[DNSOP] Fwd: Status of "let localhost be localhost"?

[william manning <[e-mail]>]

---------- Forwarded message ----------
From: william manning <[e-mail]>
Date: Sat, Aug 5, 2017 at 5:33 PM
Subject: Re: [DNSOP] Status of "let localhost be localhost"?
To: John Levine <[e-mail]>


i think the question hinges on zone completion logic and fully qualified
domain names.

when localhost appears as:

localhost   IN  AAAA   3ffe:53::53

without the trailing dot, zone completion logic should ensure that it is
NOT treated as a TLD.
whereas if I code this:

localhost.  IN AAAA  3ffe:53::53

that is a clear indication that I am running my own root zone and defining
my own view of the DNS namespace for class IN.   Shouldn't apps depend on
the DNS to serve trustworthy, correct, data?

/Wm

On Sat, Aug 5, 2017 at 2:01 PM, John Levine <[e-mail]> wrote:

> In article <CAAiTEH9=RNDrUmSOs8Rg2Ea4+as9pg=j5jnU6Y=nc8A4Z1aPog@mail.
> gmail.com> you write:
> >In the case where 'localhost' is being passed to DNS resolution software,
> a
> >validating stub (for example inside a web browser) needs a way to know
> that
> >the 'localhost' TLD should be treated as insecure.  In that case, the only
> >way to accomplish that is ...
>
>  ... by having the stub or cache treat localhost as a special case.
>
> I use unbound as my cache which as far as I know has always done that.
> Are there caches that don't?  Are there validating stubs that don't?
>
> My reading of this draft is that if you don't treat localhost as a
> special case already, it's time to get with the program.
>
> R's,
> John
>
> > with an insecure delegation at the root.
>
> _______________________________________________
> DNSOP mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/dnsop
>