Sender: Chris Newman <chrisn+@CMU.EDU>
Date: Thu, 7 Mar 1996 20:42:26 -0500 (EST)
From: Chris Newman <chrisn+@CMU.EDU>
Subject: Re: Message format document outline
To: drums@cs.utk.edu
In-Reply-To: <19960307200106.18212.qmail@koobera.math.uic.edu>
References: <19960307200106.18212.qmail@koobera.math.uic.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Message-ID: <826249346.26376.0@nifty.andrew.cmu.edu>

 djb@koobera.math.uic.edu (D. J. Bernstein) writes:
> > Invisible continuation lines must not be generated.  There then
> > becomes no requirement on how they are recognized or interpreted.
>
> This is incompatible with existing practice. Yes, it's unfortunate that
> such messages exist, but they do, and an 822bis parser shouldn't choke
> on a valid 822 message. Mail messages stick around forever.
>
> I suggest must-not-generate-but-must-interpret-properly.

DRUMS is not simply clarifying RFC822, but also changing it to reflect
reality.  The majority of mail messages with what appears to be an invisible
continuation line are actually the result of gateways that map CRLF CRLF to
CRLF SPACE CRLF.  Your proposed "must interpret properly" rule will result in
incorrect real-world behavior more often than correct real-world behavior.

In addition, if you wish to implement "interpret according to RFC822 rules"
in this case, you will still be permitted to do so.
-----
Chris Newman <chrisn+@cmu.edu>, http://www.contrib.andrew.cmu.edu/~nifty/
CMU & Compu$erve: Your freedom of expression ends at their newsfeeds.