🚀 go-pugleaf

In article <1993Jan21.234753.20583@blilly.uucp>, bruce@blilly.uucp (Bruce Lilly) writes:
> In article <1993Jan19.211424.9632@news.csuohio.edu>,
>  posted to comp.mail.headers,
>  damin@ullman.cba.csuohio.edu (Greg Boehnlein) wrote:
>> Once the mail was AT these locations, it 
>>was my job to pick up the mail files in their original format, download 
>>them to our gateway machines, parse and distribute them, converting in the
>>process between the Internet (RFC 822) standards and our FutureNet standards.
> 
> mail != files.
> 
> In order to deliver the messages, you don't want to look at the headers, you
> want to look at the envelope address. The same applies to the return
> address; it comes from the envelope, not the headers.
> 
> It is perfectly legal to have a message with no To: header and an empty Bcc:
> header.
> 
> If you must deal with the message as a stream of characters (or a disk
> file), I'd recommend using an SMTP-formatted message, where the envelope
> addresses are contained in the SMTP MAIL FROM (or SAML/SOML/SEND FROM) and
> RCPT TO commands. This is sometimes called batched SMTP. RFC821 (as amended
> by RFC1123) is the SMTP specification.

Things are getting closer to the truth. The address to whom mail is delivered
is, indeed, the envelope address which is not part of the message. This can
lead to some strange messages when the To: line and the envelope differ, but it
is NOT unusual.

The reply address MAY be the envelope "from" address, but not necessarily. The
order of prefererence for reply is rather long, but LOTS of header fields take
precidence over the envelope. RFC-822 is quite fuzzy on this, but there can be
no question that some fields (e.g. "Reply-To:) must be deferred to. Simply
using the envelope "from" field for replys is NOT adequate. The commercial
mailer I use determines the corrrect address for replys in the order:

Resent-Reply-To:
Resent-From:
Reply-To:
From:
Resent-Sender:
Sender:
envelope "from" address

The envelope "from" is pretty much reserved for occasions when there is no
useful header information. The RFCs are NOT clear on this order, so some
mailers may vary.

R. Kevin Oberman			Lawrence Livermore National Laboratory
Internet: koberman@llnl.gov		(510) 422-6955

Disclaimer: Don't take this too seriously. I just like to improve my typing
and probably don't really know anything useful about anything.