May 12, 2008 
< | 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 | >
May 12, 2008 [00:05:37] *** chondrite has joined ##OpenGL[00:11:00] *** magnet_ has quit IRC[00:17:33] *** speedy1 has joined ##OpenGL[00:24:32] *** Bollinger has quit IRC[00:26:57] *** noteventime has quit IRC[00:29:47] *** Yustme has quit IRC[00:34:24] <obmij> when I'm accessing a matrix from GLSL is it always column major?[00:36:25] <Satan_Inside> as per the OpenGL FAQ (http://www.opengl.org/resources/faq/technical/transformations.htm -- specifically section 9.005) that would be correct[00:38:40] <obmij> Satan_inside, are you a Laveyan satanist? :p[00:38:49] <obmij> just wondering :)[00:53:32] *** obmij has quit IRC[00:54:38] *** mastro has joined ##OpenGL[00:54:39] *** TheLorax has joined ##opengl[01:03:59] *** JernejL_ has joined ##OpenGL[01:07:45] *** Gonzo is now known as FrankyFatLaces[01:08:06] *** JernejL has joined ##OpenGL[01:17:01] *** Suprano has quit IRC[01:20:20] *** TheLorax has quit IRC[01:21:00] *** nerdzyboy has joined ##OpenGL[01:22:36] *** Jernej has quit IRC[01:23:09] *** JernejL_ has quit IRC[01:29:40] <nerdzyboy> {-w,0,0, -w,-w,0, 0,-w,0, 0,0,0, 0},[01:29:53] <nerdzyboy> sorry, wrong windows[01:34:46] *** MatthiasM has quit IRC[01:34:53] *** MatthiasM has joined ##opengl[01:37:40] *** odietsch has quit IRC[01:38:39] *** speedy1 has quit IRC[01:39:59] *** amalon has quit IRC[01:43:14] *** User01_ has joined ##OpenGL[01:43:16] *** replor has quit IRC[01:43:36] <User01_> hi[01:43:42] *** User01_ is now known as Ram3[01:45:52] *** replor has joined ##OpenGL[01:45:57] *** davidc__ has quit IRC[01:46:32] *** davidc__ has joined ##opengl[01:46:37] *** inktree has joined ##OpenGL[01:46:52] <inktree> is there an opengl circle or ellipsoid function[01:50:44] *** DobosCake has quit IRC[01:52:33] <neoneurone> inktree, try glu[01:54:00] *** Walt has joined ##opengl[01:55:07] *** juanmabc has quit IRC[01:55:59] *** Walt has quit IRC[01:56:59] *** speedy1 has joined ##OpenGL[02:00:42] *** TheLorax has joined ##opengl[02:00:50] *** DMINATOR has quit IRC[02:01:07] *** speedy1 has quit IRC[02:02:06] *** speedy1 has joined ##OpenGL[02:03:24] *** speedy1 has quit IRC[02:04:58] *** IJs\laptop has quit IRC[02:06:53] *** speedy1 has joined ##OpenGL[02:09:15] <inktree> how do i make lines transparent[02:13:26] *** blight_ has quit IRC[02:14:11] *** Eltran has quit IRC[02:16:46] <speedy1> like all other geometry - enable blending modes and fiddle with alpha[02:19:57] *** blbmp has quit IRC[02:20:02] *** blbmp has joined ##OpenGL[02:22:12] *** Jorachim has joined ##openGL[02:22:27] *** neoneurone has quit IRC[02:30:43] *** davidc__ has quit IRC[02:31:25] *** davidc__ has joined ##opengl[02:33:56] *** Lacerta-2 has quit IRC[02:35:15] *** Xmas| has quit IRC[02:39:34] *** dvoid_ has quit IRC[02:45:54] *** amz has joined ##opengl[02:46:46] *** slide has quit IRC[02:47:13] *** mm765 is now known as mm765^away[02:49:44] *** Ralith has left ##OpenGL[02:51:47] *** hrehf has quit IRC[02:51:54] *** Jorachim has quit IRC[02:54:00] *** TheLorax has quit IRC[02:56:26] *** speedy1 has quit IRC[02:57:03] *** m4ggus has quit IRC[03:02:36] *** mattn_ has joined ##OpenGL[03:04:16] *** enoex has joined ##OpenGL[03:14:53] *** Jorachim has joined ##openGL[03:19:59] *** mattn2|home has quit IRC[03:20:58] *** Amorphous has quit IRC[03:22:04] *** Amorphous has joined ##opengl[03:36:11] *** kaotrix has joined ##OpenGL[03:37:17] *** TheLorax has joined ##opengl[03:43:28] *** hibread has joined ##opengl[03:56:11] *** Jorachim has quit IRC[04:00:18] *** devcpp has joined ##OpenGL[04:00:23] <devcpp> hi all[04:00:55] <devcpp> is there anybody alive?[04:02:14] <rnx> somewhat ... though you shouldn't poll but just ask your question[04:03:59] <devcpp> rnx :) thx : i've downloaded tutorials from codecolony.de so they r written for msvc as i seem, and i can't make them run in dev cpp. why could it be and how can i solve the problem?[04:04:52] <devcpp> btw - do u prefer english or maybe another language i can speak[04:06:20] <devcpp> rnx linker doesn't find any procs and funcs although i've included glut.h and gl and glu[04:06:46] <rnx> including != linking[04:06:51] <devcpp> sorry if it is a stupid question[04:07:37] <rnx> #including the headers isnt enough ... you also have to link to the libraries[04:07:50] <devcpp> how?[04:08:02] <devcpp> have u used dev cpp?[04:08:13] <devcpp> what do u use if i may ask[04:09:05] <rnx> i assume there is some project setting for this but i have never used dev cpp[04:09:22] <rnx> it's manual certainly covers that[04:10:05] <rnx> personally i just write my own makefiles and use the commandline[04:10:27] <devcpp> hmm ok i'll try to seek in settings. may be u could be so nice and recommend me some ide especially convenient for opengl and c++ ?[04:10:34] *** X-Scale has left ##OpenGL[04:11:01] <devcpp> do u use gcc/g++?[04:16:46] *** mattn__ has joined ##OpenGL[04:17:17] *** sohail has quit IRC[04:17:38] *** mattn_ has quit IRC[04:22:18] *** ata2 has joined ##OpenGL[04:22:24] <ata2> hello everybody[04:22:33] <devcpp> ata2 hi[04:22:58] <ata2> if I am searching somebody to talk to with both deep knowledge in OpenGL and DirectX, is this channel the right place?[04:24:50] *** devcpp has quit IRC[04:29:05] *** sohail has joined ##OpenGL[04:29:50] <ata2> anyawya[04:29:59] <ata2> OpenGL2 is not thread safe, right?[04:30:15] <Satan_Inside> incorrect[04:30:29] <ata2> what happens if I have 2 different threads drawing on the same graphical context/buffer[04:30:38] *** Ragnarok has quit IRC[04:30:48] <hibread> kaboom[04:30:57] <ata2> then it's not thread safe[04:31:00] <Satan_Inside> you cannot submit OpenGL rendering comments from N threads into a single context[04:31:16] <Satan_Inside> but this doesn't imply the lack of "thread safety"[04:31:17] <ata2> what do you mean by comments[04:31:40] <ata2> can't I draw two triangles from 2 different threads at the same time?[04:32:11] <Satan_Inside> because being that the GL API is effectively a state machine being able to do this wouldn't be sensible[04:32:12] *** Spark has joined ##opengl[04:32:13] *** rnx has quit IRC[04:32:47] <ata2> so there is no mechanism that queues drawing request.. mhh, ok[04:32:48] <Spark> hi, if i'm using a fragment shader but no vertex shader, do i still have to implement fog?[04:32:51] <Satan_Inside> if you want to submit rendering comments from multiple threads into a single context you need to serialize those commands on the client side[04:32:58] <Spark> i'm not sure where fog gets calculated anyway[04:33:13] <Satan_Inside> yes you need to implement fog[04:33:15] <ata2> Right Satan, that is what I was thinking about[04:33:54] <Spark> out of interest, is there a complete implementation of the fixed function pipeline in glsl? and how does the performance compare[04:34:22] <Spark> it would seem to make more sense to start from a glsl implementation of the fixed function pipeline, and then tweak the code from there to add more effects / change things[04:34:42] <hibread> Spark: just code in the shaders what you need[04:34:51] <Satan_Inside> honestly though you shouldn't be submitting rendering commands from multiple threads into a single context[04:35:54] <Satan_Inside> a much better approach is to have a single context per thread for N threads[04:35:54] <Spark> ok so do i implement fog in a new vertex shaader or the fragment shader i already have[04:36:23] <hibread> Spark: for a particular "pass" impliment everything you need[04:36:31] <Spark> hibread: but where[04:36:36] <hibread> Spark: you can do it how ever you like :)[04:36:38] <hibread> there are no rules[04:36:45] <Spark> it needs to look the same as the fixed function pipeline[04:37:05] <Spark> i'm fading things out use scene blending when they reach their max view distance[04:37:29] <Spark> so there is a phase where i set scene blending, and i start using a fragment shader which lets me set the alpha per pixel[04:37:56] <Spark> but now if i turn fog on, only the entities that aren't currently fading out get fogged[04:38:18] <Spark> so i guess i should add fog to the fading shader i already have?