January 7, 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 7, 2009 [00:02:10] *** amz has joined ##opengl[00:02:23] *** neoneye has quit IRC[00:03:26] <Madsy> emzic: Query OpenGL for errors already[00:04:54] <b0000> emzic: why not query the shader for active attributes?[00:07:10] *** JernejL has joined ##OpenGL[00:09:40] *** Xantoz has joined ##OpenGL[00:12:44] *** scai has left ##opengl[00:13:25] *** mouserSVK has quit IRC[00:20:57] *** twist has quit IRC[00:22:54] *** pietia has quit IRC[00:25:26] *** Suprano has quit IRC[00:29:31] *** vade has quit IRC[00:32:40] *** emzic has quit IRC[00:41:00] *** Amorphous has quit IRC[00:41:17] *** eXtronuS has quit IRC[00:44:10] *** Amorphous has joined ##opengl[00:44:13] *** ol1veira has quit IRC[00:45:52] *** m4ggus has joined ##opengl[00:47:27] *** zigonick has quit IRC[00:56:15] *** Renderwahn has quit IRC[01:01:20] *** cplusplus has joined ##OpenGL[01:01:31] *** tokemonstah^ has quit IRC[01:01:40] *** cplusplus2 has quit IRC[01:03:18] <yates> what's a good way to determine frame rate?[01:04:09] <yates> the obvious way? (i.e., get the time after each swapbuffers and compute the deltas)?[01:04:59] <MatthiasM> count the time it took to render say 100 frames[01:05:42] <MatthiasM> this has 2 benefits - a kind of low pass filtering - and the number does not change every frame so that you can read it[01:06:21] <yates> is right after SwapBuffers a good place to sample the time?[01:06:47] <yates> (every 100th swapbuffer, i.e.)[01:07:36] <MatthiasM> yep - but I normaly get time at the start of the loop - and swap at the end of the loop[01:08:17] <yates> right[01:08:32] *** yates has left ##opengl[01:09:16] *** AlastairLynn has quit IRC[01:09:27] *** yates has joined ##OpenGL[01:10:20] <yates> is there a vsync control function in glu or glut?[01:14:06] <yates> MatthiasM?[01:14:42] <b0000> yates: no, you must use extensions[01:16:00] <yates> you mean e.g. the method you pastebin'ed last night?[01:16:23] <MatthiasM> yates: you got the method for this already[01:16:35] *** vade has joined ##OpenGL[01:17:12] <yates> i did?[01:17:29] <yates> you mean the one b0000 pastbin'ed last night?[01:18:11] <yates> http://pastebin.com/d9747966[01:19:59] <yates> i don't understand what you're talking about[01:20:51] *** m4ggus has quit IRC[01:20:54] *** m4ggus_ has joined ##opengl[01:20:56] *** m4ggus_ is now known as m4ggus[01:22:42] <yates> MatthiasM: What are you referring to?[01:23:43] <MatthiasM> that link[01:26:20] <e-roder> hey[01:26:38] <e-roder> i'm having trouble properly calculating my normals[01:26:43] <yates> that's normal[01:26:47] <e-roder> or something[01:26:47] <yates> hehe[01:27:00] <yates> are they triangles?[01:27:02] <e-roder> yeah[01:27:11] <yates> cross product. easy beezy[01:27:20] <e-roder> i have them in order, going counter clockwise[01:27:24] <e-roder> right[01:27:37] <yates> ok[01:27:46] <e-roder> but i have 3 vertices, not 2 vectors[01:27:59] <e-roder> i'm not sure which ones i need to subtract, and it seems to change for each triangle[01:28:45] <yates> it's arbitrary[01:28:56] <yates> vector1 = v2 - v1[01:29:02] <yates> vector2 = v3 - v1[01:29:02] <e-roder> i mean, i can get the normal as a +/- scalar of the correct vector[01:29:04] <yates> for example[01:29:23] <yates> you mean you need to normalize?[01:29:34] *** johndoe has quit IRC[01:29:39] <e-roder> not normalization[01:29:40] <e-roder> direction[01:29:44] <e-roder> signage[01:29:47] <yates> +/- isn't good.[01:29:55] <MatthiasM> e-roder: do you use backface culling when rendering ?[01:30:04] *** itsmonktastic has joined ##opengl[01:30:16] <yates> he should do it rright anyway[01:30:35] <e-roder> yeah[01:30:35] <e-roder> i'm trying to specify normals for lighting[01:30:45] <e-roder> first of all[01:30:48] <MatthiasM> if you don't use backface culling then you won't see if the winding of some of the triangles is wrong[01:30:53] <e-roder> can i get away specifying normals for every 3 vertices?[01:30:59] <yates> e-roder: the "right" way to do it is to have your verts CW for the surface that's "up"[01:31:19] <MatthiasM> e-roder: you can only compute a normal per triangle[01:31:42] <yates> then use the goold-old right-hand rule for the right vector order[01:31:52] <e-roder> i think my cross product is using the left hand rule[01:31:58] <yates> :([01:32:03] <MatthiasM> if you need a normal per vertex then you need to average the normals of all adjacent triangles[01:32:04] <yates> bad, left hand is bad[01:32:05] <e-roder> it seems to work for my shadow volume creation[01:32:19] <e-roder> but the normal isn't jiving with their lighting[01:32:24] <e-roder> (their = ogl)[01:32:24] <MatthiasM> e-roder: a cross-product has no left/right hand[01:32:29] <yates> that's because you've established CCW as up[01:32:51] *** SilentAssassin has quit IRC[01:34:03] <yates> MatthiasM: You're wrong on that one.[01:34:25] <MatthiasM> huh? the math for the cross product is defined[01:34:36] <yates> sure it is[01:34:45] <yates> but v1 x v2 != v2 x v1[01:36:34] <Madsy> Sure. The Cross product isn't commutative.[01:36:55] <yates> v1 x v2 = -1 * v2 x v1[01:37:16] *** ignacio has joined ##OpenGL[01:38:10] <Madsy> But it doesn't have anything to do with handness of a coordination system[01:38:25] <MatthiasM> yates: this is the use of the cross product - not the cross product itself ![01:38:47] <Madsy> It's just the rules for the multiplication, just as row and column vectors are rules for matrix multiplication,[01:39:31] *** cplusplus has quit IRC[01:39:54] <yates> Madsy: right[01:40:06] <yates> e-roder: forget my comment on "CCW as up"[01:41:03] <yates> wait[01:41:09] <yates> no, i think i'm right[01:41:34] <yates> you want your cross product vector to be in the "up" direction. so knowing which way is up is relevent[01:41:54] <Madsy> Of course it is relevant. You either get n or -n[01:42:16] <e-roder> that's what i thought my issue was[01:42:22] <e-roder> but something else funny is going on here[01:42:24] <Madsy> But you mentioned the multiplication order of the cross product as coordination system handness[01:42:27] <yates> Madsy: what did you mean by "doesn't have anything to do with the handness of a coordinate system"?[01:42:28] <Madsy> Which is incorrect[01:42:46] *** reprore_ has joined ##OpenGL[01:43:05] <yates> i did?[01:43:57] <Madsy> s/you/someone[01:44:59] <yates> did you mean the left-hand vs right-hand rule thing?[01:46:06] *** b1010 has joined ##opengl[01:46:27] *** b1010 has joined ##opengl[01:47:50] <e-roder> ok well i solved the normal problem[01:48:12] <e-roder> i was using memory addresses instead of vectors, i needed to throw in some *'s[01:48:21] <e-roder> but now everything is white[01:48:49] <e-roder> and i normalized the normal first[01:50:09] <yates> cool[01:50:20] <yates> not cool[01:50:26] <e-roder> there should be diffuse lighting (i think?)[01:50:32] <e-roder> some sort of gradient should be going on[01:57:59] *** reprore_ has quit IRC[02:00:04] *** reprore_ has joined ##OpenGL[02:00:58] <e-roder> alright this is just strange[02:01:28] <e-roder> if i specify the normal with glNormal3f(0, 0, 1); i get proper shading[02:01:51] *** b1010 has quit IRC[02:02:35] <e-roder> but if i specify the normal as glNormal3f(0, 1, 0); and i have a glRotatef(90, 1, 0, 0); preceding that, the shading is all black or white[02:04:32] *** b0000 has quit IRC[02:08:36] *** mm765 is now known as mm765^sleep[02:08:42] *** b0000 has joined ##opengl[02:10:56] *** dvoid_ has quit IRC[02:11:43] *** paul424 has joined ##OpenGL[02:15:35] <paul424> hey suppouse I use the vertex list to keep the map of a terrain.. it is possible to customize this to the roam algorithm ...... probably not so the only way is to put the vertex "by hand" into the pipeline ........[02:18:58] <paul424> quicksilver: ????[02:19:40] <paul424> MatthiasM: >>>[02:20:12] <MatthiasM> what do you want to do ?[02:20:58] <paul424> do you know ROAM algorithm ?[02:21:22] <MatthiasM> yes[02:22:09] <paul424> hmm you probably wonder whether my observer is dynamicall or statical[02:22:16] <paul424> well its dynamical[02:22:30] <paul424> so it's no use to keep the vertex lists .......[02:23:02] <paul424> or I might to keep a several smaller ones and only play with the map at which observer stays at[02:24:07] <paul424> anyway I did some benchmarks and the vertex lists is not faster than putting the vertex "by hand" .......[02:24:19] <paul424> MatthiasM: ???[02:24:22] <MatthiasM> using a VBO / VA is not only good when the data is static[02:24:50] <MatthiasM> you can also use it top stream data - in which case you save method calls compared to immediate mode[02:25:19] <MatthiasM> and don't assume just because you ask something that all people sit and wait to answer you[02:25:33] <paul424> yeap I know .....[02:26:05] <paul424> it's quite calm now .... emm there are people rather from europe or USA ?[02:27:58] <paul424> I will be gonig .... good night to the people in Europe and in the USA as well.[02:28:12] *** paul424 has quit IRC[02:28:45] *** jlongster has quit IRC[02:33:38] *** grindhold has joined ##OpenGL[02:34:23] *** reprore_ has quit IRC[02:35:20] *** pfo has quit IRC[02:40:10] *** HuntsMan has quit IRC[02:41:12] *** jfroy|work has quit IRC[02:48:40] *** ata2 has joined ##OpenGL[02:49:13] <ata2> hello there[02:49:15] <ata2> I have this issue: http://img210.imageshack.us/img210/9090/picture5do0.png[02:49:28] <ata2> as you can see backfaces of the GLU disks have a color, while I would like them to be black[02:49:45] <ata2> I set the color of the disks as: glMaterialfv(GL_FRONT, GL_AMBIENT_AND_DIFFUSE, gray);[02:49:55] <ata2> how do I set the color of the back of the disk?[02:50:30] <MatthiasM> http://www.opengl.org/sdk/[02:51:55] <ata2> uh?[02:52:27] <MatthiasM> click on reference pages[02:52:35] <Madsy> ata2: Uh.. GL_BACK?[02:52:56] <Madsy> This channel isn't a sustitute for thinking or doing research[02:53:01] <MatthiasM> Madsy: don't tell such obvoise things[02:53:07] <ata2> oh goddamit, I was doing GL_BACK but it was not working.. then I realized that I used the same color[02:53:18] <ata2> well, two different pointers but the content was the same[02:53:26] <Madsy> MatthiasM: Sorry :<[02:53:38] <Madsy> substitute*[02:54:55] <ata2> No, even doing that it doesn't help... look at this snippet of code[02:55:18] <ata2> http://rafb.net/p/kNFjJY71.html[02:55:52] <MatthiasM> ata2: did you read in the link I gave you ?[02:56:26] <ata2> the specification?[02:56:37] <MatthiasM> no - the man page[02:56:50] <MatthiasM> or reference page[02:57:07] <ata2> http://www.opengl.org/sdk/docs/man/[02:57:08] <MatthiasM> the spec is the pdf file and not so easy to read but also contains more information[02:57:15] <MatthiasM> yep[03:00:18] *** Eforen is now known as Eforen-AtSchool[03:02:55] <ata2> thanks for the link[03:03:05] <ata2> usually I google for information, I wasn't aware of that manpage[03:03:18] * MatthiasM points at the channel topic[03:03:48] *** Gorgoroth has quit IRC[03:06:22] *** Kasu| has quit IRC[03:06:45] *** androoid has quit IRC[03:12:06] *** Xmas| has quit IRC[03:16:53] *** e-roder has quit IRC[03:17:07] *** kbotnen has joined ##OpenGL[03:25:39] *** HuntsMan has joined ##opengl[03:33:39] *** jfroy has joined ##OpenGL[03:37:31] *** MatthiasM has quit IRC[03:37:38] *** MatthiasM has joined ##opengl[03:44:42] *** grindhold has quit IRC[03:45:14] *** grindhold has joined ##OpenGL[03:46:34] *** yates has quit IRC[03:48:13] <RTFM_FTW> "how can I use a depth texture?" can be answered by reading the GL_ARB_depth_texture specification[03:48:50] <RTFM_FTW> http://www.opengl.org/registry/specs/ARB/depth_texture.txt[03:53:13] *** ata2 has quit IRC[03:55:07] *** davidc__ has quit IRC[03:57:25] *** rnx has left ##opengl[04:00:45] *** m3ga has joined ##OpenGL[04:03:34] <m3ga> hi all. i'm on linux and i'm trying to figure out how to get the X window of one program to display on a texture in an openGL window. A bit like what a window managed like Beryl or Sqeak smalltalk does.[04:03:51] <m3ga> any clues about what i should look at?[04:08:00] <HuntsMan> GLX_EXT_texture_from_pixmap[04:08:43] *** grindhold has quit IRC[04:21:34] <m3ga> thanks HuntsMan. i'm googling that now[04:22:49] *** jcouture|away is now known as jcouture[04:26:22] *** mattn2|home_ has joined ##OpenGL[04:26:50] *** mattn2|home has quit IRC[04:29:39] *** jcouture has quit IRC[04:39:01] *** calav3ra_ has quit IRC[04:42:51] *** burak575 has quit IRC[04:44:21] *** b0000 has quit IRC[04:54:23] <m3ga> any opinions of GLX_EXT_texture_from_pixmap vs FBO?