I was just thinking that the Lempel–Ziv–Welch basically finds patterns and creates a dictionary of unique data strings. Which is what one would need to do if one was to post process ID's to Classes. CSS is several orders of magnitude more complex than images but computers are about 7 orders of magnitude faster than they were in 1984. The optimizing of css might just be a compression problem.:-) That and of course master pages like muse to create classes meant to be shared.
↧