[04:38:42] <Spark> one thing in particular, i see the gl_Fog glsl variable[04:38:59] <ata2> Satan_Inside: the fast is that we have a system which is used for the parallelization of game engins. Everything is organized in threads by "chores" (collections of updates to be computed)[04:39:00] <Spark> which is good, but i can't see where i find out whether the fog is LINEAR, EXP, EXP2, or NONE[04:39:46] <ata2> Satan_Inside: unless we are able to put graphical drawing in chores as well, we would need to have a locking on the other elements[04:39:56] <ata2> which from our experiments terribly slows down the system[04:40:00] <hibread> Spark: http://www.opengl.org/sdk/libs/OpenSceneGraph/glsl_quickref.pdf[04:40:08] <hibread> gl_FogParameters[04:40:12] <hibread> has scale[04:40:19] <hibread> i guess that would bethe linear/exp etc?[04:40:31] <Spark> hmm you mean if it's 0, use the other pair[04:40:41] <Spark> or do i just use an equation that does both[04:41:22] <Spark> i.e. fogginess = max(exp_method(),exp2_method(),linear_method())[04:42:10] <Satan_Inside> ata2 like I said you only need to serialize on the client side for situations involving a N:1 ratio of command submitting threads to contexts[04:42:33] <ata2> yah, I see that point[04:43:17] <Satan_Inside> and considering that you can trivially share resources between threads via context sharing supporting a 1:1 thread:context ratio for N threaded systems is also reasonably trivial[04:43:30] <hibread> Spark: i have no idea really.. haven't played with fog for a while. I probably wouldn't use those variables if i did.. i'd just hack something up that resembles what i want the effect to look like[04:43:33] <Satan_Inside> which is what you should be doing here[04:43:44] <Spark> hibread: heh[04:43:45] <TheLorax> I hear that alot in openGL, what is "client side" and what is "server side"?[04:43:58] *** amz has quit IRC[04:44:17] <Satan_Inside> think of the "client" as the application(s) you guys are writing[04:44:30] <Satan_Inside> think of the "server" as the underlying driver / GL runtime[04:44:37] <TheLorax> oh[04:44:38] *** davidc__ has quit IRC[04:44:41] <Satan_Inside> which is of course specific to a given OS, vendor, ...[04:45:47] <Spark> my experience of threading is that as soon as you start hitting gl from more than one thread, performance goes through the floor[04:45:59] <Spark> at the moment i use another thread for loading / decompressing resources[04:46:02] <Satan_Inside> that is highly platform dependent[04:46:20] <Satan_Inside> certain platforms do far better than others in the realm of shared contexts[04:46:22] <Spark> but that thread hands off to the rendering thread as soon as the data is fully available and gl calls need to be made[04:46:34] <Spark> i've used linux nv gl and win32 nv gl[04:46:36] <Satan_Inside> Mac OS X being at the "fast" end of that spectrum[04:46:43] <Spark> hmm interesting[04:46:58] <Spark> i was looking at about a 30% drop in FPS, when rendering lots of batches[04:47:05] <Satan_Inside> since nearly everything in the Apple realm uses massive amounts of context sharing[04:47:31] <Satan_Inside> applications dealing with tens of contexts aren't uncommon over here[04:47:39] <Satan_Inside> and *many* threads[04:47:41] <Spark> they must use lock free algorithms then[04:47:56] <Spark> ah you see it wasn't the contentiont hat was teh problem, it was just the linear overhead of having locks at all[04:48:01] <Satan_Inside> actually it goes far being that in the realm of the server[04:48:06] <Satan_Inside> the GL server[04:48:11] <Satan_Inside> err beyind[04:48:13] <Satan_Inside> err beyond[04:48:33] <Spark> i profiled gl on linux, it was taking several locks per batch[04:48:48] <Spark> when there was more than one context[04:48:49] <Satan_Inside> in any case its always a good idea to profile under your OSes of choice in order to see what the preferred paths happen to be[04:54:32] *** replor has quit IRC[04:54:43] *** TheLorax has quit IRC[04:58:19] *** inktree has quit IRC[05:21:04] *** hibread has quit IRC[05:31:45] *** chondrite has quit IRC[05:33:46] <geocalc> papy-gl change of name but write the same ;)[05:33:47] *** blbmp has quit IRC[05:35:49] *** blbmp has joined ##OpenGL[05:42:23] *** peda_ has quit IRC[05:42:28] *** peda_ has joined ##OpenGL[05:45:42] <Ram3> how create a more visible depth texture ?[05:53:02] *** ata2 has quit IRC[06:01:28] *** deus has joined ##OpenGL[06:02:17] <deus> Do anybody know of an ircchannel for windows development issues and visual studio?[06:10:34] *** BahamutZERO has quit IRC[06:12:23] *** blbmp has quit IRC[06:35:42] *** Maerz has quit IRC[06:36:11] *** Maerz has joined ##OpenGL[06:40:21] *** kaotrix has quit IRC[06:45:32] *** nathan_ has joined ##OpenGL[06:49:33] *** Ragnarok has joined ##OpenGL[06:49:50] <Ragnarok> how can I find out what version of OpenGL my onboars graphics support?[06:50:23] <HuntsMan> glGetString(GL_VERSION);[06:50:50] <Ragnarok> thanks[06:50:55] *** xonev_ has joined ##OpenGL[06:52:06] *** xonev has quit IRC[06:52:07] <Ragnarok> gave me 0x0[06:52:57] <HuntsMan> you need to have a OpenGL context made current[06:53:41] <Ragnarok> okay i got 0xb6ef5973[06:53:49] *** obmij has joined ##OpenGL[06:53:52] <nathan_> when should a futures contract be bought and when should it be sold?[06:54:07] <obmij> for reading vertex/fragment shader files, should I do a binary read or an ASCII read?[06:54:12] *** nathan_ has quit IRC[06:54:21] *** nathan_ has joined ##OpenGL[06:54:36] <HuntsMan> Ragnarok: it's a string[06:54:48] <obmij> HuntsMan, what do you suggest? :)[06:55:20] <Ragnarok> ah i get 2.1.2 NVIDIA 169.12 :D[06:55:21] <HuntsMan> obmij: text[06:55:24] <obmij> kay[06:55:55] <Ragnarok> wow i didn't know nVidia GeForce 6150 LE was OGL 2.1[06:56:19] * obmij has a eVGA 8800 GTX :P[06:56:30] <obmij> I wish I had saved up for a 9800 Gx2[06:56:33] <Ragnarok> imma be getting a 8800GT soon[06:56:53] <obmij> meh, I'll save up for one ;P[06:56:56] <Ragnarok> obmij, i would but my mobo doesn't support POCI-e 2.0 only x16[06:57:08] <Ragnarok> PCI-e*[06:57:10] <obmij> eh :P[06:57:17] <obmij> I had to update my PS as well[06:57:22] * obmij got a 700 watt ps[06:57:32] <obmij> it needs 2 PCI-e power cables[06:57:39] <Ragnarok> well imma be buying me a Q9450 with a EVGa mobo soon :D[06:57:44] <obmij> You know what[06:57:52] <obmij> buy a 9800 Gx2 and feel like a king[06:57:55] <Ragnarok> :D[06:57:56] <obmij> ;D[06:58:01] <Ragnarok> overclocked version too ;)[06:58:06] * obmij nods[06:58:18] <Ragnarok> and a ultra when they have it available[06:58:26] <obmij> Gx2 is the highest end[06:58:27] <Ragnarok> i won't worry about SLI it[06:58:30] <obmij> the so called 'ultra'[06:58:31] <Ragnarok> ah[06:58:35] <obmij> :)[06:58:39] <Ragnarok> :D[06:58:40] <obmij> GX is the normal one[06:58:45] <Ragnarok> i need to find me a real job[06:58:47] <Ragnarok> :([06:58:52] <obmij> eVGA sent me the email as soon as they released[06:58:57] <Ragnarok> awesome[06:59:01] <obmij> they thougt since I put the money down for a 8800 GTX[06:59:10] <obmij> I might as well do it for 9800 gx2 :D[06:59:31] <Ragnarok> lucky cunt[06:59:36] <obmij> either way, I got it for serious business so ... I'm happy so far anyway ;/[06:59:45] <obmij> hey man, I work hard for my money[06:59:51] <obmij> and I piss it off easily[06:59:53] <obmij> :/[07:00:03] <Ragnarok> hahah i need to move out[07:00:04] <Ragnarok> >.>[07:00:11] <Ragnarok> <,<[07:00:16] <obmij> Q.Q[07:00:22] <Ragnarok> X_X[07:00:25] *** mm^away has joined ##opengl[07:00:36] <Ragnarok> i have to save up $12000[07:00:39] <Ragnarok> to move out[07:01:01] <Ragnarok> thats a year of work without spending a single dime[07:02:04] * obmij goes back to the IDE[07:02:09] <Ragnarok> hahah[07:02:21] <Ragnarok> i wish gentoo would update qt4 to the lastest version[07:02:30] <obmij> God damn, you have Gentoo[07:02:35] <obmij> I had that crap on my laptop for a long time[07:02:42] <obmij> before I got to my senses and installed Debian[07:02:50] <obmij> I mean no offense[07:02:53] <Ragnarok> ewww debian[07:03:05] <Ragnarok> Linux Narf 2.6.25-gentoo-r3 #1 SMP Sun May 11 13:09:06 CDT 2008 i686 AMD Athlon(tm) 64 X2 Dual Core Processor 3800+ AuthenticAMD GNU/Linux[07:03:06] <obmij> but honestly... no way. Portage and hard/soft masks... no way. Not me, not again.