~vm/vm/8.1.x

\input texinfo
@setfilename vm.info
@settitle VM User's Manual
@dircategory Emacs
@direntry
* VM: (vm).                             A mail reader.
@end direntry

@iftex
@finalout
@end iftex
@c @setchapternewpage odd               % For book style double sided manual.
@c      @smallbook
@tex
\overfullrule=0pt
%\global\baselineskip 30pt      % For printing in double spaces
@end tex
@ifinfo
This file documents the VM mail reader.
@table @asis
@item Copyright (C) 1989, 1991, 1999 Kyle E. Jones
@item Copyright (C) 2003 - 2008 Robert Widhopf-Fenk
@item Copyright (C) 2008 Uday S. Reddy
@end table

Permission is granted to make and distribute verbatim copies of
this manual provided the copyright notice and this permission notice
are preserved on all copies.

@ignore
Permission is granted to process this file through TeX and print the
results, provided the printed document carries copying permission
notice identical to this one except for the removal of this paragraph
(this paragraph not being relevant to the printed manual).

@end ignore
@end ifinfo
@c
@include version.texinfo
@titlepage
@sp 6
@center @titlefont{VM User's Manual}
@sp 4
@center VM Version @value{VERSION}
@sp 5
@page
@vskip 0pt plus 1filll
Copyright @copyright{} 1989, 1991, 1999, 2002, 2003 Kyle E. Jones
Copyright @copyright{} 2003 - 2008 Robert Widhopf-Fenk
Copyright @copyright{} 2008 - 2009 Uday S. Reddy

Permission is granted to make and distribute verbatim copies of
this manual provided the copyright notice and this permission notice
are preserved on all copies.

@end titlepage
@page
@ifnottex
@node Top, Introduction,, (DIR)

This manual documents the VM mail reader, a Lisp program which runs as a
subsystem under Emacs.  The manual is divided into the following
chapters.

This manual corresponds to VM version @value{VERSION}.

@menu
* Introduction::        Overview of the VM interface.
* Starting Up::         What happens when you start VM.
* Selecting Messages::  How to select messages for reading.
* Reading Messages::    Previewing and paging through a message.
* Sending Messages::    How to send messages from within VM.
* Saving Messages::     How to save messages.
* Deleting Messages::   How to delete, undelete and expunge messages.
* Editing Messages::    How to alter the text and headers of a message.
* Marking Messages::    Running VM commands on arbitrary sets of messages.
* Message Attributes::  How to change and undo changes to message attributes.
* Sorting Messages::    How to make VM present similar messages together.
* Digests::             How to read digests under VM.
* Summaries::           How to view and customize the summary of a folder.
* Virtual Folders::     Blurring the boundaries of different physical folders.
* Frames and Windows::  How to customize VM's use of windows and frames.
* Toolbar::             How to configure VM's toolbar.
* Menus::               How to configure VM's menus.
* Faces::               How to configure VM's use of faces.
* Using the Mouse::     Understanding the VM mouse interface.
* Hooks::               How you can make VM run your code at certain times.
* Bugs::                How to report VM bugs
* History and Administration:: Information about VM

For developers:

* Internals::           Information on VM internals for developers

Indices:

* Concept Index::       Menus of key concepts
* Key Index::           Menus of command keys and their references.
* Command Index::       Menus of commands and their references.
* Variable Index::      Menus of variables and their references.

Rights:
* License::             Copying and distribution terms for VM.


@end menu
@end ifnottex

@node Introduction, Starting Up, Top, Top
@unnumbered Introduction

VM (View Mail) is an Emacs subsystem that allows UNIX mail to be read
and disposed of within Emacs.  Commands exist to do the normal things
expected of a mail user agent, such as generating replies, saving
messages to folders, deleting messages and so on.  There are other more
advanced commands that do tasks like bursting and creating digests,
message forwarding, and organizing message presentation according to
various criteria.

You can make VM your default mail user agent by setting @code{mail-user-agent}
to @code{vm-user-agent}, e.g. by @kbd{M-x} @code{customize-variable} @kbd{RET}
@code{mail-user-agent} @kbd{RET}.

To invoke VM, type @kbd{M-x vm}.  VM gathers any mail that has
arrived in your system mailbox and appends it to a file known as your
@dfn{primary inbox}, and visits that file for reading.  @xref{Starting Up}.
A file visited for reading by VM is called the @dfn{current folder}.

If you type @kbd{?} in a VM folder buffer you will get some help, i.e.
@code{vm-help} is called.

@findex vm-scroll-forward
@findex vm-scroll-backward
@kindex SPC
@kindex b
@kindex DEL
If there are any messages in the primary inbox, VM selects the first new
or unread message, and previews it.  @dfn{Previewing} is VM's way of
showing you part of a message and allowing you to decide whether you want
to read it.  @xref{Previewing}.  By default VM shows you the message's
sender, recipient, subject and date headers.  Typing @key{SPC}
(@code{vm-scroll-forward}) exposes the body of the message and flags the
message as read.  Subsequent @key{SPC}'s scroll forward through the
message, @kbd{b} or @key{DEL} scrolls backward.  When you reach the end
of a message, typing @key{SPC} or @kbd{n} moves you forward to preview
the next message.  @xref{Viewing}.

If you do not want to read a message that's being previewed, type
@kbd{n} and VM will move to the next message (if there is one).
@xref{Selecting Messages}.

To save a message to a mail folder use @kbd{s} (@code{vm-save-message}).
VM will prompt you for the folder name in the minibuffer.
@xref{Saving Messages}.

Messages are deleted by typing @kbd{d} (@code{vm-delete-message}) while
previewing or reading them.  The message is not removed right away; VM
makes a note that you want the message to be removed later.  If you
change your mind about deleting a message, select it and type @kbd{u}
(@code{vm-undelete-message}), and the message will be undeleted.
@xref{Deleting Messages}.  The actual removal of deleted messages from
the current folder is called @dfn{expunging} and it is accomplished by
typing @kbd{###} (@code{vm-expunge-folder}).  The message is still present
in the on-disk version of the folder until the folder is saved.

Typing @kbd{h} (@code{vm-summarize}) causes VM to display a window
containing a summary of the contents of the current folder.  The summary is
presented one line per message, by message number, listing each message's
author, date sent, line and byte count, and subject.  Also, various
letters appear beside the message number to indicate that a message is
new, unread, flagged for deletion, etc.  An arrow @samp{->} appears to
the left of the line summarizing the current message.  The summary
format is user configurable, @pxref{Summaries}.

@findex vm-save-folder
@kindex S
When you are finished reading mail the current folder must be saved, so
that the next time the folder is visited VM will know which messages
have been already read, replied to and so on.  Typing @kbd{S}
(@code{vm-save-folder}) saves the folder.  Note that deleted messages are
@emph{not} expunged automatically when you save a folder.
The next time you visit the folder any deleted
messages will still be flagged for deletion.

@vindex vm-folder-file-precious-flag
When a folder is first visited, the value of the variable
@code{vm-folder-file-precious-flag} is used to initialize a
buffer-local instance of @code{file-precious-flag}, which
determines how folders are saved.  A non-nil value causes
folders to be saved by writing to a temporary file and then
replacing the folder with that file.  A nil value causes
folders to be saved by writing directly to the folder without
the use of a temporary file.

@vindex vm-delete-empty-folders
If the folder is empty at the time you save it and the variable
@code{vm-delete-empty-folders} is non-@code{nil}, VM will remove
the zero length folder after saving it.

@findex vm-quit
@findex vm-quit-no-change
@kindex q
@kindex x
To quit visiting a folder you can type @kbd{q} (@code{vm-quit}) or
@kbd{x} (@code{vm-quit-no-change}).  Typing @kbd{q} saves the current
folder before quitting.  Also, any messages flagged new are changed to
be flagged as old and unread, before saving.  The @kbd{x} command quits
a folder without changing the status of new messages, saving or
otherwise modifying the current folder.

@vindex vm-confirm-quit
If the variable @code{vm-confirm-quit} is set to @code{t}
VM will always ask for confirmation before ending a VM
visit of a folder.  A @code{nil} value means VM will ask only
when messages will be lost unwittingly by quitting, i.e. not
removed by intentional delete and expunge.  A value that is
neither @code{nil} nor @code{t} causes VM to ask only when
there are unsaved changes to message attributes or when messages
will be lost.

@findex vm-quit-just-bury
You do not have to quit a folder to continue using Emacs for other
purposes.  (@code{vm-quit-just-bury}) buries the buffers associated with
the current folder deep in Emacs' stack of buffers, but otherwise leaves
the folder visited so that you can resume reading messages quickly.  You
can locate the folder's buffers again by using @code{list-buffers},
which is normally bound to @kbd{C-x C-b}.

@findex vm-quit-just-iconify
Another command you can use if you are using a window system like X
Windows is @code{vm-quit-just-iconify}.  This command buries the
folder's buffers like @code{vm-quit-just-bury} and also iconifies the
current frame.

@findex vm-get-new-mail
@kindex g
At any time while reading mail in any folder you can type @kbd{g}
(@code{vm-get-new-mail}) to check to see if new mail for that folder has
arrived.  If new mail has arrived it will be moved from the spool files
or maildrops associated with the current folder and merged into the
folder.  If you are not in the middle of another message, VM will also
move to the first new or unread message.

If @code{vm-get-new-mail} is given a prefix argument, it will prompt for
another file from which to gather messages instead of the usual spool
files.  In this case the source folder is copied but no messages are
deleted from it as they would be for a spool file.

By default your primary inbox has your system mailbox associated with
it, e.g. @file{/var/spool/mail/kyle}, and so typing @kbd{g} will retrieve
mail from this file.  Your system mailbox is one example of a @dfn{spool
file}, a file that the mail transport system delivers messages into.
You can associate other spool files with your primary inbox and spool
files with other folders by setting the variable
@code{vm-spool-files}.  @xref{Spool Files}.

@node Starting Up, Selecting Messages, Introduction, Top
@chapter Starting Up

@findex vm-load-init-file
@vindex vm-init-file
@kindex L
The first time VM is started in an Emacs session, it attempts to load
the file specified by the variable @code{vm-init-file}, normally
@file{~/.vm}.  If present this file should contain Lisp code, much like
the @file{.emacs} file.  Since VM has well over one hundred
configuration variables, use of the @file{~/.vm} can considerably reduce
clutter in the @file{.emacs} file.  You can reload this file
by typing @kbd{L} (@code{vm-load-init-file}) from within VM.

@findex vm
@vindex vm-primary-inbox
@vindex vm-auto-get-new-mail
@cindex primary inbox
@kbd{M-x vm} causes VM to visit a folder known as your @dfn{primary
inbox}, specified by the variable @code{vm-primary-inbox}.  If the
variable @code{vm-auto-get-new-mail} is set 
non-@code{nil}, VM will gather any new mail that has arrived
and integrate it into your primary inbox.  The default setting for your
primary inbox is the local file @file{~/Mail/inbox}, but a variety of
other options are available.

VM can work with mail folders saved on the local file system.
@xref{Local Folders}.  It can also work with mail folders stored on
remote mail servers, such as POP and IMAP servers.  @xref{Server
Folders}.  Server folders have the advantage that they can be accessed
from multiple locations on the internet.  VM might appear to have a
bias towards local folders due to its history of development.  But it
treats server folders with equal facility.

@findex vm-visit-folder
@findex vm-visit-pop-folder
@findex vm-visit-imap-folder
@kindex v
@kbd{M-x vm-visit-folder} (@kbd{v} from within VM) allows you to visit
any local mail folder.  The folder name will be
prompted for in the minibuffer.  @kbd{M-x vm-visit-pop-folder} and
@kbd{M-x vm-visit-imap-folder} perform similar function for server
folders. 

Once VM has read the folder and assimilated any new mail, the first new or
unread message will be selected, if any.  If there is no such message,
VM will select whatever the selected message was when this folder was last
saved.  If this folder has never been visited and saved by VM, then the
first message in the folder is selected.

@findex vm-mode
@kbd{M-x vm-mode} can be used on a buffer already loaded into Emacs
to put it into the VM major mode so that VM commands can be executed
on it.  This command is suitable for use in Lisp programs, and for
inclusion in @code{auto-mode-alist} to automatically start VM on a
file based on a particular filename suffix.  @code{vm-mode} skips
some of VM's start-up procedures (e.g. starting up a summary) to make
non-interactive use easier.

@vindex vm-startup-with-summary
The variable @code{vm-startup-with-summary} controls whether VM
automatically displays a summary of the folder's contents at startup.  A
value of @code{nil} gives no summary; a value of @code{t} always gives a
summary.  A value that is a positive integer @var{n} means that VM
should generate a summary if there are @var{n} or more messages in
the folder.  A negative value @var{-n} means generate a summary only if
there are @var{n} or fewer messages.  The default value of
@code{vm-startup-with-summary} is @code{t}.

@menu
* Local Folders::     Working with folders on the local file system
* Server Folders::    Working with folders on mail servers
* Thunderbird Folders:: Working with folders managed by Thunderbird
* Getting New Mail::  Retrieving mail from spool files.
* Crash Recovery::    Recovering changes after Emacs or your system dies.
@end menu

@node Local Folders, Server Folders, Starting Up, Starting Up
@section Local Folders

@cindex mbox
@cindex Babyl
@cindex Rmail
@vindex vm-default-folder-type
A local mail folder is simply a file that can be stored on the local
file system.  VM works with the Unix @dfn{mbox} format to store
messages in folders.  It can also work with the @dfn{Babyl} format
used by the Emacs Rmail package.  The subtypes of mboxes handled by VM
are listed in the documentation of the variable
@code{vm-default-folder-type}.

@cindex spool file
@vindex vm-spool-files
@cindex file locking
A @dfn{spool file} is a file where the mail transport system delivers
messages intended for you.  On Unix systems, a program called
@file{/bin/mail} or @file{/bin/mail.local} does this delivery,
although agents such as @file{procmail}, @file{filter} and
@file{slocal} can be invoked from a user's @file{~/.forward} or
@file{~/.qmail} files.  On other systems, incoming mail is typically
delivered to mailboxes on remote mail servers, from where it can be
retrieved through protocols like POP and IMAP.  No matter what the
delivery agent, what all spool files have in common is that mail is
delivered into them by one or more entities apart from VM and that all
access to spool files must therefore be accompanied by the use of some
file locking protocol.

@vindex vm-movemail-program
@vindex vm-movemail-program-switches
VM leaves the task of accessing spool files to @file{movemail}, a
program distributed with Emacs that is written for this purpose.
The variable @code{vm-movemail-program} specifies the name of the
movemail program and defaults to @samp{"movemail"}.  The variable
@code{vm-movemail-program-switches} lets you specify some initial
command line argument to pass to the movemail program.

@cindex crash box
@vindex vm-crash-box
VM transfers the mail from a spool file to a folder via a
temporary file known as the @dfn{crash box}.  The variable
@code{vm-crash-box} names the crash box file for the primary inbox.
VM first copies the mail to the crash box, truncates the system
mailbox to zero messages, merges the crash box contents into the
primary inbox, and then deletes the crash box.  If the system or Emacs
should crash in the midst of this activity, any message not present in
the primary inbox will be either in the system mailbox or the crash
box.  Some messages may be duplicated but no mail will be lost.

If the file named by @code{vm-crash-box} already exists when VM is
started up, VM will merge that file with the primary inbox before
retrieving any new messages from the system mailbox.

@menu
* Spool Files::         Specifying where mail comes from
* POP Spool Files::     How to use a POP mailbox as a spool file
* IMAP Spool Files::    How to use an IMAP mailbox as a spool file
* Index Files::         Using an index to speed up VM starting
* Folder types::        About the mail folder formats handled by VM
@end menu

@node Spool Files, POP Spool Files, Local Folders, Local Folders
@unnumberedsubsec Spool Files

Every folder, including the primary inbox, can have one or more spool
files associated with it.  You make these associations known to VM by
setting the variable @code{vm-spool-files}.

If you only want to associate spool files with your primary inbox, you
can set @code{vm-spool-files} to a list of strings.  By default, the location
of your system mailbox (the spool file that is associated with your
primary inbox) is determined heuristically based on what type of system
you're using.  VM can be told explicitly where the system mailbox is by
setting @code{vm-spool-files} like this:

@example
(setq vm-spool-files '("/var/spool/mail/kyle" "~/Mailbox"))
@end example

@noindent With this setting, VM will retrieve mail for the primary
inbox from first @file{/var/spool/mail/kyle} and then @file{~/Mailbox}.

If the value of @code{vm-spool-files} is @code{nil}, a default value for
@code{vm-spool-files} will be inherited from the shell environmental
variables MAILPATH or MAIL if either of these variables are defined.
This inheritance happens before your init file is loaded, so setting
@code{vm-spool-files} in your init file will override any environmental
variables.

If you want to associate spool files with folders other than or in
addition to the primary inbox, the value of @code{vm-spool-files} must be a
list of lists.  Each sublist specifies three entities, a folder, a spool
file and a crash box.  When retrieving mail for a particular folder, VM
will scan @code{vm-spool-files} for folder names that match the current
folder's name.  The spool file and crash box found in any matching
entries will be used to gather mail for that folder.

For example, you can set @code{vm-spool-files} like this

@example
@group
(setq vm-spool-files
      '(
        ("~/INBOX"      "/var/spool/mail/kyle"      "~/INBOX.CRASH")
        ("~/INBOX"      "~/Mailbox"                 "~/INBOX.CRASH")
        ("~/Mail/bugs"  "/var/spool/mail/answerman" "~/Mail/bugs.crash")
       )
)
@end group
@end example

@noindent The folder @file{~/INBOX} has two spool files associated
with it in this 
example, @file{/var/spool/mail/kyle} and @file{~/Mailbox}.  Another
folder, @file{"~/Mail/bugs"} has one folder
@file{/var/spool/mail/answerman} associated with it.  Note that both of
the @file{~/INBOX} entries used the same crash box.  The crash box can be
the same if the folder name is the same.  Different folders should use
different crashboxes.

@vindex vm-crash-box-suffix
@vindex vm-spool-file-suffixes
An alternate way of specifying folder/spool file associations
is to use the variables @code{vm-spool-file-suffixes} and
@code{vm-crash-box-suffix}.

The value of @code{vm-spool-file-suffixes} should be a list of string suffixes
to be used to create possible spool file names for folders.  Example:

@example
@group
(setq vm-spool-file-suffixes '(".spool" "-"))
@end group
@end example

@noindent With @code{vm-spool-file-suffixes} set this way, if you
visit a 
folder @file{~/mail/beekeeping}, when VM attempts to retrieve new mail for
that folder it will look for mail in @file{~/mail/beekeeping.spool}
and @file{~/mail/beekeeping-} in addition to scanning @code{vm-spool-files}
for matches.  The value of @code{vm-spool-files-suffixes} will not be used
unless @code{vm-crash-box-suffix} is also defined, since a crash box is
required for all mail retrieval from spool files.

The value of @code{vm-crash-box-suffix} should be a string suffix used to
create possible crash box file names for folders.  When VM uses
@code{vm-spool-file-suffixes} to create a spool file name, it will append
the value of @code{vm-crash-box-suffix} to the folder's file name to
create a crash box name.  If the value of @code{vm-spool-files-suffixes}
is @code{nil}, then the value of @code{vm-crash-box-suffix} is not used
by VM.

@vindex vm-make-crash-box-name
@vindex vm-make-spool-file-name
The idea behind @code{vm-spool-file-suffixes} and
@code{vm-crash-box-suffix} is to give you a way to have many
folders with individual spool files associated with them, without
having to list them all in @code{vm-spool-files}.  If you need
even more control of spool file and crash box names, use
@code{vm-make-spool-file-name} and @code{vm-make-crash-box-name}.
The value of both of these should be a function or the name of a
function.  When VM visits a folder, it will call the function
with the name of the folder as an argument, and the function
should return the spool file name or crash box name to be used
for that folder.

If your spool file is on another host, VM supports accessing
spool files on remote hosts using the POP and IMAP protocols.

@node POP Spool Files,IMAP Spool Files,Spool Files,Local Folders
@unnumberedsubsec POP Spool Files
@cindex POP spool files
VM can access spool files on mail servers via the Post Office Protocol
(POP).  To use a POP mailbox as a spool file, you need to use a POP
maildrop specification (@ref{maildrop specification},
@ref{Server Folders}).  Once this is done, VM will retrieve new mail
from the POP mailbox in the same way as it retrieves it from system
mailbox.  The retrieved messages can be automatically removed from the
POP mailbox or retained until a later expunge operation.

@vindex vm-pop-max-message-size
By default VM will retrieve all the messages from a POP mailbox
before returning control of Emacs to you.  If the mailbox is
large, the wait could be considerable.  If you set
@code{vm-pop-max-message-size} to a positive numeric value, VM will not 
automatically retrieve messages larger than this size.  If VM is
retrieving messages because you invoked @code{vm-get-new-mail}
interactively, then VM will ask whether it should retrieve the
large message.  If VM is retrieving messages automatically
(e.g. @code{vm-auto-get-new-mail} is set non-@code{nil}) then VM will skip the
large message and you can retrieve it later.

@vindex vm-pop-bytes-per-session
@vindex vm-pop-messages-per-session
The variable @code{vm-pop-messages-per-session} controls how many messages
VM will retrieve from a POP mailbox before returning control to
you.  Similarly, the variable @code{vm-pop-bytes-per-session} limits the
number of bytes VM will retrieve from a POP mailbox before returning
control to you.  By default, the value of both variables is nil, which
tells VM to retrieve all the messages in the POP mailbox regardless
of how many messages there are and how large the mailbox is.

@findex vm-expunge-pop-messages
After VM retrieves messages from the mailbox, the default action is to
leave the original messages on the server unchanged.  They can be
expunged from the server by running @code{vm-expunge-pop-messages};
only those messages that VM has retrieved into the current folder will
be expunged.

@vindex vm-pop-expunge-after-retrieving
@vindex vm-pop-auto-expunge-alist
If you want VM to expunge the messages automatically after retrieving
them, you can set @code{vm-pop-expunge-after-retrieving} to @code{t}.
But a better method is to set the variable
@code{vm-pop-auto-expunge-alist}, which gives you a way to specify, on
a per-mailbox basis, which POP mailboxes should have messages automatically
removed after retrieving and which ones should leave the messages on the POP
server.  The value of @code{vm-pop-auto-expunge-alist} should be a
list of POP mailboxes and values specifying whether messages should
be automatically deleted from the mailbox after retrieval.  The format
of the list is:

@example
((@var{MAILDROP} . @var{VAL}) (@var{MAILDROP} . @var{VAL}) ...)
@end example

@var{MAILDROP} should be a POP mailbox specification as described
in the documentation for the variable @code{vm-spool-files}.  If
you have the POP password specified in the @code{vm-spool-files}
entry, you do not have to specify it here as well.  Use @samp{*}
instead; VM will still understand that this mailbox is the same as
the one in @code{vm-spool-files} that contains the password.

@var{VAL} should be @code{nil} if retrieved messages should be left in the
corresponding POP mailbox, @code{t} if retrieved messages should be
removed from the mailbox immediately after retrieval.

Here is an example:

@example
(setq vm-pop-auto-expunge-alist
   '(
     ("odin.croc.net:110:pass:kyle:*" . nil)  ;; leave message on the server
     ("hilo.harkie.org:110:pass:kyle:*" . t)  ;; expunge immediately
    )
)
@end example

@node IMAP Spool Files, Index Files, POP Spool Files, Local Folders
@unnumberedsubsec IMAP Spool Files
@cindex IMAP spool files
@cindex maildrop specification
VM can also use the IMAP protocol to retrieve mail from a mail server.
As with POP, instead of specifying a local file name in the
@code{vm-spool-files} definition, you would give an IMAP maildrop
specification (@ref{maildrop specification}, @ref{Server Folders}).
Once this is done, VM will retrieve new mail from the IMAP mailbox in
the same way as it retrieves it from system mailbox.  The retrieved
messages can be automatically removed from the IMAP mailbox or
retained until a later expunge operation.

@vindex vm-imap-bytes-per-session
@vindex vm-imap-messages-per-session
The variable @code{vm-imap-messages-per-session} controls how many messages
VM will retrieve from an IMAP mailbox before returning control to
you.  Similarly, the variable @code{vm-imap-bytes-per-session} limits the
number of bytes VM will retrieve from an IMAP mailbox before returning
control to you.  By default, the value of both variables is nil, which
tells VM to retrieve all the messages in the IMAP mailbox regardless
of how many messages there are and how large the mailbox is.

@cindex expunging, IMAP messages
@findex vm-expunge-imap-messages
After VM retrieves messages from the mailbox, the default action is to
leave the original messages on the server unchanged.  They can be
expunged from the server by running @code{vm-expunge-imap-messages};
only those messages that VM has retrieved into the current folder will
be expunged.

@vindex vm-imap-expunge-after-retrieving
@vindex vm-imap-auto-expunge-alist
If you want VM to expunge the messages automatically after retrieving
them, you can set @code{vm-imap-expunge-after-retrieving} to @code{t}.
But a better method is to set the variable
@code{vm-imap-auto-expunge-alist}, which gives you a way to specify, on
a per-mailbox basis, which IMAP mailboxes should have messages automatically
removed after retrieving and which ones should leave the messages on the IMAP
server.  The value of @code{vm-imap-auto-expunge-alist} should be a
list of IMAP mailboxes and values specifying whether messages should
be automatically deleted from the mailbox after retrieval.  The format
of the list is:

@example
((@var{MAILDROP} . @var{VAL}) (@var{MAILDROP} . @var{VAL}) ...)
@end example

@var{MAILDROP} should be an IMAP maildrop specification as described
in the documentation for the variable @code{vm-spool-files}.  If
you have the IMAP password specified in the @code{vm-spool-files}
entry, you do not have to specify it here as well.  Use @samp{*}
instead; VM will still understand that this mailbox is the same as
the one in @code{vm-spool-files} that contains the password.

