Re: [v6ops] New Version Notification for draft-ipversion6-loopback-prefix-00.txt

[t.petch <[e-mail]>]

--- Original Message -----
From: "Mark Andrews" <[e-mail]>
Sent: Wednesday, February 18, 2015 11:10 AM

> In message <024f01d04b68$49ae8340$[e-mail]>, t.petch
<[e-mail]> writes:
> > ----- Original Message -----
> > From: "Mark Andrews" <[e-mail]>
> > Sent: Tuesday, February 17, 2015 1:23 AM
> >
> > > The fundemental reason for 127.0.0.0/8 was to give each node a
> > > addresses block they could use. 127.0.0.1 evolved as the
"standard"
> > > loopback address over time.  For the most part no one uses the
rest
> > > of 127.0.0.0/8 but it is useful to have available.  That said any
> > > use of the rest of 127.0.0.0/8 has to be negotiated between the
> > > users.  You can't just grab 127.0.0.2 and hope that no one else is
> > > using it for IP traffic.
> >
> > Mark
> >
> > I would refer you to RFC5782 and RFC6471 for a discussion on the use
of
> > 127.0.0.2 (and ::FFFF:7F00:2 ).
>
> RFC5782 Informational.
> RFC6471 Informational.
>
> And they actually grabbed namespaces in the DNS not addresses.
> Neither of them use those "addresses" for IP traffic which
> is what the proposed reservation of the block is about.
>
> The only reason 127.0.0.2 etc. are use for rbls is that it
> was possible to get sendmail to do a address lookup from
> sendmail.cf without hacking the code.  They returned values
> are tokens not addresses despite them being encoded in
> A/AAAA records.
>
> Raising RFC5782 and RFC6471 is a Red Herring.

Mark

Classifying those two RFC as Informational is a subtlety that those
reading this e-mail will understand but most will not (as I am reminded
of when I have seen manufacturers cite conformance to Internet
Drafts:-).   I am sure that if I looked hard enough I would find these
RFC being cited as if they were Standards, let alone a BCP.  And note
the title of RFC6471

'overview of Best email dns-based list (DNSBL) Operational Practices '
(my capitalisation).

As I recall, the reason why it is not a BCP, when it is a BCP in all but
name, is political, that the BCP stream is owned by the IETF and that
RFC is not a product of the IETF.

So, rightly or wrongly, the IETF has implicitly endorsed a usage of
127.0.0.n, where n is a small number, and it would be foolish, IMHO, to
endorse a separate usage.  (For some reason, the choice of percent for
interface identifiers, when percent was already spoken for in URIs,
comes to mind).

Last I heard, IPv6 is different, that the number of MX is small, the
namespace is enormous and so while ::FFFF:7F00:2 was posited, it is
unlikely to gain much traction.

Tom Petch

> Mark
>
> > Perhaps those RFC should have included an IANA Considerations:-(
> >
> > Tom Petch
> --
> Mark Andrews, ISC
> 1 Seymour St., Dundas Valley, NSW 2117, Australia
> PHONE: [phone]                 INTERNET: [e-mail]