[07:04:04] <Ragnarok> debian with finding over the internet for updates, grab a package, it won't install due to a dependcies, search again, find it, wow it needs another one[07:04:06] <Ragnarok> etc[07:04:09] <obmij> It was fun compiling every little thing from scratch (not really) but it kind of made you think it was fun.[07:04:23] <Ragnarok> it doesn't bother me[07:04:48] <obmij> yeah well everything has its trade offs :P[07:05:05] <obmij> Either way kisses to the fellow tux0r[07:05:06] <obmij> ;)[07:05:17] <obmij> I'm on Windows btw as we speak[07:05:47] <Ragnarok> haha i only use Vista for gaming[07:05:54] <obmij> No ... HELL no[07:05:56] <obmij> XP SP2[07:05:58] <Ragnarok> since linux fails at that[07:06:01] <obmij> how do you live with that crap?[07:06:06] <Ragnarok> XP SP3 is already released[07:06:10] <obmij> Oh they did?[07:06:11] <Ragnarok> i love vista[07:06:14] <Ragnarok> yes[07:06:15] <obmij> I should check it out[07:06:23] <obmij> Oh come on, Vista kills[07:06:26] <obmij> Vista eats babies[07:06:29] <obmij> x_x[07:06:32] <Ragnarok> i've never had any problems with it[07:06:36] <obmij> Really? odd.[07:06:42] <obmij> Glad you're happy with it :)[07:06:45] <Ragnarok> on 3 different pc's no problems[07:06:50] <obmij> Nice[07:06:54] <obmij> You got a SP for it?[07:07:01] <Ragnarok> Plus my motherboard has no drivers for Windows XP :P[07:07:07] <obmij> aww that sucks[07:07:23] <geocalc> fista[07:07:26] <Ragnarok> so even if I hated vista, I have no choice but to use it[07:07:36] <obmij> If it floats your boat, that's awesome[07:07:42] <Ragnarok> hehe[07:07:45] <obmij> but I've had so many different legit problems with it[07:07:53] <obmij> that I could easily find others complaining about[07:07:56] <obmij> too many to list honestly[07:07:57] <geocalc> refuse the cluf[07:09:23] <Ragnarok> wtf why isn't my triangle drawing :I([07:09:26] <Ragnarok> :(*[07:10:36] <geocalc> lol[07:10:43] <Ragnarok> i have glVertex2f(0.0f, 0.0f); glVertex2f(25.0f, 25.0f); glVertex2f(50.0f, 0.0f); and it won't draw :([07:10:51] <obmij> SwapBuffers?[07:10:56] <obmij> glBegin glEnd?[07:11:00] <Ragnarok> both[07:11:04] <obmij> Did you set up view?[07:11:05] <Ragnarok> yeah i have al that[07:11:06] <obmij> and projection?[07:11:17] <obmij> and look and eye[07:11:17] <obmij> ?[07:11:20] <obmij> and up?[07:11:27] <Ragnarok> http://pastie.caboo.se/195384[07:11:52] <Ragnarok> i don't have glLookUp but i never needed it when I did the nehe tutorials[07:11:57] <obmij> umm[07:12:04] <obmij> all right, it's Nehe[07:12:09] <obmij> so can't help you there[07:12:16] <Ragnarok> no no[07:12:23] <geocalc> is it on the good screen[07:12:25] <Ragnarok> I removed its stuff and trying to do it myself from meory[07:12:28] <Ragnarok> memory*[07:12:34] <obmij> Honestly[07:12:45] <obmij> the only headers file I have are gl.h and glu.h[07:12:49] *** adampetrone has quit IRC[07:12:58] <obmij> and most of my code is plaing C style[07:13:10] <obmij> even though technically it ends up being C++[07:13:41] <obmij> Wish I could help man.[07:14:06] <Ragnarok> i hate C but love C++[07:14:22] <obmij> Well, you can't love C++ without loving C[07:14:27] <obmij> that's a paradox right there[07:14:37] <Ragnarok> u don't need to know C to know C+[07:14:42] <Ragnarok> C++*[07:14:42] <geocalc> non-sense[07:14:58] <obmij> You learn a lot of C before learning C++[07:15:18] <obmij> If you don't know about functions and/or structures how would you know about member functions of a class and the concept of class itself?[07:15:23] <Ragnarok> the only C you need to know really is when you mess with C strings but thats niot recommanded to use[07:15:35] <obmij> Ragnarok, tuch tuch tuch[07:15:39] <Ragnarok> obmij, I have C# background knowledge :P[07:15:40] <obmij> you have a long way to go man[07:15:48] <obmij> loooong way[07:16:01] <Ragnarok> the only thing I have problems with is understand unions[07:16:03] * obmij goes back to his IDE[07:17:14] *** mm765^away has quit IRC[07:17:19] *** obmij has quit IRC[07:30:45] <rutski> is it normal for glGetError to not report errors until after glFinish() is called?[07:38:29] *** Lemml has joined ##OpenGL[07:55:34] <NeoThermic> No.[07:55:49] <NeoThermic> A call to glGetError should immediatly report the error status of the GL state[07:56:30] <NeoThermic> be it GL_NO_ERROR if the state is good, or one of 6 other error states that the GL engine can be in.[08:00:33] *** Walt has joined ##opengl[08:02:02] *** nathan_ has quit IRC[08:03:18] <Ragnarok> wtf, if I use gluLookAt(), my points are reversed :S[08:03:54] <Ragnarok> ex. my bottom left = my bottom right[08:04:03] <Ragnarok> and vice-versa[08:04:16] <Ragnarok> but if I use glTranslatef() instead its correct[08:23:19] *** neoneye has joined ##OpenGL[08:25:03] <madsy_> Ragnarok: You have something called an "up vector"[08:25:05] <madsy_> flip ip[08:25:09] <madsy_> it*[08:31:52] *** karabash has joined ##OpenGL[08:48:17] *** deus has quit IRC[09:13:42] *** inktree has joined ##OpenGL[09:13:50] <inktree> anyone out there?[09:14:06] <madsy_> Maybe :-)[09:14:31] <Spark> anyone in there?[09:14:40] *** Bollinger has joined ##opengl[09:14:51] <inktree> whenever i try to compile i get a crap load of errors:[09:14:52] <inktree> c:\program files\microsoft visual studio 8\vc\platformsdk\include\gl\gl.h(44) : error C2146: syntax error : missing ';' before identifier 'TexModes'[09:14:54] <inktree> 1>c:\program files\microsoft visual studio 8\vc\platformsdk\include\gl\gl.h(44) : error C4430: missing type specifier - int assumed. Note: C++ does not support default-int[09:14:55] <inktree> etc...[09:15:23] <inktree> i was able to compile many times before 10 seconds ago... so what happened?[09:15:44] <madsy_> VS8.. hah :-)[09:15:53] <madsy_> Run a clean[09:16:00] <madsy_> I bet'll work afterwords.[09:16:17] <inktree> Clean Solution?[09:16:27] <inktree> i just tried it still doesn't work :([09:16:38] <madsy_> Clean as in clean the object files.[09:16:49] <madsy_> Wait. Did the program *run* before as well?[09:17:00] <inktree> yes[09:17:03] <madsy_> Or did the compilation/linking just not give any errors?[09:17:11] <inktree> it ran perfectly a minute ago[09:17:29] <madsy_> Did you remove any of the includes?[09:17:39] <madsy_> gl.h depends on windows.h[09:18:10] <inktree> nope didn't remove any[09:18:26] <Spark> including things in the wrong order?[09:18:30] *** LordHavoc has joined ##OpenGL[09:18:54] <madsy_> Yes, or perhaps you've done changes in a header which is include before gl.h[09:18:55] <inktree> nope i didn't change include order[09:19:00] *** Jupp3 has quit IRC[09:19:09] <madsy_> Always include the standard stuff first, and you own includes last.[09:19:32] <NeoThermic> tried replacing gl.h with a fresh copy?[09:19:42] <madsy_> Okay, show us the code, I'll take a look.[09:19:47] <Spark> changed compiler params?[09:19:55] <Spark> turned on more warnings or optimisations[09:20:00] <Spark> added more #defines[09:20:03] <Spark> turned on debug mode?[09:20:23] <madsy_> Spark: Debug mode my arse :-)[09:20:28] <madsy_> This is a compile-time error[09:20:39] <Spark> yeah but it changes the way it compiles too[09:20:40] <Spark> i think[09:20:49] <Spark> different defines[09:23:41] <inktree> http://www.2shared.com/fadmin/3275634/c81cba02/proj.rar[09:24:59] *** [AD]Turbo has joined ##OpenGL[09:25:13] <inktree> i think i may have accidentally overwritten the gl.h[09:25:40] <Spark> tell it to the judge[09:25:54] <inktree> what is gl.h[09:25:57] <inktree> where can i dl a new one[09:26:22] <Spark> just debug the error like any other error[09:26:39] <Spark> what is line 44 of gl.h[09:26:44] <Spark> and the lines above and below it[09:27:09] <[AD]Turbo> yo[09:27:14] <inktree> i overwrote the entire gl.h file[09:27:15] <inktree> by accident[09:27:26] <inktree> with my own code... where can i dl a new gl.h?