Important Concepts

Fundamental af Commands

Other Ways to Deal With Messages

Commands for Mail Folder Management

How to Customise Af

Glossary and Indexes

Distribution

Af is free software; this means that everyone is free to use it and free to redistribute it on certain conditions. Af is not in the public domain; it is copyrighted and there are restrictions on its distribution, but these restrictions are designed to permit everything that a good cooperating citizen would want to do. What is not allowed is to try to prevent others from further sharing any version of af that they might get from you. The precise conditions are found in the GNU General Public License that comes with af and also appears following this section.

The simplest way to get a copy of af is from someone else who has it. You need not ask for permission to do so, or tell anyone you have done so; just copy it. If you have access to the Internet, you can get the latest version, or a patch to upgrade to the latest version, from the Af Home Page on the World Wide Web `http://www.thing.demon.co.uk/af', or by anonymous FTP from `ftp://ftp.csv.warwick.ac.uk/pub/mail/af'.

GNU GENERAL PUBLIC LICENSE

Version 2, June 1991

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.

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.

1. 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.

2. 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.

3. 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:
   1. You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change.
   2. 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.
   3. 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.)

   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.

4. 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:
   1. 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,
   2. 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,
   3. 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.)

   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.

5. 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.
6. 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.
7. 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.
8. If, as a consequence of a court judgement 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.

9. 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.
10. 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.

11. 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.
12. 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.
13. 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.

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.

one line to give the program's name and an idea of what it does.
Copyright (C) 19yy  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.

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:

Gnomovision version 69, Copyright (C) 19yy 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.

The hypothetical commands `show w' and `show c' should show the appropriate parts of the General Public License. Of course, the commands you use may be called something other than `show w' and `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:

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

signature of Ty Coon, 1 April 1989
Ty Coon, President of Vice

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.

Introduction

You are reading about af, an advanced, self-documenting, customisable, real-time display-oriented mail reader and composer. If you think that sounds rather like the Emacs editor, you're correct; af is designed to follow the Emacs paradigm as closely as possible.

We say that af is display-oriented because normally the messages being read are visible on the screen and are updated automatically as you type your commands. See section Display.

We call it a real-time mail reader because the display is updated very frequently, usually after each character or pair of characters you type. This minimises the amount of information you must keep in your head as you process your mail. See section Basic af Commands.

We call af advanced because it provides facilities that go beyond simple reading of messages: viewing multiple mail folders at once, searching and sorting mailboxes, and dealing with groups of messages in one operation.

Similarly, we have tried to take care that af complies with any relevant standards: it conforms closely to RFC822, the Internet Standard for Mail Messages, and will work with such standard Mail Transfer Agents as UUCP, sendmail, and MMDF; and supports the POP3 protocol for reading mail from a remote server. Af also has good support for reading MIME mail, and limited support for composingl MIME mail, although this is still being improved. (1).

Self-documenting means that help about af is available from within af itself. You can find out what any command does, or find all the commands that are relevant to a topic. See section Help.

Customisable means that you can change the definitions of af commands in little ways. For example, if you prefer to include the text of the original message into the text of a reply to that message, then you can tell af to do so. Another sort of customisation is rearrangement of the command set. For example, if you prefer the four basic cursor motion commands (up, down, left and right) on keys in a diamond pattern on the keyboard, you can rebind the keys that way. See section Customisation.

Acknowledgements

Several people have contributed to af's development. Here's my chance to thank them for their efforts, and for an attempt at a moment of glory.

• Malc Arnold (malc@thing.demon.co.uk) conceived af, did the initial design work, wrote almost all of the code, and argued passionately in defence of the interface.
• Kay Dekker (kay@vide.coventry.ac.uk) did a vast amount of design work, discussing and arguing about features I should support and how the interface should work. More af features than I can readily recall were originally Kay's idea. He also wrote the manual pages and the first draft of the user manual, and showed great patience when struggling with some of the less stable versions of af.
• Ian "Vato" Dickinson (idickins@fore.com) actually suggested turning af into a full-featured mailer, and had a wealth of suggestions for features, some of which I've actually implemented. He also wrote the first draught of the POP3 and SMTP support code.
• Andrew "Vic" Fry (vic@pootug.demon.co.uk) has contributed insights about the af interface, as well as enough encouragement and coffee to keep me going. He has also ported and tested several versions of af.
• Justin Murdock (justin@vide.coventry.ac.uk), Dave Berry (daveb@harlequin.co.uk), Iain Bowen (alaric@harlech.demon.co.uk), and Onno van der Linden (onno@simplex.nl) have reported many bugs and suggestions for improving af.
• Emma Kemm (emma@thing.demon.co.uk) and Iain Bowen (alaric@harlech.demon.co.uk) proofread the manuals and supplied more much-needed moral support.
• Richard Stallman (rms@gnu.mit.edu) wrote the eleventh edition of the GNU Emacs manual, from which many parts of this manual have been derived. Thanks to the Free Software Foundation for permission to distribute this manual as a "derived work".

