On 17 dec 2007, at 17:07, Dave Crocker wrote:
As a simple model, let's assume 4 administrations are in the
critical path:
1. Participant's laptop
2. IETF network ops
3. Collection of backbone operators
4. Participant's service provider
Each of these must support v6, for true dual stack to work.
IPv6 support on the laptop should be doable if the participant cares
to enable it for pretty much any OS that shipped in the last 5 years.
Obviously the IETF part should be working, and if the participant's
service provider has IPv6 connectivity then by definition they are
connected to a backbone that supports this.
Please note that dual stack is not an interesting test. Unless I'm
mistaken, the IETF meeting network has supported dual stack IPv4 +
IPv6. This simply means that people running IPv4-only don't see a
difference. Also, if we had enough address space to give everyone the
IPv4 part of dual stack we wouldn't need IPv6 in the first place. IPv6-
only is the future.
b) What is the lowly participant to do if they are a mere customer
of a real, commercial service provider or a real organizational IT
department who does not support v6 on production machines?
Untether from the mother ship for a an hour?
Use a transition mechanism such as a transport relay to access
services at home?
With respect to the latter, what information will be usefully
obtained by the blocking of these participants from their home
services that could not have been learned from a simple survey to
their home service operators?
I think this question is easier to answer afterwards than beforehand...
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf