Message-Id: <199512111628.LAA16153@info.cren.net>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Mon, 11 Dec 1995 11:32:42 +0100
To: jwn2@qualcomm.com (John Noerenberg)
From: conklin@info.cren.net (Jim Conklin)
Subject: Re: Clarify amibuities)
Cc: ietf-drums <drums@cs.utk.edu>

mailing list manager agent, and those (few!) who expect the secretary who
mailer his boss' letter to use Sender to indicate that, will lose a little
information, I agree that little real harm would come from deprecating
Sender.  I think my personal preference would be to clarify that it's only
for human use, as distinguished from use by a mail agent, but I believe
it's not a show-stopper either way.  (As a historical note of some possible
significance, Sender has been used by mailing-list managers to force
certain mail agents to send error notifications to that address instead of
the From address; do we know if any agents still in popular use incorrectly
send error notifications to Sender if it's present?)

Jim

At  2:27 AM 12/8/95 -0600, John Noerenberg wrote:
> ...
>Jim is arguing that in each context in which Reply-To is used (two have
>been mentioned, directing replies to lists and directing personal
>replies), the intended destination of a reply to a message is clear:
>the destination of a reply to this message must be sent to the value of
>the Reply-To field.
>
>
>This seems unambiguous to me, and perfectly implementable.
>
>
>Ambiguities for the interpretation of Reply-To are introduced in the
>presence of a Sender field.  The semantics of Sender have always been
>confusing to me.  Reading over the definition of Sender, I'm inclined
>to think it is an anachronism we can discard.  Discarding Sender
>eliminates ambiguities for Reply-To.
>
>
>I suggest deprecating Sender.  That kills 2 birds with one stone. :-)
> ...