit IRC[13:13:53] *** Killari_ is now known as Killari[13:14:18] *** ainaker has quit IRC[13:20:36] *** Wagoo has quit IRC[13:21:16] *** TiLex has quit IRC[13:21:25] *** Wagoo has joined ##opengl[13:35:44] *** groton___ has joined ##OpenGL[13:38:19] *** bijoo_osdev has joined ##opengl[13:38:19] *** bijoo_appdev has joined ##OpenGL[13:39:42] *** geocalc has quit IRC[13:40:18] *** geocalc has joined ##opengl[13:41:17] *** oliveira_ has joined ##OpenGL[13:42:05] *** johndoe has joined ##opengl[13:50:41] *** phrosty has quit IRC[13:53:48] *** oliveira has quit IRC[13:55:20] *** hang3r has joined ##OpenGL[13:56:22] *** wisey has joined ##OpenGL[13:58:06] *** XT95___ has joined ##OpenGL[13:59:52] *** XT95__ has quit IRC[14:06:08] *** maxton has quit IRC[14:18:11] *** UUncia_ has joined ##OpenGL[14:18:28] *** UUncia_ has quit IRC[14:19:16] *** UUncia has quit IRC[14:22:15] *** johndoe has quit IRC[14:22:36] *** johndoe has joined ##opengl[14:23:41] *** exitcode1 has joined ##OpenGL[14:25:07] *** LordHavoc has joined ##OpenGL[14:32:50] *** djork has joined ##opengl[14:33:24] *** GX has joined ##OpenGL[14:44:26] *** scai has left ##opengl[14:44:35] *** wisey has quit IRC[14:50:16] *** signalsea has quit IRC[14:58:17] *** pro-rsoft has joined ##OpenGL[14:58:43] <pro-rsoft> Hi, what can be wrong if CopyTexImage2D throws an "invalid operation" ?[15:00:17] <MatthiasM> pro-rsoft: read the docu: http://www.opengl.org/sdk/[15:00:27] <pro-rsoft> I did. I did everything it says[15:00:33] <pro-rsoft> it only happens on depth textures.[15:00:54] <MatthiasM> http://mikeash.com/getting_answers.html[15:02:01] *** geocalc has quit IRC[15:02:18] *** HuntsMan has quit IRC[15:02:24] *** eXtronuS has joined ##OpenGL[15:03:03] *** geocalc has joined ##opengl[15:05:59] *** eXtronuS has quit IRC[15:08:58] *** maxton has joined ##OpenGL[15:11:13] *** LordMetroid has joined ##OpenGL[15:14:14] *** HuntsMan has joined ##opengl[15:14:29] *** TiLex has joined ##opengl[15:27:29] *** hang3r has quit IRC[15:30:31] *** TiLex has quit IRC[15:34:02] *** wisey has joined ##OpenGL[15:36:28] *** jm|earth has joined ##OpenGL[15:41:17] *** sohail has joined ##OpenGL[15:42:16] *** GX has quit IRC[15:43:16] *** LtJax has quit IRC[15:54:32] *** jm|afk has quit IRC[15:56:16] *** Osirus has joined ##OpenGL[15:58:13] *** synchris has quit IRC[15:58:35] *** pietia has quit IRC[15:58:40] *** elite01 has joined ##opengl[16:00:49] *** pietia has joined ##OpenGL[16:01:30] *** synchris has joined ##opengl[16:05:22] *** Yustme_ has joined ##OpenGL[16:06:14] *** jcazevedo has joined ##OpenGL[16:07:32] *** calav3ra has joined ##opengl[16:08:33] *** pro-rsoft has quit IRC[16:11:01] *** huperniketes has joined ##OpenGL[16:11:05] *** huperniketes has quit IRC[16:13:00] *** lewymati has joined ##OpenGL[16:15:52] *** HuntsMan has quit IRC[16:17:06] *** mikolalysenko has joined ##OpenGL[16:18:21] *** XT95___ has quit IRC[16:20:55] *** Yustme has quit IRC[16:21:50] *** djork has quit IRC[16:23:47] *** Wagoo has quit IRC[16:25:53] <pa> MatthiasM, what you suggested was this i guesS: http:/www.j3d.org/matrix_faq/matrfaq_latest.html#Q26[16:26:01] <pa> do you have a clue what RADIANS could be?[16:29:15] <brutopia> it's probably a factor that converts degrees to radians when applied[16:29:54] *** marenz_ has joined ##OpenGL[16:35:26] *** geocalc has quit IRC[16:36:17] *** geocalc has joined ##opengl[16:36:29] *** kenws has quit IRC[16:36:41] *** m4ggus has joined ##opengl[16:38:11] *** k3t has quit IRC[16:39:25] *** dvoid_ has joined ##OpenGL[16:39:27] *** djork has joined ##opengl[16:42:28] *** kenws has joined ##OpenGL[16:44:45] *** twist has joined ##OpenGL[16:48:16] *** Wagoo has joined ##opengl[16:50:52] <sparky> pa: I didn't see RADIANS in there[16:52:03] <sparky> but radians is when you measure your angle in "actual" length along the unit circle[16:52:46] *** |Veji| has joined ##OpenGL[16:53:59] *** LordMetroid has quit IRC[16:54:49] *** LordMetroid has joined ##OpenGL[17:02:47] <pa> ah ok thanks[17:06:44] <sparky> pa: in other words 2*pi is 360 degrees[17:07:04] <sparky> since that's the length all the way around a unit circle[17:07:27] <sparky> pa: got it?[17:07:55] <sparky> this is also why you convert between radians and degrees the following way[17:08:35] <sparky> rad = deg * (PI / 180) and deg = rad * (180 / PI)[17:08:54] <sparky> since 2*pi corresponds to 360 degrees[17:11:03] <MatthiasM> pa: the sin/cos math function in math libraries of common programming languages also takes radians as parameter[17:13:06] <pa> yes i see[17:13:10] <pa> thanks : )[17:13:45] <pa> alltough he uses asin before multiplying with RADIANS[17:14:45] <MatthiasM> well - asin returns radians[17:15:22] <MatthiasM> pa: and if you read the code and the comments it's clear what this constant does[17:16:38] *** AlastairLynn has joined ##opengl[17:16:45] <pa> is it? not to me..