xd- with C/DNAME (was: Re: The Two Lookups Approach (was Re: Parsing the issuesand finding a middle ground -- another attempt))

[Erik van der Poel]

On Mon, Mar 16, 2009 at 10:20 AM, Andrew Sullivan <ajs at shinkuro.com> wrote:
> On Mon, Mar 16, 2009 at 09:54:02AM -0700, Erik van der Poel wrote:
>> CNAME/DNAME don't appear to require more than one lookup. I tried
>> www.xn--kxadbfj6eq.gr from Firefox on Windows and Ethereal reported a
>> single packet going out (the query), and a single packet coming back,
>> which included a DNAME, a CNAME and an IP address. (I'm willing to
>> believe that Ethereal is showing me fake packets due to Windows API
>> quirks.)
>
> I don't understand your report there, because you haven't given enough
> information.  What query did you send, and how, and what was the
> actual response you got back from the DNS server?

I couldn't get Ethereal on Windows to save as text, so I tried the
"dig" command on Linux, and got the same result as Ethereal did:

;; QUESTION SECTION:
;www.xn--kxadbfj6eq.gr.   IN  A

;; ANSWER SECTION:
xn--kxadbfj6eq.gr.  300 IN  DNAME xn--mxaaehw9bq.gr.
www.xn--kxadbfj6eq.gr.  0 IN  CNAME www.xn--mxaaehw9bq.gr.
www.xn--mxaaehw9bq.gr.  300 IN  A 67.15.125.21

According to Ethereal on Windows, the above info was all in one
packet. Is the above packet unusual? Illegal?

> Note that your preference for CNAME is going to make things rather
> more difficult, becuase CNAME aliases an RNAME but not things below
> the RNAME.  There are big controversies over whether _anything_ may be
> below the RNAME.

I don't know how difficult it is to set up these CNAMEs and DNAMEs,
but it seems like it should be possible to put xu-- labels in the
CNAME/DNAME, even if the question section has an xn-- name.

Erik