Errors-to: owner-drums@CS.UTK.EDU
Message-Id: <199509150316.XAA22055@wilma.cs.utk.edu>
From: Keith Moore <moore@CS.UTK.EDU>
To: Jacob Palme <jpalme@dsv.su.se>
cc: Keith Moore <moore@CS.UTK.EDU>, Eric Thomas <ERIC@SEARN.SUNET.SE>,
        ietf-drums <drums@CS.UTK.EDU>
Subject: Re: Another header-munging example 
In-reply-to: Your message of "Fri, 15 Sep 1995 04:22:54 +0200."
             <Pine.SUN.3.91.950915041637.24926L-100000@ester.dsv.su.se> 
Date: Thu, 14 Sep 1995 23:16:26 -0400
Sender: moore@CS.UTK.EDU

(from the chair)

> You say that fixing mailing lists is outside of our charter. I do
> not agree. A general and full standard for all aspects of mailing lists
> is clearly outside our charter, but problems with obscurities in the
> existing standards as relates to mailing lists are within our
> charter.

Adding new functionality is outside of our charter.  (I also
believe we've already had a ruling from our AD stating that 
"the group does not have a mandate to describe mailing list 
behaviour in any great detail...".)

My opinion is that we can clarify the language in 822 that
refers to mailing lists (or "text message teleconferencing"),
but not to add new functionality specifically to support 
mailing lists.

Our charter was deliberately limited so we could focus on certain
problems with the mail specs.  This was done because history has
demonstrated that some issues are rat-holes for which a working
group can spend arbitrary amounts of time with no resolution.
Mailing lists are one such issue.  

This is not to say that fixing mailing lists isn't important, just 
that we're not going to fix them here.

> Even trying to get an agreement for a narrower implementation of
> "Reply-To" is "fixing mailing lists", so if fixing mailing lists
> is outside our charter, we chould do nothing in this case, nor
> in the case of "Resent-".

I have to draw the line somewhere.  This is where I'm drawing it.

If we want to fix mailing lists later, we don't want to have that
group re-defining the meanings of ordinary 822 headers.  Instead 
we should nail down those definitions here, with the full expectation
that the mailing list group will eventually be defining new headers.

822 allows Reply-to to be used with mailing lists, and we have
concensus that this behavior cannot be outlawed.  So we're simply
going to clarify that language.  We might want to add some text 
that explains the consequences of having list managers mung headers, 
but this won't affect the letter of the standard.

As for Resent-*, there are at least two interpretations of what 
it's to be used for: one is for manual resends and another is for
mailing lists and/or forwarding like things.  I don't really 
endorse any of these views, but I recognize that in Resent-* 
there was a hint of something like support for mailing lists.  

People might want to make a case that a well-defined Resent-like 
mechanism is what is needed for mailing lists, and that because 
Resent- was in 822, we should consider that within our charter.
I'm simply saying that I'm not going to rule this kind of approach
out-of-scope -- at least not yet.

Keith