Traditional packet-switched networks require an end-to-end path to exist between communicating peers. Even in the case of mobility, it is assumed that the round-trip time between any two nodes be extremely short. However, such conditions don't hold for a certain number of new generation networks, including networks of portable devices carried by human users. Recent work has shown that connectivity patterns in such scenarios are different from those observed when synthetic mobility models are applied. On the other hand, some researchers have begun to think about the problem of routing in such intermittently connected networks. In this talk I motivate the necessity of incorporating the knowledge about real mobility patterns into the design of these routing protocols. At the same time, new simulation models are required in order to more accurately reflect the behavior of users in the mobile environment.