Box2d Bodies that are really close together, are getting "stuck" - java

Im using a tiled tmx map, and I created a class that adds bodies to each tile within a certain layer. This has been working great so far except for when a character or an enemy moves around on the screen, its body gets stuck on an edge between two tiles.
This seems to happen "sometimes" and in certain spots. Jumping makes u unstuck but its annoying when it happens, and i tried increasing the position iterations, but the problem keeps reoccurring.
Heres what my game looks like: http://i.stack.imgur.com/f5Igm.png
I didnt render the tiles so that its easier to see what the tile's body looks like

What happens is that an upper dynamic body's "skin" can get embedded into the shapes below it. If the upper body is then moved across the lower shapes and encounters a corner of one of those shapes, then an impulse is generated by the physics engine that's in opposition to the direction of travel.
Boom! Upper body gets stuck.
Here's a zoomed in image showing this for an upper rectangle moving to the right:
The blue dots and lines extending from them are where the Box2D manifold calculation code has determined impulses need to be applied. Note the blue dots and lines that are on the right side of the reddish/brown rectangle. Those are opposing impulses.
A solution, as Colonel Thirty Two suggested in the comments, is to use the "ghost-vertices" mechanism in edge shapes for the lower shapes (or just use a chain shape that effectively calculates the edges for you).
As background, I have my own fork of Box2D that I've been developing and I just pushed out an alternative solution for this very problem that doesn't require the lower shapes to be edge (or chain) shapes. I've also done a write-up of this alternative.

Related

Draw curved custom object in LIBGDX?

I've recently been looking into LibGDX and seem to have hit a wall, seen in the picture, the blue dot represents the users finger, the map generation it self is where i seem to get stuck, does LibGDX provide a method of dynamically drawing curved objects? I could simply generate them myself as images but then the image is hugely stretched to the point of the gap for the finger can fit 3! But also would need to be 1000's of PX tall to accommodate the whole level design.
Is it such that i should be drawing hundreds of polygons close together to make a curved line?
On a side not i'll need a way of determining when the object has from bottom to top so i can generate another 'chunk' of map.
You don't need hundreds of polygons to make a curve like you drew. You could get away with 40 quads on the left, and 40 on the right, and it would look pretty smooth. Raise that to 100 on each side and it will look almost perfectly smooth, and no modern device is going to have any trouble running that at 60fps.
You could use the Mesh class to generate a procedural mesh for each side. You can make the mesh stay in one spot, locked to the camera, and modify it's vertices and UVs to make it look like you are panning down an infinitely long corridor. This will take a fair amount of math up front but should be smooth sailing once you have that down.
Basically, your level design could be based on some kind of equation that takes Y offset as an input. Or it could be a long array of offsets, and you could use a spline equation or linear equation to interpolate between them. The output would be the UV and X coordinates which can be used to update each of the vertices of your two meshes.
You can use the vertex shader to efficiently update the UV coordinates, using a constant offset uniform parameter that you update each frame. That way you don't have to move UV data to the GPU every frame.
For the vertex positions, use your Mesh's underlying float[] and call setVertices() each frame to update it. Info here.
Actually, it might look better if you leave the UV's and the X positions alone, and just scroll the Y positions up. Keep a couple quads of padding off top and bottom of screen, and just move the top quad to the bottom after it scrolls off screen.
How about creating a set of curved forms that can be put together variably. Like the gap in the middle will at the top and bottom of each image be in the middle (with the same curvature at end and beginning points)...
And inbetween the start and end points you can go crazy on the shape.
And finally, you can randomly put those images together and get an endless world.
If you don't want to stop in the middle each time, you could also have like three entry and exit points (left, middle, right)... and after an image that ends left, you of course need to add an image that starts left, but might end somewhere else...