[17:17:06] <MatthiasM> pa: the return values are in the range 0..360 - eg degree[17:17:21] <MatthiasM> pa: and asin return ratians: -pi ... +pi[17:17:30] <pa> ok got it[17:17:31] <pa> true[17:18:36] *** vade has quit IRC[17:19:39] *** bijoo_osdev has quit IRC[17:19:39] *** bijoo_appdev has quit IRC[17:28:15] *** marenz_ has quit IRC[17:29:00] *** itewsh has joined ##OpenGL[17:33:48] *** bijoo_osdev has joined ##opengl[17:33:48] *** bijoo_appdev has joined ##OpenGL[17:37:45] *** braden has quit IRC[17:39:17] *** bijoo_osdev has quit IRC[17:39:17] *** bijoo_appdev has quit IRC[17:45:59] *** bijoo_appdev has joined ##OpenGL[17:45:59] *** bijoo_osdev has joined ##opengl[17:46:14] *** BenPa85 has joined ##opengl[17:52:10] *** mikolalysenko has quit IRC[17:52:32] *** [AD]Turbo has quit IRC[17:53:58] *** LordMetroid has quit IRC[18:02:00] *** stevosICBM has joined ##OpenGL[18:03:55] *** signalsea has joined ##OpenGL[18:04:27] *** TheLorax has joined ##opengl[18:06:03] *** ewn has quit IRC[18:08:03] <lewymati> s -m krakow.irc.pl[18:08:18] <lewymati> damn sry[18:09:33] *** emzic has joined ##OpenGL[18:11:33] *** Osirus has quit IRC[18:12:02] *** jfroy|work has joined ##OpenGL[18:13:34] *** stevosICBM has quit IRC[18:14:33] *** BenPa85 has quit IRC[18:14:52] *** stevosICBM has joined ##OpenGL[18:15:33] *** stevosICBM has quit IRC[18:21:00] *** predaeus has quit IRC[18:21:25] *** fargiolas has quit IRC[18:25:37] *** bijoo_appdev has quit IRC[18:25:37] *** bijoo_osdev has quit IRC[18:25:58] *** mikolalysenko has joined ##OpenGL[18:26:28] *** fargiolas has joined ##OpenGL[18:26:43] *** kenws_ has joined ##OpenGL[18:28:07] <|Veji|> hi[18:28:39] *** maxton has quit IRC[18:28:40] <|Veji|> does anyone know where i can find a 3D model loader?[18:29:23] <|Veji|> i just need to test my app with 3 scenes at least[18:29:37] *** k3t^ has joined ##OpenGL[18:29:42] <|Veji|> C/C++[18:32:19] *** NekoGrafico has joined ##OpenGL[18:33:00] *** NekoGrafico has quit IRC[18:33:07] <AlastairLynn> for what format?[18:33:41] *** NekoGrafico has joined ##OpenGL[18:34:41] <|Veji|> .3ds or .max[18:34:59] *** NekoGrafico has left ##OpenGL[18:35:15] <AlastairLynn> *shrug*[18:36:25] *** johndoe has quit IRC[18:37:31] <|Veji|> but any format would be great because i didn't found anything in my searches[18:38:23] *** kenws has quit IRC[18:39:05] <MatthiasM> |Veji|: a) use another nick - without |[18:39:13] <MatthiasM> b) .max is not a published format[18:39:30] <MatthiasM> c) try WaveFront OBJ - that's a text based format and easy to parse[18:40:21] *** tokemonstah has joined ##OpenGL[18:40:26] *** scai has joined ##opengl[18:40:41] *** TheLorax is now known as factorz[18:40:56] *** factorz is now known as TheLorax[18:43:01] *** |Veji| is now known as Veji_[18:43:25] *** wisey has quit IRC[18:43:41] <Veji_> ok i'm going to check that[18:43:55] *** Renderwahn has joined ##OpenGL[18:43:56] <Veji_> thanks[18:44:16] *** groton has quit IRC[18:45:05] *** pro-rsoft has joined ##OpenGL[18:49:28] *** AlastairLynn has quit IRC[18:49:55] *** AlastairLynn has joined ##opengl[18:50:15] *** TheLorax has quit IRC[18:53:15] *** Nescafe has quit IRC[18:55:32] *** Veji_ has quit IRC[18:58:49] *** mm765 is now known as mm765^away[19:00:17] *** vade has joined ##OpenGL[19:12:28] *** XT95_ has joined ##OpenGL[19:13:02] *** stringfellow has quit IRC[19:13:04] *** doub has quit IRC[19:15:02] *** GuShH_ has joined ##OpenGL[19:16:57] *** GuShH has quit IRC[19:21:05] *** predaeus has joined ##opengl[19:32:50] *** JernejL has joined ##OpenGL[19:33:00] *** TheLorax has joined ##opengl[19:36:14] *** Kasu has joined ##OpenGL[19:38:19] *** dolphin has joined ##OpenGL[19:40:20] *** zwiep` has joined ##opengl[19:40:39] *** zwiep` has quit IRC[19:41:12] *** kenws__ has joined ##OpenGL[19:41:12] *** WhatAHam has quit IRC[19:41:42] *** WhatAHam has joined ##OpenGL[19:42:54] *** pietia has quit IRC[19:44:04] *** TheLorax has quit IRC[19:44:19] *** emzic has quit IRC[19:50:09] *** Jernej has quit IRC[19:50:19] *** Jernej has joined ##OpenGL[19:51:49] *** JernejL has quit IRC[19:52:19] *** groton has joined ##OpenGL[19:53:21] *** mikolalysenko has quit IRC[19:53:28] *** kenws_ has quit IRC[19:54:35] *** GuShH_ has quit IRC[19:55:32] *** chron- has joined ##OpenGL[19:56:37] *** TuomasT has joined ##OpenGL[19:59:37] *** kenws__ has quit IRC[20:00:20] <TuomasT> Can a dithering algorithm, that processes pixels in order, and each pixel which gets processed affects the coming pixels near it, be expressed in GPU (SIMD) way?