November 24, 2014 
< | 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 | >
November 24, 2014 [00:00:01] <void256> :>[00:00:03] *** BlackPhoenix is now known as BlackFox[00:02:30] *** iKlsR has quit IRC[00:05:38] *** jdolan has quit IRC[00:07:13] *** rdgawdzi has joined ##OpenGL[00:07:35] <aethersis> when coils are wheezing in my GPU, is that bad?[00:07:47] <rdgawdzi> Is it possible to install packages such as libglu1-mesa-dev locally?[00:07:52] <rdgawdzi> *ie no sudo[00:08:03] *** Biliogadafr has joined ##OpenGL[00:08:07] <aethersis> why'd you do that, rdgawdzi[00:08:09] *** Biliogadafr has quit IRC[00:08:09] <aethersis> and sure it is possible[00:08:26] *** abs25 has quit IRC[00:08:28] <aethersis> you just find a repo and download it manually through web browser[00:08:46] <rdgawdzi> aethersis It's because I am doing a school project[00:08:58] <aethersis> just don't even dare to put them into the /lib location[00:09:19] <aethersis> if you replace existing libraries from repos with libraries from unknown source, it will cause serious problems[00:09:34] *** goliath has quit IRC[00:09:39] <rdgawdzi> so I should basically clone xorg-dev, libglu1-mesa-dev into my project? And I would be good?[00:10:04] <aethersis> yes, put these files in separate folder and you will be fine[00:10:14] <aethersis> but why are you using versions of the libraries that are not in your repo?[00:10:16] <aethersis> what's the reason?[00:10:59] <rdgawdzi> Well they don't exist in my enviroment[00:11:13] <rdgawdzi> I want to use glfw[00:11:21] <aethersis> you can add repos for glfw[00:11:25] <aethersis> do you use ubuntu?[00:11:27] <rdgawdzi> and it has a dependency on those two debian packages[00:11:31] <aethersis> ohhh[00:11:34] <rdgawdzi> Yes the enviroment I am in is ubuntu[00:11:40] <aethersis> wait[00:11:44] <aethersis> then there will be no problem with it[00:11:51] *** jdolan has joined ##OpenGL[00:11:57] *** linuxuz3r_ has quit IRC[00:12:07] *** groton has quit IRC[00:12:12] <aethersis> I have glfw 2.7.2 in my repos[00:12:18] <aethersis> and these are built in[00:12:27] <rdgawdzi> wait seriously?[00:12:29] <aethersis> I can't see what is your problem[00:12:29] <aethersis> yes[00:12:35] <rdgawdzi> Do you have a link?[00:12:41] <aethersis> sudo apt-get install glfw2 libglfw-dev[00:13:04] <aethersis> but I use xubuntu 14.04[00:13:29] <rdgawdzi> right but maybe glfw has a dependency on those two libraries I mentioned?[00:13:35] <rdgawdzi> so it downloads them?[00:13:39] <rdgawdzi> Remember I can't use sudo[00:13:48] <aethersis> even if it does, then if it is in the repos, it will download these libs automatically[00:13:50] <aethersis> why can't you use sudo[00:14:13] <rdgawdzi> I don't have an admin account, I need to build my project on a school enviroment[00:14:26] <aethersis> suuuuucks[00:14:47] <aethersis> then the only way to deal with the problem is to find out these dependencies, download all the libs needed and put them to your project's folder[00:14:54] *** Zerflag has quit IRC[00:14:55] <rdgawdzi> yup[00:15:04] <aethersis> this solution kinda sucks anyway, because linux is not designed for using anything in such the way[00:15:14] <rdgawdzi> I want to figure out how to use apt-get locally[00:15:25] <aethersis> I think you can't[00:15:40] *** Zerflag has joined ##OpenGL[00:15:42] *** Codex2 has quit IRC[00:16:01] <aethersis> oh wait[00:16:02] <aethersis> you can XD[00:16:17] <aethersis> apt-get download libglfw2 ^^[00:16:24] <aethersis> :3[00:16:31] <aethersis> I've never done it before[00:16:34] <aethersis> just read man pages lol[00:18:04] *** roboman2444 has quit IRC[00:19:13] *** foreignFunction has quit IRC[00:20:27] *** Zerflag has quit IRC[00:20:31] *** roboman2444 has joined ##OpenGL[00:24:09] *** linuxuz3r has joined ##OpenGL[00:25:15] *** Zupo has joined ##OpenGL[00:25:50] *** paul424 has quit IRC[00:25:57] *** Zupoman has quit IRC[00:27:47] *** pizthewiz has joined ##OpenGL[00:29:27] *** hahuang61 has quit IRC[00:31:18] <aethersis> I have an interesting question: when something takes 1mpix 2d texture certain ammount of time to compute on fragment shaders, will it take the same time to 3d texture of the same size ? or less, or more?[00:32:07] *** TheLorax has joined ##OpenGL[00:32:27] <HuntsMan> aethersis: 1 MP doesn't make much sense on 3D textures[00:32:37] <HuntsMan> what are the width, height and depth of such texture?[00:32:43] <aethersis> 100x100x100[00:33:03] <aethersis> will it take the same time to compute as 1000x1000 2d?[00:33:20] <derhass_> depends on the computation[00:33:42] <aethersis> almost all these computations are based on convolution, like box blur[00:33:48] <derhass_> first of all, fragment shader's aren't organized in 3d[00:34:03] <aethersis> but they can take 3d texture as sampler, right?[00:34:13] <aethersis> and output the same to fbo with 3d texture attached[00:34:33] <derhass_> well[00:34:44] <derhass_> not really[00:34:47] <aethersis> I'm just trying to estimate something[00:34:51] <aethersis> what do you mean?[00:34:56] <aethersis> Bloodlust said it's possible[00:35:22] <derhass_> possible is always relative[00:35:41] <aethersis> can they do that or not (the stuff with fbo and texture) ?[00:36:01] <derhass_> not in a terrible useful way[00:37:12] *** zajfy has quit IRC[00:37:19] <derhass_> there is actually no good reason to _try_ to do that[00:37:41] <aethersis> why?[00:37:56] <aethersis> what am I supposed to do instead? take 100x 100x100 2d texture?![00:37:58] <derhass_> because the fragment shader invocations are organzied in 2d[00:38:07] <aethersis> shit[00:38:15] <aethersis> so I will have to put 3d texture on 2d sampler?[00:38:33] <derhass_> the sampling is not the issue (at least not the main issue)[00:38:43] <derhass_> the shader invocations and the output are the issue[00:38:44] <aethersis> actually I will need 3 2d textures at once for neigbourhood in 3d[00:38:58] <aethersis> what do you mean? it won't go in order or what?[00:39:22] *** glararan has quit IRC[00:39:23] <derhass_> you'll need to render 100 quads instead of one[00:39:27] <derhass_> one for each layer[00:39:57] <aethersis> jesus...[00:40:15] <aethersis> I don't want to render any quads though :/[00:40:32] <aethersis> it's just for data, that will be processed as particles, nothing more[00:40:34] <derhass_> because you haven't understood the gl pipeline[00:40:43] <aethersis> you are right about it :([00:40:52] <derhass_> talking about fragment shader without wanting to render anything is going nowhere[00:40:58] <aethersis> I know[00:40:58] <derhass_> you might just want a compute shader[00:41:06] <aethersis> I thought maybe I can do that on a cube or something?[00:41:40] <derhass_> the key thing about the whole graphics pipeline is that is produces _2D_ images[00:41:47] <aethersis> derhass_ can't it be hacked somehow to make it as easy as possible?[00:42:04] <aethersis> I don't want to generate 100 planes in my vbo, or even more[00:42:05] <derhass_> aethersis: well, rendering 100 quads isn't particulary bad[00:42:15] <derhass_> you don't need 100 "planes"[00:42:22] <derhass_> you can do that all on the fly[00:42:37] <aethersis> it'd be actually easier to render them at once[00:42:53] <aethersis> and properly fit each "layer" of the texture to each quad[00:42:58] <aethersis> it[00:43:03] <aethersis> it's screwed up :/[00:43:19] <aethersis> also it will look like shit[00:43:40] <derhass_> why would it "look" like anything?[00:43:54] <aethersis> yeah you're right[00:44:08] <aethersis> you just broke my dreams :([00:44:11] <derhass_> you just need to generate a fragment for every texel you want to write to[00:44:35] <aethersis> that's what I was afraid of. How otherwise would it push uv coordinates to the fragment shader[00:45:20] <aethersis> I don't want to use compute shaders, because I would want to try it with WebGL[00:45:29] <derhass_> wegl?[00:45:30] <derhass_> lol[00:45:35] <derhass_> than forget about all of that[00:45:41] <aethersis> why[00:45:59] *** shintah has quit IRC[00:46:04] <derhass_> webgl is so completele ridioulus[00:46:07] <derhass_> *compeltely[00:46:10] <aethersis> why's that[00:46:15] <derhass_> it is so limited[00:46:29] <derhass_> basically a casrated version of ES 2[00:47:52] <glYoda> umm you can no op the fragment pipeline… consider raster discard[00:48:00] *** Jeanne-Kamikaze has quit IRC[00:48:39] <glYoda> assuming one just wants to execute the primitive assembly step[00:48:47] *** Crehl has quit IRC[00:48:50] *** mayankmadan_ has joined ##OpenGL[00:48:58] <aethersis> ll this opengl thing is a beast :/[00:49:10] <glYoda> otherwise CS would be the most sensible option[00:49:38] <aethersis> glYoda, what are you talking about xD[00:49:42] <derhass_> aethersis: webgl does not even support 3d textures[00:50:19] <derhass_> aethersis: you have to do all this in 2d. which of course is possible[00:50:24] <derhass_> but not really optimal[00:50:28] <glYoda> aethersis referring to an earlier comment from derhass: "the key thing about the whole graphics pipeline is that is produces _2D_ images"[00:50:51] <aethersis> god damn it :/[00:50:58] <aethersis> how did you get so pro, derhass_[00:51:00] <aethersis> what's your job?