[04:56:52] <HuntsMan> they are unrelated[04:56:57] <HuntsMan> FBO allows rendering to textures[04:57:07] <HuntsMan> GLX_EXT_tfp allows binding X pixmaps to textures[04:57:17] *** davidc__ has joined ##opengl[04:59:46] <m3ga> thanks. also, according to glxinfo, my card doesn't have FBO[04:59:57] <HuntsMan> what card is that?[05:02:47] *** bijoo_appdev has joined ##OpenGL[05:02:48] *** bijoo_osdev has joined ##opengl[05:05:39] *** bijoo_appdev_ has joined ##OpenGL[05:05:39] *** bijoo_osdev_ has joined ##opengl[05:06:26] <m3ga> HuntsMan: ATI Technologies Inc RV370 [Radeon X300SE][05:06:59] <m3ga> It does have GLX_EXT_texture_from_pixmap[05:07:02] <HuntsMan> it should support FBO, probably the drivers don't[05:07:43] *** qeed has quit IRC[05:07:51] <m3ga> machine is running up-to-date ubuntu hardy.[05:08:13] <HuntsMan> fglrx driver?[05:12:27] <m3ga> no, not fglrx. This is my work devel machine. i hadn't taken much interest in OpenGL stuff until very recently[05:13:51] *** bijoo_appdev has quit IRC[05:14:38] *** mccbaka has joined ##OpenGL[05:15:04] <HuntsMan> m3ga: you're using the free driver, try the fglrx driver[05:15:27] *** bijoo_osdev has quit IRC[05:15:53] <mccbaka> Do things like glStencilFunc and glStencilOp get recorded in display lists?[05:17:21] *** mccbaka_ has joined ##OpenGL[05:18:47] <mccbaka_> Er... sorry, got disconnected[05:20:46] <m3ga> HuntsMan: yeah, I should :-). When I first installed this machine about 3 years ago, fglrx was really flakey so I went with the free driver.[05:30:00] *** stevosICBM has joined ##OpenGL[05:33:25] *** mccbaka has quit IRC[05:34:59] *** Yuzuru has joined ##OpenGL[05:39:50] *** Dagyde has joined ##OpenGL[05:41:40] <Madsy> http://www.flickr.com/photos/23579228@N04/2335016192/[05:41:53] <Madsy> So EVE is like the vi editor[05:46:24] *** servus_ has joined ##opengl[05:49:41] *** servus has quit IRC[05:49:45] *** servus_ is now known as servus[05:50:59] *** Dagyde has quit IRC[05:53:02] *** A2intaer has quit IRC[05:53:11] *** Dagyde has joined ##OpenGL[06:09:04] *** m4ggus has quit IRC[06:09:05] *** kbotnen has quit IRC[06:09:39] *** calav3ra has joined ##opengl[06:12:27] *** b0000 has joined ##opengl[06:16:31] <sparky> evenin[06:18:22] <furrywolf> heyas[06:20:00] *** m3ga has quit IRC[06:25:19] <Dagyde> GL/glut.h: No such file or directory. Where do I get that file, and associated header and library files? I have installed all of the mesa rpm's. Linux Fedora 10.[06:34:01] *** davidc__ has quit IRC[06:45:38] <Dagyde> I have freeglut installed[06:46:44] *** b0000 has quit IRC[06:50:26] <Dagyde> rpmfind.net reports "No package found for glut.h"[06:52:23] <Dagyde> yum only has freeglut[06:52:26] *** sohail has quit IRC[06:53:04] *** sohail has joined ##OpenGL[06:55:22] *** amz has quit IRC[06:59:18] *** calav3ra_ has joined ##opengl[07:04:06] *** boghog has quit IRC[07:05:46] *** Dagyde has left ##OpenGL[07:11:49] *** mccbaka_ has quit IRC[07:15:48] *** bijoo_appdev_ has quit IRC[07:15:48] *** bijoo_osdev_ has quit IRC[07:17:17] *** Yuzuru has quit IRC[07:17:33] *** davidc__ has joined ##opengl[07:20:36] *** calav3ra has quit IRC[07:23:52] *** stevosICBM has quit IRC[07:38:26] *** calav3ra has joined ##opengl[07:39:16] *** Quentarez has joined ##OpenGL[07:46:47] *** Shel2476 has joined ##OpenGL[07:46:56] *** Shel2476 has left ##OpenGL[07:51:47] *** calav3ra_ has quit IRC[07:54:23] *** mlucassmith has quit IRC[07:59:04] *** mccbaka has joined ##OpenGL[08:25:39] *** tokemonstah has joined ##OpenGL[08:27:16] *** scai has joined ##opengl[08:29:42] <TheFlash> Term "learning curve" is misused.[08:30:11] <acetoxy> Where?[08:30:23] <TheFlash> Everywhere.[08:30:30] <TheFlash> Or almost[08:30:45] <TheFlash> In this pic also 06:42:04 Madsy> http://www.flickr.com/photos/23579228@N04/2335016192/[08:31:21] <TheFlash> If learning curve is steep, it's not a bad thing. It means it is easy to learn.[08:33:10] <acetoxy> That depends on the graph, doesn't it?[08:34:19] *** boghog has joined ##opengl[08:34:27] <TheFlash> Of course.[08:35:40] <TheFlash> But generaly, if you have skills/knowledge in vertical axis and time spent learning in vertical axis. Steep means fast learning.[08:36:29] <TheFlash> time spent in horizontal.[08:36:45] *** neoneye has joined ##OpenGL[08:39:34] <acetoxy> Oh no[08:39:37] <acetoxy> One of the bosses is here[08:39:55] <acetoxy> I haven't seen him since I went on vacation for two weeks[08:44:19] *** tokemonstah has quit IRC[08:48:59] *** scai has left ##opengl[08:50:00] *** m3ga has joined ##opengl[08:50:38] *** mccbaka has quit IRC[08:56:32] *** mccbaka has joined ##OpenGL[08:56:42] <Jupp3> acetoxy: Here? On the same channel? :)[08:59:05] *** Bollinger has joined ##OpenGL[09:05:54] <acetoxy> No, here at work.[09:06:10] <acetoxy> But I got rid of him quickly :)[09:06:14] <acetoxy> He's pretty annoying :)[09:09:58] *** korff_home has quit IRC[09:13:03] *** kornerr has joined ##OpenGL[09:13:08] <kornerr> hi[09:13:15] <kornerr> is 'shader lighting' (if such exists) produced by shaders only without actual lighting?[09:15:10] *** [AD]Turbo has joined ##OpenGL[09:15:53] <TheFlash> kornerr: Err?[09:15:54] <[AD]Turbo> yo[09:16:04] <kornerr> what err?[09:16:38] <TheFlash> Can you be more specific? What do you mean?[09:17:11] <kornerr> the usual gl lighting is not shader lighting, right?[09:17:27] <kornerr> usual being the one consisted of 8 sources[09:17:33] <TheFlash> Well...[09:18:50] <TheFlash> In modern hard ware, the gl fixed pipeline is a shader (or many of them)... But yes, it is considered to be a different.[09:19:13] <kornerr> I mean is it only shaders that alter the color of a displayed texel so light doesn't really exist[09:19:23] <kornerr> ?[09:19:49] <TheFlash> It depends your shader.[09:20:30] <TheFlash> If your shader has anything to do with lighting or lights, of course light(s) exist![09:20:41] <kornerr> hm[09:21:00] <kornerr> I mean when you need more than 8 lights[09:21:20] <TheFlash> You can have more than 8 lights with fixed pipeline also.[09:21:39] <kornerr> I remember lights were restricted to 8[09:21:51] <kornerr> limited[09:21:58] <TheFlash> Yes.[09:22:03] <TheFlash> But you can multipass.[09:22:19] <kornerr> how is it done nowadays?[09:22:46] <kornerr> shaders only?[09:22:47] <TheFlash> With shaders, it is also faster to do multi-pass render than using gazilllion lights.[09:23:02] <TheFlash> kornerr: You can do it with fixed pipeline also.[09:23:53] <TheFlash> kornerr: Just render your "ambient" then, alter depth test to GL_EQUAL and change to additive blending and render everything again with each light.[09:27:09] *** TheLorax has joined ##opengl[09:28:20] <Jupp3> And isn't that 8 the "minimum amount of lights that have to be supported" anyway?[09:28:29] <Jupp3> Well no idea really how many drivers usually support[09:29:06] <Jupp3> In OpenGL, I guess it's safe to say that light never really exists :-)[09:29:38] *** UUncia has joined ##OpenGL[09:29:51] <Jupp3> kornerr: iirc, you can glGet() max amount of lights[09:30:30] <kornerr> that would be a funny game if different players hafe different picture ))[09:31:22] *** korff_home has joined ##OpenGL[09:40:58] *** maxton has quit IRC[09:41:36] *** Nescafe has joined ##OpenGL[09:43:19] *** marenz_ has joined ##OpenGL[09:44:01] *** ewn has joined ##OpenGL[09:46:05] *** Aintaer has joined ##OpenGL[09:47:17] <kornerr> what does 'state' mean in the sentence from ogl3.0 specs: 'The state required to support polygons consists of at least two processed vertices'?[09:48:01] <ewn> Is there an extension that makes it possible to work with texture coordinates outside of the range [0 1]?[09:48:10] <kornerr> (page 20)[09:51:54] *** x0rx0r has joined ##OpenGL[09:51:56] <Jupp3> ewn: Do you need an extension for that anyway?[09:52:34] <Jupp3> You can do that, but I didn't remember you would need an extension for that[09:53:05] <ewn> Not really. But it would make it easier for me and the code would probably be easier to understand[09:53:12] <MatthiasM> ewn: read in the red book (see channel topic) about texturing[09:53:36] * kornerr thinks 'state' was for ogl implementations[09:54:11] *** Akselii has quit IRC[09:54:53] *** sohail has quit IRC[09:55:27] <Jupp3> ewn: http://msdn.microsoft.com/en-us/library/ms970772.aspx has some example about it, and I can't really see any extension being used...[09:55:45] <Jupp3> ewn: Also remember to set texture repeat / clamp which ever way you want[09:56:09] <TheFlash> kornerr: It's not the implementation.[09:56:38] <Jupp3> Scroll down to the end of the page for the example[09:56:46] <kornerr> I mean OGL implementers[09:56:47] <TheFlash> kornerr: It's something you can alter by doing glSomething().[09:56:52] <kornerr> hm[09:57:04] <kornerr> well[09:57:24] <Madsy> ewn: http://opengl.org/registry/specs/ARB/texture_rectangle.txt[09:57:25] <kornerr> I don't need to keep track of any states to draw a polygon[09:57:37] <TheFlash> glEnable(GL_LIGHTING); // This alters the state.[09:57:47] <kornerr> I meant the other state)[09:57:53] <kornerr> the one in the sentence[09:58:02] <Madsy> ewn: It's a new texture target[09:58:06] <TheFlash> It is just that what it meant.[09:59:32] <MatthiasM> kornerr: this sentence assume immediate mode - multiple glVertex calls - you need to remember atleast 2 other glVertex calls (depends on mode) to create a triangle[09:59:46] <TheFlash> kornerr: State in OpenGL is a bunch of internal variables that affects the rendering.[09:59:46] <kornerr> yeah[09:59:57] <kornerr> (yeah to mathias)[09:59:58] <kornerr> )[10:00:01] *** maxton has joined ##OpenGL[10:00:14] <kornerr> I mean that 'state' was not for me to alter[10:00:26] <kornerr> but for ogl implementation[10:01:03] <MatthiasM> you alter it calling glVertex[10:01:18] <ewn> Thanks for the links. Just to clarify what functionality I'm looking for. I'm rendering an earth sphere. I want to use the geographic coordinates as texture coordinates (x in range [-180 180], y in range [-90 90]).[10:01:23] <kornerr> yeah[10:01:26] <kornerr> but I don't keep state[10:01:26] <TheFlash> Or by any other gl-function.[10:01:53] <TheFlash> Not any (glGet*) but...[10:01:54] <MatthiasM> ewn: a) write what you want to know together with your question b) that doesn't work[10:01:55] <kornerr> I alter it, ok[10:02:11] <kornerr> I alter, but do not keep it )[10:02:29] <Jupp3> ewn: And now you tell us :)[10:02:40] <MatthiasM> ewn: http://mikeash.com/getting_answers.html[10:02:41] <Jupp3> ewn: Remember that you can apply glScale etc. to texture matrix aswell[10:02:44] *** scai has joined ##opengl[10:02:56] <Jupp3> Which is probably what you want[10:03:07] <MatthiasM> or use a shader to implement the coordinate mapping[10:04:12] <ewn> Ok. Thanks for the help![10:05:19] *** pietia has joined ##OpenGL[10:07:58] *** electrostatic has joined ##OpenGL[10:12:59] *** Renderwahn has joined ##OpenGL[10:16:07] *** x0rx0r has quit IRC[10:16:15] *** Ingenu has joined ##OpenGL[10:17:27] *** x0rx0r has joined ##OpenGL[10:17:43] *** mm765^sleep is now known as mm765[10:18:49] *** Spkka has joined ##OpenGL[10:19:40] *** Renderw has joined ##OpenGL[10:20:14] *** Renderwahn has quit IRC[10:21:16] <electrostatic> hello, quick question... I am having problems figuring out how to ensure that a PBO is done reading pixels before continuing. I checked here, and it wasn't much help in the matter. http://opengl.org/registry/specs/ARB/pixel_buffer_object.txt[10:21:33] <electrostatic> There is one line in one of the examples that mentions that if you map a buffer it should wait for the PBO to finish, but even when I map mine I have this problem...[10:21:33] *** GuShH has joined ##OpenGL[10:22:05] <electrostatic> Code: http://pastebin.com/m7c0366e6 The only way I have found around it is to put that - int i; for (i=0; i<10000000; i++); - to waste time, otherwise no matter what I do I risk ending up with only a partial data transfer before the new textures are written to.[10:22:12] <electrostatic> Any ideas? (FYI, I am using OS X 10.5 Radeon drivers with an RV360M11 chipset (9700 mobility), but It is supposed to be able to handle PBO's.[10:25:23] *** Renderwahn has joined ##OpenGL[10:26:26] <electrostatic> here is what happens when I do not arbitrarily wait, regardless of whether or not I us glFinish or glMapBuffer... http://incp.us/img/lines.png[10:26:31] <quicksilver> electrostatic: the whole point of PBOs is that the read can be done asynchronously.