  Note that this does not imply any endorsement of af by the Free Software Foundation.

Bugs

Sometimes you will run into a bug in af. Although I can't promise that I can or will fix the bug, and I may not even agree that it is a bug, I want to hear about problems you encounter. Often I will agree that they are bugs and want to fix them.

If af should ever crash (ie fall over with an operating system error message), or exit back to the shell without warning, then it is certainly a bug. Commands doing the "wrong thing" are also bugs, but you should check the manual and help entry for the command carefully to be sure that the command isn't doing what it is supposed to.

If you think that you've found a bug, it is important to report it, and to provide enough information with the report to be useful. The most useful kind of bug report (except the fix for it) is an exact description of what commands you type, from entering af until the bug manifests. It may also help if you could include a small folder which triggers the bug, since many af bugs can be related to the data in the current folders. It is also important to tell me which version of af you are using, and what machine it is running on. You can find out which version of af you are using by typing M-x af-version RET.

At present, I am also interested in any feedback on af. If you have comments, praise, criticisms or complaints, then I'd like to hear it.

The best way to send bug reports or comments is to mail them electronically to the af maintainer at `af-bug@csv.warwick.ac.uk'. I don't promise to fix the problem; but if I agree that its a bug then I'll most likely want to fix it. And remember that the clearer your bug report is, the more likely it is that the bug will get fixed quickly, or indeed at all.

Future Developments

Af has now been released, but this doesn't mean that it is complete. There are several major enhancements that af could benefit from, some of which are loosely planned to be done at some time in the future.

• Internal support for composing non-textual MIME mail.
• Commands to allow you to use PGP conveniently with af. PGP stands for Pretty Good Privacy; it is a public-key encryption package written by Phil Zimmerman.
• Support for internationalising af, so that messages are displayed in the user's native language. I would need volunteers to actually do the translations.
• Major enhancements to the Afl language, allowing people to write their own extensions to af.
• An X interface with pull-down menus, scroll bars and mouse support.
• An additional major mode for reading news, similar to the mail mode but carefully tailored for convenient news-reading.
• Ports to other operating systems, such as MS-DOS or windows.

If you have any strong preferences about the direction af should take in the future, or suggestions of other enhancements you think I might like to consider, then please mail them to me at `af-bug@csv.warwick.ac.uk'.

1. An Introduction to Electronic Mail

Electronic mail (often known as e-mail, or simply mail), is the exchange of messages between people using computer networks. In this manual, when we say mail, we mean electronic mail. A mail message is a message which has been or will be sent via electronic mail. Again, when we say message in this manual, we are referring to a mail message.

When a mail message is sent, it is transmitted from one computer to another until it reaches the correct one. A message can be addressed to more than one person, in which case a separate copy of the message is sent to each person. Once a message has reached the machine to which it was addressed, then it is usually stored in the incoming mailbox of the person it is addressed to.

An incoming mailbox, or mailbox is simply a file containing mail messages, with a special sequence of characters separating each message from the others. Each user of a system will normally have their own mailbox. Users can also store messages in files of their own, which are called folders. In this manual, the term folder will refer to any file which contains mail messages.

To read your mail, you will usually use a mail reader, a program which can understand the format of a folder and present the contents to you in a convenient form. Af is an example of a mail reader. Similarly, you would normally send mail by using a mail composer, a program which knows how to construct and send a mail message, and gives the user a convenient environment in which to compose and send their message. Most mail readers, including af, are also mail composers.

To allow a mail message to be sent to the correct person, it needs an envelope. To supply this, a mail message is divided up into two parts, the headers and the body. The headers supply the information required to send the message to the right people, while the body is the actual message you wanted to send.

1.1 Mail Headers

The headers of a mail message are held at the beginning of the message, and are separated from the message body by a blank line. Here's an example of message headers:

From malc@thing.demon.co.uk Tue Jan 23 16:20:39 1996
Date: Tue, 23 Jan 1996 16:20:36 GMT
Message-Id: <4780.199201201539@thing.demon.co.uk>
From: Malc Arnold <malc@thing.demon.co.uk>
Organization: Team Limpid
Sender: The Management <root@thing.demon.co.uk>
To: Kay Dekker <kay@thing.demon.co.uk>,
    Andrew Fry <vic@pootug.demon.co.uk>
Subject: Af manual in progress
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Note that apart from the first line, each header consists of a header name, followed by a colon and then the header text. The first line is not really a header at all; it is the marker that most Unix systems insert to mark the start of each new message. You can also see that unlike the envelope for a postal letter, the envelope information contained in the header can also be of interest to the recipient. There are actually several other headers contained in most mail messages, but by default af will filter them out when it displays a message for you to read (see section Reading Messages).

Another interesting feature is the `To:' header, which has been extended over two lines by starting the second line with a tab. You can also continue headers in the same way by starting the following line with a space. Probably it's best to try to keep headers to a single line though; if a long header line needs to be folded like this, then af will normally do so for you.

Here is a brief summary of some of the headers you are likely to see in a mail message.

`From:'

The e-mail address of the person who the mail is from. Replies to the message will be sent to this address (see section Composing and Sending Mail). It is possible for a message to be from more than one person.

`Organization:'

The organisation the sender belongs to.

`Subject:'

A brief description of what the message is about.

`To:'

The e-mail addresses of the message's primary recipients.

`Cc:'

A "carbon copy" of the message was sent to the addresses listed. People listed in the `Cc:' header should regard the mail as being a copy for informational purposes.

`MIME-Version:'

Indicates that the message is a MIME message. See section The Message Body.

`Content-Type:'

What type of data the message contains. See section Content Type.

`Content-Transfer-Encoding:'

How the data has been encoded for transport via mail. See section Transfer Encoding.

1.2 Mail Addresses

Mail addresses usually take the form `user@full.domain.name'. They may also be written as `Full Name <user@full.domain.name>', or `user@full.domain.name (Full Name)'. In all of these cases, the address of the user is the same; the addition of the full name is simply a convenience to allow mail readers to display the person's full name rather then their e-mail address.

Usually, the part of the address to the left of the "@" character (`user') identifies the user to which the mail is to be sent, while the part to the right (`full.domain.name') identifies the specific computer or organisation where that user can be found. Note that while mail addresses must be unique, a user name may be duplicated many times by different organisations; each with a different full domain name.

Lists of addresses (such as in a `To:' header) should be separated either by spaces or commas.

We shall not go into the intimate details of mail address syntax here; af will check any addresses you give it, and either correct minor problems or report any serious errors in an address.

1.3 The Message Body

Normally, the body of a mail message is just text, which is not handled in any special way. Recently, electronic mail has been extended to allow for a message body which is something other than text. These extensions are called MIME (Multipurpose Internet Mail Enhancements), and mail which makes use of them is called MIME mail.

A MIME message uses the `Content-Type:' header to define the type of data the message contains. The `Content-Transfer-Encoding:' header describes how the data was encoded in order to be sent via mail. The `Content-Disposition:' header describes how a message should be handled when read. Finally the `Content-Description:' header can be used to describe the contents of a message.

1.3.1 Body Parts

A MIME message body may consist of more than one part. Each of these body parts is like a miniature MIME message in its own right, and has its own headers specifying the MIME details.

For example, a message might contain two body parts. The first is a description of the work you've been doing; the second is an image displaying the results of that work. You can see that the two parts of the message are related, but they must go into different body parts since they contain a different type of data.

Since we can think of a normal message as having one body part, when describing MIME mail it is conventional to talk about body parts rather than the message body. We will follow this convention in this manual.

1.3.2 Content Type

The body of a MIME mail message may be textual, or it may be an image or some other non-textual material. The type of the message body is specified by the `Content-Type:' header.

The form of a `Content-Type:' header is something like:

Content-Type: type/subtype [; parameter=value ... ]

The main type is separated from a subtype by a slash. The content type may be further refined by parameters, which are separated from the type and subtype, and from each other, by semicolons.

The content types you are likely to see in mail messages are:

`multipart'

The message body consists of a set of body parts. The `mixed' subtype indicates that the body parts may contain any content type. The `digest' subtype indicates that each body part is an entire mail message, which may itself be in MIME format. The `alternative' subtype indicates that each body part contains a different form of the same data, and only the "best" should be displayed. The `parallel' subtype indicates that the body parts should be displayed simultaneously if possible. Af can handle multipart messages internally.

`message'

The body part is itself a MIME message. The `rfc822' subtype indicates that the body part is an entire MIME message. Af is capable of handling such encapsulated messages internally.

The `partial' subtype indicates that the original message has been split into a number of smaller messages, presumably to avoid limits on the size of mail messages. Af can rebuild the original message from such partial messages, if all the parts of the original message have arrived intact.

`text'

The body part is textual. Although it may contain formatting information, it should be readable as it stands. The `plain' subtype means there is no formatting information; the text should be displayed as it stands.

The `charset' parameter indicates the character set the text is written in. Common values are `us-ascii' for plain ASCII text, and `iso-8859-1', which is suitable for Western European languages. See section Character Sets.

`image'

The body part is an image, in a format specified by the subtype, such as `image/gif' or `image/jpeg'. Af will call an external program to display `image' body parts.

`audio'

The body part is audio data, in a format specified by the subtype. Af will call an external program to display `audio' body parts.

`video'

The body part is video data, in a format specified by the subtype. Af will call an external program to display `video' body parts.

`application'

The body part is raw data for some application, often specified by the subtype. One common value is `application/octet-stream', which indicates that the body part contains raw data with no meaning attached. Hopefully the context will make it clear what you are to do with the data. Af will call an external program to display `application' body parts.

1.3.3 Character Sets

One of the commonest uses of MIME (in Europe at least) is to send mail which contains international characters (such as accents, or Cyrillic characters) which aren't in the us-ascii character set.

To allow these characters in a message, textual MIME messages can specify which character set they are written in with a `charset' parameter. The character set may be either `us-ascii' (the default), or one of the `iso-8859' character sets.