[20:01:40] *** pro-rsoft has left ##OpenGL[20:01:42] <MatthiasM> unlikly - except if you add constraints on how other pixels are affected[20:01:55] <exDM69> and GPU's aren't SIMD[20:02:40] *** rorokimdim___ has joined ##opengl[20:02:44] <MatthiasM> exDM69: assuming that a lot of GPUs have vector operations you can consider them SIMD[20:02:50] *** JernejL has joined ##OpenGL[20:03:05] <exDM69> TuomasT: you may be able to do this with a multi pass algorithm[20:03:32] *** pietia has joined ##OpenGL[20:03:49] <exDM69> MatthiasM: yes, true. GPU's use SIMD too[20:04:29] *** rorokimdim___ has quit IRC[20:04:36] <TuomasT> Another thing is that I know how to in a fragment program (GLSL) do certain stuff to one pixel (in parallel), but can I process eg. four pixels in a single program[20:04:37] <exDM69> but that's only instruction level parellelism, when CPU's run several programs in parellel[20:05:12] <MatthiasM> TuomasT: a fragment shader doesn't know about neighbour pixels - butr you can use MRT[20:05:24] <TuomasT> MatthiasM: MRT?[20:05:25] *** scai has left ##opengl[20:05:25] <MatthiasM> exDM69: huh?[20:05:36] <MatthiasM> TuomasT: multiple render targets[20:06:53] *** Nescafe has joined ##OpenGL[20:07:02] *** rnx has joined ##opengl[20:07:10] <exDM69> MatthiasM: oops, GPU's have program level parallelism, too.[20:07:16] <MatthiasM> exDM69: SIMD means that one opcode executes the same operation on multiple data elements - like adding 2 vectors together[20:07:22] *** scai has joined ##opengl[20:07:40] <MatthiasM> it doesn't say anything about single/multi core, or super scalar etc[20:11:22] *** mikolalysenko has joined ##OpenGL[20:11:30] <exDM69> MatthiasM: exactly. my intent was to say that GPU programming is not best described with the term SIMD[20:11:38] *** Jernej has quit IRC[20:11:49] <TuomasT> In this dithering algorithm every other pixel in each row of pixels at a time could be processed in parallel. And the same for the next row. I guess that means it might execute fast on GPU?[20:12:12] <MatthiasM> exDM69: it does - just look at the GLSL or even the assembler like language - you have vector data types, and vector operations - so you write SIMD code[20:12:32] <TuomasT> How would I use this MTR though.. if I process one pixel and it effects another pixel then how does MTR help here?[20:12:55] <MatthiasM> TuomasT: I did not say that MRT helps your goal[20:13:16] <exDM69> MatthiasM: yeah, true. but I think that more important is the fact that it's parallel at program level[20:13:18] <MatthiasM> I don't think that you can implement dithering with a shader in a reasonable way[20:13:21] <exDM69> esp. if coding with cuda[20:13:36] <MatthiasM> exDM69: huh?[20:13:39] <TuomasT> I think the difference between CPU SIMD and GPU "SIMD" is that on CPU you have 4-16 or so parallelism and on GPU you have a lot more[20:14:13] <MatthiasM> the number of cores available doesn't matter at all for SIMD[20:14:25] <TuomasT> Thus I guess the best term for GPU is stream processing or massive SIMD[20:14:47] <MatthiasM> the only thing that matters is how many independant data values you can process with a single instruction[20:14:54] *** Walt has quit IRC[20:14:54] <TuomasT> MatthiasM: No, but SSE does 4 or 8 things in parallel[20:15:08] <TuomasT> Larrabee 12 I think[20:15:38] <exDM69> TuomasT: SIMD is instruction level parellelism. GPU's don't run the same instruction in all programs[20:15:41] <MatthiasM> TuomasT: using SSE you can process 16 bytes, 8 shorts, 4 ints/float or 2 double at once[20:15:58] <MatthiasM> exDM69: you got SIMD wrong[20:16:20] <exDM69> MatthiasM: no, I figure you understand me wrong. I agree with what you say about SIMD[20:16:21] <TuomasT> exDM69: They don't?[20:16:32] <MatthiasM> exDM69: the assembly level shaders and the GLSL level shaders allows you to write SIMD operations with up to 4 values[20:17:03] <MatthiasM> and how many programs are executed in paralell does not matter for SIMD at all[20:17:05] <exDM69> TuomasT: no, consider two shader threads running with an if statement.[20:17:32] <exDM69> MatthiasM: yes, I know.[20:17:42] <MatthiasM> exDM69: some GPUs have shared instruction sequencers - other independant - and others inbetween[20:17:44] <exDM69> adding two vectors together is SIMD[20:17:47] <MatthiasM> and this doesn't matter[20:18:52] <MatthiasM> and GF 8+ cards don't execute SIMD instructions at HW level (according to the published description on various websites)[20:19:05] <MatthiasM> but the programming interface through OpenGL is SIMD[20:19:39] <exDM69> yeah, new nvidia GPU's are scalar, not vector[20:20:24] <TuomasT> exDM69: I doubt that is true in strict sense[20:21:04] <exDM69> that's what I heard. don't know the details[20:21:33] <MatthiasM> the details don't matter at all[20:21:50] <MatthiasM> it's the job of the driver to match the API spec to the HW[20:21:52] <exDM69> TuomasT: a dithering algorithm could be written with a gpu in a multi pass algorithm[20:22:22] *** walt has joined ##OpenGL[20:22:34] <TuomasT> exDM69: How?[20:23:08] <exDM69> TuomasT: by splitting the dithering algorithm into different phases that can be made run in parallel[20:24:05] <TuomasT> exDM69: So how would you split it? I up there said you could split the algorithm so it processes each other pixel in a row in parallel, but I have no idea how to implement so one pixel affects other pixels[20:24:33] <TuomasT> Perhaps writing to a texture in each of the pixels if I know in the program what pixel it is[20:25:08] <MatthiasM> you can check the texture coordinate when reading to decide what code path to execute on that pixel[20:25:28] *** walt has quit IRC[20:26:03] <MatthiasM> but a lot of GPUs share branch control for several fragment - and mostly in 4x4 or similar blocks - so every other pixel causes both pathes to be executed per fragment[20:26:09] <MatthiasM> so it will be slower#[20:27:18] *** pietia has quit IRC[20:28:29] <TuomasT> Well one could also use an algorithm where each pixel sees two or three pixels near it. That way I think 4x4 per one pixel would not be wasting things[20:29:23] <MatthiasM> you can fetch as many texels as you like[20:31:36] <TuomasT> To divide things like "every other pixel" I need to create a new texture with eg. half the width of the full texture and then process that with fragment program?[20:32:29] <MatthiasM> no[20:33:00] <TuomasT> How then?[20:33:06] <MatthiasM> I told you[20:33:54] <TuomasT> I don't see how MRT is revelant[20:34:00] <TuomasT> relevant here[20:34:09] <MatthiasM> <MatthiasM> you can check the texture coordinate when reading to decide what code path to execute on that pixel[20:34:35] *** Walt has joined ##opengl[20:34:44] <TuomasT> Ah ok so every other pixel would just be made to terminate the program right away[20:34:53] <MatthiasM> because splitting a texture into even/odd pixels will take the same or even more time[20:35:00] <MatthiasM> no[20:35:38] <TuomasT> remember that in this algorithm you can't process a pixel until you have processed a pixel before it[20:35:40] <MatthiasM> if you want to execute a program per frament - then you can do it in one[20:36:02] <MatthiasM> if that is true for all pixels - then the last pixels depend on all pixels before it[20:36:14] <MatthiasM> and that can't be done effectivly on a GPU[20:37:22] <MatthiasM> and the last time I coded a dithering algorithm - it used error accumulation over all columns and the current row[20:37:31] <MatthiasM> eg 2d dithering[20:39:09] *** MouserSVK has joined ##OpenGL[20:39:17] <TuomasT> I guess I will have to either do this in CPU or come up with more suitable dithering algorith[20:39:18] *** XT95_ has quit IRC[20:44:17] *** _JSC415 has joined ##OpenGL[20:44:28] <_JSC415> damn .[20:44:39] <_JSC415> anyone knows how to disable those errors ?[20:44:44] <_JSC415> on gcc ?[20:44:53] <_JSC415> ;P[20:44:59] <MatthiasM> you can't disable errors :DD[20:45:06] <_JSC415> how ?[20:45:19] <_JSC415> i realy need it.[20:45:23] <MatthiasM> learn to read[20:45:30] <MatthiasM> _JSC415: http://mikeash.com/getting_answers.html[20:45:32] <_JSC415> because it makes it hard to use console sometimes[20:46:06] <_JSC415> damn. not that again.[20:46:19] <_JSC415> its going to be something like rickrol thing[20:46:30] <_JSC415> i hate to see those page.[20:46:48] <_JSC415> i just need a damn gcc option[20:47:05] <_JSC415> damn nick.[20:47:09] <MatthiasM> you don't get it - do you ?[20:47:12] *** _JSC415 is now known as _THEGOD[20:47:26] <MouserSVK> _THEGOD: This one is even better: http://www.catb.org/~esr/faqs/smart-questions.html[20:47:37] <_THEGOD> what i were hawe to get ?[20:48:06] <_THEGOD> blah.[20:48:21] <_THEGOD> is you..[20:48:40] <_THEGOD> because you dont even capable to answer a stupid gcc option wuestion .[20:48:42] <_THEGOD> hm..[20:48:44] <_THEGOD> lols.[20:48:46] <_THEGOD> sorry[20:48:52] <_THEGOD> you were using java ?[20:49:13] <_THEGOD> so you dont know it actually. ;P[20:49:15] <predaeus> _THEGOD, please stop pressing enter so often.[20:49:17] <MatthiasM> I can't answer stupid question - because they contain no useable information that can be answered[20:49:50] <_THEGOD> yeah a thing that does not exist ius not useable ;P[20:50:15] <_THEGOD> lols anyway now i dont hate you that much,[20:50:19] *** b0000 has joined ##opengl[20:50:38] <MatthiasM> _THEGOD: do you even read the shit you write ?[20:50:59] <_THEGOD> because you dont know it so you don hawe to say it.[20:51:09] <_THEGOD> i mean that gcc option.[20:51:13] <MatthiasM> http://echelog.matzon.dk/logs/browse/opengl/1231369200 <-- read what you wrote @ _THEGOD starting 20:44[20:51:49] <MatthiasM> and I answered your question[20:51:59] <MatthiasM> but you can't understand it[20:53:05] <_THEGOD> lols :P[20:53:08] <_THEGOD> [20:44:59] <MatthiasM> you can't disable errors :DD ?[20:53:32] <_THEGOD> it looks like i didnt see the 't'[20:53:35] <_THEGOD> ok.[20:53:37] <_THEGOD> sorry.[20:53:51] <_THEGOD> but if its possible to show all -Wall[20:54:11] <_THEGOD> so shouldn't it be possible to disable too ?[20:54:18] <MouserSVK> _THEGOD there's difference between "Warning" and "Error"[20:54:19] *** tuxer` has joined ##OpenGL[20:54:22] *** mikolalysenko has quit IRC[20:54:28] <_THEGOD> hm..[20:54:33] <tuxer`> Hello World![20:54:36] <MatthiasM> _THEGOD: there is also a GCC manual - but it requires reading[20:54:45] <_THEGOD> may be i asked wrong thing.[20:54:54] <_THEGOD> i nedd to disable errors ?[20:54:57] <MatthiasM> _THEGOD: http://mikeash.com/getting_answers.html <-- READ THIS![20:55:02] <_THEGOD> is this what were said before ?[20:55:03] <MouserSVK> that's why you got 2 links to web sites how to ask[20:55:18] *** pietia has joined ##OpenGL[20:55:21] <MouserSVK> you can't disable errors ;)[20:55:23] <MouserSVK> Error is error[20:55:26] <MouserSVK> end of line[20:55:33] <MouserSVK> you can't do anything about it[20:55:39] <MatthiasM> _THEGOD: I suggest to stop writing anything into this chat until you understood what is written in these links[20:55:46] <_THEGOD> yeah may be but i dont wanna disable errors.[20:55:52] <MouserSVK> it's just like if I was asking you to make gold from water[20:55:53] <_THEGOD> i want to disable warnings.[20:56:01] <MouserSVK> it is error to do such thing[20:56:02] <_THEGOD> which does not makes much sense .[20:56:03] <MouserSVK> impossible :)[20:56:31] <_THEGOD> nothing is impossible. ;P[20:56:38] <_THEGOD> its may take too long.[20:56:54] <_THEGOD> but not impossible :P[20:57:22] <_THEGOD> hm..[20:57:23] <MatthiasM> it's impossible that _THEGOD learns to read[20:57:54] <_THEGOD> may be not anything anyway. ;P[20:57:54] <MouserSVK> letmegooglethatforyou.com doesn't work :([20:57:59] <MouserSVK> anyway, look here: http://developer.apple.com/documentation/DeveloperTools/gcc-4.0.1/gcc/Warning-Options.html#Warning-Options[20:58:22] <MatthiasM> MouserSVK: don't give him links[20:58:26] <MouserSVK> next time, it will be (for all of us) faster to use google[20:59:38] <MatthiasM> it's easy for anywone half way interrested to find the answer to such questions in less then 5 minutes[21:00:06] <MouserSVK> http://letmegooglethatforyou.com/?q=gcc+suppress+warnings[21:00:13] <MouserSVK> this was the kind of link i wanted to give him[21:00:28] <MouserSVK> yeah, you're right[21:02:12] <MatthiasM> I bet he believes in nosmoke.exe too :P[21:02:25] <predaeus> hahaha, cool page[21:02:49] <_THEGOD> hm.. my gcc says unrecognised option for -Wno-#warnings[21:03:04] *** m4ggus has quit IRC[21:04:06] <MouserSVK> _THEGOD.. then you are doing something wrong[21:04:17] *** sysRPL has joined ##OpenGL[21:04:18] <MouserSVK> you should really learn to read carefully[21:04:25] <sysRPL> hello[21:04:41] <sysRPL> has any of you here looked at crayon physics deluxe?[21:04:50] <b0000> _THEGOD: when you are as big a noob as you clearly are i think you should leave warnings on, and fix them properly, warnings are designed to help noobs[21:05:54] *** stevosICBM has joined ##OpenGL[21:05:59] <MatthiasM> b0000: warnings are to help everyone[21:06:15] *** Jupp3 has quit IRC[21:06:32] <MouserSVK> sysRPL I've seen the vids[21:06:46] <b0000> MatthiasM: sure, thats why i normally compile with highest warning level, and disable just a few[21:06:50] <sysRPL> it's somewhat interesting[21:07:05] <sysRPL> i like the whole hand crayon drawn look to the graphics[21:07:13] <sysRPL> and of course the paper background[21:07:23] <b0000> MatthiasM: my point being, if you are a noob and do not know if a warning is safe or not you should probably fix it[21:07:36] <MatthiasM> yes - languages like C/C++ are way to complex to see and prevent every dangerous construct at source code level[21:07:38] <sysRPL> i was wondering if some of it uses opengl, because they have an iphone version[21:07:39] <MatthiasM> exactly[21:07:47] <sysRPL> i figured they used opengles[21:08:17] <MatthiasM> sysRPL: a physic library should