#jogamp @ irc.freenode.net - 20141112 05:06:27 (UTC)

20141112 05:06:27 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20141111050627.html
20141112 05:06:27 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20141112050627.html
20141112 06:57:19 * hharrison (~chatzilla@anon) Quit (Ping timeout: 250 seconds)
20141112 07:04:10 * jvanek (jvanek@anon) has joined #jogamp
20141112 07:31:22 * monsieur_max (~maxime@anon) has joined #jogamp
20141112 07:36:53 * eclesia (~husky@anon) has joined #jogamp
20141112 07:36:59 <eclesia> good morning
20141112 08:02:01 <[Mike]> good morning!
20141112 11:01:05 * doev (~doev@anon) has joined #jogamp
20141112 11:03:38 * xranby (~xranby@anon) has joined #jogamp
20141112 12:18:21 * jvanek (jvanek@anon) Quit (Ping timeout: 265 seconds)
20141112 12:20:26 * jvanek (jvanek@anon) has joined #jogamp
20141112 12:40:16 * jvanek (jvanek@anon) Quit (Quit: Leaving)
20141112 13:28:01 * xranby (~xranby@anon) Quit (Ping timeout: 258 seconds)
20141112 13:43:00 * xranby (~xranby@anon) has joined #jogamp
20141112 13:54:20 * doev (~doev@anon) Quit (Quit: Verlassend)
20141112 15:55:11 * juank_prada (~juank_pra@anon) has joined #jogamp
20141112 15:58:52 <rmk0> .o.
20141112 16:07:38 * eclesia (~husky@anon) has left #jogamp
20141112 16:21:21 * jvanek (jvanek@anon) has joined #jogamp
20141112 16:50:58 * jvanek (jvanek@anon) Quit (Ping timeout: 265 seconds)
20141112 16:56:59 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20141112 17:11:31 * jvanek (jvanek@anon) has joined #jogamp
20141112 17:16:40 * [Mike] (~Mike]@anon) Quit ()
20141112 17:37:22 * jvanek (jvanek@anon) Quit (Ping timeout: 265 seconds)
20141112 17:46:45 <rmk0> http://blogs.msdn.com/b/dotnet/archive/2014/11/12/net-core-is-open-source.aspx
20141112 17:48:20 * luska72 (8182120b@anon) has joined #jogamp
20141112 17:49:53 * jvanek (jvanek@anon) has joined #jogamp
20141112 17:52:20 * monsieur_max (~maxime@anon) has joined #jogamp
20141112 17:53:41 <luska72> Hey guys, it's me again (the guy having trouble with gl textures). Does anyone know why using: DebugGL2(drawable.getGL().getGL2()); would output "glGetError() = 0" after each draw call (when using -Debug/trace gl run flag") on one computer, but would output nothing at all on a different computer with the exact same code?
20141112 17:54:27 <luska72> I know "glGetError() = 0" means there was no error btw. It's just weird that the working computer outputs nothing, and the computer that doesn't work is outputting that line :P
20141112 17:55:56 <rmk0> mysterious
20141112 17:57:12 <rmk0> it... might be worth putting your code somewhere visible
20141112 17:57:37 <rmk0> trying to guess at opengl issues is basically like throwing darts blindfolded
20141112 17:57:49 <rmk0> with just as much potential for injury!
20141112 17:58:42 <luska72> That's a good point haha. I'll make a github with the code, or maybe a google doc if I can't remember how. then come back to you
20141112 18:00:05 <rmk0> github's a lot better
20141112 18:00:26 <rmk0> can keep track of the fixes that way
20141112 18:29:04 * jvanek (jvanek@anon) Quit (Quit: Leaving)
20141112 18:31:55 <luska72> Ok, I finally got github to work! https://github.com/Luska72/TiledGL Just a few notes: I'm using immediate mode while setting things up, it's temporary! Also I'm working on this project to learn, so if I do anything incorrectly or in a dumb way please let me know :).
20141112 18:32:31 <luska72> I've gotta go eat, but I'll be back soon. Sorry about that and thanks for all the help so far, I really appreciate it.