If your display is capable of displaying these character sets properly, Af can be set up so that it display's such messages properly by setting the variable viewable-charsets (see section Variables).

The person who installed af at your site may have already configured it to handle messages in the available character sets. In this case you don't need to do anything to make af handle internationalised messages correctly.

Sadly, none of these character sets is capable of displaying all the characters used in all the world's languages. So you may sometimes receive textual messages which you cannot read. The ISO-8859-1 character set, which is suitable for most Western European languages, is by far the most common character set used in MIME mail.

1.3.4 Transfer Encoding

Traditionally, the body of a mail message had to contain only plain text consisting of only ASCII characters, and much of the infrastructure which supports mail is only capable of transmitting such messages. For this reason, it may have been necessary to encode the contents of the message before they were send via mail. This encoding is specified by the `Content-Transfer-Encoding:' header.

The encodings that you may encounter in a MIME body part are:

`7bit'

The body part contains only short lines of 7 bit ASCII characters. No encoding has been applied.

`8bit'

The body part contains only short lines of 8 bit characters, not including the null character ASCII NUL. No encoding has been applied.

`binary'

The body part contains binary data, but has not been encoded. The electronic mail infrastructure is such that this encoding is unlikely to work.

`quoted-printable'

The body part has been encoded with the `quoted-printable' encoding. This encoding leaves ASCII characters untouched, but replaces international characters with sequences beginning with `='. It is intended for encoding text, since many characters will still be readable without decoding the message.

`base64'

The body part has been encoded with the `base64' encoding. This renders the body part completely unreadable, but is more compact than `quoted-printable' if the message contains a high proportion of non-ASCII characters. It is suitable for encoding binary data, such as images.

`x-uuencode'

`x-uue'

These encodings are not part of MIME at all. They represent the Unix `uuencode' encoding, which predates MIME. Some people send messages marked in this format, and af is able to handle them. These encodings are similar to `base64', but is not as robust. It is a bad idea to send messages encoded in this way except by prior arrangement with the recipient.

1.3.5 Content Disposition

Sometimes a body part isn't primarily intended to be read. It might be a program source file, or a configuration file. The `Content-Disposition:' header specifies whether a body part should be displayed inline when the message is read, or that it is an attachment, which requires confirmation before it is displayed.

You can also specify a file name in the `Content-Disposition:' header. Again, this is a hint to the recipient's mail reader that the body part should be stored in the named file if it is saved.

The form of a `Content-Disposition:' header is something like:

Content-Disposition: disposition [; filename=file ... ]

1.3.6 Content Description

Sometimes it can be useful to the recipient to have a textual description of a body part, especially if the body part contains something other than text. The `Content-Description:' header contains just that; a free text description of the contents of a body part.

1.4 Signatures

It is common for people to want to include a small amount of text at the end of each message, to give information such as their telephone number, paper-mail ("snail-mail") address, other e-mail addresses. and so on.

To aid people in doing this, af supports a feature known as signature files. These files contain text which will automagically be included at the end of each mail message you send. By default, af's uses the file .signature in your home directory as your signature file.

Some people also like to include lengthy quotations, large ASCII graphics or other amusements in signature files. This annoys many people on the net, since it costs money to transfer large signatures, and conveys no useful information. Because of this, the version of af which you use may be configured to truncate signature files that are larger than a certain size; typically 4 lines of 79 columns each. You have been warned.

2. The Organization of the Screen

Normally, the af display occupies the whole screen. When you start af, the entire screen except for the last line is devoted to the folder you are reading. This area is called the window. The last line is a special echo area or minibuffer window where prompts appear and where you can enter responses. You can subdivide the window into different windows, each of which can be used to display a different folder (see section Multiple Windows). In this manual, the word "window" always refers to the subdivisions of the screen within af.

The window that the arrow cursor is in is the selected window, in which most operations take place. Most af commands implicitly apply to the messages in the selected window. The other windows display messages for reference only, unless/until you select them.

Most windows display the details of folders, showing a header line for each message in the folder. Each header line shows a brief summary of the message's details, and is treated as a pointer to the message it describes.

Each window's last line is a mode line which describes what is going on in that window. It contains text that starts like `==== Af: something'. Its purpose is to indicate what buffer is being displayed above it in the window above it, how many messages are in the buffer, whether the buffer contains unsaved changes, and so on.

2.1 Point

Within a buffer displaying a folder, an arrow pointer shows the location at which commands will take effect. This location is called point. Many af commands move point through the buffer, so that you can execute commands at different places in it.

While the arrow appears to point at a message, you should think of point as between two messages; it lies before the message that it is pointing at. Sometimes people speak of "the cursor" when they mean "point", or speak of commands that move point as "cursor motion" commands.

If you are reading several folders in af, each in its own buffer, each buffer has its own point location. A buffer that is not currently displayed remembers where point is in case you display it again later.

When there are multiple windows on the screen, each window has its own point location. The arrow shows the location of point in the selected window. This also is how you can tell which window is selected. If the same buffer appears in more than one window, each window has its own position for point in that buffer.

