[QUOTE=Chris220;21852247]It's because of these arguments that WAYWO pages go by so fast
We're already on page 6[/QUOTE]
No, it's the awesomeness that includes Nulls grapher that means we go by so fast, he seems to have every other post and they are 99% either his grapher or constructive criticism or tips.
[QUOTE=turb_;21852870]If you chuck in stuff like finding intersect points, finding intercepts, etc. this can replace my graphics calculator :>[/QUOTE]
I want to do that but it becomes a little complex with implicit graphs so I gotta figure out how the hell to do it :v:
If you can do it just for explicit (amirite?) graphs, that would still be awesome
[QUOTE=ryandaniels;21850803]That's really cool... I was thinking something along these lines, except linearly; I was thinking how members of a list collide. But until you said this, I didn't fully realize the implications.[/QUOTE]
You could probably pull this off in a realtime application with some effort. Practically you can't use the same data set you use for drawing as you use for collision. If I was trying to do real time voxel rendering/collision etc I actually have a few techniques I've thought up.
First is use a quadtree or octree. Keep splitting the models into 8ths until you find a completely empty 8th, a completely full 8th, or reach the level of voxels themselves. You could also do an aproximation to make the rendering faster. So if 3/4ths of the 8th is filled mark it as filled etc.
Second is use a cheap lighting method before casting rays. Since your casting rays into a grid you can actually cheat a lot. First thing is when more than one voxel lies on the same plane that you're tracing against do a test for that plane once and then find the position and since it's a plane you can just do 'point in square' tests instead of recasting for all of them. Then instead of casting a ray for each ones lighting do the math for that plane and see how it'd be shaded with flat shading. You're able to cull up to 3 faces shadow tests I believe. Because obviously if the face is opposite the light it's in shadow so no test is needed.
Third is for collision and movement use a simplified voxel representation. Maybe even just boxes or spheres but using volume units obviously. Using a 512x512x512 graph with a single byte per voxel would be in the 160mb area. You could probably use indivual bits but you'd need to do bitshifting or something so it'd be slowed somewhat but use much less RAM. When rending you can make use of rasterization to speed up rendering I think. Never completely thought this through but I'm pretty sure if you used OpenGL you could render bounding boxes the size of the voxel model and then trace for that first. In most cases you'd find a collision in the octree. If not you could just do a tradiional loop through each face checking for a hit. Something about how OpenGL renders could stop that from working. I dunno just thought of that as like a passing thought.
I've put all this thought into it but I've always hit a bump before succesfully implementing much so just my 2 cents basically.
[QUOTE=Darwin226;21835415]Some people like making their own engines. Why would making a game be better?[/QUOTE]
[url]http://www.gamedev.net/community/forums/topic.asp?topic_id=522990[/url]
This whole make a game not an engine thing is so banal. Can't we respect eachothers ability to decide what they're fucking making?
Ok, I'll stop making game engines because some guy posted an article about it on his blog. (sorry if I'm supposed to know who that guy is but I don't)
It's not like every modern game is based on an engine or something.
Since there isn't a strict definition of a game engine nobody can say that making one is more complex than making a game.
I bet someone would consider 2 classes that handle user input and rendering a game engine.
If someone wants to make a game and not an engine I wouldn't tell him not to, why would someone tell me not to make an engine?
[QUOTE=Darwin226;21855070]Ok, I'll stop making game engines[/QUOTE]
Multiple ones? Aren't you kind of missing the point of having one if you need so many of them
[QUOTE=blankthemuffin;21853190]tbh building a real engine from scratch would be stupidly cool, it would totally be worth it even if you never built the car.
If you want to write an engine, write an engine. If you want to write a game, write a game.
Jesus it's not that hard.[/QUOTE]
For the analogy to hold up you'd not only build the engine, but design it from scratch, and probably with limited knowledge. Probably wouldn't be very useful, and you would have no basis for improvement.
Jesus christ, stop arguing already.
Lets not waste all the pages of this version.
[QUOTE=Jawalt;21848859]Or:
I wrote my model class specifically so it'd support as many features as possible. Skeletons, texture coords, normals, etc.[/QUOTE]
And thats great and all (if you're actually using them features) but not for someone who's learning an API.
Class Model
{
// shit.. I don't know what a model consists of..
};
So if you want a quad, you draw a quad, if you want texture coords, you add texture coords, if you want normals, you add normals. Once you know how all them work, wrap it up into a class?
[editline]04:31PM[/editline]
[QUOTE=leontodd;21855843]Lets not waste all the pages of this version.[/QUOTE]
Cool things can't be posted everyday, I see no harm in having a friendly discussion about something development related. I mean.. this is a forum after all.