[00:51:04] <glYoda> while that is the more common case it isn't a requirement[00:51:24] <derhass_> glYoda: it is still tuned to it. we have rasterization and 2d framebuffers[00:51:35] <derhass_> aethersis: glYoda is more pro than me[00:51:53] <glYoda> derhass_ I wouldn't claim it's more tuned on that front[00:52:23] *** mayankmadan has quit IRC[00:52:30] <glYoda> after all with a sensible raster discard implementation you just skip the back end of the pipeline (i.e. rasterization)…[00:52:43] <aethersis> derhass_, what's your job?[00:52:53] <derhass_> aethersis: why would that matter?[00:53:26] <glYoda> now if the underlying implementation decides push content from the front end AND continue to execute the back end of the pipeline with a transparent discard (ugh) you'd have a point :P[00:53:33] <aethersis> I just wonder if it has anything to do with Opengl, derhass_[00:53:43] <glYoda> but no one would be stupid enough to do that in shipping HW :P[00:53:53] <glYoda> *decides to[00:54:08] <derhass_> glYoda: well. if one has to bypass half of the pipeline, I still might have a point[00:54:56] <glYoda> derhass_ if one was attempting to reinvent rasterization through this process you'd have a point :P[00:55:10] <glYoda> I don't know if that is what the OP was interested in though[00:55:17] *** DolpheenDream has quit IRC[00:55:28] <derhass_> nope. he just wants to do a 3d convolution[00:55:42] <glYoda> in any case I'd consider raster discard a very useful optimization WHEN used properly[00:55:53] <glYoda> yeah in that case this wouldn't be the way to go[00:55:56] <aethersis> what is OP[00:56:02] <derhass_> aethersis: YOU[00:56:03] <glYoda> use a compute shader FFS[00:56:28] <glYoda> otherwise you'd have to abuse the fragment pipeline… and I wouldn't recommend that[00:56:31] <aethersis> what does it mean, the OP[00:57:13] <glYoda> derhass_ just told you[00:57:22] <glYoda> original poster i.e. you in this context[00:57:23] <aethersis> why did you call me OP[00:57:25] <aethersis> ohhh[00:57:26] <derhass_> aethersis: "original poster"[00:57:36] <aethersis> I thought you were laughing at me or somehting xD[00:59:15] <derhass_> why would we do that?[00:59:22] <void256> this is a SERIOUS channel[00:59:27] <void256> people never laugh in here[00:59:44] <derhass_> the only one we laugh about here is void256[00:59:45] *** Burga has quit IRC[01:00:02] <glYoda> heh[01:00:44] *** mizux has quit IRC[01:00:46] <aethersis> lol[01:00:59] *** DolpheenDream has joined ##OpenGL[01:03:26] *** pizthewiz has quit IRC[01:03:43] <derhass_> aethersis: my job sometimes has to do with gl, and sometimes it has not. it is just an api, not really important, after all[01:04:11] <glYoda> indeed[01:04:34] <aethersis> what do you do then, derhass_[01:06:41] <derhass_> aethersis: a bit software development, a bit research, a bit teaching[01:06:54] <derhass_> and lots of ugly hacks[01:07:38] <aethersis> awesome![01:07:43] <aethersis> I want to have similar job, derhass_ :3[01:08:50] <derhass_> well. one can describe it one way such that it appears awesome. or one could describe it as horrible. the thruth is in the middle[01:09:13] <derhass_> although I never said that I've only one job^^[01:09:15] <aethersis> you must be some ultra pro guy then :D[01:09:26] <void256> >_<[01:09:46] <derhass_> aethersis: I already told you, glYoda is more pro than me, at least regarding opengl[01:09:50] <void256> everyone in here is ultra pro[01:09:59] <void256> why would they be here if not?[01:10:03] <void256> doesn't make sense[01:10:07] <derhass_> void256: why are you here?[01:10:09] <aethersis> it does - like me, for learning :D[01:10:22] <glYoda> derhass_ lol[01:10:25] *** garFF has quit IRC[01:10:28] <aethersis> if everyone was ultra pro here, then it would not make sense[01:10:36] <void256> because I want to be ultra pro but currently I'm stuck at pro only :/[01:10:52] <derhass_> void256: ah, ok. that makes sense[01:10:57] <aethersis> XD[01:11:01] <glYoda> void256 you need to level up then[01:11:03] <glYoda> :P[01:11:08] <aethersis> you have to farm bits[01:11:39] * danhedron rolls for "read specification"[01:12:06] <derhass_> hehe, writing 350 simple glut-based "hello triangle" programs just to gain 1 EP each[01:12:11] *** doomlord_1 has joined ##OpenGL[01:12:25] <aethersis> in 350 different ways lol[01:12:33] <aethersis> until you get to the level of machine code xD[01:12:38] <aethersis> and leave glut alone[01:13:39] *** Guest30532 has joined ##OpenGL[01:14:15] <void256> https://pbs.twimg.com/media/B2W5HDBIgAAt5aB.png:large <-- integrated potter&duff into Nifty 2.0 - does that give any ultra pro experience points? :D[01:15:05] <aethersis> what is this[01:15:19] <aethersis> it's basically some simple CSG operations[01:15:29] <void256> if you don't understand it you're not even mini pro :>[01:15:35] *** Zerflag has joined ##OpenGL[01:15:43] <aethersis> there was a youtube channel with beeping stuff and these planes[01:16:05] *** hexagoxel has quit IRC[01:18:07] <void256> and I can render bezier paths now ;)[01:18:48] <void256> https://www.youtube.com/watch?v=Or0xEBEy3jM[01:18:57] <void256> it will BLOW YOUR MIND[01:19:00] <void256> ^^[01:19:04] <aethersis> y..uh...[01:19:06] <aethersis> no xD[01:19:18] <aethersis> if you call yourself pro with that then I must be ultra pro with my fluid solver but I'm not[01:19:24] <aethersis> I'm noob and you must be ultra noob xD[01:19:48] <void256> I'm mega über ultra pro noob :>[01:19:50] <aethersis> and you don't even measure fps properly xD[01:19:57] <aethersis> you measure it on client side[01:20:20] <aethersis> that led strip is a cool thing[01:20:21] <void256> my FPS are ultra pro[01:20:24] <aethersis> you could make a costume :D[01:20:47] <void256> 20245 fps is pretty standard[01:20:50] <void256> :>[01:20:56] <aethersis> lol yeah it's CPU time[01:21:03] <void256> nope it's render time![01:21:08] <derhass_> aethersis: how do you know that?[01:21:10] <aethersis> it could be for such the simple thing[01:21:13] <aethersis> actually[01:21:20] <aethersis> but I guess you do it incorreclty BECAUSE xD[01:21:23] *** doomlord_1 has quit IRC[01:21:28] <derhass_> aethersis: because?[01:21:29] <void256> I'm ultra pro?[01:21:31] <void256> :>[01:21:31] <aethersis> because I did the same error xD[01:21:40] <void256> no it's valid[01:21:47] <aethersis> and it defiinitely must mean you made the same error![01:22:00] <void256> it's so high because it only renders a simple texture or not at all if nothing changed xD[01:22:25] <aethersis> I'm noob :<[01:22:25] <void256> that makes me ultra pro - actually something I've learned in here! don't render if nothing changed - ultra pro tip :>[01:22:41] <aethersis> my solver always changes[01:22:46] <aethersis> chaotic system[01:22:57] <void256> you'll never get 20k fps then ;)[01:23:08] <void256> (which is actually silly but anyway ... ^^)[01:23:13] <aethersis> I will buy ultra pro expensive GPU![01:23:17] <Domx> I need to take a shit[01:23:27] <aethersis> or I will just do printf("FPS: 200 000 000"[01:24:08] *** linuxuz3r has quit IRC[01:24:41] <aethersis> anyways[01:24:44] <aethersis> goodnight everyone[01:25:15] *** Zerflag has quit IRC[01:26:12] *** tcsc has joined ##OpenGL[01:26:39] *** aethersis has quit IRC[01:26:58] *** centrinia has quit IRC[01:27:48] *** derhass_ has quit IRC[01:28:41] *** glYoda has quit IRC[01:29:13] *** glYoda has joined ##OpenGL[01:31:20] *** jdolan has quit IRC[01:31:23] *** void256 has quit IRC[01:32:12] *** glYoda has quit IRC[01:32:25] *** glYoda has joined ##OpenGL[01:32:49] *** bkre has quit IRC[01:33:11] *** jdolan has joined ##OpenGL[01:33:30] *** Misu has quit IRC[01:35:45] *** pizthewiz has joined ##OpenGL[01:36:49] *** narcan has joined ##OpenGL[01:39:57] <dahlia> if one was going to do a 3d convolution couldn't it just call a draw for each 2d plane in the output array?[01:40:10] <dahlia> or did I miss something in that conversation?