BlendMode.LAYER is still not working (right?), and I was wondering how you bypass this limitation when applying alpha to composite/nested display objects?
If you don’t know this BlendMode, well, LAYER is(/was) essential when you have to apply an alpha < 1 to an object containing more than one display object, otherwise the alpha is applied to each sub-object, revealing the overlapping areas. BlendMode.LAYER allows to automatically bake/prerender all the items into a single display objects that doesn’t suffer the described problem, but I don’t know how it was handled in memory in AS3.
A way to bypass this limitation is to create a new bitmap and bake all items in it, but it is memory-consuming and it is problematic with animations.
O.T.: the concept is similar to cacheAsBitmap, I have just realized that the underlying implementation could be very similar.
P.S.: test done, cacheAsBitmap fixes the problem, but I wonder if it works with transformations too.
P.P.S.: test done, cacheAsBitmap fixes the problem with transformations applied too, it is exactly like BlendMode.LAYER! I’m not sure that the behavior of cacheAsBitmap in AS3 was the same…