Received: from localhost by CS.UTK.EDU with SMTP (cf v2.9s-UTK)
	id WAA11234; Thu, 7 Mar 1996 22:25:36 -0500
Received: by CS.UTK.EDU (bulk_mailer v1.4); Thu, 7 Mar 1996 22:25:02 -0500
Received: from leibniz.math.psu.edu by CS.UTK.EDU with ESMTP (cf v2.9s-UTK)
	id WAA11149; Thu, 7 Mar 1996 22:24:58 -0500
Received: from augusta.math.psu.edu (augusta.math.psu.edu [146.186.132.2]) by leibniz.math.psu.edu (8.6.12/8.6.9) with ESMTP id WAA13316 for <drums@cs.utk.edu>; Thu, 7 Mar 1996 22:24:48 -0500
Received: from localhost (barr@localhost) by augusta.math.psu.edu (8.7.3/8.6.9) with ESMTP id WAA26775 for <drums@cs.utk.edu>; Thu, 7 Mar 1996 22:24:12 -0500 (EST)
Message-Id: <199603080324.WAA26775@augusta.math.psu.edu>
To: drums@cs.utk.edu
Subject: Re: Message format document outline 
In-reply-to: Your message of "Thu, 07 Mar 1996 20:42:26 EST."
             <826249346.26376.0@nifty.andrew.cmu.edu> 
References: <19960307200106.18212.qmail@koobera.math.uic.edu> <826249346.26376.0@nifty.andrew.cmu.edu> 
X-Face: $+9-wYg.[->94HJ{go[7Q]E!K&hUg7ZhLyCMyq_FU*ca0GazE>^/2BKLcK0bP-'%;Nn?M<m
 c~dP_j^A?M5Jw{+a7vcMu|M-mq"V4AzKq?mPMMNP\;f8J0n\-L]ODU)YMP;-pF>+am,jlSP>1K$iz@
 %'v'FEW{@](U&Ed/}>ju3Ctlr!XwJ27Q)7h2a%"`sz;j:/3EC[mXi@*X@HE1]'ddq$ZX"ePsMyTkeg
 >zdML.SV<t=R>vX1W`adGIUD
Date: Thu, 07 Mar 1996 22:24:11 -0500
From: Dave Barr <barr@math.psu.edu>

In message <826249346.26376.0@nifty.andrew.cmu.edu>, Chris Newman writes:
>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.

If this were true, then virtually _no_ messages generated by these
gateways would be valid, since it couldn't properly generate a correct
CRLFCRLF which marks the end of the headers and the start of the body.

I haven't seen this error condition, but I wouldn't be surprised to see
it given a message which contains a large number of trailing spaces at
the end of a long header.   Some mailer along the way (heck, even the
MUA) decides to reformat the header to make it several continued lines
instead of one long line and leaves at the end a whitespace-only line.
Even worse is a condition where it then goes through a VM system which
strips trailing spaces and suddenly part of your header is in the
body!

We have to be very careful here.

--Dave