Errors-to: owner-drums@CS.UTK.EDU
Message-Id: <199508152017.QAA22107@wilma.cs.utk.edu>
From: Keith Moore <moore@CS.UTK.EDU>
To: conklin@info.cren.net (Jim Conklin)
cc: Eric Allman <eric@CS.Berkeley.EDU>, Robert Elz <kre@munnari.oz.au>,
        ietf-drums <drums@CS.UTK.EDU>, moore@CS.UTK.EDU
Subject: Re: "Reply-To" 
In-reply-to: Your message of "Tue, 15 Aug 1995 15:57:56 BST."
             <199508151955.PAA12715@info.cren.net> 
Date: Tue, 15 Aug 1995 16:17:46 -0400
Sender: moore@CS.UTK.EDU

>   At least as common (much more common in my personal experience) is its
> use to force replies to go to a list instead of the From: address, a
> feature that's handy for lists with inexperienced networkers whose replies
> would by default go to the originator of a list message rather than to the
> list as intended.  While this is a pain for experienced folks, it's
> frequently considered helpful by less experienced users of lists.

I don't think we can stop lists from doing this, but 
I believe this "feature" needs to be officially discouraged.

The sender knows the content of the message better than the list does;
the sender should be able to specify where replies go.  

If the list adds a reply-to header, it looks as if the *sender* 
specified it; it therefore misrepresents the sender's intentions.

I'll note that other lists have the opposite problem: they want replies
to go to the sender only, not to the entire list.  (we have "announcement"
lists here that do just that.)

Maybe the right answer is to suggest that if lists muck with reply-to,
they do so ONLY if the sender did not specify a Reply-to, and also that
they put a comment to the effect that they inserted the header.
This is ugly but it might be better than a completely new mechanism.

One downside is that the sender would then have to specify Reply-To 
in cases where the 'default' interpretation (i.e. the From header 
addresses and possibly those of other recipients) would otherwise
have been sufficient.

Keith