[10:26:40] <electrostatic> When I do wait, those black lines are not there[10:26:42] <quicksilver> electrostatic: if you don't want an asynchronous read, do not use a PBO.[10:26:57] *** mccbaka has quit IRC[10:27:02] <electrostatic> quicksilver: Yes, that is one of the main reasons, and I do like that[10:27:18] <electrostatic> but what I want to know is simply how to be sure it is done[10:27:48] <electrostatic> (If you notice, I start the second read before the first one is done, as you said, asynchronously...)[10:28:33] <electrostatic> And also, I would like to be able to fill the first tex while the second is still reading, but I thought I would try to break it down and figure this out first[10:28:59] *** walt has quit IRC[10:29:27] *** Renderw has quit IRC[10:30:18] <quicksilver> mapping the buffer with glMapBufferARB will wait for any async writes on that buffer to finish[10:30:18] *** mccbaka has joined ##OpenGL[10:30:26] <quicksilver> as the spec describes[10:30:26] *** mccbaka has quit IRC[10:30:42] *** tac has left ##OpenGL[10:30:57] <electrostatic> Meh, then my driver must have a bug, because when I uncomment those lines it still does it[10:32:07] <quicksilver> well, even if you uncomment those lines you're only mapping one buffer[10:32:12] <quicksilver> the other might still not be complete[10:33:25] <electrostatic> It's that one that I am testing, and believe me, I have tried both, up to the point of absurdity, actually[10:37:42] *** Walt has joined ##opengl[10:39:21] <electrostatic> quicksilver: In this one, you can actually see the map code in each of them behind the window. It isn't quite as bad, because it had to waste some time doing that, but it is still there... http://incp.us/img/lines2.png[10:45:34] *** dvoid_ has joined ##OpenGL[10:47:57] <quicksilver> hmm[10:48:03] <quicksilver> reading the spec again it's slightly ambiguous now[10:48:12] <quicksilver> "Process partial images" is the comment[10:48:25] <quicksilver> maybe it *is* supposed to let you map even when reading isn't complete.[10:48:40] *** aucahuasi has joined ##OpenGL[10:48:45] <electrostatic> That would be highly odd[10:48:55] <electrostatic> "Place your bets!"[10:48:56] <aucahuasi> hi all[10:49:04] <electrostatic> 'ello[10:51:12] *** TheLorax has quit IRC[10:51:42] <aucahuasi> does anyone have experience plotting 3d math surfaces?[10:52:14] <electrostatic> And glGetBufferParameteriv doesn't even seem to have a query for "is busy", or I could work around it with a dirty little spin lock...[10:52:44] <quicksilver> electrostatic: I think you'll have to wait until RTFM_FTW wakes up.[10:52:47] <quicksilver> He will know the answer :)[10:53:41] <electrostatic> ok, thanks[10:53:52] <electrostatic> =)[10:54:54] <quicksilver> electrostatic: the only code I've seen which uses PBOs to read always doesn't process until the next frame.[10:55:05] <aucahuasi> to calculate the intersections between surfaces, which dta struct should I use? heightfield or polygonal mesh?[10:55:08] <quicksilver> (in which case I imagine swapping buffers will wait for the read if that's necessary)[10:55:15] <aucahuasi> ... please some ? ;-)[10:55:35] <quicksilver> intersections between heightfields is much easier.[10:55:45] <quicksilver> but not all surfaces can be represented as heightfields.[10:57:03] <aucahuasi> thanks ;) quicksilver[10:57:22] <aucahuasi> quicksilver, yes I know that, that's why I'm thinking about using polymesh[10:58:07] <electrostatic> quicksilver: ya, I know what you mean.. if you look at the code I was trying to use it to unpack YUYV data in the graphics card without relying on heavy shader code or mapping to client address space.. Just an idea I got, and if I could figure this out I'm pretty sure it would work...[10:59:19] <aucahuasi> plots in the form f(x,y) would be easy in heightmaps, but surfaces in spherical coordinates or meta surfaces ? ... :P[10:59:48] <electrostatic> I might rewrite it with just Shader -> FBO -> Texture and only use PBO's to pipeline frames, but this seemed so close to working...[11:01:56] <electrostatic> Although come to think of it, I would still need a PBO in the Y plane, but I could start that going and do both the U and V while it was reading, but that still wouldn't be what I would call "stable"[11:03:11] <electrostatic> I mean who would say, "Hey, lets have a worker thread do something incredibly useful for them, and then give them no way to figure out when it is done."[11:03:22] <electrostatic> *snicker[11:03:28] <quicksilver> electrostatic: agreed.[11:04:28] <electrostatic> anyway, 2 AM here, gtg[11:04:29] *** pietia has quit IRC[11:04:36] <electrostatic> have fun =)[11:05:27] *** electrostatic has left ##OpenGL[11:05:43] <aucahuasi> 2 am :) .. 5am here XD[11:05:51] <hackkitten> 11 AM here ^-^[11:11:06] <aucahuasi> is it right to use a polygon mesh to draw math surfaces? Is there any better struct?[11:11:22] <quicksilver> I can't think of a sensible way to answer that question.[11:11:31] <quicksilver> In opengl the only thing you can draw is polygons.[11:11:33] <aucahuasi> go ;)[11:11:46] <quicksilver> so of course you will have a polygonal mesh, at some level.[11:14:40] <aucahuasi> I understand ;)[11:15:31] <aucahuasi> thanks :)[11:17:30] *** cplusplus has joined ##OpenGL[11:23:44] *** Spkka has quit IRC[11:25:14] *** aucahuasi has left ##OpenGL[11:30:42] *** marenz_ has quit IRC[11:30:45] *** emzic has joined ##OpenGL[11:43:13] *** dvoid_ has quit IRC[11:47:05] <exDM69> http://qaq4.kyla.fi/~riku/goodenough/stencil.png <<--- I did some tests on filling convex polygons using the stencil buffer[11:47:09] <exDM69> works like a charm[11:47:29] <Ingenu> what for ?[11:47:56] <exDM69> I thaught about using that to render text[11:48:27] <quicksilver> exDM69: something's wrong. triangle 2 is much larger on the left view than the right ;)[11:48:33] <exDM69> and I also wanted to try out this texture[11:49:03] <exDM69> s/texture/technique[11:49:22] <exDM69> quicksilver: oh, so it seems. they must be from before and after I hacked with the polygon vertices[11:50:00] <Ingenu> you should fear overdraw[11:50:11] <quicksilver> to anti-alias this technique you would have to render "too large" and then scale down, I think.[11:50:13] <Ingenu> it's not something you like, wasting valuable fillrate when you have another way[11:50:15] <exDM69> http://qaq4.kyla.fi/hg/goodenough/file/dbf82b874a99/test/stencil.cpp <<-- here's the code[11:50:23] <quicksilver> no obvious way to add antialiasing in situe[11:50:50] <Ingenu> alpha to coverage & MSAA comes to mind[11:51:25] <Ingenu> and that new DX10.1 feature, forgot what it was exactly, gives nice antialiased sprites[11:52:51] <exDM69> Ingenu: why should I fear overdraw?[11:53:14] <exDM69> is there an issue with this?[11:54:00] <exDM69> I figure that I will write another renderer that uses tesselated polygons[11:56:34] <exDM69> this technique could be combined with geometry shaders to get the edges extruded into quads easily[11:57:40] *** rabbit- has joined ##OpenGL[11:58:20] <Ingenu> I don't know exactly what you intend to do[11:58:46] <Ingenu> but stencil shadow volumes were bad because of the vertex load and huge fillrate cost[11:59:14] <exDM69> this technique is similar to stencil shadow volumes, yes[12:00:19] <exDM69> but I still don't see a problem[12:00:25] <exDM69> perhaps it's a bit fillrate intensive[12:00:28] <Ingenu> yeah but it depends on the scale at which you apply it[12:00:35] <exDM69> but drawing a little bit of text can't really be the killer[12:00:41] <Ingenu> if you fill a small buffer or a small part of the screen it might be just fine[12:00:58] <Ingenu> but the whole screen like Stencil Shadow Volumes...[12:01:21] <exDM69> Ingenu: note that stencil shadow volumes do that for every light in the scene[12:01:49] <Ingenu> I don't know, it sounds like a fun experiment, but why would you want to use that instead of an alpha tested/blended texture ?[12:02:11] <exDM69> Ingenu: this is for rendering text from outlines[12:03:59] <exDM69> my intent is to do several views/renderers. outline using line strips, tesselated triangles and a stencil buffer view[12:04:24] *** hang3r has joined ##OpenGL[12:04:34] <quicksilver> I keep wondering about an efficient way to render curved 2D geometry[12:04:35] <exDM69> I don't have a tesselator at hand and I want to hack with the stencil buffer so this was the better one to get started with[12:04:40] <quicksilver> vector graphics.[12:04:47] <quicksilver> I think this is closely related to exDM69's idea.[12:05:19] <quicksilver> although one of my requirements was it be antialiased.[12:05:42] <exDM69> quicksilver: I learned about this technique from an NVIDIA paper, they were using this particular technique for high perf vector graphics on opengl es[12:05:48] <exDM69> so I think that it can't be that bad[12:05:57] <quicksilver> that's interesting.[12:06:13] <quicksilver> I was thinking of doing AA with textures like that website someone pasted here[12:06:15] <quicksilver> which I now can't find[12:06:17] <exDM69> (at least I think that they use this technique, I reverse engineered this from the bits of information I heard)[12:06:38] <quicksilver> ah, this one :[12:06:39] <quicksilver> http://homepage.mac.com/arekkusu/bugs/invariance/TexAA.html[12:09:18] <exDM69> intresting technique[12:09:23] <quicksilver> however, if you tesselated and then AA'ed as if it was triangles[12:09:27] <quicksilver> you'd see the edges.[12:09:33] <quicksilver> that would suck.[12:09:41] <quicksilver> you'd have to either tesselate VERY finely[12:09:44] <quicksilver> (probably stupid)[12:09:57] <quicksilver> or tesselate and then fit a *curved* edge to it[12:10:06] <quicksilver> like an appropriate scaled fragment of a sphere.[12:10:29] <quicksilver> you could have a single texture of an antialiased circle[12:10:37] <quicksilver> and then scale it anisotropically[12:10:44] <quicksilver> to the correct curvature ?[12:11:05] <quicksilver> assuming you had an efficient way to calculate or approximate curvature at each point.[12:12:11] *** scai has left ##opengl[12:19:18] *** Suprano has joined ##OpenGL[12:22:21] *** m3ga has quit IRC[12:27:36] *** maxton has quit IRC[12:27:37] *** LordHavoc has quit IRC[12:27:37] *** tmccrary1 has quit IRC[12:27:37] *** turbo24prg has quit IRC[12:27:37] *** Jupp3 has quit IRC[12:27:37] *** |chiz| has quit IRC[12:29:12] *** maxton has joined ##OpenGL[12:29:12] *** LordHavoc has joined ##OpenGL[12:29:12] *** tmccrary1 has joined ##OpenGL[12:29:12] *** |chiz| has joined ##OpenGL[12:29:12] *** Jupp3 has joined ##OpenGL[12:29:12] *** turbo24prg has joined ##OpenGL[12:31:16] *** wisey has joined ##OpenGL[12:33:53] *** XT95_ has joined ##OpenGL[12:41:34] *** LordMetroid has joined ##OpenGL[12:41:46] *** elite01 has joined ##opengl[12:47:14] *** Jernej has joined ##OpenGL[12:48:34] *** JernejL has quit IRC[12:49:41] *** KU0N has joined ##opengl[12:56:26] *** rnx has joined ##opengl[13:04:04] *** eXtronuS has joined ##OpenGL[13:05:33] *** HugoMaxx has joined ##OpenGL[13:18:05] <exDM69> http://qaq4.kyla.fi/~riku/goodenough/stencil.png <<-- fixed[13:18:14] *** rabbit- has quit IRC[13:18:41] <quicksilver> exDM69: what kind of font defintions are you hoping to work with?[13:18:47] <quicksilver> truetype? PS? some format of your own devising?[13:19:05] <exDM69> quicksilver: freetype[13:20:52] * quicksilver nods[13:21:00] <quicksilver> how are you going to turn the curves into polygon segments?[13:21:12] <quicksilver> does freetype already have a routine to do roughly that?[13:22:06] <exDM69> last time I did it, I just did simple interpolation routines myself[13:29:55] * quicksilver nods[13:30:07] <quicksilver> hmm[13:30:21] <quicksilver> interpreting the bezier curve yourself?[13:31:20] *** sohail has joined ##OpenGL[13:33:30] *** predaeus has joined ##opengl[13:33:52] *** jcouture has joined ##OpenGL[13:35:05] *** wisey has quit IRC[13:35:13] *** LtJax has joined ##OpenGL[13:42:09] *** kornerr has left ##OpenGL[13:42:48] *** ol1veira has joined ##OpenGL[13:56:08] *** NevroPus has joined ##OpenGL[13:59:58] <exDM69> quicksilver: yes[14:00:16] <quicksilver> cool.[14:00:25] <quicksilver> very similar to what I'm considering then.