From owner-freebsd-bugs  Mon Jan  1 22:19:01 1996
Return-Path: owner-bugs
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id WAA01712
          for bugs-outgoing; Mon, 1 Jan 1996 22:19:01 -0800 (PST)
Received: from limbic.ssdl.com (limbic.ssdl.com [206.109.78.34])
          by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id WAA01703
          Mon, 1 Jan 1996 22:18:53 -0800 (PST)
Received: (from gil@localhost) by limbic.ssdl.com (8.7.2/8.7.2) id AAA11366; Tue, 2 Jan 1996 00:18:40 -0600 (CST)
From: "Gil Kloepfer Jr." <gil@limbic.ssdl.com>
Message-Id: <199601020618.AAA11366@limbic.ssdl.com>
Subject: Re: misc/922
To: peter@freefall.freebsd.org (Peter Wemm)
Date: Tue, 2 Jan 1996 00:18:40 -0600 (CST)
Cc: peter@freefall.freebsd.org, freebsd-bugs@freefall.freebsd.org,
        eric@reference.com, sendmail@sendmail.org
In-Reply-To: <199601011236.EAA20439@freefall.freebsd.org> from "Peter Wemm" at Jan 1, 96 04:36:09 am
X-Mailer: ELM [version 2.4 PL25]
Content-Type: text
Sender: owner-bugs@FreeBSD.ORG
Precedence: bulk

I got two messages on this problem.  I now know what the code in mail.local
does, and as Eric points out, my fix breaks the intent of the code:

| To: "Gil Kloepfer Jr." <gil@limbic.ssdl.com>
| From: Eric Allman <eric@sendmail.org>
| Subject: Re: Bad From line handling in mail.local (fwd) 
| Date: Sat, 30 Dec 1995 09:37:48 -0800
| 
| The point of this is to minimize the amount of body munging that goes
| on.  I believe elm is broken, not mail.local.
| 
| eric

And as Peter stated, here is why elm is broken:

> This is actually intended behavior...
> 
> The end-of-message marker is a blank line followed by "^From "..
> ie:
> "\n\nFrom "
> 
> So, if you have a message like this: (indented for example)
> 
>   foo bar baz
>   From this I can see blah..
> 
> It should not be escaped because there is not a blank line.

I found the bug in elm and will report it there.  In case any of you are
interested, it's in newmbox.c and what needs to be done is to check for
a blank line before "From " in order to determine that it is entering
a new mail header (except, of course, on the first message).  I have a
fix, but I will let it go through the normal channels unless anyone wants
my fix.

My apologies about the mail.local bug scare.  It was a fix, just not the
right one!! :)   Thanks again for your quick feedback!  Happy New Year!

Gil.