January 9, 2009 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
January 9, 2009 [00:03:00] *** groton has quit IRC[00:04:40] *** tuxer` has quit IRC[00:05:53] <aep> hey any idea how to create an offscreen buffer with X? i tried glXCreateGLXPixmap , but the context i have from glXCreateContext breaks on intel drivers[00:06:11] <MatthiasM> try Pbuffer[00:06:12] *** scai has left ##opengl[00:06:52] <aep> thanks[00:07:24] <aep> um any idea what the command to create them is?[00:07:40] *** prak_ has joined ##OpenGL[00:08:02] <MatthiasM> aep: use a lib which handles cross platform issues like this[00:08:09] <aep> "It is an experimental extension- it may be changed in the future."[00:08:10] <MatthiasM> like GLFW[00:08:20] <aep> most libs are not cross platform at all[00:08:24] <aep> going to look at glfw[00:08:29] *** elite01 has quit IRC[00:08:40] <MatthiasM> LWJGL is :)[00:08:58] <aep> thats java[00:09:11] <MatthiasM> yep[00:09:32] <aep> well i'm writing a low level toolkit[00:09:38] <b0000> aep: use nvidia hardware. they seem to be the only vendor capable of writing good opengl support[00:09:40] <aep> so, neither of the named work for me[00:09:54] <aep> b0000: not an option.[00:10:03] <aep> other glx programs appear to run fine[00:10:14] <aep> but yeah on my nvidia machine, the code works[00:10:24] <aep> propably i shouldnt work on it. too good[00:10:37] *** TuomasT has quit IRC[00:10:46] <MatthiasM> aep: my point is that these other libs work[00:11:12] <MatthiasM> and if you don't want to use them - you can still look how they do it[00:11:23] <aep> MatthiasM: yeah good point[00:11:32] <aep> i was just hoping there was a common workaround[00:12:17] <MatthiasM> well - I only use native display + FBO and sometimes pbuffer[00:12:27] <MatthiasM> and I use it through LWJGL[00:12:40] <aep> fbo?[00:12:53] <MatthiasM> frame buffer object[00:13:02] *** oliveira_ has quit IRC[00:13:05] *** djork has quit IRC[00:13:25] <aep> oh[00:13:33] <aep> well there is a double buffer extension in Xgl[00:13:35] <aep> its just garbage[00:13:54] <aep> all i need that for is double buffering[00:14:02] <aep> becouse Xorg is garbage in the first place[00:14:04] <aep> *sigh*[00:14:11] <MatthiasM> huh?[00:14:26] *** oliveira is now known as oliveira_[00:14:43] <aep> double buffering is only needed becouse the native window system doesnt have it[00:14:47] <aep> thats the only platform that doesnt[00:14:52] *** Nescafe has quit IRC[00:14:53] <aep> (if you exclude pre stonage mac)[00:15:02] <MatthiasM> ??[00:15:13] <mlucassmith> what?[00:15:45] <aep> um i guess you just repaint a gazillion times a second?[00:16:11] <MatthiasM> the opengl window ? yes[00:16:14] <aep> i'm writing a gui toolkit. apps are supposed to run at minimal cpu/gpu. so i cant repaint until i actually have to[00:16:40] <MatthiasM> yep - me too[00:16:43] *** prak_ has quit IRC[00:16:59] <aep> so, how did you implement double buffering if not via offscreen buffer?[00:17:15] <MatthiasM> I do full redraw[00:17:18] <aep> i cant read that format[00:17:29] <aep> do no buffering at all?[00:17:33] <MatthiasM> it's java webstart - run it with "javaws <jnlp-file>"[00:17:45] *** yates has joined ##OpenGL[00:17:46] <aep> java doesnt run on my machine[00:17:51] <MatthiasM> GL bouble buffering - eg clear, render, swap[00:17:57] <MatthiasM> *double[00:18:01] <MatthiasM> why not?[00:18:17] <aep> i dont know. ask sun[00:18:33] <yates> does opengl/glx/the driver perform vertex calculations internally with floats or integers?[00:18:42] <MatthiasM> floats[00:18:47] <aep> got a windows machine, wait[00:18:48] <MatthiasM> except fot GL|ES[00:18:59] <yates> GL|ES?[00:19:02] <MatthiasM> where it uses fixed point[00:19:08] <yates> ES?[00:19:14] <MatthiasM> yates: for mobile devices[00:19:17] <yates> ah[00:19:21] *** iion has quit IRC[00:19:23] <yates> ok, thanks MatthiasM[00:19:34] <aep> MatthiasM: extremly slow on windows XP SP2 gforce 7600.[00:19:36] <aep> its unusable[00:19:44] <aep> and eats 100% cpu[00:19:47] <aep> well its java...[00:19:50] <MatthiasM> lol[00:19:51] <yates> i'm seeing something i didn't expect re: profiling[00:20:01] <MatthiasM> I have a GF7800 and the CPU load is 0%[00:20:09] <yates> a significant amount of cpu time is spent just casting integers to floats![00:20:17] <yates> (in my tessellation routine)[00:20:18] <MatthiasM> yates: heh[00:20:20] <aep> *shrug* "works for me". hear that alot[00:20:29] <yates> without even calling anything openGL![00:20:42] *** signalsea has quit IRC[00:20:54] <MatthiasM> aep: this app generates nearly no garbage[00:20:56] <yates> things like vert[n][k] = (float)n / N;[00:21:10] <MatthiasM> so there is no reason that it runs slow - except if you have a 100MHz CPU[00:21:24] <yates> any tricks for speeding up float conversions like this?[00:21:31] <aep> MatthiasM: 2Ghz Intel Dual core[00:21:36] <yates> quick tricks, that is?[00:21:39] <aep> MatthiasM: i dont know what garbage is[00:21:53] <yates> aep: go to Redmond[00:22:07] <MatthiasM> yates: that's highly unlikly that this line is your bottleneck[00:22:09] <aep> and do what?[00:22:28] <yates> aep: there's a massive garbage factory there...[00:22:43] <MatthiasM> aep: you allocate memory and let it fall down - then if there is a larger pile of garbage - the collector comes and takes it away[00:22:52] <aep> oh that. java GC[00:23:08] <MatthiasM> and normaly this takes about 1ms every few minutes[00:23:21] <aep> i guess you got a busy loop in it?[00:23:23] <yates> oh[00:23:27] <aep> something that just keeps drawing?[00:23:35] <yates> java - eeew![00:23:46] <MatthiasM> aep: maybe you forces vsync off ?[00:24:18] <aep> i dunno. sorry. not in the mood to debug a java app[00:24:33] *** Nostrass has joined ##OpenGL[00:24:34] <MatthiasM> it's your system[00:24:38] <Nostrass> Anyone here knows about an algorythm to triangle rasterization?[00:24:39] <aep> ok. whatever[00:24:42] *** prak has quit IRC[00:24:52] <MatthiasM> Nostrass: use OpenGL[00:24:55] <yates> ha[00:25:04] <aep> so, got any hints how to implement double buffering on X11 without relying on sane drivers like nvidia?[00:25:20] <Nostrass> but before use it, i need to know how i can rasterize triangles.[00:25:24] <mlucassmith> aep: use framebuffers[00:25:33] <MatthiasM> Nostrass: read the red book[00:25:44] *** yates has quit IRC[00:25:46] <aep> mlucassmith: can you tell me more? i'm not sure what you mean[00:25:54] <mlucassmith> www.google.com[00:26:12] <aep> the linux framebuffers i know are not related to X11 at all[00:26:19] <aep> and google will tell me the same[00:26:20] <mlucassmith> i'm talking about opengl frame buffer objects[00:26:26] <aep> thanks.[00:26:31] <aep> thats a better keyword ;)[00:27:15] <aep> looks like X doesnt have that[00:27:30] <MatthiasM> aep: because it's an GL extension ![00:27:32] <mlucassmith> X doesn't have to have it, it's part of GL[00:27:52] <aep> so i'm supposed to find something in Gl.h?[00:28:01] <MatthiasM> no[00:28:04] <mlucassmith> no, FBOs are explained in the GL spec[00:28:09] <aep> confusing[00:28:17] <MatthiasM> glext.h[00:28:21] <aep> aye[00:28:32] <MatthiasM> and use GLEW to handle extensions[00:28:54] <aep> yeah i read the topic[00:28:58] <aep> so i need to load that?[00:29:24] <aep> means i need to rely on the availability of the extension, right?[00:29:34] <aep> which might be problem on linux[00:29:55] <MatthiasM> no - not on linux - depends on your drivers[00:29:57] <MatthiasM> and GPU[00:30:06] <aep> ew, so its worse[00:30:07] <MatthiasM> it works on linux without issues[00:30:08] <mlucassmith> so yes, a problem on linux :)[00:30:17] <aep> everything that can be removed will be removed by some distro some day[00:30:23] <MatthiasM> I use it on linux using a nVidia GPU[00:30:28] *** HuntsMan has joined ##opengl[00:30:33] <aep> well nvidia isnt the problem here[00:30:43] <aep> nvidia can handle glxCreatePixmap[00:30:44] <mlucassmith> the mesa driver that comes with Ubuntu 8.04 is OpenGL 1.6[00:30:53] <MatthiasM> 1.6 ?!?[00:31:00] *** Amorphous has quit IRC[00:31:42] <mlucassmith> yep, 1.6[00:31:59] <MatthiasM> 1.1 - 1.5, 2.0, 2.1, 3.0[00:32:00] <mlucassmith> better than Windows :)[00:32:04] <boghog> I think with mesa (and fgglrx) you're also limited to glx 1.2[00:32:20] <mlucassmith> oh sorry, you're right, 1.5[00:32:24] *** neoneye has quit IRC[00:32:25] <MatthiasM> heh[00:35:08] <boghog> Nostrass, if you really want to know how to rasterize triangles you could look at http://chrishecker.com/Miscellaneous_Technical_Articles#Perspective_Texture_Mapping (its mostly about texture mapping but deals with rasterization at the same time)[00:35:40] *** mm765^away is now known as mm765^sleep[00:41:59] *** Amorphous has joined ##opengl[00:43:37] *** vade has joined ##OpenGL[00:45:52] <aep> meh. i fail. the documentation clearly says that DRI on GlxPixmap wont work[00:46:04] <aep> sucks[00:46:40] <aep> yet, if i disable dri, all i get is garbage[00:46:56] <MatthiasM> aep: just use a pbuffer[00:47:12] <mlucassmith> make a window and don't map it, that'll work too[00:47:26] <aep> 1)i dont know how 2) i cannot rely on a distro not removing something optional. debian loves to do that[00:47:36] <aep> mlucassmith: not really, i tried. :/[00:47:44] <aep> it doesnt paint on it if it isnt mapped[00:47:51] <aep> XCopy => garbage[00:48:06] <mlucassmith> make a hidden window and map it[00:48:07] <MatthiasM> aep: pbuffer is not optional - and a distribution doesn't decide about GL extension either[00:48:32] <aep> mlucassmith: i didnt know that is possible. how?[00:48:41] <MatthiasM> mlucassmith: an unmapped window would cause all pixel ownerchip tests to fail[00:49:08] <aep> MatthiasM: so the vendors do? isnt that even worse ?[00:49:18] <aep> dont vendors break things on a regular basis?[00:49:28] <MatthiasM> aep: ATI does[00:49:30] <mlucassmith> on RTFM_FTW breaks things :)[00:49:34] <mlucassmith> only[00:49:58] <MatthiasM> but the driver and HW defines what features are available - orMESA if no HW driver is available[00:49:58] <aep> MatthiasM: so why do you suggest t use it, if you know it will break?[00:50:33] <MatthiasM> aep: use it - because it's the only thing you can count on[00:50:41] <mlucassmith> pbuffers have been around a while, they'll work[00:50:49] <aep> ok good[00:50:51] <aep> thank you[00:50:56] <MatthiasM> aep: and you can't write software that will still run in 10 years without adjusting it[00:51:04] <aep> that sounds good[00:51:22] *** Aintaer has joined ##OpenGL[00:51:24] *** Kasu has quit IRC[00:51:32] <aep> so i'll have to learn more insider stuff of opengl, or is there a copy and paste like "tutorial" for me?