[01:46:49] *** reed has joined ##OpenGL[01:51:47] *** garFF has joined ##OpenGL[01:53:11] *** Guest30532 is now known as mintaki_[01:54:39] *** jdolan has quit IRC[01:54:54] *** paperManu_ has joined ##OpenGL[01:54:59] *** jdolan has joined ##OpenGL[01:55:46] *** pazul has quit IRC[01:56:24] *** garFF has quit IRC[01:57:03] *** mayankmadan_ has quit IRC[01:58:15] *** BitPuffi1 has joined ##OpenGL[01:58:38] *** mayankmadan_ has joined ##OpenGL[01:59:16] *** BitPuffin has quit IRC[02:08:05] *** mukunda_ has quit IRC[02:08:25] *** mukunda_ has joined ##OpenGL[02:09:34] *** jdolan has quit IRC[02:12:29] *** jdolan has joined ##OpenGL[02:15:18] *** jdolan has quit IRC[02:17:01] *** jdolan has joined ##OpenGL[02:21:04] *** Zerflag has joined ##OpenGL[02:25:42] *** Zerflag has quit IRC[02:41:57] *** pigeon has quit IRC[02:45:31] *** pigeon has joined ##OpenGL[02:51:12] *** reed has quit IRC[02:54:20] *** jdolan has quit IRC[02:55:51] *** BitPuffi1 has quit IRC[03:03:14] *** Garner has quit IRC[03:09:35] *** konom has quit IRC[03:10:18] *** robot-beethoven has joined ##OpenGL[03:10:44] *** jdolan has joined ##OpenGL[03:13:04] *** charlie5 has joined ##OpenGL[03:15:39] *** grzywacz has quit IRC[03:16:14] *** konom has joined ##OpenGL[03:18:03] *** Burga has joined ##OpenGL[03:18:19] *** TyrfingMjolnir has quit IRC[03:21:49] *** Zerflag has joined ##OpenGL[03:23:06] *** XCVB has joined ##OpenGL[03:25:08] <XCVB> Can somone tell why I can't call this function multiple times without textures messing up: http://pastebin.com/j6KnY3bV[03:25:40] <XCVB> I'm try to load a transparent png[03:26:06] <XCVB> s[03:26:09] *** Zerflag has quit IRC[03:27:28] *** slidercrank has quit IRC[03:30:07] *** Jack has joined ##OpenGL[03:30:30] *** Jack is now known as Guest57393[03:32:00] *** Guest57393 has quit IRC[03:34:32] <pink-rg> XCVB: you may want to ensure you set up the texture filetering and number of mips[03:34:40] <pink-rg> https://www.opengl.org/wiki/Common_Mistakes#Creating_a_complete_texture[03:36:48] *** tcsc has quit IRC[03:37:06] *** tcsc has joined ##OpenGL[03:46:37] <XCVB> pink-rg: what are mips[03:48:08] *** MLM has joined ##OpenGL[03:49:17] *** Hunts has joined ##OpenGL[03:50:00] *** urraka has quit IRC[03:51:09] *** clincher has quit IRC[03:51:59] <pink-rg> mip maps[03:52:57] *** HuntsMan has quit IRC[03:53:00] *** clincher has joined ##OpenGL[04:05:09] *** DolpheenDream has quit IRC[04:05:44] *** DolpheenDream has joined ##OpenGL[04:07:38] *** phao has left ##OpenGL[04:09:13] *** TyrfingMjolnir has joined ##OpenGL[04:17:15] *** CainJacobi has joined ##OpenGL[04:19:23] *** xissburg has quit IRC[04:19:53] *** samrat has joined ##OpenGL[04:20:54] *** paperManu_ has quit IRC[04:22:34] *** Zerflag has joined ##OpenGL[04:23:48] *** Orion] has quit IRC[04:27:15] *** lenarhoyt has quit IRC[04:27:30] *** Zerflag has quit IRC[04:27:57] *** luxigo has joined ##OpenGL[04:35:16] *** LifeBlood has joined ##OpenGL[04:37:57] *** shingshang has joined ##OpenGL[04:43:43] *** HaltingState has quit IRC[04:46:32] *** karswell has quit IRC[04:46:36] *** mintaki_ has quit IRC[04:47:17] *** rishin458 has joined ##OpenGL[04:48:00] *** karswell has joined ##OpenGL[04:48:05] <rishin458> hi opengl[04:49:49] <rishin458> is the structure of opengl4 is diff from opengl2[04:50:27] *** Sos has quit IRC[04:51:10] <Stragus> It's very similar, with more flexibility over various parts of the pipeline[04:51:55] <roboman2444> eh, its much much closer to gl3[04:51:59] <roboman2444> gl3 and gl2 are large leaps[04:52:54] *** Cabanoss- has joined ##OpenGL[04:55:53] *** Cabanossi has quit IRC[04:55:53] *** Cabanoss- is now known as Cabanossi[04:56:00] *** rishin458 has quit IRC[04:56:14] *** HaltingState has joined ##OpenGL[05:09:46] *** jdolan has quit IRC[05:10:26] *** jdolan has joined ##OpenGL[05:11:44] *** centrinia has joined ##OpenGL[05:12:45] *** qeed has quit IRC[05:17:23] *** DolpheenDream has quit IRC[05:17:28] *** TheLorax has quit IRC[05:19:26] *** konom has quit IRC[05:19:44] *** Keniyal has joined ##OpenGL[05:20:40] *** fatalhalt has joined ##OpenGL[05:22:42] *** Khlorghaal_ has quit IRC[05:23:23] *** Zerflag has joined ##OpenGL[05:28:19] *** Zerflag has quit IRC[05:28:49] *** samrat has quit IRC[05:33:25] *** pizthewiz has quit IRC[05:45:01] *** XCVB has quit IRC[05:49:19] *** Keniyal_ has joined ##OpenGL[05:51:03] *** Keniyal has quit IRC[05:52:36] <mat^2> hey - for some reason, if I don't perform any draw calls while drawing to an FBO (I only call glClear), when I try to draw that FBO to screen, its texture is not getting updated[05:53:09] <mat^2> I can see in gDEBugger that the FBO texture is cleared, but the front/back buffers are not getting updated[05:53:35] <mat^2> if I do a dummy "glBegin(GL_QUADS); glEnd();" before unbinding the FBO, it works[05:54:26] *** Hadrien has joined ##OpenGL[05:54:33] <mat^2> any ideas?[05:57:32] <mat^2> hnm, looks like it happens on my AMD Radeon only, probably a driver bug then[05:57:34] <mat^2> oh well :([05:58:57] <Stragus> That very much sounds like a drivers bug, indeed[05:58:58] *** Hadrien has quit IRC[05:59:14] *** TyrfingMjolnir has quit IRC[06:00:25] <slime> mat^2: are your drivers up-to-date?[06:00:37] <mat^2> slime, they're very recent[06:01:45] <mat^2> I think they're the latest 14.9.2 beta drivers, but I'm doing very simple stuff[06:01:55] <mat^2> I'd be surprised if the 'beta' part is causing it[06:04:06] <mat^2> I guess I'll just have to live with a dummy draw call for now[06:06:02] *** TyrfingMjolnir has joined ##OpenGL[06:06:10] *** Keniyal has joined ##OpenGL[06:09:43] *** Keniyal_ has quit IRC[06:10:09] *** samrat has joined ##OpenGL[06:13:08] *** MLM has quit IRC[06:13:21] *** charlie5 has quit IRC[06:17:37] *** luxigo has quit IRC[06:19:17] *** pizthewiz has joined ##OpenGL[06:26:41] *** Garner has joined ##OpenGL[06:27:07] *** bb010g has quit IRC[06:28:47] *** tcsc has quit IRC[06:31:14] *** Zupo2 has joined ##OpenGL[06:32:40] *** Zupo has quit IRC[06:38:00] *** __zoot__ has quit IRC[06:44:56] *** BreadProduct has joined ##OpenGL[06:50:33] *** charlie5 has joined ##OpenGL[06:54:10] *** Codex_ has joined ##OpenGL[06:58:50] *** telex has quit IRC[07:00:39] *** telex has joined ##OpenGL[07:04:15] *** CainJacobi has quit IRC[07:04:50] *** charlie5 has quit IRC[07:07:06] *** foreignFunction has joined ##OpenGL[07:07:37] *** rishin458 has joined ##OpenGL[07:09:31] *** hahuang61 has joined ##OpenGL[07:10:35] *** rishin458 has quit IRC[07:12:03] *** Zerflag has joined ##OpenGL[07:14:40] *** tcsc has joined ##OpenGL[07:16:35] *** Zerflag has quit IRC[07:16:47] *** DrBenway has quit IRC[07:32:46] *** fatalhalt has quit IRC[07:36:20] *** shintah has joined ##OpenGL[07:39:15] *** pizthewiz has quit IRC[07:39:48] *** tcsc has quit IRC[07:47:05] *** narcan has quit IRC[07:57:03] *** Twinklebear has quit IRC[07:57:36] *** mayankmadan__ has joined ##OpenGL[07:59:36] *** foreignFunction has quit IRC[08:01:42] *** mayankmadan_ has quit IRC[08:04:56] *** luxigo has joined ##OpenGL[08:09:17] *** luxigo has quit IRC[08:15:29] *** luxigo has joined ##OpenGL[08:18:55] *** SleekoNiko has quit IRC[08:23:46] *** neure has joined ##OpenGL[08:26:18] *** Zupo2 has quit IRC[08:30:50] *** DrSkyLizard has joined ##OpenGL[08:31:51] *** luxigo has quit IRC[08:31:58] *** hexagoxel has joined ##OpenGL[08:32:13] *** Zerflag has joined ##OpenGL[08:34:49] *** puerum has quit IRC[08:36:40] *** Alina-malina has quit IRC[08:37:16] *** Alina-malina has joined ##OpenGL[08:37:17] *** Alina-malina has quit IRC[08:37:17] *** Alina-malina has joined ##OpenGL[08:45:11] *** luxigo has joined ##OpenGL[08:45:26] *** Beetny has joined ##OpenGL[08:54:25] *** bjz has joined ##OpenGL[08:55:17] *** Burga has quit IRC[08:55:42] *** hexagoxel has quit IRC[08:59:28] *** linuxuz3r has joined ##OpenGL[09:00:35] *** slidercrank has joined ##OpenGL[09:01:19] *** clauswitt has joined ##OpenGL[09:06:36] *** Gamecubic has quit IRC[09:07:33] *** Jaus has joined ##OpenGL[09:10:29] <Jaus> Hello, I am trying to use a FrameBufferObject to do post processing effects. I am drawing to the framebuffer and then performing another draw with different shaders. The first I draw, the effect works perfectly[09:10:49] <Jaus> The second pass and onwards all my polygons seem to be rending on the same planw.[09:10:54] <Jaus> plane*[09:12:24] *** redeemed has joined ##OpenGL[09:12:43] *** shintahW has joined ##OpenGL[09:13:45] <Adrinael> Depth testing on?[09:13:47] *** zzuegg has quit IRC[09:13:51] <Adrinael> You attached a depth buffer?[09:14:08] <Jaus> depth buffer is on and depth testing is enabled[09:14:58] *** zzuegg has joined ##OpenGL[09:15:17] <Adrinael> Show a picture[09:15:22] <Bloodust> umm :D[09:15:22] <Adrinael> And some code[09:16:43] <Jaus> one sec[09:17:04] <Jaus> Drawing to the screen I turn off the depth test[09:17:21] <Jaus> and turn it back on when drawing to the frame buffer[09:19:11] <Jaus> http://pastebin.com/SLzjciQC[09:19:59] *** tm604 has quit IRC[09:20:50] <Jaus> http://i.imgur.com/oW5vJek.png[09:20:56] *** Folkol has joined ##OpenGL[09:20:59] <Jaus> http://i.imgur.com/mdK0ySW.png[09:21:25] <Jaus> And those are the images of the first run through the draw loop[09:21:35] <Jaus> and the second run through the draw loop[09:21:50] <Jaus> >=second run[09:22:49] <Bloodust> :D[09:23:13] *** Waynes has joined ##OpenGL[09:23:23] <Jaus> Yeah, Bloodust?[09:23:41] *** zzuegg has quit IRC[09:23:43] *** zzuegg2 has joined ##OpenGL[09:25:33] *** tm604 has joined ##OpenGL[09:34:21] <Adrinael> Are you sure your main scene render goes to the proper target? On the first draw[09:36:09] *** tm604 has quit IRC[09:36:17] *** luxigo has quit IRC[09:36:28] <Jaus> I will prematurely swap my buffers to check now[09:36:30] *** tm604 has joined ##OpenGL[09:37:25] <Jaus> yes[09:37:39] *** clauswitt has quit IRC[09:37:39] *** BitPuffi1 has joined ##OpenGL[09:37:50] <Jaus> my main scene always renders to the FB[09:38:14] *** tm604 has quit IRC[09:38:14] *** tm604 has joined ##OpenGL[09:38:27] *** Hunts has quit IRC[09:39:09] *** Waynes has quit IRC[09:39:26] *** BitPuffi1 is now known as BitPuffin[09:42:31] *** zzuegg2 has quit IRC[09:43:21] *** linuxuz3r has quit IRC[09:43:28] *** Majiet has quit IRC[09:43:43] *** zzuegg has joined ##OpenGL[09:45:27] *** Jeanne-Kamikaze has joined ##OpenGL[09:48:27] <Adrinael> Did your vertex buffers get b0rked on the second draw?