[IPv6]Re: Call for adoption: draft-kumari-ipv6-loopback-02 (Ends 2026-08-05)

[Michael Richardson <[e-mail]>]

I have read kumari-ipv6-loopback, and I support adoption.
I agree with others that the name should probably be "hostlocal"
(which is akin to 'localhost'...)

I have noticed how v4 (docker!) uses 127.0.1.1, within 127.0.0.0/8,
and I never noticed that all of ::/96 wasn't equally useable.
In practice, I've regularly changed the prefix-length ("netmask") of lo, and
I've run services on ::2, ::3, ::4...

I am not convinced that ::/96, (with ::/128 understood to be cutout), can not
work, but that's a discussion for post-adoption.

The IANA special purpose registry is at
https://www.iana.org/assignments/iana-ipv6-special-registry/iana-ipv6-special-registry.xhtml
Generally that means something within: 2001::/23 as per RFC2928, and amended
later on.   2001:db8::/32 (Documentation Prefix) is there.

If this prefix will be within 2001::/23, then let's make sure that:
a) it's easy to type.
b) it's quite visually distinct from 2001:db8:  so that humans seeing it
   for the first time are minimally surprised.

I won't start a beauty contest here on that, I have many ideas.
I would like some variation of the above two points to be included in the
IANA Considerations.

(I'm reminded of: "Why is prisoner.iana.org attacking me?")

--
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-                      *I*LIKE*TRAINS*