20141112 18:35:29 <rmk0> will look at it shortly
20141112 18:55:49 <luska72> cool, no rush or anything. If i'm asking you too much please let me know, I don't wanna be bother
20141112 18:56:15 <rmk0> this is why we're here
20141112 18:57:47 <rmk0> i see a call to glMatrixMode(GL_PROJECTION), but no call to glMatrixMode(GL_MODELVIEW)
20141112 18:59:39 <rmk0> and you set the near plane to 0, which... is bad
20141112 19:00:55 <luska72> I'm trying to setup the view to be 2d (hence the glOrtho() I believe), what should the near plane be for that?
20141112 19:02:09 <rmk0> i'm not actually sure what that does to the math, it might be ok
20141112 19:03:53 <luska72> Either way I'll look into that, I got it from a tutorial so I need to make sure everything makes sense in the project :P. Do I need a glMatrixMode(GL_TEXTURE) or MODELVIEW anywhere? I removed that because it didn't seem to effect anything at the time.
20141112 19:04:22 <rmk0> it just tells opengl which matrix the following operations will affect
20141112 19:04:30 <rmk0> if you're not touching the modelview matrix, it won't do anything
20141112 19:05:30 <rmk0> i don't see anything obviously wrong, but it's been a long time since i looked at the fixed function stuff
20141112 19:05:33 <rmk0> it's awful
20141112 19:05:51 <luska72> fixed function? Are you talking about the immediate mode?
20141112 19:05:55 <rmk0> yeah
20141112 19:06:18 <rmk0> i'm mostly suspicious that you're drawing stuff at z = 0, with the near clipping plane also being 0
20141112 19:06:34 <rmk0> could try drawing at z = -0.5, or something
20141112 19:07:13 <luska72> Yeah, I actually have a code sample that uses VBOs, but the same texture problem applied so I'm using immediate mode while I try to fix it. I'll try z = -0.5 real quick and tell you what happens
20141112 19:10:07 <luska72> yeah that didn't change anything :(. Just as a note, remember that this code does display a texture on my laptop, but displays a black screen on this desktop. So I think that means I coded something that doesn't work across graphics cards?
20141112 19:11:35 * luska72a (~luska72@anon) has joined #jogamp
20141112 19:11:45 <rmk0> what type of gpu is it on both systems?
20141112 19:11:50 <rmk0> got a feeling i already know...
20141112 19:16:03 * luska72a (~luska72@anon) Quit (Ping timeout: 264 seconds)
20141112 19:26:25 * luska72a (~luska72@anon) has joined #jogamp
20141112 19:27:48 <luska72a> Sorry about the delay, I had to walk to class. Lucky for me I put my gpu info in a pastebin the other day! http://pastebin.com/b0Di5ns1
20141112 19:28:25 <rmk0> hehe, yeah... works on nvidia, doesn't work on amd
20141112 19:28:28 <rmk0> same old story!
20141112 19:28:40 <rmk0> nvidia is very permissive of mistakes, amd isn't at all
20141112 19:29:35 <luska72a> Nuts. So what can I do to circumvent this? I'd like to write code that works on basically all platforms
20141112 19:29:58 <rmk0> i notice you don't actually enable texturing anywhere
20141112 19:30:12 <rmk0> glEnable(GL_TEXTURE_2D)
20141112 19:30:37 <rmk0> all you can do is follow the specification to the letter, use the type system to prevent as many mistakes as possible, and develop on AMD first
20141112 19:31:20 <rmk0> i think the fixed function pipeline is harder to learn and manage, frankly
20141112 19:31:33 <rmk0> there's... so much more of it, and so much more state you have to keep track of
20141112 19:34:22 <luska72a> I actually just tried that before I left, and enabling it didn't work :(. Can you recommend me something I can read to learn about what you're describing? And I don't plan on using immediate mode as anything but a placeholder for vbo code :)
20141112 19:40:17 <rmk0> i'd do one of the "modern opengl" tutorials
20141112 19:40:35 <rmk0> http://ogldev.atspace.co.uk/
20141112 19:41:08 <rmk0> be sure to ask for "core" context... GLProfile.get(GLProfile.GL3) and use DebugGL3
20141112 19:41:50 <rmk0> once you start getting the hang of the concepts, you start building a safe layer of types over the existing weakly-typed API
20141112 19:42:04 <rmk0> so that you're prevented from passing in wrong values to the API, mixing up all the various integers, etc
20141112 19:42:17 <rmk0> otherwise you'll generally make the same mistakes over and over and over, because the API happily allows it
20141112 19:43:08 <rmk0> it's really not an API that's friendly to experimentation
20141112 19:43:16 <rmk0> you need to really understand it, or it'll just hurt you endlessly
20141112 19:44:28 <rmk0> http://mvn.io7m.com/io7m-jcanephora is one i wrote and use every day... i'd not recommend using it yet as the API's not stable, but looking at the implementation might help
20141112 19:44:44 <rmk0> it cleanly separates all the different parts of the API and puts a strong layer of types over the top
20141112 19:44:47 <rmk0> difficult to make mistakes
20141112 19:45:46 <luska72a> Makes sense, and will this tutorial also talk about gpu specification at all? Tbh I don't know what that even is haha.