[09:52:58] *** bjz has quit IRC[09:54:50] *** bjz has joined ##OpenGL[09:58:00] <Jaus> they are not touched[09:58:02] <Jaus> so idk[09:58:28] *** goliath has joined ##OpenGL[09:58:31] *** groton has joined ##OpenGL[10:00:04] *** shintahW has quit IRC[10:01:25] *** luxigo has joined ##OpenGL[10:04:53] *** zzuegg has quit IRC[10:05:37] <Jaus> Adrinael, I just modified it so that the VBOs would get updated with the same data every single draw.[10:05:44] <Jaus> And it made no difference.[10:06:45] <Adrinael> And you're correctly handling the brick wall texture vs. the framebuffer texture?[10:07:11] *** clauswitt has joined ##OpenGL[10:07:32] *** zzuegg has joined ##OpenGL[10:07:36] <Jaus> the samething occurs with just colored cubes[10:08:43] <Adrinael> I'm still waiting for Bloodust to say something other than ":D"[10:08:56] <Jaus> http://i.imgur.com/7yXF1ua.png <- colored cube example[10:09:38] <Bloodust> eh[10:09:53] <Bloodust> do you clear the depth buffer(s) ?[10:11:04] <Jaus> I believe I clear them on both the FBO and the screen buffer[10:14:06] <Jaus> Bloodust, glClear should do it right?[10:14:13] <Bloodust> yep[10:14:50] <Jaus> I now clear color and depth bits after beinding thr frame buffer[10:21:21] *** rishin458 has joined ##OpenGL[10:25:51] *** rishin458 has quit IRC[10:26:23] *** LifeBlood has quit IRC[10:27:52] *** miyabe has joined ##OpenGL[10:33:10] *** narcan has joined ##OpenGL[10:34:28] *** karab44 has joined ##OpenGL[10:41:29] *** shintahW has joined ##OpenGL[10:41:30] *** navitsu has joined ##OpenGL[10:41:31] *** erhan_ has joined ##OpenGL[10:46:10] *** BtbN has quit IRC[10:46:45] *** BtbN has joined ##OpenGL[10:48:34] *** mat^2 has quit IRC[10:49:23] *** wutata- has joined ##OpenGL[10:51:10] *** mayankmadan__ has quit IRC[10:52:04] *** DrSkyLizard has quit IRC[11:01:31] *** Morzan has joined ##OpenGL[11:03:03] *** Zerflag has quit IRC[11:04:02] *** garFF has joined ##OpenGL[11:05:04] *** garFF has quit IRC[11:07:12] *** luxigo has quit IRC[11:08:49] *** Morzan has quit IRC[11:10:37] *** Drakon has quit IRC[11:14:00] *** zzuegg has quit IRC[11:18:26] *** robot-beethoven has quit IRC[11:18:51] *** garFF has joined ##OpenGL[11:19:39] *** Lemml has joined ##OpenGL[11:21:16] *** TyrfingMjolnir has quit IRC[11:23:18] *** slidercrank has quit IRC[11:36:05] *** paul424 has joined ##OpenGL[11:36:54] <paul424> http://www.ogre3d.org/forums/viewtopic.php?f=1&t=82078[11:36:59] <paul424> my culling idea[11:37:03] *** JoshSyn has joined ##OpenGL[11:40:39] *** Demon_Fox has quit IRC[11:47:11] <Bloodust> how about http://www.umbrasoftware.com/en/[11:47:57] <Bloodust> although its probably crazy expensive[11:48:03] <Bloodust> but you could do what they do :P[11:50:52] *** snakenerd has joined ##OpenGL[11:54:02] *** zzuegg has joined ##OpenGL[11:54:48] *** goliath has quit IRC[11:54:52] *** Crehl has joined ##OpenGL[11:58:03] *** HuntsMan has joined ##OpenGL[12:01:41] *** snakenerd has quit IRC[12:06:01] *** narcan has quit IRC[12:11:58] *** wiky has joined ##OpenGL[12:12:27] *** DrBenway has joined ##OpenGL[12:13:07] *** wiky has quit IRC[12:13:51] *** Zerflag has joined ##OpenGL[12:16:29] *** juanmabc has joined ##OpenGL[12:18:39] *** Zerflag has quit IRC[12:19:22] *** razieliyo has joined ##OpenGL[12:20:06] <paul424> Bloodust: are yopu talking to me ?[12:20:45] <Bloodust> yes[12:20:52] *** Drakon has joined ##OpenGL[12:24:34] *** HuntsMan has quit IRC[12:29:07] *** Garmine42 has quit IRC[12:29:07] *** Garmine42 has joined ##OpenGL[12:29:14] *** Garmine42 is now known as Garmine[12:32:22] *** karab44 has quit IRC[12:32:28] *** Match has joined ##OpenGL[12:35:05] *** groton_ has joined ##OpenGL[12:36:30] *** groton has quit IRC[12:36:32] *** groton_ is now known as groton[12:39:23] *** garFF has quit IRC[12:39:53] *** Crehl has quit IRC[12:41:21] *** razieliyo has quit IRC[12:46:33] *** goliath has joined ##OpenGL[12:54:34] *** samrat has quit IRC[12:56:38] *** navitsu has quit IRC[12:56:53] <paul424> http://www.gamasutra.com/view/feature/164660/sponsored_feature_next_generation_.php?print=1[12:59:01] *** razieliyo has joined ##OpenGL[12:59:01] *** razieliyo has joined ##OpenGL[12:59:33] *** shintahW2 has joined ##OpenGL[13:01:24] *** paul424 has quit IRC[13:02:53] *** shintahW has quit IRC[13:02:53] *** shintahW2 is now known as shintahW[13:03:57] *** miyabe has left ##OpenGL[13:14:21] *** Madsy has quit IRC[13:23:18] *** Codex_ has quit IRC[13:24:41] *** paperManu_ has joined ##OpenGL[13:43:03] *** samrat has joined ##OpenGL[13:46:17] *** zajfy has joined ##OpenGL[13:50:39] *** groton has quit IRC[13:55:15] *** Codex_ has joined ##OpenGL[13:58:24] *** clauswitt has quit IRC[13:58:38] *** urraka has joined ##OpenGL[13:59:59] *** clauswitt has joined ##OpenGL[14:01:05] *** Jeanne-Kamikaze has quit IRC[14:02:48] *** snakenerd has joined ##OpenGL[14:10:48] *** xissburg has joined ##OpenGL[14:14:30] *** Beetny has quit IRC[14:17:42] *** jeaye has quit IRC[14:21:53] *** Madsy has joined ##OpenGL[14:21:53] *** Madsy has joined ##OpenGL[14:23:00] *** jeaye has joined ##OpenGL[14:23:16] *** BitPuffin has quit IRC[14:24:42] *** huhlig has quit IRC[14:25:09] *** ocharles_ has quit IRC[14:25:31] *** devbug has joined ##OpenGL[14:25:39] *** DaGardner has quit IRC[14:25:40] *** Shadda has quit IRC[14:25:43] *** oberstein has quit IRC[14:25:44] *** skarn has quit IRC[14:25:44] *** edenist has quit IRC[14:26:30] *** ocharles_ has joined ##OpenGL[14:26:39] *** edwardk has quit IRC[14:27:11] *** edenist has joined ##OpenGL[14:27:12] *** skarn has joined ##OpenGL[14:27:38] *** huhlig has joined ##OpenGL[14:27:56] *** garFF has joined ##OpenGL[14:29:30] *** devbug_ has quit IRC[14:29:38] *** edwardk has joined ##OpenGL[14:30:02] *** goliath has quit IRC[14:30:33] *** oberstein has joined ##OpenGL[14:32:54] *** garFF has quit IRC[14:34:04] *** Shadda has joined ##OpenGL[14:37:26] *** DaGardner has joined ##OpenGL[14:39:22] *** goliath has joined ##OpenGL[14:44:09] *** Zupo2 has joined ##OpenGL[14:46:44] *** Waynes has joined ##OpenGL[14:47:42] *** drmugg123 has joined ##OpenGL[14:48:01] <drmugg123> okay so i need some advice[14:48:12] <drmugg123> what would be faster for 2d rendering of animated sprites[14:48:49] <drmugg123> instancing with no indices and a model matrix per sprite(which is recalculated every frame) and updating UV coordinates[14:49:16] <drmugg123> or constantly reuploading modified vertices and UVs with glBufferSubData?[14:49:51] <drmugg123> I mean instancing would allow me to batch the sprites easily[14:51:28] *** goliath has quit IRC[14:51:47] <drmugg123> if i send the new vertices i would only update at max 12 vertices per frame[14:52:07] *** DaGardner has quit IRC[14:52:07] <drmugg123> using instancing i would send and calculate a 4*4 matrix and calculate and send 6 vertices[14:52:16] *** cmr has quit IRC[14:52:24] <drmugg123> but i could easily batch it[14:53:06] <drmugg123> am i going to be fillrate limited before i hit any other bottleneck?[14:53:54] *** cmr has joined ##OpenGL[14:53:54] *** cmr has joined ##OpenGL[14:54:37] <ClaudiusMaximus> drmugg123: i'd go with what is easiest to implement, then if there's a performance problem profile and optimize the hot spots[14:56:15] <urraka> for 2d you dont need a 4x4 matrix.. you can do with 3x2 (unless you do some weird stuff that i dont know about)[14:58:09] <drmugg123> okay[14:58:23] *** PasNox has joined ##OpenGL[14:58:24] <drmugg123> i'll try the "recalc vertices" way[15:00:46] *** DaGardner has joined ##OpenGL[15:01:03] *** PasNox has quit IRC[15:01:57] *** bjz has quit IRC[15:02:17] *** bjz has joined ##OpenGL[15:02:50] *** PasNox has joined ##OpenGL[15:04:16] *** garFF has joined ##OpenGL[15:04:22] *** PasNox has quit IRC[15:04:45] *** PasNox has joined ##OpenGL[15:10:19] *** bjz has quit IRC[15:10:36] *** bjz has joined ##OpenGL[15:12:54] *** clauswitt has quit IRC[15:13:52] *** razieliyo has joined ##OpenGL[15:15:01] *** PasNox has quit IRC[15:16:07] *** ruuns has joined ##OpenGL[15:20:16] *** tehrain has joined ##OpenGL[15:20:16] *** tehrain has joined ##OpenGL[15:26:24] *** BitPuffin has joined ##OpenGL[15:26:49] *** Ad1_RnR has joined ##OpenGL[15:27:23] *** xishem has joined ##OpenGL[15:30:23] *** xishe has quit IRC[15:30:24] *** Ad1_RN has quit IRC[15:31:52] *** drmugg123 has quit IRC[15:31:58] *** barosl has joined ##OpenGL[15:43:44] *** lenarhoyt has joined ##OpenGL[15:44:52] *** neurre has joined ##OpenGL[15:52:11] *** telex has quit IRC[15:53:14] *** neure has quit IRC[15:54:39] *** telex has joined ##OpenGL[15:55:08] *** konradsurf has joined ##OpenGL[15:57:07] *** kriskropd has joined ##OpenGL[15:58:12] *** DrBoneStrong has joined ##OpenGL[16:00:54] *** DrBoneStrong has left ##OpenGL[16:02:42] *** TyrfingMjolnir has joined ##OpenGL[16:04:06] *** suppahsrv has quit IRC[16:05:09] *** DrSkyLizard has joined ##OpenGL[16:07:43] <AbiBuccaneer> my experiences with coding metal so far: being repeatedly told by apple's documents that all the shader compilation happens offline, and then getting runtime errors that say "Internal compiler error" when I try and load those "offline" shaders[16:07:50] *** PasNox has