I just came across this blog posting by Giovanna Carofiglio, which seems to have hit the Internet today.
One of my biggest qualms with the notion of ICN/CCN/NDN, and what they’ve all been called, is that they’ve called for full-stack replacement of IP with ICN – and, if we are to learn any lessons from the 20+ year adventure of IPv4-to-IPv6 transitioning, then it is that “full-stack replacement is hard” (not to say “impossible”). RFC5218 even talks about “Incremental Deployability“, as being one of the basic success factors for initial success.
hICN, which Giovanna’s blog posting presents seems an interesting approach at integrating clever ideas from ICN into a network where they can be deployed – i.e., into any IPv6 network. Thus, the RFC5218 “incremental deployability” box ticked.
Also, really nice going, Giovanna and team at Cisco PIRL, to open-source your in-house research and upstream it to fd.io.
While I still have reservations with respect to ICN, having ticked the “incremental deployability” box, and having an open-source implementation, is a great springboard for further research – and for experiments, whose results may, of course, confirm or dispel my reservations – as is the nature of research.