Message-Id: <199607040720.DAA07469@wilma.cs.utk.edu>
From: Keith Moore <moore@cs.utk.edu>
To: djb@koobera.math.uic.edu (D. J. Bernstein)
cc: drums@cs.utk.edu, moore@cs.utk.edu
Subject: Re: Multiple 'To:" fields (was: Re: Microsoft's Messaging Cont acts) 
In-reply-to: Your message of "04 Jul 1996 03:43:28 -0000."
             <19960704034328.11443.qmail@koobera.math.uic.edu> 
Date: Thu, 04 Jul 1996 03:20:21 -0400
Sender: moore@cs.utk.edu

> There is no evidence of this. qmail unquotes and requotes reliably.

Good for you.  Are you willing to assert that all SMTPs do as good a job?
 
> In contrast, we've already seen that ``simply copying the local part''
> leads to all sorts of violations of RFC 821 and RFC 822. 

I was referring to the specific case of relaying a message already
in SMTP.  A relaying MTA should always use the same envelope RCPT
addresses that it received.  This will introduce no protocol violations
that weren't already there in the first place.  And if there are 
protocol violations, having the relaying MTA try to "fix" them
just makes it harder to track down and find the real source of the
problem when the violation causes a failure.

Translation of 822 addresses to SMTP addresses is a separate issue.

Keith