joined ##OpenGL[16:10:38] <bkc_> hehe[16:16:45] *** PasNox_ has joined ##OpenGL[16:17:44] *** konradsurf has quit IRC[16:17:54] *** PasNox has quit IRC[16:26:29] *** PasNox has joined ##OpenGL[16:27:36] *** HuntsMan has joined ##OpenGL[16:30:17] *** PasNox_ has quit IRC[16:33:50] *** AbiBuccaneer is now known as AbigailBuccaneer[16:47:14] *** PasNox has quit IRC[16:47:33] *** PasNox has joined ##OpenGL[16:48:32] *** slidercrank has joined ##OpenGL[16:49:50] *** redeemed has quit IRC[16:49:55] *** DrSkyLizard has quit IRC[16:50:42] *** Waynes1 has joined ##OpenGL[16:52:52] *** Waynes has quit IRC[16:57:48] *** Zupo2 has quit IRC[16:58:10] *** Keniyal has quit IRC[16:58:56] *** MLM has joined ##OpenGL[16:59:35] *** goliath has joined ##OpenGL[16:59:57] *** wutata- has quit IRC[17:00:08] *** Zupo2 has joined ##OpenGL[17:03:08] *** [H]unt3r has joined ##OpenGL[17:07:56] *** LifeBlood has joined ##OpenGL[17:13:54] *** Gamecubic has joined ##OpenGL[17:14:21] *** Zupo3 has joined ##OpenGL[17:16:12] *** Zupo2 has quit IRC[17:20:57] *** Folkol has quit IRC[17:22:29] *** PasNox_ has joined ##OpenGL[17:22:33] *** Chunk2 has quit IRC[17:23:32] *** pazul has joined ##OpenGL[17:24:17] *** PasNox has quit IRC[17:26:42] *** Chunk2 has joined ##OpenGL[17:27:15] *** PasNox_ is now known as PasNox[17:29:44] *** Ad1 has joined ##OpenGL[17:30:32] *** Zerflag has joined ##OpenGL[17:34:20] *** Zupo3 has quit IRC[17:34:41] *** Zupo2 has joined ##OpenGL[17:35:11] *** Zupoman has joined ##OpenGL[17:38:17] *** kaleid0 has joined ##OpenGL[17:39:51] *** shintahW2 has joined ##OpenGL[17:42:23] *** pazul has quit IRC[17:42:52] *** centrinia has quit IRC[17:43:10] *** shintahW has quit IRC[17:43:10] *** shintahW2 is now known as shintahW[17:45:49] <Jaus> about 8 hous later and I still have no idea what is going on[17:47:21] *** Kane has joined ##OpenGL[17:51:28] *** Zerflag has quit IRC[17:54:41] <Codex_> i had similar problem few days ago, some sprites just disappeared when i added more sprites to a scene; it turned out that two ways of blitting sprites used the same variables...[17:55:27] <Codex_> they didnt have independent state any longer[17:56:13] <Codex_> it was a quick fix though, just renaming variables etc[17:57:03] <Jaus> did they dissapear or were they drawing strangely?[17:57:33] <Codex_> just disappeared[17:58:12] <Jaus> mine are drawing strangely[18:01:29] <Codex_> i did this 2 minutes ago: http://sivut.koti.soon.fi/~terop/face.png[18:02:00] <Codex_> good for my google+ profile pic[18:02:09] <Bloodust> mspaint 101[18:02:32] <Codex_> i had to write some opengl code to do it[18:12:31] *** Orion] has joined ##OpenGL[18:12:44] *** TyrfingMjolnir has quit IRC[18:16:46] *** japro has joined ##OpenGL[18:18:00] *** linuxuz3r has joined ##OpenGL[18:21:51] <Jaus> an update to my previous issue[18:22:13] <Jaus> I now realize that only the first orimative in every VBO is being draw[18:22:18] <Jaus> primative[18:22:40] *** doev has joined ##OpenGL[18:22:47] <Codex_> so, primitive count is wrong?[18:23:02] <Jaus> primitive count is static[18:23:08] <Codex_> or you lose the attribute data?[18:23:28] *** stefkos has joined ##OpenGL[18:23:29] <Jaus> maybe, I don't know how I would have lost the attribute data[18:23:36] *** jorj has joined ##OpenGL[18:24:47] <japro> maybe you dropped it?[18:25:05] *** konom has joined ##OpenGL[18:25:06] <AbigailBuccaneer> yes check down the back of the sofa[18:25:12] *** SugarCube has quit IRC[18:26:24] <japro> has anyone made experiments about he cost of degenerate triangles?[18:27:05] <AbigailBuccaneer> my local conservative councillor says degenerate youth are going to ruin the nation. i don't know if triangles have the same effect but it sounds pretty costly[18:27:06] *** garFF has quit IRC[18:28:58] <Codex_> if it's a triangle, its evil anyway[18:29:26] <AbigailBuccaneer> the most evil primitive since GL_PENTAGRAMS[18:29:53] <AbigailBuccaneer> (though i'm wearing top that literally just has a giant inverted pentagram on it, so what do i know)[18:30:20] * japro is dressed as a tree[18:31:01] <AbigailBuccaneer> ah, the dying generation of ghillie suit coders...[18:31:41] <japro> related... https://twitter.com/JakobProgsch/status/532299060049633281[18:32:37] *** Renderwahn has quit IRC[18:33:04] *** mockillo has quit IRC[18:34:47] <japro> anyway the degenerate triangle was because i'm currently doing stencil shadows and instead of generating te volumes in a GS which seems the hip thing now i just put a index buffer over a unindexed vbo that contains collapsed quads for each edge[18:35:04] <japro> that way i can just extrude all vertices with a back facing normal along the light direction[18:35:25] <japro> i just means there is plenty of degenerate triangles in the non expanded edges[18:35:34] <AbigailBuccaneer> japro, that's sneaky[18:36:25] *** tcsc has joined ##OpenGL[18:36:34] <Codex_> oh, I should check if I can do shadow for the face.png :)[18:36:50] <Codex_> it sounds like an easy task, and could be useful later too[18:37:42] *** shingshang has quit IRC[18:37:44] *** erhan_ has quit IRC[18:38:16] *** SugarCube has joined ##OpenGL[18:40:51] *** ManDay has joined ##OpenGL[18:41:11] *** neurre has quit IRC[18:42:16] *** japro has quit IRC[18:45:43] *** CainJacobi has joined ##OpenGL[18:48:30] *** Misu has joined ##OpenGL[18:51:34] *** thegrid has joined ##OpenGL[18:51:47] *** linuxuz3r has quit IRC[18:52:31] <thegrid> hello all, can i upload the image to the gpu, then convert it? im reading this good post, and im told thats not how opengl works[18:52:36] *** stoopdapoop has quit IRC[18:52:44] <thegrid> am i crazy?[18:52:48] <Bloodust> you can[18:52:52] <Bloodust> depends what convering is tho[18:53:02] <thegrid> so i have a bmp[18:53:16] <thegrid> it can be uploaded to the gpu? what[18:53:28] <Bloodust> well you have to parse it first on the cpu side[18:53:32] <Bloodust> opengl doesnt understand file formats[18:54:21] <thegrid> yeah i thought so[18:54:31] <thegrid> this article is worded funny[18:54:52] <AbigailBuccaneer> thegrid, OpenGL supports various uncompressed texture formats (not file formats like .bmp, .tga etc., but raw arrangements of bytes, like GL_RGBA8_UNORM, etc.)[18:55:10] *** Khlorghaal has joined ##OpenGL[18:55:23] <Bloodust> opengl also supports a few compressed texture formats[18:55:23] <AbigailBuccaneer> additonally, there are various compressed texture formats which you can upload to the GPU as-is and it'll stay compressed in GPU memory and will be essentially decompressed on-the-fly when you do a texture access[18:56:02] <AbigailBuccaneer> the cool and hip format at the moment is ASTC... though being an OpenGL ES coder, I don't know what version of desktop OpenGL that's available in[18:56:08] <Jaus> thegrid, SOIL is a tiny public library that can parse a lot of file formats. You could use it to make an OpenGL texture.[18:56:18] <Bloodust> I personally use freeimage[18:56:32] *** faty has quit IRC[18:56:35] <Bloodust> oh and GLI[18:56:41] <Bloodust> works nice with glm[18:56:44] <AbigailBuccaneer> i export everything as .KTX files, because i have a proper asset export pipeline ;)[18:56:47] <thegrid> yeah ok, and how many 120x90 RGBA images can i store on an gl es 2.0 compatible chip?[18:56:57] <thegrid> just an estimate[18:57:09] <Bloodust> TheFlash how much vram does gl es 2.0 compatible chip have?[18:57:11] <thegrid> it says its a videocore chip[18:57:47] <thegrid> and what if i have 50 textures to display at once?[18:57:49] *** abs25 has joined ##OpenGL[18:58:06] <AbigailBuccaneer> thegrid, if you want hardware details about mobile devices, http://gfxbench.com/result.jsp?benchmark=gfx27 is a good place to look[18:58:21] <Jaus> you can draw them in sequence?[18:59:14] <AbigailBuccaneer> for memory: pessimistically assume that your 120x90 image will be stored as 128x128. (this may not be that pessimistic, i can't remember the limitations on NPOT and non-square textures in ES2 off the top of my head so you may have to pad them to that anyway)[18:59:23] *** JoshSyn has quit IRC[18:59:35] *** mockillo has joined ##OpenGL[19:00:00] <thegrid> thx for the tip[19:00:01] <AbigailBuccaneer> i assume by 'RGBA' you mean 'RGBA8', ie. four bytes per pixel, which gives a size of 128*128*4 = 65536 bytes = 16KiB per texture[19:00:41] <AbigailBuccaneer> *64KiB per texture, whoops[19:02:15] *** Zerflag has joined ##OpenGL[19:02:32] *** Misu2 has joined ##OpenGL[19:02:39] <AbigailBuccaneer> remembering numbers off the top of my head, one of our games uses about 40-50MiB of texture memory on low-end devices, i think[19:02:54] *** Misu has quit IRC[19:03:00] <AbigailBuccaneer> which means you'd be able to store 640 of those textures[19:03:25] <AbigailBuccaneer> so realistically it sounds like the answer is "store as many as you want, you won't run out of space unless you're doing something weird"[19:03:32] *** hahuang61 has quit IRC[19:03:42] <AbigailBuccaneer> anything that's not a low-end device will have a lot more memory than that available for you to use[19:04:49] <AbigailBuccaneer> if you want 50 textures to display at once on screen then you should be able to do that just fine by just drawing them sequentially[19:05:17] <AbigailBuccaneer> if that's not fast enough you could store your textures in a texture atlas (ie. one large texture which contains several smaller images as subrectangles)[19:05:31] <AbigailBuccaneer> and then you can batch your draw calls better and get better cache usage[19:05:35] *** navitsu has joined ##OpenGL[19:05:45] *** Misu has joined ##OpenGL[19:05:46] <AbigailBuccaneer> thegrid, ^[19:06:59] *** Zerflag has quit IRC[19:08:25] *** lenarhoyt has quit IRC[19:09:30] *** Misu2 has quit IRC[19:10:06] *** plash has quit IRC[19:11:03] *** japro has joined ##OpenGL[19:11:14] <japro> what did i miss?