@var{VAL} should be @code{nil} if retrieved messages should be left in the
corresponding IMAP mailbox, @code{t} if retrieved messages should be
removed from the mailbox immediately after retrieval.

Here is an example:

@example
(setq vm-imap-auto-expunge-alist
   '(
     ;; leave message on the server
     ("imap:odin.croc.net:143:inbox:login:kyle:*" . nil)
     ;; expunge immediately
     ("imap:hilo.harkie.org:143:inbox:login:kyle:*" . t)
    )
)
@end example

@unnumberedsubsubsec Multiple access to IMAP spool files

A principal idea behind the IMAP protocol is that messages can be
retained on the server so that you can read them from multiple
locations, e.g., from office and home, or from other places on the
Internet while you travel.  If you access your IMAP mailbox from
multiple locations then you would need to plan your strategy for
expunging messages carefully.  For instance, if you access your work
mailbox from home, and both your office machine and home machine
expunge messages after retrieving them, then some of your mail will
end up on your office machine and some on your home machine.  That is
unlikely to be a successful strategy.

The best way to access IMAP mailboxes from multiple locations is to
use the facility of IMAP folders.  (@xref{POP and IMAP Folders}.)
However, if you prefer to download all mail to local folders, then
your best bet is to designate one of your machines as the principal
location for downloading mail and treat the other machines as
temporary mail reading sites.  In that case, you should set the
principal downloading location to expunge messages on the server and
set the other reading sites to leave the messages on the server
intact.  You can also manually run @code{vm-expunge-imap-messages} if
you are careful to remember which site should expunge messages and
which site should retain them.

@cindex X-VM-IMAP-Retrieved header
VM remembers the messages you have downloaded from an IMAP spool file
so that it can avoid downloading them again on your next visit.  The
list of these messages is written into a special mail header titled
@code{X-VM-IMAP-Retrieved} in your mail folder.  When you expunge IMAP
messages, their entries are deleted from the list.  However, when you
designate one of your machines as a reading site and never expunge
messages from there, then the @code{X-VM-IMAP-Retrieved} header on
that machine will only grow over time.  When the list gets excessively
long, it will slow down the saving of folders.

@findex vm-prune-imap-retrieved-list
To avoid the problem, you should prediodically run the command
@code{vm-prune-imap-retrieved-list}.  It will examine the IMAP server
to see which messages still exist and retain only their information in
the @code{X-VM-IMAP-Retrieved} header.


@node Index Files, Folder types,IMAP Spool Files, Local Folders
@unnumberedsubsec Index Files
@cindex index file

VM can create an index file which describes the messages contained in
a folder.  If such an index file exists and is up to date, then VM
will read the contents of the index file first while starting up in
order to quickly form the summary of the folder.

@vindex vm-index-file-suffix
To use this feature, set the variable @code{vm-index-file-suffix} to a
file name extension, e.g., 

@example
(setq vm-index-file-suffix "idx")
@end example


@node Folder types,,Index Files, Local Folders
@unnumberedsubsec Folder types
@cindex folder types

@vindex vm-default-folder-type
VM can handle a variety of formats for mail folders, which differ in
details.  The variable @code{vm-default-folder-type} can be used to
set the default format that is suitable for your environment.  This
setting is used when VM creates new folders.

When VM reads a folder from the file system, it examines contents of the
folder to determine what format it is stored in and decodes it
appropriately.  (However, such inference is not fully automatic.  See
below.)

@findex vm-change-folder-type
After a folder is loaded into VM, you can convert it to a different
format using the command @code{vm-change-folder-type}.  It is a good
idea to keep all your mail folders in a single format in order to avoid
incompatibilities.

The system default format is referred to as @code{From_}.  It is the
Unix mbox format described RFC 4155.  In this format, a leading
separator line and a trailing separator line are added to each message.
The leading separator line starts with the string ``From ''.  The
trailing separator line is a blank line.  VM actually adds two blank lines at
the end for clarity.

A variant of this format is referred to as @code{BellFrom_}.  It has a
leading separator line that starts with the string ``From ''.  However,
it does not have a trailing blank line.

Since VM cannot reliably infer whether a mail folder is of type
@code{From_} or @code{BellFrom_}, you must tell VM which one your system
uses by setting the variable @code{vm-default-From_-folder-type}.  Some
of the old folders created by VM prior to 2000 were in the
@code{BellFrom_} format.  If you will be using both @code{From_} and
@code{BellFrom_} style folders, it is not possible to choose an
appropriate setting for this variable.  It is recommended that you
convert all the old @code{BellFrom_} folders to the @code{From_} format using
the command @code{vm-change-folder-type}.

Solaris, System V and AIX operating systems use another variant of the
mbox format where the content-length is specified in the ``From '' line.
VM refers to this format as @code{From_-with-Content-Length}.  Since the
content lengths may be unreliable, you must also set the variable
@code{vm-trust-From_-with-Content-Length} to a non-Nil value in order to
convince VM that you really want to use this format.

Two additional formats are @code{mmdf} used by MMDF systems and
@code{babyl} used by the Emacs Rmail mode.  These formats are recognized
automatically when read from the file system.


@node Server Folders, Thunderbird Folders, Local Folders, Starting Up
@section Server Folders

@cindex primary inbox
@cindex maildrop specification
@vindex vm-primary-inbox
VM supports accessing remote mailboxes on mail servers via the Post
Office Protocol (POP) and the Internet Message Access Protocol (IMAP).
Instead of a local file name, you can set the @code{vm-primary-inbox} to
a string that tells VM how to access a server mailbox. Called a
@dfn{maildrop specification}, the string is of one of the
following formats:

@example
``pop:@var{HOST}:@var{PORT}:@var{AUTH}:@var{USER}:@var{PASSWORD}''
``imap:@var{HOST}:@var{PORT}:@var{MAILBOX}:@var{AUTH}:@var{USER}:@var{PASSWORD}''@end example

@noindent Remote mailboxes accessed by VM in this fashion are referred
to as @dfn{server folders} (and @dfn{POP folders} or @dfn{IMAP
folders}, more specifically).

VM retrieves mail from the server folders into internal Emacs buffers
for its normal operation.  It also saves copies of the folders on the
local file system for speed of operation.  However, the @emph{only}
permanent copies of the folders are on the mail server.  This should
be contrasted with using server mailboxes as spool files (
@ref{POP Spool Files} and @ref{IMAP Spool Files}), where the permanent
folders are on the @emph{local} file system and only incoming mail is
held on the servers.

Server folders have the advantage that they can be transparently
accessed from multiple locations on the internet.  However, you must
ensure that you have access to enough storage on the mail server to
store all your email.

@anchor{maildrop specification}
@unnumberedsubsec Maildrop specification
The format of a POP or IMAP maildrop specification is as follows:

@example
``pop:@var{HOST}:@var{PORT}:@var{AUTH}:@var{USER}:@var{PASSWORD}''
``imap:@var{HOST}:@var{PORT}:@var{MAILBOX}:@var{AUTH}:@var{USER}:@var{PASSWORD}''
@end example

@noindent Replace @samp{pop} in the example with @samp{pop-ssl} to
have VM speak POP over an SSL connection.  Use @samp{pop-ssh} to use
POP over an SSH connection.  Similarly, replace @samp{imap} with
@samp{imap-ssl} or @samp{imap-ssh}, as needed.

For SSL, you must have the stunnel program installed and the
variable @code{vm-stunnel-program} must name it in order for
POP/IMAP over SSL to work.  The default value of this variable,
@samp{"stunnel"}, should be sufficient if the program is
installed in your normal command search path.

For SSH, you must have the ssh program installed and the variable
@code{vm-ssh-program} must name it in order for POP/IMAP over SSH to
work.  When VM makes the SSH connection it must run a command on
the remote server so that the SSH session is maintained long enough
for the POP/IMAP connection to be established.  By default that command
is @samp{"echo ready; sleep 10"}, but you can specify another
command by setting @code{vm-ssh-remote-command}.  Whatever
command you use must produce some output and hold the connection
open long enough for VM to establish a port-forwarded connection
to the mail server.  (SSH must be able to authenticate without a password,
which means you must be using .shosts authentication or RSA.)

@var{HOST} is the host name of the mail server.  

@var{PORT} is the TCP port number to connect to.  The normal port
numbers are:
@multitable @columnfractions 0.20 0.80
@item 110 @tab for POP
@item 995 @tab for POP over SSL
@item 143 @tab for IMAP
@item 993 @tab for IMAP over SSL
@end multitable

@var{MAILBOX} is the name of the mailbox on the IMAP server.  This should
be @samp{"inbox"}, to access your default IMAP mailbox on the
server.  No @var{MAILBOX} component is needed for POP maildrops
because POP does not support multiple mailboxes.

@vindex vm-pop-md5-program
@var{AUTH} is the authentication method used to convince the
server you should have access to the mailbox.  Acceptable
values for POP are @samp{pass}, @samp{rpop} and @samp{apop}.  For
@samp{pass}, the @var{PASSWORD} is sent to the server with
the POP PASS command.  For @samp{rpop}, the @var{PASSWORD}
should be the string to be sent to the server via the RPOP
command.  In this case the string is not really a secret;
authentication is done by other means.  For @samp{apop}, an
MD5 digest of the @var{PASSWORD} appended to the server
time-stamp will be sent to the server with the APOP command.
If Emacs does not have built in MD5 support, you will have
to set the value of @code{vm-pop-md5-program} appropriately
to point at the program that will generate the MD5 digest
that VM needs.

@vindex vm-imap-session-preauth-hook
Acceptable values of @var{AUTH} for IMAP
are @samp{"preauth"}, @samp{"cram-md5"}, and @samp{"login"}.
@samp{"preauth"} causes VM to skip the authentication stage of
the protocol with the assumption that the session was
authenticated in some way external to VM.  The hook
@code{vm-imap-session-preauth-hook} is run, and it is expected to
return a process connected to an authenticated IMAP session.
@samp{"cram-md5} tells VM to use the CRAM-MD5 authentication
method as specified in RFC 2195.  The advantage of this method
over the @samp{"login"} method is that it avoids sending your
password over the net unencrypted.  Not all IMAP servers support
@samp{"cram-md5"}; if you're not sure, ask your mail
administrator or just try it.  The other value, @samp{"login"},
tells VM to use the IMAP LOGIN command for authentication, which
sends your user name and password in clear text to the server.

@var{USER} is the user name used in authentication methods that
require such an identifier.  @samp{"login"} and @samp{"cram-md5"}
use it currently.

@var{PASSWORD} is the
secret shared by you and the server for authentication purposes.  How
it is used depends on the value of the @var{AUTH} parameter.  If the
@var{PASSWORD} is @samp{*}, VM will prompt you for the password the
first time you try to retrieve mail from the mailbox.  If the password
is valid, VM will not ask you for the password again during this Emacs
session.


@menu
* POP Folders::         How to use mailboxes on POP servers
* IMAP Folders::        How to use mail folders on IMAP servers
@end menu

@node POP Folders, IMAP Folders, Server Folders, Server Folders
@unnumberedsubsec POP Folders
@cindex POP
@cindex message attributes


@findex vm-visit-pop-folder
The command @code{vm-visit-pop-folder} allows you to visit a POP
mailbox as a folder.  When you visit a POP folder, VM will download
copies of the messages that it finds there for you to read.  These
messages are saved locally in the file system, in the directory
specified by @code{vm-pop-folder-cache-directory} (or
@code{vm-folder-directory} if the former is not defined).
@vindex vm-pop-folder-cache-directory
If you delete and expunge messages in the folder, the
corresponding messages on the POP server will be removed when you
save the changes with @code{vm-save-folder}.

Message attributes (new, replied, filed, etc.) and labels cannot be
stored on the POP server but they will be maintained in the local
copy.  This means that if you access the same POP mailbox from
multiple locations on the internet, you will see different attributes
at different locations.  To be able to store message attributes and
labels on the server, you should use IMAP folders (@ref{IMAP Folders})
resident on an IMAP server.

@vindex vm-pop-folder-alist
In order for VM to know about POP folders that you can access, you
must declare them by setting the variable @code{vm-pop-folder-alist}.
The variable's value should be an associative list of the form:

@example
 ((@var{POPDROP} @var{NAME}) ...)
@end example

@var{POPDROP} is a POP maildrop specification (@ref{maildrop specification}).

@var{NAME} is a string that should give a less cumbersome name that you
will use to refer to this maildrop when using @code{vm-visit-pop-folder}.

For example:

@example
(setq vm-pop-folder-alist
      '(
         ("pop:pop.mail.yahoo.com:110:pass:someuser:*" "Yahoo! mail")
         ("pop:localhost:110:pass:someuser:*" "local mail")
       )
)
@end example

@samp{Yahoo! mail} and @samp{local mail} are what you would type
when @code{vm-visit-pop-folder} asks for a folder name.  There is no
need to specify the password for POP accounts in this definition.

@node IMAP Folders,, POP Folders, Server Folders
@unnumberedsubsec IMAP Folders
@cindex IMAP
@cindex message attributes 
@cindex message labels

@findex vm-visit-imap-folder
The command @code{vm-visit-imap-folder} allows you to visit an IMAP
mailbox as a folder.  The name of the IMAP mailbox should be input via
the minibuffer in the format account-name:folder-name.  Here,
``account-name'' is the name of an IMAP account declared in
@code{vm-imap-account-alist} and ``folder-name'' is the name of an IMAP
mailbox in this account.

When you visit an IMAP folder, VM will
download copies of the messages that it finds there for you to read.
These messages are saved locally in a folder on the disk, in the 
directory specified by @code{vm-imap-folder-cache-directory} (or
@code{vm-folder-directory} if the former is not defined).
@vindex vm-imap-folder-cache-directory
If you delete and expunge messages, these changes are made to both the
local copy of the folder and the folder on the IMAP server when saved
with @code{vm-save-folder}.

Message attributes (new, replied, filed, etc.) are stored on the IMAP
server and are also cached locally.  Message labels are also stored on
the IMAP server as user-defined permanent flags.  (This assumes that
the IMAP server has the ability to store user-defined permanent
flags.)

@vindex vm-imap-account-alist
In order for VM to know about IMAP accounts that you can access, you
must declare them by setting the variable @code{vm-imap-account-alist}.
The variable's value should be an associative list of the form:

@example
 ((@var{IMAPDROP} @var{NAME}) ...)
@end example

@var{IMAPDROP} is an IMAP maildrop specification (@ref{maildrop
  specification}). 

@var{NAME} is a string that should give a less cumbersome name that you
will use to refer to this maildrop when using @code{vm-visit-imap-folder}.
For example:

@example
(setq vm-imap-account-alist
      '(
        ("imap-ssl:mail.foocorp.com:993:*:login:becky:*" "becky")
        ("imap:crickle.lex.ky.us:143:*:login:becky:*" "crickle")
       )
)
@end example

@noindent The mailbox and password fields (@samp{*} in the example) are
ignored.  When @code{vm-visit-imap-folder} asks for a folder name, you
enter an account name followed by ``:'' and a folder name.  Any folder
that is accessible to you on the IMAP server can be specified.  For
example, @code{becky:inbox} or @code{crickle:drafts}.

@vindex vm-imap-server-list
The customization variable @code{vm-imap-server-list}, used in older
versions of VM, is deprecated.  Please use @code{vm-imap-account-alist}
instead.

@anchor{IMAP Synchronization}
@unnumberedsubsec IMAP Synchronization
The local copy and the folder on the IMAP server are partially
synchronized every time @code{vm-get-new-mail} is invoked.  This
involves (i) writing the changed attributes and labels to the server,
(ii) updating the attributes and labels in the local copy based on the
server data, (iii) expunging messages in the local copy that have been
expunged on the server (through perhaps a concurrent access to the
IMAP folder), and finally, (iv) retrieving any new messages on the
server.  The variable @code{vm-imap-sync-on-get} specifies whether
such synchronization should be done as part of @code{vm-get-new-mail}.
@vindex vm-imap-sync-on-get
If the variable is set to nil then @code{vm-get-new-mail} simply
retrieves any new messages.

The local copy and the folder on the IMAP server are also synchronized
every time @code{vm-save-folder} is invoked.  This involves (i)
writing the changed attributes and labels to the server, (ii) updating
the attributes and labels in the local copy based on the server data,
(iii) expunging messages in the local copy that have been expunged on
the server (through perhaps a concurrent access to the IMAP folder),
(iv) deleting the locally expunged messages on the server, and
finally, (v) saving a copy of the folder on the file system.

@findex vm-imap-synchronize
The command @code{vm-imap-synchronize} can always be used to perform
full synchronization with the server.

After fetching messages from the IMAP server into the local copy, it
is possible to visit the local copy as if it were a normal folder.  VM
can operate on the local copy without contacting the server.  This
allows offline operation on the mail folders.  When the IMAP server is
connected again, one should do the @code{vm-imap-synchronize} command
with a prefix argument.  This causes @emph{all} the message attributes
and labels to be written to the server, since it may not be known
which of them have actually changed during the offline operation.




@node Thunderbird Folders, Getting New Mail, Server Folders, Starting Up
@section Thunderbird Folders

@cindex Thunderbird
VM can work with local folders managed by Mozilla Thunderbird.  You can
find the location of Thunderbird's folders by examining the Account
Settings for ``Local Folders'' inside Thunderbird.

Thunderbird stores the folders in the @code{From_} folder type.
@xref{Folder types}.  Within such folders, Thunderbird stores the
message status flags (message attributes such as whether a message is
read, replied to, deleted etc.) under special header fields called
@code{X-Mozilla-Status} and @code{X-Mozilla-Status2}.  In addition to
these headers, Thunderbird also stores a quick copy of the message
status flags in a separate file with the extension @code{.msf}.

When you visit a Thunderbird folder, VM reads the status flags stored in
the special headers and uses them for processing.  As you make changes
to the folder by reading messages, replying to them or deleting them,
the changes are propagated to the Thunderbird status flags and written
to the disk when saved.  VM also deletes the @code{.msf} file maintained
by Thunderbird so that Thunderbird will recompute the status information
from the headers.  Thus, the changes made to the Thunderbird folders
will be visible inside Thunderbird.

@vindex vm-sync-thunderbird-status
The variable @code{vm-sync-thunderbird-status} controls how VM deals
with Thunderbird folders.  The default value @code{t} gives the behavior
described above.  You can also set it to @code{'read-only}, in which
case VM reads the Thunderbird status flags, but makes no changes to
them.  So, the changes made to the folders will be lost after you quit
VM.  If you set it to @code{nil}, then VM refrains from reading and
writing the Thunderbird status flags.  In this case, the changes made to
the folders are visible inside VM even after revisiting, but they will have
no effect for Thunderbird.  

WARNING: Keep in mind that all this applies to changes to message
attributes only.  If you @i{expunge} a folder, then the deleted messages
are physically purged from the folder.  They will be lost both inside VM
as well as Thunderbird.

The variable @code{vm-sync-thunderbird-status} is a buffer-local
variable.  You may set its default value in your @code{.vm} file.  To
change it in a running Emacs session, you must use @code{setq-default}.
@xref{Locals,, Local Variables, emacs, Gnu Emacs manual}.

@node Getting New Mail, Crash Recovery, Thunderbird Folders, Starting Up
@section Getting New Mail

@findex vm-get-new-mail
Pressing @kbd{g} runs @code{vm-get-new-mail}, which will retrieve
mail from all the spool files associated with the current folder.
@xref{Local Folders}.  For POP folders, any newly arrived messages
at the POP server will be incorporated into the local copy of the
POP folder.

@vindex vm-auto-get-new-mail
If the value of the variable @code{vm-auto-get-new-mail} is non-@code{nil} VM
will retrieve mail for a folder whenever the folder is visited.  If the
value is a positive integer @var{n}, VM will also check for new mail
every @var{n} seconds for all folders currently being visited.  If new
mail is present, VM will retrieve it.

@vindex vm-mail-check-interval
If the value of the variable @code{vm-mail-check-interval} is a
positive integer @var{n}, VM will check for new mail every @var{n}
seconds, but instead of retrieving mail, the word ``Mail'' will
appear on the Emacs mode line of folders that have mail waiting.

@node Crash Recovery,, Getting New Mail, Starting Up
@section Crash Recovery
@cindex message attributes
@cindex message labels

@cindex autosave
When Emacs crashes, its last action before dying is to try to
write out an autosave file that contains changes to files that
you were editing.  VM folders are file buffers inside Emacs, so
folders are autosaved also.  Changes, with regard to VM folders,
means attribute changes, label additions and deletions, message
edits, and expunges.  VM keeps track of whether a message is new
or old, whether it has been replied to, whether it is flagged
for deletion and so on, by writing special headers into the
folder buffer.  These headers are saved to disk when you save
the folder.  If Emacs crashes before the folder has been saved,
VM may forget some attribute changes unless they were written to
the autosave file.

Note that when VM retrieves mail from spool files it @emph{always}
writes them to disk immediately and at least one copy of the message is
on disk at all times.  So while you can lose attribute changes from
crashes, you should not lose messages unless the disk itself is
compromised.

When you visit a folder, VM checks for the existence of an
autosave file that has been modified more recently than the
folder file.  If such an autosave file exists, there is a good
chance that Emacs or your operating system crashed while VM
was visiting a folder.  VM will then write a message to the echo
area informing you of the existence of the autosave file and
visit the folder in read-only mode.  Visiting the folder in
read-only mode prevents you from modifying the folder, which
in turn prevents Emacs from wanting to write new changes to
the autosave file.  VM will not retrieve new mail for a folder
that is in read-only mode.  VM also skips summary
generation and MIME decoding to help catch your attention.

@findex vm-recover-folder
@findex recover-file
If you want to recover the lost changes, run @kbd{M-x
vm-recover-folder} or use the Recover Folder entry in Folder menu.  At
the @samp{Recover File: } prompt press @kbd{RET}.  (Emacs's built-in
@kbd{recover-file} command is not recommended for this purpose because
VM is unable to obtain reliable data regarding the mail folders from
Emacs.)  Emacs will then display a detailed directory listing showing
the folder file and the autosave file and ask if you want to recover
from the autosave file.  A good rule of thumb is to answer ``yes'' if
the autosave file is larger than the folder file.  If the autosave
file is significantly smaller, Emacs may not have completed writing
the autosave file.  Or it could be that the smaller autosave file
reflects the results of an expunge that you had not yet committed to
disk before the crash.  If so, answering ``no'' means you might have
to do that expunge again, but this is better than not knowing whether
the autosave file was truncated.

Assuming you answered ``yes'', the folder buffer's contents will be
replaced by the contents of the autosave file and VM will re-parse the
folder.  At this point the contents of the folder buffer and the disk
copy of the folder are different.  Therefore VM will not get new mail
for this folder until the two copies of the folder are synchronized.
When you are satisfied that the recovered folder is whole and intact,
type @kbd{S} to save it to disk.  After you do this, VM will allow you
to use @kbd{g} to retrieve any new mail that has arrived in the spool
files for the folder.

Assuming you answered ``no'' to the recovery question, you should type
@kbd{C-x C-q}, which is bound to @code{vm-toggle-read-only} in VM folder
buffers.  The folder will be taken out of read-only mode and you can
read and retrieve your mail normally.

@findex vm-revert-folder
@findex revert-file
If you have made changes to a mail folder which you would like to
cancel and go back to the version currently on the disk, you can use
the function @kbd{vm-revert-folder} or the Revert Folder entry in the
Folder menu.  (Emacs's built-in @kbd{revert-file} is not recommended.)

@node Selecting Messages, Reading Messages, Starting Up, Top
@chapter Selecting Messages

@findex vm-next-message
@findex vm-previous-message
@kindex n
@kindex p
@vindex vm-skip-deleted-messages
@vindex vm-skip-read-messages
In order to read, delete, or do anything to a message, you need to
select it.  In other words, make the message the @dfn{current message}.

The primary commands for selecting messages in VM are @kbd{n}
(@code{vm-next-message}) and @kbd{p}
(@code{vm-previous-message}).  These commands move forward and
backward through the current folder.  By default, these commands
skip messages flagged for deletion.  This behavior can be
disabled by setting the value of the variable
@code{vm-skip-deleted-messages} to @code{nil}.  These commands
can also be made to skip messages that have been read; set
@code{vm-skip-read-messages} to @code{t} to do this.

The commands @kbd{n} and @kbd{p} also take prefix arguments that
specify the number of messages to move forward or backward.  If
the magnitude of the prefix argument is greater than 1, no
message skipping will be done regardless of the settings of the
skip variables.

@vindex vm-circular-folders
The variable @code{vm-circular-folders} determines whether VM folders
will be considered circular by various commands.  @dfn{Circular} means VM
will wrap from the end of the folder to the start and vice versa when
moving the message pointer, deleting, undeleting or saving messages
before or after the current message.

A value of @code{t} causes all VM commands to consider folders circular.
A value of @code{nil} causes all VM commands to signal an error if
the start or end of the folder would have to be passed to complete the
command.  For movement commands, this occurs after the message pointer
has been moved as far as it can go.  For other commands the error occurs
before any part of the command has been executed, i.e. no deletions, saves,
etc. will be done unless they can be done in their entirety.  A value
other than @code{nil} or @code{t} causes only VM's movement
commands to consider folders circular.  Saves, deletes and undeletes
will behave as if the value is @code{nil}.  The default value of
@code{vm-circular-folders} is @code{nil}.

@vindex vm-follow-summary-cursor
You can also select messages by using the summary window.
@xref{Summaries}.  Move the cursor to the summary line for the message
you want to select and press @kbd{RET}.  VM will select this message.
Instead of pressing @kbd{RET} you could run some other VM command that
operates based on the notion of a `current message'.  VM will select the
message under the cursor in the summary window before executing such
commands.  Example, if you type @kbd{d}, VM will select the message
under the cursor and then delete it.  Note that this occurs @emph{only}
when you execute a command when the cursor is in the summary buffer
window and only if the variable @code{vm-follow-summary-cursor} is
non-@code{nil}.

@vindex vm-jump-to-unread-messages
@vindex vm-jump-to-new-messages
When a folder is visited or when you type @kbd{g} and VM retrieves some
mail, the default action is to move to the first new or unread message
in the folder.  New messages are favored over old but unread messages.
If you set @code{vm-jump-to-new-messages} to @code{nil}, VM will favor old,
unread messages over new messages if the old, unread message appears
earlier in the folder.  If you set @code{vm-jump-to-unread-messages} to
@code{nil} also, VM will not search for new or unread messages.

@cindex searching
Other commands to select messages:

@table @kbd
@findex vm-goto-message
@kindex RET
@item RET (@code{vm-goto-message})
Go to message number @var{n}.  @var{n} is the prefix argument, if
provided, otherwise it is prompted for in the minibuffer.
@findex vm-goto-message
@kindex TAB
@item TAB (@code{vm-goto-message-last-seen})
Go to message last previewed or read.
@findex vm-next-message-no-skip
@findex vm-previous-message-no-skip
@kindex N
@kindex P
@item N (@code{vm-next-message-no-skip})
@itemx P (@code{vm-previous-message-no-skip})
Go to the next (previous) message, ignoring the settings of the skip
control variables.
@findex vm-next-unread-message
@findex vm-previous-unread-message
@kindex M-n
@kindex M-p
@item M-n (@code{vm-next-unread-message})
@itemx M-p (@code{vm-previous-unread-message})
Move forward (backward) to the nearest new or unread message.
@findex vm-isearch-forward
@findex vm-isearch-backward
@kindex M-s
@comment @kindex M-r
@vindex vm-search-using-regexps
@item M-s (@code{vm-isearch-forward})
@item M-x vm-isearch-backward
These work just like Emacs' normal forward and backward incremental
search commands, except that when the search ends, VM selects the
message containing point.  If the value of the variable
@code{vm-search-using-regexps} is non-@code{nil}, a regular expression
may be used instead of a fixed string for the search pattern; VM
defaults to the fixed string search.  If a prefix argument is given,
the value of @code{vm-search-using-regexps} is temporarily reversed for
the search.
@xref{Incremental Search,,,emacs, the GNU Emacs Manual}.
@end table

@node Reading Messages, Sending Messages, Selecting Messages, Top
@chapter Reading Messages

Once a message has been selected, VM will show it to you.  By default,
presentation is done in two stages: @dfn{previewing} and @dfn{paging}.

@menu
* Previewing::                  Customizing message previews.
* Viewing::                     Viewing the current message.
* MIME Messages::               Using VM's MIME display features.
@end menu

@node Previewing, Viewing, Reading Messages, Reading Messages
@section Previewing

@dfn{Previewing} means showing you a small portion of a message
and allowing you to decide whether you want to read it.  Typing
@key{SPC} exposes the body of the message, and from there you can
repeatedly type @key{SPC} to page through the message.

By default, the sender, recipient, subject and date headers are shown
when previewing; the rest of the message is hidden.  This behavior may
be altered by changing the settings of three variables:
@code{vm-visible-headers}, @code{vm-invisible-header-regexp} and
@code{vm-preview-lines}.

@vindex vm-preview-lines
If the value of @code{vm-preview-lines} is a number, it tells VM how
many lines of the text of the message should be visible.  The default
value of this variable is 0.  If @code{vm-preview-lines} is @code{nil},
then previewing is not done at all; when a message is first presented it
is immediately exposed in its entirety and is flagged as read.  If
@code{vm-preview-lines} is @code{t}, the message body is displayed fully
but the message is not flagged as read until you type @key{SPC}.

@vindex vm-visible-headers
The value of @code{vm-visible-headers} should be a list of regular
expressions matching the beginnings of headers that should be made
visible when a message is presented.  The regexps should be listed in
the preferred presentation order of the headers they match.

@vindex vm-invisible-header-regexp
If non-@code{nil}, the variable @code{vm-invisible-header-regexp}
specifies what headers should @emph{not} be displayed.  Its value should
be a string containing a regular expression that matches all headers you
do not want to see.  Setting this variable non-@code{nil} implies that
you want to see all headers not matched by it; therefore the value of
@code{vm-visible-headers} is only used to determine the order of the
visible headers in this case.  Headers not matched by
@code{vm-invisible-header-regexp} or @code{vm-visible-headers} are
displayed last.

If you change the value of either @code{vm-visible-headers} or
@code{vm-invisible-header-regexp} in the middle of a VM session the
effects will not be immediate.  You will need to use the command
@code{vm-discard-cached-data} on each message (bound to @kbd{j} by
default) to force VM to rearrange the message headers.  A good way to do
this is to mark all the messages in the folder and apply
@code{vm-discard-cached-data} to the marked messages
@xref{Marking Messages}.

@vindex vm-highlighted-header-regexp
@vindex vm-highlighted-header-face
Another variable of interest is @code{vm-highlighted-header-regexp}.
The value of this variable should be a single regular expression that
matches the beginnings of any header that should be presented in inverse
video when previewing.  For example, a value of
@samp{"^From\\|^Subject"} causes the From and Subject headers to be
highlighted.  Highlighted headers will be displayed using the face
specified by @code{vm-highlighted-header-face}, which defaults to
'bold.

@vindex vm-preview-read-messages
By default, VM will not preview messages that are flagged as read.  To
have VM preview all messages, set the value of
@code{vm-preview-read-messages} to @code{t}.

@findex vm-expose-hidden-headers
Typing @kbd{t} (@code{vm-expose-hidden-headers}) makes VM toggle
between exposing and hiding headers that would ordinarily be hidden.

@node Viewing, MIME Messages, Previewing, Reading Messages
@section Viewing

@vindex vm-auto-next-message
Typing @key{SPC} during a message preview exposes the body of the
message.  If the message was new or previously unread, it will be
flagged ``read''.  At this point you can use @key{SPC} to scroll
forward, and @kbd{b} or @key{DEL} to scroll backward a windowful of
text at a time.  A prefix argument @var{n} applied to these commands
causes VM to scroll forward or backward @var{n} lines.  Typing space
at the end of a message moves you to the next message.  If the value
of @code{vm-auto-next-message} is @code{nil}, @key{SPC} will not
move to the next message; you must type @kbd{n} explicitly.

If the value of @code{vm-honor-page-delimiters} is non-@code{nil}, VM
will recognize and honor page delimiters.  This means that when you
scroll through a document, VM will display text only up to the next page
delimiter.  Text after the delimiter will be hidden until you type
another @key{SPC}, at which point the text preceding the delimiter will
become hidden.  The Emacs variable @code{page-delimiter} determines what
VM will consider to be a page delimiter.

You can ``unread'' a message (so to speak) by typing @kbd{U}
(@code{vm-unread-message}).  The current message will be flagged
unread.

@cindex longlines.el
@cindex filling paragraphs
@cindex visual line mode
@vindex vm-paragraph-fill-column
@vindex vm-fill-paragraphs-containing-long-lines
@vindex vm-word-wrap-paragraphs
Sometimes you will receive messages that contain lines that are
too long to fit on your screen without wrapping.  Setting
@code{vm-word-wrap-paragraphs} to t will cause VM to use the 
@code{longlines.el} library by Grossjohann, Schroeder and Yidong to
carry out word wrapping.  You must have this library installed
somewhere on your @code{load-path}.  Another way to deal with the
problem is to use the @code{visual-line-mode} in Emacs 23.  You can
activate it automatically for viewing messages by adding the function
@code{turn-on-visual-line-mode} to the
@code{vm-presentation-mode-hook}. 

If you are unable to use either of these solutions, then you can use
Emacs's paragraph filling facility.  If you set
@code{vm-fill-paragraphs-containing-long-lines} to a positive numeric
value @var{N}, VM will call @code{fill-paragraph} on all paragraphs that
contain lines spanning @var{N} columns or more.  You can also set this
variable to the symbol @code{window-width}, in which case the width of
the current window is used the limiting width beyond which paragraph
filling is invoked.  As with other things that VM does for presentation
purposes, this does not change the message contents.  VM copies the
message contents to a ``presentation'' buffer before altering them.  The
fill column that VM uses is controlled by
@code{vm-paragraph-fill-column}.  Unlike the Emacs variable
@code{fill-column}, this variable is not buffer-local by default.


@node MIME Messages,, Viewing, Reading Messages
@section Reading MIME Messages

@cindex MIME
@vindex vm-display-using-mime
@dfn{MIME} is a set of extensions to the standard Internet message
format that allows reliable transmission of arbitrary data including
images, audio and video, as well as ordinary text in different
languages.  By default, VM will recognize MIME encoded messages and
display them as specified by the various MIME standards specifications.
This can be turned off by setting the variable
@code{vm-display-using-mime} to @code{nil} and VM will then display MIME
messages as plain text messages.

@vindex vm-mime-base64-decoder-program
@vindex vm-mime-base64-encoder-program
@vindex vm-mime-base64-decoder-switches
@vindex vm-mime-base64-encoder-switches
@vindex vm-mime-qp-decoder-program
@vindex vm-mime-qp-decoder-switches
@vindex vm-mime-qp-encoder-program
@vindex vm-mime-qp-encoder-switches
@vindex vm-mime-uuencode-decoder-program
@vindex vm-mime-uuencode-decoder-switches
At its most basic MIME is a set of transfer encodings used to ensure
error free transport, and a set of content types.  VM understands the
two standard MIME transport encodings, Quoted-Printable and BASE64, and
will decode messages that use them as necessary.  VM also will
try to recognize and decode messages using the UNIX ``uuencode''
encoding system.  While this is not an official MIME transfer
encoding and never will be, enough old mailers still use it
that it is worthwhile to attempt to decode it.
VM has Emacs-Lisp based Quoted-Printable and BASE64 encoders and
decoders, but you can have VM use external programs to perform
these tasks and the process will almost certainly be faster.
The variables @code{vm-mime-qp-decoder-program},
@code{vm-mime-qp-decoder-switches},
@code{vm-mime-qp-encoder-program},
@code{vm-mime-qp-encoder-switches},
@code{vm-mime-base64-decoder-switches},
@code{vm-mime-base64-encoder-switches},
@code{vm-mime-base64-decoder-program},
@code{vm-mime-base64-encoder-program},
tell VM which programs to use
and what command line switches to pass to them.  There are C
programs at VM's distribution sites on the Internet to handle BASE64
and Quoted-Printable.  VM does not have a built-in ``uuencode''
decoder, so @code{vm-mime-uuencode-decoder-program} must be set
non-@code{nil} for VM to decode uuencoded MIME objects.

@menu
* Viewing MIME::                Decoding MIME for viewing
* Attachments::                 Operating on MIME attachments
* Internal display::            Viewing attachments internally in Emacs
* External display::            Viewing attachments with external viewers
* Displaying images::           Using Emacs facilities for images
* MIME type conversion::        Converting external attachments to internal
* Character sets::              MIME character sets
* multipart/alternative::       MIME content in alternative formats
* Bad MIME::                    Dealing with bad MIME
@end menu


@node Viewing MIME, Attachments,, MIME Messages
@unnumberedsubsec Viewing MIME messages
By default VM will display as many content types as possible
within Emacs.  Images and audio are also supported if
support for images and audio has been compiled in.  Types that
cannot be displayed internally within Emacs can be converted to a
type that can, or be displayed using an external viewer.

@vindex vm-auto-decode-mime-messages
@vindex vm-mime-decode-for-preview
The first step in displaying a MIME message is decoding it to
determine what object types it contains.  The variable
@code{vm-auto-decode-mime-messages} controls when this happens.
A value of @code{t} means VM should decode the message as soon as
the message body is exposed, or during previewing if
@code{vm-mime-decode-for-preview} is also set non-@code{nil}.  A
@code{nil} value means wait until decoding is explicitly
requested.  Type @kbd{D} (@code{vm-decode-mime-message}) to
manually initiate MIME decoding.

@vindex vm-mime-button-format-alist
When VM does not display a MIME object immediately, it displays a
button or tag line in its place that describes the object and what you
have to do to display it.  The value of @code{vm-mime-button-format-alist}
determines the format of the text in those buttons.

@node Attachments, Internal display, Viewing MIME, MIME Messages
@unnumberedsubsec  Operating on MIME attachments

@cindex attachments
@kindex $ |
@kindex $ d
@kindex $ RET
@kindex $ s
@kindex $ w
@kindex $ p
@kindex $ d
@kindex $ e  
@vindex vm-auto-displayed-mime-content-types
@findex vm-mime-reader-map-pipe-to-command
@findex vm-delete-mime-object
@findex vm-mime-reader-map-display-using-default
@findex vm-mime-reader-map-display-object-as-type
@findex vm-mime-reader-map-save-message
@findex vm-mime-reader-map-save-file
@findex vm-mime-reader-map-pipe-to-printer
@findex vm-delete-mime-object
@findex vm-mime-reader-map-display-using-external-viewer
@findex vm-mime-attach-object-from-message
After decoding you will see either the decoded MIME objects or
button lines that must be activated to attempt display of the
MIME object.  

The variable @code{vm-auto-displayed-mime-content-types} specifies the
types that are displayed immediately.  Its value should be a list of
MIME content types that should be displayed immediately after decoding.
Other types will be displayed as a button that you must activate to
display the object.  

To activate a button, either click the middle mouse
button over it, or move the cursor to the line and press RET.  If you
are running under a window system, you can use the right mouse button
over a MIME button to display a menu of actions you can take on the MIME
object.  If you prefer using keyboard commands, you can save the MIME
object with @kbd{$ w}, print it with @kbd{$ p}, or pipe it to a shell
command with @kbd{$ |}.  Use @kbd{$ s} to append an encapsulated message
or USENET news article to a folder.  If you want to display the object
with its characters displayed using Emacs' default face, use @kbd{$
RET}.  To display the object using an external viewer, type @kbd{$ e}.

@multitable @columnfractions .20 .80
@item $ w   @tab @code{vm-mime-reader-map-save-file}
@item $ s   @tab @code{vm-mime-reader-map-save-message}
@item $ p   @tab @code{vm-mime-reader-map-pipe-to-printer}
@item $ |   @tab @code{vm-mime-reader-map-pipe-to-command}
@item $ RET @tab @code{vm-mime-reader-map-display-using-default}
@item $ e   @tab @code{vm-mime-reader-map-display-using-external-viewer}
@item $ v   @tab @code{vm-mime-reader-map-display-object-as-type}
@item $ d   @tab @code{vm-delete-mime-object}
@item $ a   @tab @code{vm-mime-attach-object-from-message}
@end multitable

@vindex vm-mime-delete-after-saving
@vindex vm-mime-attachment-save-directory
@vindex vm-mime-confirm-delete
The MIME attachments can be saved to disk with @kbd{$ w}
(@code{vm-mime-reader-map-save-file}).  They can be deleted at the
same time by setting the variable @code{vm-mime-delete-after-saving}.
In this case, the attachment is deleted and replaced by a MIME part
that refers to the saved copy.  The variable
@code{vm-mime-attachment-save-directory} specifies the default
directory to save the attachments in.  The MIME attachments can also
be deleted directly from the message bodies with @kbd{$ d}
(@code{vm-delete-mime-object}).  The variable
@code{vm-mime-confirm-delete} controls whether a confirmation is asked
for.  

It is a good idea to use @code{vm-mime-delete-after-saving} to delete
saved attachments instead of deleting them manually, because with the
former approach the message will have a handle to the saved copy,
which can be retrieved when desired.

Saving attachments to the file system and deleting them from message
bodies has the beneficial effect of reducing the size of VM folders.
That leads to a better utilization of the computer resources and
usually a faster operation of VM.

@findex vm-mime-save-all-attachments
@findex vm-mime-delete-all-attachments
@vindex vm-mime-deletable-types
@vindex vm-mime-deletable-type-exceptions
@vindex vm-mime-savable-types
@vindex vm-mime-savable-type-exceptions
The commands @code{vm-mime-save-all-attachments} and
@code{vm-mime-delete-all-attachments} can be used to save or delete
@i{all} the attachments in a message.  An "attachment" in this context
is any MIME part that has "attachment" as its content-disposition or
simply has a file name.  In addition, all MIME parts that have types
matching @code{vm-mime-savable-types} or @code{vm-mime-deletable-types}
(but not the corresponding @code{-exceptions}) are included.

@node Internal display, External display, Attachments, MIME Messages
@unnumberedsubsec Internal display of MIME attachments

@vindex vm-auto-displayed-mime-content-types
A value of t for @code{vm-auto-displayed-mime-content-types} means that
all types should be displayed immediately.  A nil value means
never display MIME objects immediately; only use buttons.  If
the value of @code{vm-auto-displayed-mime-content-types} is a list, it
should be a list of strings, which should all be MIME types or
type/subtype pairs.  Example:

@example
(setq vm-auto-displayed-mime-content-types '("text" "image/jpeg"))
@end example

@noindent If a top-level type is listed without a subtype, all
subtypes of that type are assumed to be included.  The example above
specifies that all text types are displayed immediately, but only JPEG
images are displayed this way.

@vindex vm-auto-displayed-mime-content-type-exceptions
The variable @code{vm-auto-displayed-mime-content-type-exceptions}
should be a list of MIME content types that should not be
displayed immediately after decoding.  This variable acts as
an exception list for @code{vm-auto-displayed-mime-content-types};
all types listed there will be auto-displayed except those in
the exception list.

The value of @code{vm-auto-displayed-mime-content-type-exceptions}
should be either nil or a list of strings.  The strings should
all be types or type/subtype pairs.  Example:

@example
(setq vm-auto-displayed-mime-content-type-exceptions '("text/html"))
@end example

@noindent Again, if a top-level type is listed without a subtype, all
subtypes of that type are assumed to be included.

@vindex vm-mime-internal-content-types
The variable @code{vm-mime-internal-content-types} specifies
which types should be displayed internally within Emacs.  Like
@code{vm-auto-displayed-mime-content-types} its value should be a
list of MIME content types.  A value of t means that VM
should always display an object internally if possible.  VM
knows which object types can be displayed internally, so you
can specify the types you want without worrying about errors
if Emacs can't handle them.  A @code{nil} value means never
display MIME objects internally, which means VM will have to
run an external viewer to display all MIME objects.

If the value is a list, it should be a list of strings.  Example:

@example
(setq vm-mime-internal-content-types '("text" "message" "image/jpeg"))
@end example

@noindent If a top-level type is listed without a subtype, all
subtypes of that type are assumed to be included.  Note that multipart
types are always handled internally regardless of the setting of this
variable.

@vindex vm-mime-internal-content-type-exceptions
The variable @code{vm-mime-internal-content-type-exceptions} serves as
the exception list for @code{vm-mime-internal-content-types}.  Its value 
should be a list of types that should not be displayed internally.

@cindex HTML
@vindex vm-mime-text/html-handler
The HTML content in text/html MIME parts can be displayed in Emacs using
a variety of packages.  VM knows about:

@multitable @columnfractions .15 .85
@item lynx   
  @tab  The lynx browser used externally to convert HTML to plain text
@item w3m    
  @tab  The w3m browser used externally to convert HTML to plain text
@item emacs-w3 
  @tab  The Emacs/W3 browser used internally in Emacs
@item emacs-w3m 
  @tab  The Emacs/W3M browser used internally in Emacs
@end multitable

You can set the variable @code{vm-mime-text/html-handler} to one of
these values to use the appropriate package.  A value of
@code{auto-select} causes VM to select the best package available.  A
value of @code{nil} asks VM not to display HTML content internally.


@node External display, Displaying images, Internal display, MIME Messages
@unnumberedsubsec External display of MIME attachments

@vindex vm-mime-external-content-types-alist
For types that you want displayed externally, set the value
of @code{vm-mime-external-content-types-alist} to specify external
viewers for the types.  The value of this variable should be an
associative list of MIME content types and the external programs
used to display them.  If VM cannot display a type internally or
a type is not listed in @code{vm-mime-internal-content-types} VM will
try to launch an external program to display that type.

The alist format is a list of lists, each sublist having the form

@example
(@var{TYPE} @var{PROGRAM} @var{ARG} @var{ARG} ... )
@end example

@noindent or

@example
(@var{TYPE} @var{COMMAND-LINE})
@end example

@noindent @var{TYPE} is a string specifying a MIME type or
type/subtype pair.  For example ``text'' or ``image/jpeg''.  If a
top-level type is listed without a subtype, all subtypes of that type
are assumed to be included.

In the first form, @var{PROGRAM} is a string naming a program to
run to display an object.  Any @var{ARG}s will be passed to the
program as arguments.  The octets that compose the object will be
written into a temporary file and the name of the file can be
inserted into an @var{ARG} string by writing @samp{%f} in the
@var{ARG} string.  In earlier versions of VM the filename was
always added as the last argument; as of VM 6.49 this is only done
if @samp{%f} does not appear in any of the @var{ARG} strings.

If the @var{COMMAND-LINE} form is used, the program and its
arguments are specified as a single string and that string is
passed to the shell ("sh -c" typically) for execution.  Since
the command line will be passed to the shell, you can use shell
variables and input/output redirection if needed.  As with the
@var{PROGRAM/ARGS} form, the name of the temporary file that
contains the MIME object will be appended to the command line if
@samp{%f} does not appear in the command line string.

In either the @var{PROGRAM/ARG} or @var{COMMAND-LINE} forms, all the
program and argument strings will have any %-specifiers in
them expanded as described in the documentation for the
variable @code{vm-mime-button-format-alist}.  The only difference
is that @samp{%f} refers to the temporary file VM creates to store
the object to be displayed, not the filename that the sender
may have associated with the attachment.

Example:

@example
(setq vm-mime-external-content-types-alist
      '(
         ("text/html"   "netscape")
         ("image/gif"   "xv")
         ("image/jpeg"  "xv")
         ("video/mpeg"  "mpeg_play")
         ("video"       "xanim")
       )
)
@end example

The first matching list element will be used.

No multipart message will ever be sent to an external viewer.

External viewer processes are normally killed when you select
a new message in the current folder.  If you want viewer
processes to not be killed, set
@code{vm-mime-delete-viewer-processes} to a @code{nil} value.

Any type that cannot be displayed internally or externally or
converted to a type that can be displayed, will be displayed as a
button that allows you to save the body to a file.

@vindex vm-mime-external-content-type-exceptions
As with the internal type list, there is an exception list that
you can use to specify types that you do not want displayed
externally.  When VM is considering whether it should
automatically launch an external viewer, it will consult the
variable @code{vm-mime-external-content-type-exceptions}.  If the 
type to be displayed is listed, VM will not launch a viewer.
This allows you to setup viewers for types that ordinarily you
would not want VM to display or for types that you normally want
to convert to some other type using @code{vm-mime-type-converter-alist}.
You can still display such a type with an external viewer by using 
@kbd{$ e}.

@vindex vm-mime-attachment-auto-suffix-alist

When a MIME object is displayed using an external viewer VM must
first write the object to a temporary file.  The external viewer
then opens and displays that file.  Some viewers will not open a
file unless the filename ends with some extension that it
recognizes such as @samp{.html} or @samp{.jpg}.  You can use the
variable @code{vm-mime-attachment-auto-suffix-alist} to map MIME
types to extensions that your external viewers will recognize.
The value of this variable should be a list of type and suffix
pairs.  The list format is:

@example
((@var{TYPE} . @var{SUFFIX}) ...)
@end example

@var{TYPE} is a string specifying a MIME top-level type or a type/subtype pair.
If a top-level type is listed without a subtype, all subtypes of
that type are matched.

@var{SUFFIX} is a string specifying the suffix that should be used for
the accompanying type.

Example:

@example
(setq vm-mime-attachment-auto-suffix-alist
      '(
        ("image/jpeg"           .       ".jpg")
        ("image/gif"            .       ".gif")
        ("image/png"            .       ".png")
        ("text"                 .       ".txt")
       )
)
@end example

@noindent VM will search the list for a matching type.  The suffix 
associated with the first type that matches will be used for the
temporary filename.

@node Displaying images, MIME type conversion, External display, MIME Messages
@unnumberedsubsec Displaying inline images in messages

@cindex images
Most versions of Emacs can display images when used on graphical
screens.  You can verify if the Emacs version is able to do so by
calling the function @code{display-images-p}.  However, Emacs relies
on external libraries to create graphical images, which are specified
through the variable @code{image-library-alist}.  Even if Emacs has
the ability to display some image type, it cannot display such images
unless appropriate libraries are installed and specified to Emacs.  You
can verify which image types are really available by calling the
function @code{image-type-available-p} with an image type such as
@samp{tiff} or @samp{gif} as the argument.

