On Fri, 13 Jun 1997, era eriksson wrote:
Date: Fri, 13 Jun 1997 09:42:08 +0300 (EET DST)
From: era eriksson <era(_at_)iki(_dot_)fi>
To: procmail(_at_)Informatik(_dot_)RWTH-Aachen(_dot_)DE
Subject: Re: spurious "from" - whose bug is it?
Resent-Date: Fri, 13 Jun 1997 08:47:06 +0200 (MET DST)
Resent-From: procmail(_at_)Informatik(_dot_)RWTH-Aachen(_dot_)DE
[...]
*From* what I've seen while following the problems of the guy whose
Procmail digest would split always when the digest contained an
unquoted From_ line (hi Lance, how you going? :-), I am now under the
impression that this is not frequently done on outgoing mail. Instead,
it would seem that most sites (including my own) escape the From_
lines in the delivery phase.
There seems to be a lot of confusion about who should escape a leading
From_ in the message body. On the outgoing side, pine adds a '>',
elm doesn't, if I remember correctly. On the delivery side, sendmail
adds a '>', procmail v3.11pre3 doesn't always. From `man procmail':
If there is no Content-Length: field and procmail appends to
regular mailfolders, any lines in the body of the message
that look like postmarks are prepended with `>' (disarms
bogus mailheaders). The regular expression that is used to
search for these postmarks is:
`\nFrom '
I have been seriously bitten by this. Somebody once sent me as
postmaster a message which contained a sample message (complete with
From_ header of course) from my site about which he complained. His MUA
(elm, IIRC) generated a Content-Length: field and didn't escape the
From_. My procmail saw the Content-Length: and didn't either. My pine
ignores Content-Length:, but it doesn't take just any for '^From '. As
far as I know, it requires a preceding empty line and something that
looks like an address and a date and nothing else on that line. But that
was of course the case. So I saw some rather confusing messages in my
incoming mail, which added to the confusion my mail system was in at
that time already.
I am not sure whether I should blame procmail or pine or elm or all of
them.
--
Klaus Wacker wacker(_at_)Physik(_dot_)Uni-Dortmund(_dot_)DE
Experimentelle Physik V http://www.physik.uni-dortmund.de/~wacker
Universitaet Dortmund Tel.: +49 (231) 755 3587
D-44221 Dortmund Fax: +49 (231) 755 3569