Now I realized I have seen a naturally occurring test case: Twitter user pages with and without static background images. Everything else is the same, yet the ones with the static background image (CSS => background-attachment:fixed) are much slower.
Now I realized I have seen a naturally occurring test case: Twitter user pages with and without static background images. Everything else is the same, yet the ones with the static background image (CSS => background-attachment:fixed) are much slower.