[dhcwg] Re: Erik Kline's Yes on draft-ietf-dhc-addr-notification-12: (with COMMENT)

[Erik Kline <[e-mail]>]

On Tue, May 14, 2024 at 8:00 PM Jen Linkova <[e-mail]> wrote:

> Hi Erik,
>
> On Wed, May 15, 2024 at 11:14 AM Erik Kline via Datatracker
> <[e-mail]> wrote:
> > ### S3
> >
> > * "After successfully assigning...configured IPv6 address" ->
> >   "After successfully assigning...configured valid IPv6 address"
> >
> >   This would match the RFC 4862 "valid addresses" specificity used in
> S4.2.
>
> I think as per RFC4862, "invalid address - an address that is not
> assigned to any interface.", so I guess it's not possible to assign
> "invalid address".
>
> What do you think?
>

The main point I was after with the "valid address" language, and what I
had assumed was meant by its use in S4.2, is that addresses that are in
non-usable states like "tentative" or "dadfailed" must not be registered.
Generally addresses in these states won't appear as "assigned to the
interface", but I was thinking it just might be a bit more precise (also an
answer to the question "what does it mean to be successfully assigned":
"it's a 'valid address' per 4862").

I approved the candidate PR with the "valid" text addition.

> ## Nits
> >
> > ### S3
> >
> > * "the client MUST NOT register any addresses" ->
> >   "the client MUST NOT register any addresses using the mechanism in this
> >    specification"
>
> Ack, will be fixed in -13.
>
> --
> Cheers, Jen Linkova
>