[...]

From RFC 821
http://www.faqs.org/rfcs/rfc0821.html --> CURRENT ACCEPTED

3.5. OPENING AND CLOSING
3.7. DOMAINS
4.1.1. COMMAND SEMANTICS
4.1.2. COMMAND SYNTAX
4.5. DETAILS
4.5.1. MINIMUM IMPLEMENTATION
APPENDIX F

-----------------------------------------------------------------------
From RFC2821
http://www.faqs.org/rfcs/rfc2821.html --> CURRENT IN PRACTICE

2.3.4 Host
2.3.5 Domain
3.6 Domains
4.1.1.1 Extended HELLO (EHLO) or HELLO (HELO)
>
> | And Greg is not the only one that requires host information | to
> valid in many different ways. --
>
> True, I'm sure. Which is not the same as saying it's a good idea. FWIW,
> when I experimented with much less strict consistency checking here
> (IP->PTR->sameIP), I got about 25% false positives. I have a hard time
> picturing any company that could afford to use such checks to reject
> mail.
>

I never said that I currently implement IP->PTR->sameIP, you will get far
more than a 25% false positive rate.  I do however require that your
connecting IP meet one of these:
(yes PTR OR no PTR then A/CNAME == connecting IP)
        AND
(A/CNAME == resolvable)

Believe it or not, most spam/virus get kicked on those alone, and that
does not include any other checks done on connectingIP DNS information or
the information gleaned from the HELO command.

Greg IS more strict than I am, but if you follow RFC2821 then
HELO blahblahblah MUST have IP==PTR==HostA/CNAME
--