Well it is trying to render the web, which is no small feat in itself given that rather creative ways in which people construct web pages. One of the more complex parts of a web crawler is the part that tries to extract from the "html" the actual content on the page :-) That some pages render at all is pretty amazing to me.
