Invisible Sprite elements are still considered part of the parent sprites bounding box.

techdojo

I'm not sure if this is a bug or a "feature".

But if you have a parent sprite with several children, the sprites width and height reflect the extent of ALL of the children, even if one or more of them has been "hidden" with Sprite:setVisible(false) - and as such can lead to confusion with positioning, touch events (especially if you use Sprite:hitTestPoint() ) etc.

I haven't tried it yet with Sprite:getBounds() so I don't know if that exhibits the same "issue"

Personally I'd say it was a bug - but there might be a good reason why they are included, anybody care to shed some light on the issue?

atilim

Yes, this is on purpose. The main reason is mimicking Flash's behavior. Also I had another good reason of this decision but now I cannot remember what it is

techdojo

Fair enough - I raised the point just in case it wasn't or if it was causing anyone any grief.

I suppose if need's be I could always store a copy of the offending element somewhere else and actually remove it from the parent group.

alexzheng

Talking about bounding box of a sprite, will it be easy to make hitTestPoint function one step further to check the pixel at the touch point to make it support check only Opaque area of the sprite?

I'm not meaning the invisible children of the sprite.

Maybe defined as
Sprite:hitTestPoint(x, y, bOpaque)

if bOpaque is true, make further check after the orignal hitTestPoint return true.

atilim

@alexzheng yes. I was thinking exactly as your suggestion.

alexzheng

@alexzheng yes. I was thinking exactly as your suggestion.

Great minds think alike

:-))

atilim

yeah! \m/

OZApps

Maybe defined as
Sprite:hitTestPoint(x, y, bOpaque)

yes. I was thinking exactly as your suggestion.

Does that mean that one can test for the opacity of the point using the hitTest function? Like a discussion in another thread? Maybe this can also be used to create an array of points in the image that are transparent or opaque by running the hitTest function for 1..width and 1..height. Am I getting this right or is this going to be used differently?

Mells

If that's the case, that could be very useful for pathfinders and even better... alpha maps? (if the level of opacity is a number and not a boolean).
Also, does it mean that getPixel() and setPixel() 's availability would become one step closer?

techdojo

Oy - I'm the official *squeaker* here, get your own job