I've been watching this debate for a while, and all the previous
debates every time this subject comes up.
IETF is no longer a single subject group. Very few participants are
up to date with everything that IETF is working on, and many newer
participants are participating in what I would characterize as "edge"
areas: capwap, manet, and the like.
I suspect (and I fall into this category) that many of these newer
participants are not IP wizards; they're here to do work in other
areas. Requiring them to participate in testing a new infrastructure
that they are not familiar with would be counterproductive to getting
their work done.
Is it OK to have an IPv6 infrastructure available to play with and
test? Absolutely! Just don't mandate it yet.
On 7/3/07, Hallam-Baker, Phillip <pbaker(_at_)verisign(_dot_)com> wrote:
Why can't the people who want to use IPv4 use that and the people who want to
use IPv6 use that.
If IP and appletalk can survive in the same network then IPv4 should be able to
survive with IPv6.
Sounds to me as if what is really being proposed here is to shut down the IPv4
connectivity. That will cause a serious loss of productivity for me. I am going
to Chicago to work and not to participate in a LAN party or an interop. There
are plenty of venues where that can take place.
With 1200 attendees paying $600 plus travel to attend the meeting is costing
well over a million dollars.
I agree that we should do dog food. The fact that this experiment cannot take
place is the data point.
People have to accept the data point and think through the deployment strategy
Back when laptops first came out you have to buy one power brick for the US and
a second one for Europe. The modern multi-voltage power supply only appeared
when an Apple vice president got angry when he discovered he dd not have the
right adaptor, returned to the engineering dept and told them that from now on
the machine would work on any mains electricity anywhere in the world.
> -----Original Message-----
> From: Thierry Ernst [mailto:thierry(_dot_)ernst(_at_)inria(_dot_)fr]
> Sent: Tuesday, July 03, 2007 4:05 AM
> To: ietf(_at_)ietf(_dot_)org
> Subject: Re: chicago IETF IPv6 connectivity
> Did we forget the IETF motto about running code ?
> I do understand that there is a need to alleviate risks in
> the operational network, so a traditional operational IPv4
> network has to be maintained until a vast majority of IETFers
> have assessed the proper operation of IPv6.
> But this doesn't prevent to operate a more advanced
> technology simultaneously. Someone has to start and
> experiment the shortcomings, if the IETF is not doing it no
> body is going to do it or nobody is going to be put any trust
> in a new technology. IETF has to pioneer this as it did so
> far (multicast, security, IPv6, and hopefully more will
> experimented at the IETF).
> Anyway, IPv6 is not experimental, it runs alright in
> operational networks.
> What I think the IETF should experiment now at a wider scale
> is the transition tools between IPv6 and IPv4. This would
> only be meaningful if results of the experiments are
> published, for instance in the IETF journal.
> On Mon, 02 Jul 2007 16:18:11 -0700
> Dave Crocker <dcrocker(_at_)bbiw(_dot_)net> wrote:
> >Jun-ichiro itojun Hagino wrote:
> >>>> The IETF network is not, and never has been, for
> >>>> showing off new technology, or making political
> statements. Please
> >>>> keep it that way.
> >>> +1
> >> RFC1883 is not new.
> >Neither is X.400, TP0, or many other specs. But then,
> that's not the issue.
> >The issue is operational risk.
> > Dave Crocker
> > Brandenburg InternetWorking
> > bbiw.net
> >Ietf mailing list
> Thierry ERNST, PhD
> INRIA Rocquencourt France Project-Team IMARA / JRU LARA
> http://www.lara.prd.fr +33 1 39 63 59 30 (office)
> The coming end of the IPv4 world: http://penrose.uk6x.com
> Ietf mailing list
Ietf mailing list
Clint (JOATMON) Chaplin
Corporate Standardization (US)
Ietf mailing list