Hi Thomas, I tried your approach. Much simpler than mine, indeed, but you actually lose the dynamic behavior of the interactors. Only after the mouse is released, the other layers are transformed.
In my approach, although you manage to keep the dynamic behavior for all canvases, synchronization is not perfect, meaning that there's a little lag in the repaint time. This is not at all unexpected, since I basically iterate in a list of canvases calling each one's transformation method. Maybe threading would alleviate this, but I doubt that it would be sufficient to keep the synchronization I want. So, I'm moving on to the overlay idea. I'm thinking, to do that I'll have to rasterize the svgs first, right? And then, in the paint() method of the original canvas, get the rendering transform of the canvas and use it to set the transform on the raster. Does that sound ok? More coding ahead... --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
