.TH MH-MAIL 5 RAND
.SH NAME
mh-mail \- Message format for MH
.SH DESCRIPTION
.I MH,
the Rand message-handling system, processes messages in
a particular format.  It should be noted that neither
the Berkeley mailer nor the Bell mail program produces
message files in precisely the correct format for 
.I mh.
.PP
Each user possesses a mail drop box which initially receives
all messages processed by
.I deliver(8).
.I Inc(1)
will read from that drop box and incorporate the new messages
found there into the user's own mail folders.   The mail drop
box consists of one or more messages.  To facilitate the
separation of messages, each message begins and ends with
a line consisting of nothing but four Control-A (octal 001)
characters.
.PP
Each message consists of two parts: a header, consisting of
one or more header lines, and a body, containing the actual
text of the message.  These parts are separated by an empty
line, i.e. two consecutive Newline characters.
Within MH, the header and body may be separated by a line
consisting of dashes.
.PP
Each header item can be viewed as a single logical line of ASCII
characters.
If the text of a header item extends across several
real lines, the continuation lines are indicated by leading
spaces or tabs.
.PP
Each header item is called a component and is composed of a
keyword or name, along with associated text.
The keyword begins at the
left margin, may contain spaces or tabs, may not exceed 63
characters, and is terminated by a colon (:).
Certain
components (as identified by their keywords) must follow rigidly
defined formats in their text portions.
.PP
The text for most formatted components (e.g., ``Date:'' and ``Message-Id:'')
is produced automatically.
The only ones entered by the
user are address fields such as ``To:'', ``cc:'', etc.
ARPA addresses
are assigned mailbox names and host computer specifications.
The
rough format is ``mailbox at host'', such as ``Borden at Rand-Unix''.
Multiple addresses are separated by commas.
A missing host is
assumed to be the local host.
.PP 
Following is a list of header components that are considered
meaningful to various MH programs.
.TP 15
.I Component
.I Usage
.br
.TP
Date:
Added by
.I deliver(8),
contains date and time of delivery.
.br
.ns
.TP
From:
Added by
.I deliver(8),
contains the userid of the sender.  But see "Sender:"
.br
.ns
.TP
Sender:
Added by
.I deliver(8)
in the event that the user has supplied a "From:" line,
contains the userid of the sender.
.br
.ns
.TP
To:
Contains address of recipient of message.
.br
.ns
.TP
cc:
Contains addresses of additional recipients.
.br
.ns
.TP
Bcc:
Still more recipients.  However, the Bcc line is not
copied onto the message as delivered, so these recipients
are not listed.
.br
.ns
.TP
fcc:
Causes
.I deliver(8)
to copy the message into the specified folder.
.br
.ns
.TP
Subject:
Sender's commentary.  It is displayed by
.I scan(1).
.br
.ns
.TP
Message-ID:
A unique message identifier glued on by 
.I deliver
if the
.B \-msgid
flag is set.
.br
.ns
.TP
In-Reply-To:
A commentary line added by 
.I repl(1)
when replying to a message.
.br
.ns
.TP
Replied:
Reply date added by
.I repl(1)
under the "annotate" option.
.br
.ns
.TP
Forwarded:
Forwarding date added by
.I forw(1)
under the "annotate" option.
.br
.ns
.TP
Distribute-to:
New recipients for a redistributed message processed
by
.I dist(1).
.br
.ns
.TP
Distribute-cc:
Still more recipients.  See "cc:".
.br
.ns
.TP
Distribute-bcc:
Even more recipients.  See "bcc:.
.br
.ns
.TP
Distribute-fcc:
Copy distributed message into a folder.  See "fcc:".
.br
.ns
.TP
Distribute-from:
Added when redistributing a message.
.br
.ns
.TP
Distribution-date:
Is to distribute-from as Date: is to From:.
.br
.ns
.TP
Distribution-id:
Added if msgid option is in effect.
.br
.ns
.TP
Distributed:
Annotation for
.I dist(1)
.SH FILES
/usr/spool/mail/*
.SH SEE ALSO
\fIThe MH Message Handling System: User's Manual\fR
by B. Borden, R. Gaines, and N. Shapiro
.br
\fIStandard for the Format of ARPA Network Test Messages\fR
by D. Crocker, J. Vittal, K. Pogran, and D. Henderson, Jr.