[19:12:06] <Bloodust> nothing[19:12:21] *** hexagoxel has joined ##OpenGL[19:12:41] <AbigailBuccaneer> japro, full log * japro has quit (Quit: Leaving)[19:12:42] <AbigailBuccaneer> * CainJacobi (~CainJacob at 24-113-58-44 dot wavecable.com) has joined[19:12:42] <AbigailBuccaneer> * Misu (~misu at ner95-3-78-192-196-78 dot fbxo.proxad.net) has joined[19:12:42] <AbigailBuccaneer> * thegrid (45f6fa73 at gateway/web/freenode/ip dot 69.246.250.115) has joined[19:12:42] <AbigailBuccaneer> * linuxuz3r has quit (Quit: Leaving)[19:12:42] <AbigailBuccaneer> <thegrid> hello all, can i upload the image to the gpu, then convert it? im reading this good post, and im told thats not how opengl works[19:12:45] *** AbigailBuccaneer has left ##OpenGL[19:12:47] *** AbigailBuccaneer has joined ##OpenGL[19:12:49] <AbigailBuccaneer> * stoopdapoop has quit ()[19:12:51] <AbigailBuccaneer> <thegrid> am i crazy?[19:12:53] <AbigailBuccaneer> <Bloodust> you can[19:12:54] <Bloodust> please no[19:12:55] <AbigailBuccaneer> <Bloodust> depends what convering is tho[19:12:57] <AbigailBuccaneer> <thegrid> so i have a bmp[19:12:59] <AbigailBuccaneer> <thegrid> it can be uploaded to the gpu? what[19:13:01] <AbigailBuccaneer> <Bloodust> well you have to parse it first on the cpu side[19:13:03] <Bloodust> oh ffs[19:13:03] <AbigailBuccaneer> <Bloodust> opengl doesnt understand file formats[19:13:05] <AbigailBuccaneer> <thegrid> yeah i thought so[19:13:10] *** AbigailBuccaneer has quit IRC[19:13:24] *** AbigailBuccaneer has joined ##OpenGL[19:13:31] <Codex_> hmm, some kind of shadow: http://sivut.koti.soon.fi/~terop/face_shadow.png[19:13:44] <AbigailBuccaneer> gawd i'm so sorry, i meant to post a link that i'd just pastebinned. http://pastebin.com/0FfDhJSx i hope i didn't spam too hard[19:13:54] <Codex_> didnt take too long time[19:13:55] <Bloodust> only half a page[19:13:59] <japro> lol[19:14:20] <AbigailBuccaneer> well, as somebody who could never do anything wrong, i'd like to blame japro for my incompetence. :)[19:15:00] <japro> who could argue with a pirate... especially since i forgot my cutlass[19:17:44] <japro> the "upload bmp to gpu" thing seems popular right now?[19:18:12] *** tcsc has quit IRC[19:18:16] <japro> also... who uses bmps?[19:18:22] <japro> like for anything?[19:18:30] <Jaus> parsing them is easy?[19:18:42] <AbigailBuccaneer> japro, people who know that compressed formats are marginally harder to upload to the GPU, and inexplicably don't have a .TGA viewer[19:19:01] *** thegrid has quit IRC[19:19:28] <japro> i'm going to make my own no-bs image format :D (i will be all bs obviously)[19:20:00] <japro> i suggest using only rgba8 format in landscape format and width and height need to be prime[19:20:00] <Codex_> make it compressed ascii[19:20:13] <japro> no need to story any metadata that way[19:20:16] <japro> store[19:20:18] <AbigailBuccaneer> japro, i was just about to suggest that you could do exactly that![19:20:56] <AbigailBuccaneer> for real though if anybody needs an image format for working with OpenGL, khronos has your back with the KTX file format. https://www.khronos.org/opengles/sdk/tools/KTX/file_format_spec/[19:20:56] <Bloodust> japro make it assume its always square texture, that way you save space by not saving width and height to the file[19:21:09] <japro> Bloodust, uh, i dont :D[19:21:27] <japro> if th dimensions are prime you only have to factor the file size/4 :)[19:21:50] <AbigailBuccaneer> graphics programmers: "it's easy, you only have to factor this integer"[19:21:58] <AbigailBuccaneer> computer security programmers: *flinches*[19:22:35] <japro> hrhr... that by the time that becomes an issue you are looking at stupidly large images though ;)[19:23:14] <japro> i played with some general number sieve once, that factored a 300bit number in like 2 hours[19:24:16] <AbigailBuccaneer> japro, modifcation to your file spec: if a portrait image is required, this can be specified by transposing the pixels and rotating the user's monitor[19:24:29] *** __zoot__ has joined ##OpenGL[19:24:30] <ManDay> japro: and you were watching it all the two hours, right[19:24:40] *** sedeki has joined ##OpenGL[19:24:46] <japro> ManDay, are you suggesting it cheated?[19:24:48] *** ShadowIce has joined ##OpenGL[19:25:08] <ManDay> japro: :-D[19:25:25] <ManDay> it was lucky[19:26:04] <japro> the funny thing is those general number sieves are probablistic[19:26:29] <japro> like they compute those funky binary vectors and each of those gives a 1/2 chance to find a factorization[19:26:34] *** [H]unt3r has quit IRC[19:27:07] <Stragus> Was that 300 bits number a product of two huge primes? It's supposed to take longer[19:27:19] <japro> maybe it was 4 hours[19:27:25] <AbigailBuccaneer> Stragus, no, it was 2^300[19:27:26] <ManDay> Stragus: I bet he gave it a prime :-P[19:27:51] <japro> http://cado-nfs.gforge.inria.fr/ <- this program[19:28:02] *** mat^2 has joined ##OpenGL[19:28:14] <ManDay> There is nothing INRIA can't do :-D[19:28:40] <AbigailBuccaneer> INRIA can't beat me in a fist-fight. the cowards[19:30:06] <japro> they quote 44h for rsa-120 (390bit)[19:30:26] <ManDay> I believe no one here can beat you in a fist-fight AbigailBuccaneer :)[19:30:33] <AbigailBuccaneer> <3 ManDay you're so sweet[19:30:43] <ManDay> amnt I ^^[19:30:48] *** BearishMushroom has joined ##OpenGL[19:31:49] <ManDay> japro: how does that not comprise security of the typical encription schemes?[19:31:55] *** Match has quit IRC[19:32:10] <japro> well those have way longer keys[19:32:14] <ManDay> I thought I had heard it would take decades to crack TLS et al?[19:32:30] <ManDay> oh[19:32:32] <japro> iirc the factorization records are in the order of 768bit and that takes months... on clusters[19:32:57] <japro> the smallest keys in use are 1024bit iirc. which is multiple orders of magnitudes harder[19:33:06] <ManDay> yeah, got it[19:33:16] <ManDay> i had some idea of keys being 256bits long in my head[19:33:31] <japro> for symmetric ciphers i guess[19:33:41] <ManDay> hm[19:34:02] <japro> but factoring those doesn't do anthing :D[19:35:18] *** hexagoxel has quit IRC[19:35:36] <ManDay> has anyone started on a lookup table for keys yet?[19:35:44] <ManDay> i know there was that project for md5 lookup[19:36:03] <ManDay> i mean how many storage would it require to hash all 1024bit keys...[19:36:03] <japro> that is called "rainbow tables" i think?[19:36:10] <ManDay> rainbow?[19:36:11] *** hahuang61 has joined ##OpenGL[19:36:13] <japro> at least for he purpose of password cracking[19:36:26] <Jaus> salt your passwords kids[19:36:36] <Jaus> you don't want to get rainbow tabled[19:37:15] *** [H]unt3r has joined ##OpenGL[19:37:29] <ManDay> they still won't know my password[19:37:37] <japro> password123?[19:37:40] <ManDay> it contains an embarassing secret[19:37:50] <ManDay> japro: not quite[19:37:52] <ManDay> but close![19:38:02] <ManDay> (in the md5 metric that is)[19:40:18] *** ShadowIce has quit IRC[19:42:34] *** ShadowIce has joined ##OpenGL[19:43:28] *** plash has joined ##OpenGL[19:45:42] *** __zoot__ has quit IRC[19:48:01] <ManDay> japro, have any new interesting videos of one of your projects to show us? ^^[19:48:59] *** foreignFunction has joined ##OpenGL[19:49:25] *** __zoot__ has joined ##OpenGL[19:50:03] *** foreignFunction has quit IRC[19:55:34] *** ShadowIce has joined ##OpenGL[19:55:42] *** ShadowIce has quit IRC[19:56:13] *** Puffin has joined ##OpenGL[19:56:35] *** Puffin has quit IRC[19:56:54] *** drag0nius has joined ##OpenGL[19:58:02] *** doboo has joined ##OpenGL[19:58:44] *** SugarCube has quit IRC[19:59:07] <japro> ManDay, not really[19:59:52] *** SugarCube has joined ##OpenGL[20:00:06] <japro> apart from the old ones on youtube... and i can't make one from the current thing on this crapbook (also it wouldn't be that interesting anyway since it's just lots of bricks with shadows...)[20:00:15] <drag0nius> dumb question, but how do i make light's orientation align with environment (which's every point is multiplied by camera's rotation matrix) instead of camera?[20:00:30] *** PasNox has quit IRC[20:03:30] *** __zoot__ has quit IRC[20:03:32] <japro> either do the lighting in worldspace or apply the inverse view transform to it?[20:04:16] *** Renderwahn has joined ##OpenGL[20:04:30] *** foreignFunction has joined ##OpenGL[20:05:06] *** __zoot__ has joined ##OpenGL[20:06:10] *** grid_ has joined ##OpenGL[20:07:03] <grid_> so i have about 50 textured quads to draw: i will try multiple calls to glDrawElements? will they be in the same frame?