It is possible to move point past the last message in a buffer. This is deliberate, to allow several other commands to work properly in an "Emacs-like" way. While it may seem strange at first if you are used to other mail readers, it soon becomes familiar; and it allows you to use many af features much more easily. Most commands that deal with messages will report an error if you have moved point past the last message in the buffer, just as they would if there were no messages at all.

2.2 The Echo Area

The line at the bottom of the screen (below the mode line) is the echo area. It is used to display small amounts of text for several purposes.

Echoing means displaying the characters that you type. Outside af, the operating system normally echoes all your input. Inside af things work a little differently.

Single-character commands do not echo in af, and multi-character commands echo only if you pause while typing them. As soon as you pause for more than a second in the middle of a command, af echoes all the characters of the command so far. This is to prompt you for the rest of the command. Once echoing has started, the rest of the command echoes immediately as you type it. This behaviour is designed to give confident users fast response, while giving hesitant users maximum feedback. You can change this behaviour by setting the variable echo-keystrokes (see section Variables).

If a command cannot be executed, it may print an error message in the echo area. Error messages are accompanied by a beep.

Some commands print informative messages in the echo area. These messages look much like error messages, but they are not announced with a beep. Sometimes the message tells you what the command has done, when this is not obvious from looking at the screen. Sometimes the sole purpose of a command is to print a message giving you specific information--for example, C-x = prints a message describing the position of point in the buffer. Commands that take a long time often display messages ending in `...' while they are working, and add `done' at the end when they are finished.

The echo area is also used to display the minibuffer, a window that is used for reading arguments to commands, such as the name of a file to be read. When the minibuffer is in use, the echo area begins with a prompt string that usually ends with a colon; also, the cursor appears in that line because it is the selected window. You can always get out of the minibuffer by typing C-g (see section The Minibuffer).

2.3 The Header Lines

Most of the lines in a window which displays a folder will be header lines, which present a one-line summary of a mail message. By default a header line will look something like this:

ptr tags  originator                  subject

ptr is the pointer which indicates the position of point in the window; the ptr will be represented as `=>' only on the line that point lies before, and as blank on all other lines. The line on which the pointer is present is sometimes referred to as the current line, and the message it represents as the current message, or the message at point.

tags are the tags of the message. Normally, this will only show system tags which af sets to show information about a message's status (see section Tags).

originator is the sender of the message. If the sender's real name is available in the message headers then it will be shown here, otherwise their e-mail address will be displayed.

subject is the subject of the message.

It is possible to change the layout of the header lines by setting the variable header-line-format (see section Variables). Also, the arrow pointer can be changed by setting the variable header-line-arrow.

2.4 The Mode Line

Each window's last line is a mode line which describes what is going on in that window. When there is only one window, the mode line appears right above the echo area. The mode line starts and ends with dashes, and it contains text like `Af: something'.

Normally, the mode line looks like this:

=ch= Af: buf == count == (modes) == pos =

This gives information about the buffer being displayed in the window: the buffer's name, what modes are in use, whether the buffer has been changed, and how far down the buffer you are currently looking.

ch contains two stars `**' if the buffer has been changed (the buffer is "modified"), two plus signs `++' if the buffer contains messages whose status has changed, or `==' if the buffer has not been changed. For a read-only buffer, it is `%*' if the buffer is modified, and `%%' otherwise.

buf is the name of the window's buffer. In most cases this is the same as the name of a folder you are processing. (see section Using Multiple Buffers)

count is the number of messages in the window's buffer.

The buffer displayed in the selected window (the window that the arrow cursor is in) is also af's selected buffer, the one that most commands operate on. When we speak of what some command does to "the buffer", we are talking about the currently selected buffer.

pos tells you whether there are more messages above the top of the window, or below the bottom. If your buffer is small and it is all visible in the window, pos is `All'. Otherwise, it is `Top' if you are looking at the beginning of the buffer, `Bot' if you are looking at the end of the buffer, or `nn%', where nn is the percentage of the buffer above the top of the window.

modes lists the major mode and any minor modes which are in effect in the buffer. At any time, each buffer is in one and only one of the possible major modes. The major modes available include Mail mode (for reading folders), Typeout mode (for displaying information) and Minibuffer mode (for asking the user for input). Each major mode may be supplemented by one or more minor modes, which change the mode's behaviour in some small way. (see section Major Modes).

It is possible to change the layout of the mode line by setting the variable mode-line-format (see section Variables).

3. Characters, Keys and Commands

ASCII consists of 128 character codes. Some of these codes are assigned graphic symbols such as `a' and `='; the rest are control characters, such as Control-a (usually written C-a for short). C-a gets its name from the fact that you type it by holding down the CTRL key while pressing a.

Some control characters have special names, and special keys you can type them with: for example, RET, TAB, LFD, DEL and ESC. The space character is usually referred to below as SPC, even though strictly speaking it is a graphic character whose graphic happens to be blank.

On ASCII terminals, there are only 32 possible control characters. These are the control variants of letters and `@[]\^_'. In addition, the shift key is meaningless with control characters: C-a and C-A are the same character, and af cannot distinguish them.