3D voxel angled plane

I'm trying to draw a flat surface out of voxels, the goal is to draw it filled and I'm having a lot of trouble. Everything I try results in holes on the surface. The surface has 4 corners, but I'd like to be able to use the same method for triangles too.
Here's what I've tried:
Draw along from one parallel side to the other
Draw only in one direction (z direction) along a side of the plane
I've had the most success with 2 but it fails when I add any pitch or roll to the plane (any elevation present).
Any tips? There's no code because I'm sure my implementations are all correct, it's just the choice of algorithm that's wrong.
EDIT:
On a side note, though number 2 had less holes, the planes were distorted and didn't appear flat.
EDIT2:
I'm sticking with my first decision, but now the question is, how do I detect when there will be a hole? By observation I notice there's the same amount of holes per plane regardless of pitch and roll. Yaw is the culprit here.
EDIT3:
I'm leaving this question up but I decided to just test a nearby block to see if it's empty. I didn't want to do it, but yeah. If you have a more elegant solution I'm all ears.
A plane, being infinite, does not have corners. Are you talking about a four-sided polygon? Does it have square corners?
For a polygon, I would certainly start off with a triangle, since you can construct any other polygon out of triangles, not the other way around.
Then, a good start for filling a triangle would probably be to come up with an accurate test of whether a given voxel should be filled or not. Here's an example of two different point-in-triangle tests.
After you have that you can proceed in different ways. For example, although not the most efficient, you could region-grow from the center, testing each neighboring voxel and recursing with a stack.

Android opengl vertex array

Simple question that i've always pondered about. When I first got into opengl I had to find a way to draw "tiles" or a bunch of triangles with 1 opengl draw call to improve performance vastly. I did that by putting all the vertices into an array and drawing the array. One problem that occurred was whenever I scrolled the tiles, I would see random placed pixels showing on and off depending on how much I scrolled the map. Around 5-8 on a 10x10 map. Now, I got back into opengl again and and this time I drew using GL_LINE_LOOP instead of GL_TRiANGLES. This never occurred to me back then but what I see when using GL_LINE_LOOP is all the triangles but when it finishes(it goes from bottom left to top right, so top right) there is a line connecting where I ended and where I began. Would the cause of those random pixels be because of this? Or does this have nothing to do with it. Does that line connecting the end and beginning appear because of GL_LINE_LOOP mode or does that also have nothing to do with and have to do with the way I created the map?
GL_LINE_LOOP is a completely different drawing mode - and no that won't be the reason why you're dropping pixels on tiles.
Most likely reason for dropping pixels when drawing a tight mesh of tiles is that you're not computing the vertices consistently. The principle is that if a vertex is shared by two adjacent triangles (or quads, lines, whatever), then the floating point coordinates of that vertex must be 100% identical for every draw call. If you do that, then you're guaranteed to render a tight mesh without any gaps in between tiles.
Your problem may have been something else though... perhaps Z fighting.. but my guess is the first thing I mention here.

Vertices selection and state of model after rotation

I'm currently writing an application that actually acts as a "cut" tool for 3D meshes. Well, I had some problems with it now which I am clueless on how to solve, since it is my first application.
I have loaded a model from an object file onto the canvas, then on the same canvas, I use the mouse drag event to draw lines to define the cutting point.
Let us say I want to cut a ball into half and I draw the line in the middle. How do I detect the vertices of the ball under the line.
Secondly, if I rotate/translate the ball, would all the the vertices information change?
Think of what you'd do in the real world: You can't cut a ball with a line, you must use a knife (a line has no volume). To cut the ball, you must move the knife through the ball.
So what you're looking after is a plane, not a line. To get such a plane, you must use some 3D math. What you have is the canvas orientation and the "side view" of the plane (which looks like a line).
So the plane you're looking for is perpendicular to the canvas. A simple way to get such a plane is to take the canvas orientation and create a plane which has the same orientation and then rotate the plane around the line by 90°.
After that, you can visit all edges of your model and determine on which side of the plane they are. For this, determine on which side of the plane the end points of the edge are. Use the cross product. If they are on the same side (both results of the cross products will have the same sign), you can ignore the edge. Otherwise, you need to determine the intersection point of the edge and plane. Create new edges and connect them accordingly.
See this page for some background on the math. But you should find some helper methods for all this in your opengl library.
if I rotate / translate the ball, would all the the vertices information change
Of course.
It's not going to be that easy.
I assume the line you are drawing induces a plane which then cuts the sphere.
To do so, you have to calculate the intersecting area of the sphere and the plane.
This is not a trivial task and I suggest using an existing framework for this or if you really want to do this yourself, read about basic intersection problems to get a feeling for this kind of problem. This paper offers a good introduction to various intersection tests.
In general boundary represended volumes, as in your case, are difficult to handle when it comes to more advanced manipulations. Cutting a sphere in half is easy compared to burring a small hole into it. Sometimes it's better to use a volume representation, like tetrahedral meshes or CSG.
Regarding your second question, you shouldn't rotate or translate the sphere, rotate and translate the camera.

