I am porting as3iso to use with my abstract display list but I fail to see the reason why the author decided to use a secondary display list (the node hierarchy). I cannot think of the necessity of this apart from being able to get the children array of a display object. Is it really that big a performance hit to do for(i in 0…numChildren) getChildAt vs getting the children array? This can also be mitigated by maintaining a children array while overriding.
The only other reason is that some properties of flash.display.DisplayObject are final and wanting to have x and y correlate to isoX/isoY, but worldX/worldY is a reasonable API for this.
I have no problem with the secondary display list but I fail to see the point and am leaning toward having my iso display objects extend from my DisplayObject.
Also, I had presumed that marking objects visible=false that are not on the screen or outside of a scrollRect is not necessary with Flash and in rendering in general as this is/should be handled at the renderer level? Is this the case with the flash software renderer?
This way the library doesn’t rely on the rendering system. You can use the traditional flash display list, Stage3D or a custom solution.