@vindex vm-mime-internal-content-types
@vindex vm-auto-displayed-mime-content-types
Assuming that a particular image type, say @samp{tiff} is available,
you can include its MIME type in
@code{vm-mime-internal-content-types}, e.g., 
@example
(add-to-list 'vm-mime-internal-content-types "image/tiff")
@end example
You can also add the MIME type to the variable
@code{vm-auto-displayed-mime-content-types} so that VM will
automatically display all images of the type.  
If the type is not included among the auto-displayed types, then the
image is initially shown as a button with a thumbnail image.  Clicking on the
button with the middle mouse button expands the image to its full size.

@cindex ImageMagick
@vindex vm-imagemagick-identify-program
@vindex vm-imagemagick-convert-program
Once an image is displayed, you can use the right mouse button to do
various image manipulations on it, such as enlarging/reducing it,
rotating it etc.  To do such operations, VM uses the ImageMagick
graphics manipulation software.  You can install ImageMagick on your
system and specify the location of its @code{identify} and
@code{convert} programs to VM via the variables
@code{vm-imagemagick-identify-program} and
@code{vm-imagemagick-convert-program}.

@vindex vm-mime-use-image-strips
By default, VM displays images by slicing them into contiguous
horizontal strips and displaying the strips in order.  This
facilitates vertical scrolling within an image.  The variable
@code{vm-mime-use-image-strips} controls whether VM uses strips for
image display.  It is @samp{t} by default.

VM also uses the ImageMagick's @code{convert} program to convert
between image formats, so that an image that is not displayable in
Emacs is converted to another format that is displayable.  You can
turn off such conversion by setting the variable
@code{vm-imagemagick-convert-program} to @samp{nil}.


@node MIME type conversion, Character sets, Displaying images, MIME Messages
@unnumberedsubsec MIME type conversion

@vindex vm-mime-type-converter-alist
Types that cannot be displayed internally or externally are
checked against an associative list of types that can be converted to other
types.  If an object can be converted to a type that VM can
display, then the conversion is done and the new object is
subject to the auto-display rules which determine whether the
object is displayed immediately or a button is displayed in its
place.  The conversion list is stored in the variable
@code{vm-mime-type-converter-alist}.

The alist format is

@example
( (START-TYPE END-TYPE COMMAND-LINE ) ... )
@end example

@var{START-TYPE} is a string specifying a MIME type or type/subtype pair.
Example @samp{"text"} or @samp{"image/jpeg"}.  If a top-level type is
listed without a subtype, all subtypes of that type are assumed
to be included.

@var{END-TYPE} must be an exact type/subtype pair.  This is the type
to which @var{START-TYPE} will be converted.

@var{COMMAND-LINE} is a string giving a command line to be passed to
the shell.  The octets that compose the object will be written to
the standard input of the shell command.

Example:

@example
(setq vm-mime-type-converter-alist
       '(
         ("image/jpeg"  "image/gif"     "jpeg2gif")
         ("text/html"   "text/plain"    "striptags")
        )
)
@end example

@noindent The first matching list element will be used.

@node Character sets, multipart/alternative, MIME type conversion, MIME Messages
@unnumberedsubsec MIME character sets

For text type messages, MIME also requires that a character set
be specified, so that the recipient's mail reader knows what
character glyphs to use to display each character code.  To
display a message properly VM needs to know how to choose a font
for a given character set.

@vindex vm-mime-default-face-charsets
The variable @code{vm-mime-default-face-charsets} tells VM what character
sets your default face can display.  For most American and European
users using X Windows, Emacs' default face displays the ISO-8859-1
and US-ASCII characters, US-ASCII being a subset of ISO-8859-1.  The
value of @code{vm-mime-default-face-charsets} must be a list of strings
specifying the character sets that your default face can display.
This variable is useful for making bogus, unregistered character sets
that are slight variants of ISO-8859-1 visible.
Example:

@example
(add-to-list 'vm-mime-default-face-charsets "Windows-1251")
(add-to-list 'vm-mime-default-face-charsets "Windows-1252")
(add-to-list 'vm-mime-default-face-charsets "Windows-1257")
@end example

@noindent Messages sent using such character sets would normally be
considered undisplayable by VM, and a button would be displayed
that offers to save the message body to a file.

Please note that for character sets listed in this variable, VM's MIME
decoding is bypassed.  So don't add charsets like \"utf-8\" that require
additional decoding.

@vindex vm-mime-charset-converter-alist
Sometimes a charset that VM cannot display can be converted to a
one that VM can display.  An example would be a message encoded
using UTF-8 but in fact only contains Japanese characters.  In
that case the message text could probably be converted to
iso-2022-jp which VM running on a MULE-enabled Emacs could
display.

VM offers a way to do such conversions.  The variable
@code{vm-mime-charset-converter-alist} is an associative list of MIME
charsets and programs that can convert between them.  If VM
cannot display a particular character set, it will scan this list
to see if the charset can be converted into a charset that it can
display.

The alist format is:

@example
 ( ( START-CHARSET END-CHARSET COMMAND-LINE ) ... )
@end example

@var{START-CHARSET} is a string specifying a MIME charset.
Example @samp{"iso-8859-1"} or @samp{"utf-8"}.

@var{END-CHARSET} is a string specifying the charset to which
@var{START-CHARSET} will be converted.

@var{COMMAND-LINE} is a string giving a command line to be passed to
the shell.  The characters in @var{START-CHARSET} will be written to the
standard input of the shell command and VM expects characters
encoded in @var{END-CHARSET} to appear at the standard output of the
@var{COMMAND-LINE}.  @var{COMMAND-LINE} is passed to the shell, so you can
use pipelines, shell variables and redirections.

Example:

@example
(setq vm-mime-charset-converter-alist
      '(
         ("utf-8" "iso-2022-jp" "iconv -f utf-8 -t iso-2022-jp")
       )
)
@end example

The first matching list element will be used.

@vindex vm-mime-charset-font-alist
The variable @code{vm-mime-charset-font-alist} tells VM what font to use 
to display a character set that cannot be displayed using
the default face.  The value of this variable should be an
assoc list of character sets and fonts that can be used to display
them.  The format of the list is:
 
( (@var{CHARSET} . @var{FONT}) ...) 
 
@var{CHARSET} is a string naming a MIME registered character set such 
as @samp{"iso-8859-5"}.
 
@var{FONT} is a string naming a font that can be used to display
@var{CHARSET}.
 
An example setup might be: 
 
@example
(setq vm-mime-charset-font-alist
  '(
    ("iso-8859-5" . "-*-*-medium-r-normal-*-16-160-72-72-c-80-iso8859-5")
   )
)
@end example

@noindent This variable is only useful for character sets whose
characters can all be encoded in single 8-bit bytes.  Also multiple
fonts can only be displayed if you're running under a window system
e.g. X Windows.  So this variable will have no effect if you're
running Emacs on a tty.

Note that under FSF Emacs 19 any fonts you use must be the
same height as your default font.  XEmacs and Emacs 21 do not
have this limitation.  Under Emacs 20 and beyond, and under
any XEmacs version compiled with MULE support, the value of
@code{vm-mime-charset-font-alist} has no effect.  This is
because all characters are displayed using fonts discovered by
MULE and VM has no control over them.

@node multipart/alternative, Bad MIME, Character sets, MIME Messages
@unnumberedsubsec MIME multipart/alternative

MIME allows a message to be sent with its content encoded in multiple
formats, simultaneously, in the same message.  Such messages have a
content type of multipart/alternative.  The idea is that the sender
might have different MIME decoding or display capabilities than some
of his recipients.  For instance, the sender may be able to compose a
message using fancy text formatting constructs like tables, italics
and equations but some of the recipients may only be able to display
plain text.  The multipart/alternative type message is the solution
to this dilemma.  Such a message would contain at least two text
subparts, one in plain text and the other in the full featured text
formatting language that the sender used.

@vindex vm-mime-alternative-select-method
To control how VM displays multipart/alternative messages, you must
set the variable @code{vm-mime-alternative-select-method}.  Its value must be
a symbol.  A value of @code{best} tells VM to display the message
using the subpart closest in appearance to what the sender used to
compose the message.  In the example above this would mean displaying
the fully featured text subpart, if VM knows how to display that type.
VM will display the type either internally or externally.  A
value of @code{best-internal} tells VM to use the closest subpart that
it can display internally.  External viewers won't be used in this
case.  A value of @code{all} asks VM to display all the alternatives.

The value can also be a list of the form
@example
  (favorite TYPE ...)
@end example
with the first element of the list being the symbol @code{favorite}.
The remaining elements of the list are strings specifying MIME types.
VM will look for each TYPE in turn in the list of alternatives and
choose the first matching alternative found that can be displayed.  If
instead of the symbol @code{favorite}, @code{favorite-internal} is used
then the first TYPE that matches an alternative that can be displayed
internally will be chosen.

@findex vm-mime-nuke-alternative-text/html
Messages with multiple alternatives use up extra file space and slow
down the operation of vm.  If you would like keep the text/plain
alternatives but erase the text/html alternatives, you can use the
'vm-mime-nuke-alternative-text/html' command.

@node Bad MIME, , multipart/alternative, MIME Messages
@unnumberedsubsec Dealing with bad MIME

@vindex vm-infer-mime-types
Some mailers incorrectly use the generic
@samp{application/octet-stream} type when sending files that
really have a specific MIME type.  For example, a JPEG image
might be sent using @samp{application/octet-stream} type instead
of @samp{image/jpeg}, which would be the correct type.  In many
cases the filename sent along with the mistyped file
(e.g. @file{foo.jpg}) suggests the correct type.  If the variable
@code{vm-infer-mime-types} is set non-@code{nil}, VM will attempt to use 
the filename sent with a MIME attachment to guess an attachment's 
type if the attachment is of type @samp{application/octet-stream}.

@node Sending Messages, Saving Messages, Reading Messages, Top
@chapter Sending Messages

When sending messages from within VM, you will be using the
standard Mail major mode provided with GNU Emacs, plus some
extensions added by VM.  @xref{Mail Mode,,,emacs, the GNU Emacs
Manual}.  However, mail composition buffers created by VM have some
extra command keys.

@table @kbd

@findex vm-yank-message
@findex vm-yank-message-other-folder
@kindex C-c C-y
@item C-c C-y (@code{vm-yank-message})
Copies a message from the folder that is the parent of this
composition into the mail composition buffer.
The message number is read from the minibuffer.  By default, each line of
the copy is prepended with the value of the variable
@code{vm-included-text-prefix}.  All message headers are yanked along
with the text.  Point is left before the inserted text, the mark after.
Any hook functions bound to @code{mail-yank-hooks} are run, after inserting
the text and setting point and mark.  If a prefix argument is given,
this tells VM to ignore @code{mail-yank-hooks}, don't set the mark, don't prepend the
value of @code{vm-included-text-prefix} to every yanked line, and don't yank
any headers other than those specified in
@code{vm-visible-headers} and @code{vm-invisible-headers}.  

@item @code{M-x vm-yank-message-other-folder}
This allows one to yank a message from a different folder than the
parent of this composition.

@kindex C-c C-v
@item C-c C-v <Any VM command key>
All VM commands may be accessed in a VM Mail mode buffer by prefixing them
with C-c C-v.
@kindex C-c C-a
@vindex vm-send-using-mime
@item C-c C-a (@code{vm-mime-attach-file})
Attaches a file to the composition.  When you send the message, VM 
will insert the file and MIME encode it.  The variable
@code{vm-send-using-mime} must be set non-@code{nil} for this command to work.
You will be asked for the file's type, and a brief description of 
the attachment.  The description is optional.  If the file's type
is a text type, you will also be asked for the character set
in which the text should be displayed.
The new attachment will appear as a highlighted tag in the
composition buffer.  You can use mouse button 3 on this tag
to set the default content disposition of the attachment.  The
content disposition gives a hint to the recipient's mailer how to 
treat the attachment.  Specifically the disposition will indicate 
whether the attachment should be displayed along with the message 
or saved to a file.  Any text in the composition that appears
before the tag will appear in a MIME text part before the
attachment when the message is encoded and sent.  Similarly, any
text after the tag will appear after the attachment in the
encoded message.  If you change your mind about using the
attachment, you can remove it from the composition with @key{C-k}.
If you want to move the attachment to some other part of the message,
you can kill it @key{C-k} and yank it back with @key{C-y}.

@kindex C-c C-m
@item C-c C-m (@code{vm-mime-attach-message})
Attaches a mail message to the composition.  If invoked with a
prefix argument, the name of a folder read from the minibuffer and
the message or messages to be attached are copied from that
folder.  You will be prompted for the message number of the
message to be attached.  If you invoke the command on marked
messages by running
@code{vm-next-command-uses-marks} first, the marked messages in
the selected folder will be attached as a MIME digest.
@kindex C-c C-b
@item C-c C-b (@code{vm-mime-attach-buffer})
Attaches an Emacs buffer to the composition.

@findex vm-mime-encode-composition
@kindex C-c C-e
@item C-c C-e (@code{vm-mime-encode-composition})
Encodes the composition using MIME, but does not send it.  This
is useful if you want to use PGP to sign a message before sending
it.  After signing the message, you would use C-c C-c as usual to
send the message.  Emacs' @code{undo} command can be used to undo
the encoding, so that you can continue composing the unencoded
message.

@findex vm-preview-composition
@kindex C-c C-p
@item C-c C-p (@code{vm-preview-composition})
Previews the current composition.  The message is copied into a
temporary folder and you can read the message and interact with
it using normal VM mode commands to see how it might look to a
recipient.  Type @key{q} to quit the temporary folder and resume
composing your message.

@findex vm-postpone-composition
@kindex C-c C-d
@item C-c C-d (@code{vm-postpone-composition})
Postpones the current composition.  The message is stored in the
folder specified in @code{vm-postponed-folder}.  You can continue composing
of messages by visiting @code{vm-postponed-folder}, selecting a massage
and @key{m} or by directly calling the function on any message in any folder
@kbd{M-x vm-continue-postponed-message}.  When called with a prefix argument
@code{vm-postpone-composition} will ask you for the folder to save the draft
to, i.e. you might also save it to your inbox.

@end table

@findex vm-mail
@kindex m
The simplest command is @kbd{m} (@code{vm-mail}) which sends a mail
message much as @kbd{M-x mail} does but allows the added commands
described above.

@code{vm-mail} can be invoked outside of VM by typing @kbd{M-x vm-mail}.
However, only (@code{vm-yank-message-other-folder}) will work; all the
other commands require a parent folder.

If you send a message and it is returned by the mail system
because it was undeliverable, you can resend the message by
typing @kbd{M-r} (@code{vm-resend-bounced-message}).  VM will
extract the old message and its pertinent headers from the
returned message, and place you in a VM Mail mode buffer.  A
Resent-To header will be added, which you can fill in with
the corrected addresses of the recipients that bounced.  You
can also added a Resent-Cc header, which has the same meaning
as a Cc header in a normal message.  Mail will only be sent to
the addresses in the Resent-To and Resent-Cc headers unless
you delete both of those headers.  In that case the To and Cc
headers will be used.

@menu
* Sending Options::     Variables that control mail sending.
* MIME Composition::    Sending a message using MIME attachments.
* Replying::            Describes the various ways to reply to a message.
* Forwarding Messages:: How to forward a message to a third party.
@end menu

@node Sending Options, MIME Composition, Sending Messages, Sending Messages
@section Mail Sending Options

As already mentioned, VM uses Emacs @ref{Mail Mode,,,emacs, the Gnu
Emacs Manual} for sending email.  Therefore, Mail Mode options
should be set to configure the mail sending.  The extra options
provided by VM are described below.

@vindex vm-mail-header-from
@vindex vm-mail-mode-hidden-headers
When a mail composition buffer is created, VM initializes it with
header lines that you can fill in.  The @code{From} header is usually
standard and contains your email address.  You can have VM fill it in
for you
automatically by setting the variable @code{vm-mail-header-from}.  (It is
@code{nil} by default.)  The variable
@code{vm-mail-mode-hidden-headers} can be used to hide some of the
header lines from the mail composition buffer.  By default, the
headers ``References'', ``In-Reply-To'' and ``X-Mailer'' are hidden. 

@vindex vm-mail-header-insert-date
@vindex vm-mail-header-insert-message-id
@vindex vm-mail-reorder-message-headers
@vindex vm-mail-header-order
Additional header lines are created by VM before the composed message
is sent.  The variable @code{vm-mail-header-insert-date} can be set to
@code{t} (which is the default value) asking VM to insert a Date
header into a message before it is sent.  You should set it to
@code{nil} if you would like to insert a Date header yourself.  The
variable @code{vm-mail-header-insert-mesasge-id} asks VM to insert a
Message-ID header before sending the message. The variable
@code{vm-mail-reorder-message-headers} asks VM to reorder the message
headers into a particular order before sending.  The order is
determined by the variable @code{vm-mail-header-order}.


@node MIME Composition, Replying, Sending Options, Sending Messages
@section MIME Composition

@vindex vm-send-using-mime
To use VM's MIME composition features, you must have
@code{vm-send-using-mime} set to a non-@code{nil} value.  With MIME composition
enabled, VM will allow you to add file attachments to your
composition and will analyze your message when you send it and
MIME encode it as necessary.

@menu
* MIME attachments::    Sending a message using MIME attachments.
* MIME characters::     Sending a message with MIME-encoded characters.
* MIME headers::        Sending a message with MIME-encoded headers.
* MIME preview::        Previewing a MIME message before sending.
@end menu

@node MIME attachments, MIME characters, MIME Composition, MIME Composition, 
@unnumberedsec MIME attachments

@kindex C-c C-a
To attach a file to your composition, use @kbd{C-c C-a}
(@code{vm-mime-attach-file}).  VM will ask you for the name of the file, 
its type, a brief description and its character set if it is a
text attachment.  The attachment will be represented in the
composition as a tag line like this

  [ATTACHMENT ~/sounds/chronophasia_scream.au, audio/basic]

You can type text before and after this tag and it will appear
before or after the text in the final MIME message when VM encodes
it.  You can kill the tag with @kbd{C-k} and yank it back with
@kbd{C-y} to move it to another place in the message.  You can
yank back the tag multiple times to duplicate the attachment in
the message.  Or you can leave the tag killed and the attachment
won't appear in the message when it is sent.

If you click the right mouse button on the attachment tag, a menu
will appear that allows you to change the content disposition of
the attachment.  The @dfn{content disposition} of a MIME object
gives a mail reader a hint as to whether an object should be
displayed inline or as an inert tag or button that you must
activate in some fashion.  @dfn{Inline} display usually means
that the object will be display within or alongside the message
text, if that is possible.  @dfn{Attachment}, when used as a
content disposition, means that the object will likely be
displayed as a tag.  By default, VM specifies an inline
disposition for all MIME types except @samp{application} and
@samp{model} types.

@kindex C-c C-b
To attach a buffer instead of a file, use @kbd{C-c C-b} (normally 
bound to @code{vm-mime-attach-buffer}.  You must not kill the
buffer that you attach until after the message has been sent.

@node MIME characters, MIME headers, MIME attachments, MIME Composition
@unnumberedsec MIME characters

@vindex vm-mime-7bit-composition-charset
By default, when you type text into a composition buffer VM
assumes that if all the character codes are less than 128, you
are using the US-ASCII character set and that is the character
set declared in the encoding of the message when it is sent.  If
you are using some other character set, you must specify it by
setting the variable @code{vm-mime-7bit-composition-charset}.  The
value of this variable should be a string specifying the character
set.

@vindex vm-mime-8bit-composition-charset
If there are character codes in the composition greater than 128, the
variable @code{vm-mime-8bit-composition-charset} tells VM what character 
set to assume when encoding the message.  The default is
@samp{iso-8859-1}.

Character codes greater than 128 may not be transported reliably
across the Internet in mail messages.  Some machines will refuse
to accept messages containing such characters and some will accept
them but zero the eighth bit, garbling the message.  To avoid
these problems, VM transfer encodes 8-bit text by default.

MIME has two transfer encodings that convert 8-bit data to 7-bit data
for safe transport.  @dfn{Quoted-printable} leaves the text mostly
readable even if the recipient does not have a MIME-capable mail
reader.  @dfn{BASE64} is unreadable without a MIME-capable mail
reader.

@vindex vm-mime-8bit-text-transfer-encoding
VM's text transfer encoding behavior is controlled by the
variable @code{vm-mime-8bit-text-transfer-encoding}.  Its value should
be a symbol that specifies what kind of transfer encoding to do
for 8-bit text.  A value of @samp{quoted-printable}, means to use
quoted-printable encoding.  A value of @samp{base64} means to use
BASE64 encoding.  A value of @samp{8bit} means to send the message as
is.  Note that this variable usually only applies to textual MIME
content types.  Images, audio, video, etc. typically will have
some attribute that makes VM consider them to be ``binary'',
which moves them outside the scope of this variable.  For
example, messages with line lengths of 1000 characters or more
are considered binary, as are messages that contain carriage
returns (ASCII code 13) or NULs (ASCII code 0).

@node MIME headers, MIME preview, MIME characters, MIME Composition
@unnumberedsec MIME headers

The internet standards specify that the header lines of messages should
always be in 7 bit ASCII, even if the body of a message can use an
8 bit character set.  If you use other non-ASCII characters in typing
the headers then VM encodes their words using the MIME encoded-word
syntax, which is of the form @code{=?charset?encoding?encoded text?=}.  

@vindex vm-mime-encode-headers-regexp
@vindex vm-mime-encode-headers-type
The variable @code{vm-mime-encode-headers-regexp} specifies which
headers should be encoded in this way.  By default, @samp{Subject},
@samp{Organization}, @samp{From}, @samp{To}, @samp{CC}, @samp{Bcc} and
@samp{Resent-} header lines encoded.  The words are encoded using
quoted-printable encoding (@code{Q}).  You can ask VM to use the base64
encoding by setting the variable @code{vm-mime-encode-headers-type}.

@vindex vm-mime-encode-words.regexp
@vindex vm-mime-encode-headers-words-regexp
The variables @code{vm-mime-encode-words.regexp} and
@code{vm-mime-encode-headers-words-regexp} control what is 
meant by a ``word'' for VM for the purpose of encoding.  By default, the
words are those containing any 8 bit character and delimited by white
space characters.

@node MIME preview, , MIME headers, MIME Composition
@unnumberedsec MIME preview

@kindex C-c C-p
To preview what a MIME message will look like to a recipient,
use @kbd{C-c C-p} (@code{vm-mime-preview-composition}).  VM
will encode a copy of the message and present it to you in a
temporary mail folder.  You can scroll through the message
using normal VM mail reading commands.  Typing @kbd{q} in this
folder will return you to your composition where you can make
further changes.

@kindex C-c C-e
To encode a MIME message without sending it, use @kbd{C-c C-e}
(@code{vm-mime-encode-composition}).  This is useful if you use
PGP and want to sign a message before sending it.  VM will encode
the message for transport, inserting all necessary headers and
boundary markers.  You can then sign the message and send it with
C-c C-c and be confident that VM won't invalidate the signature
by making further modifications to the message.  Or if you want
to resume editing the message you can run the Emacs @code{undo}
(normally bound to @kbd{C-x u}) command which will revert the
encoded MIME bodies back to tags and you can continue entering
your composition.


@node Replying, Forwarding Messages, MIME Composition, Sending Messages
@section Replying

@vindex vm-reply-subject-prefix
VM has special commands that make it easy to reply to a message.  When a
reply command is invoked, VM fills in the subject and recipient headers
for you, since it is apparent whom the message should be sent to and
what the subject should be.  There is an old convention of prepending
the string @samp{Re: } to the subject of replies if the string isn't
present already.  You can customize the string to be prepended in this
way by setting the variable @code{vm-reply-subject-prefix}.  Its value
should be a string to prepend to the subject of replies, if the string
isn't present already.  A @code{nil} value means don't prepend anything
to the subject (this is the default).  In any case you can edit any of
the message headers manually, if you wish.

@vindex vm-included-text-prefix
VM also helps you cite material from the message to which you are
replying, by providing @dfn{included text} as a feature of some of the
commands.  @dfn{Included text} is a copy of the message being replied to
with some prefix to each line so that the included text
can be distinguished from the text of the reply.  By default, the
prefix string is @samp{> }.  This can be customized via the variable
@code{vm-included-text-prefix}.

The reply commands are:

@table @kbd
@findex vm-reply
@kindex r
@item r (@code{vm-reply})
Replies to the author of the current message.
@findex vm-reply-include-text
@kindex R
@item R (@code{vm-reply-include-text})
Replies to the author of the current message and provides included text.
@findex vm-followup
@kindex f
@item f (@code{vm-followup})
Replies to the all recipients of the current message.
@findex vm-followup-include-text
@kindex F
@item F (@code{vm-followup-include-text})
Replies to the all recipients of the current message and provides
included text.
@end table

These commands all accept a numeric prefix argument @var{n}, which if
present, causes VM to reply to the next (or previous if the argument is
negative) @var{n-1} messages as well as the current message.  Also, all
the reply commands set the ``replied'' attribute of the messages to
which you are responding, but only when the reply is actually sent.  The
reply commands can also be applied to marked messages. (@pxref{Marking
Messages}.) 

@vindex vm-reply-ignored-addresses
If you are one of multiple recipients of a message and you use @kbd{f}
and @kbd{F}, your address will be included in the recipients of the
reply.  You can avoid this by judicious use of the variable
@code{vm-reply-ignored-addresses}.  Its value should be a list of
regular expressions that match addresses that VM should automatically
remove from the recipient headers of replies.  The default value is
@code{nil}, which means that no addresses are removed.