20141112 19:46:08 <rmk0> oh, i meant the opengl specification
20141112 19:46:41 <rmk0> i generally write towards the opengl 3.1 specification, as that's the bare minimum you can currently get away with (mostly limited by the intel drivers on linux that still only support 3.1 on a lot of hardware)
20141112 19:46:57 <rmk0> 3.1 is a very small sane core... much smaller than any of the versions that came before it
20141112 19:47:02 <rmk0> good place to start learning, i think
20141112 19:47:26 <rmk0> i'd... do the tutorials first, and the specification later
20141112 19:47:31 <rmk0> it can be an intimidating read
20141112 19:48:40 <luska72a> Yeah that makes sense. I'll go through them over the next few weeks, luckly I have a break coming up haha
20141112 19:52:18 <luska72a> Well thanks for the help, i really appreciate it. I'll probably come back to my texture problem after reading the tutorial, hopefully I'll know how to fix it by then. If you don't mind me asking, what's your job? Do you work on graphics professionally or something?
20141112 19:52:55 <rmk0> self employed
20141112 19:53:12 <rmk0> currently building the rendering portion of a game i've been trying to put together for a few years
20141112 19:53:59 <rmk0> will be open source, but commercial
20141112 19:54:11 <rmk0> pay for the game content, like the way quake is distributed now
20141112 19:54:27 <luska72a> That's awesome, what's the game about?
20141112 19:55:07 <rmk0> not sure how to describe it without giving an endless list of labels
20141112 19:55:36 <rmk0> sort of a utopian sci-fi open world adventure game
20141112 19:55:55 <rmk0> sort of 2.5Dish
20141112 19:56:18 <rmk0> the intention is to release it with as many editing tools as possible and build a community of people that write new storylines and scenarios for it
20141112 19:57:13 <luska72a> 2.5d, meaning like a 2d game that looks 3d? Also that's really awesome, so people can make their own levels and characters and stuff?
20141112 19:57:37 <rmk0> hehe, it's... more sort of 3D that looks 2D /o\
20141112 19:57:50 <rmk0> the camera's fixed on a plane, can move up/down/left/right, but not in or out of the scene
20141112 19:58:05 <rmk0> but the scenes have depth, so the characters can move into the foreground/background
20141112 19:58:26 <rmk0> suppose the scene mechanics are a bit like that south park stick of truth game
20141112 19:58:33 <rmk0> "not exactly 2D"
20141112 19:58:44 <rmk0> not sure if you're familar with Nifflas and the games he produces
20141112 19:59:02 <rmk0> http://nifflas.ni2.se
20141112 20:00:21 <rmk0> he released a game called Knytt Stories
20141112 20:00:32 <rmk0> very simple game, beautiful minimalist aesthetics
20141112 20:00:47 <rmk0> but he released an editor with it, and people set about making a ridiculous number of new scenarios for it
20141112 20:00:57 <rmk0> as in... tens of thousands
20141112 20:01:01 <luska72a> I'll check them out when I get back to my dorm room, because I'd really like to see how all this works. I don't suppose you've seen stabyourself's games? I might be remember that name wrong :P
20141112 20:01:14 <rmk0> never heard of it
20141112 20:02:03 <rmk0> now i have
20141112 20:02:12 <luska72a> I'll have to find his website, he makes games in love2d (a language I coded in for a while) his most famous game is mari0, it's Mario with a portal gun.