[14:00:48] <exDM69> http://glttf.sourceforge.net/ <<--- this is what I came up with last time[14:01:26] <quicksilver> cool![14:03:33] *** itewsh has joined ##OpenGL[14:04:30] <exDM69> quicksilver: hah, I just thaught of an intresting property this rendering method has[14:04:42] <exDM69> if you look at polygons from the back side, they will appear negative[14:04:52] <exDM69> or depending on the winding rule they might[14:08:16] <quicksilver> yes, you will have to be careful of such things[14:09:52] <exDM69> I am just hacking together a version with many contours and winding rules[14:23:06] <exDM69> quicksilver: btw, did you take a look at the source code?[14:24:24] *** hang3r has quit IRC[14:25:50] <quicksilver> no, haven't had a chance[14:25:56] <quicksilver> sadly I have real work to do too ;)[14:29:55] <exDM69> just tell me how you like it if you get a chance to look at it[14:33:11] *** jimi_hendrix has joined ##OpenGL[14:33:31] <jimi_hendrix> hi...if i want to make a 2d game in OpenGL will i still need to use SDL for anything[14:33:41] <jimi_hendrix> or can OpenGL take care of keypresses and stuff[14:34:47] *** eXtronuS has quit IRC[14:40:54] <tmccrary> OpenGL is purely for drawing things on the screen[14:41:04] <tmccrary> SDL will be your input layer[14:41:07] <jimi_hendrix> ok thanks[14:41:24] <tmccrary> You should check the red book and perhaps the superbible[14:41:49] <tmccrary> You can check out an old version o fhte red book that's freely online, or buy a recent version[14:48:22] *** elektr1k_ has joined ##opengl[14:51:11] *** scai has joined ##opengl[14:53:56] <jimi_hendrix> red book?[14:54:16] <jimi_hendrix> and its libopenGL correct...or libgl?[14:55:01] <GuShH> jimi_hendrix : look at the topic[14:55:08] *** groton has joined ##OpenGL[14:55:09] *** johndoe has joined ##opengl[14:55:47] <jimi_hendrix> ok that solves redbook question[14:56:07] <Jupp3> Yeah, it's not the chinese one[14:56:17] <jimi_hendrix> ok and whats the lib called[14:56:47] <Jupp3> libgl[14:56:55] <Jupp3> if you mean how to link it, which is -lgl[14:58:17] <jimi_hendrix> na i wanted the name[14:58:23] <jimi_hendrix> trying to find it in cygwin installer[14:58:58] <jimi_hendrix> ok not there...so i should compile from source?[15:00:24] *** calav3ra has quit IRC[15:00:35] *** elektr1k has quit IRC[15:00:37] <jimi_hendrix> nvm found it in a diff section[15:00:47] *** kbotnen has joined ##OpenGL[15:02:58] *** kenws has joined ##OpenGL[15:03:41] *** saTTY has joined ##OpenGL[15:04:20] <saTTY> guys ! i am not able to run ShaderDesigner[15:04:26] <saTTY> why ??[15:04:29] <saTTY> plz help[15:05:04] *** JernejL_ has joined ##OpenGL[15:07:01] <saTTY> anybody alive here ?[15:07:35] <scai> of course not[15:07:37] <LtJax> no, everyone died over the holidays[15:07:41] <jimi_hendrix> so how should i link it?[15:07:47] *** elektr1k_ has quit IRC[15:07:47] <Jupp3> What do you expect us to say? "You can't run it becouse of X"?[15:07:51] <kbotnen> what is "not able to run"?[15:07:52] *** sysRPL has joined ##OpenGL[15:07:56] <sysRPL> hello[15:08:01] <saTTY> in windows[15:08:04] <quicksilver> sysRPL: probably because you don't have a computer.[15:08:06] <Jupp3> X being whatever reason we find with our 1337 mind reading skills[15:08:10] <scai> jimi_hendrix: -lGL[15:08:12] <quicksilver> or perhaps because you have run out of coal.[15:08:19] <saTTY> :p[15:08:30] <sysRPL> quicksilver: ??/[15:08:33] <saTTY> quicksilver, i am talking about windows[15:08:41] <jimi_hendrix> couldent find it[15:08:55] <jimi_hendrix> but -lopengl32 works[15:09:12] <scai> oh, windows, ok[15:09:36] <saTTY> after installation it refuse to run ![15:09:39] <kbotnen> saTTY, you have to explain better. what have you tried, what error if any do you get etc.[15:09:39] *** b0000 has joined ##opengl[15:10:18] <quicksilver> sysRPL: actually that was ment for saTTY, sorry ;)[15:10:55] <saTTY> i downloaded ShaderDesigner for windows ! ans installed ; but after installation it refuse to run[15:11:08] <jimi_hendrix> scai: so -lopengl32 is the same?[15:11:18] <scai> jimi_hendrix: i dunno, seems so[15:11:27] *** A2intaer has joined ##OpenGL[15:11:48] <GuShH> saTTY : that's hardly a mainstream application anymore...[15:12:05] <GuShH> for shader authoring I recommend any of the ati / nv tools[15:12:17] *** maxton has quit IRC[15:12:29] <GuShH> although I just use "shadermaker" which is the most simplistic thing ever, although you can't do anything advanced with it (mrts, etc)[15:12:30] <saTTY> any alternative ! GuShH[15:12:41] <GuShH> I just said... try ati's or nv's authoring tools[15:12:47] <GuShH> RenderMonkey, etc... christ, google it.[15:12:58] <sysRPL> i need an algorithm to tessellate ALL polygons, be they concave, convex, simple, or complex ... and the algorithm should handle different winding modes[15:13:00] <sysRPL> TIA[15:13:14] <GuShH> Good luck :)[15:13:26] <saTTY> thanks GuShH[15:13:29] <kbotnen> saTTY, worked fine here, dl, install and run. so now its up to you to provide more information about how you screwed up.[15:13:31] <kbotnen> :)[15:13:44] <GuShH> Perhaps it's a driver issue[15:15:15] *** Yuzuru has joined ##OpenGL[15:16:47] <saTTY> kbotnen, this message come to me " Shader Designer can't find multitexture support "[15:17:04] <saTTY> what i suppose to do next[15:17:05] <sysRPL> is that the typhoon labs app?[15:17:19] <saTTY> yeah ![15:17:24] <sysRPL> well[15:17:32] <sysRPL> are you on a laptop?[15:18:02] <saTTY> no on desktop[15:18:12] <sysRPL> well it's a driver issue[15:18:35] <sysRPL> write an opengl app and look for multitexture in the strings[15:18:36] <saTTY> what kind of driver ![15:18:49] <sysRPL> are you on windows?[15:19:00] <saTTY> yeah ![15:19:18] <sysRPL> i *could* send you a program to lsit your driver extensions[15:19:27] <sysRPL> or you could write you own program to do that[15:19:36] <saTTY> ok ![15:19:39] <sysRPL> or you could downlaod some other program to do it[15:19:43] <LtJax> or download the extension viewer..[15:19:44] <sysRPL> those are your choices[15:20:17] <sysRPL> what LtJax said[15:20:59] <saTTY> means sysRPL[15:21:18] <sysRPL> ?[15:21:31] *** batbrat has joined ##opengl[15:21:35] <LtJax> then again, that prolly wont solve his issue[15:21:41] <sysRPL> systemRPL = system reverse polish logic = a programming language[15:21:41] <LtJax> it might diagnose it[15:22:44] *** saTTY has left ##OpenGL[15:22:45] *** Jernej has quit IRC[15:23:43] <sysRPL> saTTY: what you probably need to do is download the latest driver for your video card (a non beta version), remove the hardware in the device manager, restart the pc, choose the VGA Adapater at next startup, restart, then run the driver install[15:25:09] <sysRPL> your current driver probably says Microsoft[15:25:55] <batbrat> I have been trying to understand the Redbook's polarView example since the past few days. With the help of the people on this forum, my grandad and some serious thinking, I understood it(contd...)[15:27:21] *** Aintaer has quit IRC[15:27:25] *** Aintaer has joined ##OpenGL[15:27:28] <batbrat> It turns out that while I understood the use of +azimuth in that example, I had misunderstood the order of rotation and translation (a total newbie mistake).[15:28:49] <batbrat> I have since gotten the order of translation and rotation straightened out, but I am still confused as to how it ends up showing the right view.[15:30:59] <batbrat> I have written some code to test it, and it works. Still I don't get it fully. So, I'll be grateful, if anyone can help me. I'll post my exact problem next. I'm sorry to tax everybody's patience, however, I am determined to learn things right.[15:32:02] *** A2intaer has quit IRC[15:32:29] <batbrat> Currently, the code goes like: glTranslated followed by three calls to glRotated.(For twist(about z) elevation(about x) azimuth(about z)).(contd...)[15:33:42] <batbrat> I take this to mean, interpreting w.r.t a global coordinate system that the camera is rotated about the grand origin and then translated along the grand z axis.[15:36:21] *** jm|afk has joined ##OpenGL[15:36:44] <Jupp3> batbrat: Why double?[15:37:03] *** Yuzuru has quit IRC[15:37:21] <Jupp3> Well of course it does make difference, in some cases not only "being slower" either[15:37:22] <batbrat> My problem is that this doesn't seem to be right.For instance, when the rotation is 180 degrees (azimuth), the camera would end up moving along the z axis and pointed away from the object to be viewed. However, this doesn't happen in the code.I'd be grateful if someone points out what I am doing wrong and explains the right way to look at things.[15:38:05] <batbrat> Jupp3: I am just using the code in the redbook. I could have used float, but I just used double because the redbook did.[15:38:44] <batbrat> Did my previous post appear fully?[15:39:07] <batbrat> Jupp3: Sorry, but what would make a difference? float vs. double?[15:39:23] <batbrat> *The post on 180deg rotation?[15:42:34] <batbrat> *** Some posts ago I said forum. Sorry, it should be channel :)[15:43:04] <Jupp3> Well I guess many implementations use float internally[15:44:14] <batbrat> Jupp3: Ok. Thanks. What do you think, is my interpretation wrong or am I misunderstanding?[15:45:11] <Jupp3> There is no movable camera in OpenGL[15:45:19] <Jupp3> BUT there are functions that will fake one for you[15:46:00] <sparky> mornin fuckers[15:46:12] <sparky> I missed you[15:47:00] <batbrat> Agreed. Thats what the code I use does. It does the reverse movements to everything else. I am rewriting the redbook code.[15:47:19] <batbrat> *I mean I am just writing the same code thats in the redbook.[15:48:20] <Jupp3> if you do glTranslate(), you move "everything you will draw after that", NOT the camera[15:48:24] <batbrat> sparky: Good morning. I don't mean to irritate you or anyone else. If I should do thing s differently, do tell me. I'd be happy to learn.[15:48:41] <Jupp3> But you know, when you move room away from the camera, it appears as if the camera was moving backwards[15:48:46] <batbrat> *things => with regard to behaviour on the forum.[15:48:56] <batbrat> Agreed Jupp3.[15:49:26] <batbrat> Still it seems to me that the code means something diffrent. I'll post it if you like.[15:49:32] <sparky> batbrat: I gave this very lecture to caine only about a week ago :)[15:49:53] <Jupp3> batbrat: Also remember that with glRotate, the 3 "one-or-zero" arguments actually define the rotation angle[15:50:03] <Jupp3> So you can define any angle around which to rotate[15:50:23] <sparky> batbrat: I don't use opengl for any of this crap[15:50:29] <sparky> I just create my own input matrices[15:50:33] <Jupp3> Not related to your current problem probably, but that can be useful sometimes[15:50:38] <sparky> pass the 16 floats per matrix[15:50:41] <sparky> to my shader[15:51:05] <batbrat> sparky: I haven't learnt shaders yet. I should do that post-haste.[15:51:06] <sparky> that was not the lecture though[15:51:23] <sparky> I explained to him about transformation hierarchies[15:51:33] <batbrat> Jupp3: I follow that I can define the axis and angle. I don't fully get the meaning though.[15:51:51] <sparky> which will de-mystify the issues of cameras relative to world[15:52:26] <batbrat> sparky: The world is the topmost node in the transformation hierarchy. I've benn told. But I didn't understand :(. Sorry[15:53:15] <batbrat> Does it mean that the camera rotates and not the world. Even so, I still have the same question. Wont the camera get messed at 180?[15:53:41] *** jm|earth has quit IRC[15:53:45] <Jupp3> The camera doesn't rotate[15:53:53] <batbrat> Jupp3: Can you please explain what you mean by 0ne or zero arguments?[15:53:57] <Jupp3> But when you rotate the world, it can look like the camera rotated instead[15:54:27] <Jupp3> batbrat: The 3 numbers define a 3D unit vector (length==1) around which the rotation will happen[15:54:39] <batbrat> Jupp3: Ok. Thats what I thought. As of now, I don't get the transformation hierarchy.[15:54:42] <Jupp3> So they are not used to "just pick x, y or z"[15:55:12] <batbrat> Ok. That I get. The axis is the unit vector from the origin to point x,y,z specified in the call.[15:55:33] *** stringfellow has quit IRC[15:55:40] <sparky> batbrat: for now I'll let jupp3 help you with your current issue. So you don't have two people feeding you advice[15:56:32] *** rorokimdim___ has joined ##opengl[15:56:39] <Jupp3> batbrat: Anyway, the hierarchy... Let's say you have a camera, room, and people moving around the room[15:56:44] <batbrat> sparky: I am sorry to impose on you. But can you tell me the transformation hierarchy thing? I promise once I get it I'll put it on my blog along with an explanation of all the stuff I learnt. So people can look it up there.