Re: Services and top-level DNS names

"Frank Ellermann" <> Sat, 05 July 2008 07:59 UTC

Return-Path: <>
Received: from [] (localhost []) by (Postfix) with ESMTP id 8804D3A6A05; Sat, 5 Jul 2008 00:59:28 -0700 (PDT)
Received: from localhost (localhost []) by (Postfix) with ESMTP id 737423A6A0B for <>; Sat, 5 Jul 2008 00:59:27 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: 2.196
X-Spam-Level: **
X-Spam-Status: No, score=2.196 tagged_above=-999 required=5 tests=[FROM_LOCAL_NOVOWEL=3.196, RCVD_IN_DNSWL_LOW=-1]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id Sn6G7A5tPRYj for <>; Sat, 5 Jul 2008 00:59:26 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id EC6323A6A05 for <>; Sat, 5 Jul 2008 00:59:24 -0700 (PDT)
Received: from list by with local (Exim 4.43) id 1KF2fw-000625-W0 for; Sat, 05 Jul 2008 07:59:20 +0000
Received: from ([]) by with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <>; Sat, 05 Jul 2008 07:59:20 +0000
Received: from hmdmhdfmhdjmzdtjmzdtzktdkztdjz by with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <>; Sat, 05 Jul 2008 07:59:20 +0000
From: "Frank Ellermann" <>
Subject: Re: Services and top-level DNS names
Date: Sat, 5 Jul 2008 10:01:43 +0200
Organization: <URL:>
Lines: 63
Message-ID: <g4n9kf$7o7$>
References: <> <E709066D4201B0B9202841FC@p3.JCK.COM>
Mime-Version: 1.0
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2800.1914
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1914
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: IETF-Discussion <>
List-Unsubscribe: <>, <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

John C Klensin wrote:

> For years, we managed to dodge that problem with conventional
> subdomains (e.g., nic.example.), but we have not, to my knowledge,
> ever promoted those uses via, e.g., a BCP that strongly recommends
> them (unlike the email case, where RFC2142 does just that).

I'd be happy if we can salvage abuse@ from the remains of RFC 2142,
anything else is IMO hopeless:  The netnews folks flat out refused
to get as much as an informative reference to RFC 2142 for the old
news@ / usenet@ / netnews@ mess.  

Compare <> (Caveat:
errata links on the IETF Tools server are unreliable at the moment,
reported bug)

IOW we can say that spam killed RFC 2142, and maybe indirectly an
IETF WG.  RFC 2142 was published before RFC 2277, today its lack of
I18N considerations is rather peculiar.  I don't see what RFC 2142
has to do with reserved TLDs for technical reasons.
>     RCPT TO:<somemailbox@example.>
> is a syntax error.   RFC 2821 permitted it

Nope, RFC 2821 *required* more than one label, but it didn't allow
a trailing dot.  2821bis stripped the requirement.  In other words
somemailbox@example *is* an RFC 2821 syntax error until the second
when 2821bis is approved, after that it's only a bad idea.  

"One dot required" (but no trailing dot) also made it in RFC 4871,
among others, so far for TLDs trying to use DKIM directly at their
root.  SPF cannot talk about TLDs, but TLDs can have SPF records.
TLDs starting a news server will coordinate their activities with
the UUCP world map project (this involves time travel).

> So, much as I'd like it if we could say "Single label names are
> local in scope...does not work", I fear that it is unrealistic
> in practice unless we can somehow turn the clock back 15 years
> or so.

[ hours later you wrote...]

> let me note that 1536 is an informational document.  We generally
> don't claim that systems are expected to be compliant with those.

[ITYM s/1536/1535/g for those search lists (?)]
The complete concept of IDN TLDs including the running experiment
with the eleven IDN test TLDs, builds on an informational RFC 3696,
or on RFC 1123 errata, in a certain sense; often the status of an
RFC is unrelated to its importance.  

> Review BCP17/RFC2219, be sure that we still believe it

Not sure, it talks about WKS, it is apparently older than SRV, it
has no proper IANA registry; "archie", "finger", "gopher", "ph",
"rwhois", and "wais" are dead (but you can't say that in an RFC,
because some folks including me would whine wrt "gopher")

Apart from being odd names, what is wrong with a TLD ".abuse" or
a TLD ".archie" ?  TLD ".finger" could be funny for some Finnish-
German organization.  


Ietf mailing list