20141112 20:03:01 <luska72a> The reason I bring him up us because he has an orthobot game which does some cool 3d/2d stuff too
20141112 20:03:30 <rmk0> anyway, er... point was that the knytt engine was horribly primitive, and the editor was "easy" but very limited
20141112 20:03:42 <rmk0> wondered what could be done if an advanced engine was written with really good editing tools
20141112 20:03:56 <rmk0> one that's actually designed with ease of content production in mind
20141112 20:04:25 <rmk0> seems like almost no developers actually try to take advantage of the modding communities they end up with
20141112 20:05:16 <luska72a> That'd be really cool, mods are usually crazy good. So you make good modding tools and get a community and it just expands from there? :)
20141112 20:05:26 <rmk0> yeah
20141112 20:05:38 <rmk0> ideally set up a small marketplace where people can earn money from their creations
20141112 20:06:25 <rmk0> i'll take a small percentage \o/
20141112 20:07:11 <luska72a> That's really smart, something like blender's store but not impossible to use?
20141112 20:07:17 <rmk0> hehe
20141112 20:07:23 <luska72a> I meant to say unity, not blender
20141112 20:07:34 <rmk0> not familiar with unity's store
20141112 20:08:17 <rmk0> think steam may do something like this already with the steam workshop
20141112 20:08:39 <rmk0> last eighteen months have been non-stop work on a good modern deferred renderer, because one just doesn't exist for java
20141112 20:08:52 <rmk0> is basically done now, just documenting and finishing any minor pieces
20141112 20:10:03 <luska72a> That's awesome, I'd love to see it when it's done. Deferred rendering is the same as deferred shading right?
20141112 20:10:13 <rmk0> think so
20141112 20:10:35 <rmk0> rendering all of the geometry in the scene into a buffer, and then performing lighting on that buffer in screen-space
20141112 20:10:58 <rmk0> the renderer only supports dynamic lighting, so i wanted to have a lot of them
20141112 20:11:10 <rmk0> static lighting ups the cost of content production too much
20141112 20:11:16 <rmk0> nobody's going to sit there and wait for lightmaps to compile
20141112 20:12:13 <luska72a> So you calculate the lighting every time one moves? Then make a light map of the scene at that time?
20141112 20:12:49 <rmk0> http://waste.io7m.com/2014/11/12/io7m-r1-documentation-0.9.0/p2s12.xhtml#st200_p2s12
20141112 20:13:15 <rmk0> lighting is calculated from nothing every frame
20141112 20:13:38 <rmk0> this is the standard way to do things now in games
20141112 20:14:57 <luska72a> That's crazy complex haha. I've read up a bit on 2d lighting, but 3d seems too hard for me
20141112 20:15:16 <rmk0> it's not too bad
20141112 20:15:40 <rmk0> there's pretty much a straight path you can follow
20141112 20:15:58 <luska72a> Are you coding the game in Java btw?
20141112 20:16:04 <rmk0> yeah
20141112 20:16:29 <luska72a> And yeah it's a straight path, but if something goes wrong that would be hard :P
20141112 20:16:47 <rmk0> yeah, there are an endless series of traps
20141112 20:16:56 <rmk0> nobody documents their conventions
20141112 20:17:14 <rmk0> most documentation on techniques is beyond awful
20141112 20:17:26 <rmk0> i've had to piece things together and then document them so that i don't forget
20141112 20:17:40 <luska72a> If everyone followed the Java doc format life would be a lot easier haha
20141112 20:17:59 <rmk0> just some basic static typing helps
20141112 20:18:12 <luska72a> I'm leaving class, so I'll brb in like 10 min or so..
20141112 20:18:16 <rmk0> urhur
20141112 20:18:27 <rmk0> VectorI3F → RVectorI3F<RSpaceEye>
20141112 20:18:52 <rmk0> one of the most horrible and frequent mistakes in opengl is passing in a position that's in one coordinate system to something that's expecting another system
20141112 20:19:05 <rmk0> let the fucking compiler catch those errors!