[15:57:05] <batbrat> I've given up after google searches didn't yeild many explanations.[15:57:10] *** rorokimdim___ has quit IRC[15:57:18] <sparky> batbrat: it's not really that complicated so I'll give it a go[15:57:25] <batbrat> sparky: As far as two people helping, I am glad to have all the help I can get.[15:57:30] <Jupp3> So first you move the room so that it matches the camera movement / rotation. Then you store the matrix (glPushMatrix()) transform to the location of first moving thing, draw it, do glPopMatrix(), and go on to the next one[15:57:38] <sparky> let's say you have a collection of objects obj[i] indexed by i[15:57:54] <sparky> and let's say they are all children of the world node[15:57:56] <sparky> this is a simple case[15:58:14] <sparky> so each one has a transformation that will transform it's vertices into the world[15:58:19] <sparky> fair enough right?[15:58:26] <batbrat> Fair enough :)[15:58:37] <sparky> We'll call the transformatino Mobj_to_world[i][15:58:47] <batbrat> Ok[15:58:56] <sparky> now the point is, the same goes for other things, such as a camera or even a light[15:59:33] <batbrat> Right so everything gets positioned by its transform into world space?[15:59:34] *** maxton has joined ##OpenGL[15:59:37] <sparky> So let the camera also be a child of the world in which case Ccam_to_world[16:00:11] <batbrat> Ok[16:00:13] <sparky> but to render you have to pass something called the modelview matrix[16:00:13] *** stringfellow has joined ##opengl[16:00:19] <sparky> with me so far?[16:00:31] <batbrat> With you so far. :)[16:00:42] <sparky> the model view matrix takes vertices from the object and tranforms them into the camera coordinate system[16:01:07] <batbrat> Ok :)[16:01:22] <sparky> now the transformations we have will move us up through the tree. But to get down you have to invert a matrix[16:01:56] <sparky> for instance to get from world to camera you need Cworld_to_cam = Invert(Ccam_to_world)[16:02:15] *** dvoid_ has joined ##OpenGL[16:02:29] <batbrat> Ok. What is inverting? Using opposite translations and rotations i.e. with minus signs in the calls, roughtly?[16:02:36] <batbrat> *roughly.[16:03:11] <sparky> because: Mmv = Cworld_to_cam * Mobj_to_world[i] = Invert(Ccam_to_world) * Mobj_to_world[i][16:03:14] <quicksilver> that's how you invert translations and rotations, yes.[16:03:20] <sparky> no, wait with the inverting[16:03:23] <sparky> I'll get to that[16:03:26] <quicksilver> inverting a general matrix is hard.[16:03:33] <sparky> for now, just accept that you can get one[16:03:35] <quicksilver> but it doesn't matter, ot understand what sparky is saying.[16:03:43] <sparky> I want you to understand the concept initially[16:03:51] <batbrat> Ok. I'll accept it for now.[16:03:55] *** bijoo_appdev_ has joined ##OpenGL[16:03:55] *** bijoo_osdev_ has joined ##opengl[16:04:10] <sparky> now you could also have deeper hierarchies[16:04:18] *** UUncia has quit IRC[16:04:26] <Jupp3> batbrat: Also I recommend you to do some kind of object / model system from the start[16:04:31] <sparky> with objects as children of other objects or even cameras as children of other objects which are children etc[16:04:39] <batbrat> Jupp3: Thanks. Will do.[16:04:45] <Jupp3> batbrat: object defining the position, orientation and a model for itself[16:04:49] <sparky> but when you want to see the view from a certain camera in your scene[16:05:08] <sparky> you have to create the transformation sequence all the way down to the camera[16:05:12] <Jupp3> While model doesn't care about orientation and position at all, except for relative positions for all vertices from the object origo[16:05:29] <sparky> and to get up, you use the assigned matrix, to get down you invert. With me?[16:05:35] *** flazz has joined ##OpenGL[16:05:55] <batbrat> Not with this bit.One mo while I reread the posts. please.[16:07:11] <WhatAHam> question about blending:[16:07:22] <sparky> batbrat: let me know what you are ready[16:07:40] <batbrat> sparky: Will do. :)[16:07:58] <sparky> the main thing is, apply the matrix when moving up and use the inverse when moving down[16:08:11] <WhatAHam> if i have a source texture with colors: 1,1,1,.5 and i blend it with a background using glBlendFunc(GL_SRC_ALPHA, GL_SRC_ALPHA_MINUS_ONE), is it normal for the result to be somewhat grayish?[16:08:39] <WhatAHam> i am assuming the .5 alpha results in a sort of grayening of the resulting fragments?[16:08:41] <MatthiasM> WhatAHam: yes[16:08:58] <MatthiasM> WhatAHam: you basicly get 50% of the backgropund and 50% of the white color[16:09:06] <WhatAHam> which can look gray, depending, right?[16:09:16] <MatthiasM> or your texture / white[16:09:20] <MatthiasM> yes[16:09:33] <WhatAHam> ok that explains why my clouds are grayish :)[16:09:43] <MatthiasM> if you upload both texture I can check[16:09:47] <WhatAHam> ok[16:09:55] <WhatAHam> where should i send them to?[16:09:57] <WhatAHam> (thanks btw!!!)[16:10:06] <MatthiasM> look at the channel topic[16:10:09] <MatthiasM> or use imagebin[16:10:18] <WhatAHam> ah heh[16:10:18] <batbrat> sparky: Moving up means positioning child node in the world. Moving down is what? I don't follow.[16:10:43] <sparky> bat: ok, let me try that part again[16:10:49] <sparky> in a tree you have nodes[16:10:53] <sparky> the world is the top node[16:10:53] <batbrat> Ok thanks. :)[16:11:02] <sparky> and the objects, cameras and lights are nodes[16:11:19] <batbrat> Ok.[16:11:22] <sparky> typically the matrices are stored such that they take a child node to its parent[16:11:31] <sparky> moving up in the tree[16:11:41] <sparky> to get from the parent to the child you must use the inverse[16:11:56] <WhatAHam> http://img187.imageshack.us/my.php?image=snowframe1xi5.png[16:12:11] <WhatAHam> MatthiasM: that's the foreground texture[16:12:12] <sparky> for example in the case of the objects you have Mobj_to_world[i] which takes the child node to the parent node[16:12:39] <sparky> to get from the world to the child, ie., the object you'd use Invert(Mobj_to_world[i])[16:12:39] <batbrat> sparky: So typically they position the child relative to the parent. So I must invert to position the parent w.r.t child. Right?[16:12:51] <sparky> forget about how you invert for now[16:12:56] <WhatAHam> MatthiasM: http://img442.imageshack.us/my.php?image=levelsnowke7.png <-- that's the bg[16:13:05] <sparky> just imagine you have some magic routine that can take almost any 4x4 matrix and invert[16:13:07] *** elektr1k has joined ##opengl[16:13:11] <sparky> even if you filled it up with random floats[16:13:18] <MatthiasM> WhatAHam: I get only a network error instead of an image[16:13:23] <MatthiasM> try imagebin[16:13:29] <batbrat> I will not think about how to invert. I'll imagine the magic routine.[16:13:35] <sparky> good[16:13:46] <sparky> so this is the concept of the hierarchy[16:13:49] <MatthiasM> imageshack is often overloaded / buggy[16:13:51] <batbrat> sparky: 1 second please.[16:14:10] <WhatAHam> ok trying[16:14:58] <sparky> so in the same way, in the simple and most common case where the camera is also simply a child of the top/world node you design it's transformation relative to the top node just like you do the objects. That makes it simpler to visualize it what you are doing when you create it's matrix. With me so far?[16:15:27] *** korff_home has quit IRC[16:15:46] <batbrat> sparky: I position the cam just like an object. Take the camera to the world :)[16:15:55] <sparky> yes![16:15:57] <sparky> same for lights[16:16:06] <sparky> design your transformations to take them to their parents[16:16:12] <sparky> which in the typical case is the top node[16:16:15] *** korff_home has joined ##OpenGL[16:16:24] <sparky> that makes it conceptually easier to work with[16:16:40] <batbrat> And then use the invert to build the modelview matrix.[16:16:46] <sparky> yes![16:17:06] <batbrat> To take the world to the cam :). Thats the hierarchy. Yay![16:17:23] <WhatAHam> MatthiasM: the bg is here: http://imagebin.org/35247[16:17:27] <sparky> what you'd effectively be passing to modelview is: Mmv = Cworld_to_cam * Mobj_to_world[i] = Invert(Ccam_to_world) * Mobj_to_world[i][16:17:46] <MatthiasM> WhatAHam: is this a photo or rendered ?[16:17:56] <sparky> what people do to support more complex hierarchies assuming you want to know that to[16:17:59] <WhatAHam> that's rendered but i am making a 2d game so that's the texture in the game[16:18:03] <WhatAHam> http://imagebin.org/35248[16:18:03] <batbrat> sparky: Can you hold on a sec. I have to attend to a call. I have a small question. Please wait. Sorry to impose.[16:18:07] <sparky> is simply start at the camera node you wish to see from[16:18:14] <WhatAHam> that's the foreground that is supposed to be snow, but when blended with the bg i provide.. it looks grayish[16:18:15] <sparky> ok, don't be too long[16:18:28] <batbrat> sparky: Be right back. (2 mins)[16:18:51] <quicksilver> WhatAHam: you don't want to make the alpha 0.5 everywhere.[16:19:05] <WhatAHam> right well in parts it's .5[16:19:10] <WhatAHam> in some parts it's 1[16:19:12] <WhatAHam> etc[16:19:18] <WhatAHam> lots of gradients of alpha[16:19:23] <batbrat> 20:48 < sparky> what people do to support more complex hierarchies assuming you want to know that to[16:19:27] <batbrat> [B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[16:19:33] <tmccrary1> nice[16:19:33] <quicksilver> ok that's fine but the colour appears to be gray not white?[16:19:33] <sparky> wow[16:19:34] <batbrat> B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[B[[16:19:36] <sparky> ha ha[16:19:41] <batbrat> impose.[16:19:43] <batbrat> 20:48 < WhatAHam> http://imagebin.org/35248[16:19:44] <quicksilver> key stuck down?[16:19:46] <batbrat> 20:48 < sparky> is simply start at the camera node you wish to see from[16:19:56] <sparky> batbrat: I want a urin sample from you[16:19:58] <WhatAHam> quicksilver: the color values are white.. but yeah.. with a <1 alpha you get.. gray![16:20:11] <quicksilver> WhatAHam: the colour values look grey to me[16:20:11] <sparky> wtf is going on[16:20:20] <MatthiasM> batbrat: what are you pasting here ?!?[16:20:21] <WhatAHam> that's because it's white with .5 alpha in places :)[16:20:22] <quicksilver> or did you paste the post-composted version?[16:20:29] <WhatAHam> no no that's the original[16:20:37] <tmccrary1> I vote for spyware[16:20:37] <batbrat> Sorry everybody.Key stuck down. Very sorry :([16:20:38] <WhatAHam> so i guess white iwth .5 alpha, when blended with anything, will gray it[16:20:47] <sparky> lol[16:20:50] <WhatAHam> which stands to reason if you look at the math behing blending..[16:20:50] <quicksilver> you want some parts to be fully opaque[16:20:55] <quicksilver> so they show up as white[16:20:58] <WhatAHam> right[16:20:59] <WhatAHam> and they are[16:21:05] <WhatAHam> just the non-opaque parts look gray[16:21:06] <WhatAHam> as you can see[16:21:10] <quicksilver> and then around the edges it should be a bit transparent[16:21:14] <quicksilver> so you can see through[16:21:15] <batbrat> sparky: I'm still on call one minute more please.[16:21:20] <sparky> batbrat: ok[16:21:21] <quicksilver> and then the rest totally transparent[16:21:23] <MatthiasM> WhatAHam: can you also show a screenshot ? the image of the snow looks ok[16:21:26] <quicksilver> so you see the underneath[16:21:28] <WhatAHam> sure[16:21:29] <WhatAHam> hang on[16:22:08] <MatthiasM> wait[16:22:17] <MatthiasM> WhatAHam: I just composed them in GIMP[16:22:21] <WhatAHam> http://imagebin.org/35249[16:22:28] <WhatAHam> actually to me that's not that gray[16:22:30] <MatthiasM> you snow texture also has some grey parts in it[16:22:31] <batbrat> sparky: I'll post this and my explanation for the polarView on my blog. Would you be so kind as to check it after I put it up. I'd be extremely grateful.[16:22:36] <WhatAHam> but the designer i am working with is bitching it's too gray and is blaming my code[16:22:41] <sparky> sure[16:22:44] <sparky> but let me finish though[16:22:46] <sparky> almost done[16:22:56] <batbrat> quicksilver, all the others: I'd be glad if you took a look too.