Message-Id: <v02140404ad09fbec8f80@[165.227.40.21]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Fri, 29 Dec 1995 12:33:32 -0800
To: drums@cs.utk.edu
From: paulh@imc.org (Paul Hoffman)
Subject: Re: Header to label lists (was untitled)

At 4:56 AM 12/29/95, Robert Elz wrote:
>Of Mailing-List Group and Mail-Group I much prefer the first.

I too am in favor of "Mailing-List" over the others, given that that is
what end users call them.

>However much they are currently mangled together, newsgroups and
>mailing lists are not the same thing, and anything that tends to
>even subliminally suggest they are should be avoided.

Completely agree. Point-to-point messages have a different set of problems
than flood messages, and the one of the main desires for this header (to
help direct error messages) comes from the differences.

>I am not sure we should be even considering defining new headers
>here however, especially not for a new purpose like this, a WG
>to consider mailing list issues would be a better forum.

The drums WG charter says "New functionality is expressly inappropriate."
That doesn't say forbidden. :-) [ducking Keith's flinging of an overhead
marker]

>But while on the topic of headers, is there anything we can do to
>try to stamp out the practice of mailers advertising themselves in
>X-* headers.  I rather doubt it, and I expect there are probably
>people who believe this somehow helps them debug problems (oh,
>that was sent from the dead-fish mailer, that always mangles the
>foobar header, so ignore that value...) - but it would be nice to
>make the practice stop.

I strongly disagree here. These X-headers have proven invaluable in
debugging, and if we aren't going to give the manufacturers of buggy
clients a real header to announce themselves with, we shouldn't discourage
them from telling the world who broke the message.

--Paul Hoffman
--Internet Mail Consortium