[20:07:41] <japro> sure[20:08:10] <japro> usually your frame is defined by what happens between two "SwapBuffers" calls[20:08:57] *** hexagoxel has joined ##OpenGL[20:09:14] *** Jeanne-Kamikaze has joined ##OpenGL[20:11:32] *** __zoot__ has quit IRC[20:13:03] *** __zoot__ has joined ##OpenGL[20:14:18] *** bkre_ has joined ##OpenGL[20:15:08] <ManDay> japro: did you just suggest that you're using a mac?[20:16:00] <ManDay> japro: i have a feeling your bricks and shadows are exciting anyway[20:16:05] *** foreignFunction has quit IRC[20:16:11] <ManDay> how many bricks are there?[20:16:39] <ManDay> If I can't see it, describe it to me! I'm craving for something interesting! :-D[20:18:13] *** Biliogadafr has joined ##OpenGL[20:18:41] <japro> lol[20:19:19] <drag0nius> meh, can't think today[20:25:36] *** paul424 has joined ##OpenGL[20:27:38] *** suppahsrv has joined ##OpenGL[20:30:37] *** doboo has quit IRC[20:31:00] *** roboman2444 has quit IRC[20:32:35] <japro> ManDay, http://gfycat.com/BluePoorAfricanbushviper[20:32:41] *** pazul has joined ##OpenGL[20:33:28] <japro> just about runs on this celeron netbooks HD graphics[20:33:43] <foobaz> looks nice[20:33:59] <drag0nius> i am not doing it in opengl, in each frame i: 1. clone whole environment 2. translate -> rotate -> project everything 3. i do z-buffer, then for each displayed point i calculate flat-shading, but how the heck do i preserve orientation (i've light orientation as normal vector, camera rotation in radians and camera rotation matrix)[20:34:02] *** DolpheenDream has joined ##OpenGL[20:34:41] <drag0nius> hmm, might be that i should do the light calculation before projecting?/[20:35:25] *** derhass has joined ##OpenGL[20:36:19] <japro> i should try if carmacks reverse has different performance here[20:37:09] <japro> i'm using the uh depth pass (?) algorithm and fix the camera-in-shadow problem by "preloading" the stencil buffer before any geometry is rendered to the depth buffer[20:37:20] <japro> so i effectively draw the shadow volumes twice[20:38:14] <japro> interestingly the expensive part is the deferred resolve which is singlehandedly more expensive the both stencil passes together[20:38:59] *** pazul has quit IRC[20:39:04] *** mm765 has quit IRC[20:39:54] <Bloodust> bricks for shadows are cool[20:42:10] *** doev has quit IRC[20:45:44] *** __zoot__ has quit IRC[20:46:10] *** __zoot__ has joined ##OpenGL[20:47:42] <japro> i also learned that on the hd graphics decadently using RGBA32F g-buffers has quite a hefty impact (compared to using RGBA16F)[20:48:24] <japro> my desktop cards didn't seem to care too much about that[20:48:29] *** samrat has quit IRC[20:48:55] <ManDay> japro: that flickering annoys me[20:49:03] <ManDay> apart from that, nice city[20:49:18] *** soulz has quit IRC[20:49:42] *** soulz has joined ##OpenGL[20:50:51] <japro> hmm, that might be a gif artifact... at least the wonky color changes on the tall red building are[20:51:11] <ManDay> gifs are so 1995 anyway[20:51:23] <ManDay> webm baby, webm :-D[20:51:44] *** pazul has joined ##OpenGL[20:51:48] <ManDay> eh, i wish i had some time to do some coding right now. but probably wont in a year or so[20:51:54] <ManDay> had to put everything on ice :-(((([20:52:18] *** lenarhoyt has joined ##OpenGL[20:52:57] <japro> drag0nius, after applying the projection matrix pretty much all bets are off[20:55:14] <drag0nius> worst case i'll just say it works xD[20:55:26] <drag0nius> if i won't have enough time[20:55:29] <drag0nius> to rewrite it[20:56:23] *** HuntsMan has quit IRC[20:57:10] <drag0nius> i guess it would be very hard to shade just the visible parts?[20:58:11] *** roboman2444 has joined ##OpenGL[20:58:24] <japro> "when in doubt the program is right and reality is wrong"[21:01:42] <drag0nius> http://grafika.drag0nius.net/ looks pretty convincing to me xD[21:02:48] *** BearishMushroom_ has joined ##OpenGL[21:02:52] *** [H]unt3r has quit IRC[21:02:57] *** Crehl has joined ##OpenGL[21:02:58] *** [H]unt3r has joined ##OpenGL[21:04:16] *** BearishMushroom has quit IRC[21:04:50] <grid_> glUniform1i(_textureUniform, 0); << this is a texture uniform passing to shader, why is the 2nd argument 0? im passing to a sampler2d[21:05:31] <roboman2444> likely because you are setting that sampler to texture unit 0[21:06:51] <grid_> they have used glGenTexture to get an id, shouldnt they use that?[21:07:14] <grid_> sampler2d uses id 0 for the texture right?[21:08:01] <grid_> k just read it, thanks[21:09:33] <japro> naaaaaah, using the actual texture would be too simple[21:09:40] *** meoblast001 has joined ##OpenGL[21:09:45] <japro> we need at least one indirection via binding points in opengl[21:13:24] *** __zoot__ has quit IRC[21:13:26] *** Zupoman has quit IRC[21:13:43] *** Zupoman has joined ##OpenGL[21:13:43] *** Zupoman has joined ##OpenGL[21:13:55] <roboman2444> grid_, no[21:14:02] <roboman2444> grid_, you have texture units[21:14:08] <roboman2444> these range from 0-whatever[21:14:19] <roboman2444> you set your sampler2ds in a shader to these texture units[21:14:21] <grid_> yeah i see them, im ready to go[21:14:39] <roboman2444> then you bind a texture (say textureid 34) to a texture unit( say 0 for most cases)[21:14:45] *** ManDay has quit IRC[21:15:11] *** __zoot__ has joined ##OpenGL[21:15:12] <grid_> oh ok[21:15:29] <grid_> thats very helpful, i need that bit[21:15:55] <grid_> how do you bind to a texture unit[21:16:06] <derhass> glBindTexture[21:16:09] *** __zoot__ has quit IRC[21:16:16] <Bloodust> glActiveTexture and then glBindTexture[21:16:34] *** cmr has left ##OpenGL[21:16:35] <grid_> activetexture(tex_unit)[21:16:43] <grid_> ill read it thanks[21:18:28] *** pazul has quit IRC[21:22:11] *** chirpsalot has quit IRC[21:28:10] *** chirpsalot has joined ##OpenGL[21:31:13] *** erhan_ has joined ##OpenGL[21:33:30] *** pizthewiz has joined ##OpenGL[21:34:35] *** Keniyal has joined ##OpenGL[21:37:00] *** bjz_ has joined ##OpenGL[21:37:00] *** bjz has quit IRC[21:39:21] *** sedeki has quit IRC[21:40:56] *** HuntsMan has joined ##OpenGL[21:42:10] *** grid_ has quit IRC[21:48:03] *** Black0range has joined ##OpenGL[21:49:03] <Black0range> Hello ladies and gents, just wondering whats the absolute minimum number of verticies required to make a cube? And how? :)[21:49:11] *** drag0nius has quit IRC[21:50:40] <derhass> Black0range: that deoends on a lot of thingas[21:50:43] *** snakenerd has quit IRC[21:51:22] <foobaz> to figure that out, unfold the cube[21:51:32] <foobaz> i suspect you could do it with one or two triangle strips[21:51:33] *** snakenerd has joined ##OpenGL[21:51:44] <derhass> Black0range: keeping the non-obviuos solutions like 3 or even 1 aside, you'll get 8.[21:51:45] <foobaz> but only if faces can share vertexes[21:52:05] <foobaz> usually, you need different normals for different faces, so you can’t share vertexes[21:52:06] <Bloodust> Black0range give nothing to opengl, generate 3 faces[21:52:26] <Bloodust> in geometry shader or vertex shader[21:52:27] *** snakenerd has quit IRC[21:52:57] <foobaz> how to unfold cube: http://gwydir.demon.co.uk/jo/solid/cube.htm[21:53:41] <derhass> foobaz: the unfolding isn't really relevant, since the question is only about vertices. and we have indexed rendering[21:54:20] <foobaz> ah true[21:55:07] <Black0range> Hmm well without indexing :)[21:55:55] <Black0range> can you do it with one triangle strip using front_face clockwise?[21:56:09] <foobaz> no[21:56:21] <Black0range> soo two triangle strips are the minimum?[21:56:38] <foobaz> can you share vertexes between faces?[21:56:58] <Black0range> sure, why not :)[21:57:02] <foobaz> normals are why not[21:57:33] <Black0range> normals? :)[21:57:42] * foobaz facepalm[21:58:19] <Black0range> Sorry i'm neither native english or very good at opengl :)[21:58:34] <foobaz> will the cube have lighting applied to it?[21:58:46] <Black0range> should be able to, yes[21:58:57] <foobaz> then you can’t share vertexes between faces[21:59:05] <foobaz> because you need different normals[21:59:25] <Black0range> :( so one strip per face?[21:59:31] <derhass> you'll end up with the standrad 24 vertices cube[21:59:34] <Bloodust> no strips[21:59:35] <Bloodust> just triangles[21:59:37] <derhass> *standard[21:59:54] <foobaz> why no strips? face culling?[21:59:54] <derhass> and of course you want to use indexed rendering[22:00:39] <foobaz> hmm yea, one strip per face means the two triangles have different winding order, so you can’t use it with face culling[22:00:50] <derhass> what?[22:01:01] <foobaz> am i wrong?[22:01:05] <derhass> yes[22:01:21] <derhass> strips would be totally useless if the winding order would change[22:01:38] *** snakenerd has joined ##OpenGL[22:01:41] <Black0range> well i'm using face culling and using strips[22:01:54] <Black0range> or i'm not doing something correctly :)[22:02:08] <foobaz> what order would the vertexes be in? like this?