[6bone] RFC2772 rewrite -- bigger scope goals

[Bob Fink]

Pekka,

At 10:03 PM 11/14/2002 +0200, Pekka Savola wrote:
>Hello,
>
>First, before we try to modify RFC2772, I think we need to have an idea
>what we're trying to do:

I totally agree. That's why I called for ideas/issues etc. When I don't see 
any comments on the list, however, I wonder if anyone cares (I do know that 
you and some others do).

I would like to see an open and vigorous debate/discussion on what the 
6bone's role should become in the next phase of a transition to v6.


>  - policy for new allocations and for new organizations only?
>  - policy for new and old pTLA holders alike (any policy we agree on MUST
>also by implemented by current pTLA's, as an item they agreed to in
>RFC2772)?

I would think the latter. We have been skirting the issue of making 
existing pTLA holders clean up or change for several years. I believe it is 
time to get fairly specific.


>Follow-up questions: is there a need for the policies for the old and the
>new be the same (IMO not necessarily)?  What if don't conform to the new
>rules, the revocation process?

We would need to address this. Peer group pressure, if there is a real 
consensus, can be quite powerful, meaning we can yank prefixes.


>The next question would be whether we want to keep 6bone de-facto free and
>open, and a "big mess", or try to do something about it.  Views differ on
>this one; the options are basically (I hope I didn't miss any):
>  1) keep 6bone routing as it is, build totally separate competing v6
>Internet for "production"
>  2) try to move 6bone-style routing off to the edges of the network
>     a) try to clean up the current mess, or
>     b) prevent any further mess in new-pTLA rules
>  3) kill 6bone

The 6bone will be killed sooner or later. It's primarily a question of how 
articulate we are in making the case for a longer than a shorter life 
(which includes real plans, not just talk). However, sometime in the next 
1-5 years the 6bone prefix will get yanked. Let's use whatever the time to 
useful purposes.

So, we need to have a good discussion on just the issues you raise.


>Only after there's some rough consensus on these, we can proceed to the
>details in how to really revise RFC2772.

Agree, but, absent any other discussion, we will publish something to 
address what Rob and others think need be done.

So let's have a wide ranging discussion in Atlanta on what is really best 
to do with the 6bone.


Thanks,

Bob