Get Actual Collision X,y?

Physics
#21

I guess this was one I had converted from a Runtime listener, as the instantiation of the listener is this:

bg2:addEventListener( "collision", onEm3Collision )

Even if I convert it to this:

bg2.collision = onEm3Collision bg2:addEventListener( "collision", bg2 )

and change the listener function definition to this:

local function onEm3Collision( self, event )

I still get event.x and event.y values that don’t correspond to the global coordinates.

I believe I saw similar behavior with the Runtime listener, but I will have to go back and try it again.

#22

Which build has this?  The build I am currently using 2013.1135 always returns 0 for event.x and event.y.

#23

since brent’s post was in april and 1135 was released in june i would say its in 1135

#24

Hi @tonygod,

Can you post your collision code? Did you remember to add this line around where you instantiate physics?

[lua]

physics.setReportCollisionsInContentCoordinates( true )

[/lua]

http://docs.coronalabs.com/api/library/physics/setReportCollisionsInContentCoordinates.html

Brent

#25 
local function onEm3Collision( event ) print( "onEm3Collision: " .. event.phase ) if ( event.phase == "began" ) then audio.play( sounds["splash"] ) -- TRIGGER THE EMITTERS em3.x = event.x em3.y = event.y Particles.StartEmitter("em3", true) elseif ( event.phase == "ended" ) then bg2:removeEventListener( "collision", onEm3Collision ) waitingForSplash = false end end

I was not previously using physics.setReportCollisionsInContentCoordinates( true ), as I was unaware it was necessary.  The documentation merely states “The x and y position can be influenced by physics.getAverageCollisionPositions() and physics.setReportCollisionsInContentCoordinates().”  

I don’t know what influenced means. For me, it seems to mean I get event.x and event.y coordinates that do not match what I would expect to be the global x/y coordinates when I set it to true and I get 0 when it is not there.

Maybe the fact that I am using scenes or DisplayGroups is affecting this in some way?

#26

Hi @tonygod,

At first glance, it appears you’re not utilizing the proper parameters in the collision listener. Since this is a “local” collision listener (not a Runtime one), the listener supports both “self” and “event”… you’re using just “event”, but that should be the second additional parameter.

Please see the guide on “Local collision listeners” here:

https://developer.coronalabs.com/content/game-edition-collision-detection

#27

I guess this was one I had converted from a Runtime listener, as the instantiation of the listener is this:

bg2:addEventListener( "collision", onEm3Collision )

Even if I convert it to this:

bg2.collision = onEm3Collision bg2:addEventListener( "collision", bg2 )

and change the listener function definition to this:

local function onEm3Collision( self, event )

I still get event.x and event.y values that don’t correspond to the global coordinates.

I believe I saw similar behavior with the Runtime listener, but I will have to go back and try it again.

#28

Any followup re gottcha’s here?

I’m getting 0’s returned also for event.x, event.y, but I note the local object is within larger display groups.  I have:

* called “setReportCollisionsInContentCoordinates” after physics start, although from the sample physics tutorial “CollisionDetection” you don’t seem to need it

* have followed the same layout for setting up the local listener

    function myCollList (self, event)        print("x/y: ", event.x, event.y)      end      myObject.collision = myCollList     myObject:addEventListener( "collision", myObject )  

* noted that the “event.target” and “event.other” return what I expect fine.  It’s just the event.x and event.y that are giving zero’s back

#29

I could not get this to reliably return correct coordinates for event.x and event.y and just ended up using x and y coordinates for one of the objects.

#30

Hi Greg, Tony,

I just tested this and it appears to work fine. The collision point of the falling box is accurate in my tests.

[lua]

physics.setReportCollisionsInContentCoordinates( true )

local function myCollList( self, event )

    if (event.phase == “began”) then

        print( event.x,event.y )

    end

end

local t = display.newRect( 100,0,50,50)

physics.addBody( t, “dynamic” )

t.rotation = 45

t.collision = myCollList

t:addEventListener( “collision”, t )

local tx = display.newRect( 100,200,50,50)

physics.addBody( tx, “static” )

print(tx.x,tx.y)

[/lua]

Brent

#31

Worked in a basic test for me too. However in my app where I have embedded display objects within groups etc it wasn’t working. I’ve just worked around by using “target” and “other” position info

#32

Hi Greg,

I assume you weren’t moving around the display groups independently of each other?

#33

It was a player object hitting an obstacle object both within a larger level display group. The larger level group gets moved, horizontal scrolling game, but I think in the case I had it wasn’t being moved at the time. Hard to track something like this down when the function seems to work in the most basic scenario.

#34

Hi Greg,

I just tried the sample that I wrote in nested display groups, and it works the same (correctly). So, I think there must be some other issue here beyond the Corona APIs. Let me know if you figure it out…

Brent

#35

Ok. Spent of bit of time already so might need to move on with the work around for the moment.

#36

Any followup re gottcha’s here?

I’m getting 0’s returned also for event.x, event.y, but I note the local object is within larger display groups.  I have:

* called “setReportCollisionsInContentCoordinates” after physics start, although from the sample physics tutorial “CollisionDetection” you don’t seem to need it

* have followed the same layout for setting up the local listener

    function myCollList (self, event)        print("x/y: ", event.x, event.y)      end      myObject.collision = myCollList     myObject:addEventListener( "collision", myObject )  

* noted that the “event.target” and “event.other” return what I expect fine.  It’s just the event.x and event.y that are giving zero’s back

#37

I could not get this to reliably return correct coordinates for event.x and event.y and just ended up using x and y coordinates for one of the objects.

#38

Hi Greg, Tony,

I just tested this and it appears to work fine. The collision point of the falling box is accurate in my tests.

[lua]

physics.setReportCollisionsInContentCoordinates( true )

local function myCollList( self, event )

    if (event.phase == “began”) then

        print( event.x,event.y )

    end

end

local t = display.newRect( 100,0,50,50)

physics.addBody( t, “dynamic” )

t.rotation = 45

t.collision = myCollList

t:addEventListener( “collision”, t )

local tx = display.newRect( 100,200,50,50)

physics.addBody( tx, “static” )

print(tx.x,tx.y)

[/lua]

Brent

#39

Worked in a basic test for me too. However in my app where I have embedded display objects within groups etc it wasn’t working. I’ve just worked around by using “target” and “other” position info

#40

Hi Greg,

I assume you weren’t moving around the display groups independently of each other?