To: drums@cs.utk.edu
Subject: Re: Message format document outline 
In-Reply-To: Your message of "Fri, 08 Mar 1996 11:12:15 -0000."
             <oO1PbDAPYBQxQAwG@turnpike.com> 
Date: Sat, 09 Mar 1996 01:22:53 +1100
Message-Id: <1721.826294973@munnari.OZ.AU>
From: Robert Elz <kre@munnari.oz.au>

    Date:        Fri, 8 Mar 1996 11:12:15 +0000
    From:        Paul Overell <paulo@turnpike.com>
    Message-ID:  <oO1PbDAPYBQxQAwG@turnpike.com>

    As an implementor of an RFC822 parser I strongly agree with the view
    expressed elsewhere in this thread that any message that is legal under
    RFC822 should remain legal under RFC822bis.

I don't believe we need to go this far.  If something is obsolete,
we need not be afraid to say so.

We should not, however, then go ahead and re-use the syntax,
or anything that wold be incompatible with the old syntax, for
some new purpose.

Eg: say we had decided that the group addressing syntax was
obsolete (which was NOT done, it isn't, this is just the easiest
example I can think of right now), we could simply remove
the

     group       =  phrase ":" [#mailbox] ";"

rule (rather than fixing it by removing the []) from the grammar
completely.   We could not decide to use ':' ';' type addresses
for some other purpose however.

Note, this does not detract from anyone's ability to parse ancient
messages if they want to - you can still parse 822 syntax in your
implementation if you so desire, and thus allow users to still
correctly process whatever ancient crud is decided to be deleted
from the spec.

Further, this is not at all new, even though 822 section C.5.5
excplicitly says that " at " is no longer equivalent to "@"
(and " at " is no longer in the grammar in 822), people's mailers
have been supporting parsing of "user at host" ever since (though
this practice is gradually dying).   That is, if you have users
with ancient 733 format messages still in their mailboxes, you're
still permitted to process the things, there's no requirement
that your mailer complain and refuse.

So, please let's stick to common sense here, the DRUMS output,
whatever it is called, should document today's internet mail, and
not be overly constrained with remaining compatible with every
nuance of 822.

kre