TILTBALL!
Tiltball is my first game for the calc, written in axe.  

I asked a question for the math of this game earlier, but now i realize that i will probably need to ask more questions...

a brief overview of the game:

The ball falls from the cieling, due to gravity. the goal is to get the ball to the floor, without hitting the walls. There will be lines in the way of the ball, which the ball bounces off of. The only thing you can control is the rotation of the lines, as shown in the pictures below.

The original level:

Now, after you press the -> arrow on the keyboard:

all of the lines have been rotated counterclockwise around the center point.

That is the general idea.

I have the rotation of the lines worked out, but still have two of questions:

1. Is there any way to clip the lines without sacrificing too much speed?
2. How to calculate angle of bounce? so if the ball is coming in at velocity of xv, yv, and slope of the line m, how to bounce off?

Thanks in advance!

Lol me too!

Looks nice. A solution to the collision issues would be to make the ball move slower and make the lines twice as thick (probably by drawing 2 copies of them, the second copy being offset by 1 pixel below. Maybe add a 3rd copy offset 1 pixel to the right or left so lines remains thick even if rotated at 90° angle or something.

One thing to make sure is that the ball cannot fall at a speed higher than 1 pixel per frame. If speed becomes an issue then, just refresh the LCD content only every 3-4 frame.

Another question...
How would you go about doing trigonometric functions in axe, like tan^-1() or sin^-1(), to calculate the reverse of the tangent or sine of an angle. I looked all over the internet, but all I could find was that tan(X) = sin(X)/cos(X).

Well, actually collision testing is done by comparing the three points (first line point, second line point, and ball location) and testing if they are collinear, not by pxl-checking. but the 3-4 ticks before refresh is a good idea!

Actually, I think that pxl-check() is way faster. And drawing the screen every 4 frames saves more than 4 ticks I think O.o

I ended up posting in the Axe section (thank you DJ_O), and I did get a response, from Quigibo. I'll quote it here in case anyone wants to see the formula.

A surprisingly accurate approximation for arctangent is:

Code: [Select]

atan(x) ≈ (π/2)*x/(abs(x)+1)
Which, when converted from radians, can be implemented in fixed point notation as x*64//(abs(x)+256) however this has overflow problems in the numerator.  The best way to fix this issue is to split the equation into 2, one for large numbers, and one for small:

Code: [Select]

:Lbl atan
:If abs(r1)+256→r2>768
:  Return 64-(16384//r2)
:End
:Return r1*64//r2