[00:51:38] <mlucassmith> MatthiasM: sure you can, someone will make an emulator :)[00:51:45] <MatthiasM> aep: read the red book[00:52:06] <aep> i had a look. didnt really read it. does it cover extensions?[00:52:12] <MatthiasM> mlucassmith: well - if the environment doesn't change then it will keep on working[00:52:22] <MatthiasM> but I assumed upgrades to OS/HW/libs etc[00:52:43] <MatthiasM> aep: it covers GL core - and the latest book covers also some extensions[00:52:46] *** urnotE` has joined ##OpenGL[00:52:49] <mlucassmith> I'm just assuming it'll work 2 years, break for 8, then the emulator appears and it works again :)[00:53:17] <MatthiasM> heh[00:53:25] <MatthiasM> yeah - likely[00:53:48] *** McLovin` has quit IRC[00:53:56] *** zwiep` has joined ##opengl[00:57:15] *** urnotE has quit IRC[00:58:17] *** essoe- has joined ##opengl[00:58:38] *** mRCUTEO has joined ##OpenGL[01:00:26] *** soc has joined ##OpenGL[01:02:02] <soc> hi[01:02:14] <soc> what's the current status of opengl?[01:02:15] *** mRCUTEO has left ##OpenGL[01:02:27] <MatthiasM> soc: website up and running[01:04:07] *** korff_home has quit IRC[01:04:38] *** Nostrass has left ##OpenGL[01:04:38] <soc> i meant the spec ... after the luke-warm reception of opengl, things got pretty quit ...[01:05:18] <soc> is there any work being done to deliver what was promised? or is that now cancelled alltogether now?[01:05:26] <MatthiasM> soc: a) http://mikeash.com/getting_answers.html b) http://www.opengl.org/[01:06:19] *** yates has joined ##OpenGL[01:06:28] *** iion has joined ##OpenGL[01:08:03] <tmccrary> soc: You should need to research opengl more, you don't seem to understand what it is[01:08:10] <yates> when glDrawArrays() is given, does it transfer ALL enabled arrays, e.g., the vertex array and texture coordinate array (assuming these have been glEnableClientState()'ed and glVertexPointer() and glTexCoordPointer() have been called)?[01:08:16] <tmccrary> err, you should research opengl more ;)[01:08:24] *** _JSC415 has joined ##OpenGL[01:08:29] <_JSC415> so.[01:08:30] *** AlastairLynn has quit IRC[01:08:40] <_JSC415> did anyone used tokamak once before ?[01:08:45] <yates> that is, when ONE glDrawArrays() command is given?[01:08:47] <soc> tmccrary: mhh, could you explain a bit further?[01:08:53] <_JSC415> i need to know what that ne.h is ?[01:09:00] <_JSC415> i cant find it on samples[01:09:17] <_JSC415> and google says its neh.[01:09:22] <_JSC415> and not ne.h[01:09:25] <tmccrary> "OpenGL (Open Graphics Library) is a standard specification defining a cross-language cross-platform API for writing applications that produce 2D and 3D computer graphics."[01:09:30] <tmccrary> http://en.wikipedia.org/wiki/OpenGL[01:09:35] <_JSC415> or can anyone help me about searching ne.h ?[01:09:36] <tmccrary> I think that answers your original question nicely[01:09:50] <soc> yes, what was wrong with my question?[01:10:14] <tmccrary> It doesn't make much sense[01:10:17] <soc> mhh, i try to rephrase, maybe i wasn't clear ...[01:10:40] <tmccrary> The opengl specification was released a while a go[01:10:47] <yates> "Boot to the head" beginning line from Ed Grubermann: "ahh - sir, excuse me..."[01:10:59] <MatthiasM> yates: when you issue an array draw call - it uses all enabled arrays/buffers - so make sure you set valid pointers which are large enough[01:11:38] <yates> so it uses the same "first" and "count" for all enabled arrays?[01:11:45] <MatthiasM> yates: yes[01:11:48] <yates> cool[01:11:54] *** paul424 has joined ##OpenGL[01:12:14] <paul424> is there more generall channel on computer graphics[01:12:15] <soc> before the release of opengl 3.0 khronos promised a "new, shiny" api, but when the spec was published the old api wasn't replaced, but profiles and extensions were added, features marked as deprecated. my question now is: is there any work being done to deliver what was promised? or is that now cancelled alltogether now?[01:12:43] <MatthiasM> soc: read the forum[01:12:58] <tmccrary> yes its all cancelled[01:13:07] <soc> mhh, i read quite a lot, most of the people were flaming ...[01:13:19] <soc> so opengl will add extensions and so on but the api won't really change?[01:13:30] <tmccrary> Did you read the wikipedia page???[01:13:39] <yates> http://www.digitalsignallabs.com/BootToTheHead.mp3[01:14:30] <paul424> is there more generall channel on computer graphics ?[01:15:16] <MatthiasM> paul424: ask for the channel list :DD "/list channel" or so[01:16:18] <paul424> server load to heavy blahabalbahlbhalhbla, thx ....... ;([01:17:14] <paul424> ok anyway, what is the simplest roam algorithm you know ?[01:17:18] <b0000> MatthiasM: i am trying to export tangents from blender...but they are all the inverse of what i expect...any ideas?[01:17:43] <MatthiasM> b0000: no idea - I create them myself when I convert the model into my engine format[01:17:48] <soc> ah, ok, i missed "The proposed object model was not included, and no plans have been announced to include it in any future revisions."[01:17:55] <soc> that basically answers my question ...[01:18:02] <soc> sorry, for asking[01:18:22] *** Suprano has quit IRC[01:19:54] * yates thinks Boot To The Head is the best damn radio skit ever devised![01:20:15] * yates , having learned nothing from the lesson of Ed Gruberman, returns to his OpenGL vertex arrays...[01:22:58] *** Kasu has joined ##OpenGL[01:23:20] <b0000> MatthiasM: do you calculate tangents as vector3s and store the bitangent parity in the 4th component?[01:23:39] <MatthiasM> umm - let me check[01:24:18] *** dvoid_ has quit IRC[01:24:18] <MatthiasM> yep[01:24:59] <b0000> MatthiasM: do you have some text based format that you could nopaste a cube for me to compare with please?[01:25:12] <MatthiasM> engine format ?[01:25:43] <MatthiasM> tbh - I don't know the format that my engine uses :P[01:26:16] <MatthiasM> http://www.matthiasmann.de/worldscape but take a look if you want[01:28:31] *** essoe- has quit IRC[01:29:54] *** soc has left ##OpenGL[01:30:00] <paul424> MatthiasM: heap could we discuss the matter of roam ?[01:30:09] <MatthiasM> heap?[01:31:04] <paul424> its like please :D[01:31:04] *** HuntsMan has quit IRC[01:31:39] <MatthiasM> I only read about roam and tried the demos - never implemented it or used it[01:31:43] <MatthiasM> I think it's too slow[01:32:00] <MatthiasM> and uses too much memory[01:33:10] *** jfroy|work is now known as jfroy[01:33:38] *** jfroy is now known as jfroy|work[01:34:48] <paul424> look here http://www.vterrain.org/LOD/Papers/hoppe04.jpg, this seems to be enough simple idea for me to implement, but the paper which explains the implementation is to werid for me ......just I think how one could avoid cracks here on the edges .......[01:35:27] <MatthiasM> did you tried the link I posted above ?[01:35:40] *** DeforMobile has joined ##opengl[01:35:51] <MatthiasM> (ignore the missing trees - they are MIA :DD )[01:36:32] *** RTFM_FTW has joined ##OpenGL[01:37:25] <paul424> o_O that;s nice but what have to do with the ROAM brb[01:37:53] <MatthiasM> I don't use roam[01:38:05] <paul424> hmm the O_o should be like a more confused face ;)[01:38:11] <MatthiasM> I use something based on the idea from a gamasutra atricle[01:39:13] <MatthiasM> the O_o and O_o are displayed here as a smily with one big eye[01:39:19] <MatthiasM> *o_O[01:39:31] <paul424> that;s fine if you gpu can handle maximum resolution terrain, but my lab is to prepare one implementation ......... if you could take a look at the picture I provided[01:39:45] <MatthiasM> I use LOD a lot :DD[01:40:00] <MatthiasM> you can use the detail slider on the lower left[01:41:07] *** kbotnen has joined ##OpenGL[01:41:27] <MatthiasM> the difference is easiest visible on hills against sky[01:42:59] <paul424> ok but how do i fire up the demo ....... but I have only geforce2mx so it's to weak anyway ...;([01:43:17] <MatthiasM> http://www.matthiasmann.de/java/worldscape/btttest.jnlp[01:54:48] *** Spkka has quit IRC[01:56:44] <yates> why doesn't smoothing utilize normal information?[01:57:02] <yates> it would be nice if it did, so that triangles wouldn't be so flat...[01:57:25] *** Walt has quit IRC[01:57:44] <MatthiasM> yates: huh ?[01:58:25] *** Walt has joined ##opengl[01:58:45] <yates> brain-fart, i guess[01:58:50] <yates> excuse me[01:59:39] *** oliveira1 has joined ##OpenGL[02:02:10] <yates> can you OR glEnableClientState() arguments together and just make one call?[02:02:44] <MatthiasM> no[02:02:50] <MatthiasM> you can't OR enums[02:02:57] <MatthiasM> only things that end in _BIT[02:04:35] <paul424> MatthiasM: hmm you aren't english native, how do you get with all this technical jargon ?[02:05:03] <MatthiasM> paul424: well - you have to start[02:06:08] <MatthiasM> *spelling[02:12:49] *** oliveira_ has quit IRC[02:14:00] <paul424> yes, but I mean something diffrent for example http://www.vterrain.org/LOD/Papers/index.html like those articles are too difficult not because of my programming/math background but the terms their operate[02:15:39] <MatthiasM> use a dictionary[02:17:11] <paul424> :D[02:17:18] <yates> do the various array pointers (glVertexPointer(), etc.) remain in effect across glDrawArrays() calls, or do you have to set them for each call?[02:17:36] *** MatthiasM has quit IRC[02:17:42] *** MatthiasM has joined ##opengl[02:19:02] <MatthiasM> yates: every state is keeped - except when noted otherwise in the spec[02:19:22] *** hibread has quit IRC[02:19:43] *** hibread has joined ##opengl[02:22:19] <yates> does glDrawArrays() have to be embedded in a glBegin/glEnd?[02:23:09] <MatthiasM> NO![02:23:22] <MatthiasM> yates: http://www.opengl.org/sdk/ <-- READ IT![02:23:24] <yates> well, they aren't working[02:25:36] *** johndoe has quit IRC[02:26:31] *** cs_student has joined ##OpenGL[02:27:02] <yates> if GL_TEXTURE_COORD_ARRAY was enabled, does the glDrawArrays() get the number of dimensions from the preceding glBindTexture() dimensions?[02:28:36] *** _JSC415 has quit IRC[02:29:53] <MatthiasM> yates: did you even read the docu ?!?[02:30:13] <yates> MatthiasM: fuck off.