--On Monday, 16 June, 2003 21:39 -0400 Keith Moore <moore@cs.utk.edu> wrote:
The think I find mindboggling about all this is that I have yet to see a concise explanation of how the great transition to IPv6 is going to be managed and what the incentive for early adopters is going to be.
There isn't going to be a great transition to IPv6 in the sense that you seem to mean. IPv4 and IPv6 will coexist for a long time. The most popular IPv4 applications - web and email - will be the last to abandon IPv4, and they won't do so until IPv6 is ubiquitous.
The incentive for IPv6 adopters is obvious - they'll use IPv6 to do things they cannot do with IPv4. Those things include: deploying lots of distinctly addressible pieces of hardware (e.g. things that get monitored or controlled remotely), alleviating an actual or imposed, global or local, shortage of IPv4 addresses (this applies both to China and home networks), talking to IPv6-only devices (that use IPv6 because they cannot reliably get enough IPv4 addresses), and any apps that cannot reliably operate through NAT. ...
And, if IETF gets its act sufficiently together early enough, the capability for someone who doesn't qualify for PI space under current IPv4 rules to do multihoming -- perhaps to compensate for two incompetent ISPs by hoping they won't be out of service at the same time-- at plausible total costs. My own guess is that, as reliable Internet connections become more important to people, and if the trend toward lowering skill levels at ISPs continues, small enterprise and SOHO multihoming may turn out to be one of the driving applications for IPv6. If we get our act sufficiently together...
john