not use a graphic library - only examples should[21:08:56] <sysRPL> MatthiasM: the game isn't a library[21:09:17] <sysRPL> i was referring to the visuals[21:10:03] *** sysRPL has quit IRC[21:12:15] *** chron- has quit IRC[21:15:10] *** mikolalysenko has joined ##OpenGL[21:17:28] *** chron- has joined ##OpenGL[21:17:36] *** Spkka has joined ##OpenGL[21:17:37] *** chron- has quit IRC[21:21:54] *** KU0N has quit IRC[21:29:35] *** druggy has joined ##opengl[21:32:30] *** Spkka has quit IRC[21:33:01] *** cplusplus3 has joined ##OpenGL[21:35:54] *** druggy__ has joined ##opengl[21:36:37] *** stringfellow has joined ##opengl[21:38:16] *** druggy has quit IRC[21:38:59] *** GuShH has joined ##OpenGL[21:45:35] *** druggy_ has quit IRC[21:46:03] *** dusted has quit IRC[21:48:00] *** Jupp3 has joined ##OpenGL[21:48:18] *** maxton has joined ##OpenGL[21:48:24] *** Gorgoroth has joined ##OpenGL[21:49:53] *** Spkka has joined ##OpenGL[21:50:22] *** cplusplus2 has quit IRC[21:55:41] <TuomasT> MatthiasM: Could you direct me to how I can know in fragment shader whether "this" is even or odd pixel (in x and y)?[21:56:01] <MatthiasM> you have the texture coordinate - the rest is math[21:56:44] <TuomasT> MatthiasM: You mean this: gl_TexCoord[0].st ?[21:57:12] <TuomasT> My shader starts like ( vec4 color = texture2D(tex,gl_TexCoord[0].st); )[21:57:37] <MatthiasM> yep[21:58:05] <TuomasT> Is that .st an integer value from 0..last pixel?[21:59:01] <MatthiasM> depends[22:02:29] <TheFlash> .st is same as .xy[22:02:49] <TheFlash> Or .bg[22:02:57] <MatthiasM> TheFlash: the same components but different semantic[22:03:02] <MatthiasM> and no - rg[22:03:19] <TheFlash> ?[22:03:22] *** Quentarez has joined ##OpenGL[22:03:25] <MatthiasM> .x == .r[22:03:47] <_THEGOD> kububtu sucks.[22:03:49] <TheFlash> .y == .g[22:03:56] *** MouserSVK has quit IRC[22:04:11] <_THEGOD> do anyone knows how to install gl.h to kubuntu ?[22:04:23] <TheFlash> .xy == .rg, or is it illegal to get only 2 component out of vector?[22:04:25] <_THEGOD> i downloaded it but its ill.[22:04:28] <_THEGOD> no g++[22:04:33] <_THEGOD> just gcc[22:04:38] <MatthiasM> TheFlash: you can get any combinations out of a vector[22:04:50] <MatthiasM> TheFlash: you can also do .raag[22:04:58] <_THEGOD> im about to download knoppix dvd. 5.3[22:05:00] <TheFlash> So, .xy would be .rg ?[22:05:06] <TheFlash> +same as[22:05:18] <_THEGOD> and i know it has all i need.[22:05:21] <MatthiasM> _THEGOD: stop posting shit into the channel[22:05:33] <_THEGOD> when i look too statistical referances ;P[22:05:43] <MatthiasM> TheFlash: yes - but you can't do - .ry[22:05:52] <TheFlash> I know.[22:05:53] <_THEGOD> .ry ?[22:06:05] <_THEGOD> what that means ?[22:06:12] <MatthiasM> we talk about GLSL[22:06:31] <_THEGOD> shading effects ?[22:06:39] <TheFlash> You can't mix them.[22:06:40] <MatthiasM> no - the language[22:07:33] <TuomasT> MatthiasM: So should this be a valid "x is even" check: if (mod(gl_TexCoord[0].s,2) == 0) ?[22:07:41] <MatthiasM> _THEGOD: you are a) unable to read manuals b) use google c) the correct IRC channel for the questions you ask - so STFU[22:08:01] <MatthiasM> TuomasT: it's float point math - so better to compare < 0.5[22:08:04] *** pietia has quit IRC[22:08:32] <TuomasT> MatthiasM: But isn't gl_TexCoord[0].s integer?[22:08:33] <TheFlash> TuomasT: Texels are 0..1 floats, by default.[22:08:35] <MatthiasM> TuomasT: but GL_TEXTURE_2D uses coords from 0..1[22:08:51] <MatthiasM> TuomasT: how did you declare it ?[22:08:58] <TheFlash> Oh that was coordinate. Not a sampler value.[22:09:04] <MatthiasM> ahh - wait built ins - they are float[22:09:21] <TuomasT> In vertex shader I have: gl_TexCoord[0] = gl_MultiTexCoord0; ..[22:09:43] <MatthiasM> yep - get the GLSL spec - it has the definition of all built in variables[22:11:02] <TuomasT> Hmm. So I should check from 0..1 pixel value whether something is odd or even? I don't see how[22:11:15] <MatthiasM> try ARB_texture_rectangle[22:11:27] <TheFlash> Multiply it to another domain.[22:11:29] <MatthiasM> TuomasT: you would need to multiply by the texture size first[22:11:52] <TuomasT> Do I need to supply that manually from program?[22:12:06] <TheFlash> No.[22:12:13] <TheFlash> Of course not.[22:12:15] <_THEGOD> hey do anyone knows any coll things as a opengl programmer about kubuntu or any ubuntu ?[22:12:35] <MatthiasM> TuomasT: yes you need to pass the texture size for GSLSL < 1.3[22:12:36] <TheFlash> If you mean your main program that runs on CPU, not on GPU.[22:12:37] <_THEGOD> there is a empire about ubuntu.peoples who thinks its awesome[22:12:38] <MatthiasM> *GLSL[22:12:47] <_THEGOD> and i cant see any better about it ;P[22:12:48] <TuomasT> MatthiasM: I'm using 2.0[22:12:56] <TuomasT> Or rather, I require 2.0[22:13:06] <_THEGOD> i was goint to use it as peoples says.