[02:30:18] <yates> hell yes[02:30:24] <yates> i've read three fuckinjg documents[02:30:44] <MatthiasM> then you should see where the dimension is specified[02:30:57] <yates> no i don't[02:31:14] <MatthiasM> then read the reference page for glTexCoordPointer again[02:31:18] <yates> i don't even see an example of a vertex arrays with texture coordinates[02:32:03] <yates> why this bullshit? why not just refer me to the function instead of acting like i'm an idiot?[02:32:38] <MatthiasM> I already gave you all the link - you just need 1 click[02:32:58] <MatthiasM> I'm not gona read it for you[02:33:06] *** cplusplus2 has joined ##OpenGL[02:33:20] <yates> again, the entire sum of the world's knowledge is "just 1 click"[02:33:24] <yates> just read, right?[02:33:27] <yates> bullshit[02:33:34] <yates> sometimes reading need interpration[02:33:49] <yates> and a little direction[02:33:49] *** cs_student is now known as poseidon[02:34:12] <MatthiasM> yates: <MatthiasM> then read the reference page for glTexCoordPointer again[02:34:26] *** jcazevedo has quit IRC[02:34:41] <kbotnen> whats the anger about?[02:35:11] *** Xmas| has quit IRC[02:35:22] <MatthiasM> asking tons of simple questions which are all stated in one easily to acess documentation[02:36:17] <yates> simple? bs[02:36:37] <yates> easy-to-read? again, bs[02:37:39] <MatthiasM> yates: well - if you can't work with the reference pages in the OpenGL SDK - then I wonder what you will do when you get to other libs[02:37:57] <yates> AFM - wife has goodies![02:38:58] *** calav3ra has quit IRC[02:39:35] *** amz has quit IRC[02:42:30] <kbotnen> guess its about learning how to interpret. the sdk is fairly consise if you know how to interpret. but I find it strange the hostility yates.[02:43:49] <kbotnen> but I tried to google one of your guestions: "does glDrawArrays() have to be embedded in a glBegin/glEnd?" you get a lot of hint with that too.[02:43:55] <MatthiasM> well - maybe he asked too much at the wrong time :/[02:46:26] <RTFM_FTW> this makes *no sense* and if the OP had actually bothered to read through the API specification he / she would have realized that :P[02:46:43] <RTFM_FTW> as per usual RTFM[02:46:46] <tmccrary> RTFM_FTW: fuck off, he clearly needs a simple english version of the documentations[02:47:00] <tmccrary> like those menus they have at macdonalds for illiterates[02:47:02] <RTFM_FTW> oh of course... how could I be so stupid :P[02:47:15] <MatthiasM> heh[02:47:27] <MatthiasM> "OpenGL for runaways"[02:49:25] *** cplusplus has quit IRC[02:54:23] *** Walt has quit IRC[02:55:09] <boghog> my head hurts[02:55:26] <yates> are you guys done mentally jerking off?[02:55:35] <RTFM_FTW> never[02:55:45] <yates> ha[02:55:50] <kbotnen> what have you donw with it lately?[02:55:52] * boghog erases his GLX code and starts reading glfw documentation[02:55:54] <kbotnen> *done[02:56:12] <MatthiasM> boghog: good idea[02:56:22] <yates> oh, i get it - it's a grammar contest.[02:58:17] <yates> and i realized my error while i was eating my cookies.[02:58:19] <yates> so there[02:58:24] <yates> pfffft![02:59:08] <yates> seriously, anyone who thinks this is a simple matter needs to get out from in front of the computer more often.[02:59:16] <yates> yes, the docs are fairly well-written[02:59:35] <RTFM_FTW> since when does development imply "simple"?[02:59:37] *** iion has quit IRC[02:59:50] <MatthiasM> well - reading manuals and data sheets is part of my job[03:00:03] <yates> mine too[03:00:16] <RTFM_FTW> fuck you are in the wrong business (or doing the wrong thing for fun) if you want that to be the case :P[03:00:24] <MatthiasM> and most of the time it's easier to collect the information yourself then asking for it[03:00:32] <RTFM_FTW> exactly[03:00:38] <yates> most of the time, yes[03:00:49] <yates> do i ask you about every fucking thing in opengl?[03:00:55] <tmccrary> yes, you really fucking do[03:01:03] <RTFM_FTW> haha[03:01:05] <tmccrary> I've never seen anyone ask for such easily gotten details[03:01:08] <yates> ha ha[03:01:25] <yates> oh my[03:01:34] <tmccrary> like "could someone download the openglred book and cat it into the irc channel so I can read it"[03:01:38] <tmccrary> seriously[03:02:06] <yates> tmccrary: i'm surprised you got off your mother long enough to write that long of a sentence.[03:02:24] <tmccrary> I don't even know what that means[03:02:54] <MatthiasM> yates: you should keep your mouth in check[03:03:10] <RTFM_FTW> meh like that ever stopped anyone else in here :P[03:03:14] <RTFM_FTW> heheh[03:03:38] <kbotnen> so yates . what app are you working on now? game or another kind of app?[03:04:00] <yates> kbotnen: another kind - a scientific visualization tool[03:04:15] *** cplusplus2 has quit IRC[03:04:18] <yates> a dynamic waterfall spectrum analyzer[03:04:20] <kbotnen> oo.. cool :)[03:04:35] <kbotnen> what can you do with a dynamic waterfall spectrum analyzer?[03:05:03] <yates> kbotnen: do you know what a waterfall plot is?[03:05:20] <yates> magnitude vs. time vs. frequency[03:05:31] <yates> but a plot is a static, one-shot image[03:05:57] <kbotnen> never heard about it, but there was a wikipedia article about it :)[03:06:03] <yates> the plan here is to run a file, say, a .wav file, through and analyze it while you're listening.[03:07:01] <yates> the higher-level plan is to develop an algorithm that can separate respiratory and heart rate algorithms and estimate heart rate and breathing rate[03:07:16] <yates> s/heart rate algorithms/heart rate signals/[03:07:42] <yates> and please don't tell me i'm reinventing the wheel...[03:07:55] <tmccrary> you're reinventing the wheel[03:07:58] <kbotnen> hehe.. I dont know. You lost me at hello :)[03:08:01] <tmccrary> ;)[03:08:20] <kbotnen> but sounds like you have something to puzzle with for quite a few days :)[03:08:27] *** HuntsMan has joined ##opengl[03:08:37] <yates> yup[03:09:26] <yates> kbotnen: see here: http://en.wikipedia.org/wiki/File:Spectrogram.png[03:10:27] <yates> and imagine that plot brought to life, moving as time passes.[03:12:07] <kbotnen> I can imagine it yes. But I would prefer a good movie I think :)[03:13:38] <yates> for that we have netflix...[03:13:55] <yates> currently on the DVD player: "What Dreams May Come"[03:26:09] *** davidc__ has quit IRC[03:28:32] *** HuntsMan has quit IRC[03:32:47] *** iion has joined ##OpenGL[03:36:16] *** MatthiasM has quit IRC[03:36:22] *** MatthiasM has joined ##opengl[03:41:39] *** b0000 has quit IRC[03:44:22] *** HuntsMan has joined ##opengl[03:47:07] <yates> kbotnen: what are you working on?[03:48:20] *** qeed has quit IRC[03:49:15] <kbotnen> Im not quite sure, hehe. A app that let the user display models (3ds obj whatever) and navigate through them. I know it can easily be done with things such as jme but I want to learn openGL on the way. And if I manage it I will extend the app to other uses. So its placed in the archviz section.[03:51:21] *** jfroy|work has quit IRC[03:58:23] *** poseidon has quit IRC[03:59:18] <yates> opengl is cool[03:59:54] <kbotnen> I hope so :)[04:00:30] <kbotnen> I previously come from Director and Shockwave 3D programming. So this is something reeally different :)[04:01:57] <yates> little lower level?[04:02:31] <kbotnen> openGL is lower yes.[04:02:33] <vade> oh man I missed irc flames! >:-O[04:02:35] <vade> argh[04:02:51] <kbotnen> hehe.. there will always be a new one.[04:02:56] <kbotnen> coming to a # near you.[04:06:17] <vade> :P[04:08:14] <yates> is anyone here running linux on a relatively new gpu?[04:08:24] *** mattn2|home__ has joined ##OpenGL[04:13:56] *** PhrostByte has joined ##opengl[04:14:07] *** dv_ has quit IRC[04:17:22] *** djork has joined ##opengl[04:20:06] *** jfroy has joined ##OpenGL[04:23:50] *** mattn2|home_ has quit IRC[04:30:24] <yates> according to here, http://www.ogre3d.org/forums/viewtopic.php?t=40050&highlight=&sid=ce193664e1d3d7c4af509e6f4e2718c6[04:31:50] *** phrosty has quit IRC[04:35:33] <yates> b000: do you still have that vsync code around?[04:37:22] <boghog> yates, running linux with a radonehd4850[04:39:51] <yates> boghog: do you know how to turn vsync off ?[04:40:45] <boghog> hrm, I'm afraid I don't, I think there are GLX extensions that can deal with vsync[04:42:15] *** fargiola` has joined ##OpenGL[04:43:42] <yates> i'm not sure it can be done under glx[04:43:51] <yates> i've been searching and haven't found it.[04:43:59] <boghog> my ATI drivers support GLX_MESA_swap_control, GLX_OML_swap_method, GLX_OML_sync_control, GLX_SGI_video_sync. i guess one of those can do it[04:44:04] *** UUncia has joined ##OpenGL[04:44:39] <yates> where is the documentation for those extension functions?[04:44:46] <yates> i also have GLX_SGI_video_sync[04:44:48] <yates> http://pastebin.com/m1493803f[04:44:57] <yates> it's not in the glx manual[04:45:23] <boghog> http://www.opengl.org/registry/[04:45:40] <yates> nice! thanks.[04:47:05] *** GinoMan has quit IRC[04:47:18] *** GinoMan has joined ##OpenGL[04:48:08] *** GinoMan has quit IRC[04:48:28] *** GinoMan has joined ##OpenGL[04:51:36] <yates> given immediate mode code with vertex arrays is too slow, what's the next, least complicated thing to try? display lists are not valid since the model data is dynamic.[04:53:59] <yates> no lighting is required, but a 1d texture map is[04:54:47] *** fargiolas has quit IRC[04:56:56] <yates> RTFM_FTW: surely you haven't gone to bed yet - it's only 8pm in Nicasio ...[04:58:34] *** Kasu has quit IRC[05:02:02] <yates> boghog: do you know of a 3d graphics performance benchmark you can run quickly to get a rough idea of your system's performance? and one i would have avialable as well? i'm running fedora 8 with a GeForce FX 5200[05:02:21] <yates> i'm thinking that my system just isn't all that powerful.[05:03:13] <yates> i'm trying to push through about 100,000 vertices per frame and at only 500x500 i'm only getting 25 fps with cpu maxed out.[05:04:04] <boghog> hrm, i don't really know of any benchmark, I guess you could try some opensource quake3-based game (warsow, nexuiz) and record a demo and play it back with the timedemo command[05:04:20] <replaced> you are using linux[05:04:28] <yates> yes[05:04:31] <replaced> waht do you expect :p[05:04:50] <yates> i expect performance better than windows given the same hardware :)[05:05:17] <yates> actually it's about 190,000 vertices per frame.[05:05:58] <yates> i think the GeForce FX 5200 is somewhat dated - is the perhaps par performance for that technology?