[09:34:33] *** sohail has quit IRC[09:35:01] <inktree> can someone here please[09:35:08] <inktree> give me the contents of gl.h[09:35:30] <inktree> c:\program files\microsoft visual studio 8\vc\platformsdk\include\gl\gl.h[09:35:31] <inktree> if you have VC it's located in above... it's a mesa lib file that i accidentally overwrote[09:41:12] *** groton has joined ##OpenGL[09:49:44] *** dvoid_ has joined ##OpenGL[10:01:23] *** Yustme has joined ##opengl[10:04:24] *** rutski has quit IRC[10:11:05] <MatthiasM> inktree: install VC again[10:18:47] *** Roderic has joined ##OpenGL[10:18:48] *** Roderic is now known as Ingenu[10:19:17] *** hibread has joined ##opengl[10:24:13] *** neunon has joined ##opengl[10:37:24] *** Suprano has joined ##OpenGL[10:56:09] *** neoneurone has joined ##OpenGL[11:00:14] *** BahamutZERO has joined ##OpenGL[11:00:36] *** hubbe3 has joined ##OpenGL[11:12:36] *** charlie_zzz is now known as charlie5[11:22:54] *** Ademan has joined ##OpenGL[11:26:46] *** LtJax has joined ##opengl[11:26:52] *** LordMetroid has joined ##OpenGL[11:51:11] *** Castigador has joined ##Opengl[12:10:19] *** enoex has quit IRC[12:10:46] *** enoex has joined ##OpenGL[12:31:32] *** Killari has quit IRC[12:34:40] *** karabash has quit IRC[12:35:01] *** sanjin has joined ##OpenGL[12:42:39] *** Jorachim has joined ##openGL[12:52:26] *** SaphireS has joined ##OpenGL[12:57:51] *** joakim_ has joined ##OpenGL[13:05:09] *** prophile has joined ##opengl[13:12:30] *** sanjin has left ##OpenGL[13:16:19] *** joakim__ has quit IRC[13:16:38] *** Ademan has quit IRC[13:17:42] *** Ademan has joined ##OpenGL[13:22:36] *** blight_ has joined ##opengl[13:41:55] *** dv_ has joined ##OpenGL[13:49:31] *** Suprano has quit IRC[13:51:14] *** LtJax has quit IRC[13:54:57] *** neunon has quit IRC[13:55:02] *** LtJax has joined ##OpenGL[13:57:03] *** sanjin has joined ##OpenGL[14:05:47] *** mastro has quit IRC[14:06:29] *** mastro has joined ##OpenGL[14:21:54] *** SaphireS has left ##OpenGL[14:25:52] *** Jorachim has quit IRC[14:26:38] *** Eltran has joined ##OpenGL[14:27:44] *** nathan_ has joined ##OpenGL[14:34:46] *** [AD]Turbo has quit IRC[14:37:19] *** [AD]Turbo has joined ##OpenGL[14:39:09] *** gaffa has joined ##opengl[14:49:25] *** replor has joined ##OpenGL[14:55:19] *** Jorachim has joined ##OpenGL[14:58:57] *** odietsch has joined ##OpenGL[15:11:59] *** rnx has joined ##opengl[15:16:07] *** PainBank has joined ##OpenGL[15:20:49] *** LordMetroid has quit IRC[15:21:46] *** HuntsMan has quit IRC[15:24:20] *** Nai has joined ##OpenGL[15:24:32] *** elite01 has joined ##opengl[15:27:34] <Nai> hi, i am following an opengl programming guide, and have opengl window opening in unix with my perl/tk script, however it seems as if i'm reinventing the wheel, should I be using some kind of open source game engine or a some kind of low level interface to the opengl[15:28:59] <hibread> Nai: totally depends on what you're aiming for... or do you mean just a basic API to help you set up a window?[15:30:05] <Nai> well i would love to make a tiny interactive game, and the way im doing it seems the long way[15:30:55] <hibread> so the setting up of the window with user inputs etc is no problem? you just want to further develop this game?[15:31:13] <Nai> ideally I would like to export shapes or environments from Maya into the opengl somehow[15:31:51] <Nai> well at the moment i have the hexagonal opengl test window with the coords, and im reading about how to make the shapes.. and im just considing how it's going to work with the animation and binding to keys[15:32:03] <hibread> you could either write that yourself... or maybe check out Ogre or something? That may have importing type features for some 3D meshes[15:32:57] <Nai> hmm[15:33:29] <Nai> another thing im considering is events happening in time, how opengl switches the focus, and how it's going to interact with the back end coding[15:33:49] *** Jorachim has left ##OpenGL[15:34:47] <hibread> i dont follow sorry[15:34:56] <Nai> neither[15:35:01] <Nai> reading about ogre now[15:36:48] *** filmor has joined ##OpenGL[15:37:06] <Spark> ogre is good for getting small things up and running[15:37:32] <Spark> if you want to do something as large and complex as an actual modern computer game, you can end up writing a lot of stuff yourself though[15:37:35] <Spark> but that's to be expected[15:37:37] <Spark> ogre is a good effort[15:37:53] <Nai> by switching the focus i mean.. for example when i did Perl/Tk gui programming i had to use "select" to keep the MainLoop running while interpreting network data input at the same time, there was also the fork option, which forked off a seperate instance of the script[15:38:16] <Nai> Spark, can i use it with Perl?[15:40:17] <Nai> the featured projects are hilarious[15:41:16] <quicksilver> I think if you plan to use any toolkit which uses opengl underneath[15:41:25] <quicksilver> you will probably benefit from understanding opengl's underlying model first[15:41:30] <quicksilver> at least to some extent.[15:41:41] <quicksilver> until I did some low-level opengl there was quite a bit I wouldn't have understood.[15:42:14] <hibread> yep good call quicksilver[15:42:41] <Spark> Nai: no perl[15:42:45] <Spark> you can use c++[15:42:47] <Spark> and maybe python[15:43:41] <quicksilver> (e.g. opengl doesn't support shadows, reflections or bump maps, "directly", but it does give you the tools to make them)[15:44:15] <quicksilver> initially I was quite surprised at how 'little' opengl can do directly.[15:44:17] <Spark> ogre supports shadows out of the box, the rest you have to do yourself[15:44:42] <Spark> but there are useful utility methods for working out where to put your camera for a reflection rtt[15:44:48] <Nai> I would prefer something that works with perl, otherwise i will have to learn another lang[15:45:01] <quicksilver> but then I saw that it gives you the bits and pieces to do lots of things you can invent yourself[15:45:13] <Spark> and there are some shaders in the source tree that do bump mapping[15:46:02] <hibread> Nai: learn a real language :)[15:46:26] <Spark> yeah you're not going to get very far with perl[15:46:31] <Spark> it will quickly become unmanagable[15:46:46] <quicksilver> don't think there's any reason perl need be more unmanagable than C++.[15:46:53] <quicksilver> good perl is just as good as good C++[15:47:01] <quicksilver> and bad C++ is about as bad as the pits of hell.[15:47:13] <Nai> :)[15:47:35] <hibread> quicksilver: and that reminds me, i should really read "Effective C++" again[15:49:25] <hibread> isn't there going to be a new release of c++? c++0x or something?[15:49:40] <Ingenu> yeah[15:49:43] <Ingenu> c++ worst[15:49:48] <Ingenu> like c++ isn't bad enough[15:50:22] <Spark> written a 300kloc perl program? :)[15:50:33] <hibread> Ingenu: save it :) C++ has stood up against the test of time[15:50:42] <Ingenu> C++ sucks anyway[15:51:11] <Ingenu> the simple fact that people don't acknowledge it and don't even think about finding the right tool for their needs tells a lot about the state of the software industry[15:51:33] <hibread> acknowledge what?[15:51:38] <hibread> thats its "Shit"?[15:51:41] <MatthiasM> there are still a lot of places were only C is used[15:52:07] <cupe> i use opengl with lisp :)[15:52:22] <Ingenu> I'm considering moving to D[15:52:31] <quicksilver> and I use opengl with haskell.[15:52:41] <quicksilver> so the non-C-family users outnumber the C-family users.[15:52:42] <quicksilver> we win![15:52:43] <quicksilver> ;)[15:52:50] <Ingenu> lol[15:52:53] <Ingenu> :)[15:52:54] * MatthiasM uses Java[15:53:06][15:53:37] <cupe> haskell would be my second choice[15:54:42] <Ingenu> I really like D[15:54:47] <Ingenu> I've not used it yet though[15:54:54] <quicksilver> I think D has some very interesting ideas, but I've never used it.[15:55:01] <MatthiasM> not enough tool and library support yet[15:55:08] <Ingenu> but I've been reading their website for a while now, and it's really interesting[15:55:09] <MatthiasM> (for D)[15:55:25] <Ingenu> that's what I feared :([15:56:09] <Spark> I'm thinking of using Java for the next thing i do[15:56:14] <Spark> it performs the best i think[15:56:15] <quicksilver> the proprietary compiler backend (for D) puts me off somewhat.