A large number of streaming applications use reliable transport protocols such as TCP to deliver content over the Internet. However, head-of-line blocking due to packet loss recovery can often result in unwanted behavior and poor application layer performance. Transport layer coding can help mitigate this issue by helping to recover from lost packets without waiting for retransmissions. We consider the use of an on-line network code that inserts coded packets at strategic locations within the underlying packet stream. If retransmissions are necessary, additional coding packets are transmitted to ensure the receiver's ability to decode. An analysis of this scheme is provided that helps determine both the expected in-order packet delivery delay and its variance. Numerical results are then used to determine when and how many coded packets should be inserted into the packet stream, in addition to determining the trade-offs between reducing the in-order delay and the achievable rate. The analytical results are finally compared with experimental results to provide insight into how to minimize the delay of existing transport layer protocols.