[05:06:14] <yates> s/is the/is this/[05:07:24] <yates> even with GL_CULL_FACE enabled, it still maxes out before 30fps (my goal)[05:07:51] <yates> replaced: what kind of performance do you get and what hardware are you using?[05:08:56] <replaced> man, my fastest card is an r200[05:08:56] <replaced> :D[05:09:35] <yates> has anyone used the new SLI technology?[05:10:05] <HuntsMan> new?[05:10:21] <yates> oh, sorry, it's older than 3 months...[05:10:23] <replaced> its like 10 years old[05:10:27] <yates> is it?[05:10:37] <HuntsMan> from NVIDIA it's at least 3 years old[05:10:44] <HuntsMan> from 3Dfx, 10 years old[05:10:49] <replaced> prehistoric voodoo card had it[05:11:07] <yates> oh[05:11:25] <RTFM_FTW> multi-GPU rendering is old period... hell ATI had a SLI based Rage 128 board back in the day :)[05:11:29] <replaced> oh i miss my voodoo3[05:11:42] <yates> (as the news lady on SNL used to say, "never mind")[05:12:06] <yates> so what the hey - is my performance par for my hardware?[05:12:39] <replaced> tell us a q3 fps :D[05:12:52] <yates> q3?[05:13:01] <yates> quake ?[05:13:06] <yates> sorry, i'm not a gamer[05:13:19] *** rnx has quit IRC[05:13:51] <yates> ok, wow - it's in the fedora rpos - loading now...[05:14:19] *** rnx has joined ##opengl[05:14:33] <replaced> you dont have to be a gamer to play quake[05:14:54] <yates> you need to install Quake III Arena in order to pla[05:15:04] <yates> is openarena good enough?[05:15:08] <boghog> yates, try warsow or nexuiz[05:15:14] <boghog> or openarena :D[05:15:25] <replaced> or tremulous[05:15:26] <boghog> or urban terror[05:15:28] <boghog> hehe[05:15:31] <replaced> or that[05:15:42] <replaced> hmm ishould play one of those[05:16:00] <boghog> used to play urbanterror a bit[05:16:09] <boghog> but meh, im too old of quirky fps games[05:16:15] <replaced> yeah, hehe[05:16:19] <replaced> we're getting old[05:16:20] <replaced> :)[05:16:20] <boghog> not good at it and it makes my hand hurt :D[05:18:16] <yates> how do you get the frame rate to display in openarena?[05:18:51] <boghog> hrm, not sure if I rememebr correctly, try com_drawfps 1[05:18:59] <boghog> in the console (open with tilde)[05:19:11] <boghog> or was it ctrl+alt+tilde[05:20:18] *** braden has joined ##OpenGL[05:20:26] <replaced> cg_drawfps 1[05:20:36] <replaced> and youmay have to turn off the limiter[05:21:28] <replaced> and i can tell you how does my junk hw perform in an hour[05:22:47] <yates> how do you get out of the console?[05:22:53] <yates> quit exits the whole game[05:23:01] <yates> is there a help for the console?[05:23:07] <yates> help / ? didn't work[05:25:13] <boghog> press the same key that got you into the console again[05:25:21] <boghog> its a toggle thing[05:26:20] <yates> i'm getting about 90 fps[05:26:37] <yates> in 640x480 mode / fullscreen[05:29:18] <replaced> huh[05:29:56] <replaced> /set com_maxfps 0[05:29:56] <replaced> ?[05:30:41] <yates> why?[05:32:12] <kbotnen> to remove any fps maxrate constraint maybe?[05:34:00] <yates> now it's somewhere between the high 200's and 360 fps or so[05:34:30] <replaced> that looks ok[05:34:56] <replaced> i dont thin its underpowered[05:35:03] <replaced> maybe with todays standards[05:35:29] <replaced> but who cares about today? :)[05:35:36] *** iion has joined ##OpenGL[05:37:16] <yates> ok, thanks[05:37:50] <sparky> evenin[05:38:41] <yates> i'm not sure how this game compares with my app's complexity: should i be getting way more than 25 fps at 64000 triangle strips per frame?[05:42:24] *** PhrostByte is now known as phrosty[05:42:25] <yates> is this perhaps just about as far as I can take things in immediate mode code?[05:43:25] <yates> would switching to GLSL get me a lot of performance increase ?[05:43:40] <yates> one important point i've neglected to mention[05:43:52] <yates> my frame rate depends heavily on the object orientation.[05:44:20] <yates> when i'm looking at the objecte from the side, it smes to take the most cpu[05:44:53] <yates> turning it so i see the top and two sides makes it a good bit less cpu intensive - down to 14 percent or so from 100 percent[05:46:37] <kbotnen> I asked for this yesterday, but still havent manage to sort things out so Ill try again. When in ortho mode, what are the coordinates for upper left and lower right for my Display window?[05:47:28] <yates> the code is here: http://galois.digitalsignallabs.com/wxwaterfall[05:47:45] <kbotnen> Im entering ortho with this line: GL11.glOrtho(0, 640, 480, 0, -1, 1); and want to draw a frame around (inside) my window.[05:47:53] <kbotnen> an innerframe if you want.[05:50:01] *** Quentarez has joined ##OpenGL[05:50:18] <yates> kbotnen: get the OpenGL Superbible. excellent reference resource[05:50:48] <yates> tmccrary: hey, can we kiss and make up now?[05:50:56] <kbotnen> When I get money I will purchase the red book and other good literature. but that is not now.[05:51:05] <yates> i hear ya'[05:51:35] <replaced> heh[05:51:39] <kbotnen> I have read through relevant chapters in the red book now, but It seems that something is not quite correct with my way of doing it. my frame get drawn in the middle of the appwindow when using 0,0[05:52:02] *** mikolalysenko has quit IRC[05:53:32] <yates> kbotnen: sorry, i use perspective mode[05:54:02] <replaced> night[05:55:22] <yates> sleep tight[05:56:20] <kbotnen> night.[05:57:09] <kbotnen> I found it. at last. now I only have to find a way to make my texture (512image) fit to a quad of 640x480.[05:57:23] <kbotnen> :)[05:58:08] <kbotnen> one step further. This is going to take a while, but learning on the way :)[05:59:45] *** wisey has joined ##OpenGL[06:00:53] <yates> goog[06:00:57] <yates> good[06:01:04] *** tokemonstah has quit IRC[06:02:08] *** djork has quit IRC[06:07:07] *** iion has quit IRC[06:07:08] <kbotnen> but now Im off. see you guys later.[06:07:18] *** edwardk has joined ##OpenGL[06:09:12] *** neoneye has joined ##OpenGL[06:12:29] *** kbotnen has quit IRC[06:12:49] *** paul424 has quit IRC[06:19:07] <yates> RTFM_FTW: Ever go to Big Basin State Redwood Park? Stop at Skyline / Hwy 9 and get a hot dog?[06:22:39] <RTFM_FTW> yates I'm not anywhere near CA :D[06:22:46] <RTFM_FTW> wrong time zone :P[06:25:35] <yates> ah[06:28:31] <yates> is the only alternative left practically to use the glsl pipeline?[06:30:53] <yates> the real question is, even if i optimized with glsl, would i realize a significant performance benefit?[06:31:57] <yates> are extensions just program objects?[06:34:53] *** yates has quit IRC[06:44:00] *** oliveira_ has joined ##OpenGL[06:44:45] *** ectropy has joined ##OpenGL[06:46:06] *** oliveira1 has quit IRC[06:51:17] *** wisey has quit IRC[07:23:37] *** TheLorax has joined ##opengl[07:42:29] * sparky slaps synchris around a bit with a large trout[07:50:57] *** rnx has left ##opengl[07:54:10] *** korff_home has joined ##OpenGL[07:56:37] *** pragma__ has joined ##opengl[07:58:46] *** pragma_ has quit IRC[07:59:04] *** pragma__ is now known as pragma_[08:17:06] *** ectropy has quit IRC[08:20:30] *** GinoMan has quit IRC[08:25:15] *** UUncia has quit IRC[08:30:30] *** scai has joined ##opengl[08:36:36] *** edwardk has left ##OpenGL[08:39:46] *** LiraNuna has quit IRC[08:39:49] *** LiraNuna has joined ##OpenGL[08:47:13] *** scai has left ##opengl[08:55:02] *** TheLorax has quit IRC[08:55:03] <exDM69> My FreeType reader/wrapper is getting together, I can't wait to plug it together with my stencil buffer polygon renderer[08:55:45] *** avacore has quit IRC[08:56:11] *** avacore has joined ##OpenGL[09:02:19] *** UUncia has joined ##OpenGL[09:03:51] *** Bollinger has joined ##OpenGL[09:08:05] *** Bollinger has quit IRC[09:08:27] *** kenws has joined ##OpenGL[09:09:15] *** dvoid_ has joined ##OpenGL[09:09:56] *** calav3ra has joined ##opengl[09:10:12] *** Bollinger has joined ##OpenGL[09:11:14] *** [AD]Turbo has joined ##OpenGL[09:13:33] *** fargiolas has joined ##OpenGL[09:13:38] *** jezek2 has quit IRC[09:13:38] *** replaced has quit IRC[09:13:38] *** pgas has quit IRC[09:13:38] *** arkx has quit IRC[09:13:38] *** cmang_ has quit IRC[09:13:39] *** stef has quit IRC[09:13:49] *** replaced has joined ##OpenGL[09:13:49] *** cmang_ has joined ##OpenGL[09:13:49] *** pgas has joined ##OpenGL[09:13:49] *** jezek2 has joined ##OpenGL[09:13:49] *** stef has joined ##OpenGL[09:13:49] *** arkx has joined ##OpenGL[09:14:37] <[AD]Turbo> yo[09:14:50] *** Adrinael has quit IRC[09:15:16] *** pwned has quit IRC[09:15:31] *** pwned has joined ##opengl[09:15:55] *** fargiola` has quit IRC[09:21:49] *** groton has joined ##OpenGL[09:21:50] *** Adrinael has joined ##opengl[09:27:12] *** eXtronuS has joined ##OpenGL[09:34:44] *** Yustme has joined ##OpenGL[09:41:43] *** Renderwahn has joined ##OpenGL[09:50:19] *** _JSC415 has joined ##OpenGL[09:50:27] *** _JSC415 is now known as _THEGOD[09:50:38] <_THEGOD> hi...[09:50:48] <_THEGOD> i compiled tiokamak src.[09:51:02] <_THEGOD> but i see a missing /undefined referance.[09:51:22] <_THEGOD> which i guess points to one module that werent compiled.[09:51:41] <_THEGOD> can i find somewhere libtokamak.a ?[09:51:59] <_THEGOD> like you can guess. i cant find them using google.[09:52:09] <_THEGOD> hello ?[09:52:24] <_THEGOD> can anyone reading those words ?[09:53:38] <_THEGOD> MatthiasM ?[09:53:56] <_THEGOD> MatthiasM youre gone ? tats rydiculous .[09:54:06] *** uster has joined ##OpenGL[09:54:27] *** predaeus has joined ##opengl[09:54:38] * MatthiasM kicks _THEGOD[09:55:15] <predaeus> :-D[09:55:45] <_THEGOD> lols.[09:56:04] <_THEGOD> how in the hell you are allmost allways awake and there ?[09:56:33] <_THEGOD> do you hawe a bell ringing when a new messae apperas there ?[09:57:15] <_THEGOD> hm.. ok mirc has it .[09:57:23] <_THEGOD> im not using mrc andyway ;P[09:57:36] <_THEGOD> so i compiled tikamak.[09:57:42] <MatthiasM> STFU[09:57:51] <_THEGOD> but there is some other modules which are on other folders.[09:57:51] <MatthiasM> this is an OpenGL help chennel[09:58:12] <MatthiasM> not a "I'm too stupid to use tools"[09:58:14] <_THEGOD> tokamak is a opengl 3d engine ?[09:58:21] <MatthiasM> and ?[09:58:31] <MatthiasM> then ask the authors[09:58:41] <_THEGOD> authors ?[09:58:43] <_THEGOD> where they are ?[09:59:04] <_THEGOD> are they allways awake ,like you ?[09:59:34] *** marenz_ has joined ##OpenGL[10:00:55] <_THEGOD> whatewer ill try mixing all necessary modules on on folder.[10:01:10] <_THEGOD> or use -I"" for all folder.