[22:13:09] <TheFlash> TuomasT: GL3 ?[22:13:18] <MatthiasM> TuomasT: GLSL version != GL version - but it's likely that you only have GLSL 1.2 or even 1.1[22:13:27] <MatthiasM> _THEGOD: RTFM !!!!![22:13:46] <_THEGOD> RTFM ?[22:13:50] <_THEGOD> what that means ?[22:13:55] <b0000> _THEGOD: google it[22:14:01] <MatthiasM> read the f*** manual ![22:14:05] <_THEGOD> run time file memory ?[22:14:09] <MatthiasM> and the important part is READ[22:14:17] <MatthiasM> but you can't read[22:14:19] <TuomasT> Why stars? Its read the fine manual[22:14:40] <MatthiasM> TuomasT: heh :DD[22:14:44] <tmccrary1> read the fearsome manual[22:15:18] <_THEGOD> the ubuntu manuals says ubuntu has no root password ;P[22:15:20] <_THEGOD> lols.[22:15:28] <TheFlash> TuomasT: What do you want to do with texels that have odd s-coordinate. (just if we can come up with some alternatives)[22:15:31] <_THEGOD> then how could it be used ?[22:15:55] <MatthiasM> _THEGOD: are you so stupid ? read it ![22:16:02] <TheFlash> _THEGOD: Does this look like #ubuntu?[22:16:03] <TuomasT> TheFlash: Implement ordered dithering. Trying out 2x2 first.[22:16:07] <b0000> _THEGOD: how did you get so stupid?[22:16:20] <_THEGOD> i dunno.[22:16:32] <_THEGOD> damn. usas awerage is stupid them me anyway ;P[22:16:47] <_THEGOD> they chosed a foreigner as their minister ;P[22:17:08] <b0000> _THEGOD: you are by far the stupidest person ive seen in a long time[22:17:12] <_THEGOD> thats the most stupid a nation or banch of animals ewer can do ;P[22:17:27] <_THEGOD> the most stupid thing i mean[22:17:39] <_THEGOD> yeah may be.[22:17:53] <_THEGOD> we cahe to do a cechnmark test for that[22:18:03] <_THEGOD> who is the stupidest man you ewer saw ?[22:18:17] <b0000> _THEGOD: i find your racism highly offensive too[22:18:29] <_THEGOD> racism ?[22:18:43] <TheFlash> I thought I had a high typo rate...[22:18:52] <_THEGOD> its not offensive .[22:19:05] <_THEGOD> if you are not allready offended by default ;P[22:19:09] <b0000> TheFlash: typo tends to suggest one knows how to spell the word in the first place[22:19:24] *** Jernej has joined ##OpenGL[22:19:49] <MatthiasM> TheFlash: tell me how it's possible to misstype a V as a W ?[22:20:09] <_THEGOD> well. ve call them booth wwwwwwww.[22:20:13] <_THEGOD> vvvvvvvvvv. ;P[22:20:19] <oliveira_> Oh just shut up[22:20:21] <MatthiasM> no[22:20:35] <TheFlash> MatthiasM: Maybe with some odd keyboard layout?[22:20:38] <b0000> ./ignore _THEGOD[22:20:41] <MatthiasM> heh - yeah :)[22:20:44] <_THEGOD> and visually too they looks some when we look .[22:21:51] * vade sighs[22:21:55] <vade> on inter-tubez[22:21:58] <vade> how you amuse me[22:22:21] *** groton has quit IRC[22:22:32] <TuomasT> Can I in fragremt program say: "print this variable to command line" (for debugging)?[22:22:32] <TheFlash> TuomasT: Is it important that your dithering thing is... "looking just correct"?[22:23:01] <vade> TuomasT: not that I am aware[22:23:14] <TuomasT> TheFlash: Well ordered-dithering is not very great way to do dithering to begin with[22:23:23] <vade> the best way to debug is to set the gl_FragmentColor to the variable you want to debug[22:23:27] <vade> so you can 'visualize' its output[22:23:29] <TheFlash> TuomasT: If not, you could just multiply that .s with some integer and take frac() of that product and compare it with < 0.5.[22:23:41] <b0000> TuomasT: unfortunately not, but you could write it to gl_FragColor[22:23:52] <vade> er yes gl_FragColor :D[22:24:23] <b0000> shame opengl doesnt have anything close to NV PerfHUD[22:24:40] <MatthiasM> there are shader debuggers[22:25:20] <b0000> MatthiasM: please direct me towards one that even slightly compares to PerfHUD on windows, and doesnt cost money[22:25:23] <TheFlash> Printing out debug values to console would be great...[22:25:37] <MatthiasM> you may want to get a 30 day trail of gDEBugger[22:25:47] <TheFlash> gdb like debuggers would also...[22:25:53] <vade> Shader Builder on OS X? (maybe?)[22:25:54] <_THEGOD> anyway... i found pelib. on my pc and wen.[22:26:01] * vade shrug[22:26:07] <vade> and OpenGL profiler is sexy as fuck on OS X[22:26:13] <_THEGOD> sad i dont hawe to find missing files again from web ;([22:26:13] <TheFlash> b0000: What does that PerfHUD do?[22:26:14] <MatthiasM> well - use MESA software rasterisor and single step it[22:26:18] *** JernejL has quit IRC[22:26:22] <_THEGOD> time.[22:26:36] <b0000> MatthiasM: i wasted my gDEBugger 30 day liscence years ago[22:26:49] <MatthiasM> b0000: maybe it works again ?