For the record, maybe you won't believe me as an SMTP author, but maybe you
would believe an EXIM software author, a very popular system on Unix boxes.
The issue is discussed pretty much all over the documentation in regards to
HELO or HOST lookup, but nothing is better than looking at the directly at
source code itself where the inner thoughts, ups and downs of the programmer
wrote about:
see host.c source code, inline comment:
/*************************************************
* Find sender host name by gethostbyaddr() *
*************************************************/
/* This used to be the only way it was done, but it turns out that not all
systems give aliases for calls to gethostbyaddr() - or one of the modern
equivalents like getipnodebyaddr(). Fortunately, multiple PTR records are
rare,
but they can still exist. This function is now used only when a DNS lookup
of
the IP address fails, in order to give access to /etc/hosts.
......
*/
Further down the source code, you will find even more detail comments:
/*
......
This function does a thorough check that the names it finds point back to
the
incoming IP address. Any that do not are discarded. Note that this is relied
on
by the ACL reverse_host_lookup check.
On some systems, get{host,ipnode}byaddr() appears to do this internally, but
this it not universally true. Also, for release 4.30, this function was
changed
to do a direct DNS lookup first, by default[1], because it turns out that
that
is the only guaranteed way to find all the aliases on some systems. My
experiments indicate that Solaris gethostbyaddr() gives the aliases for but
Linux does not.
[1] The actual order is controlled by the host_lookup_order option.
.....
*/
So as I said, as EXIM learned by release 4.30, it is better to do a direct
DNS lookup first because quote "it turns out that that is the only
guaranteed way to find all the aliases on some systems." and it goes on to
say, gethostbyaddr() works differently for various flavors of Unix.
Now, the reality about gethostbyaddr() is that it all depends on the OS
networking style (TCP/IP or NETBIOS or other) and HOSTS file. It has
nothing to do with WINDOZE or Unix system. Lots of software still use the
legacy methods, and even if the software was updated, there is no urgent
need to upgrade for this reason only because 99.99% of the systems do not
REJECT on a HELO A record mismatch.
Not convinced yet?
--
Hector Santos, Santronics Software, Inc.
http://www.santronics.com
----- Original Message -----
From: "Mark" <admin(_at_)asarian-host(_dot_)net>
Newsgroups: spf.-.sender.policy.framework.discussion
To: <spf-discuss(_at_)v2(_dot_)listbox(_dot_)com>
Sent: Wednesday, November 23, 2005 3:07 PM
Subject: RE: [spf-discuss] Re: SPF adoption statistics
-----Original Message-----
From: Hector Santos [mailto:spf-discuss(_at_)winserver(_dot_)com]
Sent: woensdag 23 november 2005 19:52
To: spf-discuss(_at_)v2(_dot_)listbox(_dot_)com
Subject: Re: [spf-discuss] Re: SPF adoption statistics
Hector, "If the server sends 'EHLO mail.example.com' then this means
that there must be a DNS A record" is as true a statement as you will
ever come across.
Servers don't send EHLO/HELO, clients do.
Uh-uh. So, my server never sends a HELO/EHLO, is that what you'r saying?
You're too funny. :)
Outlook used the socket call gethostbyaddr() which will return the
netbios computer name "hdev1" for my home office machine.
I am not, in the least, interested in the brokenness of your Windoze
client software.
So what are you doing in an world-wide open discussion forum that is
constributed to the email system? This isn't about your SMALL unix wie-
nie world only or what you call about.
So, now RFC correctness is limited to UNIX only? And the rest is, what,
exempt? And why? Because you never figured out how to have your software
do a proper HELO? 'Fraid not, dude. If you cannot get your software to
send a proper HELO, then you need to stay the hell away from sending mail.
And to answer your question, yes, I have written SMTP server and
client software.
Oh I find that hard to believe. You have not because you wouldn't be so
close minded and would understand the long 20+ years real WORLD inter-
face issues and understand why there is a REAL reason why its not reli-
able and if you keep BLAMING on WINDOWS, then you are a FOOL.
Actually, I am not blaming Windoze, I'm blaming you. It is, after all, you
how are incapable of configuring your software properly. Now, granted, in
the Windoze MCSE world, with its 'click-and-play' attitude, no high
demands
are made on its operators. And you are a fine specimen of that. You just
start sending mail from your box without even settings things up properly.
And then you give me some lame-ass story about a gethostbyaddr() call
being responsible for your wrong HELO. Yeah, real mature way of taking
responsibility.
If you want to write to me, keep the BULLSHIT out, ok?
I have no intention of writing you at all. Save to say you should take
your pills. And after you did, fix your HELO. And stop getting angry at
people who do not buy into your "But my software is so broken, and I am so
incompetent, that you will just have to let me use a non-FQDN for HELO."
- Mark
System Administrator Asarian-host.org
---
"If you were supposed to understand it,
we wouldn't call it code." - FedEx
-------
Sender Policy Framework: http://www.openspf.org/
Archives at http://archives.listbox.com/spf-discuss/current/
To unsubscribe, change your address, or temporarily deactivate your
subscription,
please go to
http://v2.listbox.com/member/?listname=spf-discuss(_at_)v2(_dot_)listbox(_dot_)com
-------
Sender Policy Framework: http://www.openspf.org/
Archives at http://archives.listbox.com/spf-discuss/current/
To unsubscribe, change your address, or temporarily deactivate your
subscription,
please go to
http://v2.listbox.com/member/?listname=spf-discuss(_at_)v2(_dot_)listbox(_dot_)com