[10:01:17] <_THEGOD> lets see if it works. ;P[10:01:42] <MatthiasM> you are just too stupid to use a computer or book[10:02:00] <_THEGOD> lols. and the best thing about opensource is. i guess[10:02:15] <_THEGOD> you dont ewen need to compile a libx.a[10:02:25] <_THEGOD> you can use sourcecode directly ;P[10:03:05] <_THEGOD> i heard somewhere also i can use dlls as includes ...[10:03:10] <_THEGOD> lols amszing[10:03:19] <_THEGOD> i tryed it last day.[10:03:23] <_THEGOD> on one sample[10:03:26] <_THEGOD> it compiles[10:03:27] *** LtJax has joined ##opengl[10:03:36] <_THEGOD> but didnt worked.[10:03:47] <_THEGOD> it reported some king of runtime errors.[10:04:32] <_THEGOD> anyway its still good to know that gcc knows dlls and can integrate with them :P[10:04:33] <MatthiasM> "Ciritical Error: User is missing brain"[10:05:38] <_THEGOD> well. ill try compiling one of the samples by collecting all modules . ;P[10:06:13] <_THEGOD> last time i tryed this i realised there some other modules which needs to be imported.[10:06:20] <_THEGOD> but after 7-8th one .[10:06:26] <_THEGOD> i gave up[10:06:37] <_THEGOD> well. may be i can doo it today.[10:09:13] *** sohail has quit IRC[10:10:38] *** Nescafe has joined ##OpenGL[10:12:42] *** synchris_m6300 has joined ##opengl[10:14:05] *** synchris has quit IRC[10:14:29] *** synchris_m6300 is now known as synchris[10:21:23] <_THEGOD> well.. anyway.[10:21:37] <_THEGOD> greetings to all.[10:21:49] <_THEGOD> the god knows anytinh.[10:21:58] <_THEGOD> including basics of cc[10:22:06] <_THEGOD> and ewen more ;p[10:22:37] <_THEGOD> i can read ur brain. ;P[10:22:41] <_THEGOD> lols.[10:23:34] <_THEGOD> i hawe to ask why this were too hard to answer ?[10:23:53] <_THEGOD> what all u guys were hawe to write is -w[10:24:33] <_THEGOD> but u uys wrote dozens of how to ask liks.[10:25:01] <_THEGOD> u gus hawe to find a site about how to answer questions . ;P[10:25:43] <_THEGOD> u gyus knows asking ,you guess NOT ME ! UR ANYONEELSE ![10:26:37] <_THEGOD> but U gus hawe to hawe lear hUw to unswe QuestiUns.[10:33:01] *** uster has quit IRC[10:34:21] *** JernejL_ has joined ##OpenGL[10:36:26] *** hackkitten has quit IRC[10:42:33] *** s_ariga has joined ##OpenGL[10:44:49] <_THEGOD> i was idling.[10:45:16] <_THEGOD> is there a tool that remoces extra directory entys from #include definitios[10:45:22] <_THEGOD> thats waste of time.[10:45:40] <_THEGOD> eg #include "folder1/folder2/folder3/gl.h"[10:46:00] <_THEGOD> so i hawe to define spesific include path dor all headers sometimes.[10:46:08] <_THEGOD> or remove those definitions[10:46:12] <_THEGOD> damn[10:46:17] <_THEGOD> i hate this[10:47:50] <_THEGOD> i dont need tokamak as resource.[10:48:01] <_THEGOD> i need it for making things faster ;P[10:48:21] <_THEGOD> or just for refusing to write some codes ;P[10:48:42] <_THEGOD> i realised that writing those codes alone will take w bit longer.[10:48:53] <_THEGOD> or i need a killer robot .[10:49:01] <_THEGOD> who will write those for me.[10:52:43] *** Jernej has quit IRC[11:00:49] *** dizekat has joined ##OpenGL[11:01:25] <dizekat> i have weird problem with FBO....[11:01:32] *** geocalc has quit IRC[11:01:38] <dizekat> rendering to FBO, then drawing textured quad[11:01:50] <dizekat> and about every 100th frame[11:02:11] <dizekat> takes like 10x longer than other frames[11:02:22] <dizekat> which looks very jerky[11:02:36] <dizekat> has anyone encountered this?[11:02:40] <_THEGOD> ne_luawrapper.h friend declaration requires class-key, i.e. `friend class CLuaState' ?[11:03:33] <_THEGOD> may be ur FBO device is software provided FBO defice not VGA mode fbo.[11:04:13] <dizekat> nvidia 8800 GT here[11:04:20] <dizekat> should support all that[11:04:38] <dizekat> its not that its slow[11:04:45] <dizekat> its that every about 100th frame or so[11:04:46] <dizekat> is slow[11:04:50] <_THEGOD> i dont know . may be its drivers fault.[11:04:54] <dizekat> yes[11:04:58] <dizekat> sounds like that actually[11:05:00] <predaeus> dizekat, does it all fit in video ram? might have to swap to RAM from time to time[11:05:03] <dizekat> but i do need workaround :/[11:05:14] <dizekat> i'm pretty sure i'm not leaking anything[11:05:15] <_THEGOD> afaik. my old 6600 too were using sometimes cpu too much.[11:05:24] <dizekat> and geometry stays same...[11:05:46] <dizekat> it works fine when i render to screen not fbo[11:05:52] <_THEGOD> well. anyway .[11:06:04] <_THEGOD> if you update fbo wery common.[11:06:08] <_THEGOD> too much.[11:06:18] <_THEGOD> its not much different then software rendering afaik.[11:06:31] <dizekat> well it IS different[11:06:35] <dizekat> it does render quick[11:06:43] <dizekat> 99 frames out of 100 thats it[11:07:05] <dizekat> i'm totally sure its not having software fallback on majority of frames....[11:07:06] <_THEGOD> it depends on how much work is there for cpu.[11:07:11] <MatthiasM> dizekat: can you paste a test case ?[11:07:16] <_THEGOD> eg prepairing fbo is work of cpu[11:07:19] <_THEGOD> not gpu[11:07:25] <dizekat> too big code actually here[11:07:31] <dizekat> didnt make small testcase yet[11:07:37] <MatthiasM> dizekat: make a reduced test case[11:07:44] <dizekat> yea[11:07:57] <dizekat> thats what i'm going to eventually do, if something, for reporting to nvidia[11:08:06] <MatthiasM> what language do you use `?[11:09:07] *** tokemonstah has joined ##OpenGL[11:09:31] *** Jernej has joined ##OpenGL[11:10:07] <dizekat> c++[11:10:29] <predaeus> dizekat, make sure you measure time the correct way too, unless you also see the difference, measuring could be wrong.[11:10:50] <MatthiasM> don't use getTickCount on windows[11:11:47] <dizekat> using linux here...[11:11:48] <dizekat> actually[11:12:02] <MatthiasM> gettimeofday ?[11:12:02] <dizekat> it seems something gone screwy otherwise[11:12:11] <dizekat> testing with other programs now[11:12:17] <dizekat> maybe driver just leaked something again[11:12:26] <dizekat> yes, gettimeofday...[11:12:29] <dizekat> its quite good[11:12:39] <MatthiasM> 1us precision on linux[11:15:15] *** geocalc has joined ##opengl[11:15:47] <_THEGOD> can anyone know why this aerror might happen on a opensource library ?ne_luawrapper.h friend declaration requires class-key, i.e. `friend class CLuaState'[11:16:04] <_THEGOD> lola.[11:16:11] <_THEGOD> its called luavrapper ?[11:16:14] <_THEGOD> hm..[11:16:20] <_THEGOD> does this means i need lua ?[11:16:35] <_THEGOD> may be i need some lua moduler ?[11:16:40] <_THEGOD> s[11:16:55] <_THEGOD> damn.[11:17:23] <_THEGOD> i hate those spider web style dependencies of opensource tools ;P[11:17:43] <_THEGOD> i wonder where itll end.[11:18:18] <_THEGOD> also i hawe this[11:18:21] <_THEGOD> 36 C:\Dev-C++\DOWNLOADED LIBRARIES\tokamak_sample_1_2\samples\sample2\sample2 .cpp could not convert `CLuaValue(0)' to `CLuaValue&'[11:18:37] <_THEGOD> what that migh be meaning ?[11:18:46] <_THEGOD> `CLuaValue(0)' to `CLuaValue&' ?[11:18:53] <predaeus> _THEGOD, please stop spamming the channel with nonsense, some people try to learn from others here or ask/answer questions about OpenGL.[11:19:28] <_THEGOD> try learning how to ask questions ;P[11:19:48] <_THEGOD> look there-->>http://mikeash.com/getting_answers.html[11:19:52] <MatthiasM> _THEGOD: this channel is for questions regarding the OpenGL API ![11:20:01] <MatthiasM> _THEGOD: not for questions about some library[11:20:13] <MatthiasM> _THEGOD: or on how to use a compiler[11:20:21] <_THEGOD> you mean only opengl library ?[11:20:32] <MatthiasM> _THEGOD: and you should READ and UNDERSTAND this link you just posted[11:20:52] <MatthiasM> _THEGOD: go and play with sand and water[11:21:02] <_THEGOD> but tokamak uses opengl library as part of its core.[11:21:30] <_THEGOD> anyway couldnt it be easy to make U guys learn how to answer correctly ?[11:21:40] <_THEGOD> ;P[11:21:43] *** belou has joined ##OpenGL[11:22:15] <_THEGOD> and let anyone know the answeres. instead how to ask questons ?[11:22:58] <_THEGOD> and letting peoples to find what they searching for only by clicking im not felling luck icon on google ;P[11:23:02] <dizekat> hmm i think its either problem with my code or i need a reboot[11:23:16] <dizekat> jerky fps now without fbo[11:23:18] <dizekat> lol[11:23:22] <MatthiasM> dizekat: try a reboot[11:23:26] <dizekat> yea[11:23:33] <dizekat> drivers leak like sieve :/[11:23:34] *** LtJax has quit IRC[11:25:27] *** dizekat has quit IRC[11:25:38] <_THEGOD> do anyone knows what the hell friend means ?[11:26:29] *** k3t^ is now known as k3t[11:26:52] <MatthiasM> _THEGOD: you need to stop using computers - because your are incompatible with them[11:27:02] *** JernejL_ has quit IRC[11:27:04] *** sohail has joined ##OpenGL[11:27:23] <_THEGOD> yeah but i need them for creating compatible computers with me ;P[11:27:41] <MatthiasM> that's impossible[11:27:54] <MatthiasM> you can't lower the IQ of a computer this far[11:28:51] <_THEGOD> like nigger nike says on baseball shoe advertising , notinh is impossible ;P[11:29:37] <_THEGOD> its impossible for a idiot to controll 100million+ peoples[11:29:43] <_THEGOD> but like we all see.[11:29:48] <_THEGOD> it is ;P[11:30:27] <_THEGOD> so we end at the starting point. N.I.I.P.[11:30:46] <_THEGOD> NIIP=Nothing Is Impossible ?[11:30:50] <_THEGOD> lols.[11:30:59] <_THEGOD> where those p come from ;P[11:31:26] <_THEGOD> well. like i allwasys say. sometimes is anyway ;P[11:32:05] *** Ingenu has joined ##OpenGL[11:33:15] *** dizekat has joined ##OpenGL[11:33:21] *** s_ariga has left ##OpenGL[11:34:07] *** Osirus has joined ##OpenGL[11:34:19] *** marenz_ has quit IRC[11:34:32] *** Osirus has quit IRC[11:36:40] <ville> MatthiasM: Next time try sending something that begins with either ville or Adrinael to the channel. Guess he hasn't been kicked out of here enough times.[11:37:16] <MatthiasM> good idea :)[11:37:44] <dizekat> i figured it out[11:37:52] <dizekat> when my program is using very little cpu[11:37:59] <dizekat> which i made it do for test sake[11:38:04] <dizekat> it jerks MORE[11:38:17] <dizekat> could be related to me having quadcore cpu[11:38:54] <dizekat> btw speaking of moderation, i sometimes had idea to name myself "fuck" as moderator[11:38:59] <dizekat> so that i get ring[11:39:59] <ville> at least my irc clients lets me specify whatever words to act on such as highlight it or beep[11:40:10] <dizekat> well[11:40:11] <ville> s/clients/client/[11:40:19] <dizekat> the idea was that it'd be obvious[11:40:25] <dizekat> like more obvious[11:41:41] <ville> harder to catch automatically when someone suffers for verbal diarrhea and makes no, profanity is easy[11:41:50] <dizekat> well yes[11:42:03] <dizekat> i knew a mod that only responded to profanity[11:42:34] <dizekat> he got deopd or something later on because it was game channel[11:42:46] <dizekat> where profanity's officially kinda ok[11:42:56] <dizekat> anyways[11:43:22] <dizekat> anyone has seen program fps jerk more when program uses less cpu time?