[22:26:55] <_THEGOD> gDEBugger ?[22:27:05] <MatthiasM> last time I tried gDEBugger it crashed once my app started :([22:27:09] <_THEGOD> i had wstudio 8 pro ;P[22:27:11] <b0000> TheFlash: profiles dx, allows you to single step draw calls, debug shaders and many other features[22:27:17] <_THEGOD> its web designer is cool.[22:27:20] <_THEGOD> i like it.[22:27:31] <MatthiasM> _THEGOD: you are the stupiest thing on IRC that I have seen[22:27:36] <TheFlash> b0000: Sounds handy...[22:27:37] <_THEGOD> but c++ editor works but not good enough.[22:27:48] <_THEGOD> but still can help a bit for some works[22:27:49] <b0000> MatthiasM: i could apply for an extension i think but id rather have something i can use for a reasonable amount of time[22:27:56] *** elite01 has quit IRC[22:28:18] <MatthiasM> b0000: I heared of a GLSL debugger for linux - which allows single stepping - but I don't remember the name :([22:28:34] <predaeus> TuomasT, there is also FragCoord if you want screen relative and not primitive relative coordinates.[22:28:36] <_THEGOD> yeah remembering is pain ;P[22:28:41] <MatthiasM> http://www.vis.uni-stuttgart.de/glsldevil/ :DD[22:29:10] <MatthiasM> I would avoid gl_FragCoord - some ATI drivers returned inverted y coordinates[22:29:12] <b0000> MatthiasM: i tried it before, it was slow as hell[22:29:19] <b0000> MatthiasM: ill try it again now though[22:29:24] <MatthiasM> b0000: well - slow is better then nothing at all :)[22:29:36] <b0000> MatthiasM: it was unbearably slow[22:29:49] <TuomasT> This works: " if (fract(gl_TexCoord[0].s * 1024.0/2.0) < 0.5) " for getting a striped result in x ( the texture is 1024 width, don't know why it has to be divided)[22:29:51] <b0000> MatthiasM: like 400x slower than normal :/[22:30:54] <b0000> MatthiasM: ill try it again on my uber rig though, and maybe its matured somewhat[22:31:28] <MatthiasM> TuomasT: because .s * 1024 will give you the texel location - if you divide by 2 you get 2 texels in the range 0...1[22:32:00] <MatthiasM> b0000: I think you need to strip your app down to the minimum - eg using glScissor etc[22:32:30] <predaeus> how advanced are GLSL compilers? maybe better use * 0.5 instead of divide.[22:33:55] <MatthiasM> predaeus: you can assume that this basic constant folder is performed[22:35:13] *** predaeus has quit IRC[22:36:22] *** qeed has joined ##opengl[22:37:26] *** dolphin has quit IRC[22:37:57] <b0000> MatthiasM: ok glslDevil has improved a lot since i last used it, or maybe its just my pc[22:38:07] *** elite01 has joined ##opengl[22:38:10] <MatthiasM> good to know :)[22:38:36] <b0000> MatthiasM: give it another year and it might actually come close to perfhud :D[22:40:45] *** prak has joined ##OpenGL[22:41:10] <prak> does it matter which opengl development tutorial i use to get myself started, considering that there are different types of api packages?[22:41:14] *** lewymati has quit IRC[22:42:01] <MatthiasM> prak: OpenGL itself will not change between bindings - only how you create the window and receive input events[22:42:56] <prak> that means any type of opengl tutorials would be fine?[22:43:29] <MatthiasM> prak: try the red book[22:43:38] <prak> the red book?[22:44:05] <prak> http://www.opengl.org/documentation/red_book/?[22:44:57] <MatthiasM> yes - the online version is an old version[22:45:03] <MatthiasM> but it's a good start[22:47:05] <prak> ok[22:47:07] <prak> thanks a lot[22:55:05] *** dv_ has joined ##opengl[22:57:39] *** Eforen-Sleeping is now known as Eforen[23:03:34] *** Gorgoroth has quit IRC[23:09:01] *** _THEGOD has quit IRC[23:11:36] *** Renderwahn has quit IRC[23:12:33] *** phrosty has joined ##opengl[23:12:37] *** phrosty has quit IRC[23:13:00] *** phrosty has joined ##opengl[23:15:52] *** Quentarez has quit IRC[23:15:58] *** Eforen is now known as Eforen-atSchool[23:20:33] *** groton has joined ##OpenGL[23:21:57] *** phrosty has quit IRC[23:22:14] *** phrosty has joined ##opengl[23:24:13] *** cplusplus has joined ##OpenGL[23:28:21] *** Ingenu has quit IRC[23:29:31] *** Yustme_ has quit IRC[23:29:49] *** Yustme has joined ##OpenGL[23:34:34] *** johndoe has joined ##opengl[23:37:43] *** vade has quit IRC[23:38:34] *** RTFM_FTW has quit IRC[23:39:21] *** mlucassmith has quit IRC[23:39:40] *** mlucassmith has joined ##OpenGL[23:41:12] *** cplusplus3 has quit IRC[23:41:34] *** iion has joined ##OpenGL[23:42:00] *** Bollinger has quit IRC[23:42:51] *** twist has quit IRC[23:47:46] *** stevosICBM has quit IRC[23:50:08] *** itewsh has quit IRC[23:55:54] *** Yustme has quit IRC[23:59:04] *** oliveira has joined ##OpenGL[23:59:33] *** amz has joined ##opengl[23:59:33] *** Xmas| has joined ##OpenGL