One af character set extension is that characters have an additional modifier, called Meta. Every character has a Meta variant; examples include Meta-a (normally written M-a, for short), M-A (not the same character as M-a, but those two characters normally have the same meaning in af), M-RET, and M-C-a. For reasons of tradition, we usually write C-M-a rather than M-C-a; logically speaking, the order in which the modifier keys CTRL and META are mentioned does not matter.

Some terminals have a META key, and allow you to type Meta characters by holding this key down. Thus, Meta-a is typed by holding down META and pressing a. The META key works much like the SHIFT key. Such a key is not always labelled META, however, as this function is often a special option for a key with some other primary purpose.

If there is no META key, you can still type Meta characters using two-character sequences starting with ESC. Thus, to enter M-a, you could type ESC a. To enter C-M-a, you would type ESC C-a. ESC is allowed on terminals with META keys, too, in case you have formed a habit of using it.

ASCII terminals represent function keys as a special sequence of ASCII characters. Because of this, it is possible, although not simple, to configure af to handle function and arrow keys. (see section Customisation)

3.1 Keys

A key sequence (key, for short) is a sequence of characters that are meaningful as a unit--as "a single command." Some af command sequences are just one character; for example, just C-n is enough to move down one line. But af also has commands that take two or more characters to invoke.

If a sequence of events is enough to invoke a command, it is a complete key. Examples of complete keys include C-n, RET, C-x C-f and C-x 4 C-f. If it isn't long enough to be complete, we call it a prefix key. The above examples show that C-x and C-x 4 are prefix keys. Every key sequence is either a complete key or a prefix key.

Many single characters constitute complete keys in the standard af command bindings. A few of them are prefix keys. A prefix key combines with the following character to make a longer key sequence, which may itself be complete or a prefix. For example, C-x is a prefix key, so C-x and the next character combine to make a two-character key sequence. Most of these key sequences are complete keys, including C-x C-f and C-x b. A few, such as C-x 4 and C-x r, are themselves prefix keys that lead to three-character key sequences. There's no limit to the length of a key sequence, but in practice people rarely use sequences longer than four characters.

By contrast, you can't add more characters onto a complete key. For example, the two-character sequence C-n C-k is not a key, because the C-n is a complete key in itself. It's impossible to give C-n C-k an independent meaning as a command. C-n C-k is two key sequences, not one.

All told, the prefix keys in af are C-h, C-t, C-x C-x 4, and ESC. But this is not cast in concrete; it is just a matter of af standard key bindings. If you customise af, you can make new prefix keys, or eliminate these (see section Customising Key Bindings).

If you do make or eliminate prefix keys, that changes the set of possible key sequences. For example, if you redefine C-n as a prefix, C-n C-k automatically becomes a key (complete, unless you define it too as a prefix). Conversely, if you remove the prefix definition of C-x 4, then C-x 4 f (or C-x 4 anything) is no longer a key.

3.2 Keys and Commands

This manual is full of passages that tell you what particular keys do. But af does not assign meanings to keys directly. Instead, af assigns meanings to named commands, and then gives keys their meanings by binding them to commands.

Every command has a name chosen by a programmer. The name is usually made of a few English words separated by dashes; for example, next-line or open-message. The bindings between keys and commands are recorded in various tables called keymaps (see section Customising Key Bindings).

When we say that "C-n moves down vertically one line" we are glossing over a distinction that is irrelevant in ordinary use but is vital in understanding how to customise af. It is the command next-line that is programmed to move down vertically. C-n has this effect because it is bound to that command. If you rebind C-n to the command end-of-buffer then C-n will move to the end of the buffer instead. Rebinding keys is a common method of customisation.

In the rest of this manual, we usually ignore this subtlety to keep things simple. To give the information needed for customisation, we state the name of the command which really does the work in parentheses after mentioning the key that runs it. For example, we will say that "The command C-n (next-line) moves point vertically down," meaning that next-line is a command that moves vertically down and C-n is a key that is standardly bound to it.

While we are on the subject of information for customisation only, it's a good time to tell you about variables. Often the description of a command will say, "To change this, set the variable foobar." A variable is a name used to remember a value. Most of the variables documented in this manual exist just to facilitate customisation: some command or other part of af examines the variable and behaves differently according to the value that you set. Until you are interested in customising, you can ignore the information about variables. When you are ready to be interested, read the basic information on variables, and then the information on individual variables will make sense (see section Variables).

3.3 Major Modes

Af has several modes of operation; known as major modes. A major mode is in effect a subset of af, where a command may have a different effect, and different commands may be available (This is probably the most pronounced difference between af and Emacs; Emacs major modes just alter its behaviour in minor ways, such as changing key bindings; while af major modes may have a completely different set of commands and behaviour).

There are three major modes currently available in af. Mail mode is the mode used for buffers displaying mail messages; most af buffers use mail mode. Minibuffer mode is used to support the minibuffer; it has no commands to handle mail messages, but many extra commands to allow editing of text. Typeout mode is the mode used for displaying text to the screen. Again, it has no commands to handle mail messages, it simply allows you to browse text displayed on the screen.

