The operational cost of supporting both v4 and v6 from the network perspective not great, based on our experience (although the support load for v6 clients to content hosted in our network is currently much lower than for v4 clients, as you'd expect).
I'd be very happy to share more details about what we're running with people who have interest.
Gentlemen,
let agree IETF is lacking formal interfaces with the real world of users and the real world of operators. John Klensin's official participation to the ICANN BoD is a first good step towards formal links with operators. However, the lack of relation of ICANN with users (@large) and the US oriented nature of ICANN (the problem is not the USA but a single country), leads world towards ITU. I support the principle of this move as a multilateralism American structures have some difficulties with, but not a move towards ITU-T. We need an ITU-I where IETF may very well fit through a good and fruitful MoU or more.
Each "sector" (ICANN/NICs, @large, IETF/IAB, ITU/GAC) in this chain suffers from its imperfections and its lack of cooperation. For too long each said "the fault is with the other". IMHO the fault is with every of us and we all are to cooperate. When ICANN started the ERC process, I made sure in calling on everyone (Vin may recall that he contributed) that there was a consensus about the need to reform ICANN (read Governance) and that the solution had to be consensual to succeed. The solution is not consensual and we saw Geneva WSIS positions as a result. We now have 2 years to correct that, before UN says "I take over".
White House identified the weaknesses of the system too. One of their answers is IPv6. http://whitehouse.gov/pcipb asks all the US administrations to move to IPv6. And DoD is obeying the Chief. In France, the Research Ministry switched to IPv6 and the research network for France (Renater) is the IPv6 leader (the same for Europe (Geant) as far as I understand). We all know what Asia is doing for IPv6. So, IPv6 could be a uniting factor to reorganize the Governance, the "Technicance" and the Brainware (the ways users use the system) in a concerted way, instead of bickering each other.
As a user/brainware person for a long, I see several issues where I need the Technicance and the Governance to help. Not to commit anyone I will say a popous "I", but it is actually several "we"s I currently see(share in) structuring
1. I am not technical enough on this to comment about IPv6 as an architecture and as a system. But I do not trust a system the architecture of which is to be dig into hundredth of RFCs and confused closed discussion lists. I need first an Internet architecture document (equivalent to what the W3C published) to understand what is the Internet network system application and how it relates with other network system applications such as telephone, television, OSI, etc. so I am sure everyone understands what we want to do together.
2. I am not a government but I have a IQ high enough to understand that Govs will not buy IPv6 the way it is proposed today for two main reasons.
- IPv6.001 (001 numbering plan) is unique and IPv6 is to support up to 6 numbering plans. If IPv6 is to be deployed (real life tested) with only one numbering plan, we will face a Y2K potential threat since no one will be able to warranty the world a new numbering plan will be compatible with the then existing equipment base or even if IPv6 can support it. Nor that it will not endanger existing world operations. We have that exact example with the DNS. To introduce a new TLD is considered as a potential technical threat (while it should be able to support millions of it).
- IPv6.001 is flat. Both in routing and in addressing (and confuses both - some wanting to add identification/authetification). As such it offers no control nor surety to countries. Also, it creates many identification, bandwidth, recovery and economical problems.
I need a second numbering plan (IPv6.010) to be accepted in the RFCs and discussed. And I do propose this plan to be defined and maintained by ITU. For several reasons. (1) to enforce the ITU-I concept and protect us from the Telcos [to aggregate the Telephone and X.121 numbering plan the ITU-I will have to talk with ITU-T] (2) to get the support of the Govs and show them what is possible and what is not in term of international Technicance. (3) to get a good PR for IPv6. The TF IPv6 submitted a document to the French Research Minister to promote IPv6: I tried to make them understand that technical arguments were low interest, what was of interest was to propose tax cuts for corporations switching to IPv6 to compensate for the equipment upgrade/change. If Govs are involved, they will understand better why to advertise IPv6 and to support to the change.
Also, I miss words like dynamic routing, anonymous network presence, domotic/teleurbanism support, service/addressing continuity support, mobile identification, intelligent ubiquity support and I know the societal demand is for them. If they are not addressed, I able also able to understand there will be no funding for R&D programs by Europe or others. I am therefore not excited as this will not take off quick and I suppose that in the meanwhile someone may comes with an NGN universal addressing plan (COULD have been IPv6.010).
3. as user I find IPv6 totally unfriendly. The main reason why IPv4 will stay around IMHO is that I can manually use any ASCII editor and manage a DNS db file. The semantic of an IPv6 address is user unfriendly and its support obscure to me as a user. I want a 0-Z numbering plan first, to be able to quote an IPv6 address on the phone and an explanation manual in three pages that everyone will understand. If it is needed it must be simply understood and accepted.
Please give me a one page explaining why Music Lovers can universally take advantage from an IPv6 oriented TLD in a way where stakeholders may share in the control (through the naming used?) and I will tell you how we can quick make IPv6 ... universal.
We are in a real life world. Lobbying is of use towards Govs. and we (all of us together) need some. Only adhesion, interest, personal advantage can influence the users brainware. An IPv6 mandatory Napster is what we need.
Short of that IPv6 may never take off (what for? tell your kid why they need IPv6 and copy us. Here is what people may wantto listen to).
jfc