[ On Saturday, March 19, 2005 at 16:14:27 (-0800), Fred Viles wrote: ]
> Subject: Re: [exim] Exim server behind NAT router (and HELO)
>
> On 18 Mar 2005 at 20:06, Greg A. Woods wrote about
> "Re: [exim] Exim server behind NAT r":
>
> | The RFCs actually do require all SMTP clients to identify themselves
> | truthfully and accurately. The HELO/EHLO parameter _MUST_ be a valid
> | canonical hostname that resolves to the address its connection
> | originates from. There are no if's, and's, or but's about it.
>
> Quoting from section 4.1.4 in RFC-2821, dated April 2001:
Please keep in mind that RFC 2821 is not yet an IETF "STD" -- STD 10,
the standard describing SMTP, is still RFC 821.
Regardless RFC 1123 is also still the main part of the mandatory host
requirements standard (part of STD 3), the single most important (though
as I said, somewhat dated and politically flawed) requirements document
for all Internet participants. It still reads, in part:
5.2.5 HELO Command: RFC-821 Section 3.5
The sender-SMTP MUST ensure that the <domain> parameter in a
HELO command is a valid principal host domain name for the
client host. [[ .... ]]
There are, still, no if,s, and,s, or but's about it. Period. It really
cannot get much simpler. (though most of us do allow the silly
waste-of-time IP literal excuse, provided it looks exactly right and
matches the source address... :-)
The other part about what the server does or does not do, or "MUST" or
"MUST NOT" do, is irrlevant here; unless of course you're the owner or
operator of an SMTP client that lies about its name and you're trying to
deliver mail to some server that won't put up with such B.S. ;-)
As for NATs, well they must be invisible to the public net if they're to
approach any degree of correctness in their functioning, regardless of
what any given peer they meet does or does not do to/with/for them.
--
Greg A. Woods
H:+1 416 218-0098 W:+1 416 489-5852 x122 VE3TCP RoboHack <woods@???>
Planix, Inc. <woods@???> Secrets of the Weird <woods@???>