Each major mode allows you to bind keys independently of the other major modes (just like in Emacs), so keys may be bound in only one mode, or even be bound to different things in different modes (just as they can in Emacs). For example, in mail mode SPC just clears the echo area, in minibuffer mode it inserts a space into the text you are editing, and in typeout mode it scrolls the text up to display the next page.

Each af command may be available in all modes, or just in some of them. For example the command open-message is only available in mail mode, since in the other modes you are not dealing with mail messages and the command wouldn't make sense. So you can't run the command open-message in typeout mode even by typing M-x open-message.

A command may also have slightly different behaviour in different modes. For example the command search-forward in mail mode searches through the text of each mail message, looking for some text. In typeout mode it searches through the text you are viewing instead.

This distinction is only really important when customising af; when you are using af it is (hopefully) quite obvious what commands are likely to work. The important thing to remember is that keys may do very different things depending on whether you are browsing a list of messages, entering an argument in the minibuffer, or viewing text.

3.4 Minor Modes

A minor mode is a particular feature which can be turned on or off. For example, Read Only mode selects whether you are allowed to modify the contents of a buffer. Unlike the major modes, the minor modes are independent of each other, although they may only be valid in some major modes.

Minor modes are usually specific to a buffer, so they may be turned on in some buffers and off in others. Normally, minor modes are turned on or off automatically in response to some situation. There may also be a command to turn the minor mode off if it is on, or on if it is off; this will be formed from the prefix `toggle-' followed by the name of the mode. So the command to turn Read Only mode on or off is toggle-read-only.

4. Entering Af

The usual way to invoke af is with the shell command `af'. Af clears the screen and then displays an initial help message and copyright notice while it processes any startup files (see section The Startup File, `~/.afrc'), and then reads your incoming mailbox. (If your mailbox is small, you may not see the initial help message and copyright; it would be very inconvenient for experienced users if af were to pause long enough to let you read the message.) When af has finished reading the mailbox, it displays a final count of how many messages it read, and displays the buffer containing the messages.

Some operating systems discard all type-ahead when af starts up; they give af no way to prevent this. Therefore, it is advisable to wait until af clears the screen before typing your first command.

It is possible to specify folders to be visited, or details of a message you want to send, by giving af arguments in the shell command line (see section Command Line Arguments). But you can do these things from within af too; whether you run af once and suspend it, or separately each time you want to read or send mail is simply a matter of taste.

Arguments starting with `-' are options. Other arguments specify addresses to send mail to. If you specify addresses (or use any of the `-c', `-b', or `-E' options), then af doesn't read any folders; it simply allows you to send a single mail message to the addresses you supplied, and then exits.

You can use options to specify various other things, such as which folders to read, the subject for mail you're sending, and so on. A few options support advanced usage, such as running afl functions in batch mode. The sections of this chapter describe the available options, arranged according to their purpose.

Most options specify how to initialise af, or set parameters for the af session. We call them initial options. A few options specify things to do: for example, read folders or load afl programs (see section Afl). These are called action options. These and file names together are called action arguments. Af processes all the action arguments in the order they are written.

4.1 Command Line Arguments

Here is a table of the arguments and options that af accepts:

`address'

Send mail to address. See section Composing and Sending Mail. If addresses are supplied, then many of the other arguments are disabled, since af will not be visiting any folders. If the standard input is not a terminal, then af will silently send the message as requested, rather than expecting you to edit the text of the message.

`-E'

This option has two effects. Firstly, it makes af start up to send a single mail message even if no addresses were given as arguments. Secondly, it forces af to allow you to edit the text of the message you send, even if the text is being obtained from the standard input. This doesn't really seem very useful, but occasionally it can be very handy in allowing you to write flexible scripts which send mail using af.

`-H'

Force af to present the user with the headers of an outgoing message when editing it, or to force translation of headers in mail sent from the standard input. In effect, it causes af's edit-initial-headers variable to be set to true regardless of the contents of the user's `.afrc' file (see section Variables).

`-c cc'

Specifies one or more addresses to send a carbon copy of a message to. If the addresses contain spaces then you should enclose them in double quotes.

`-b bcc'

Specifies one or more addresses to send a blind carbon copy of a message to. If the addresses contain spaces then you should enclose them in double quotes.

`-s subject'

Specifies the Subject: for an outgoing message. Only one `-s' option may be given, and it only takes effect if addresses (or the `-c', `-c', or `-E' option) were given on the command line. If the subject contains spaces then you should enclose it in double quotes.

`-C content-type'

Specifies the MIME Content-Type: for an outgoing message. Only one `-C' option may be given, and it only takes effect if addresses (or the `-E' option) were given on the command line.

`-f folder'

Read folder in place of your incoming mailbox. Multiple `-f' options may be given, in which case each folder is read into its own buffer. This option is ignored if addresses or the `-E' option were given.

