spf-discuss
[Top] [All Lists]

[spf-discuss] Re: advice wrong, or is it?

2007-12-22 12:19:49
David Woodhouse wrote:
 
Note:  they say "~all" is good, "-all" is not.

Assuming that they say this (I'm not convinced):
 
They would say it because they believe, like many others,
that the fundamental principle on which SPF is based --
that forwarding does not happen -- is completely wrong.

The fundamental principle of SPF is RFC 821, with a clear
responsibility for accepting and forwarding mail, noted in
the reverse path.  RFC 821 also offers a simple solution
for folks not wishing to take this responsibility:  551.

And thus that publishing a '-all' record invites people
to throw away genuine mail.

That's rubbish, '-all' emulates 551 as good as possible
for receivers rejecting FAIL.  The original senders are
supposed to know why the they publish FAIL.

OTOH with '~all' it's more likely that mail ends up in a
black hole resulting in a loss of "genuine mail".  Even
where receivers handle SOFTFAIL with greylisting, later
they'll accept it as "suspicious" => spam folder => loss.

With 'FAIL => reject => forwarder bounces => sender makes
another plan' all is fine.  

For the normal definition of 'genuine', that is, not
the SPF NewSpeak definition.

The normal definition of responsibility is RFC 821, not
the 1123 5.3.6(a) bug where "responsibility" in the case
of forwarding to 3rd parties mysteriously vanished.  SPF
emulates the status quo antea wrt 1123 5.3.6(a), that's
no NewSpeak simply because it isn't "new". 

 Frank

-------------------------------------------
Sender Policy Framework: http://www.openspf.org
Archives: http://v2.listbox.com/member/archive/735/=now
RSS Feed: http://v2.listbox.com/member/archive/rss/735/
Modify Your Subscription: 
http://v2.listbox.com/member/?member_id=2183229&id_secret=78832188-714596
Powered by Listbox: http://www.listbox.com