Simple question - how is text-zooming implemented on an Android/iPhone device? Do they pre-compute frequently used bitmaps of a font and replace the text as the scale changes? Or do they extract the contours from the font files and render the text as vector graphics?