While the Internet is hardly "broken", it has proved unable to integrate new ideas, new architectures, and provide paths for future integration of data, voice, rich media and higher reliability. The reason is that the basic concept of the Internet as an end-to-end packet delivery service has made its middle layer, networking services through TCP/IP, untouchable. If we wish to see any disruptive enhancements to security, routing flexibility and reliability, and robust quality of service guarantees in coming years, we will need to move towards an Internet in which networking environments offering differing strengths can coexist on a permanent basis. This view is gaining currency in the US, advocated by the FIND/GENI initiative [7, 6] and in Europe, where it forms the heart of the activities reviewed by ARCADIA. The ARCADIA activity, sponsored by COST [1] has been chartered to look at critical areas in which research on fundamentals in the Internet's architecture and protocols, supported by accurate experiment, can unlock some of the Internet impasses. This paper attempts to describes the insight gained and conclusions drawn from the first ARCADIA workshop on the Future of the Internet, organized around the main themes of Virtualization, Federation and Monitoring/Measurement.