[11:43:51] <predaeus> dizekat, how do you limit cpu time? nice, or change clock speed or how?[11:44:00] <ville> vsync issue?[11:44:23] <dizekat> i have no cpu frequency scaling at the moment[11:44:36] <dizekat> program just draws stuff on screen while doing almost no calculations[11:44:46] <dizekat> so that almost no cpu time is spent in program itself[11:44:48] *** cplusplus has joined ##OpenGL[11:45:13] <ville> or possibly you're aliasing but in fact drawing stuff faster?[11:45:28] <dizekat> doesnt look like fps aliasing[11:45:37] <dizekat> jerks a ton more visually[11:45:49] <dizekat> and when i write out seconds per frame to console[11:45:57] <dizekat> some frames take 10x longer than other[11:46:08] <dizekat> maybe it is switching cpu core or something?[11:46:21] <dizekat> like process goes to another core[11:47:04] <ville> would figure the OS be smarter than that, tell it to use one core?[11:47:44] <dizekat> yea[11:47:52] <dizekat> well i know windoze isnt[11:48:13] <dizekat> common advice to tell to set core affinity of process in taskmanager...[11:48:44] <dizekat> i actually have no idea how to set process core affinity on linux[11:49:26] <dizekat> well with luck wont matter because final ver of that program will be calculating more on each frame[12:12:06] *** Eforen-atSchool is now known as Eforen-SleepTime[12:13:35] *** sohail has quit IRC[12:18:15] *** _THEGOD has quit IRC[12:23:19] *** RTFM_FTW has quit IRC[12:23:34] *** RTFM_FTW has joined ##OpenGL[12:24:00] *** Gorgoroth has joined ##OpenGL[12:34:40] *** HugoMaxx_ has joined ##OpenGL[12:36:50] *** UUncia has quit IRC[12:38:51] *** HugoMaxx has quit IRC[12:40:30] *** hang3r has joined ##OpenGL[12:41:13] *** wey has joined ##opengl[12:41:24] *** hibread has quit IRC[12:41:39] *** hibread has joined ##opengl[12:48:14] *** KU0N has joined ##opengl[12:57:13] *** jsvuorin has joined ##OpenGL[12:59:00] *** Loack has joined ##OpenGL[13:17:04] *** marenz_ has joined ##OpenGL[13:21:35] *** scai has joined ##opengl[13:24:10] *** Loack- has joined ##OpenGL[13:25:37] *** gallahad has joined ##OpenGL[13:26:07] *** gallahad has quit IRC[13:27:11] *** twist has joined ##OpenGL[13:28:44] *** rabbit- has joined ##OpenGL[13:30:01] *** Burga has quit IRC[13:33:03] *** Loack has quit IRC[13:39:08] *** NevroPus has quit IRC[13:41:43] *** maxton has quit IRC[13:47:04] *** kbotnen has joined ##OpenGL[13:53:33] *** djork has joined ##opengl[14:03:11] *** twist_ has joined ##OpenGL[14:05:23] *** reprore_ has joined ##OpenGL[14:06:26] *** Loack- has quit IRC[14:07:12] *** reprore__ has joined ##OpenGL[14:07:21] *** kenws has quit IRC[14:08:23] *** reprore_ has quit IRC[14:08:33] *** djork has quit IRC[14:13:02] *** wey has quit IRC[14:14:36] *** marenz_ has quit IRC[14:22:00] *** twist has quit IRC[14:23:31] *** reprore__ has quit IRC[14:36:15] *** NevroPus has joined ##OpenGL[14:36:42] *** UUncia has joined ##OpenGL[14:37:32] *** stringfellow has quit IRC[14:38:38] *** UUncia has quit IRC[14:38:43] *** UUncia has joined ##OpenGL[14:38:46] *** maxton has joined ##OpenGL[14:41:20] *** stringfellow has joined ##opengl[14:44:22] *** pietia has joined ##OpenGL[14:50:34] *** tokemonstah has quit IRC[14:53:40] *** lewymati has joined ##OpenGL[15:07:28] *** ectropy has joined ##OpenGL[15:08:35] *** phrosty has quit IRC[15:09:06] *** hang3r has quit IRC[15:30:00] *** HuntsMan has quit IRC[15:35:45] *** UUncia has quit IRC[15:36:51] *** rabbit- has quit IRC[15:36:54] *** jm|afk has joined ##OpenGL[15:42:01] *** jm|earth has quit IRC[15:46:59] *** mikolalysenko has joined ##OpenGL[15:57:57] *** dvoid_ has quit IRC[16:01:48] *** nitrotrigger has joined ##OpenGL[16:06:23] *** AlastairLynn has joined ##opengl[16:09:40] <lewymati> hi guys[16:09:45] <lewymati> i need help[16:10:13] <lewymati> i have texture atlas[16:10:22] <MatthiasM> gz[16:10:52] <lewymati> and i want to put "subtexture" on some triangle/quad[16:10:59] <lewymati> but with repeating[16:11:12] <lewymati> is that possible?[16:11:15] <MatthiasM> that's gona be a headache[16:11:40] <MatthiasM> you can try to use a shader and some modulo arithmetic - but texture filter will cause issues[16:11:46] <dizekat> yea[16:11:47] <lewymati> meh[16:12:06] <lewymati> so ill have to do it with normal textures :E[16:12:12] *** Suprano has joined ##OpenGL[16:12:22] <MatthiasM> you would need to use the new sampler function which takes an explicit derivate[16:12:41] <dizekat> speaking of which i wonder how much worse its for performance to have a lot of small textures[16:12:46] <RTFM_FTW> textureGrad is your friend :)[16:12:48] <MatthiasM> or if you need repeat only in one axis - put the texture in a way into the atlas the it's the full width/height[16:13:08] <lewymati> good idea mat :D[16:13:08] *** Ingenu has quit IRC[16:13:13] <MatthiasM> or adjust your model to not need this repeating[16:13:33] <lewymati> well that wont be possible[16:13:59] <lewymati> btw its only 2d game but i'd like to make it in most efficient way[16:14:20] <MatthiasM> do you need to zoom ?[16:14:32] <lewymati> hmm[16:14:39] <lewymati> not rly[16:14:39] <MatthiasM> if not you could set it to nearest - then filtering is not causing issues[16:15:15] <lewymati> i'm quite newbie, idk what you mean by "set it to nearest"[16:15:25] <MatthiasM> texture filtering[16:15:35] <MatthiasM> look it up in the red book[16:24:14] *** Ingenu has joined ##OpenGL[16:24:45] <RTFM_FTW> GL_MIN_FILTER, GL_MAG_FILTER[16:26:13] <lewymati> so is it going to repeat specified rect from texture?[16:26:22] <MatthiasM> no[16:26:35] <MatthiasM> it's texture filtering control[16:26:40] <lewymati> so what it will do[16:26:53] <MatthiasM> read the docu[16:26:54] <lewymati> what's the difference between that and LINEAR[16:27:16] <MatthiasM> lewymati: http://www.opengl.org/sdk/ and http://www.glprogramming.com/red/[16:27:39] <MatthiasM> lewymati: min/mag are keys and linear is a value[16:27:49] <MatthiasM> look at glTexParameter[16:32:47] *** elite01 has joined ##opengl[16:36:01] *** k3t has quit IRC[16:36:06] *** k3t has joined ##OpenGL[16:36:45] <lewymati> oh, so NEAREST is faster[16:37:16] <MatthiasM> depends on HW - mostly no noticable difference[16:37:40] <MatthiasM> but NEAREST will always only access a single texel[16:38:01] <MatthiasM> while LINEAR will also access neighbour texels[16:38:09] <MatthiasM> and sometimes you don't want that[16:45:00] *** nettikkcah has joined ##opengl[16:45:20] *** nettikkcah is now known as hackkitten[16:47:20] <lewymati> okay, i see[16:48:05] <lewymati> 16:13:12 <dizekat> speaking of which i wonder how much worse its for performance to have a lot of small textures[16:48:19] <lewymati> so what's your opinions/facts[16:48:19] <lewymati> ?[16:48:37] *** sohail has joined ##OpenGL[16:48:52] <RTFM_FTW> NEAREST being faster will depend upon the format(s) being fetched[16:48:55] <MatthiasM> lewymati: why do you need repeat at all for 2d ?[16:49:13] <lewymati> well[16:49:19] <lewymati> do you know soldat?[16:49:24] <MatthiasM> no[16:49:35] <RTFM_FTW> in many cases as MatthiasM stated the performance difference between NEAREST and LINEAR is effectively nonexistent[16:49:40] <lewymati> http://www.freemmorpglist.com/pics/foto_soldat.jpg[16:49:49] <lewymati> terrain is made with triangles[16:49:51] <RTFM_FTW> since the texture pipe can fetch at the same performance rate in both cases[16:49:55] <lewymati> and its covered with texture[16:50:17] <MatthiasM> lewymati: the use a single texture for the background - or do manual tiling[16:50:21] *** rorokimdim___ has joined ##opengl[16:50:25] <RTFM_FTW> float targets (i.e. FP16, FP32) are the cases where you are likely to notice a performance win in the NEAREST case[16:51:11] *** rorokimdim___ has quit IRC[16:51:13] <lewymati> MatthiasM: have you looked at that screenshot?[16:51:18] <MatthiasM> yes[16:51:53] <MatthiasM> not all textures are used in repeat mode - keep these in simple textures - and put other things - like your soldats - into a sprite sheet[16:52:17] <lewymati> okay[16:52:36] <lewymati> i already wanted to do it like this way, i just was wondering[16:53:09] <lewymati> and im quite disapointed about FPS[16:53:24] <MatthiasM> do you use immediate mode ?[16:53:25] <lewymati> i have like 500 FPS when i dont draw anything :E[16:53:45] <MatthiasM> idle FPS is meaning less[16:53:58] <Ingenu> yep[16:54:12] <Ingenu> and FPS isn't really relevant, measure frame duration, it's more interesting[16:54:51] <lewymati> what do you mean by immediate mode?[16:54:51] <MatthiasM> but measure over several frames - at minimum 10 frames[16:54:59] <MatthiasM> lewymati: glBegin/glEnd[16:55:04] <lewymati> ah :D[16:55:08] <lewymati> yes, im using[16:55:15] <sparky> meth?[16:55:28] <lewymati> i don't rly get display list yet[16:55:43] <Ingenu> fine for prototyping or GUI stuff[16:56:14] <lewymati> i want to achieve soldat-like gfx[16:57:03] <lewymati> ah, something more[16:57:19] <MatthiasM> lewymati: DLs are easy: you generate an id using glGenLists, then you start the DL compiler using glNewLIst(id, GL_COMPILE), then you render everything that should end in the DL list, you call glEndList() - and now you can just call glCallList to render it again[16:57:29] <MatthiasM> eg for your background[16:58:05] <lewymati> hmm[16:58:57] <lewymati> so its good for objects which are displayed all time in game, yes?[16:59:12] <MatthiasM> DL list are for things that are not changing[16:59:20] <MatthiasM> but you can still move and scale them[16:59:49] <MatthiasM> if you don't include a glLoadIdentity in it - then it will use the current matrix when it's executing[16:59:55] <RTFM_FTW> or you can just submit all of your vertex data as a static VBO[17:00:02] <sparky> matt: do you know if the only way to avoid copying of vertices per frame is to use a VBO?[17:00:02] <RTFM_FTW> which is very trivial[17:00:14] <sparky> or is using glVertexPointer inside a new list enough?