@vindex vm-in-reply-to-format
The variable @code{vm-in-reply-to-format} specifies the format of the
In-Reply-To header that is inserted into the header section of the reply
buffer.  Like @code{vm-included-text-attribution-format},
@code{vm-in-reply-to-format} should be a string similar to that of
@code{vm-summary-format}.  A @code{nil} value causes the In-Reply-To
header to be omitted.  If the format includes elements that can contain
non-ASCII characters, then @samp{In-Reply-To} should be added to
@code{vm-mime-encode-headers-regexp}.

@vindex vm-strip-reply-headers
The recipient headers generated for reply messages are created by
copying the appropriate headers from the message to which you are
replying.  This includes any full name information, comments, etc. in
these headers.  If the variable @code{vm-strip-reply-headers} is
non-@code{nil}, the recipient headers will be stripped of all information
except the actual addresses.

@unnumberedsubsec Included text

As mentioned above, the commands @code{vm-reply-include-text} and
@code{vm-followup-include-text} provide ``included text'' from the
original message in your reply.  In addition, you can use @kbd{C-c C-y}
(@code{vm-yank-message}) inside a mail buffer to include text from any
desired mail message.  This is a more general mechanism for citing
message text in the composed message.  (The composed message does not
have to be a reply.  Neither do the cited messages have to be the
messages you are replying to.)

Citing message text is tricky business because the original message
could be a MIME message with encoded text or formatted text along with
embedded images and attachments.  By default, VM uses its MIME
displaying mechanism to extract the included text to be cited in
replies.  The cited text is then identical to what appears in the
message Presentation buffer.

@vindex vm-included-mime-types-list
If the default MIME display is not enough for your purposes, you can
set the variable 
@code{vm-included-mime-types-list} to selectively include the MIME
type/subtype pairs
that should be covered in the included text.  The list should have
@code{text/plain} at a minimum and could include other types such as
@code{text/enriched} and @code{message/rfc822}.  Setting the variable to nil
returns you to the default behaviour.  

@vindex vm-include-text-from-presentation
Another possible way to extract
included text is to use the contents of the Presentation buffer.  You
can ask VM to do so by setting the variable
@code{vm-include-text-from-presentation} to t.

@vindex vm-fill-paragraphs-containing-long-lines-in-reply
@vindex vm-fill-long-lines-in-reply-column
If the included text contains long lines, i.e., lines longer than the
normal window width, you might want to fill paragraphs.  You can invoke
automatic filling of paragraphs by setting the variable
@code{vm-fill-paragraphs-containing-long-lines-in-reply}.  Like its
namesake used in message presentation (@pxref{Viewing}), it should be
set to a positive numerical value N or the symbol @code{window-width}.
Setting it to @code{nil} disables paragraph filling.  If filling is
used, the fill column is controlled by the variable
@code{vm-fill-long-lines-in-reply-column}.

Alternatively, you can fill individual paragraphs manually using
@kbd{C-c C-q} (@code{mail-fill-yanked-message}).

@unnumberedsubsec Options

@vindex vm-included-text-attribution-format
The variable @code{vm-included-text-attribution-format} specifies the
format for the attribution of the included text.  The @dfn{attribution} is a
line of text that tells who wrote the text that is to be included; it
will be inserted before the included text.  If non-@code{nil}, the value
of @code{vm-included-text-attribution-format} should be a string format
specification similar to @code{vm-summary-format}.  @xref{Summaries}.  A
@code{nil} value causes the attribution to be omitted.

@vindex vm-included-text-headers
@vindex vm-included-text-discard-header-regexp
VM normally includes only the body text from the cited messages.  If you
wish, you can include also the message headers by customizing
the variables @code{vm-included-text-headers} and
@code{vm-included-text-discard-header-regexp}.

@node Forwarding Messages,, Replying, Sending Messages
@section Forwarding Messages

VM has three commands to forward messages: @kbd{z}
(@code{vm-forward-message}), @key{@@} (@code{vm-send-digest}) and
@kbd{B} (@code{vm-resend-message}).

@findex vm-forward-message
@kindex z
@vindex vm-forwarding-digest-type
@vindex vm-forwarding-subject-format
Typing @kbd{z} puts you into a VM Mail mode buffer just like @kbd{m},
except the current message appears as the body of the message in the VM
Mail mode buffer. The forwarded message encapsulated as specified by the
variable @code{vm-forwarding-digest-type}.  Recognized values are
@samp{"rfc934"}, @samp{"rfc1153"} and @samp{"mime"}.  If the variable
@code{vm-forwarding-subject-format} is non-@code{nil} it should specify
the format of the Subject header of the forwarded message.  A @code{nil}
value causes the Subject header to be left blank.  The forwarded message
is flagged ``forwarded'' when the message is sent.

@findex vm-send-digest
@vindex vm-digest-send-type
@kindex @@
The command @key{@@} (@code{vm-send-digest}) works like @kbd{z} except
that a digest of all the messages in the current folder is made and
inserted into the VM Mail mode buffer.  Also, @code{vm-send-digest} can
be applied to just marked messages.  @xref{Marking Messages}.  When applied
to marked messages, @code{vm-send-digest} will only bundle marked
messages, as opposed to the usual bundling of all messages in the
current folder.  The message encapsulation method is specified by the
variable @code{vm-digest-send-type}, which accepts the same values as
@code{vm-forwarding-digest-type}.  All the messages included in the digest will
be flagged ``forwarded'' when the digest message is sent.

@vindex vm-digest-preamble-format
@vindex vm-digest-center-preamble
If you give @code{vm-send-digest} a prefix argument, VM will insert a
list of preamble lines at the beginning of the digest, one line per
digestified message.  The variable @code{vm-digest-preamble-format}
determines the format of the preamble lines.  If the value of
@code{vm-digest-center-preamble} is non-@code{nil}, the preamble lines
will be centered.

@findex vm-resend-message
@kindex B
If you wish to forward a message and want to send it without the
encapsulation used by @code{vm-forward-message}, use @kbd{B}
(@code{vm-resend-message}).  Instead of encapsulating the message, VM
will use essentially the same message and headers and add a Resent-To
header that you should fill in with the new recipients.  Use @kbd{C-c
C-c} as usual to send the message.  The resent message will be flagged
as ``redistributed''.

@vindex vm-forwarded-headers
@vindex vm-unforwarded-header-regexp

You can control which header lines are included in forwarded messages
via the variables @code{vm-forwarded-headers} and
@code{vm-unforwarded-header-regexp}.  By default, both of these
variables are nil, which means that all headers should be included.
If @code{vm-unforwarded-header-regexp} is set to a regular expression
then all the headers matching the regular expression are omitted in
forwarded messages.  If this variable is nil but
@code{vm-forwarded-headers} is set to a list of headers, then only
these headers are included in the forwarded messages.  See the
documentation of these variables for further information.

@node Saving Messages, Deleting Messages, Sending Messages, Top
@chapter Saving Messages

@cindex file locking
Mail messages are normally saved to files that contain only mail
messages.  Such files are called @dfn{folders}.  Folders are
distinguished from spool files in that VM does not expect other
programs to modify them while VM is visiting them.  This is
important to remember.  VM does no locking of folders when
visiting them.  If the disk copy of a folder is modified behind
VM's back, Emacs will complain with the dreaded ``File changed
on disk'' message when you try to save the folder.

@findex vm-save-message
@kindex s
The VM command to save a message to a folder is @kbd{s}
(@code{vm-save-message}); invoking this command causes the current
message to be saved to a folder whose name you specify in the
minibuffer.  If @code{vm-save-message} is given a prefix argument
@var{n}, the current message plus the next @var{n-1} messages are saved.
If @var{n} is negative, the current message and the previous @var{n-1}
messages are saved.  Messages saved with @code{vm-save-message} are
flagged ``filed''.

@findex vm-save-message-to-imap-folder
One can save messages to remote IMAP folders by
@code{vm-save-message-to-imap-folder}; the name of the folder can be
entered from the minibuffer.

@vindex vm-confirm-new-folders
If the value of the variable @code{vm-confirm-new-folders} is
non-@code{nil}, VM will ask for confirmation before creating a new
folder on interactive saves.

@vindex vm-imap-save-to-server
@cindex IMAP
A non-@code{nil} value of @code{vm-imap-save-to-server} means that
messages from IMAP folders should be saved in other IMAP folders on the
same server.  Otherwise, the messages are saved to local folders. 

@vindex vm-folder-directory
If you have a directory where you keep all your mail folders, you should
set the variable @code{vm-folder-directory} to point to it.  If this
variable is set, @code{vm-save-message} will insert this directory name
into the minibuffer before prompting you for a folder name; this will save
you some typing. The default value of @code{vm-folder-directory} is 
@code{~/Mail}.

@vindex vm-auto-folder-alist
Another aid to selecting folders in which to save mail is the variable
@code{vm-auto-folder-alist}.  The value of this variable should be a
list of the form:

@display
((@var{header-name}
   (@var{regexp} . @var{folder-name}) ...)
  ...)
@end display

where @var{header-name} and @var{regexp} are strings, and
@var{folder-name} is a string or an s-expression that evaluates to a
string.

If any part of the contents of the message header named by
@var{header-name} is matched by the regular expression
@var{regexp}, VM will evaluate the corresponding
@var{folder-name} and use the result as the default when
prompting for a folder to save the message in.  The value of
@var{folder-name} can be the absolute path name of a local folder, a
relative path name -- relative to @code{vm-folder-directory} or the
@code{default-directory} of the currently visited folder -- or the
maildrop specification of an IMAP folder.

When @var{folder-name} is evaluated, the current buffer will contain only
the contents of the header named by @var{header-name}.  It is safe to
modify this buffer.  You can use the match data from any @samp{\( @dots{}
\)} grouping constructs in @var{regexp} along with the function
@code{buffer-substring} to build a folder name based on the header information.
If the result of evaluating @var{folder-name} is a list, then the list will
be treated as another auto-folder-alist and will be descended
recursively.

@vindex vm-auto-folder-case-fold-search
Whether matching is case sensitive depends on the value of the variable
@code{vm-auto-folder-case-fold-search}.  A non-@code{nil} value makes
matching case insensitive.  The default value is @code{t}, which means
matching is case insensitive.  Note that the matching of header names is
always case insensitive because the Internet message standard RFC 822
specifies that header names are case indistinct.

@vindex vm-visit-when-saving
VM can save messages to a folder in two distinct ways.  The message can be
appended directly to the folder on disk, or the folder can be visited as
Emacs would visit any other file and the message appended to that
buffer.  In the latter method you must save the buffer yourself to change
the on-disk copy of the folder.  The variable @code{vm-visit-when-saving}
controls which method is used.  A value of @code{t} causes VM to always
visit a folder before saving message to it.  A @code{nil} value causes VM
to always append directly to the folder file.  In this case VM will not
save messages to the disk copy of a folder that is being visited.  This
restriction is necessary to insure that the buffer and on-disk copies of
the folder are consistent. If the value of @code{vm-visit-when-saving} is
not @code{nil} and not @code{t} (e.g. 0, the default), VM will append to
the folder's buffer if the buffer is currently being visited, otherwise VM
will append to the file itself.

@vindex vm-delete-after-saving
@vindex vm-delete-after-archiving
After a message is saved to a folder, the usual thing to do next is to
delete it.  If the variable @code{vm-delete-after-saving} is
non-@code{nil}, VM will flag messages for deletion automatically after
saving them.  This applies only to saves to folders, not for the @kbd{w}
or @kbd{A} commands (see below).  The variable @code{vm-delete-after-archiving}
works like @code{vm-delete-after-saving} but applies to the @kbd{A}
(@code{vm-auto-archive-messages}) command instead.

Other commands:

@table @kbd
@findex vm-save-message-sans-headers
@kindex w
@item w (@code{vm-save-message-sans-headers})
Saves a message or messages to a file without their headers.  This
command responds to a prefix argument exactly as @code{vm-save-message}
does.  Messages saved this way are flagged ``written''.
@findex vm-auto-archive-messages
@kindex A
@item A (@code{vm-auto-archive-messages})
Save all unfiled messages that auto-match a folder via
@code{vm-auto-folder-alist} to their appropriate folders.  Messages that
are flagged for deletion are not saved by this command.  If invoked with a
prefix argument, confirmation will be requested for each save.
@findex vm-pipe-message-to-command
@kindex ||
@item || (@code{vm-pipe-message-to-command})
Runs a shell command with some or all of the current message as input.
By default, the entire message is used.  However, the leading and
trailing message separator lines are not included.  When applied to
multiple messages, the command is invoked on each message individually.@* 
If invoked with one @t{C-u} the text portion of the message is used.@*
If invoked with two @t{C-u}'s the header portion of the message is used.@*
In invoked with three @t{C-u}'s the visible headers and the text
portions of the message are used.@*
If the shell command generates any output, it is displayed in a
@samp{*Shell Command Output*} buffer.  The message itself is not altered.
@findex vm-pipe-message-to-command-discard-output
@kindex |d
@item |d (@code{vm-pipe-message-to-command-discard-output})
Runs a shell command with some or all of the current message as input,
like the above, but will not display the output.
@findex vm-pipe-messages-to-command
@vindex vm-pipe-messages-to-command-start
@vindex vm-pipe-messages-to-command-end
@kindex |s
@item |s (@code{vm-pipe-messages-to-command})
Runs a shell command using as input the current message or marked
messages in the mbox format.  In contrast to
@code{vm-pipe-message-to-command}, the leading and trailing separator
lines are included.  This behaviour can be altered using the variables
@code{vm-pipe-messages-to-command-start} and
@code{vm-pipe-messages-to-command-end}. 
@findex vm-pipe-messages-to-command-discard-output
@kindex |n
@item |n (@code{vm-pipe-messages-to-command-discard-output})
Runs a shell command using as input the current message or marked
messages in the mbox format, but will not display the output.
@end table

@vindex vm-berkeley-mail-compatibility
A non-@code{nil} value of @code{vm-berkeley-mail-compatibility}
means to read and write BSD @i{Mail(1)} style Status: headers.
This makes sense if you plan to use VM to read mail archives
created by @i{Mail}.


@node Deleting Messages, Editing Messages, Saving Messages, Top
@chapter Deleting Messages

In VM, messages are flagged for deletion, and then are subsequently
@dfn{expunged} or removed from the folder.  The messages are not removed
from the on-disk copy of the folder until the folder is saved.

@table @kbd
@findex vm-delete-message
@kindex d
@item d (@code{vm-delete-message})
Flags the current message for deletion.  A prefix argument @var{n}
causes the current message and the next @var{n-1} messages to be flagged.
A negative @var{n} causes the current message and the previous @var{n-1}
messages to be flagged.
@findex vm-undelete-message
@kindex u
@item u (@code{vm-undelete-message})
Removes the deletion flag from the current message.  A prefix argument @var{n}
causes the current message and the next @var{n-1} messages to be undeleted.
A negative @var{n} causes the current message and the previous @var{n-1}
messages to be undeleted.
@findex vm-kill-subject
@kindex k
@item k (@code{vm-kill-subject})
Flags all messages with the same subject as the current message (ignoring
``Re:'') for deletion.
@findex vm-delete-duplicate-messages
@item @code{vm-delete-duplicate-messages}
Flags duplicate messages for deletion.  The duplicate messages are
detected by comparing message ID's.
@findex vm-delete-duplicate-messages-by-body
@item @code{vm-delete-duplicate-messages-by-body}
Flags duplicate messages for deletion.  The duplicate messages are
detected by comparing message bodies.
@findex vm-expunge-folder
@kindex ###
@item ### (@code{vm-expunge-folder})
Does the actual removal of messages flagged for deletion in the current
folder.
@end table

@vindex vm-move-after-deleting
@vindex vm-move-after-killing
@vindex vm-move-after-undeleting
Setting the variable @code{vm-move-after-deleting} non-@code{nil} causes
VM to move past the messages after flagging them for deletion.  Setting
@code{vm-move-after-undeleting} non-@code{nil} causes similar movement
after undeletes.  Setting @code{vm-move-after-killing} non-@code{nil}
causes VM to move after killing messages with @code{vm-kill-subject}.
Note that the movement is done by calling @code{vm-next-message} which
means that the value of @code{vm-circular-folders} applies to the
post-command motion as for a motion command, not as for a non-motion
command.

@node Editing Messages, Marking Messages, Deleting Messages, Top
@chapter Editing Messages

To edit a message, type @kbd{e} (@code{vm-edit-message}).  The current
message is copied into a temporary buffer, and this buffer is selected
for editing.  The major mode of this buffer is controlled by the
variable @code{vm-edit-message-mode}.  The default is Text mode.

Use @kbd{C-c ESC} (@code{vm-edit-message-end}) when you have finished
editing the message.  The message will be inserted into its folder,
replacing the old version of the message.  If you want to quit the edit
without your edited version replacing the original, use @kbd{C-c C-]}
(@code{vm-edit-message-abort}), or you can just kill the edit buffer
with @kbd{C-x k} (@code{kill-buffer}).

If you give a prefix argument to @code{vm-edit-message}, then the
current message will be flagged unedited.

As with VM Mail mode buffers, all VM commands can be accessed from
the edit buffer through the command prefix @kbd{C-c C-v}.

@node Marking Messages, Message Attributes, Editing Messages, Top
@chapter Marking Messages

@cindex marking
@cindex searching
@cindex virtual folders
VM provides a way to mark selected messages so that subsequent
operations can be applied to them.  This is similar to marking in other
parts of Emacs, e.g., @xref{Marks vs Flags, Dired Marks, Dired Marks,
emacs}, but arguably more powerful.  For example, one can mark all
messages from a particular sender and save them to a folder, or mark all
messages with a particular subject and print them.  One can also mark
messages by searching for particular strings in their text.

To mark the current message, type @kbd{M M}
(@code{vm-mark-message}).  If you give a numeric prefix argument
@var{n}, the next @var{n-1} messages will be marked as well.  A negative
prefix argument means mark the previous @var{n-1}.  An asterisk
(@samp{*}) will appear to the right of the message numbers of all marked
messages in the summary window.

To remove a mark from the current message, use @kbd{M U}
(@code{vm-unmark-message}).  Prefix arguments work as with
@code{vm-mark-message}.

Use @kbd{M m} to mark all messages in the current folder; @kbd{M u}
removes marks from all messages.

Other marking commands:

@table @kbd
@findex vm-mark-matching-messages
@kindex M C
@item M C (@code{vm-mark-matching-messages})
Mark all messages matched by a virtual folder selector.
@xref{Virtual Folders}.
@findex vm-unmark-matching-messages
@kindex M c
@item M c (@code{vm-unmark-matching-messages})
Unmark all messages matched by a virtual folder selector.
@findex vm-mark-thread-subtree
@kindex M T
@item M T (@code{vm-mark-thread-subtree})
Mark all messages in the thread tree rooted at current message.
@xref{Threading}.
@findex vm-unmark-thread-subtree
@kindex M t
@item M t (@code{vm-unmark-thread-subtree})
Unmark all messages in the thread tree rooted at current message.
@findex vm-mark-messages-same-subject
@kindex M S
@item M S (@code{vm-mark-same-subject})
Mark messages with the same subject as the current message.
@findex vm-unmark-messages-same-subject
@kindex M s
@item M s (@code{vm-unmark-same-subject})
Unmark messages with the same subject as the current message.
@findex vm-mark-messages-same-author
@kindex M A
@item M A (@code{vm-mark-same-author})
Mark messages with the same author as the current message.
@findex vm-unmark-messages-same-author
@kindex M a
@item M a (@code{vm-unmark-same-author})
Unmark messages with the same author as the current message.
@end table

While the above commands can be used in any VM buffer, the following
commands can be used in a Summary buffer to mark or unmark a region of
message summary lines.

@table @kbd
@findex vm-mark-summary-region
@kindex M R
@item M R (@code{vm-mark-summary-region})
Mark all messages in the current region in a Summary buffer
@findex vm-unmark-summary-region
@kindex M r
@item M r (@code{vm-unmark-summary-region})
Unmark all messages in the current region in a Summary buffer
@end table

To apply a VM command to all marked messages you must prefix it with the
key sequence @kbd{M N} (@code{vm-next-command-uses-marks}).  The next VM
command will apply to all marked messages, provided the command can be
applied to such messages in a meaningful and useful way.  Unfortunately,
as of this writing, this mechanism works only if the next command
invoked is a keyboard command.  Commands invoked by @code{M-x} are
unable to access the marked messages.  So, to invoke a complex command,
you might temporarily bind it to an unused key, e.g.,

@example
M-x local-set-key C vm-forward-message-all-headers
M N C
@end example

@noindent forwards marked messages with all headers included.

@node Message Attributes, Sorting Messages, Marking Messages, Top
@chapter Message Attributes
@cindex message attributes

Each message in a folder has a set of attributes that VM will remember
from session to session.  Various VM commands set and unset these
attributes.  Here are the attributes maintained by VM.

@table @code
@item new
The message was retrieved from a spool file during this
visit of the current folder.
@item unread
The message was retrieved from a spool file during some
past visit of the folder but is still unread.
@item filed
The message has been saved to some folder.
@item written
The body of the message has been saved to a file.
@item edited
The message has been altered (with @code{vm-edit-message}) since it arrived.
@item deleted
The message is deleted and will be removed from the folder
at the next expunge.
@item forwarded
The message has been forwarded with either
@code{vm-forward-message} or @code{vm-send-digest}.
@item redistributed
The message has been forwarded with the
@code{vm-resend-message} command.
@item replied
The message has been replied to.
@end table

You can set and unset these attributes directly by using @kbd{a}
(@code{vm-set-message-attributes}).  You will be prompted in the
minibuffer for names of the attributes and you can enter them with
completion.  Every attribute has an ``un-'' prefixed name you can use
to unset the attribute, excepting ``new'' and ``unread'', which are both
negated by ``read''.  You can use a prefix argument with this command to
affect multiple messages, and you can apply this command to marked
messages with @kbd{M N}.

@findex vm-undo
@kindex C-x u
@kindex C-_
@cindex undo
VM provides a special form of undo which allows changes to message
attributes to be undone.  Typing @kbd{C-x u} or @key{C-_}
(@code{vm-undo}) undoes the last attribute change.  Consecutive
@code{vm-undo}'s undo further and further back.  Any intervening command
breaks the undo chain, after which the undo's themselves become undoable
by subsequent invocations of @code{vm-undo}.

Note that expunges, saves and message edits are @emph{not} undoable.

@findex vm-add-message-labels
@findex vm-delete-message-labels
@kindex l a
@kindex l d
@cindex message labels
@dfn{Labels} are user-defined message attributes.  They can have any
name and be assigned any meaning by you.  Labels are added with
@kbd{l a} (@code{vm-add-message-labels} and @kbd{l e}
(@code{vm-add-existing-message-labels}, and are removed by @kbd{l d}
(@code{vm-delete-message-labels}).  BABYL format folders use labels to
store basic attributed like ``deleted'' and ``unread''.  When visiting a
BABYL folder VM uses these labels also in order to be compatible with
other BABYL mailers.  The labels used are ``recent'', ``unseen'',
``deleted'', ``answered'', ``forwarded'', ``redistributed'', ``filed'',
``edited'' and ``written''.  If (and only if) you are using BABYL format
folders, you should not use these label names for your own purposes.

@vindex vm-flush-interval
@cindex autosave
All message attributes are stored in the folder.  In order for
attribute changes to be saved to disk, they must be written to
the folder's buffer prior to the buffer being saved.  The
variable @code{vm-flush-interval} controls how often that is done.  A
value of @code{t} means write the new attributes to the folder
buffer whenever a change occurs.  A value of @code{nil} means
wait until just before the folder is saved before writing out the
attributes.  VM will work faster with this setting, but if Emacs
or your system crashes, the autosave file will contain no useful
data pertaining to message attribute changes.  The autosave file
will still reflect message edits and expunges.  @xref{Crash
Recovery}.  A positive integer value @var{n} instructs VM to write
out attribute changes every @var{n} seconds.  The default value
of this variable is @code{t}.

@node Sorting Messages, Digests, Message Attributes, Top
@chapter Sorting Messages

@cindex sorting
@findex vm-sort-messages
@vindex vm-move-messages-physically
@kindex G
In order to make numerous related messages easier to cope with, VM
provides the command @kbd{G} (@code{vm-sort-messages}), which sorts
all messages in a folder using one or more sort keys.
By default the actual order of the messages in the folder is not
altered; that is, if you looked at the folder file outside of VM the
message order would be unchanged.  VM numbers and presents the messages
in a different order internally.  If you want the message order to be
changed in the folder so that other programs can see the change, you
can either invoke @code{vm-sort-messages} with a prefix argument, or
you can set @code{vm-move-messages-physically} non-@code{nil} before
sorting.  Either way, VM will shift the actual messages around in the
folder buffer, and when you save the folder, the order change will be
visible to other programs.

