My latest project for work was about urls and how to parse them (an issue of longstanding interest to me). One thing that I found remarkable was how few of the domain names I was working with involved ambiguities that were likely to create parsing problems in practice; in most ambiguous cases one option was so much more likely that for practical purposes the other could be disregarded entirely. The unlikely options could be quite interesting, however. This site, for instance, could have been about congressional attitudes toward sex crimes (or something like that), and this one could have been about the seamy underbelly of online dating. Unfortunately, the actual sites at those locations are considerably less intriguing.