What about writing an engine for a sandbox game or an engine which was made from the ground up to be used by others to make game? Take for example LÖVE. Surely something like that needs its own engine to give the game developer an extensive interface to make games.
Sure but eskil isn't learning OpenGL :P
That's why you add to the class as you learn.
git is better than svn
I've decided I'm going to start learning OpenGL, seeing all this cool 3D stuff has made my mind up for me.
Wish me luck :ohdear:
[QUOTE=i300;21844140]No, Now lets see this bot![/QUOTE]
Not much to show.
[img]http://filesmelt.com/dl/Untitled1234.gif[/img]
[QUOTE=Maurice;21857399]Not much to show.
[img]http://filesmelt.com/dl/Untitled123.gif[/img][/QUOTE]
I'm interested in how you're actually communicating with the game.
Oh I'm bad at anything so I do cheap pixel reading.
[QUOTE=Maurice;21857482]Oh I'm bad at anything so I do cheap pixel reading.[/QUOTE]
It's probably the same way I'd go about it, seems fast enough anyway.
Made [URL="http://cid-a0e819e0c7d7deb6.skydrive.live.com/self.aspx/.Public/Firefly.rar"]this[/URL].
Could you guys try it and say how the controls feel?
(requires .NET 4.0)
Exit with ESC key since it traps the mouse.
[QUOTE=Darwin226;21857574]Made [URL="http://cid-a0e819e0c7d7deb6.skydrive.live.com/self.aspx/.Public/Firefly.rar"]this[/URL].
Could you guys try it and say how the controls feel?
(requires .NET 4.0)
Exit with ESC key since it traps the mouse.[/QUOTE]
I like it. Controls would be better using WASD though IMO. Also, can you tell me how you positioned the Console window right next to your game window? I always need to rearrange mine.
It will be WASD later.
I used PInvokes for the console window, here's the code.
Outside of the function:
[cpp]
[DllImport("kernel32.dll", ExactSpelling = true)]
private static extern IntPtr GetConsoleWindow();
private static IntPtr MyConsole = GetConsoleWindow();
[DllImport("user32.dll", EntryPoint = "SetWindowPos")]
public static extern IntPtr SetWindowPos(IntPtr hWnd, int hWndInsertAfter, int x, int Y, int cx, int cy, int wFlags);[/cpp]
Inside the function:
[cpp]
SetWindowPos(MyConsole, 0, 814, 0, 0, 0, 1);[/cpp]
You need to include System.Runtime.InteropServices
[QUOTE=Darwin226;21858204]It will be WASD later.
I used PInvokes for the console window, here's the code.
Outside of the function:
[cpp]
[DllImport("kernel32.dll", ExactSpelling = true)]
private static extern IntPtr GetConsoleWindow();
private static IntPtr MyConsole = GetConsoleWindow();
[DllImport("user32.dll", EntryPoint = "SetWindowPos")]
public static extern IntPtr SetWindowPos(IntPtr hWnd, int hWndInsertAfter, int x, int Y, int cx, int cy, int wFlags);[/cpp]
Inside the function:
[cpp]
SetWindowPos(MyConsole, 0, 814, 0, 0, 0, 1);[/cpp]
You need to include System.Runtime.InteropServices[/QUOTE]
Thanks, you the man.
[QUOTE=voodooattack;21857375]git is better than svn[/QUOTE]
It doesn't actually matter now because github.com takes both! (as does beanstalk.com apparently)
So now we can just argue over providers of free repositories!
Still learning C++ .Also SFML made my life easier with that C++ is much easier to learn for me :v:
[QUOTE=turb_;21853776]If you can do it just for explicit (amirite?) graphs, that would still be awesome[/QUOTE]
Actually, no, I'll do it for implicit graphs, too. I want to be able to intersect two circles and stuff.
It won't be 100% precise because it's numerical solving instead of analytical, but that's OK
[editline]03:38PM[/editline]
I also want to add differentiation of formulas. But that's for later. Won't do any expression simplification or anything, it'll depend on you entering the correct simplified expression.
Shit, I'm getting addicted to my own game :v:
[img]http://www.cubeupload.com/files/291200mclone4.png[/img]
(disabled grass at the moment, I messed it up again)
[editline]10:48PM[/editline]
Ok guys, I need some help. I have no clue how to make platformer physics for the player entity. Anyone wanna guide me in the right direction? I know how to do collisions (just go through every block, and if it intersects the player boundary box, collision has occured) but I don't know how to do the platformer velocity stuff...
Sorry, you need to Log In to post a reply to this thread.