@cindex spam
Valid sort keys are:
@multitable @columnfractions 0.35 0.60
@item date @tab reversed-date
@item author @tab reversed-author
@item subject @tab reversed-subject
@item recipients @tab reversed-recipients
@item line-count @tab reversed-line-count
@item byte-count @tab reversed-byte-count
@item physical-order @tab reversed-physical-order
@item spam-score @tab reversed-spam-score
@end multitable

@vindex vm-subject-ignored-prefix
@vindex vm-subject-ignored-suffix
When sorting by subject (or threading using subjects, or killing
messages by subject) the subject of the message is
@dfn{normalized} before comparisons are done.  A @dfn{normalized}
subject has uninteresting prefixes and suffixes stripped off, and
multiple consecutive white space characters are collapsed to a single
space.  The variable @code{vm-subject-ignored-prefix} should be
a regular expression that matches all strings at the beginning of
a subject that you do not want to be considered when message
subjects are compared.  A @code{nil} value means VM should not ignore
any prefixes.  The analogous variable for subject suffixes is
@code{vm-subject-ignored-suffix}.

@vindex vm-subject-significant-chars
Once the subject has been normalized, the variable
@code{vm-subject-significant-chars} controls how much of what
remains is considered significant for matching purposes.  The
first @code{vm-subject-significant-chars} will be considered
significant.  Characters beyond this point in the subject string
will be ignored.  A @code{nil} value for this variable means all
characters in the subject are significant.

If you want to move messages around by hand, use @kbd{C-M-n}
(@code{vm-move-message-forward}) and @kbd{C-M-p}
(@code{vm-move-message-backward}).  The default is to move the current
message forward or backward by one message in the message list.  A
prefix argument @var{n} can specify a longer move.  The value of
@code{vm-move-messages-physically} applies to these commands.

@menu
* Threading::   Using subjects and message IDs to group messages.
@end menu

@node Threading,,, Sorting Messages
@section Threading
@cindex threading
A @dfn{thread} is a group of messages that are either related by subject
or that have a common ancestor.  @dfn{Threading} is the process of
determining the relationship between such messages and displaying them so
that those relationships are evident.

@findex vm-toggle-threads-display
@vindex vm-summary-thread-indent-level
To enable and disable threading, type @kbd{C-t}
(@code{vm-toggle-threads-display}.  In the summary buffer related
messages are grouped together and the subject part of the summary
listings of messages are indented to show hierarchical relationships.
Parent messages are displayed before their children and children are
indented a default two spaces to the right for each level of descendence
from their ancestors.  The amount of indentation per level is controlled by the
variable @code{vm-summary-thread-indent-level}.

@vindex vm-thread-using-subject
Message relationships are discovered by examining References,
In-Reply-To, and Subject headers.  The first two headers are more
reliable sources of information but not all mailers provide them.
If you don't want VM to use Subject headers, set the variable
@code{vm-thread-using-subject} to @code{nil}.

If you want VM to always display messages using threads, you should set
the default value of the variable @code{vm-summary-show-threads}
non-@code{nil} in your VM init file.  Example:

@example
(setq-default vm-summary-show-threads t)
@end example

@noindent Do not use @code{setq}, as this will only set the value of
the variable in a single buffer.  Once you've started VM you should
not change the value of this variable.  Rather you should use
@kbd{C-t} to control the thread display.

Note that threading is really a specialized form of sorting, and so the
value of the variable @code{vm-move-messages-physically} applies.

@node Digests, Summaries, Sorting Messages, Top
@chapter Digests

A @dfn{digest} is one or more mail messages encapsulated within another
message.

VM supports digests by providing a command to ``burst'' them into their
individual messages.  These messages can then be handled like any other
messages under VM.

@findex vm-burst-digest
@kindex *
The command @kbd{*} (@code{vm-burst-digest}) bursts a digest into its
individual messages and appends them to the current folder.  These
messages are then assimilated into the current folder as new messages.
The original digest message is not altered, and the messages extracted
from it are not part of the on-disk copy of the folder until a save is
done.  You will be prompted for the type of digest to burst.  VM
understands three formats, ``rfc934'', ``rfc1154'' and ``mime''.  If you
don't know what kind of digest you've received, type ``guess'' and VM
will try to figure out the digest type on its own.  VM is pretty smart
about digests and will usually make the correct choice if the digest is
properly formatted.

@node Summaries, Virtual Folders, Digests, Top
@chapter Summaries

@findex vm-summarize
@vindex vm-auto-center-summary
@vindex vm-summary-arrow
@kindex h
Typing @kbd{h} (@code{vm-summarize}) causes VM to display a summary of
contents of the current folder.  The information in the summary is
automatically updated as changes are made to the current folder.  An
arrow @samp{->} appears to the left of the line summarizing the current
message.  The variable @code{vm-auto-center-summary} controls whether VM
will keep the summary arrow vertically centered within the summary
window.  A value of @code{t} causes VM to always keep the arrow
centered.  A value of @code{nil} (the default) means VM will never
bother centering the arrow.  A value that is not @code{nil} and not
@code{t} causes VM to center the arrow only if the summary window is not
the only existing window.  You can change what the summary arrow looks
like by setting vm-summary-arrow to a string depicting the new arrow.
You should set this variable before VM creates the summary buffer.

@vindex vm-summary-format
The variable @code{vm-summary-format} controls the format of each
message's summary.  Its value should be a string.  This string should
contain printf-like ``%'' conversion specifiers which substitute
information about the message into the final summary.

Recognized specifiers are:
@table @code
@item a
attribute indicators (always four characters wide)
@*
The first char is  `D', `N', `U' or ` ' for deleted, new, unread
and read messages respectively.
@*
The second char is `F', `W' or ` ' for filed (saved) or written
messages.
@*
The third char is `R', `Z' or ` ' for messages replied to,
and forwarded messages.
@*
The fourth char is `E' if the message has been edited, ` ' otherwise.
@item A
longer version of attributes indicators (seven characters
wide).@*
@*
The first char is  `D', `N', `U' or ` ' for deleted, new, unread
and read messages respectively.
@*
The second is `r' or ` ', for message replied to.
@*
The third is `z' or ` ', for messages forwarded.
@*
The fourth is `b' or ` ', for messages redistributed.
@*
The fifth is `f' or ` ', for messages filed.
@*
The sixth is `w' or ` ', for messages written.
@*
The seventh is `e' or ` ', for messages that have been edited.
@item P
indicator for a message with attachments.
The variable @vindex vm-summary-attachment-indicator
@code{vm-summary-attachment-indicator} is the inserted string, by default a @code{$}.
@item p
indicator for a postponed message. 
The variable @vindex vm-summary-postponed-indicator
@code{vm-summary-postponed-indicator} is the inserted string, by default a @code{P}.
@item c
number of characters in message (ignoring headers)
@item S
human readable size of the message
@item d
numeric day of month message sent
@item f
author's address
@item F
author's full name (same as f if full name not found)
@item h
hour:min:sec message sent
@item H
hour:min message sent
@item i
message ID
@item I
thread indentation
@item l
number of lines in message (ignoring headers)
@item L
labels (as a comma list)
@item m
month message sent
@item M
numeric month message sent (January = 1)
@item n
message number
@item s
message subject
@item t
addresses of the recipients of the message, in a comma-separated list
@item T
full names of the recipients of the message, in a comma-separated list
If a full name cannot be found, the corresponding address is used
instead.
@item U
user defined specifier.  The next character in the format
string should be a letter.  VM will call the function
vm-summary-function-<letter> (e.g. vm-summary-function-A for
``%UA'') in the folder buffer with the message being summarized
bracketed by (point-min) and (point-max).  The function
will be passed a message struct as an argument.
The function should return a string, which VM will insert into
the summary as it would for information from any other summary
specifier.
@item w
day of the week message sent
@item y
year message sent
@item z
timezone of date when the message was sent
@item *
`*' if the message is marked, ` ' otherwise
@item (
starts a group, terminated by %).  Useful for specifying
the field width and precision for the concatenation of
group of format specifiers.  Example: \"%.35(%I%s%)\"
specifies a maximum display width of 35 characters for the
concatenation of the thread indentation and the subject.
@item )
ends a group.
@end table

Use ``%%'' to get a single ``%''.

A numeric field width may be specified between the ``%'' and the
specifier; this causes right justification of the substituted string.  A
negative field width causes left justification.  The field width may be
followed by a ``.'' and a number specifying the maximum allowed length
of the substituted string.  If the string is longer than this value, it
is truncated.

@vindex vm-summary-uninteresting-senders
@vindex vm-summary-uninteresting-senders-arrow
If you save copies of all your outbound messages in a folder and
later visit that folder, the @samp{%F} format specifier will normally
display your own name.  If you would rather see the recipient
addresses in this case, set the variable
@code{vm-summary-uninteresting-senders}. This variable's value,
if non-@code{nil}, should be a regular expression that matches
addresses that you don't consider interesting enough to appear in
the summary.  When such senders would be displayed by the @samp{%F} or
@samp{%f} summary format specifiers VM will substitute the value of
@code{vm-summary-uninteresting-senders-arrow} (default "To: ")
followed by what would be shown by the @samp{%T} and @samp{%t} specifiers
respectively.

The summary format need not be one line per message but it must end with
a newline, otherwise the message pointer will not be displayed correctly
in the summary window.

@findex vm-fix-my-summary!!!
Summary lines are precomputed and cached in the folder buffer.  If you
change the @code{vm-summary-format}, you need to force the cache to be
updated.  You can do this by the command@code{vm-fix-my-summary!!!}.

Every folder can have its own summary format.  The format is written
into the folder and saved on the disk.  When you visit the folder
again, you can reuse the saved summary format.  Set the variable
@code{vm-restore-saved-summary-format} to t to achieve this effect.

You can have a summary generated automatically at VM startup
by setting the variable @code{vm-startup-with-summary} non-nil.
@xref{Starting Up}.

@vindex vm-follow-summary-cursor
All VM commands are available in the summary buffer just as they are in
the folder buffer itself.  If you set @code{vm-follow-summary-cursor}
non-@code{nil}, VM will select the message under the cursor in the
summary window before executing commands that operate on the current
message.  Note that this occurs @emph{only} when executing a command
from the summary buffer window.

@vindex vm-gargle-uucp
A non-@code{nil} value of @code{vm-gargle-uucp} means to use a crufty
regular expression that does surprisingly well at beautifying UUCP
addresses that are substituted for @samp{%f} and @samp{%t} as part
of summary and attribution formats.

@node Virtual Folders, Frames and Windows, Summaries, Top
@chapter Virtual Folders

@cindex searching
@cindex virtual folders

A @dfn{virtual folder} is a mapping of messages from one or more
real folders into a container that in most ways acts like a
real folder but has no real existence outside of VM.  You can have a
virtual folder that contains a subset of messages in a real folder
or several real folders.  A virtual folder can also contain a
subset of messages from another virtual folder.

A virtual folder is defined by its name, the folders that it contains
and its selectors.  The variable @code{vm-virtual-folder-alist} is a list of
the definitions of all named virtual folders.  In order to visit a
virtual folder with the @code{vm-visit-virtual-folder} (@kbd{V V}) command,
a virtual folder must have an entry in vm-virtual-folder-alist.

Each virtual folder definition should have the following form:

@example
(@var{VIRTUAL-FOLDER-NAME}
  ( (@var{FOLDER} ...)
    (@var{SELECTOR} [@var{ARG} ...]) ... )
  ... )
@end example

@var{VIRTUAL-FOLDER-NAME} is the name of the virtual folder being defined.
This is the name by which you and VM will refer to this folder.

@var{FOLDER} should be the specification of a real folder: a file path
for a local folder or a maildrop specification for a POP/IMAP folder.  There may
be more than one @var{FOLDER} listed, the @var{SELECTOR}s within
that sublist will apply to them all.  If @var{FOLDER} is a
directory, VM will assume this to mean that all the folders in
that directory should be searched.

The @var{SELECTOR} is a Lisp symbol that tells VM how to
decide whether a message from one of the folders specified by
the @var{FOLDER-NAME}s should be included in the virtual
folder.  Some @var{SELECTOR}s require an argument @var{ARG};
unless otherwise noted, @var{ARG} may be omitted.

@table @code
@item any
matches any message.
@item author
matches message if @var{ARG} matches the author; @var{ARG} should be a
regular expression.
@item author-or-recipient
matches message if @var{ARG} matches the author of
the message or any of its recipients; @var{ARG}
should be a regular expression.
@item recipient
matches message if @var{ARG} matches any part of the recipient
list of the message.  @var{ARG} should be a regular expression.
@item outgoing
matches message if your are the author of it, i.e. if the author matches
@code{vm-summary-uninteresting-senders}
@item in-bbdb
matches if the addresses from the message are in the BBDB. With an option
first argument you might select the address class and with an optional second
argument if only the first address should be checked.  
@example
(in-bbdb authors)
@end example
@example
(in-bbdb recipients t)
@end example
@item subject
matches message if @var{ARG} matches any part of the message's
subject; @var{ARG} should be a regular expression.
@item text
matches message if @var{ARG} matches any part of the text
portion of the message; @var{ARG} should be a
regular expression.
@item sent-after
matches message if it was sent after the date @var{ARG}.
A fully specified date looks like this:
@example
``31 Dec 1999 23:59:59 GMT''
@end example
@noindent although the parts can appear in any order.
You can leave out any part and it will
default to the current date's value for that
part, with the exception of the @samp{hh:mm:ss}
part which defaults to midnight.
@item sent-before
matches message if it was sent before the date @var{ARG}.
A fully specified date looks like this:
@example
``31 Dec 1999 23:59:59 GMT''
@end example
@noindent although the parts can appear in any order.
You can leave out any part and it will
default to the current date's value for that
part, with the exception of the hh:mm:ss
part which defaults to midnight.
@item older-than
matches message if it is more than @var{ARG} days old
@item newer-than
matches message if it is at most @var{ARG} days old
@item header
matches message if @var{ARG} matches any part of the header
portion of the message; @var{ARG} should be a
regular expression.
@item header-or-text
matches message if @var{ARG} matches any part of the
headers or the text portion of the message;
@var{ARG} should be a regular expression.
@item spam-score
matches message if its spam score is at least @var{ARG}.  See
@code{vm-vs-spam-score-headers} for configuration.
@item spam-word
matches message if it contains a word from @code{vm-spam-words-file}
@item deleted
matches message if it is flagged for deletion.
@item undeleted
matches message if it has not been deleted.
@item edited
matches message if it has been edited.
@item unedited
matches message if it has not been edited.
@item filed
matches message if it has been saved with its headers.
@item unfiled
matches message if it has not been saved with its
headers.
@item written
matches message if it has been saved without its headers.
@item new
matches message if it is new.
@item recent
matches message if it is new.  Same as the @code{new} selector.
@item read
matches message if it is neither new nor unread.
@item unread
matches message if it is not new and hasn't been read.
@item unseen
matches message if it is not new and hasn't been read.
Same as the @code{unread} selector.
@item replied
matches message if it has been replied to.
@item answered
matches message if it has been replied to.  Same as the @code{replied}
selector. 
@item unreplied
matches message if it has not been replied to.
@item unanswered
matches message if it has not been replied to.
Same as the @code{unreplied} selector.
@item forwarded
matches message if it has been forwarded using
a variant of @code{vm-forward-message} or @code{vm-send-digest}.
@item unforwarded
matches message if it has not been forwarded using
@code{vm-forward-message} or @code{vm-send-digest} or one
of their variants.
@item redistributed
matches message if it has been redistributed using
@code{vm-resend-message}.
@item unredistributed
matches message if it has not been redistributed using
@code{vm-resend-message}.
@item marked
matches message if it is marked, as with
@code{vm-mark-message}.
@item attachment
matches if a message contains an attachment resp. its text matched
@code{vm-vs-attachment-regexp}
@item less-chars-than
matches message if message has less than @var{ARG}
characters.  @var{ARG} should be a number.
@item less-lines-than
matches message if message has less than @var{ARG}
lines.  @var{ARG} should be a number.
@item more-chars-than
matches message if message has more than @var{ARG}
characters.  @var{ARG} should be a number.
@item more-lines-than
matches message if message has more than @var{ARG}
lines.  @var{ARG} should be a number.
@item eval
matches message if evaluating the sexpr @var{ARG} yields @code{t}.
@item and
matches the message if all its argument
selectors match the message.  Example:
@example
(and (author "Derek McGinty") (new))
@end example
@noindent matches all new messages from Derek McGinty.
@code{and} takes any number of arguments.
@item not
matches message only if its selector argument
does NOT match the message.  Example:
@example
(not (deleted))
@end example
@noindent matches messages that are not deleted.
@item or
matches the message if any of its argument
selectors match the message.  Example:
@example
(or (author "Dave Weckl") (subject "drum"))
@end example
@noindent matches messages from Dave Weckl or messages
with the string ``drum'' in their Subject header.
@code{or} takes any number of arguments.
@item folder-name
matches message if it is from a folder matching @code{ARG}
@item virtual-folder-member
matches message if the message is already a
member of some virtual folder currently
being visited.
@item vm-mode
matches the message if in vm-mode and one of its argument
selectors match the message.
@item mail-mode
matches the message if in mail-mode and one of its argument
selectors match the message.
@end table

Here is an example that you may find useful as a template to
create virtual folder definitions.

@example
(setq vm-virtual-folder-alist
   '(
     ;; start virtual folder definition
     ("virtual-folder-name"
      (("/path/to/folder" "/path/to/folder2")
       (header "foo")
       (header "bar")
      )
      (("/path/to/folder3" "/path/to/folder4")
       (and (header "baz") (header "woof"))
      )
     )
     ;; end of virtual folder definition
   )
)
@end example

The @code{text} selector provides a particularly effective way to search
for strings in messages.  It is better than the
@code{vm-isearch-forward/backward} functions because it avoids searching
inside encoded attachments, hence faster.

Again, you visit virtual folders you have defined in
@code{vm-virtual-folder-alist} with @kbd{V V}.  Once you've
visited a virtual folder most VM commands work as they do in a
normal folder.  There are exceptions.  If you use @kbd{S}
(@code{vm-save-folder}, the folder save command will be invoked
on each real folder in turn.  Similarly if you use @kbd{g}
(@code{vm-get-new-mail} in a virtual folder, mail is retrieved
from the spool files associated with each of the real folders.
If any of the retrieved messages are matched by the virtual
folder's selector, they will be added to the virtual folder.

These commands will signal an error when invoked if the current
folder is a virtual folder:

@display
    vm-save-buffer
    vm-write-file
    vm-change-folder-type
    vm-expunge-imap-messages
    vm-expunge-pop-messages
@end display

Normally messages in a virtual folder share attributes with the
underlying real messages.  For example, if you delete a message
in a virtual folder, it is also flagged as deleted in the real
folder.  If you then run @code{vm-expunge-folder} in the virtual folder,
the deleted message is expunged from the virtual folder and from
the real folder.  Labels are shared between virtual and real
messages.  However virtual folders have their own set of message
marks.

To make virtual folders not share message attributes with real
folders by default, set the variable @code{vm-virtual-mirror} to nil.
This should be done in your VM init file and you should use
@code{setq-default}, as this variable is automatically local to all
buffers.

@example
(setq-default vm-virtual-mirror nil)
@end example

@noindent If you want to change whether the currently visited virtual 
folder shares attributes with the underlying real folders, use the 
command @code{vm-toggle-virtual-mirror} (bound to @kbd{V M}).  If the
virtual folder is currently sharing attributes it will no longer
be.  If it is not sharing attributes with the underlying folders
then it will be.

The command @code{vm-create-virtual-folder} (bound to @kbd{V C}) lets
you interactively create a virtual folder from the messages of
the current folder, using exactly one selector to choose the
messages.  If you type @kbd{V C header RET pigs}, VM will create
a folder containing only those messages that contain the string
@samp{pigs} in the header.

The command @code{vm-apply-virtual-folder} (bound to @kbd{V X}) tries
the selectors of a named virtual folder against the messages of
the current folder and creates a virtual folder containing the
matching messages.

The keys @kbd{V S} and @kbd{V A} invoke
@code{vm-create-virtual-folder-same-subject} and
@code{vm-create-virtual-folder-same-author} which create virtual folders
containing all the messages in the current folder with the same
subject or author as the current message.

The keys @kbd{V O} @code{vm-virtual-omit-message} will omit a message from the
virtual folder and @kbd{V U} @code{vm-virtual-update-folders} will force an
update of the virtual folder.

@kbd{M-x vm-virtual-check-selector-interactive} allows to test selectors
interactively and will emit debug information when called with a prefix
argument.

Automatic marking of messages for deletion based on a selector can be
achieved with @code{vm-virtual-auto-delete-message} for interactive use and the
function and @code{vm-virtual-auto-delete-messages} when added to the VM hook
@code{vm-arrived-messages-hook}.  This can be quite handy for marking spam for
deletion.

@kbd{M-x vm-virtual-save-message} can be used to save messages to the 
folder corresponding to the first matching selector and the function 
@kbd{vm-virtual-auto-archive-messages} can file messages based on
selectors (see also @code{vm-auto-archive-messages}). 

@node Frames and Windows, Toolbar, Virtual Folders, Top
@chapter Frames and Windows

VM uses Emacs frames and windows to display messages and summaries
and to provide a place for you to compose messages.  Using VM's
frame configuration facilities you can control when VM creates new
frames and the size and attributes associated with new frames.
Inside each frame you can associate different window setups with
commands and classes of commands by using VM's window configuration
facilities.

@vindex vm-mutable-frames
To use VM's frame configuration features, the variable
@code{vm-mutable-frames} must be set non-@code{nil}.  This is the default.  If
@code{vm-mutable-frames} is set to nil VM will only use the current
frame, and VM will not create, delete or resize frames.

@vindex vm-mutable-windows
To use window configurations, the variable @code{vm-mutable-windows}
must be set non-@code{nil}.  If @code{vm-mutable-windows} is set to nil, VM
will only use the selected window, and will not create, delete or
resize windows.

@menu
* Frame Configuration::         How to configure frame use and appearance.
* Window Configuration::        How to configure window use and appearance.
@end menu

@node Frame Configuration, Window Configuration, Frames and Windows, Frames and Windows
@section Frame Configuration

VM has a set of variables that let you specify when VM creates
frames and what attributes the new frames will have.

@vindex vm-frame-per-folder
If @code{vm-frame-per-folder} is set non-@code{nil}, when you visit a folder,
VM will create a new frame and display that folder in the new
frame.  When you quit the folder, VM will delete the frame.

@vindex vm-frame-per-summary
If @code{vm-frame-per-summary} is set non-@code{nil}, the @code{vm-summarize}
command will create a new frame in which to display a folder's summary
buffer.  This works best if a full-screen window configuration has
been assigned to the @code{vm-summarize} command.  When you quit the folder
or kill the summary, VM will delete the frame.

@vindex vm-frame-per-composition
Setting @code{vm-frame-per-composition} non-@code{nil} causes VM to create a
new frame for the composition buffer when you run any of VM's
message composition commands.  E.g. @code{vm-reply-include-text},
@code{vm-mail}, @code{vm-forward-message}.  When you finish editing the
composition and send it, or when you kill the composition buffer, 
the frame will be deleted.

@vindex vm-frame-per-edit
The variable @code{vm-frame-per-edit}, if non-@code{nil}, tells VM to create a
new frame when the vm-edit-message command is run.  When you
finish editing the message, or abort the edit, the frame will be
deleted.

@vindex vm-frame-per-help
If @code{vm-frame-per-help} is set non-@code{nil}, VM will create a new frame
to display any help buffer produced by the vm-help command.

@vindex vm-frame-per-completion
If @code{vm-frame-per-completion} is set non-@code{nil}, VM will create a new
frame on mouse initiated completing reads.  A mouse initiated
completing read occurs when you invoke a VM command using the
mouse, either with a menu or a toolbar button.  That command
must then prompt you for information, and there must be a
limited set of valid responses.  If these conditions are met
and @code{vm-frame-per-completion}'s value is non-@code{nil}, VM will
create a new frame containing a list of responses that you can
select with the mouse.

@vindex vm-search-other-frames
When VM is deciding whether to create a new frame, it checks
other existing frames to see if a buffer that it wants to display in a
frame is already being displayed somewhere.  If so, then VM will
not create a new frame.  If you don't want VM to search other
frames, set the variable @code{vm-search-other-frames} to @code{nil}.  VM will 
still search the currently selected frame and will not create a
new frame if the buffer that it wants to display is visible there.

@vindex vm-frame-parameter-alist
The variable @code{vm-frame-parameter-alist} allows you to specify the
frame parameters for newly created frames.

The value of @code{vm-frame-parameter-alist} should be of this form

@example
((@var{SYMBOL} @var{PARAMLIST}) (@var{SYMBOL2} @var{PARAMLIST2}) ...)
@end example

@var{SYMBOL} must be one of ``completion'', ``composition'', ``edit'',
``folder'', ``primary-folder'' or ``summary''.  It specifies the type
of frame that the following @var{PARAMLIST} applies to.

@table @code
@item completion
specifies parameters for frames that display lists of
choices generated by a mouse-initiated completing read.
(See @code{vm-frame-per-completion}.)
@item composition
specifies parameters for mail composition frames.
@item edit
specifies parameters for message edit frames
(e.g. created by @code{vm-edit-message-other-frame})
@item folder
specifies parameters for frames created by `vm' and the
@code{vm-visit-} commands.
@item primary-folder
specifies parameters for the frame created by running
@code{vm} without any arguments.
@item summary
specifies parameters for frames that display a summary buffer
(e.g. created by @code{vm-summarize-other-frame})
@end table
@var{PARAMLIST} is a list of pairs as described in the documentation for
the function @code{make-frame}.

