Errors-to: owner-drums@CS.UTK.EDU
From: <Harald.T.Alvestrand@uninett.no>
Message-Id: <199509181255.IAA28208@CS.UTK.EDU>
To: Jacob Palme <jpalme@dsv.su.se>
cc: Keith Moore <moore@CS.UTK.EDU>, ietf-drums <drums@CS.UTK.EDU>
Subject: Re: Resent- facility
In-reply-to: Your message of "Fri, 15 Sep 1995 04:10:28 +0200." <Pine.SUN.3.91.950915040635.24926I-100000@ester.dsv.su.se>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 18 Sep 1995 14:55:00 +0200
Sender: Harald.T.Alvestrand@uninett.no

I've found Resent: useful in one single instance:

When a message is in my mailbox, and should have been in the mailbox
of an automatic processor, I want to forward it, but make it appear
as if it came from the original sender.

In this case, the Resent-* headers are the most useful headers for
documenting what I did to the message.
The "forward" option (which I *heartily* support recommending as The Way
to move messages to someone who needs to see it) would not work in this case,
or would require me to implement a MIME parser in the auto-responder.

I would change RFC 822 in such a way that:
- Resent-* headers are only described in a single section
- The section starts something like this:

     When an UA wishes to send a message that has arrived in its mailbox
     to another mailbox while retaining the illusion that it came directly
     from the original sender, it MUST (SHOULD?) document this action
     by adding Resent-To:, Resent-Date: and Resent-From: fields to show
     what was done to the message.

     When an user wishes to bring a message to another user's attention,
     this method SHOULD NOT be used; rather, MIME encapsulation of the
     message is the recommended method.

What do you think?


                        Harald A