The Unicon programming language is twelve years old. This talk will present some of the things learned during Unicon's development, illustrate some of the current challenges, and discuss some of the research on Unicon that is likely to be coming down the pike in the nearer and farther terms. This applied research is often challenged to invent ways to scale the implementation to larger programs and larger data sets, and to port the language to new platforms and new types of I/O hardware while maximizing backward compatibility.