[16:23:08] <sparky> <sparky> what you'd effectively be passing to modelview is: Mmv = Cworld_to_cam * Mobj_to_world[i] = Invert(Ccam_to_world) * Mobj_to_world[i][16:23:11] <batbrat> sparky: I'll hear you out fully. Glad to learn[16:23:12] <sparky> <sparky> what people do to support more complex hierarchies assuming you want to know that to[16:23:20] <sparky> did you see those comments?[16:23:27] <batbrat> I saw those.[16:23:32] <MatthiasM> WhatAHam: in Gimp it looks much better[16:23:41] <sparky> ok, good. Now for a more complex tree[16:23:48] <batbrat> Ok :)[16:23:49] <WhatAHam> MatthiasM: yeah i know.. and i can't figure out why that would be..[16:23:55] <WhatAHam> this is an iphone app btw.. so it's opengles[16:24:00] *** vampirefrog has joined ##OpenGL[16:24:07] <sparky> if you want to view everything from some arbitrary camera that could be down somewhere deep in the tree structure[16:24:20] *** vampirefrog has quit IRC[16:24:30] <MatthiasM> WhatAHam: try GL11.glBlendFunc(GL11.GL_SRC_ALPHA, GL11.GL_ONE_MINUS_SRC_ALPHA);[16:24:47] <WhatAHam> MatthiasM: i am using that .. what's GL11?[16:24:52] <sparky> the idea is start at this very camera node and work your way out through the tree recursively. Using the matrix as is when moving up to parents[16:24:57] <MatthiasM> that's the GL version in LWJGL[16:24:59] *** vampirefrog has joined ##OpenGL[16:25:01] <sparky> and applying the inverse when moving down[16:25:02] <MatthiasM> just ignore iot[16:25:10] <WhatAHam> yeah i am using that blendFunc[16:25:24] <batbrat> sparky: Got it. Continue :)[16:25:26] <MatthiasM> but you use SRC_ALPHA_MINUS_ONE[16:25:26] <WhatAHam> weird that in gimp it looks better.. well one thing is the iphone is 16-bit color[16:25:28] <sparky> because you want the transformation sequence for any object in the tree to the camera[16:25:33] <MatthiasM> (according to what you wrote above)[16:25:42] <sparky> so you see how this recursion would solve this right?[16:25:48] <WhatAHam> MatthiasM: oh i typo'd.. it's GL_ONE_MINUS_SRC_ALPHA[16:25:55] <vampirefrog> iphone has opengl?[16:26:00] <WhatAHam> vampirefrog: yeah.. opengles[16:26:04] <sparky> it would give you in some cases a long transformation sequence M1 * M2 * .... * Mn[16:26:05] <WhatAHam> which is a subset of opengl[16:26:06] <vampirefrog> niccccccce[16:26:12] <WhatAHam> yeah it's a sweet fucking platform[16:26:18] <batbrat> sparky: Cool. Isee. It would do the matrix mult in the right order and with the right matrix. :)[16:26:25] <sparky> but the recursion will give you the right sequence for each object in the tree to get to your chosen camera[16:26:30] <vampirefrog> awesome... i was gonna get a new phone.. any other phones with opengl[es]?[16:26:30] <sparky> yes[16:26:40] <sparky> if you get this then you get the concept of the tree[16:26:55] <WhatAHam> vampirefrog: not sure offhand.. i would imagine some o fthe windows me phones might be.. maybe the G1 is too?[16:26:55] <sparky> the nodes move relative to their parents[16:27:01] <sparky> which is why the top node doesn't move[16:27:03] <quicksilver> vampirefrog: most of the newish colour graphics phones have opengl games on.[16:27:05] <sparky> and is referred to as world[16:27:16] <vampirefrog> and they have some sort of hardware accel, right?[16:27:21] <WhatAHam> yeah[16:27:21] <quicksilver> yes.[16:27:22] <WhatAHam> they do[16:27:24] <quicksilver> although that's relative.[16:27:26] <vampirefrog> awesome[16:27:27] <quicksilver> their CPUS are so slow[16:27:27] <tmccrary1> batbrat: transform hierarchies are used in a lot of places, skeletal animation uses the exact same concept[16:27:28] <quicksilver> they need it :)[16:27:35] <MatthiasM> WhatAHam: soory - no idea - might be bad rounding of the color values - try to make a snow a bit whiter[16:27:35] <WhatAHam> the problem iwth the g1 is.. it's a java platform so that can really mess with your senisibilities as a game programmer..[16:27:36] <vampirefrog> oh ok[16:27:47] <vampirefrog> hmm[16:27:52] <vampirefrog> you mean java slows things down?[16:27:55] <WhatAHam> MatthiasM: well the snow is pure white.. but yeah.. i think the alpha needs to be higher ..[16:28:02] <sparky> so you can also see why for the flat tree structure where they are all children of the top node[16:28:08] <tmccrary1> the G1 doesn't use Java, it uses "Dalvik"[16:28:11] <MatthiasM> WhatAHam: no - it also has some grey areas in it[16:28:12] <sparky> this because the result of the recursion[16:28:14] <sparky> Invert(Ccam_to_world) * Mobj_to_world[i][16:28:19] <batbrat> sparky,quicksilver : Won't recursing miss the other objects in the scene?[16:28:19] <sparky> which is passed to modelview[16:28:20] <WhatAHam> vampirefrog: yeah java is slow.. i don't care what they say about JIT.. things like garbage collection come at a huge cost.. you get random framerate drops for no good reason[16:28:20] <tmccrary1> which is like Java from 1998[16:28:26] <MatthiasM> these are 128-160[16:28:29] <WhatAHam> (when the GC kicks in)[16:28:37] <tmccrary1> WhatAHam: Only if you're a newb and create tons of garbage[16:28:39] <vampirefrog> nasty[16:28:42] <sparky> bat: you also recurse up[16:28:45] <sparky> not just down[16:28:46] <tmccrary1> If you don't create garbage, the garbage collectors don't run[16:28:49] <tmccrary1> It's like any other language[16:28:54] <sparky> you run through the entire tree[16:28:56] <MatthiasM> WhatAHam: well - I'm not sure which Java you used - but I had never GC issues in my games[16:29:09] <sparky> starting from the camera node[16:29:09] <tmccrary1> You have to be mindful of memory usage, although a lot of newbies think you can ignore it with Java[16:29:12] <tmccrary1> which is NOT the case[16:29:13] <vampirefrog> and is there a market for iPhone 3D graphics apps?[16:29:19] <sparky> batbrat: got it?[16:29:25] <batbrat> sparky: You run through the tree starting at whichever camera node you want. Right?[16:29:27] <WhatAHam> vampirefrog: errr.. where have you been? YES! soe people are making millions.[16:29:28] <tmccrary1> Graphics apps maybe, but the iphone is a poor general gaming platform[16:29:30] <sparky> yes[16:29:33] <batbrat> sparky: If so I got it :)[16:29:34] <vampirefrog> awsome[16:29:36] <sparky> so no objects will be missing[16:29:38] <quicksilver> vampirefrog: yes, there is a market but the market is for high volume cheap stuff[16:29:41] <sparky> because you are going through all of it[16:29:46] <sparky> including up[16:29:46] <batbrat> sparky: Cool.[16:29:48] <WhatAHam> tmccrary1: it's arguably a better gaming platform than other phones[16:29:51] <quicksilver> vampirefrog: most of the stuff is under $5 nearly all under $10.[16:29:55] <vampirefrog> can iPhone run quake 3 for example?[16:29:58] <WhatAHam> vampirefrog: yes[16:30:01] <sparky> apply when up, apply invert when down[16:30:03] <vampirefrog> at what fps?[16:30:08] <batbrat> sparky: Awesome.[16:30:08] <quicksilver> they have a version of cube which runs on it[16:30:09] <tmccrary1> WhatAHam: The lack of buttons = game over[16:30:16] <quicksilver> a decent FPS as long as not much is going on.[16:30:23] <vampirefrog> cube = nice. i've been playing sauerbraten for a while[16:30:24] <sparky> so this results in Invert(Ccam_to_world) * Mobj_to_world[i] for the flat tree[16:30:26] <sparky> with me?[16:30:37] <sparky> if you are, then the only thing remaining is the invert[16:30:37] <vampirefrog> hm[16:30:43] <quicksilver> tmccrary1: well it's a "different" gaming platform.[16:30:48] <WhatAHam> tmccrary1: yeah but the accelorometer is sort of an analog joystick[16:30:51] <quicksilver> tmccrary1: it's a poor platform for straight ports of PC games.[16:30:52] <WhatAHam> yeah it's different[16:30:59] <WhatAHam> eah if you port a pc game to it you're doing it wrong[16:31:01] <quicksilver> tmccrary1: galcon is awesome, though :)[16:31:02] <WhatAHam> you have to rethink the controls[16:31:15] <batbrat> sparky: With you. Tell me more :)[16:31:28] <tmccrary1> Well, for any game that would benefit from buttons, the iPhone is a poor platform for[16:31:37] *** sysRPL has quit IRC[16:31:38] <tmccrary1> Which is 99% of games[16:31:38] <vampirefrog> isn't there a mouse for iPhone? XD[16:31:51] <sparky> batbrat: so you are fully clear now right?[16:31:58] <sparky> on the modelview part[16:32:04] <vampirefrog> or a kboard for iPhone[16:32:06] <WhatAHam> tmccrary1: you can be creative with that.. you can do cool stuff with touch sometimes.. depends on the game man.. seriously.. that's a closed-minded point of view you are taking there, frankly.[16:32:22] <tmccrary1> Yeah, whatever, go make some "new creative" game without buttons[16:32:23] <tmccrary1> It's not fun[16:32:35] <vampirefrog> maybe u can make a porn game where u touch the lady[16:32:39] <sparky> batbrat: there are a lot of different ways you can choose to invert[16:32:39] <vampirefrog> and she likes it[16:32:40] <WhatAHam> tmccrary1: ok i will.. as long as you stay out of that market that's 1 less person to compete with.. so thanks for being closed minded![16:32:59] <batbrat> sparky: With you. Tell me.[16:33:02] <WhatAHam> vampirefrog: lol apple wouldn't let it i the store.. but you can always distribute outside the store..[16:33:08] *** m4ggus has joined ##opengl[16:33:09] <exDM69> http://qaq4.kyla.fi/~riku/goodenough/stencilheart.png <<--- more stencil fun[16:33:16] <sparky> but most of the simple ones require that you know a lot of stuff about the buildup of your matrices[16:33:21] <quicksilver> tmccrary1: 99% of existing games were designed with buttons in mind.[16:33:37] <sparky> for instance, do you know that the transpose of an orthogonal matrix is the inverse?[16:33:39] <quicksilver> tmccrary1: so, yes, absolutely, the iPhone is a poor platform for those.[16:33:41] <tmccrary1> Yeah, because buttons are critical[16:33:53] <tmccrary1> 99% of games expect a graphics display of some kind[16:33:54] <quicksilver> however, people are designing new games for the interface[16:33:58] <quicksilver> and sure, some of them are shit[16:34:01] <quicksilver> many of them, even.[16:34:03] <quicksilver> but some are great.[16:34:04] <WhatAHam> yeah i agree with quicksilver[16:34:17] <sparky> batbrat?[16:34:20] <tmccrary1> Without the tactile feedback, it's game over[16:34:31] <vampirefrog> u can imagine it as a mouse-only system[16:34:32] <WhatAHam> that's your opinion.[16:34:34] <batbrat> sparky: If you don't mind, can I come back and learn that tomorrow. Sorry. I've developed a debilitating headache, I'm feeling very sick suddenly. So sorry to bother you. Can you please do that?[16:34:39] *** LordMetroid has quit IRC[16:34:46] <batbrat> sparky: I am also nauseous. Sorry.[16:34:50] <WhatAHam> first of all the accel is a VERY good analog joystick[16:34:51] <vampirefrog> lol[16:34:57] <tmccrary1> vampirefrog: mice have tactile feedback as well[16:35:03] <vampirefrog> o shiz[16:35:07] *** k3t has quit IRC[16:35:07] <sparky> batbrat: it's ok, I can have that effect on people :)[16:35:15] *** k3t has joined ##OpenGL[16:35:24] <batbrat> sparky: Would you be so kind as to tell me the remaining tomorrow. Extremely sorry. It isn't you. I think its my lunch...[16:35:31] <WhatAHam> and you can design a game where the lack of tactile feedback is ok.[16:35:52] <vampirefrog> sure you can[16:35:56] <sparky> batbrat: np, tend to a speedy recovery[16:36:00] <sparky> sorry you're feeling ill[16:36:02] <WhatAHam> it's a different way of thinking about controls is all.. if you can see that, great, if you can't, i am sick of talking about this.[16:36:09] <vampirefrog> i don't think the problem is the phone input, but the designers/progs[16:36:09] <batbrat> Bye for now. My brother will close the connection in a few minutes. When he gets back. I'll appear online till then. Thanks for your help[16:36:11] <tmccrary1> I should point out, that an iPhone with a couple of buttons would be awesome[16:36:13] <vampirefrog> proggers*[16:36:26] <batbrat> quicksilver, sparky , Jupp3 thanks to you all.[16:36:28] <batbrat> Bye.