At 23:46 27/11/2008, Olivier MJ Crepin-Leblond wrote:
To tell you the truth, I wasn't quite sure about the initial reasons for the heated debate between the people for NAT66 and the people against NAT66. I took 24h to think about it and for a moment, came to the same conclusions about multiple DNS roots & multiple IPv6 realms. With NAT66, the concept of network number uniqueness which was a key proposed feature of IPv6, disappears.
To tell you the truth I did not yet made my mind about what they exactly have in mind and what the implications. From what I grasp, they want this to be actually assessed? So, my priority is to set-up a dialog space between IETF and Internet users and the IETF Internet standard process. I think it is pretty urgent because with IPv6, DNSSEC, etc. i feel that the polylogue between all the concerned parties become totally out of reasonable control.
Then I thought again - and was taken back to the days when the Internet was not as well integrated as it is today. You know, the days when you needed to route your email through gateways & the like. Sure, there were many other networks out there, but you know what I found to be most amusing is that historically they all merged with the Internet.
There certainly were cons and pros. We should try first to seat and not to duplicate the cons and focus on the pros. But there is such a legacy ...
I am therefore neither "for" or "against" the idea of NAT66. I've seen several alternative DNS roots rise and fall. If organisations wish to launch their own Internet, their own root servers, their own interfacing with the current Internet through IPv6 NAT, sure, go ahead! Let the market decide what's best.
I do not think we are actually engaged in any "alternative DNS" (there never were alternative DNS except in IETF/ICANN fears and Vint Cerf ".biz" trick) but a progressive understanding of the meaning of ICANN ICP-3. france@large engaged into the community "dot-root" test-bed, respecting the ICP-3 request. Our conclusion is that there is a unique virtual root matrix to be supported that corresponds to the distributed nature of the Internet. Also, that the current IETF vision is not ready to accept and support it, but that the technology can probably support it, as well as a progressive transition. But all that must now be worked on, discussed, proposed, verified, documented, tested along new ways of thinking adapted to our present time. This is NOT that easy! The problem is in the people heads, more than in the machines and protocols.
Just: 1. don't ask me to pay for it, whether through my taxes or whatever
Correct. This is why we must stop the IPv6 political campaigning and replace it by s self-financing approach. I think this is possible. But the first thing is to convince ISOC that an ISOC IPv6 Users Chapter would really help discuss this. The IPv6 TF supports the idea. ISOC is afraid. The motto should be "IPv6 what for?" instead of "someone must pay for IPv6"
2. don't restrict the freedom I currently have in using today's Internet
I am afraid that IETF is currently precisly doing it unwillingly. This is where we need to help. Multitier Internet threat is everywhere. IMHO we can protect our freedom, but we need to understand what is our _user_ architecture. IMHO this is really what @larges should be at. Anywhay, I hope my IETF list will modestly help.
3. don't impose your alternative network on me
I do not think there is any alternative nework. There is a mess. Who really knows today what IPv6 is? Margaret Wasserman and Fred Baker try to specify 66NATs against a 66HyperNATs diversity. It is probably a good idea in order to smooth the transition. However, what they probably show is that the very nature of the Internet is to be an InterNAT (however the word NAT is just for one of the functions to be required on Smart "Interboxes").
4. don't take me back to the dark days of telco monopoly
:-) I was the one who started to field unlock monopolies. I was very happy the day I obtained the first "triangle", where traffic from three monopolies were mixed - together with private traffic (US, Belgium, Netherlands and ESA). That was more than 25 years ago. I did not expect the US deregulation to create something worst than telco monopolistic centralisation. I hope we can free the Internet from the US consortia decentralisation before I die :-) jfc