@node Window Configuration,, Frame Configuration, Frames and Windows
@section Window Configuration

Window configurations allow you to specify how the windows within
a frame should look for a particular command or class of
commands.  Each command can have a configuration associated with
it and you can also associate a configuration with command
classes like ``reading-message'' or ``composing-message''.  To
setup a window configuration, first use Emacs' window management
commands (@code{split-window}, @code{enlarge-window}, etc.)  to make the
windows in the frame look the way you want.  Then use the
switch-to-buffer command to put the buffers you want to see into
the windows.  Next type @kbd{W S}, which invokes the
@code{vm-save-window-configuration} command.  Type the name of the
command or class of commands to which you want the configuration
to apply.  Nearly all VM commands can be entered here.  Valid
classes are:

@display
    default
    startup
    quitting
    reading-message
    composing-message
    marking-message
    searching-message
@end display

When a VM command is executed, window configurations are searched 
for as follows.  First, a command specific configuration is
searched for.  If one is found, it is used.  Next a class
configuration is searched for.  Not all commands are in command
classes.  Message composition commands are in the
``composing-message'' class.  All the @code{vm-quit*} commands are in the
``quitting'' class.  All the VM commands that set and clear
message marks are in the ``marking-message'' class, and so on.
If such a class configuration is found it is used.  If no
matching class configuration is found, the ``default'' class
configuration is used, if it is defined.

Note that when a window configuration is saved the selected
window at that time will be the selected window when that window
configuration is used.  So if you prefer for the cursor to be in
a particular window, make sure you invoke
@code{vm-save-window-configuration} window from that window.  Remember
that you can invoke the command with @kbd{M-x} if VM's normal
key map is not in effect.

To delete a window configuration, use @kbd{W D} which is bound to
@code{vm-delete-window-configuration}.  You will be prompted for the
name of the configuration to delete.

To see what an existing configuration looks like, type @kbd{W W}
which invokes @code{vm-apply-window-configuration}.

@vindex vm-window-configuration-file
VM saves information about your window configurations in the file
named by the variable @code{vm-window-configuration-file}.  The default
location of the configuration file is @samp{"~/.vm.windows"}.
Do not make @code{vm-window-configuration-file} point to the same
location as @code{vm-init-file}, as the window configuration save
commands will then overwrite the content of your init file.

@node Toolbar, Menus, Frames and Windows, Top
@chapter Toolbar

VM can display a toolbar that allows you to run VM commands with
a single mouse click.  By default the toolbar is displayed on the
left of the Emacs frame and is only visible if you're running
under a window system like X Windows or Microsoft Windows.

@vindex vm-use-toolbar
To make VM not display the toolbar, set @code{vm-use-toolbar} to nil.
To configure what buttons are displayed on the toolbar, you must
change the value of @code{vm-use-toolbar}.  If non-@code{nil}, the value of
@code{vm-use-toolbar} should be a list of symbols and integers, which
specify which buttons appear on the toolbar and the layout of the
buttons.  These are the allowed symbols along with the buttons
they represent.

@table @code
@item autofile
The AutoFile button.  Clicking on this button runs the command
@code{vm-toolbar-autofile-message}.  This command will save the current
message into the folder matched by @code{vm-auto-folder-alist}, if there
is a match.
@item compose
The Compose button.  Clicking on this button runs the command
@code{vm-toolbar-compose-command}.  This command is normally just an
alias for the @code{vm-mail} command.  If you want the Compose button to 
do something else, redefine @code{vm-toolbar-compose-command} using
either @code{fset} or @code{defun}.
@item delete/undelete
The Delete/Undelete button.  If the current message is marked for 
deletion, this button displays as an Undelete button.  Otherwise
it displays as a Delete button.
@item file
The File button.  Clicking on this button runs the command
@code{vm-toolbar-file-command}.  This command is normally just an
alias for the @code{vm-mail} command.  If you want the File button to 
do something else, redefine @code{vm-toolbar-file-command} using
either @code{fset} or @code{defun}.
@item getmail
The Get Mail button.  Clicking on this button runs the command
@code{vm-toolbar-getmail-command}.  This command is normally just an
alias for the @code{vm-get-new-mail} command.  If you want the
Get Mail button to 
do something else, redefine @code{vm-toolbar-getmail-command} using
either @code{fset} or @code{defun}.
@item help
The Helper button.  Clicking on this button runs the command
@code{vm-toolbar-helper-command}.  This command normally just runs
@code{vm-help}, but it also does context specific things under certain
conditions.  If the current message is a MIME message that needs
decoding, the Helper button becomes the Decode MIME button.  If the 
current folder has an autosave file that appears to be the result 
of an Emacs or system crash, the Helper button becomes the Recover 
button.  Clicking on the Recover button runs @code{vm-recover-folder}, 
so you can recover your folder from an existing autosave file.
@item mime
The Decode MIME button.  Clicking on this button runs the command
@code{vm-toolbar-mime-command}.  This command is normally just an
alias for the @code{vm-decode-mime-message} command.
@item next
The Next button.  Clicking on this button runs the command
@code{vm-toolbar-next-command}.  This command is normally just an
alias for the @code{vm-next-message} command.  If you want the Next button to 
do something else, redefine @code{vm-toolbar-next-command} using
either @code{fset} or @code{defun}.
@item previous
The Previous button.  Clicking on this button runs the command
@code{vm-toolbar-previous-command}.  This command is normally just an
alias for the @code{vm-previous-message} command.  If you want the Previous button to 
do something else, redefine @code{vm-toolbar-previous-command} using
either @code{fset} or @code{defun}.
@item print
The Print button.  Clicking on this button runs the command
@code{vm-toolbar-print-command}.  This command is normally just an
alias for the @code{vm-print-message} command.  If you want the
Print button to 
do something else, redefine @code{vm-toolbar-print-command} using
either @code{fset} or @code{defun}.
@item quit
The Quit button.  Clicking on this button runs the command
@code{vm-toolbar-quit-command}.  This command is normally just an
alias for the @code{vm-quit} command.  If you want the Quit button to 
do something else, redefine @code{vm-toolbar-quit-command} using
either @code{fset} or @code{defun}.
@item reply
The Reply button.  Clicking on this button runs the command
@code{vm-toolbar-reply-command}.  This command is normally just an
alias for the @code{vm-reply-include-text} command.  If you want
the Reply button to
do something else, redefine @code{vm-toolbar-reply-command} using
either @code{fset} or @code{defun}.
@item visit
The Visit button.  Clicking on this button runs the command
@code{vm-toolbar-visit-command}.  This command is normally just an
alias for the @code{vm-visit-folder} command.  If you want the Visit button to 
do something else, redefine @code{vm-toolbar-visit-command} using
either @code{fset} or @code{defun}.
@item nil
If nil appears in the list, it must appear exactly once.  The
buttons associated with symbols that appear after nil in the
list will be display flushright for top and bottom toolbars, and
flushbottom for left and right toolbars.
@end table

If an positive integer appears in the the @code{vm-use-toolbar} list, it
specifies the number of pixels of blank space to display between
the button that comes before and the button that comes after the
integer.

@vindex vm-toolbar-orientation
The variable @code{vm-toolbar-orientation} controls on which side of the 
frame the toolbar is displayed.  E.g.

@example
(setq vm-toolbar-orientation 'top)
@end example

@noindent causes the toolbar to be displayed at the top of the frame.  The
@code{top} in the example can be replaced with @code{bottom},
@code{right} and @code{left} to make the toolbar appear in those
places instead.

@vindex vm-toolbar-pixmap-directory
VM finds the images for the toolbar in the directory specified by
@code{vm-toolbar-pixmap-directory}.  This variable should already be set
properly by whoever installed VM on your system, so you should
not need to set it.

@node Menus, Faces, Toolbar, Top
@chapter Menus

@vindex vm-popup-menu-on-mouse-3
VM uses Emacs' menu bar and pop-up menus when they are available to 
give you access to more of VM's commands.  By default VM puts a
context sensitive pop-up menu on mouse button 3 (usually the
rightmost mouse button).  If you don't want this menu, set the
variable @code{vm-popup-menu-on-mouse-3} to nil.

@vindex vm-use-menus
If you set @code{vm-use-menus} to nil, VM will not generate a menu bar
for VM folder buffers and VM won't use pop-up menus either.  If
you set @code{vm-use-menus} to @samp{1}, VM will add a single @samp{VM}
entry to the existing menu bar instead of using the whole menu bar
for its purposes.  That single entry will have all the VM command
submenus under it.

To make VM use the whole menu bar, you must set variable @code{vm-use-menus}
to a list of symbols.  The symbols and the order in which they are listed
determine which menus will be in the menu bar and how they are ordered.
Valid symbol values are:
@table @code
@item dispose
This is menu of commands that are commonly used to dispose of a
message.  E.g. reply, print, save, delete.
@item emacs
This is actually a menu button that causes the menu bar to change
to the global Emacs menu bar.  On that menu bar you will find
a VM button that will return you to the VM menu bar.
@item folder
This is a menu of folder related commands.  You can visit a
folder, save a folder, quit a folder and so on.
@item help
This is a menu of commands that provide information for you if
you don't know what to do next.
@item label
This is a menu of commands that let you add and remove message
labels from messages.
@item mark
This is a menu of commands that you can use to mark and unmark
messages based on various criteria.  @xref{Marking Messages}.
@item motion
This is a menu of commands to move around inside messages and
inside folders.
@item send
This is a menu of commands you use to compose and send messages.
@item sort
This is a menu of commands to sort a folder by various criteria.
@item undo
This is a menu button that invokes the @code{vm-undo} command.
@item virtual
This is a menu of commands that let you visit and create virtual
folders.
@item nil
If nil appears in the list, it should appear exactly once.  All
menus after nil in the list will be displayed flushright in
the menu bar.
@end table

@node Faces, Using the Mouse, Menus, Top
@chapter Faces

VM uses Emacs faces to emphasize text in the folder and summary
buffers.  Instead of defining VM specific faces, VM's face usage
is controlled by customization variables that can point to faces.
This allows you to use standard Emacs faces, or to create your
own.  So when you want to change which face is used, write code
like this:

@example
(setq vm-summary-highlight-face 'bold-italic)
@end example

@noindent In the summary buffer, VM displays the summary entry for the
current message using the face specified by the
@code{vm-summary-highlight-face} variable.  The value of this variable
should be a symbol that names a face, or nil which means don't display
the summary entry of the current message in a special way.

@vindex vm-mouse-track-summary
The variable @code{vm-mouse-track-summary} controls whether summary
entries are highlighted when the mouse pointer passes over
them.  The highlighting is done using the standard Emacs
@code{highlight} face.

@vindex vm-highlighted-header-regexp
@vindex vm-highlighted-header-face
In the folder buffer, the header contents of headers matched by
the @code{vm-highlighted-header-regexp} variable are displayed using
the face named by @code{vm-highlighted-header-face}.  This variable is
ignored under XEmacs if @code{vm-use-lucid-highlighting} is non-@code{nil}.
The XEmacs @code{highlight-headers} package is used instead.  See the
documentation for the function @code{highlight-headers} to find out
how to customize header highlighting using this package.

@vindex vm-highlight-url-face
@vindex vm-url-search-limit
URL's that occur in message bodies are displayed using the face
named by @code{vm-highlight-url-face}.  Searching for URLs in a
large message can take a long time.  Since URLs often occur near
the beginning and near the end of messages, VM offers a way to
search just those parts of a message for URLs.  The variable
@code{vm-url-search-limit} specifies how much of a message to search.
If @code{vm-url-search-limit} has a positive numeric value @var{N}, VM 
will search the first @math{@var{N} / 2} characters and the last
@math{@var{N} / 2} characters in the message for URLs.

@vindex vm-mime-button-face
The face named by @code{vm-mime-button-face} is used to display the
textual buttons that trigger the display of MIME objects.

@node Using the Mouse, Hooks, Faces, Top
@chapter Using the Mouse

VM uses the following layout for the mouse buttons in the folder
and summary buffers.

@table @asis
@item button-1 (left button usually)
Unchanged.
@item button-2 (middle button usually)
Activate.  If you click on a summary entry, that message will be
selected and become the current message.  If you click on a
highlighted URL in the body of a message, that URL will be sent
to the browser specified by @code{vm-url-browser}.
@item button-3 (right button usually)
Context Menu.  If the mouse pointer is over the contents of the
From header, button-3 pops up a menu of actions that can be taken 
using the author of the message as a parameter.  For instance,
you may want to create a virtual folder containing all the
messages in the current folder written by this author.  If the
mouse pointer is over the contents of the Subject header, a menu
of actions to be performed on the current message's subject is
produced.  If button-3 is clicked over a highlighted URL, a menu
of Web browsers is produced.  Otherwise the normal VM mode
specific menu is produced.
@end table

In mail composition buffers only mouse button-3 is affected.
Context sensitive menus are produced when that button is clicked.

@node Hooks, Bugs, Using the Mouse, Top
@chapter Hooks

VM has many hook variables that allow you to run functions when
certain events occur.  Here is a list of the hooks and when they
are run.  (If you don't write Emacs-Lisp programs you 
can skip this chapter.)

@table @code
@vindex vm-select-new-message-hook
@item vm-select-new-message-hook
List of hook functions called every time a message with the ``new''
attribute is made to be the current message.  When the hooks are run, the
current buffer will be the folder containing the message and the
start and end of the message will be bracketed by (point-min) and
(point-max).

@item vm-select-unread-message-hook
@vindex vm-select-unread-message-hook
List of hook functions called every time a message with the ``unread''
attribute is made to be the current message.  When the hooks are run, the
current buffer will be the folder containing the message and the
start and end of the message will be bracketed by (point-min) and
(point-max).

@item vm-select-message-hook
@vindex vm-select-message-hook
List of hook functions called every time a message
is made to be the current message.  When the hooks are run, the
current buffer will be the folder containing the message and the
start and end of the message will be bracketed by (point-min) and
(point-max).

@item vm-arrived-message-hook
@vindex vm-arrived-message-hook
List of hook functions called once for each message gathered from the
system mail spool, or from another folder with @code{vm-get-new-mail},
or from a digest with @code{vm-burst-digest}.  When the hooks are run,
the current buffer will be the folder containing the message and the
start and end of the message will be bracketed by (point-min) and
(point-max).

@item vm-spooled-mail-waiting-hook
@vindex vm-spooled-mail-waiting-hook
List of functions called when VM first notices mail is spooled
for a folder.  The folder buffer will be current when the hooks are
run.

@item vm-arrived-messages-hook
@vindex vm-arrived-messages-hook
List of hook functions called after VM has gathered a group
of messages from the system mail spool, or from another
folder with @code{vm-get-new-mail}, or from a digest with
@code{vm-burst-digest}.  When the hooks are run, the new
messages will have already been added to the message list
but may not yet appear in the summary.  When the hooks are
run the current buffer will be the folder containing the
messages.

@item vm-reply-hook
@vindex vm-reply-hook
List of hook functions to be run after a Mail mode composition
buffer has been created for a reply.  VM runs this hook and then
runs @code{vm-mail-mode-hook} before leaving you in the Mail
mode buffer.

@item vm-forward-message-hook
@vindex vm-forward-message-hook
List of hook functions to be run after a Mail mode
composition buffer has been created to forward a message.  VM
runs this hook and then runs @code{vm-mail-mode-hook} before leaving the
user in the Mail mode buffer.

@item vm-resend-bounced-message-hook
@vindex vm-resend-bounced-message-hook
List of hook functions to be run after a Mail mode
composition buffer has been created to resend a bounced message.
VM runs this hook and then runs @code{vm-mail-mode-hook} before leaving
you in the Mail mode buffer.

@item vm-resend-message-hook
@vindex vm-resend-message-hook
List of hook functions to be run after a Mail mode composition
buffer has been created to resend a message.  VM runs this hook
and then runs @code{vm-mail-mode-hook} before leaving you in
the Mail mode buffer.

@item vm-send-digest-hook
@vindex vm-send-digest-hook
List of hook functions to be run after a Mail mode composition
buffer has been created to send a digest.  VM runs this hook and
then runs @code{m-mail-mode-hook} before leaving you in the Mail
mode buffer.

@item vm-mail-hook
@vindex vm-mail-hook
List of hook functions to be run after a Mail mode
composition buffer has been created to send a non specialized
message, i.e. a message that is not a reply, forward, digest,
etc.  VM runs this hook and then runs @code{vm-mail-mode-hook} before
leaving you in the Mail mode buffer.

@item vm-summary-update-hook
@vindex vm-summary-update-hook
List of hook functions called just after VM updates an existing
entry in a folder summary buffer.

@item vm-summary-redo-hook
@vindex vm-summary-redo-hook
List of hook functions called just after VM adds or deletes
entries from a folder summary buffer.

@item vm-visit-folder-hook
@vindex vm-visit-folder-hook
List of hook functions called just after VM visits a folder.
It doesn't matter if the folder buffer already exists, this hook
is run each time @code{vm} or @code{vm-visit-folder} is called interactively.
It is @emph{not} run after @code{vm-mode} is called.

@item vm-retrieved-spooled-mail-hook
@vindex vm-retrieved-spooled-mail-hook
List of hook functions called just after VM has retrieved
a group of messages from your system mailbox(es).  When these
hooks are run, the messages have been added to the folder buffer
but not the message list or summary.  When the hooks are run, the
current buffer will be the folder where the messages were
incorporated.

@item vm-edit-message-hook
@vindex vm-edit-message-hook
List of hook functions to be run just before a message is edited.
This is the last thing @code{vm-edit-message} does before leaving you
in the edit buffer.

@item vm-mail-mode-hook
@vindex vm-mail-mode-hook
List of hook functions to be run after a Mail mode
composition buffer has been created.  This is the last thing VM
does before leaving you in the Mail mode buffer.

@item vm-mode-hook
@vindex vm-mode-hook
List of hook functions to run when a buffer enters @code{vm-mode}.
These hook functions should generally be used to set key bindings
and local variables.

@item vm-mode-hooks
@vindex vm-mode-hooks
Old name for @code{vm-mode-hook}.
Supported for backward compatibility.
You should use the new name.

@item vm-summary-mode-hook
@vindex vm-summary-mode-hook
List of hook functions to run when a VM summary buffer is created.
The current buffer will be that buffer when the hooks are run.

@item vm-summary-mode-hooks
@vindex vm-summary-mode-hooks
Old name for @code{vm-summary-mode-hook}.
Supported for backward compatibility.
You should use the new name.

@item vm-virtual-mode-hook
@vindex vm-virtual-mode-hook
List of hook functions to run when a VM virtual folder buffer is created.
The current buffer will be that buffer when the hooks are run.

@item vm-presentation-mode-hook
@vindex vm-presentation-mode-hook
List of hook functions to run when a VM presentation buffer is
created.  The current buffer will be the new presentation buffer
when the hooks are run.  Presentation buffers are used to display
messages when some type of decoding must be done to the message
to make it presentable.  E.g. MIME decoding.

@item vm-quit-hook
@vindex vm-quit-hook
List of hook functions to run when you quit VM.
This applies to any VM quit command.

@item vm-summary-pointer-update-hook
@vindex vm-summary-pointer-update-hook
List of hook functions to run when the VM summary pointer is updated.
When the hooks are run, the current buffer will be the summary buffer.

@item vm-display-buffer-hook
@vindex vm-display-buffer-hook
List of hook functions that are run every time VM wants to
display a buffer.  When the hooks are run, the current buffer will
be the buffer that VM wants to display.  The hooks are expected
to select a window and VM will display the buffer in that
window.

If you use display hooks, you should not use VM's built-in window
configuration system as the result is likely to be confusing.

@item vm-undisplay-buffer-hook
@vindex vm-undisplay-buffer-hook
List of hook functions that are run every time VM wants to
remove a buffer from the display.  When the hooks are run, the
current buffer will be the buffer that VM wants to disappear.
The hooks are expected to do the work of removing the buffer from
the display.  The hook functions should not kill the buffer.

If you use undisplay hooks, you should not use VM's built-in
window configuration system as the result is likely to be
confusing.

@item vm-iconify-frame-hook
@vindex vm-iconify-frame-hook
List of hook functions that are run whenever VM iconifies a frame.

@item vm-menu-setup-hook
@vindex vm-menu-setup-hook
List of hook functions that are run just after all menus are initialized.

@item vm-mime-display-function
@vindex vm-mime-display-function
If non-@code{nil}, this should name a function to be called inside 
@code{vm-decode-mime-message} to do the MIME display of the current
message.  The function is called with no arguments, and at the
time of the call the current buffer will be the @dfn{presentation
buffer} for the folder, which is a temporary buffer that VM uses
for the display of MIME messages.  A copy of the current message
will be in the presentation buffer at that time.  The normal work
that @code{vm-decode-mime-message} would do is not done, because this
function is expected to subsume all of it.

@item vm-mail-send-hook
@vindex vm-mail-send-hook
List of hook functions to call just before sending a message.
The hooks are run after confirming that you want to send the
message (see @code{vm-confirm-mail-send} but before MIME encoding
and FCC processing.

@item mail-yank-hooks
@vindex mail-yank-hooks
Hooks called after a message is yanked into a mail composition buffer.

(This hook is deprecated, you should use mail-citation-hook instead.)

The value of this hook is a list of functions to be run.  Each
hook function can find the newly yanked message between point
and mark.  Each hook function should return with point and mark
around the yanked message.

See the documentation for @code{vm-yank-message} to see when VM will run
these hooks.

@item mail-citation-hook
@vindex mail-citation-hook
Hook for modifying a citation just inserted in the mail buffer.
Each hook function can find the citation between (point) and (mark t).
And each hook function should leave point and mark around the citation
text as modified.

If this hook is entirely empty, i.e. @code{nil}, a default action is taken
instead of no action.
@end table

@node Bugs, History and Administration, Hooks, Top
@chapter Reporting Bugs
@cindex bug reports

VM has a sophisticated bug reporting system in order to provide the
VM maintainers with adequate information about the state of VM when the
error situation occurred.  However, it is still important for the
users to give as full an explanation of the problem as possible.
@xref{Bugs,,,emacs, the GNU Emacs Manual}.


@findex vm-submit-bug-report
The command @code{M-x vm-submit-bug-report} should be invoked from the
VM folder buffer in which a problem is encountered.  This creates a
mail buffer with information about the state of VM pre-filled.  Insert
suitable text to explain the problem and send the bug-report message.

@findex vm-pop-start-bug-report
@findex vm-pop-submit-bug-report
@findex vm-imap-start-bug-report
@findex vm-imap-submit-bug-report
For mail server-associated problems dealing with POP/IMAP spool files
or POP/IMAP folders, the cause of the problem might be in the
interaction with the mail server.  To identify the cause, it may be
necessary for the VM maintainer to look at the server interactions
during the problem occurrence.  To capture the server interactions,
run @code{vm-pop-start-bug-report}/@code{vm-imap-start-bug-report}
before the problem occurrence and @code{vm-pop-submit-bug-report}
/@code{vm-imap-submit-bug-report} after the problem occurrence.  All
the server interactions during the interval are captured and
automatically included in the bug-report.



@node History and Administration, Internals, Bugs, Top
@chapter History and Administration

@cindex Kyle Jones
VM was developed by Kyle Jones, starting in early 1989.  The first
public release of VM was version 4.10, released in June of that year.
The original development environment was GNU Emacs 18.52.  

@cindex Wonderworks
The copyright for the code was retained by Kyle Jones.  Hence, the
package was never included GNU releases, which only contain code
copyrighted by the Free Software Foundation.  However, Lucid/XEmacs
shipped VM starting with version 19.9.  The other users obtained VM from
the Wonderworks web site, which hosted Kyle Jones's work.  The home page
of VM at this site is @uref{http://www.wonderworks.com/vm}.

The last version released by Kyle Jones was 7.19, in September 2004,
which can be found on the Wonderworks web site and its mirror sites.

@cindex Robert Widhopf-Fenk
After this release, Robert Widhopf-Fenk picked up the maintenance of
VM, by releasing a series of patches under a separate distribution.  He
also acquired a number of add-on's contributed by various developers,
including himself, and included them in his distribution.  Kyle Jones
agreed to hand over the maintenance of VM to Robert Fenk in February,
2007.  Further releases were made by Robert Fenk under the @code{8.0.x}
series. 

@cindex Savannah
All these releases are available from the new project page of VM hosted
by Savannah, at the URL
@uref{http://savannah.nongnu.org/projects/viewmail/}. 
According to the project page, ``this site exists to continue VM development
after version 7.19 as a community project.''

@cindex Ulrich Müller
@cindex Uday S Reddy
Currently, VM is maintained by a ``VM Development Team,'' consisting of
Robert Widhopf-Fenk, Ulrich Müller and Uday S Reddy.  Other potential members
are warmly welcomed.  Robert Fenk has been inactive since November, 2008
but he continues to be an official member of the team.  The new
releases made by the team are numbered @code{8.1.0} and up.

@cindex Launchpad
The project code base is maintained at the Launchpad web site
@uref{http://launchpad.net/vm}.  The ``VM Development Team'' can be
reached here using the email address @code{vm@@launchpad.net}.

@unnumberedsubsec Savannah project site

The changes made in each of the releases is described in the @samp{NEWS}
file, which can be found in the source code repository.  The changes made
in versions up to 7.19 are described in the @samp{CHANGES} file.

The @code{Download} link on the Savannah project page, takes you to the
downloads area where all the recent releases are available.  Under the
@code{Source Code} menu, the @code{Browse Sources Repository} takes you
to the source files, which include, among others, the @samp{NEWS} and
@samp{CHANGES} files mentioned above.