[15:56:24] <MatthiasM> a few people in hear use D for OpenGL and always complain about not beeing able to use extension X because of ... - and D even claims to be binary compatible with C[15:56:25] <quicksilver> I don't think that's a good attitude to encourage people to adopt your language.[15:56:32] <MatthiasM> you can use GCC[15:56:54] <cupe> Spark: performance is secondary, fun is primary[15:56:57] <MatthiasM> Spark: take a look at jME and LWJGL[15:57:02] <Ingenu> mmh it's supposedly C compatible[15:57:08] <Spark> fighting for every lost drop of performance is not fun[15:57:09] <Ingenu> so it shouldn't be a problem using OpenGL with it[15:57:27] <cupe> Spark: yep. don't do it :)[15:57:43] <Spark> i want 75fps, not 60fps[15:57:44] <quicksilver> java performs the best? that's an unusual claim.[15:57:58] <Spark> it has some hardcore optimisations[15:57:59] <Ingenu> and not realistic[15:58:03] <quicksilver> I'd be very surprised if java opengl code can ever outperform C opengl code[15:58:07] <Spark> and it's much better at garbage collecting than say boost::shared_ptr[15:58:08] <quicksilver> I can't imagine how that could happen.[15:58:17] <Ingenu> that's for sure[15:58:21] <MatthiasM> the only points that I had to optimize yet were the way I render the world and the algorithms - only the things that you need to care ablout in every language[15:58:22] <Spark> scene management[15:58:26] <Spark> that's the key issue[15:58:33] <Ingenu> when you start relying on shared pointers and the like you have missed the whole point of C/C++[15:58:44] <Ingenu> so better move to some other language[15:58:58] <Spark> is the whole point of c/c++ that "c++ sucks anyway" ?[15:59:13] <hibread> Im always curious to know why these conversions play out... they never really seem to get anywhere[15:59:22] <Ingenu> C++ adds more pitfall to C[15:59:25] <Ingenu> so it's aweful[15:59:26] <MatthiasM> quicksilver: most work is done by the driver and the GPU anyway[15:59:35] <quicksilver> MatthiasM: maybe. that depends.[15:59:44] <Spark> Ingenu: a fairly absurd claim[15:59:50] <quicksilver> MatthiasM: it's not uncommon to see CPU bound instead of GPU bound code[15:59:57] <MatthiasM> and things like scene graph management - which uses inheritance and virtual function - is an area where Java performs best[16:00:07] <Spark> the number of batches you can render on even a modern machine is pretty pitiful[16:00:17] <quicksilver> MatthiasM: e.g. a poor animation or particle engine.[16:00:28] *** uchitoru has joined ##OpenGL[16:00:29] <Spark> probably 50000 batches a second or something[16:00:45] <Spark> that's what... 100000 cycles per batch?[16:01:00] <Ingenu> it's almost all about algorithms and data management[16:01:03] <quicksilver> inheritance and virtual functions will be no fast in java than they are in C++.[16:01:14] <Ingenu> code execution time isn't that critical (of course there are exceptions)[16:01:18] <MatthiasM> quicksilver: java can inline virtual functions[16:01:21] <Spark> quicksilver: hotspot binds calls statically[16:01:26] <Spark> and that leads to inlining[16:01:33] <Spark> java has the most advanced optimisations of any language i think[16:01:49] <MatthiasM> Spark: no GCC 4 has better vectorisation[16:01:51] <quicksilver> true but I will be impressed if you can find real world cases where that makes a measurable difference.[16:01:55] <Ingenu> I would have thought C had[16:01:58] <Ingenu> given how long it's been around[16:02:19] <quicksilver> tehre are a few cases where JIT allows you to do really clever stuff which a static compiler can't do.[16:02:25] <quicksilver> however, theses cases are rare[16:02:29] <quicksilver> and hotspot doesn't get them all right.[16:02:32] <MatthiasM> C/C++ only has static and profile based optimisation - which can adapt to the runtime use for virtual function calls[16:02:46] <MatthiasM> *which can't[16:02:48] <quicksilver> and there are plenty of other cases where C/C++ have a substantial speed advantage[16:02:54] <quicksilver> all that pointer following and unboxing.[16:02:59] <Spark> profile-based optimisation is all well and good, but it's a massive hassle compared to JIT which gives you the same for free[16:03:11] <quicksilver> except, it doesn't.[16:03:12] <Spark> and JIT adapts to the particular hardware, whereas te profile-based approach will optimise for whatever machine you profiled it on[16:03:22] <quicksilver> you're talking about things that JIT might do.[16:03:29] <MatthiasM> quicksilver: pointers are also one of the major reason why certain optimisations can't be done[16:03:30] <quicksilver> in practice it does those things to a very limited extent.[16:03:39] <quicksilver> there is lots of interesting work on what a JIT *could* do in theory.[16:03:52] <quicksilver> In practice the JITs out there don't do very many of them, and don't do it that well.[16:03:58] <MatthiasM> quicksilver: I suggest that you first check what the Java Hotspot already does[16:04:10] <MatthiasM> we don't talk about the .net VM[16:04:11] <Spark> i benchmarked some uncontended locks in boost, pthreads, and java hotspot[16:04:12] *** filmor has quit IRC[16:04:18] <quicksilver> I have a pretty good idea what hotspot does.[16:04:19] <Spark> java won by about 150%[16:04:25] <Spark> followed by pthreads[16:04:30] <Spark> and then finally boost was very slow[16:04:40] <MatthiasM> Spark: with the latest JDK 7 or older one ? because they added a lot additional optimisation[16:04:53] <Spark> only 1.6[16:05:02] *** TheLorax has joined ##opengl[16:05:06] <Spark> this is down to doug lea though i think[16:05:10] <Spark> he just wrote very fast locks[16:05:19] <MatthiasM> with 1.7 they use EA to remove locks completely if an object can never be accessed by multiple threads[16:05:29] <Spark> but the locks are a heavy OO framework[16:05:37] <Spark> the optimisations would have to be good for all that to be invisible[16:05:48] <Spark> that's been around for a while[16:06:01] <MatthiasM> Spark: the java.util.concurrent locks are for heavy multitasking application on large multi CPU system[16:06:16] <MatthiasM> for your standard desktop application / games it's better to use synchronized[16:06:20] <Spark> not really[16:06:37] <Spark> i tested both i just can't remember the figures for synchronized[16:06:41] <MatthiasM> or use the lockless data structures if you expect contended access[16:06:50] <Ragnarok> blah this site sucks[16:07:29] <Spark> right, meeting time[16:09:10] <Ragnarok> holy shit[16:09:11] <Ragnarok> http://youtube.com/watch?v=bPTcYBSygjU&feature=related[16:09:30] *** TheLorax has quit IRC[16:14:04] <Ingenu> it's legal in many countries[16:14:09] <Ingenu> or at least they don't care[16:14:31] <hibread> "a more harmful drug in alcohol than a much safer drug in marajuana". Thats very debatable[16:15:14] <Ingenu> alcohol is worst indeed[16:15:27] <Ingenu> doesn't make marijuana good still[16:15:31] *** TheLorax has joined ##opengl[16:17:44] *** sanjin has left ##OpenGL[16:17:45] <hibread> Ingenu: i tend to agree with that... but marijuana has some serious psychological side effects that can effect people quite drastically[16:18:31] <hibread> in a different way than alcohol...[16:18:34] *** nathan_ has quit IRC[16:18:41] <Ragnarok> weed gave me a bad side effect 2 times :([16:18:46] <Ragnarok> still suffer from one :([16:18:58] <hibread> from one of the episodes?[16:19:03] <Ragnarok> ?[16:19:16] <hibread> you still suffer from one of your bad experiences?[16:19:17] *** Jorachim has joined ##OpenGL[16:19:18] *** _boto has joined ##opengl[16:19:43] *** IJs\laptop has joined ##OpenGL[16:20:03] <Ragnarok> yes :([16:20:13] <madsy_> Like.. coughing?[16:20:14] <Ragnarok> i can't say in here, too embarrassed[16:20:16] <Ragnarok> no[16:20:36] <madsy_> Heh, you don't beat me anyway[16:20:39] <madsy_> So, shoot[16:21:07] <madsy_> Last time I tried weed, I crapped from every opening of my body[16:21:12] <hibread> Ragnarok: is it something you did whilst under the influence? Or a mind altering type effect that stays with you to this day?