To: djb@koobera.math.uic.edu (D. J. Bernstein)
Cc: drums@cs.utk.edu
Subject: Re: comments from a newcomer 
In-Reply-To: Your message of "29 Feb 1996 19:54:33 GMT."
             <19960229195433.20101.qmail@koobera.math.uic.edu> 
Date: Fri, 01 Mar 1996 18:18:29 +1100
Message-Id: <22515.825664709@munnari.OZ.AU>
From: Robert Elz <kre@munnari.oz.au>

    Date:        29 Feb 1996 19:54:33 GMT
    From:        djb@koobera.math.uic.edu (D. J. Bernstein)
    Message-ID:  <19960229195433.20101.qmail@koobera.math.uic.edu>

    (It wouldn't be so harmful in a completely MIME-ified world---it'd be a
    small waste of bandwidth, nothing more---but that's not the real world.)

One could argue that it isn't the real world as people insist on
making their mailers break the standards, and send 8 bit.  If
all mailers would simply strip the top bit, as they should,  then
I suspect that MIME (or extended 8 bit mode SMTP) would grow in
popularity quite quickly.

    Why are you deliberately making de jure standards that don't match the
    de facto standards?

The de facto standard is 7 bit.   That's what we all use.
That some fraction violate it doesn't make them any kind
of standard, or not until it is an overwhealming percentage.

    The job of an MTA is to communicate. Users understand this, and yell
    at people who corrupt messages. Unfortunately, those people use your
    7-bit restriction as an excuse for continuing the corruption.

No, it is those who insist on sending 8 bits, in violation of
all practice, and the standards, who are causing e-mail
corruption.

    If you're running SMTP over a 36-bit protocol, for example, you should
    be allowed to transmit 36-bit data. Anything else is poor engineering.

No, this would be stupid.  The correct way is for SMTP to
define its character set, and then for that to be mapped into
whatever the underlying protocol is able to support, in the
most appropriate way.

    A relay shouldn't have to worry about conversions when it has the same
    transport protocol on input and output.

And if capabilities of the destination are the same as those
of the sender, transport protocols aren't all that matter,
in fact, they're the easy part.

This discussion is a waste of time, it's been held before, it
has changed nothing before, and won't again now.  In fact, now
it's even less likely, as users who need 8 bit characters have
a method (MIME), and MTA's that want to send 8 bits have a
method to negotiate that (ESMTP).   The "just send 8" argument
is even more feeble and absurd now than it ever was.

kre