Mobile hosts in a wireless network experience highly variable network performance. The applications running on the hosts, can give better performance if they are aware of the available bandwidth. These types of systems, which are aware of the bandwidth, typically operate in various modes. The granularity at which adaptation happens is coarse. This is not always suitable. So, modeless adaptation, in which we can achieve finer granularity, becomes important. For this, Adaptive Transport Protocol (ATP) - a library for bandwidth aware communication has been developed by Benjamin Atkin, a Ph.D. student at Cornell (he is my advisor too). The key insight behind this is that, not all messages are important. When the bandwidth is low, only the important messages can be sent and when there is enough bandwidth all the messages can be sent. ATP provides a way of doing this. It can be used for prioritized communication. Each packet can be assigned a priority and a deadline. ATP delivers the messages according to the priorities of the messages in its queue. The ability to assign priorities to individual messages gives a finer control over what is important and what is not and thus the applications can adapt to changes in bandwidth effectively. Also, having this functionality at the transport level, takes off the burden from the applications to implement such functionality in them. This project aims at developing a modeless streaming video client/server using ATP and evaluating it.