[16:21:21] <hibread> hahaha[16:21:28] <Ragnarok> no a body effect[16:21:37] <hibread> "Crapped".. nice use of the word[16:21:42] <Ingenu> never tried[16:21:43] <prophile> i can guess[16:21:46] <madsy_> Luckily the party host had the toilet and wash very close[16:21:56] <Ingenu> there are a number of things I'm not even interested in trying :p[16:21:57] <prophile> i get the slight feeling that this is a little off topic[16:22:14] <prophile> unless there's now a GL_ARB_marijuana extension I haven't been informed about[16:22:24] <Ingenu> yes it's on probation ;)[16:22:38] <prophile> oh, neat[16:22:44] <Ingenu> in fact it's GL3_Khronos_marijuana[16:22:47] <Ingenu> ;p[16:22:54] <prophile> I guess that's what glGenBongsARB is[16:22:57] <prophile> i never had that figured[16:22:58] <madsy_> I woke up next morning inside a basket with dirty clothes.[16:23:31] <madsy_> Between me and the exit, I had to go through the host's bedroom. And he was asleep with his gf[16:23:39] <madsy_> I was sick like fuck[16:23:58] <madsy_> When I was about to leave, I was noticed by his dog. A big schaefer.[16:24:07] <Ingenu> lol[16:24:08] <madsy_> And it started yelloing like hell[16:24:25] <hibread> a dog that yells? nice one :)[16:24:37] <madsy_> barks then[16:24:40] <madsy_> Whatever[16:24:59] <madsy_> It's one of the worst things I've experienced. No more drugs for me[16:25:20] <hibread> I've resorted to coffee[16:25:23] <madsy_> Except for the occational cigarette[16:25:28] <madsy_> Yeah, and coffeee[16:25:33] *** amz has joined ##opengl[16:25:40] <hibread> actually, and tea... try to mix them up a bit[16:25:45] <Ingenu> no coffee for me either[16:25:45] *** Jorachim has left ##OpenGL[16:26:33] *** sanjin has joined ##OpenGL[16:30:08] <hibread> how do the dFdx dFdy functions work in glsl? Is there some inter-fragment reading going on? Checking value of other fragments around the subject fragment?[16:32:32] <hibread> or is the slope/gradiant etc calculatable during the post vertex shader processing?[16:35:02] <LtJax> I think the latter[16:35:17] <LtJax> it's used for perspective correction if I'm not mistaken[16:35:23] <prophile> no, it's the former[16:35:28] <prophile> parallel processing in the fragments[16:35:33] <prophile> means you can check values from either side[16:35:45] <prophile> it's a simple linear derivitive estimate[16:35:53] <prophile> if you take a dFdx of a dFdx you will get 0[16:36:31] <hibread> no worries guys. Cheers[16:37:06] <hibread> i think it will help me reconstruct XYZ values of pixels around the subject pixel using the depth buffer and a simple quad[16:40:04] *** DarkUranium has joined ##OpenGL[16:40:18] *** sanjin has left ##OpenGL[16:40:34] *** neoneye has quit IRC[16:43:25] *** mattn__ has quit IRC[16:46:46] *** Killari has joined ##OpenGL[16:50:53] *** PainBank_ has joined ##OpenGL[16:51:20] *** PainBank has quit IRC[16:52:41] *** dv_ has quit IRC[16:56:49] *** BahamutZERO has quit IRC[16:59:25] *** sohail has joined ##OpenGL[16:59:37] <Ragnarok> is there a way to call a function to render a cube, sphere etc with regular opengl, since i don't use glut[16:59:50] <hibread> glut is not opengl[17:00:20] <hibread> what function are you using to render a cube? some gluQuadric or something?[17:01:14] *** hubbe3 has quit IRC[17:02:11] *** dv_ has joined ##OpenGL[17:02:56] *** hubbe3 has joined ##OpenGL[17:07:23] <Ragnarok> i know glut isn't opengl[17:07:40] *** LtJax has quit IRC[17:07:57] <hibread> Ragnarok: how you are drawing your cube/spheres?[17:08:14] <hibread> i think you're probably using glu functions[17:08:17] <Ragnarok> manually[17:08:25] <Ragnarok> gl*[17:08:53] <hibread> i dont understand your question then. You're asking to rendering something with "regular opengl"... it sounds like you are doing so[17:09:31] <Ragnarok> i wanna know if there is a regular gl function to render a cube, sphere etc[17:09:54] <hibread> ah ok sorry. Check out the glu functions[17:11:31] <Ragnarok> no cube found sphere and cylinder[17:11:56] <hibread> cylinder with only 4 sides?[17:12:38] *** BahamutZERO has joined ##OpenGL[17:13:11] <hibread> I guess they assume that 8 vertices is not hard to specify.. unless ofcourse you want to tesselate that cube[17:13:41] <hibread> seems to be a whole bunch of gear there http://www.opengl.org/documentation/specs/man_pages/hardcopy/GL/html/glu/[17:14:06] <MatthiasM> Ragnarok: start blender - create your cube - export in the format of your choice - import it into your app - render it[17:14:22] <MatthiasM> and then later you can change your cube into a tree etc[17:14:25] <Ragnarok> blah vthats way too much work right there xD[17:14:37] <Ragnarok> but yeah i haven't gotten that far into graphics[17:14:56] *** neunon has joined ##opengl[17:15:07] *** TheLorax has quit IRC[17:15:19] *** TheLorax has joined ##opengl[17:18:42] *** cdleonar2 has joined ##OpenGL[17:18:45] *** cdleonar2 has left ##OpenGL[17:18:50] *** cdleonar2 has joined ##OpenGL[17:18:57] *** cdleonar2 has left ##OpenGL[17:19:10] *** amz has quit IRC[17:37:31] *** Ram3 has quit IRC[17:40:44] *** Castigador has quit IRC[17:41:42] *** bbeausej has joined ##OpenGL[17:43:38] *** Scutter has quit IRC[17:49:28] *** DMINATOR has joined ##OpenGL[17:58:22] *** [AD]Turbo is now known as TurboAWAY[18:02:45] *** basro has joined ##OpenGL[18:06:37] *** Eltran has quit IRC[18:18:46] <hibread> "The Guinness-recognized record for remembered digits of pi is 67,890 digits"[18:26:04] <Ingenu> it's reaaly worth remembering that[18:29:50] *** Jupp3 has joined ##OpenGL[18:30:38] *** Jupp3 has quit IRC[18:30:39] *** Gudy has joined ##OpenGL[18:33:32] <Ragnarok> hibread, i heard PI can reach 1 million digits[18:33:57] <hibread> Ragnarok: keep on going...[18:34:03] <Ragnarok> ?[18:34:44] <hibread> Ragnarok: there is no "end" to the nmber of decimal digits of pi[18:34:57] <Ragnarok> hehe[18:35:27] <Ragnarok> image in thery had the 1 million precision PI in a GL header :P[18:36:34] *** HuntsMan has joined ##opengl[18:39:55] *** groton has quit IRC[18:49:31] <basro> if I wanted to render video in 3d, I guess I would need to transfer the image to a texture each frame, is this bad?[18:49:43] <basro> :P[18:54:51] *** Jupp3 has joined ##OpenGL[19:04:12] *** sanjin has joined ##OpenGL[19:12:25] <sanjin> why is the output not 1 2 3 4 but 0 0 0 0 instead ? http://cpp.ninjacodemonkeys.org/4184[19:13:18] <HuntsMan> basro: no[19:17:28] <basro> I was told uploading a texture is slow[19:17:33] <basro> how slow is it?[19:17:36] <basro> :P[19:17:37] <basro> I mean[19:17:59] <bobbens> slow as in don't do it every freaking frame :)[19:18:17] <basro> say I made a game which supports animated textures, those textures would change each frame, thats bad, isnt it?[19:18:30] <bobbens> preload textures and blit different ones[19:18:36] <HuntsMan> no, you can upload every frame at init time[19:18:41] <HuntsMan> i do that :)[19:18:45] <basro> yes, but what if the animation is procedural?[19:18:48] <HuntsMan> and just change the current texture[19:18:54] <sanjin> can someone help me with my problem?[19:18:56] <bobbens> pregenerate them at the beginning if possible[19:19:17] <sanjin> is is possible to use a texture which you set up via glTexImage2D(... data=0) for a framebuffer ?[19:19:26] <bobbens> there's ways to update textures and such, but I'm not an opengl expert :)[19:20:16] <bobbens> i'd look at what them video players do[19:20:39] <sanjin> the one answering my question on 19:12 will get a nice fresh cookie :>[19:20:57] <basro> time is relative to your position in earth[19:21:23] <sanjin> sorry ^^[19:21:32] <basro> ;O[19:21:35] <sanjin> :D[19:22:51] *** amalon has joined ##opengl[19:23:05] *** amalon has quit IRC[19:23:21] *** inktree has quit IRC[19:24:00] *** LiQuiDninja has joined ##OpenGL[19:27:24] *** Obfuscate has quit IRC[19:27:42] <sanjin> so did someone take a look at my code?