[17:00:21] <MatthiasM> think of it like a macro recorder - it will just do what ýour code would do when executed in the place where you call the list[17:00:22] <RTFM_FTW> as this code snippet demonstrates: http://paste.lisp.org/display/72087[17:00:36] <sparky> or RTFM for that matter[17:00:50] <MatthiasM> sparky: a DL will also store the vertex data inside OpenGL somewhere[17:01:27] <RTFM_FTW> "do you know if the only way to avoid copying of vertices per frame is to use a VBO?" ...yes[17:01:27] <MatthiasM> sparky: and while you can record VAs inside a display list - you should not use VBOs in a display list - that's somehow strange[17:01:39] <tmccrary1> Aren't display lists "emulated" somewhat on todays hardware?[17:01:59] <RTFM_FTW> DLs due to their immutable nature enforce a copy of any and all data members within[17:02:12] <RTFM_FTW> this isn't what you probably want :P[17:02:13] <tmccrary1> Do they actually have a separate implementation from static VBOs on the server side?[17:02:13] <MatthiasM> DLs are handled by the OpenGL runtime - atleast you got the data out of your memory into that of the driver[17:02:34] <sparky> RTFM: ok, so glVertexPoint within a DL won't do the trick is what you are saying[17:02:46] <RTFM_FTW> well on a current driver DLs are effectively a shallow wrapper around (static VRAM) VBO[17:02:46] <sparky> I have to use a VBO to avoid the per-frame-copy[17:02:50] <MatthiasM> tmccrary1: as DLs can also contain matrix operations - yes[17:02:54] <tmccrary1> thats what I figured[17:03:04] <lewymati> >im wondering if i should bind the textures of objects which are out of screen for performance[17:03:13] <lewymati> i mean if i shouldnt[17:03:14] <RTFM_FTW> its up to the DL optimizer to make sure that the (V)BOs generated are optimal[17:03:15] <sparky> RTFM: ok, so it will get rid of the per-frame-copy then?[17:03:19] <RTFM_FTW> for the underlying HW[17:03:22] <MatthiasM> lewymati: if you know it's outside - then don't draw it[17:03:35] <lewymati> okay[17:03:52] <sparky> which of course I assume glVertexPointer without the dl will not[17:04:21] <sparky> correct?[17:04:22] <MatthiasM> sparky: when you use VAs - then the data is copied everytime you execute a draw call[17:04:40] <RTFM_FTW> in the non VBO case yes[17:04:47] *** neoneye has quit IRC[17:05:05] <RTFM_FTW> this is what the various *Pointer calls do[17:05:20] <MatthiasM> as VBOs are very easy to use - it should be trivial to use them when you already have VA code working[17:05:36] *** KU0N has quit IRC[17:05:42] <RTFM_FTW> yep[17:05:45] <MatthiasM> RTFM_FTW: the glXYZPointer call does not copy the data - glDrawXYZ does[17:05:58] <RTFM_FTW> hell just use the snippet I posted above as a template :P[17:06:24] <sparky> so to me it sounds like the new list won't change that? --> <MatthiasM> sparky: when you use VAs - then the data is copied everytime you execute a draw call[17:06:32] <sparky> wouldn't it do the same thing when you call the list[17:06:40] <RTFM_FTW> *Pointer references (obtains a pointer to) toe data[17:06:48] <RTFM_FTW> where it is copied is irrelevant[17:06:59] <RTFM_FTW> err the[17:07:34] <MatthiasM> RTFM_FTW: it does - as you can change the data between draw calls without calling glXYZPointer again[17:07:39] <RTFM_FTW> in any case you BOs allow one to skip that step at render time which is where the performance difference between VBO and naive VA comes from[17:08:03] <sparky> yes but in the case where glVertexPointer is in a DL[17:08:05] <RTFM_FTW> since you have effectively decoupled the submit / update step from rendering[17:08:06] <lewymati> okay, thanks guys[17:08:07] <sparky> what happens when you call the list[17:08:10] <lewymati> bye[17:08:10] <MatthiasM> sparky: a DL stores what is executed - not as commands - but the result/effect - so glDrawXYZ copies the arrays into the DL[17:08:12] <sparky> will it copy it at the call[17:08:18] *** lewymati has quit IRC[17:08:24] <RTFM_FTW> exactly[17:08:24] <sparky> I see[17:08:32] <sparky> so the copy will take place when the list is created[17:08:35] <RTFM_FTW> hence the reason why I said what I said :P[17:08:37] <sparky> and not every time it is called[17:08:47] <RTFM_FTW> yep[17:08:48] <sparky> correct?[17:08:52] <sparky> ok, thanks[17:08:53] <MatthiasM> sparky: but if you already have VA - just go the step to VBOs[17:08:55] <tmccrary1> long story short, use VBO instead[17:09:04] *** rnx has joined ##opengl[17:09:04] <sparky> matt: I know, just wanted to know anyway[17:09:08] <MatthiasM> ok[17:09:10] <tmccrary1> DLs are deprecated now anyway[17:09:35] <sparky> I was just curious if the copy would take place when new list is made or when it is called[17:09:48] <AlastairLynn> why would you want DLs now?[17:10:07] <MatthiasM> I still use DLs when I create GL11 code[17:11:22] <AlastairLynn> k[17:11:26] <sparky> alas: well it still gives ogl a chance to process some of the things you need to do up front. More specifically what that is I'm not able to tell you[17:11:27] *** AlastairLynn has quit IRC[17:11:31] <sparky> I don't write the drivers[17:11:43] *** LordMetroid has joined ##OpenGL[17:11:47] <sparky> replace the word 'need' with 'intend'[17:13:12] <sparky> but if you are using VBOs I'm sure the need is not equally significant[17:13:17] <sparky> to use DL[17:13:41] <sparky> what's your opinion RTFM?[17:13:51] <sparky> is there a point to DLs when you are already using VBOs[17:16:07] *** jfroy has quit IRC[17:16:14] *** jfroy has joined ##OpenGL[17:34:37] *** belou has quit IRC[17:35:15] *** ectropy has quit IRC[17:35:42] *** jfroy has quit IRC[17:38:49] *** mathias__ has quit IRC[17:41:53] *** dusted has joined ##OpenGL[17:42:40] <dusted> Hmm, someway of: a.) Hinting to the hardware to enable vsync at a certain freq, b.) Detect if vsync is on, and for how long it waits?[17:42:52] *** rutski_ has joined ##OpenGL[17:43:18] *** [AD]Turbo has quit IRC[17:44:55] <MatthiasM> dusted: enable vsync and measure the time[17:45:27] <dusted> MatthiasM, okay, measuring the time is how I'm doing it atm, but, anyway to tell the driver to enable vsync?[17:45:43] <MatthiasM> yep - pltform dependant[17:45:55] <dusted> okay, thanks :)[17:46:21] <MatthiasM> in LWJGL it's call "Display.setVSyncEnabled(true);"[17:46:25] <MatthiasM> *called[17:47:13] <dusted> seems SDL have one too, I'll try that[17:48:29] <dusted> wohoo, works :) thank you[17:49:04] <MatthiasM> :)[17:49:12] <MatthiasM> but be prepared that vsync may not work[17:49:37] <MatthiasM> atleast on nVidia/windows you can force it off in the driver settings[17:49:40] <dusted> yep, I will be, and fall back on usleep(16666)[17:49:50] <MatthiasM> :)[17:50:32] <MatthiasM> well - better to take one time stamp at the start of your loop - and a 2nd after the swap buffer call - and then sleep if you still have time remaining[17:50:41] <dusted> I also tried measuring the time it takes to swap buffers, and subtract that from the sleeping time[17:51:14] <MatthiasM> don't measure a single call[17:51:19] <MatthiasM> measure everything you do per frame[17:51:35] <dusted> oh, yes, ofcause[17:54:36] *** LordMetroid has quit IRC[17:54:54] *** vade has quit IRC[17:55:02] *** vade has joined ##OpenGL[17:58:09] *** jfroy|work has joined ##OpenGL[17:58:14] *** jfroy|work has quit IRC[17:58:24] *** jfroy|work has joined ##OpenGL[18:06:58] *** emzic has joined ##OpenGL[18:07:58] *** Walt has joined ##opengl[18:10:38] *** Kasu has joined ##OpenGL[18:18:39] *** Gorgoroth has quit IRC[18:19:34] *** maxton has quit IRC[18:34:17] *** LordMetroid has joined ##OpenGL[18:34:23] *** GuShH has quit IRC[18:40:10] *** vade has quit IRC[18:45:07] *** groton has quit IRC[18:50:47] <predaeus> MatthiasM, in your Java engine, do you have to allocate buffers that you use repeatedly for large buffers or do you solely let the garbage collector take care of memory allocation?[18:51:22] <MatthiasM> I try to minimize dynamic allocation[18:51:23] <predaeus> MatthiasM, I mean client side buffers[18:51:28] <MatthiasM> but I don't use pools[18:52:11] <MatthiasM> eg when I create an object like an UFO then it won't need to allocate additional memory as long as it flys around[18:52:24] <predaeus> ok, so that seems to be a necessity. Ic, well I am having my first experiences with larger buffers and GCs and thought it would be handle by the GC.[18:52:27] <MatthiasM> only when it's destroyed then it's replaced with another object[18:52:53] <MatthiasM> well - what buffer do you allocate ?[18:53:23] <predaeus> Using D here atm. Thought it's GC might be much worse than Java's. I allocate client side memory for updating mesh data to pass to the server VBO.[18:53:44] <predaeus> having dynamic meshes that is.[18:53:56] <MatthiasM> the GC from Sun's Java is one of the most advanced[18:54:12] <MatthiasM> well - when I need to update a VBO I'll just map it[18:54:18] <MatthiasM> no need to allocate anything[18:55:15] <predaeus> Hm, so it would be better to map even when updating every frame in some circumstances?[18:55:31] <MatthiasM> yes[18:55:55] <predaeus> Ok thanks, I had thought it would pose problems somehow.[18:56:02] <MatthiasM> but when you update every frame you should call glBufferData(,..., NULL); with the same size as the buffer before mapping[18:56:27] <MatthiasM> this was in some GDC presentations from nVidia and ATI[18:56:38] <predaeus> ok thank you for the tip![18:58:16] *** itewsh has joined ##OpenGL[19:01:08] *** HuntsMan has joined ##opengl[19:02:44] *** Quentarez has quit IRC[19:03:00] *** davidc__ has joined ##opengl[19:21:49] *** NightVisio has joined ##OpenGL[19:23:44] <dusted> if I timestamp from start of loop, and then again after swapping buffers, if i then sleep the remaining time, and the next loop is just a bit longer, then it will skip a frame ?[19:24:16] <MatthiasM> huh ?[19:24:47] <TheFlash> No?[19:25:43] <dusted> uhm, swapbuffers will wait for vsync, so, if I use "exactly" the time that one frame takes, and the code drawing the next frame is longer, then swapbuffers will miss the sync and have to wait for the next[19:25:54] <dusted> use (ie sleeping)[19:26:20] <MatthiasM> if your vsync rate is 60HZ - then you should aim your own delay at ~61HZ[19:26:49] <MatthiasM> that way you won't sleep when vsync is on[19:27:32] <dusted> ah, good idea[19:30:54] *** Walt has quit IRC[19:34:34] *** HuntsMan has quit IRC[19:34:56] *** sporks has joined ##OpenGL[19:35:46] *** b0000 has joined ##opengl[19:37:38] *** pietia has quit IRC[19:43:07] *** lewymati has joined ##OpenGL[19:49:11] *** sparky has quit IRC[19:50:00] *** Kasu has quit IRC[19:53:12] *** Nescafe has quit IRC[20:02:23] *** Magicblaze007 has joined ##OpenGL[20:02:37] <Magicblaze007> how do I get the color values of a pixel in opengl?