Received: from localhost by CS.UTK.EDU with SMTP (cf v2.9s-UTK)
	id DAA06599; Mon, 24 Jul 1995 03:19:58 -0400
X-Resent-To: drums@CS.UTK.EDU ; Mon, 24 Jul 1995 03:19:57 EDT
Errors-to: owner-drums@CS.UTK.EDU
Received: from domen.uninett.no by CS.UTK.EDU with SMTP (cf v2.9s-UTK)
	id DAA06591; Mon, 24 Jul 1995 03:19:54 -0400
Received: from troms2.or.uninett.no by domen.uninett.no with SMTP (PP) 
          id <25029-0@domen.uninett.no>; Mon, 24 Jul 1995 09:19:12 +0200
Received: from dale.uninett.no (localhost [127.0.0.1]) 
          by dale.uninett.no (8.6.9/8.6.9) with ESMTP id TAA00144 
          for <drums@cs.utk.edu>; Sun, 23 Jul 1995 19:13:59 +0200
Message-Id: <199507231713.TAA00144@dale.uninett.no>
From: Harald.T.Alvestrand@uninett.no
To: drums@CS.UTK.EDU
Subject: Harald's list of DRUMS documents
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <141.806519637.1@dale.uninett.no>
Date: Sun, 23 Jul 1995 19:13:58 +0200
Sender: hta@dale.uninett.no

Hi,
while out walking today, I thought about the documents for DRUMS
in my mind, and decided I wanted to share it with you.
Here's a sketch of what could reasonably be written:

A: The EBNF grammar.
   A short document describing *exactly* the EBNF grammar, without
   reference to BNF or other, non-named external documents.
   Reason: The EBNF grammar is used in too many RFCs.
   It deserves a proper base.

B: Simple Mail Transfer Protocol.
   RFC 821 revisited and updated, with extensions folded in.

C: Internet Mail Message Format.
   RFC 822 revisited and updated, with a parseable EBNF specification.

D: Mail routing in the Connected Internet.
   RFC 974 revisited and updated.

The rest of my ideas are informationals:

E: An EBNF compiler compiler.
   Source code for a program that is able to take an EBNF grammar and
   produce a parser for the syntax thereby defined.
   We need to prove that this can be done!

F: Common good-faith errors in mail messages.
   For each error that we consider and decide to continue outlawing,
   we should list it and the reason behind keeping it outlawed here.

G: Internet Mail outside the Connected Internet.
   Things to think about when running mail behind a firewall, within
   an UUCP network, or similar places. Mainly a place to keep arguments
   for (for instance) keeping all names with FQDNs even locally

I've also got some ideas for statements I would like to put into various
pieces, like:

 "All correct Internet Mail Messages can be parsed by this EBNF grammar.
 However, not all messages parsed by this EBNF grammar are correct
 Internet Mail Messages; in addition, the constraints for each component
 given in section <n> must be satisfied".

One reason why I want a machine-parsable EBNF is that I want the definition
of "legal" messages to be a matter of specification, not a matter of taste!

                           Harald A