[16:36:32] <WhatAHam> i agree with vampirefrog[16:36:55] <vampirefrog> ah well. in the meanwhile i enjoy building stuff in sauerbraten with my mouse and kb :D[16:37:08] <vampirefrog> wanna see a torus i made out of cubes? XD[16:37:24] <WhatAHam> tmccrary1: yeah and pigs with wings would be awesome. you know we have to work with what we got in reality now.. not what we wish we had in some alternate universe.. as is the iphone itself is a sweet platform with a huge userbase and people are in the mindset of PAYING FOR APPS .. it's a good opportunity ofr developers[16:37:28] <sparky> batbrat: just in case you're still here, hte there's a 4x4 matrix invert in here --> http://playstation2-linux.com/project/download_single_file.php/103_648_14/mat44.cpp[16:37:51] <sparky> that will work for arbitrary matrices[16:37:58] <sparky> as always when the determinent is non zero[16:38:46] <sparky> batbrat: it's the code in the operator ~[16:39:04] <vampirefrog> he's dead from migraine[16:39:30] <sparky> yeah, poor dude[16:44:01] *** Entelin has quit IRC[16:47:03] *** kenws has quit IRC[16:49:27] *** doub has joined ##opengl[16:50:28] *** Entelin has joined ##OpenGL[16:56:35] *** djork has joined ##opengl[17:01:52] *** maxton has quit IRC[17:03:25] *** bijoo_appdev_ has quit IRC[17:03:25] *** bijoo_osdev_ has quit IRC[17:04:46] *** vampirefrog has quit IRC[17:07:15] *** NightVisio has joined ##OpenGL[17:09:10] <pa> if i have a matrix which is combination of translations and rotations, can i get back the transformations?[17:09:38] <pa> and can i get back 2 matrices where i have all the translations in the first, and all the rotations around x , y, and z axis in the second?[17:10:03] <pa> or maybe 4 matrices, first for the translation, then for the rotations around the signle axes[17:10:39] *** pietia has joined ##OpenGL[17:10:51] <tmccrary1> I'm not 100% sure what you're asking, but you could add the matrices in that case[17:11:41] *** itewsh has quit IRC[17:13:18] *** jimi_hendrix has quit IRC[17:14:30] <batbrat> batbratĖs brother logging of ..[17:14:33] <batbrat> quit[17:14:36] <batbrat> uit[17:14:51] <tmccrary1> it[17:15:12] <batbrat> \quit[17:15:22] *** batbrat has quit IRC[17:17:04] <quicksilver> pa: the short answer is no.[17:17:26] <quicksilver> oh, you didn't want to disentangle the rotation part[17:17:31] <quicksilver> then yes, it's possible.[17:18:42] <pa> ah[17:18:50] <pa> any hint/link?[17:18:58] <MatthiasM> pa: the translation is the 4th column[17:19:04] <pa> exactly[17:19:15] <pa> the rest is rotation then? : )[17:19:49] <MatthiasM> not exactly - the 4th row must be "0, 0, 0, 1" for a model view matrix[17:20:00] <MatthiasM> the top left 3x3 part is the rotation/scale/shre etc[17:20:07] <MatthiasM> *sheer[17:20:20] <MatthiasM> + mirror[17:21:36] <MatthiasM> but why would you want to "split" a matrix in the first place ?[17:22:06] <MatthiasM> if you need the parts - then remember them[17:24:21] *** Yustme has joined ##OpenGL[17:24:24] *** Dagyde has joined ##OpenGL[17:24:34] <pa> it's because i have the data in this format and a program which wants it in separate format.. the alternative is to modify the program, but i dont have the source..[17:24:52] <MatthiasM> ok[17:25:40] <pa> i guess that if i could split a matrix into an "only translation" part and an "only rotation" then it would be possible to get rotation for each axis fromt he second[17:26:16] <MatthiasM> you can convert the 3x3 part of a 4x4 matrix into a quaternion - and then to euler angles[17:27:19] <pa> but first do i have to separate the translation?[17:28:53] <MatthiasM> the problem is if you have a translate + rotate + translate - then the first translation changes the origin of the rotation - I'm not sure if you can get that back[17:29:44] *** NightVisio has quit IRC[17:35:42] *** _Rangar_ has joined ##OpenGL[17:36:08] <exDM69> http://qaq4.kyla.fi/~riku/goodenough/stencilholes.png <<--- more convex polygons filled using the stencil buffer[17:36:19] <exDM69> now with holes[17:39:41] *** reprore_ has joined ##OpenGL[17:43:19] *** jfroy has quit IRC[17:46:08] *** pwned has quit IRC[17:46:16] *** pwned_ has joined ##opengl[17:48:00] *** Rangar has quit IRC[17:52:28] <emzic> hi guys, what do you think of my normal-mapping implementation: http://vimeo.com/2750907[17:52:29] <TheFlash> exDM69: Is it easier to do convex polys with stencil than it is by breaking it down to multiple concave polys?[17:53:06] *** x0rx0r has quit IRC[17:53:48] <exDM69> TheFlash: I'd say yes[17:54:04] <exDM69> but it's also a helluva lot of slower (or at least that's what ppl have said to me)[17:54:14] <exDM69> and it can't do antialiasing that nicely[17:54:52] *** Dagyde_ has joined ##OpenGL[17:55:44] *** Dagyde has quit IRC[18:00:45] *** KU0N has quit IRC[18:02:38] *** [AD]Turbo has quit IRC[18:05:27] *** AlastairLynn has joined ##opengl[18:05:55] <quicksilver> TheFlash: don't you have convex/concave backwards in that sentence? :)[18:06:07] <quicksilver> convex is the easy one.[18:06:21] <quicksilver> concave, self-intersecting, and with holes, are the hard ones.[18:07:06] <quicksilver> MatthiasM: if you know it is exactly of that form, then you can.[18:07:11] <exDM69> I said convex too, while I meant concave[18:07:17] <quicksilver> MatthiasM: find the fixed point - that is your center of rotation.[18:07:27] <quicksilver> MatthiasM: then you have the translation which you can subtract out.[18:07:33] *** jfroy|work has joined ##OpenGL[18:07:37] <MatthiasM> quicksilver: I think the problem is when you don't know how the matrix is composed[18:07:41] <quicksilver> right[18:07:47] <quicksilver> for a general matrix, you're pretty stuffed.[18:07:59] <quicksilver> you can diagonalise it though.[18:08:05] <quicksilver> (which requires complex numbers)[18:08:19] <MatthiasM> I'm using matrix -> quaternion -> matrix to get the rotation only part for the background/sky sphere[18:08:22] * quicksilver nods[18:08:55] *** bijoo_appdev has joined ##OpenGL[18:08:55] *** bijoo_osdev has joined ##opengl[18:08:57] *** ewn has left ##OpenGL[18:09:01] <TheFlash> quicksilver: Oh, yes.[18:11:07] *** bijoo_osdev_ has joined ##opengl[18:11:07] *** bijoo_appdev_ has joined ##OpenGL[18:13:41] *** groton has quit IRC[18:16:53] *** bijoo_osdev has quit IRC[18:17:11] *** bijoo_appdev has quit IRC[18:30:12] *** urbanmonk has joined ##OpenGL[18:31:03] *** ignacio has quit IRC[18:35:41] *** calav3ra has joined ##opengl[18:36:27] *** urbanmonk has quit IRC[18:50:45] *** HuntsMan has quit IRC[18:52:59] *** XT95__ has joined ##OpenGL[18:58:33] *** pwned_ is now known as pwned[19:00:56] *** b0000 has quit IRC[19:04:25] *** prak has joined ##OpenGL[19:04:41] *** groton has joined ##OpenGL[19:04:59] <prak> how do i get the latest opengl api tools on a computer and get it installed?[19:05:12] <prak> i went to the opengl sdk and couldn't find anything helpful[19:05:45] <MatthiasM> prak: what OS ?[19:05:52] <prak> MatthiasM, windows xp[19:06:07] <MatthiasM> prak: use GLEW and install latest drivers[19:06:32] <prak> MatthiasM, are the latest drivers on the GLEW website?[19:06:38] <MatthiasM> no[19:07:34] <prak> what kind of drivers are you talking about? video card drivers?[19:07:42] <MatthiasM> yes[19:08:01] *** XT95_ has quit IRC[19:08:28] *** b0000 has joined ##opengl[19:08:53] <prak> ok[19:08:57] <prak> thanks[19:09:03] <prak> got confused for a minute there[19:14:16] *** reprore_ has quit IRC[19:15:29] *** dusted has joined ##OpenGL[19:15:38] *** elite01 has quit IRC[19:15:44] *** elite01 has joined ##opengl[19:16:06] *** vade has quit IRC[19:17:03] *** reprore_ has joined ##OpenGL[19:23:37] <emzic> hello, does anyone have a tutorial on how to achieve a glow-effect using GLSL?[19:23:54] *** reprore_ has quit IRC[19:24:06] <Renderwahn> developers.nvidia.com[19:24:15] <Plagman> there are a ton[19:24:17] <Renderwahn> gpu gems 1 or 2, don't know which one[19:24:24] <emzic> ok thanks[19:24:30] <Plagman> not necessarily using GLSL, but the logic is pretty easy to port[19:25:13] <Plagman> basically put the stuff you want to glow in a separate layer, either by using MRT or a second pass, then do a gaussian blur pass and composite the two buffers together[19:25:15] <emzic> yeah, i am basically looking for the theory only[19:25:30] <emzic> thanks! what is MRT?[19:25:37] <Plagman> Multiple Render Targets[19:25:44] <emzic> oh, a framebufferobject?[19:26:02] <Plagman> yeah, with one output for the regular scene and another color output for the glowing stuff only[19:26:14] <emzic> ok thanks![19:27:10] <Plagman> people usually use a smaller-sized target for the glow so you may just want to use two passes[19:29:35] *** reprore_ has joined ##OpenGL[19:31:41] <emzic> ok, stupid question: how do i create something like a gaussian kernel for blurring in GLSL?[19:33:16] <Plagman> people usually just hardcode it in the shader with a constant array and leave the rest to the compiler[19:33:32] <Plagman> you can always pass it as a uniform through OpenGL, though[19:34:20] *** zstars has joined ##OpenGL[19:34:30] *** Xmas| has joined ##OpenGL[19:34:43] *** reprore_ has quit IRC[19:34:50] *** itewsh has joined ##OpenGL[19:35:15] *** Dagyde_ has quit IRC[19:36:01] *** reprore_ has joined ##OpenGL[19:36:26] <emzic> ok[19:37:01] <MatthiasM> you can use the bilinear filter to get nearly half of your taps for free[19:37:01] *** Dagyde_ has joined ##OpenGL[19:37:31] <emzic> what do you mean by taps?[19:37:41] <groton> the coefficents[19:37:55] <emzic> of the kernel?[19:37:57] *** prak has quit IRC[19:38:08] <emzic> the gaussian convolution kernel?[19:38:26] <MatthiasM> read about FIR filters[19:39:01] *** reprore_ has quit IRC[19:42:51] <emzic> ok thanks[19:47:24] *** Nescafe has quit IRC[19:47:59] *** ignacio has joined ##OpenGL[19:50:26] *** mlucassmith has joined ##OpenGL[19:51:10] *** doub has quit IRC[20:06:46] *** dolphin has joined ##OpenGL[20:15:10] <emzic> what could be the reason, that glActiveTexture(GL_TEXTURE1); causes an GL_INVALID_VALUE error?[20:15:38] *** twist has joined ##OpenGL[20:15:46] *** prak has joined ##OpenGL[20:16:21] <MatthiasM> did you check for multi texturing support ?[20:16:28] <prak> for installing glew on windows xp to work with ms visual studio 2005, i can't seem to find the platform sdk folder according to the instructions stated http://glew.sourceforge.net/install.html[20:16:45] <prak> where would those files be copied in this case?[20:17:45] <MatthiasM> just keep them somewhere[20:17:59] <MatthiasM> and tell your "wants to be an IDE" where it is[20:19:05] <emzic> MatthiasM: yeah it is a GF9. can it have something to do with framebufferobjects?[20:19:23] <MatthiasM> did you read the docu ?[20:20:30] <emzic> for glActiveTexture? yes, and the strange thing is, according to the docu it can not even cause a GL_INVALID_VALUE but only a GL_INVALID_ENUM[20:20:31] <prak> MatthiasM, thanks, actually found where i need to copy the files to now[20:23:43] <MatthiasM> anyone with a ATI Radeon Mobility 9600 here ? I get strange belnding error in my app - I render different subsections of one texture and some blend ok and other get a green border[20:25:58] <emzic> looks fine over here. is that your own opengl gui library?[20:26:06] <MatthiasM> yes[20:26:12] <emzic> cool[20:26:18] <emzic> do you know fenggui?[20:26:26] <MatthiasM> the default renderer uses only GL 1.1[20:26:29] <MatthiasM> yep[20:26:50] <MatthiasM> but it was not flexible enough[20:26:54] *** Kasu- has joined ##OpenGL[20:27:01] <emzic> yeah :( we used it at work a while back[20:27:15] <emzic> but it turned out to be too buggy[20:27:27] <emzic> what is the name of your library?[20:27:35] <MatthiasM> I called it GLUI[20:27:43] <MatthiasM> GL + UI :)[20:28:05] <emzic> are you going to make it available? i.e. what license?[20:28:14] <MatthiasM> BSD[20:28:29] <emzic> cool[20:29:16] <MatthiasM> and it generates zero garbage while rendering - only thinks like resizing creates garbage[20:29:30] <emzic> nice! is it skinnable?[20:29:44] <MatthiasM> fully - the code has no idea at all how to render :)[20:29:45] *** pietia has quit IRC[20:30:00] <MatthiasM> except some special widget's like the graph[20:30:27] *** ignacio has left ##OpenGL[20:30:41] <emzic> cool. how "feature-complete" would you think it is?[20:31:25] <MatthiasM> edit field is missing selections[20:32:05] <emzic> which opengl-binding does it use?[20:32:06] *** eXtronuS has joined ##OpenGL[20:32:14] <emzic> (sorry if i am asking so many questions ;) )[20:32:22] <MatthiasM> LWJGL[20:32:32] <MatthiasM> the renderer is fully abstracted[20:32:45] <MatthiasM> but the input handling uses the keycode constants from LWJGL[20:32:58] <emzic> ok[20:33:13] *** pietia has joined ##OpenGL[20:33:19] *** Xmas| has quit IRC[20:33:36] <MatthiasM> but as I see LWJGL as the only stable OGL binding :DD it's not an issue :)[20:35:17] <emzic> hmmm i am using jogl, maybe that is the reason for the GL_INVALID_OPERATION error?