Scratch

MoreGamesNow wrote:

I'm not sure why you'd need trigonometry (unless you're bouncing off of sloped walls?).  With only right angles, all you need is velocities.

when gf clicked
forever
receive inputs
change x by (xvel)
if<touching [wall v]?>
set [xvel v] to ((xvel)*[-1])
change x by (xvel)
end
change y by (yvel)
if<touching [wall v]?>
set [yvel v] to ((yvel)*[-1])
change y by (yvel)
end

Yes I am bouncing off sloped walls...
Is there a way to do this without a wall sensor, because I hate programming wall sensors...

MoreGamesNow wrote:

coolhogs wrote:

Yes I am bouncing off sloped walls...
Is there a way to do this without a wall sensor, because I hate programming wall sensors...

If you want to do it without sensing and if the walls are sloped, things could get tricky.

You'll need to know the slope of the wall you're touching first.  I'm not sure how to do that without sensing 

It's ok... I can use wall detecting... Can you give me the script?

I'll have to get the ol' paper and pencil out.  I can tell you this though: trying to approximate the slope of a wall, even with sensing, will be just that, an approximate (seriously, like 30 degrees of error).



Basically, you find the angle of the wall (the thick line) and the direction the sprite is moving (dashed line).  Then you reset the x and y velocities with the same magnitude/hypotenuse, but with values according to the new angle. To be honest, I've tried this once or twice but haven't been successful.

Last edited by MoreGamesNow (2012-04-10 13:29:21)

AtomicBawm3 wrote:

http://scratch.mit.edu/projects/AtomicBawm3/2184464

Here you go, I did this a while back...hope that helps!

This is a great way to calculate the redirection of the bounce.  But you still need to find the slope of the wall.  What I've tried (admittedly, it failed) was to change the sprite's y position and x position by ones and try to find out the slope based on how many x/y positions it can change before hitting the wall again (I hope that made sense).  Doing this all real time is, of course, a problem.  Maybe you could make some sort of "slope of wall" sensing costume?

This should help!
http://scratch.mit.edu/projects/coolhogs/2460631

If I know the direction of an object which is vertical at 90 degrees, what is the slope relative to the direction?  I'm trying to make this with a separate slope sensing sprite but it is a bit slow at the moment.

Ok, What do I do with the slope in order to get the bounce?  Something with X and Y velocity?

MoreGamesNow wrote:

Code:

(2 x (angle of wall)) - (angle of initial velocity) = (angle of final velocity)

This seems to work.  To find the new velocities, do this

point in direction (((angle of wall)*(2)) - (direction))
set [c v] to ([sqrt v] of (((xvel)*(xvel))+((yvel)*(yvel))))
set [xvel v] to ((c)*([sin v] of (new angle)))
set [yvel v] to ((c)*([cos v] of (new angle)))

I hope that works, I don't have Scratch on the computer I'm using 

What is

(new angle)

Last edited by LiquidMetal (2012-04-12 15:44:51)

It is not working!

See it here.

EDIT: Wait is angle of wall the slope or the direction?
EDIT: Maybe because I did not put in the rotation thing?

Last edited by LiquidMetal (2012-04-12 17:28:03)

Here you go.  Press SPACE to render the walls (Flash player wouldn't detect the walls when they were rendered every frame     ).

coolhogs wrote:

MoreGamesNow wrote:

Here you go.  Press SPACE to render the walls (Flash player wouldn't detect the walls when they were rendered every frame     ).

no like a platformer...

I just remixed LiquidMetal's project.  Do you want me to add gravity?

@LiquidMetal: do you have a separate question from coolhogs' question?  If so, it's better to start your own thread than to piggy back onto another one     I was just horribly confused; I didn't realize this thread had evolved into two different questions xD.

Try my link again.  Should be good now