Message-Id: <199604012357.SAA08411@wilma.cs.utk.edu>
From: Keith Moore <moore@cs.utk.edu>
To: Eric Thomas <ERIC@vm.se.lsoft.com>
cc: drums@cs.utk.edu, moore@cs.utk.edu
Subject: Re: Null Return Path -and- Resent-* 
In-reply-to: Your message of "Mon, 01 Apr 1996 23:47:15 +0100."
             <199604012305.SAA17751@CS.UTK.EDU> 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Mon, 01 Apr 1996 18:56:56 -0500
Sender: moore@cs.utk.edu

> On 01 APR 96 14:40 RANDY@MPA15AB.mv.Unisys.COM said:
> 
> >Why do we need to be concerned with specifying behavor (how lists should
> >treat null-return-path mail) that is expected to fade away?
> 
> More importantly,  why do we  always come back  to trying to  specify the
> behaviour and  "proper" configuration of RFC1123  "mailing lists"? 

Eric,

Your message is sufficiently vague that I can't tell what the heck you are 
talking about.

1. I'm not aware of anything that has obsoleted RFC 1123's requirement (in 
section 5.3.6) that mailing lists rewrite the return-path header to point to 
the list maintainer.  I *am* aware of several public mailing lists that 
violate this rule, having experienced numerous problems as a result of their 
failure to honor the rule and/or their attempts to solve the problems in other 
ways.  I don't care whether it is widespread practice or not, public mailing 
lists that leave original sender as the return address are broken.

I could, however, see clarifying this language to allow automatic parsing
and pre-screening of bounced messages.  

2. 1123's requirement that the message header MUST be left unchanged does seem 
a bit strong.  There are too many legitimate reasons to change the header, 
though many such changes are needed because of other protocol and/or layering 
violations (like removing precedence and return-receipt-to fields).

Keith