ietf-smtp
[Top] [All Lists]

Re: [ietf-smtp] Dotless domains and email

2013-06-21 12:58:25
In practice, reliable delivery It requires at least two labels in the FQDN of a 
mail address.

Quite a few of the mail systems I've tested break when a dotless domain is 
used.  So there's an operational issue with this use.
Although I've looked, I've never been able to find a current RFC that requires 
a dot on the RHS of SMTP addresses.

Alex

-----Original Message-----
From: ietf-smtp-bounces(_at_)ietf(_dot_)org 
[mailto:ietf-smtp-bounces(_at_)ietf(_dot_)org] On Behalf Of John Levine
Sent: Friday, June 21, 2013 10:55 AM
To: ietf-smtp(_at_)ietf(_dot_)org
Cc: dcrocker(_at_)bbiw(_dot_)net
Subject: Re: [ietf-smtp] Dotless domains and email

Does anyone have an explanation for the assessment in the ICANN report
that at least two levels are required?

a) they goofed

2) MTAs running on Unix-ish systems typically use POSIX
gethostbyname() or res_query() to resolve host names.  Most implementations of 
those routines when presented with an unqualified name will first try the name 
with a locally configured default domain, and only if that fails will try the 
bare name.  There is a parameter in the resolver config file to change that, 
but I don't know anyone who changes it.

So it's not in the RFC, but it's sure in the code, and the way it's 
implemented, a local host name will mask a dotless TLD, leading to highly 
unpredictable results.

The counter-counter argument is that a fair number of ccTLDs have MX records 
for the TLD and the world hasn't come to an end, such as the world's shortest 
e-mail address n@ai.  For a good time, do an MX lookup on gt.

R's,
John
_______________________________________________
ietf-smtp mailing list
ietf-smtp(_at_)ietf(_dot_)org<mailto:ietf-smtp(_at_)ietf(_dot_)org>
https://www.ietf.org/mailman/listinfo/ietf-smtp

_______________________________________________
ietf-smtp mailing list
ietf-smtp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf-smtp