[20:35:41] <MatthiasM> ealier it was a GL_INVALID_VALUE :P[20:36:44] <emzic> yeah, that one[20:37:16] <MatthiasM> are you sure that the error is from this call and not from another call ?[20:38:21] *** XT95__ has quit IRC[20:38:59] *** Ademan has quit IRC[20:39:43] <MatthiasM> did you check how many texture unit are supported ?[20:40:27] <emzic> good question, it is the jogl-error-reporting that tells me this call is causing the error[20:40:38] <emzic> and i think 2 units should be supported on a GF9[20:40:44] *** Ademan has joined ##OpenGL[20:40:59] <MatthiasM> from the HW yes - but do you have the correct drivers ?[20:41:05] *** Nescafe has joined ##OpenGL[20:41:53] <emzic> pretty sure yes, any multitexturing has been working just before. only now that i use FBOs to achieve that glow effect, the error has occured[20:41:59] <emzic> *and[20:45:38] *** pfo has joined ##OpenGL[20:57:40] *** mlucassmith has quit IRC[20:58:39] *** vade has joined ##OpenGL[20:59:52] *** predaeus has quit IRC[21:01:18] *** madeyes has joined ##OpenGL[21:02:04] *** madeyes has left ##OpenGL[21:05:06] *** vade has quit IRC[21:09:20] *** reprore_ has joined ##OpenGL[21:10:39] <Andon> emzic, there are any number of reasons for that. Some pixel formats used for FBOs don't support a depth buffer or stencil buffer for instance, and operations that manipulate them while bound will result in an GL_INVALID_OPERATION.[21:11:12] <Andon> Likewise, it's possible that a certain pixel format doesn't offer the same number of texture units. Which might be why GL_TEXUTRE_1 is invalid.[21:13:52] <emzic> thanks for the information, Andon[21:14:39] <Andon> I'd suggest querying some of the MAX attributes with your FBO bound.[21:14:54] <emzic> ok[21:17:10] <MatthiasM> it highly unlikly that the number of texture units / samplers change based on FBO[21:17:19] <emzic> well it still says 32 if i query GL_MAX_COMBINED_TEXTURE_IMAGE_UNITS[21:21:12] *** bijoo_osdev_ has quit IRC[21:21:12] *** bijoo_appdev_ has quit IRC[21:22:35] *** bijoo_osdev has joined ##opengl[21:22:38] *** bijoo_appdev has joined ##OpenGL[21:33:48] *** MouserSVK has joined ##OpenGL[21:34:35] *** kbotnen has quit IRC[21:37:21] *** mm765 is now known as mm765^away[21:38:35] *** qeed has joined ##opengl[21:38:47] *** _JSC415 has joined ##OpenGL[21:39:13] <_JSC415> anyone knows a livecd which has opengl dev tolls ready ?[21:39:30] <_JSC415> i tryed rails live.[21:39:33] <_JSC415> its awesome.[21:39:38] <_JSC415> and knoppix[21:39:49] <_JSC415> rails looks like has sdl ready.[21:40:03] <_JSC415> but i cant find sdl.h[21:40:15] <_JSC415> can anyone find it for me ?[21:41:01] <Ingenu> don't know of any[21:41:09] *** _JSC415 is now known as _THEGOD[21:41:11] <Ingenu> but never looked for one either[21:43:21] *** bijoo_appdev_ has joined ##OpenGL[21:43:37] *** bijoo_osdev__ has joined ##opengl[21:44:02] *** bijoo_appdev has quit IRC[21:46:55] <_THEGOD> hm...[21:47:10] <_THEGOD> also my repository points to 2004 edition of pcoslinux[21:47:19] <_THEGOD> while im using rails livecd[21:47:46] <_THEGOD> i wonder if there is a 2007 edition of railslivecd[21:48:05] <_THEGOD> or if im using 2008 or 2004 editions of those cd[21:48:40] <_THEGOD> railslivecd is best livecd i ewer tryed after knoppix[21:48:45] <emzic> if i am about to forward a texture to the shader using glUniform1i, what do i put in the second argument? the id of the texture or the number of the unit?[21:49:04] <Ingenu> unit #[21:49:09] <Ingenu> AFAIR[21:49:09] <emzic> thanks![21:49:12] <_THEGOD> if i can find a way to install vine on rails lived ill prefer it instead knoppix[21:49:14] *** prak has quit IRC[21:49:48] <Ingenu> I think they talk about that in the specs btw, they seem to indicate they intent to use the texture ID later on[21:50:12] <_THEGOD> glUniform1i ?[21:50:15] <Ingenu> maybe when we get OpenGL 5.0 (which wouldn't look like a crappy dinosaur API)[21:50:17] *** bijoo_osdev__ has quit IRC[21:50:17] *** bijoo_appdev_ has quit IRC[21:50:17] <_THEGOD> what the hell is that ?[21:50:20] *** Gorgoroth has joined ##OpenGL[21:50:25] <_THEGOD> i heard it first time.. ;P[21:50:27] <Ingenu> programmable hardware[21:51:11] <_THEGOD> blah. opengl is just a little closed source api.[21:51:26] <_THEGOD> opengl ... blah blah. whatwer...[21:51:32] <_THEGOD> its just bunch of code[21:51:42] <_THEGOD> on hardware or software.[21:51:48] *** bijoo_appdev has joined ##OpenGL[21:51:52] *** bijoo_osdev_ has joined ##opengl[21:51:56] <_THEGOD> its allways a little bucket of code.[21:52:11] <_THEGOD> and nothing more.[21:52:44] <MatthiasM> you should choose a more fitting nick @ _THEGOD[21:53:02] <_THEGOD> blah...[21:53:06] <_THEGOD> i like this nick.[21:53:25] <Ingenu> I think his first idea was "thepenis", but he changed his mind[21:53:33] <Ingenu> :p[21:53:43] <_THEGOD> ewen after its got identifiead as something bad.[21:54:02] <_THEGOD> but still coolll .[21:54:24] <_THEGOD> ewen petter the the pennis or whatewer fck.[21:54:48] <_THEGOD> damn.[21:54:53] *** bijoo_appdev_ has joined ##OpenGL[21:54:54] *** bijoo_osdev__ has joined ##opengl[21:55:13] <_THEGOD> i learned about how to use devc+++[21:55:31] <_THEGOD> it has a auto ident code option.[21:55:35] <emzic> ok when i use glUniform1i and i put the unit number in the second parameter do i put "GL_TEXTURE0" or "0" ?[21:55:50] <emzic> probably just 0...[21:57:08] <_THEGOD> just by hittin shit+space[21:57:21] <_THEGOD> it identificates all valid variables.[21:57:23] *** bijoo_osdev has quit IRC[21:57:40] <_THEGOD> and autooffers most significant code.[21:57:43] <MatthiasM> _THEGOD: and ?[21:57:50] <_THEGOD> so helps saving time.[21:58:07] <MatthiasM> and why do you spam the channel with such things ?[21:58:10] <_THEGOD> while you dont need to waste your expensive time by creating a project.[21:58:20] <_THEGOD> spam ?[21:58:56] <_THEGOD> i just tryed to inform peoples about some kinds i expierenced ower time.[21:59:16] <Ingenu> emzic, I think it's zero[21:59:22] <_THEGOD> and try to help them not doing bad and wrong things i did outofcontroll.[21:59:29] <emzic> yeah, that was my problem all along... -.-[21:59:49] <emzic> i was putting GL_TEXTUREn , which causing the gl-errors[21:59:58] <Ingenu> ^^[22:00:31] <_THEGOD> i found all info i need to start my cool 3d games.[22:00:31] <Ingenu> bniding the texture id would be so much easier[22:00:35] <_THEGOD> but damn time.[22:00:42] <_THEGOD> its newer enough.[22:02:26] <_THEGOD> hm..[22:02:47] <_THEGOD> i hawe to do full check if rails has sdl.h somewhere else ..[22:02:59] <Ingenu> let me guess, you'll make Doom IV or WoWII ?[22:03:30] <emzic> doom4 is already in the works[22:03:32] <emzic> by id[22:03:50] <Ingenu> I thought they askeda studio to do it, as they are busy with Rage[22:04:03] <MatthiasM> well - if you really want a live cd - then install ubuntu - install all required packages - and create a new live cd[22:04:20] <emzic> well i dont know that, but i know it will use the rage-engine[22:04:30] <emzic> maybe the outsource it like quake4[22:05:05] <Ingenu> I'd think so[22:05:09] <Ingenu> but I don't remember[22:06:05] <_THEGOD> hm..[22:06:08] <_THEGOD> ubuntu ?[22:06:10] <_THEGOD> ok.[22:06:12] *** reprore_ has quit IRC[22:06:15] <_THEGOD> ill check it thanks.[22:06:59] <_THEGOD> i tryed also kde livecd develop studio or whatewer.[22:07:02] <_THEGOD> but it sucks.[22:07:11] <_THEGOD> it does not hawe ewen gcc too.[22:07:15] <_THEGOD> thats stupid.[22:07:20] *** mlucassmith has joined ##OpenGL[22:07:25] <_THEGOD> they calls it developer cd[22:07:39] <_THEGOD> but they did not provide a gcc[22:07:57] <_THEGOD> and its wery wery slow.[22:08:01] <MatthiasM> have - not hawe[22:08:12] <MatthiasM> same for even[22:08:14] <groton> wery[22:08:57] <_THEGOD> ok thanks.. i sometimes forget that .[22:09:00] <MatthiasM> and don't complain in ##opengl about linux distributions[22:09:24] <_THEGOD> you means linux distros does make make sense for opengl ?[22:09:48] <_THEGOD> but some distros does not supports opengl correctly ? do dhey ?[22:10:25] <_THEGOD> ok ok.. i hawe to find a live kde ubuntu.[22:10:51] <_THEGOD> and check if i can waste enough time trying installing gcc and opengl libs on it ;P[22:11:08] <_THEGOD> do you hawe a short ling about that ?[22:11:18] <_THEGOD> *ling=link[22:11:19] *** TiLex has joined ##opengl[22:11:59] <_THEGOD> MatthiasM do you hawe a link for that ?[22:12:09] <MatthiasM> this channel is about developing opengl based apps - not to discus your likes and don't likes of linux distribution[22:12:30] *** TiLex has quit IRC[22:12:46] <_THEGOD> but ewen if i like i cand do opengl based apps on any linux :P[22:13:24] <MatthiasM> meh[22:14:35] <MatthiasM> other people are able to read the manual and usage google - so they have no issue to develop OpenGL apps under linux[22:14:39] *** dusted has quit IRC[22:14:42] <MatthiasM> *use[22:14:52] <Xantoz> _THEGOD: Also, spelling... ewen => even[22:14:57] <_THEGOD> i dont think so ;P[22:15:23] *** asdfa has joined ##OpenGL[22:15:41] <MatthiasM> I'll put you on ignore - because a) you are annoying b) don't even try to laern c ) you spell stuff wrong[22:15:43] <_THEGOD> there is not much succeful programmers on our earth i know .[22:15:50] *** bijoo_appdev has quit IRC[22:16:26] *** bijoo_osdev_ has quit IRC[22:17:13] <emzic> besides thegod of course ;)[22:17:32] *** Jupp3 has quit IRC[22:17:44] *** _JSC415 has joined ##OpenGL[22:17:47] <MatthiasM> lol[22:18:35] <_JSC415> well. last one .[22:18:52] *** Waldir has joined ##OpenGL[22:18:54] <_JSC415> do i hawe to download default ubuntu or kde ubuntu ?[22:19:00] <_JSC415> is there any difference ?[22:19:05] <MatthiasM> yep[22:19:15] <MatthiasM> reading will tell you[22:19:23] *** Waldir has left ##OpenGL[22:19:24] *** bijoo_osdev__ has quit IRC[22:19:25] *** bijoo_appdev_ has quit IRC[22:20:01] *** Jupp3 has joined ##OpenGL[22:22:05] *** NinZine has joined ##OpenGL[22:24:05] *** _JSC415 is now known as __THEGOD[22:25:37] <__THEGOD> hm..[22:26:00] <__THEGOD> looks like google outocomplete options is not working for me ?[22:26:14] <__THEGOD> does it works for you kiddies ?[22:27:16] <__THEGOD> ok forget it it started working now. ;P[22:27:26] <__THEGOD> im downloading kubuntu livecd[22:27:31] <__THEGOD> if i can find it.[22:27:42] <__THEGOD> and check if it has gl.h and sdl.h[22:27:52] <__THEGOD> or if repositories are stable ;P[22:27:57] *** tokemonstah has joined ##OpenGL[22:31:46] <__THEGOD> does the kubuntu-5.10-live-i386.iso is the lastest edition ?[22:32:55] <__THEGOD> where can i find 8.10 livecd ?[22:33:10] <__THEGOD> http://www.kubuntu.org/getkubuntu offers only one option i see[22:33:31] <__THEGOD> and they does not informs about if its a livecd or not 'P ?[22:34:16] *** _THEGOD has quit IRC[22:37:31] *** Walt has quit IRC[22:37:59] *** _JSC415 has joined ##OpenGL[22:38:29] *** _JSC415 is now known as _THEGOD[22:41:14] *** LtJax has quit IRC[22:43:46] *** _JSC415 has joined ##OpenGL[22:44:10] *** cplusplus2 has joined ##OpenGL[22:45:24] *** Nescafe has quit IRC[22:46:22] *** MouserSVK has quit IRC[22:46:37] *** walt has joined ##OpenGL[22:47:37] *** walt has quit IRC[22:48:37] *** [1]_JSC415 has joined ##OpenGL[22:49:52] *** Spkka has joined ##OpenGL[22:50:43] *** x0rx0r has joined ##OpenGL[22:50:43] *** walt has joined ##OpenGL[22:51:06] *** cplusplus has quit IRC[22:53:38] *** dolphin has quit IRC[22:54:05] *** walt has quit IRC[22:54:57] *** __THEGOD has quit IRC[22:55:06] *** elite01 has quit IRC[22:56:32] *** Ingenu has quit IRC[23:00:34] *** elite01 has joined ##opengl[23:02:32] *** _JSC415 has quit IRC[23:02:33] *** [1]_JSC415 is now known as _JSC415[23:05:07] *** Jernej has joined ##OpenGL[23:06:02] *** _THEGOD has quit IRC[23:08:54] *** jfroy|work has quit IRC[23:10:06] *** Ademan has quit IRC[23:10:29] *** _JSC415 is now known as _THEGOD[23:20:23] *** _THEGOD has left ##OpenGL[23:23:20] *** JernejL_ has quit IRC[23:25:29] *** itewsh has quit IRC[23:28:03] *** calav3ra_ has joined ##opengl[23:29:53] *** Quentarez has quit IRC[23:29:58] *** groton has quit IRC[23:34:01] *** Xmas| has joined ##OpenGL[23:36:22] *** dv_ has joined ##opengl[23:36:35] *** Renderwahn has quit IRC[23:37:41] *** elite01 has quit IRC[23:39:55] *** iion has joined ##OpenGL[23:44:15] *** calav3ra has quit IRC[23:45:24] *** Yustme has quit IRC[23:51:22] *** bijoo_appdev has joined ##OpenGL[23:51:23] *** bijoo_osdev has joined ##opengl[23:55:51] *** Gorgoroth has quit IRC