[Mailman-Users] short question, probably an FAQ that I've overlooked.

[John W. Baxter]

On 6/2/05 5:50 PM, "Larry Stone" <lstone19 at stonejongleux.com> wrote:

> On 6/2/05 2:27 PM, Brad Knowles at brad at stop.mail-abuse.org wrote:
> 
>> Proper mailing list management systems don't use
>> blind-carbon-copies.  The equivalent that they do use is to list
>> multiple envelope recipients, whose addresses are not necessarily
>> referred to on the "To:" or "Cc:" headers.
> 
> Brad, I realize this is getting a little off of Mailman but how is that
> different from what a BCC does? A BCC "header" is local to the originating
> MUA - it doesn't appear in the transmitted message but the BCC recipients do
> become envelope recipients. So from the view of the MTA, isn't what you say
> Mailman does exactly the same as what the MUA does with a BCC?

Not really.  Some MUAs preserve the full Bcc header for anyone in that
header, but removed it for the To: and Cc: recipients.  Others set up a Bcc:
header with just the individual recipient's address (clearly these talk to
the MTA separately for each Bcc: recipient).  And most remove it.  All these
behaviors are sanctioned by the RFCs (clearly a compromise in the RFC
development process).

Also, there are more knobs--outside Mailman--that can be messed with.  I
don't know how in Postfix; I never will know how in sendmail (I'm 66--I
won't live long enough to learn sendmail).  In Exim, it's a matter of
settings in a transport, with routers feeding appropriate domains to the
transport.  So you could arrange to send large batches most places, batches
of one to AOL, and batches of 5 to xyz.example.com, for example.

Also, I don't believe that the original posting's 5xx response meant that
the receiving site thought that your machine is an open relay--it guessed it
might be because of too many invalid recipient addresses in one connection,
which is typical of spamming through open relays (and spamming generally).

  --John