If you have obtained VM through a secondary distribution that does not
include all the sources, you can browse and download the sources from
the @code{Source Code} menu.  The @code{Use Bazaar} entry in the menu
takes you to a page that lists various version of VM source code, and
gives instructions for downloading it via @samp{Bazaar} (@code{bzr}).

@unnumberedsubsec Technical support

VM has a dedicated usenet newsgroup @code{gnu.emacs.vm.info} in which
the developers and the active users participate.  This is the first port
of call for getting help with VM.  The archives of the newsgroup can be
found at the Google Groups site
@uref{http://groups.google.com/group/gnu.emacs.vm.info/topics}.

A second newsgroup @code{gnu.emacs.vm.bug} is dedicated to
discussing bugs or potential bugs in VM.  However, this is a place for
discussion, not reporting bugs or getting them fixed.  

The easiest way to report bugs that need fixing is to use the command
@code{M-x vm-submit-bug-report} within VM.  This prepares an email
message by including a state of your VM program which will allow the
developers to reproduce your problem.  (Potentially sensitive information
such as passwords are not included in this state.)  Please include a
detailed description of the problem and how it arose.  The developers
may need to ask you for further information or ask you to try
alternative approaches to narrow down the problem.

The best way to report bugs is via the Launchpad bug tacker.  See below.

@unnumberedsubsec Get Involved

VM is now supported and maintained by the user community.  So, as an
active user, your participation is key to keep the project going.  

Consider registering as a user of the Launchpad development site
@uref{http://launchpad.net/vm}.  This 
allows you to communicate with the developers and other users using a
private Launchpad email address.  In particular, you can contribute bug
reports and participate in the bug report discussions.

You can download the development versions of VM and act as an ``alpha''
tester.  This will allow you to shape the new developments and features
and make suggestions that will be valuable to the developers.

To download the development version, identify the ``branch'' that you
would like to download, and use Bazaar version control system with an
appropriate Launchpad URL.  For example, the command
@command{bzr get lp:vm} can be used to download the main development
branch. 

You can also make change to the branch you have downloaded, and submit
them to the developers for inclusion in the project.  The @code{README}
file in the distribution explains how to do this.  Alternatively, you
can create a separate branch in your own space on the Launchpad web
site, and submit your changes to that branch.  The developers can review
and merge your branch with the main development when your changes
are ready.

@unnumberedsubsec Contributors

Contributions to the code from the following members of the VM community
are gratefully acknowledged:

@itemize
@item Aidan Kehoe
@item Glenn <unknown last name>
@item Jens Gustedt
@item John J Foerch
@item Kevin Rogers
@item Kyle Jones
@item Rob Hodges
@item Robert Marshall
@item Robert P. Goldman
@item Katsumi Yamaoka
@item Julian Bradfield
@item Samuel Bronson
@item Brent Goodrick
@item Noah Friedman
@end itemize

Please let us know if any other contributors have been missed out.

@unnumberedsubsec Selected Releases of Kyle Jones
@itemize
@item Version 4.10, released in 1989.
@item Version 5.00, released in 1990.
@item Version 6.00, released 6 January, 1997.
@item Version 7.00, released 2 December, 2001.
@item Version 7.10, released 5 March, 2003.
@item Version 7.15, released 3 May, 2003.
@item Version 7.16, released 26 May, 2003.
@item Version 7.17, released 6 July, 2003.
@item Version 7.18, released 2 November, 2003.
@item Version 7.19, released 29 September, 2004.
@end itemize


@unnumberedsubsec Releases of Robert Widhopf-Fenk
@itemize
@item Version 8.0.0, released 31 May, 2007.
@item Version 8.0.1, released 29 June, 2007.
@item Version 8.0.2, released 25 July, 2007.
@item Version 8.0.3, released 15 August, 2007.
@item Version 8.0.4, released 2 November, 2007.
@item Version 8.0.5, released 3 November, 2007.
@item Version 8.0.6, released 2 January, 2008.
@item Version 8.0.7, released 5 January, 2008.
@item Version 8.0.8, released 11 February, 2008.
@item Version 8.0.9, released 20 February, 2008.
@item Version 8.0.10, released 22 June, 2008.
@item Version 8.0.11, released 11 August, 2008.
@item Version 8.0.12, released 5 November, 2008.
@item Version 8.0.13, released 29 November, 2009.
@item Version 8.0.14, released 16 December, 2009.
@end itemize


@unnumberedsubsec Releases of VM development team
@itemize
@item Version 8.1.0, released 21 March, 2010.
@item Version 8.1.1, planned for release on 10 April, 2010.
@end itemize





@node Internals, Concept Index, History and Administration, Top
@chapter VM Internals

This section gives a sketchy overview of the VM internals for the
developers/programmers. 

@menu
* Folder Internals::      Structure of the folders
* Message Internals::     Structure of the message data structure
@end menu

@node Folder Internals, Message Internals , , Internals
@section Folder Internals

By default, VM stores mail folders in the Unix @code{mbox} format,
described in the RFC 4155 specification of the Internet Engineering Task
Force.  In this format, the mail folder is a text file consisting of a
sequence of messages, with each message consisting of a series of
headers followed by a message body.  The beginning of each message is
delineated by a separator line starting with the string ``From '' and
the end by a blank line.  To make sure that such a line exists, VM adds
its own header line of the form ``From VM ...'' where the ``...''
records the time at which VM first saw the message.  The format of the
individual messages is as per the RFC 2822 specification, except that
Line-Feed characters may be used to delineate the end of lines in the
"Unix" format.  

In addition to the Unix mbox format (which is called the @code{From_}
format), VM also handles the MMDF format and two versions of the System
V Unix format and the Emacs Rmail's Babyl format.  The variable
@code{vm-folder-type} stores the type of the folder being used.

To every message, VM adds a header with the field name
``X-VM-v5-Data:'' and stores in it the information about the message it
wishes to remember between sessions.  This header line is used merely
for speed of processing.  If, for any reason, a VM folder is
corrupted, it is safe to delete all the X-VM-v5-Data headers, and VM
will generate them afresh when the folder is visited.

The first message of the VM folder file contains additional headers that
VM uses for remembering information between sessions.

@itemize
@item
X-VM-Bookmark. This header stores the position of the cursor, as a
message number, in effect when VM saved the folder.  Upon revisiting the
folder, VM attempts to put the cursor back at this position.
@item
X-VM-Last-Modified. 
@item
X-VM-Message-Order.  This header lists the order in which the messages
should be listed.
@item
X-VM-Labels. This header lists the message labels that have been used in
the folder.  
@item
X-VM-VHeader. This header lists the visible headers that should be
displayed in message listings.
@item
X-VM-Summary-Format.  This header stores the format string for the
summary lines.
@item
X-VM-POP-Retrieved. This header lists all the messages that have
been retrieved from POP servers together with the identifying
information for the POP servers.  VM refrains from retrieving these
messages again in future in order to avoid duplication.
@item
X-VM-IMAP-Retrieved. This header lists all the messages that have
been retrieved from IMAP servers together with the identifying
information for the IMAP servers.  VM refrains from retrieving these
messages again in future in order to avoid duplication.
@end itemize

Internal to Emacs, VM stores the folder as simply a text buffer.  However, it
remembers a variety of data about the message contents in the buffer
through internal variables.

@itemize
@item
@code{vm-message-list}.  A list of message data structures for all the
messages in the buffer.
@item
@code{vm-message-pointer}.  A sublist of vm-message-list starting from
the current message that the cursor is on.  So, the first element of
vm-message-pointer is the current message.
@item
@code{vm-last-message-pointer}.  Whenever the cursor is moved, the
previous value of vm-message-pointer is remembered in this variable.
@item
@code{vm-folder-type}.  The type of the current folder indicating how
the messages are stored: one of 'babyl, 'From_, 'BellFrom_,
'From_-with-Content-Length and 'mmdf.
@item
@code{vm-folder-access-method}.  The method for accessing the server
message store: 'pop for pop-folders as well as local folders retrieving
messages from POP servers, 'imap for imap-folders as well as local
folders retrieving messages from IMAP servers, and nil for all other folders.
@item
@code{vm-folder-access-data}.  A vector of data for accessing the server
message store.  The first two elements of the vector are the maildrop
specification for the mail server and a reference to the process
connecting to the mail server.  For the 'pop access method, that is all
there is.  But, for the 'imap access method, the vector has 9 other
entries detailing various pieces of data about the IMAP server.
@item
@code{vm-virtual-folder-definition}.  If the current folder is virtual,
then this variable holds the data constituting its definition.
@item
@code{vm-real-buffers}.  If the current folder is virtual, then this
variable is a list of all the real folder buffers involved in
constructing it.
@item
@code{vm-virtual-buffers}.  A list of all the virtual folder buffers
that the current buffer is involved in.
@end itemize



@node Message Internals, , Folder Internals, Internals
@section Message Internals

The message data structure is a vector containing various pieces of
data:

@itemize
@item
Location data.  This data about the location of the various parts of the
message in the folder buffer is calculated after a folder is loaded and
parsed. 

@itemize
@item
start.  The starting position of the message, at which a leading
separator line begins.
@item
headers. The position in the folder buffer where the headers of the message
start.
@item
vheaders.  The position in the folder buffer where the visible headers of the
message start.  (The headers are rearranged in such a way that all the
visible headers are towards the end of the headers region.)
@item
text. The position in the folder buffer where the text of the message starts.
@item
text-end. The position in the folder buffer where the text of the message ends
@item
end. The position in the folder buffer where the message ends.
@end itemize

@item
Soft data.  This vector contains other calculated data about the
message. 

@itemize
@item
number. The message number as an integer.
@item
padded number. The message number as a padded string.
@item
mark. Flag that indicates if the message has been marked (via
@code{vm-mark-message}).
@item
su-start. The position in the summary buffer where the summary line of
the message starts.
@item
su-end.  The position in the summary buffer where the summary line of
the message ends.
@item
real-message-sym.  If the message is in a virtual folder, then its
corresponding ``real message'' is the underlying message in another
folder which is described by a message data structure similar to the
current one.  The real message data structures are interned in an
obarray and represented by ``symbols''.  This field stores the symbol
representing the real message of the current message.
@item
reverse-link.  Link to the previous message in the message list.
@item
message-type.  A symbol indicating the type of the message according to
its folder type, one of @code{BellFrom_}, @code{From_} and
@code{From_-with-Content-Length}.
@item
message-id-number.  The unique id of the message.
@item
buffer. The folder buffer of the message.
@item
thread-indentation. Indentation level of the message in its message
thread.
@item
babyl-frob-flag.
@item
saved-virtual-attributes.  Saved attributes if the message switched from
unmirrored to mirrored.
@item
saved-virtual-mirror-data.  Saved mirror data, if the message was
switched from unmirrored to mirrored.
@item
virtual-summary.  Summary for unmirrored virtual message.
@item
mime-layout.  MIME layout information; types, ids, positions, etc of
all MIME entities.
@item
mime-encoded-header-flag. Flag that indicates if the headers of the
message are MIME encoded.
@item
su-summary-mouse-track-overlay.
@item
message-access-method.  The access-method to be used for the message,
inherited from its real folder.
@end itemize

@item
Attributes.  All the hard-wired message attributes are stored in this
vector.  They also get saved as part of the @code{X-VM-v5-Data} header
field when the folder is saved to disk.

@itemize
@item
new-flag.  Flag to indicate if the message is ``new''.
@item
unread-flag.  Flag to indicate if the message is unread.
@item
deleted-flag.  Flag to indicate if the message has been deleted.
@item
filed-flag.  Flag to indicate if the message has been filed.
@item
replied-flag.  Flag to indicate if the message has been replied to.
@item
written-flag.  Flag to indicate if the message has been saved.
@item
forwarded-flag.  Flag to indicate if the message has been forwarded.
@item
edited-flag.  Flag to indicate if the message has been edited.
@item
redistributed-flag.  Flag to indicate if the message has been
redistributed. 
@end itemize

@item
Cached Data.  The data that is cached for the message and stored in the
folder on disk as the @code{X-VM-v5-Data} header field.  Even though
this vector is only supposed to have data that can be calculated from
the message itself, the fields pop-uidl, imap-uid and imap-uid-validity
form an exception.  They are really hard data that cannot be calculated
from anything else.

@itemize
@item
byte-count.  The size of the message in bytes.
@item
weekday, monthday, month, year, hour, zone.  Data indicating the date of
the message.
@item 
full-name.  The full name of the author of the message.
@item
from.  The email address of the author of the message.
@item
message-id.  The unique id of the message.
@item
line-count. The number of lines in the message.
@item
subject.  The subject string of the message.
@item
vheaders-regexp.  A regular expression that can be used to find the
start of the already ordered headers.
@item
to.  Addresses of the recipients of the message in a comma separated
string.
@item
to-names. The full names of the recipients in a comma separated string.
Addresses are used if full names are not available.
@item
month-number.  Numeric month of the sent date.
@item
sortable-datestring.  Date string of the sent date for sorting purposes.
@item
sortable-subject.  The subject string for sorting purposes.  (Prefixes
such as ``re:'' are removed.)
@item
summary.  The summary string for the message.
@item
parent.  The parent of the message in the message thread, which is
another message data structure.
@item
references.  Message IDs listed in the References header of the message.
@item
headers-to-be-retrieved.  Flag that indicates whether the headers of the
message have not been retrieved from the mail server (for POP or IMAP
folders).
@item
body-to-be-retrieved.  Flag that indicates whether the body of the
message has not been retrieved from the mail server.
@item
pop-uidl.  The UIDL id of the message on the POP server.
@item
imap-uid.  The UID of the message on the IMAP server.
@item
imap-uid-validity.  The UID-VALIDITY value of the message on the IMAP
server. 
@item
spam-score.  The spam score of the message.
@end itemize

@item
Mirror Data.  Extra data shared by virtual messages if vm-virtual-mirror
is non-nil.

@itemize
@item 
edit-buffer.  If the message is being edited, this is the buffer being
used.
@item
virtual-messages.  List of virtual messages mirroring the current real
message. 
@item
stuff-flag.  Flag to indicates if the attribute changes have been
``stuffed'' into the folder buffer.
@item
labels.  List of labels attached to the message.
@item
label-string.  The string of labels attached to the message.
@item
attribute-modflag.  Flag to indicate if the attributes of the message
have been modified since the last save.
@end itemize

@end itemize



@node Concept Index, Key Index, Internals, Top
@unnumbered Concept Index
@printindex cp

@node Key Index, Command Index, Concept Index, Top
@unnumbered Key Index
@printindex ky

@node Command Index, Variable Index, Key Index, Top
@unnumbered Command Index
@printindex fn

@node Variable Index, License, Command Index, Top
@unnumbered Variable Index
@printindex vr

@node License,, Variable Index, Top
@unnumbered GNU GENERAL PUBLIC LICENSE
@center Version 2, June 1991

@display
Copyright @copyright{} 1989, 1991 Free Software Foundation, Inc.
675 Mass Ave, Cambridge, MA 02139, USA

Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
@end display

@unnumberedsec Preamble

  The licenses for most software are designed to take away your
freedom to share and change it.  By contrast, the GNU General Public
License is intended to guarantee your freedom to share and change free
software---to make sure the software is free for all its users.  This
General Public License applies to most of the Free Software
Foundation's software and to any other program whose authors commit to
using it.  (Some other Free Software Foundation software is covered by
the GNU Library General Public License instead.)  You can apply it to
your programs, too.

  When we speak of free software, we are referring to freedom, not
price.  Our General Public Licenses are designed to make sure that you
have the freedom to distribute copies of free software (and charge for
this service if you wish), that you receive source code or can get it
if you want it, that you can change the software or use pieces of it
in new free programs; and that you know you can do these things.

  To protect your rights, we need to make restrictions that forbid
anyone to deny you these rights or to ask you to surrender the rights.
These restrictions translate to certain responsibilities for you if you
distribute copies of the software, or if you modify it.

  For example, if you distribute copies of such a program, whether
gratis or for a fee, you must give the recipients all the rights that
you have.  You must make sure that they, too, receive or can get the
source code.  And you must show them these terms so they know their
rights.

  We protect your rights with two steps: (1) copyright the software, and
(2) offer you this license which gives you legal permission to copy,
distribute and/or modify the software.

  Also, for each author's protection and ours, we want to make certain
that everyone understands that there is no warranty for this free
software.  If the software is modified by someone else and passed on, we
want its recipients to know that what they have is not the original, so
that any problems introduced by others will not reflect on the original
authors' reputations.

  Finally, any free program is threatened constantly by software
patents.  We wish to avoid the danger that redistributors of a free
program will individually obtain patent licenses, in effect making the
program proprietary.  To prevent this, we have made it clear that any
patent must be licensed for everyone's free use or not licensed at all.

  The precise terms and conditions for copying, distribution and
modification follow.

@iftex
@unnumberedsec TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
@end iftex
@ifnottex
@center TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
@end ifnottex

@enumerate 0
@item
This License applies to any program or other work which contains
a notice placed by the copyright holder saying it may be distributed
under the terms of this General Public License.  The ``Program'', below,
refers to any such program or work, and a ``work based on the Program''
means either the Program or any derivative work under copyright law:
that is to say, a work containing the Program or a portion of it,
either verbatim or with modifications and/or translated into another
language.  (Hereinafter, translation is included without limitation in
the term ``modification''.)  Each licensee is addressed as ``you''.

Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope.  The act of
running the Program is not restricted, and the output from the Program
is covered only if its contents constitute a work based on the
Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does.

@item
You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any warranty;
and give any other recipients of the Program a copy of this License
along with the Program.

You may charge a fee for the physical act of transferring a copy, and
you may at your option offer warranty protection in exchange for a fee.

@item
You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:

@enumerate a
@item
You must cause the modified files to carry prominent notices
stating that you changed the files and the date of any change.

@item
You must cause any work that you distribute or publish, that in
whole or in part contains or is derived from the Program or any
part thereof, to be licensed as a whole at no charge to all third
parties under the terms of this License.

@item
If the modified program normally reads commands interactively
when run, you must cause it, when started running for such
interactive use in the most ordinary way, to print or display an
announcement including an appropriate copyright notice and a
notice that there is no warranty (or else, saying that you provide
a warranty) and that users may redistribute the program under
these conditions, and telling the user how to view a copy of this
License.  (Exception: if the Program itself is interactive but
does not normally print such an announcement, your work based on
the Program is not required to print an announcement.)
@end enumerate

These requirements apply to the modified work as a whole.  If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works.  But when you
distribute the same sections as part of a whole which is a work based
on the Program, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote it.

Thus, it is not the intent of this section to claim rights or contest
your rights to work written entirely by you; rather, the intent is to
exercise the right to control the distribution of derivative or
collective works based on the Program.

In addition, mere aggregation of another work not based on the Program
with the Program (or with a work based on the Program) on a volume of
a storage or distribution medium does not bring the other work under
the scope of this License.

@item
You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:

@enumerate a
@item
Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections
1 and 2 above on a medium customarily used for software interchange; or,

@item
Accompany it with a written offer, valid for at least three
years, to give any third party, for a charge no more than your
cost of physically performing source distribution, a complete
machine-readable copy of the corresponding source code, to be
distributed under the terms of Sections 1 and 2 above on a medium
customarily used for software interchange; or,

@item
Accompany it with the information you received as to the offer
to distribute corresponding source code.  (This alternative is
allowed only for noncommercial distribution and only if you
received the program in object code or executable form with such
an offer, in accord with Subsection b above.)
@end enumerate

The source code for a work means the preferred form of the work for
making modifications to it.  For an executable work, complete source
code means all the source code for all modules it contains, plus any
associated interface definition files, plus the scripts used to
control compilation and installation of the executable.  However, as a
special exception, the source code distributed need not include
anything that is normally distributed (in either source or binary
form) with the major components (compiler, kernel, and so on) of the
operating system on which the executable runs, unless that component
itself accompanies the executable.

If distribution of executable or object code is made by offering
access to copy from a designated place, then offering equivalent
access to copy the source code from the same place counts as
distribution of the source code, even though third parties are not
compelled to copy the source along with the object code.

@item
You may not copy, modify, sublicense, or distribute the Program
except as expressly provided under this License.  Any attempt
otherwise to copy, modify, sublicense or distribute the Program is
void, and will automatically terminate your rights under this License.
However, parties who have received copies, or rights, from you under
this License will not have their licenses terminated so long as such
parties remain in full compliance.

@item
You are not required to accept this License, since you have not
signed it.  However, nothing else grants you permission to modify or
distribute the Program or its derivative works.  These actions are
prohibited by law if you do not accept this License.  Therefore, by
modifying or distributing the Program (or any work based on the
Program), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Program or works based on it.

@item
Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions.  You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License.

@item
If, as a consequence of a court judgment or allegation of patent
infringement or for any other reason (not limited to patent issues),
conditions are imposed on you (whether by court order, agreement or
otherwise) that contradict the conditions of this License, they do not
excuse you from the conditions of this License.  If you cannot
distribute so as to satisfy simultaneously your obligations under this
License and any other pertinent obligations, then as a consequence you
may not distribute the Program at all.  For example, if a patent
license would not permit royalty-free redistribution of the Program by
all those who receive copies directly or indirectly through you, then
the only way you could satisfy both it and this License would be to
refrain entirely from distribution of the Program.

If any portion of this section is held invalid or unenforceable under
any particular circumstance, the balance of the section is intended to
apply and the section as a whole is intended to apply in other
circumstances.

It is not the purpose of this section to induce you to infringe any
patents or other property right claims or to contest validity of any
such claims; this section has the sole purpose of protecting the
integrity of the free software distribution system, which is
implemented by public license practices.  Many people have made
generous contributions to the wide range of software distributed
through that system in reliance on consistent application of that
system; it is up to the author/donor to decide if he or she is willing
to distribute software through any other system and a licensee cannot
impose that choice.

This section is intended to make thoroughly clear what is believed to
be a consequence of the rest of this License.

@item
If the distribution and/or use of the Program is restricted in
certain countries either by patents or by copyrighted interfaces, the
original copyright holder who places the Program under this License
may add an explicit geographical distribution limitation excluding
those countries, so that distribution is permitted only in or among
countries not thus excluded.  In such case, this License incorporates
the limitation as if written in the body of this License.

@item
The Free Software Foundation may publish revised and/or new versions
of the General Public License from time to time.  Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.

Each version is given a distinguishing version number.  If the Program
specifies a version number of this License which applies to it and ``any
later version'', you have the option of following the terms and conditions
either of that version or of any later version published by the Free
Software Foundation.  If the Program does not specify a version number of
this License, you may choose any version ever published by the Free Software
Foundation.

@item
If you wish to incorporate parts of the Program into other free
programs whose distribution conditions are different, write to the author
to ask for permission.  For software which is copyrighted by the Free
Software Foundation, write to the Free Software Foundation; we sometimes
make exceptions for this.  Our decision will be guided by the two goals
of preserving the free status of all derivatives of our free software and
of promoting the sharing and reuse of software generally.

@iftex
@heading NO WARRANTY
@end iftex
@ifnottex
@center NO WARRANTY
@end ifnottex

@item
BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW@.  EXCEPT WHEN
OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
PROVIDE THE PROGRAM ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED
OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE@.  THE ENTIRE RISK AS
TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU@.  SHOULD THE
PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING,
REPAIR OR CORRECTION.

@item
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR
REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES,
INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING
OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED
TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY
YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER
PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE
POSSIBILITY OF SUCH DAMAGES.
@end enumerate

@iftex
@heading END OF TERMS AND CONDITIONS
@end iftex
@ifnottex
@center END OF TERMS AND CONDITIONS
@end ifnottex

@page
@unnumberedsec How to Apply These Terms to Your New Programs

  If you develop a new program, and you want it to be of the greatest
possible use to the public, the best way to achieve this is to make it
free software which everyone can redistribute and change under these terms.

  To do so, attach the following notices to the program.  It is safest
to attach them to the start of each source file to most effectively
convey the exclusion of warranty; and each file should have at least
the ``copyright'' line and a pointer to where the full notice is found.

@smallexample
@var{one line to give the program's name and an idea of what it does.}
Copyright (C) 19@var{yy}  @var{name of author}

This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE@.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
@end smallexample

Also add information on how to contact you by electronic and paper mail.

If the program is interactive, make it output a short notice like this
when it starts in an interactive mode:

@smallexample
Gnomovision version 69, Copyright (C) 19@var{yy} @var{name of author}
Gnomovision comes with ABSOLUTELY NO WARRANTY; for details
type `show w'.  This is free software, and you are welcome
to redistribute it under certain conditions; type `show c' 
for details.
@end smallexample

The hypothetical commands @samp{show w} and @samp{show c} should show
the appropriate parts of the General Public License.  Of course, the
commands you use may be called something other than @samp{show w} and
@samp{show c}; they could even be mouse-clicks or menu items---whatever
suits your program.

You should also get your employer (if you work as a programmer) or your
school, if any, to sign a ``copyright disclaimer'' for the program, if
necessary.  Here is a sample; alter the names:

@smallexample
@group
Yoyodyne, Inc., hereby disclaims all copyright
interest in the program `Gnomovision'
(which makes passes at compilers) written 
by James Hacker.

@var{signature of Ty Coon}, 1 April 1989
Ty Coon, President of Vice
@end group
@end smallexample

This General Public License does not permit incorporating your program into
proprietary programs.  If your program is a subroutine library, you may
consider it more useful to permit linking proprietary applications with the
library.  If this is what you want to do, use the GNU Library General
Public License instead of this License.

@summarycontents
@contents
@bye