[22:02:12] <foobaz> 1 2[22:02:14] <foobaz> 3 4[22:02:29] <foobaz> 123 is clockwise, 234 is anticlockwise[22:02:48] <Bloodust> counter clockwise :P[22:02:53] <Bloodust> Ill show myself out now[22:03:19] *** meoblast001 has quit IRC[22:05:34] <derhass> foobaz: look at the gl spec[22:06:06] <derhass> the have this nice picture[22:06:47] <foobaz> whoa[22:06:51] <foobaz> “Therefore, the first vertex is stored as vertex A, the second stored as vertex B, the third stored as vertex A, and so on. Any vertex after the second one sent forms a triangle from vertex A, vertex B, and the current vertex (in that order).”[22:09:20] <foobaz> Bloodust: why do you suggest triangles instead of strips?[22:09:22] *** tehrain has quit IRC[22:10:19] <Bloodust> because the optimization is pointless[22:10:28] <Bloodust> if you draw one cube you wont see a difference[22:10:33] <Black0range> resustance is futile![22:10:47] <Bloodust> if you draw hundreds of thousands then you shouldnt be drawing cubes in the first place but cube faces[22:10:53] <japro> how fitting, given we are talking about cubes...[22:10:59] *** bookmark has joined ##OpenGL[22:11:05] <bookmark> hi[22:11:28] <bookmark> check out what i did: http://youtu.be/hJnGSCeD3UY[22:11:30] <japro> bookmark, hi, you have to decide this dispute... triangles or triangle strips[22:11:34] <japro> choose wisely![22:11:53] <bookmark> i chose triangles long ago[22:12:00] <Bloodust> bookmark haha :D[22:12:04] <chrisf> foobaz: only reduces the size of your index list, and restart indices or degenerate tris erode even most of that benefit.[22:12:10] <Bloodust> those are some effed up textures[22:12:33] <bookmark> bloodust: yeah im just setting up cameras to start modeling[22:12:38] <foobaz> i can accept that strips aren’t faster, but they’re not slower either, and they make your code simpler[22:12:46] <bookmark> i wrote a script that should fix them too[22:12:52] <japro> that is like post card type images projective textured on some pop up book geometry?[22:13:08] <bookmark> yeah[22:13:13] <bookmark> off google images[22:13:14] <Bloodust> foobaz I wouldnt say so, strips are a special case[22:13:16] <japro> foobaz, i find strips more of a brain teaser than just triangles[22:14:22] <bookmark> i already about lost my mind writing a simple wavefront obj loader[22:14:41] <bookmark> just regular triangals[22:14:48] <bookmark> angles[22:14:50] <bookmark> lol[22:14:56] <chrisf> foobaz: strip order is not the best you can do for vertex cache optimization; they actually are slower.[22:15:01] *** stefkos has quit IRC[22:15:08] <japro> my current approach to those is to reformat them into C arrays and put them in a header :o)[22:15:17] <derhass> I get the impression that people here generally lose or blow their minds to often^^[22:15:37] <japro> yeah, all that brain matter is kinda getting out of hand[22:15:40] <foobaz> it is a common side effect of working with opengl[22:15:46] * japro gets the mop[22:16:08] <japro> i wonder where it is, must have put it with the torches and pitchforks[22:16:11] <bookmark> that shanghai scene was accomplished in one night[22:16:21] <bookmark> it should be good in about a month[22:16:55] <bookmark> but i thought it was worth showin[22:18:57] <bookmark> the problem i have with vertex cache optimization is that when it comes to light mapping i would lose my mind all over again[22:19:16] <bookmark> becuase you deal with 2 uv sets[22:19:24] <bookmark> or more[22:19:28] <bookmark> probably more[22:20:06] <Black0range> Btw! :D[22:20:31] <Black0range> How much should you avoid using "if"s in the shaders?[22:20:42] <japro> about this much: <-------->[22:20:51] <japro> that's to scale btw[22:20:54] <Black0range> hmm not <---------> that much?[22:21:22] <japro> well, use them if you really need them[22:21:36] <derhass> avoid non-uniform control flow[22:22:00] *** [H]unt3r has quit IRC[22:22:06] *** paul424 has quit IRC[22:22:08] <bookmark> I use them just to test stuff but wouldn't it be faster to use independent shaders.[22:22:31] <Black0range> if you want to draw a square, and we want said square to have a border,[22:22:42] <japro> not every use of ifs is of the "ubershader" type[22:22:55] <japro> the fxaa shader is essentially one giant nested if[22:23:01] <Black0range> are there any nice magic ways to do this?[22:23:45] <bookmark> i would use a texture[22:23:48] <Black0range> did something allong the ways of if (abs(diff.x) >= width || abs(diff.y) >= height) :/[22:24:11] <japro> those type of things can usually be written nicer[22:24:32] <bookmark> you could write if statements based off the UVs i guess[22:24:34] <japro> step(0.9, max(abs(x,y)))[22:24:36] <japro> or so[22:24:56] <japro> step(0.9, max(abs(x),abs(y))))[22:25:31] <chrisf> Black0range: the cases that really hurt are where the condition is nonuniform, and you have a significant amount of code in both the taken and not-taken cases.[22:26:05] *** Jeanne-Kamikaze has quit IRC[22:26:31] <japro> out of battery... bye[22:26:48] <Black0range> ooh japro thats a nice ne[22:26:54] <Black0range> one*[22:27:13] *** bb010g has joined ##OpenGL[22:31:51] *** japro has quit IRC[22:32:07] *** razieliyo has quit IRC[22:33:51] *** paul424 has joined ##OpenGL[22:34:11] *** urraka has quit IRC[22:34:45] *** Lemml has quit IRC[22:36:41] *** LifeBlood has quit IRC[22:38:30] *** LifeBlood has joined ##OpenGL[22:52:08] *** paul424 has quit IRC[22:56:06] *** telex has quit IRC[22:58:02] *** lenarhoyt has quit IRC[22:58:28] *** snakener1 has joined ##OpenGL[22:58:39] *** telex has joined ##OpenGL[22:59:56] *** qeed has joined ##OpenGL[22:59:57] *** abs25 has quit IRC[23:02:04] *** snakener1 has quit IRC[23:03:26] *** erhan_ has quit IRC[23:05:19] *** DolpheenDream has quit IRC[23:05:20] *** paul424 has joined ##OpenGL[23:06:27] *** ibqn has joined ##OpenGL[23:07:18] *** ij has joined ##OpenGL[23:10:03] *** BearishMushroom_ has quit IRC[23:10:24] <ij> In http://sprunge.us/ZYSN from http://www.falloutsoftware.com/tutorials/gl/gl3.htm there's the glP{ush,op}Matrix called. They have no use there — it would work the same without them, right?[23:12:09] <derhass> throw away that deprecated code[23:12:47] <derhass> © 2014 OpenGL Tutorials. OMG[23:13:25] *** Dudi has joined ##OpenGL[23:14:57] <chrisf> ij: yes, the push/pop are useless.[23:15:34] *** Keniyal has quit IRC[23:16:31] *** Stephans_ has joined ##OpenGL[23:17:03] <chrisf> ij: do yourself a favor and learn from http://open.gl or similar instead though[23:17:19] <ij> Will do.[23:17:54] *** Misu has quit IRC[23:18:32] *** garFF has joined ##OpenGL[23:19:47] *** barosl has left ##OpenGL[23:21:00] *** Biliogadafr has quit IRC[23:22:47] *** boustrophedon has joined ##OpenGL[23:23:26] *** navitsu has quit IRC[23:24:41] *** Life-Blood has joined ##OpenGL[23:25:51] *** LifeBlood has quit IRC[23:30:36] <boustrophedon> should glEnableVertexAttribArray be called once per shader/program binding, or something else? I'm using webgl, but isn't it part of a VAO's state as well?[23:31:42] <chrisf> boustrophedon: it is indeed VAO state[23:33:23] *** RyanPridgeon has joined ##OpenGL[23:33:25] <boustrophedon> the current version of webgl doesn't have VAOs though. so is it essentially per shader binding?[23:33:51] <chrisf> no, in that case it's just context state.[23:34:04] <chrisf> is independent of the shader[23:34:14] <roboman2444> boustrophedon, its per VBO change[23:34:24] <roboman2444> shader is seperate[23:34:38] <boustrophedon> that is true, but if I change the shader presumably I should change the attrib locations as well, right?[23:34:44] <roboman2444> no[23:35:39] <chrisf> well... it would be a good idea to have the attrib locations consistent across shaders as much as possible[23:36:04] <roboman2444> yeah, that[23:37:34] <boustrophedon> i don't think we have layout(location = ...) in webgl though[23:37:59] <slime> you can use glBindAttribLocation before linking the shader program[23:38:20] *** Demon_Fox has joined ##OpenGL[23:40:33] <boustrophedon> ah. I have been using getAttribLocation, but I get what you're saying[23:41:50] *** SugarCube has quit IRC[23:42:18] <boustrophedon> i can make my position data always be location 0, my normal data always location 1, etc regardless of what shader i'm using[23:42:44] <boustrophedon> thank you[23:45:24] *** TunnelCat has joined ##OpenGL[23:48:34] *** Stephans_ has quit IRC[23:49:23] <ij> chrisf, open.gl is pretty good.[23:52:21] <Black0range> hey guys how do i do a bitwise-not in glsl?[23:53:37] *** nido has quit IRC[23:53:58] <foobaz> you could xor with -1[23:54:30] *** ruuns has quit IRC[23:54:46] *** PasNox has joined ##OpenGL[23:55:36] *** nido has joined ##OpenGL[23:55:49] <derhass> Black0range: the same way you do it in C :)[23:55:59] *** aethersis has joined ##OpenGL[23:56:03] <aethersis> hi[23:56:06] <aethersis> I've got a silly question[23:56:14] <derhass> Black0range: there is the unary ~ operator[23:56:34] <aethersis> if I have a fragment shader with one texture sampler and I bind its only input texture to texture unit1, will it work?[23:57:00] <derhass> if you set the sampler uniform to 1, of course[23:57:23] *** snakener1 has joined ##OpenGL[23:57:46] <aethersis> how do I do that?[23:58:10] <derhass> glUniform1i[23:58:15] <aethersis> glUniform1i(displayShader("textureMap"), 0);[23:58:16] <aethersis> like this?[23:58:23] *** linuxuz3r has joined ##OpenGL[23:58:32] <derhass> more or less