ietf
[Top] [All Lists]

Re: How the IPnG effort was started

2004-11-09 05:20:33
at 17:17 08/11/2004, Aaron Falk wrote:

I'd like to suggest that this thread move to the internet-history
list.

(For those unfamiliar with this list, information is available at
http://www.postel.org/internet-history.htm)

Dear Aaron,
The work you do at www.postel.org is not only great to pay a tribute to pioneers. It gives past experience in a system where continuity is of the essence.

This threat over the past is extremely important for the future. It plainly shows something those who were involved apparently forget and other did not know.

It shows that the Internet community, as the international community had done ten years before, identified the same that: - OSI brought an incitation to some developments and called for interfacing, but was not a replacement. Consensus.
- and a large addressing space. Consensus.

But that IETF having not to directly implement it, IETF overlooked the real problems: that the most important is not the number of addresses but their structure toward an operational and innovative numbering scheme and what it permits. A recent interview of Vint in Asia shows that has now the real questions we met 20 years ago. The French equivalent to the FCC runs right now a questionnaire with the same type of concerns.

The formula "IPv6 == just IPv4 with larger addresses" now accepted by two key persons who shared in the debate 15/10 years ago is a key point. It tells what the target was.

1) as Harald put it: this target has been roughly completed. So, IETF will not change IPv6, except if there is a new round of probably the same magnitude and duration.

2) as the world shows it in not buying it: that objective was _wrong_. Truly IPv6 offers larger addresses. But thats all. Harald is wrong when he talks (as most of us did) of an IPv6 Internet. IPv6 was "just" intended to be a smarter patch to IPv4 than NATs - and it is just that with some grounds lost to NATs because of its lateness. We still are under IPv4 management and the world see no difference and feel (for the time being) that NATs deliver more.

May be Mobile Internet could be a new Internet? But Harald gave his point of view of this.

Now, where is the world? This threats shows that:

- we have 4 to 5 years to go before doom (the internet has taken such an importance that an addressing management failure is a world major problem like energy, ozone, water, etc.). I fully understand why Harald does not want to be reelected. I am interested to know who is to take the responsibility in front of the world and bare the blame (or hopefully save the day).

- 15 years ago - in this Internet technology - the problem of the addressing space blocking innovation was properly identified (you cannot innovate on imprecise grounds). It tells that it took at least 10 years to deliver the then chosen solution. The current threats pleading for IPv6 then for at least IPv4 at the Washington IETF meeting, show that all this is not yet stabilized enough (even at IPv4) to be perceived as the smart-plug solution the world decision makers think it is. So, they just focus on the namespace 'which is a minor issue when compared to the addressing).

So it roughly tells that we are going into the wall.

There are three history proven ways out:

- IAB/IETF/IESG delivers a solution in the coming months. Can we see the motivation (RFC 3774), the imagination and the consensus.

- the ITU imposes its administration in the coming one or two years. The proposition would be late in the governance study scheme of UN. The opposition would be strong so it would call for the Govs to step in. They took three or four years to respond ICANN's call (may be through the final Tunis resolution). I do not see the weak Telcos alone, even at the ITU, to be able to impose IPv6 when they failed with ENUM.

- the users (market, atlarge, Govs) impose their solutions (USA have started with the DoD decisions over IPv6, China, Korea, .... but what will be the real impact. Possible new US Communication Act will only add to confusion as such revision usually impact). Patchwork and balkanization. They will be a blend of NAT, superNAT, OPES, patch over patches, and basically a superIPv8+IPv9. We had the 1985 move to X.121, the 1996 move to Internet IPv4 addressing, the 2005 ???

I am not sure I like this.

Question: Internet is catenet's concept of a networking of networks. If I come and tell RIRs that I have a non-IPv4, non-IPv6, non conflicting address, what will they say? As a user the response will probably the difference between a worldwide clean and quick continuity and a NAT/IPv8-9-etc. Interpatch.
jfc



_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf