Re: ULA clashes don't matter [Re: ULA Registration]

[Erik Kline <[e-mail]>]

Agreed.

The doc is clear about what is required in order to minimize the chance of
collision.

In the inevitable event two non-randomly generated fd00:: networks merge,
that's their problem.  If this situation were to be cited as a motivation
for standards action, the correct answer IMHO is "they're doing it wrong;
not the IETF's problem".

On 24 March 2017 at 13:31, james woodyatt <[e-mail]> wrote:

> On Mar 24, 2017, at 11:37, Brian E Carpenter <[e-mail]>
> wrote:
>
>
> Do I need to justify that? Oh, OK: […]
>
>
> 4. In the event of the simultaneous merge of a sufficient number of
> ULA-numbered networks that the statistical chance of Global ID collisions
> arising would approach a significant level, the entity performing that
> merge can make its own private arrangements to manage the numbering
> requirements for the merge. For one thing, there is no apparent and
> immediate need for a public service to do it, and for another, establishing
> a registry is not obviously the best way to construct such a service in the
> event the need arises for one in the unforeseeable future.
>
> --james woodyatt <[e-mail]>
>
>
>
>
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> [e-mail]
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
>
>