[19:29:22] <bobbens> i did[19:29:33] <bobbens> but i haven't used fbo and too tired to read :P[19:29:36] <bobbens> the lines blur together![19:31:13] <sanjin> no problem ;) thanks though[19:33:40] *** rutski has joined ##OpenGL[19:34:00] <sanjin> ah, i just read: "glDrawPixels() requires a shader to be bound on NV3x. Use glTexSubImage() instead (which is recommended for all NVIDIA GPUs anyway)."[19:34:22] <HuntsMan> sanjin: where?[19:34:28] <sanjin> http://www.mathematik.uni-dortmund.de/~goeddeke/gpgpu/tutorial.html#issues[19:34:40] *** mm^away is now known as mm765[19:34:56] <HuntsMan> lol[19:34:59] <HuntsMan> weird requerimient[19:35:02] *** cami has joined ##OpenGL[19:35:03] <HuntsMan> as GL doesn't require it[19:35:34] <sanjin> yup[19:35:52] <sanjin> it also sais: "texture download with glDrawPixels() fails on ATI when texturing is enabled. This is not a bug, but an issue that can be very hard to debug.", ill give that a try[19:36:11] <HuntsMan> LOL![19:36:24] <sanjin> lol?[19:36:56] <HuntsMan> Laugh out loud[19:37:15] <sanjin> em... yes... but why?[19:37:37] <HuntsMan> well, this is not a bug... sounds funny[19:37:46] <sanjin> yes, thats right[19:38:54] <HuntsMan> well where i work we don't use ATI for it's outstanding bugs with it's opengl implementation[19:40:36] <sanjin> do you have an idea why the download with glDrawPixels() fails in my code huntsman?[19:41:57] <HuntsMan> no, but you "upload" a texture with glDrawPixels, not download one AFAIK[19:42:50] <sanjin> well, okay[19:43:36] <HuntsMan> try rendering into the FrameBuffer[19:43:43] <HuntsMan> as paint a color dot[19:43:56] <sanjin> yes, but i want to use glDrawPixels...[19:44:20] <HuntsMan> yeah, but it will be useful to debug where it's the problem[19:44:52] <sanjin> the point is... if i use glSubTexImage2D to upload data to the texture, it works[19:46:40] <HuntsMan> yeah but you're uploading it directly to the texture[19:47:28] <sanjin> and i want a way to indirectly upload it to the texture via the framebuffer...[19:48:22] <HuntsMan> and if you draw a textured quad?[19:49:03] *** admax88 has joined ##OpenGL[19:49:18] <sanjin> i could - but i want to avoid changing render states, modelview etc.[19:49:40] <sanjin> i mean - it has to be possible without rendering[19:50:05] *** LtJax has joined ##opengl[19:50:36] <HuntsMan> sure, but you should do it as a debugging way, not a final method[19:50:52] *** mastro has quit IRC[19:51:06] <HuntsMan> BTW, did you check for framebuffer completeness?[19:51:29] <sanjin> yes i did, its complete[19:51:42] <sanjin> okay i will try it as a debugging way[19:52:44] *** mastro has joined ##OpenGL[19:55:46] <sanjin> it works if i put the data into a 2nd texture and render it[19:57:21] <sanjin> http://cpp.ninjacodemonkeys.org/4185[19:58:10] <HuntsMan> what card do you have?[19:58:40] <sanjin> nvidia geforce 7800 GS+[19:59:07] <HuntsMan> and glDrawPixels does generate a glError()?[20:00:31] <sanjin> no glGetError() is GL_NO_ERROR[20:02:34] <HuntsMan> can you put a complete program so i can test here?[20:02:43] <HuntsMan> here == GeForce 8400 GS[20:03:32] <sanjin> http://cpp.ninjacodemonkeys.org/4186[20:04:16] <sanjin> sorry, this was the one with the renderquad[20:05:03] <sanjin> now its ok[20:05:17] *** TheLorax has quit IRC[20:08:01] *** TheLorax has joined ##opengl[20:09:29] *** Suprano has joined ##OpenGL[20:10:01] *** Obfuscate has joined ##opengl[20:12:25] *** nplus has joined ##OpenGL[20:14:04] <sanjin> do you get the same results?[20:15:12] *** PainBank_ has left ##OpenGL[20:19:46] <HuntsMan> wait, installing glew :P[20:20:53] *** DarkUranium has quit IRC[20:22:25] <HuntsMan> yeah, the same, 0 0 0 0[20:22:44] <sanjin> thanks for testing[20:22:58] <sanjin> but you also think it has to work?[20:23:10] <HuntsMan> why do you need to use glDrawPixels BTW?[20:25:20] *** groton has joined ##OpenGL[20:25:46] *** Xmas| has joined ##OpenGL[20:29:10] <sanjin> i want to write a library and it should have the possibility to upload pixels to a framebuffer without rendering[20:33:32] *** SmokingRope has quit IRC[20:33:59] *** phrosty has quit IRC[20:34:35] *** cami has quit IRC[20:39:02] *** TheLorax has quit IRC[20:39:26] *** smiler has joined ##opengl[20:39:33] *** phrosty has joined ##opengl[20:41:51] *** m4ggus has joined ##opengl[20:43:53] *** Jupp3 has quit IRC[20:51:31] <basro> is there some AND operation for writting into the stencil buffer?[20:53:16] *** sanjin has left ##OpenGL[21:01:22] *** HuntsMan has quit IRC[21:06:49] *** Jupp3 has joined ##OpenGL[21:07:53] *** SmokingRope has joined ##OpenGL[21:12:26] *** Ingenu has quit IRC[21:12:52] *** mattn2|home has joined ##OpenGL[21:19:04] *** groton has quit IRC[21:19:08] *** Jorachim has joined ##OpenGL[21:20:41] *** Ademan_ has joined ##OpenGL[21:24:26] *** groton has joined ##OpenGL[21:25:19] *** blbmp has joined ##OpenGL[21:33:08] *** nplus has quit IRC[21:36:15] *** Ademan has quit IRC[21:42:03] *** speedy1 has joined ##OpenGL[21:44:04] *** neunon has quit IRC[21:44:36] *** neunon has joined ##opengl[21:47:58] *** neunon has quit IRC[21:50:48] *** DobosCake has joined ##OpenGL[21:52:46] *** SmokingRope has quit IRC[21:53:24] *** admax88 has quit IRC[21:57:16] *** elite01 has quit IRC[22:04:54] *** fargiolas has joined ##OpenGL[22:08:09] *** m4ggus has quit IRC[22:13:05] *** m4ggus has joined ##OpenGL[22:21:48] <rutski> I'm rendering fonts into a texture, and then slapping that texture onto a polygon[22:22:03] *** Walt has quit IRC[22:22:17] <rutski> I check the size of the viewport to make sure to made the polygon the proper width/height so it'll be displayed as exactly as many pixels as are in the texture[22:22:54] <rutski> however, when I place the polygon I place it with glVertex2f into a floating point orthographic projection that runs from -1 to 1 on both the X and Y axis[22:23:28] <rutski> given that the polygon is pixel perfect size, and with a pixel perfect texture; will placing it into a non-pixel perfect location cause problems?[22:24:56] <Xmas|> no[22:25:03] <Xmas|> precision is high enough for that[22:32:42] <speedy1> rutski - to hit a certain pixel you need to have resolution bigger then 1/xres and 1/yres which is not a big problem for 32bit floats[22:33:29] <rutski> right, but the point is that I'm not choosing floats that correspond to exact pixels[22:33:51] <rutski> so lets say that the viewport is 400x400, and the ortho projection is -200 -> 200 on both the X and the Y[22:33:56] *** groton_ has joined ##OpenGL[22:34:20] <rutski> and then I use glVertex2f(50.5, 70.24) to place the textured polygon[22:34:35] <rutski> will that screw things up?[22:35:45] <rutski> given that the size of the polygon is exactly, lets say, 10 by 20 pixels, as is it's texture[22:36:01] <rutski> oh, sorry; I meant to say that I place the top left of the polygon at glVertex2f(50.5, 70.24)[22:36:13] <rutski> and then the other points at +/-10 and +/-20[22:36:51] <speedy1> 50.5 could be problematic as it is exactly on border between pixels[22:37:12] <speedy1> 70.24 will work OK[22:38:14] <rutski> huh?[22:38:17] <rutski> wait.. why is that?[22:38:26] <rutski> because 70.24 will pick the closest pixel?[22:38:36] *** groton_ has quit IRC[22:38:59] <speedy1> yes you hit a pixel which lies from 69.5 to 70.5[22:39:11] <rutski> what pixel is that?[22:39:24] <speedy1> pixel 70[22:39:34] <rutski> that's not what I described; no[22:39:45] <rutski> the viewport is 400x400, and the projection is -200 to 200 on both axis[22:40:01] <rutski> so doesn't the space 69.5 to 70.5 touch[22:40:03] <rutski> oh... wait[22:40:04] <rutski> I see what you mean[22:40:08] <Xmas|> rutski, why don't you use exact floats?[22:40:09] <speedy1> :)[22:40:41] <rutski> Xmas|: because the way I have it coded at the moment was the most natural way; without exact floats[22:40:54] <rutski> and I'm not thinking if it's worth doing the extra work to make them exact[22:41:02] <rutski> or if OpenGL already does that for me under the hood[22:41:19] <rutski> hmm, does the fact that the texture is set to GL_NEAREST for interpolation going to have any effect here?[22:41:24] <Xmas|> just make sure to use nearest sampling[22:41:37] <speedy1> xmas: that's for reading from textures[22:41:40] <Xmas|> yes[22:41:41] <rutski> doesn't that guarantee pixel perfect mapping?[22:41:49] <rutski> see, the problem is this; let me take a screen shot[22:42:37] <speedy1> keep in mind to not confuse reading from textures and writing to screen[22:42:41] <Xmas|> if you want to just blit a texture to the screen then you should use nearest sampling[22:42:44] <rutski> http://www.rutski89.com:31597/upload/Picture%2012.png[22:42:51] <rutski> note the weirdness on the bottom curve of the 'B'[22:43:33] <rutski> speedy1: wait... what? what was that about GL_NEAREST being used for reading?[22:43:46] <rutski> speedy1: I only learned how to use texturing yesterday by the way[22:44:34] <speedy1> GL_nearest should give you point sampling from texture[22:44:38] <rutski> I thought GL_NEAREST being set on GL_TEXTURE_{MAG,MIN}_FILTER made it such that each fragment in the viewport is mapped to exactly one texture pixel[22:45:07] <rutski> hmm, but then what's causing the 'B' to be broken like that? http://www.rutski89.com:31597/upload/Picture%2012.png[22:45:08] <speedy1> nono texture filters and nearest stuff is only for reading from textures[22:45:44] <speedy1> loosely said, writing (rasterizing) is allways GL_NEAREST[22:45:52] <rutski> that's the part I don't get; I don't know what you mean by "for reading"[22:46:22] <speedy1> simple - you read from textures and display results to screen[22:46:35] <Xmas|> textures are sampled (or "read") to calculate a fragment color, which is written to the framebuffer[22:46:41] <rutski> right[22:46:46] <speedy1> you write results to screen[22:46:49] <rutski> and fragment == pixel; right?[22:47:04] <Xmas|> for the purpose of this discussion, yes[22:47:08] <rutski> right[22:47:24] <rutski> so I don't understand the meaning behind saying "GL_NEAREST is only for reading"[22:47:31] <rutski> what else is there other than reading?[22:47:38] <speedy1> writing to screen[22:47:40] <rutski> what did you think that _I thought_ it was for?[22:47:49] <rutski> hm[22:47:56] <rutski> not sure what you mean by that[22:48:12] <rutski> you mean to say that the way which fragments are written to the screen can be controls via some options?[22:48:25] <speedy1> one could say that writing to screen is allways in GL_NEAREST[22:48:28] <speedy1> mode[22:48:32] <rutski> ah, right[22:48:43] <rutski> ok; so you were just using clearer terms for the same idea then[22:48:55] <rutski> ok, now that that confusion is cleared away[22:49:01] <rutski> any ideas about the broken 'B'?[22:49:02] <rutski> http://www.rutski89.com:31597/upload/Picture%2012.png[22:49:26] <Xmas|> do you have antialiasing enabled?[22:49:52] <Xmas|> and are you sure the polygon you're mapping the B to is the right size?[22:49:58] <rutski> I'm not sure; I only learned freetype yesterday as well[22:50:11] *** sanjin has joined ##OpenGL[22:50:19] <rutski> Xmas|: I'm fairly sure it's the right size yes; but maybe I let some small inaccuracy creep into the math. I'll double check.[22:50:38] *** sanjin has quit IRC[22:50:42] *** kiras has joined ##OpenGL[22:56:14] <odietsch> rutski, from what I've read a glTranslate(0.375,0.375,0.0) will give correct rasterization[22:56:19] <odietsch> though I'm not really sure[22:56:40] <rutski> odietsch: did you pick those numbers for a good reason; or are they just random?[22:56:52] <rutski> are you saying that any floating point numbers will give correct rasterization; or just those in that specific case?[22:56:59] <odietsch> just those[22:57:05] <rutski> ah[22:57:17] <rutski> now the question is; why wouldn't others give correct rasterization?[22:57:22] <odietsch> I caught that from a graphics programming book, talking a bit about GL[22:57:45] <odietsch> maybe because it's the magic bit missing :-)[22:58:04] <odietsch> fire it right after your glOrtho, which I suspect you to be using[22:58:46] *** Yustme has quit IRC[22:59:09] *** Stefan100 has joined ##OpenGL[22:59:14] <Stefan100> :o[22:59:16] <rutski> ?[22:59:22] <Stefan100> o:[22:59:24] <rutski> ?[22:59:33] <Stefan100> ![22:59:34] <rutski> ?[22:59:42] <Stefan100> .[22:59:43] <rutski> ?[22:59:54] *** groton has quit IRC[23:00:03] <Stefan100> ._.[23:00:04] <odietsch> zomg[23:00:05] <rutski> ?[23:00:19] <Stefan100> Did I make your bot go into a loop? :|[23:01:47] <Stefan100> hmm[23:02:03] *** karabash has joined ##OpenGL[23:02:18] *** Walt has joined ##opengl[23:06:56] *** bbeausej_ has joined ##OpenGL[23:07:28] *** bbeausej has quit IRC[23:08:13] *** bbeausej_ is now known as bbeausej[23:08:20] <odietsch> rutski, out of curiosity, was my hint of any use?[23:14:15] *** mm765 is now known as mm765^away[23:16:00] *** Lemml has quit IRC[23:16:41] *** speedy1 has quit IRC[23:21:18] *** HuntsMan has joined ##opengl[23:23:16] *** DMINATOR has quit IRC[23:31:54] <rutski> odietsch: nope, I didn't get to try it; those were just made up numbers[23:32:08] <rutski> odietsch: plus I don't understanding the reasoning behind it anyway[23:33:35] <odietsch> the problem of yours is that some segments of a font are missing or too thin, right?[23:34:33] *** EvilGuru has joined ##OpenGL[23:34:49] *** Ademan_ has quit IRC[23:35:09] <odietsch> those numbers just present a really tiny translations not actually visible to the users eyes, but that small bit is thought to be kicking in when having a problem with wrong rasterization when using texture-filtering[23:35:51] *** madsy_ is now known as Madsy[23:35:53] <EvilGuru> I need to draw an image to an OpenGL context several times a second. The image is generated by a software rendering library (cairo). I initially did it by uploading the image as a texture and drawing a rectangle, however found glDrawPixels to yield better performance. Are there any other methods which could be used for this task?[23:37:05] <odietsch> rutski, also the redbook says that after googling: http://fly.cc.fer.hr/~unreal/theredbook/appendixh.html <- search there for that _magic_ number[23:37:22] *** scy has joined ##opengl[23:38:43] <rutski> odietsch: right, see the B http://www.rutski89.com:31597/upload/Picture%2012.png[23:39:47] <rutski> oh, heh; 0.375 is like 1/6th of the viewport though[23:39:47] <odietsch> Isn't that exactly the same shot as before?[23:40:07] <rutski> yea, it is; I just didn't know if you'd seen it yet[23:40:11] <rutski> plus I'm not using rasterization functions[23:40:16] <rutski> I'm slapping textures onto polygons[23:40:34] <odietsch> well, that's it, with rasterization I meant what's happening on the low-level[23:40:41] *** PhrostByte has joined ##opengl[23:40:44] <odietsch> ... when slapping textures onto polygons[23:40:47] *** Stefan100 has left ##OpenGL[23:40:47] <rutski> I read somewhere that it's common for textured polygons to have petter performance than rasterization functions[23:41:28] <odietsch> the glTranslate will help in exactly the case you have[23:41:40] <odietsch> it's just one line, it won't hurt :)[23:41:46] <kiras> can someone help me figure out what exactly is happening here? i'm new to opengl... so i wrote some code to load a .obj file and i'm using a header i found to load a .tga texture... i think my mapping is correct, but for some reason some weird stuff is happening. it looks better than it did before (seemed i had the texture upside-down) but i have no idea what it's doing now. http://img217.imageshack.us/img217/3865/cottage09qx3.pn[23:41:46] <kiras> g it *should* look more like this http://i51.photobucket.com/albums/f366/Zeffer17/cottage.jpg[23:41:52] <rutski> I'd have to scale it down hugely though[23:42:01] <rutski> 0.3 is 1/6th of the viewport[23:42:28] <odietsch> if you're using the "default" glOrtho and friends for setting up 2D, you'll be fine without scaling[23:44:12] <rnx> kiras: depth test would be my first guess ... beyond that: testcase[23:45:27] <kiras> rnx: thanks again for your help the other day. what exactly do you mean by "depth test?" do you think i'm too close to the cottage?[23:51:28] *** scy has left ##opengl[23:54:03] *** phrosty has quit IRC[23:54:55] *** EvilGuru has left ##OpenGL