[cairo] Text scaling and hinting

[Jon Smirl]

I believe MS Avalon is animating by drawing the text into a very large
sized monochrome bitmap that is expanded 3x in width. Each step of the
animation hardware transforms and filters the large bitmap to the
desired size without redrawing the text. The 3x monochrome width is
used to achieve subpixel rendering when converted to RGB. There is a
paper on the MS site about this that I can help locate if you are
interested. The scheme should work for animations but it isn't too
good for static text.

-- 
Jon Smirl
jonsmirl at gmail.com