Pacman maze in Java

So I'm building the pacman game in Java to teach myself game programming.
I have the basic game window with the pacman sprite and the ghost sprites drawn, the pacman moves with the arrow keys, doesn't move beyond the walls of the window, etc. Now I'm trying to build the maze, as in this picture:
Without giving me the direct/complete solution to this, can someone guide me as to how this can be built? I'm talking only about the boundaries and the pipes('T' marks) here which you can't go through and you have to go around. Not the dots which the pacman eats yet.
Here are my questions:
1) What's the most efficient algorithm/method for creating this maze? Will it have to be drawn each time the paint() method is called or is there a way to draw it only at the start of the game and never again?
2) How will this actually be drawn to the screen? I assume the fillRect() will be used?
3) Any hints on collision detection (so the pacman/ghosts can't go through the walls) would be helpful.
4) Any hints on how the vacant space between the pipes will be calculated so the dots can be filled between them will also be very helpful.
Thanks
I wouldn't do it that way.
I'd draw the graphical map and then create a 2D data array which represents the map. The data map would be responsible for determining collisions, eating dots, where candy is and where the ghosts are. Once all the logic for everything is handled just use the 2D array to display everything in their proper pixel coordinates over the graphical map.
For example the user is pressing the left key. First you determine that pacman is at element 3, 3. Element 3, 2 contains information denoting a wall so you can implement the code to make him ignore the command.
EDIT:
Each element would represent about where a dot could be. For example:
No, looking at the board I would say the array would look something like this.
d,d,d,d,d,d,d,d,d,d,d,d,w,w,d,d,d,d,d,d,d,d,d,d,d,d
d,w,w,w,w,d,w,w,w,w,w,d,w,w,d,w,w,w,w,w,d,w,w,w,w,d
p,w,w,w,w,d,w,w,w,w,w,d,w,w,d,w,w,w,w,w,d,w,w,w,w,p
d,w,w,w,w,d,w,w,w,w,w,d,w,w,d,w,w,w,w,w,d,w,w,w,w,d
d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d,d
And so on. You might want to pick a more flexible data structure than just characters however since some areas need to contain a bunch of different information. IE even though the ghost spawning area is blank, pacman isn't allowed in there. The movement of the ghosts and pacman is different for the side escapes, the candy spawn point is a blank spot but if you want to remain flexible you'll want to denote where this is on a per map basis.
Another thing you'll want to remember is that pacman and the ghosts are often inbetween points so containing information that represents a percentage of a space they're taking up between 1,2 and 1,3 is important for collision detection as well as determining when you want to remove dots, powerups and candy from the board.
You can paint the map into a BufferedImage and just drawImage that on every paint(). You'll get quite reasonable performance this way.
If you are happy with the walls being solid, you can draw each square wall block with fillRect. If you wish to get the same look as in the picture, you need to figure how to draw the lines in the right way and use arcs for corners.
The Pacman game map is made of squares and Pacman and the ghosts always move from one square to the neighbouring square in an animated step (i.e. you press right, the pacman moves one square to the right). That means that collision detection is easy: simply don't allow moves to squares that are not empty.
I do not understand what you are trying to ask here.
1) Just to give my advice on redrawing. Something that you can do if you find redrawing the entire image is slow, is determine only the elements that have changed on the screen and redraw those. An approach for this would be the following: Determine the sprites that have moved. Determine (approximate) a rectangle around those sprites. Redraw those rectangles only. This way you are only refreshing parts of the screen and not the whole screen. This should result in an increase in performance over redrawing the entire screen.
The other answers have been reasonable for the other questions you have asked.

Categories

Resources