[20:03:31] <MatthiasM> Magicblaze007: http://www.opengl.org/sdk/ look through the reference pages[20:03:45] <MatthiasM> you should easily find the function based on it's name[20:04:45] <dizekat> lol just had weirdest thing ever with points[20:05:03] <dizekat> wrote shader for distance attenuation....[20:05:19] <dizekat> using 1/distance^2 formula it didnt work right[20:05:26] <dizekat> and using 1/distance it worked...[20:05:42] *** pietia has joined ##OpenGL[20:05:43] <dizekat> after some "wtf wtf" figured it was scaling alpha by 1/distance too[20:05:58] <dizekat> so attenuation was 1/distance^2 when using 1/distance in shader lol[20:06:08] <MatthiasM> heh[20:06:15] <dizekat> but for brief ten minutes or so[20:06:16] *** mm765^sleep is now known as mm765[20:06:28] <dizekat> i thought i gone crazy and cant even figure how to attenuate brightness with distance[20:06:29] <dizekat> :P[20:06:38] <Magicblaze007> I can find a function that reads pixels in a rectangular area...glReadPixels?[20:06:48] *** elite01 has quit IRC[20:06:50] <Magicblaze007> but can't find one that can just read one pixel color from a location[20:07:06] <MatthiasM> Magicblaze007: read a 1x1 rectangle[20:11:17] <TheFlash> Reading pixels back is not very wise, in many cases.[20:12:52] *** NightVisio has quit IRC[20:18:06] <Magicblaze007> of course[20:23:40] *** cplusplus2 has joined ##OpenGL[20:29:26] *** groton has joined ##OpenGL[20:30:07] *** vade has joined ##OpenGL[20:31:40] <predaeus> dusted, does glXSwapBuffers block? From what I've read now it is not clear to me, they say it swaps the buffer then issues a glFlush. So timing SDL_GL_SwapBuffers might not give you the time it takes to render.[20:35:22] <emzic> hello, is there something like a software implementation for opengl?[20:35:31] <emzic> like pixomatic for directx?[20:35:46] <predaeus> emzic, mesa[20:35:56] <emzic> ok thanks[20:36:27] *** Spkka has joined ##OpenGL[20:36:57] <emzic> are there precompiled windows libraries for mesa?[20:39:38] <Ingenu> I'd think so, but why would you want to use MESA instead of your hardware ?[20:40:14] <emzic> because my hardware isnt capable of newer stuff[20:41:18] *** cplusplus has quit IRC[20:50:06] *** mm765 is now known as mm765^away[20:58:44] *** elite01 has joined ##opengl[20:59:24] *** jcouture has quit IRC[21:00:02] *** stringfellow has quit IRC[21:00:04] *** Magicblaze007 has left ##OpenGL[21:06:56] *** Nescafe has joined ##OpenGL[21:07:00] *** KingGuru has joined ##OpenGL[21:07:10] <KingGuru> Hi[21:24:43] *** Quentarez has joined ##OpenGL[21:29:36] <emzic> hi, is it ok to ask mesa3d questions here?[21:35:10] <KingGuru> try.. it seems to be rather dead in here.. friday.. you know[21:39:41] *** predaeus has quit IRC[21:40:26] <emzic> i just figured it out myself... my app wasnt loading the opengl32.dll i compiled[21:40:38] <emzic> but i had it in the wrong directory... XD[21:45:08] <KingGuru> fair enough[21:48:45] *** nitrotrigger has quit IRC[21:55:08] *** Spkka has quit IRC[21:55:49] *** dv_ has joined ##opengl[21:57:06] *** GuShH has joined ##OpenGL[21:59:10] *** e_roder has joined ##OpenGL[21:59:22] <e_roder> hey[21:59:41] <tmccrary1> Its you![21:59:56] <e_roder> how can i set up an ambient lighting color for surfaces so they have color even if they're not hit by a light?[22:00:12] <MatthiasM> e_roder: read the red book[22:00:18] <MatthiasM> it covers this in great detail[22:00:27] <e_roder> hm[22:00:34] <e_roder> it's working with textured surfaces[22:00:37] <e_roder> but not non-textured ones[22:00:57] <MatthiasM> e_roder: http://mikeash.com/getting_answers.html <-- read point #1[22:01:09] *** cplusplus has joined ##OpenGL[22:01:42] <e_roder> working meaning a non-lit untextured surface appears totally black[22:01:46] <e_roder> er[22:01:50] <e_roder> that's not working[22:02:22] <MatthiasM> you got your answer - go and read[22:02:23] <e_roder> where-as a textured non-lit surface displays the texture with a darker color[22:03:14] <e_roder> then why did you bother pointing me to that link?[22:03:32] <e_roder> you clearly had no intention of helping whether my question was clearly stated or not[22:04:18] <oliveira_> e_roder: give the surface a color and disable lighting when rendering[22:04:41] <e_roder> thanks[22:04:57] *** Zemyla has joined ##OpenGL[22:05:15] <e_roder> there's no way to do it with lighting enabled?[22:05:43] <Zemyla> Is there a simple program that I can download on Windows to see what my graphics card supports? Because I don't know whether this one supports shaders.[22:05:48] <oliveira_> e_roder: read about materials and emission, that should cover it[22:06:33] *** groton has quit IRC[22:06:33] <e_roder> Zemyla: you should be able to find out what your card supports on wiki or the card manufacturer's website[22:07:25] *** groton has joined ##OpenGL[22:11:01] <dizekat> utter wtf: glGenerateMipmap does not use box filter for me[22:11:08] <dizekat> using half-float texture format[22:11:25] <dizekat> it just scales down discarding pixels, it seems[22:11:35] <dizekat> need it for bloom... :/[22:12:57] <KingGuru> Anyone in here been working with Cuda as well ?[22:13:03] <dizekat> i'm going to[22:13:08] <KingGuru> :)[22:13:09] <dizekat> but not yet[22:13:20] <dizekat> having enough pain with opengl at the moment heh[22:14:32] <KingGuru> well.. atm I'm trying to make a volume ray tracer, where I load in a CT scan.. so far so good.. then I have to figure out how to display the data I get from the ray tracing part.... this is where the chain falls off..[22:14:39] <KingGuru> for me anyway[22:16:41] <dizekat> cool[22:16:51] <dizekat> theres examples in cuda sdk, no?[22:16:59] <dizekat> how to convert cuda stuff to opengl texture[22:17:01] *** neoneye has joined ##OpenGL[22:17:01] <dizekat> and vice versa[22:17:22] <dizekat> actually i want to do something kinda related[22:17:42] <dizekat> i want to get opengl texture (to which i render) into cude[22:17:44] <dizekat> *cuda[22:17:58] <dizekat> then do FFT based convolution on it[22:18:00] *** cplusplus2 has quit IRC[22:18:03] <dizekat> for quality lens flare effect[22:18:10] <dizekat> then put it back to gl and draw[22:19:24] <KingGuru> ahh.. ok[22:19:37] <KingGuru> well.. yes.. there IS an example in the SDK..[22:19:43] *** groton has quit IRC[22:20:53] <KingGuru> maybe I should read that one through in the morning.. it's 22:20 here.. and the couple upstairs is having a party..[22:22:21] <e_roder> lol[22:22:37] <e_roder> go upstairs and join the party[22:22:45] <e_roder> 3's a threesome as i always say[22:23:04] <KingGuru> well.. it sounds like a lot more than 2 :)[22:24:24] <dizekat> solved that[22:24:28] <dizekat> glHint(GL_GENERATE_MIPMAP_HINT, GL_NICEST);[22:24:30] <dizekat> lol[22:24:49] <Zemyla> Huh. Why is this crazy thing not compiling in cygwin?[22:25:14] <Zemyla> It says "undefined reference to '_glWhatever'" for pretty much every symbol it imports.[22:25:33] <e_roder> you have to put links in from your project file[22:25:41] <e_roder> let me try to explain that again[22:25:52] <MatthiasM> Zemyla: try mingw instead of cygwin[22:25:55] <e_roder> your project has to reference the opengl directory[22:26:14] <MatthiasM> it comes with all required files[22:26:40] <Zemyla> Why isn't it working with an example I downloaded specifically entitled "cygwin example"!?[22:30:05] *** vade_ has joined ##OpenGL[22:32:13] *** sporks has quit IRC[22:32:29] <dizekat> you need to link with libopengl32 or something similar, no?[22:33:10] <MatthiasM> Zemyla: try -lopengl32[22:33:12] <Zemyla> libopengl32 is in my library list, and the directory the library is in is on the command line with a -L.[22:33:43] *** Goran_ has joined ##OpenGL[22:33:45] *** AlastairLynn has joined ##opengl[22:34:32] <Plagman> well it's still not linking against it judging by your errors[22:34:38] *** qeed has joined ##opengl[22:34:49] <Plagman> or is it just failing to link extension symbols?[22:34:53] <Plagman> pastebin the build output[22:35:28] *** Goran_ has quit IRC[22:35:47] *** Goran_ has joined ##OpenGL[22:36:35] *** davidc__ has quit IRC[22:38:03] *** Renderwahn has quit IRC[22:38:07] <Zemyla> http://pastebin.ca/1304617[22:38:48] *** Zemyla has quit IRC[22:38:55] *** davidc__ has joined ##opengl[22:39:09] *** neoneye has quit IRC[22:39:18] *** Zemyla has joined ##OpenGL[22:39:21] <Plagman> and libopengl32 is in /usr/lib/w32api?[22:39:26] <Plagman> same for libglut32[22:39:56] *** vade has quit IRC[22:40:03] <Zemyla> libopengl32.a, libglu32.a, and libglut32.a are all in that directory.[22:40:17] *** vade_ has quit IRC[22:40:21] *** vade has joined ##OpenGL[22:41:27] *** KingGuru has quit IRC[22:51:27] *** groton has joined ##OpenGL[22:53:32] *** lewymati has quit IRC[22:53:55] <Zemyla> And nm shows that the symbols are in the files.[22:57:38] *** e_roder has quit IRC[23:02:29] *** elektr1k has quit IRC[23:11:27] *** dolphin has joined ##OpenGL[23:12:05] *** Luc_le_Pro has joined ##OpenGL[23:12:17] <Luc_le_Pro> hello :-)[23:13:07] <cplusplus> which language do you program a gpu?[23:13:22] <Luc_le_Pro> glsl[23:13:33] <tmccrary1> or Cg[23:14:18] *** hibread has quit IRC[23:15:51] <Zemyla> Huh. Why the hell doesn't this MinGW installation have GL/glut.h?[23:15:54] *** Walt has joined ##opengl[23:16:33] *** mm765^away is now known as mm765[23:18:39] *** msh07 has joined ##OpenGL[23:19:15] <mlucassmith> interesting: http://rss.slashdot.org/~r/Slashdot/slashdot/~3/ZJMWPFp-i_I/article.pl[23:25:05] <Zemyla> I found GLUT for MinGW.[23:26:09] <Zemyla> Okay, now it compiles.[23:28:33] *** Walt_ has joined ##opengl[23:32:35] *** msh07 has quit IRC[23:32:52] *** msh07 has joined ##OpenGL[23:37:52] *** autonomy has quit IRC[23:38:15] *** autonomy has joined ##OpenGL[23:43:34] *** Walt has quit IRC[23:43:45] *** groton has quit IRC[23:48:02] *** cplusplus2 has joined ##OpenGL[23:48:46] *** groton has joined ##OpenGL[23:51:58] *** Walt_ is now known as Walt[23:55:55] *** Deformalite has quit IRC[23:58:58] *** itewsh has quit IRC[23:59:25] *** vade has quit IRC