20141112 20:19:11 <rmk0> saves endless misery
20141112 20:19:38 <rmk0> seems like nobody is even trying any of this stuff... the average graphics programmer thinks stuff that's completely eliminated at compile time is somehow still too expensive
20141112 20:19:48 * rmk0 kicks the average graphics programmer
20141112 20:19:56 <zubzub> you should kick the average programmer
20141112 20:20:00 * rmk0 kicks the average graphics programmer
20141112 20:20:18 <rmk0> is an endless source of frustration
20141112 20:20:22 <zubzub> most programmers have no idea what type safety is
20141112 20:20:25 <rmk0> particularly with shadow mapping... the coordinate systems REALLY MATTER
20141112 20:20:37 <rmk0> if you get it wrong, you'll get bizarre results onscreen, and no way of knowing where the problem is
20141112 20:20:52 <zubzub> most programmers will do whatever the language allows them to do
20141112 20:20:57 <rmk0> hehe
20141112 20:22:22 * luska72b (~luska72@anon) has joined #jogamp
20141112 20:22:22 * luska72a (~luska72@anon) Quit (Ping timeout: 240 seconds)
20141112 20:22:32 * juank_prada (~juank_pra@anon) Quit (Ping timeout: 250 seconds)
20141112 20:28:43 * luska72a (~luska72@anon) has joined #jogamp
20141112 20:29:04 * luska72b (~luska72@anon) Quit (Ping timeout: 255 seconds)
20141112 20:32:35 <luska72> I'm back, and java is type-safe, correct? So incorrect passing isn't something to worry about?
20141112 20:33:02 <rmk0> it's relatively safe as these things go
20141112 20:33:23 <rmk0> is still easy to fuck things up with opengl, as opengl is all integers, all the time
20141112 20:34:00 <luska72> wait, what? What about glSomething3f() functions? Are those casted to ints?
20141112 20:34:22 <rmk0> no, i'm referring to any function that takes a GLEnum parameter
20141112 20:34:35 <rmk0> or any function that, for example, allocates and returns new IDs for different types
20141112 20:34:41 <rmk0> like allocating new textures or vertex buffer objects
20141112 20:34:45 <luska72> Ah, kk
20141112 20:34:53 <rmk0> they're all integers, you can freely mix them up or pass in invalid values
20141112 20:35:16 * luska72a (~luska72@anon) Quit (Remote host closed the connection)
20141112 20:37:24 * hharrison (~chatzilla@anon) has joined #jogamp
20141112 21:11:50 * [Mike] (~Mike]@anon) has joined #jogamp
20141112 21:21:34 * luska72a (~luska72@anon) has joined #jogamp
20141112 21:37:58 * luska72 (8182120b@anon) Quit (Ping timeout: 246 seconds)
20141112 21:58:03 * monsieur_max (~maxime@anon) Quit (Ping timeout: 264 seconds)
20141112 22:12:34 * monsieur_max (~maxime@anon) has joined #jogamp
20141112 22:21:20 * luska72 (~luska72@anon) has joined #jogamp
20141112 22:23:55 * luska72a (~luska72@anon) Quit (Ping timeout: 265 seconds)
20141112 22:28:24 * luska72 (~luska72@anon) Quit (Ping timeout: 255 seconds)
20141112 22:32:22 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20141112 23:41:25 * kermyt (~kermyt@anon) has left #jogamp
20141113 02:15:51 * rmk0 (~rmk0@anon) Quit (Ping timeout: 244 seconds)
20141113 02:18:00 * zzuegg (~zzuegg@anon) Quit (Quit: Nettalk6 - www.ntalk.de)
20141113 02:48:36 * rmk0 (~rmk0@anon) has joined #jogamp
20141113 02:48:36 * rmk0 (~rmk0@anon) Quit (Changing host)
20141113 02:48:36 * rmk0 (~rmk0@anon) has joined #jogamp
20141113 03:01:04 * [Mike] (~Mike]@anon) Quit ()
20141113 04:23:41 * juank_prada (~juank_pra@anon) has joined #jogamp
20141113 04:44:31 * juank_prada (~juank_pra@anon) Quit (Quit: Leaving)
20141113 05:06:28 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20141113050628.html