`-u user'

Read user's incoming mailbox in place of your own. You must have permissions to read user's mailbox. Multiple `-u' options may be given, and they may be mixed with `-f' options. This option is ignored if addresses or the `-E' option were given.

`-F'

Interpret any arguments other than options as folders to be read in, rather than addresses to send mail to. This can be very useful if you want to read in a set of folders with `af -F *'.

`-w'

Create enough windows to display all the folders given with the `-f', `-u' or `-F' options, subject to the maximum number of windows that will fit on the screen (usually about four on a 24-line display). This option is ignored if addresses or the `-E' option were given.

`-e'

Check whether there is mail in your default mailbox (or, if `-f' or `-u' are given, any specified mailboxes) and then exit. The exit status is zero if there is mail in any of the mailboxes; one if not. This option is ignored if addresses or the `-E' option were given.

`-z'

Makes af terminate immediately with an exit status of 1 if your incoming mailbox (or, if `-f' or `-u' are given, any specified folders) is empty. Otherwise, af will start up normally. This option is ignored if addresses or the `-E' option were given.

`-n'

Ignore the file `.afrc' in the user's home directory. See section The Startup File, `~/.afrc'.

`-v'

Prints the version of af and then exit. Ignores the effect of any other options.

`-l loadfile'

Specifies a file which should be read and executed on startup before any buffers are created. Used with the `-n' option, this allows the user to specify an alternate startup file, rather than the default of `.afrc' in their home directory.

`-S file'

Names a file of commands in af's internal language afl. The commands in the file are read and executed exactly as if af had been started up interactively and then the af command load-file had been executed. When all the commands in file have been executed, or if an error is encountered in the file, then af terminates. If the `-S' option is given, then af will not display the screen; only messages will be printed to the standard output. Only one `-S' option may be given, it is ignored if destinations or the `-E' option were specified.

4.2 Command Line Examples

Here are a few examples of af command line usage, with brief explanations.

Read two folders `foo' and `bar' into separate windows:

af -f foo -f bar -w

Check if there is any mail in folders foo and bar:

af -e -f foo -f bar

Send mail to `fred' on the local system with subject `Hi':

af -s Hi fred

Take a current directory listing, and use it as a base for some mail to user `fred@foobar.co.uk':

ls -l | af -E fred@foobar.co.uk

Run the afl script in file `script.afl' with folder `foobar' as the default buffer:

af -S script.afl -f foobar

4.3 Environment Variables

Af uses several environment variables. An environment variable is a string passed from the operating system to af, and the collection of environment variables is known as the environment. Environment variable names are case sensitive and it is conventional to use upper case letters only.

Because environment variables come from the operating system there is no general way to set them; it depends on the operating system and especially the shell that you are using. For example, here's how to set the environment variable ORGANIZATION to `Utter Chaos' using bash:

export ORGANIZATION="Utter Chaos"

and here's how to do it in csh or tcsh:

setenv ORGANIZATION "Utter Chaos"

It should be noted that the environment variables are used to customise af, not to configure it. None of these variables are required; and af should work correctly without them. They simply serve to tailor af to your taste.

Here is a list of the environment variables af uses, with a brief description of what they are used for.

MAIL

The full path to your incoming mailbox.

HOME

Your home directory.

FOLDER

The directory where your mail folders are stored. Defaults to `~/Mail'.

SAVEDIR

The directory where your news folders are stored. Defaults to `~/News'.

TMPDIR

The directory where temporary files should be created.

VISUAL, EDITOR

The editor to use for editing outgoing mail (or messages if you choose to edit them). VISUAL overrides EDITOR if they are set to different values.

PAGER

The program to use when you ask af to display a message via an external pager.

NAME

Your real name. Used for generating the headers on the messages you send.

ORGANIZATION

your organisational affiliation. If set, af will generate an Organization: header on the messages you send.

4.4 Exiting af

There are two commands for exiting af because there are two kinds of exiting: suspending af and killing af.

Suspending means stopping af temporarily and returning control to its parent process (usually a shell), allowing you to resume reading mail later in the same af job, with the same buffers, same kill ring, and so on.

Killing af means destroying the af job. You can run af again later, but you will get a fresh af; there is no way to resume the same mail reading session after it has been killed.

C-z

Suspend af (suspend-af)

C-x C-c

Kill af (save-buffers-kill-af).

M-z

Kill af, automatically saving any changed buffers (save-all-kill-af).

To suspend af, type C-z (suspend-af). This takes you back to the shell from which you invoked af. You can resume af with the shell command `%af' in most common shells.

On systems that do not support suspending programs, C-z starts an inferior shell that communicates directly with the terminal (shell). Af waits until you exit the inferior shell. (The way to do that is probably with C-d or `exit', but it depends on which shell you use). The only way on these systems to get back to the shell from which af was run (to log out, for example) is to kill af.

To kill af, type C-x C-c (save-buffers-kill-af). A two-character key is used for this to make it harder to type. This command first offers to save any modified file-visiting buffers. If you do not save them all, it asks for confirmation with yes before killing af, since any changes not saved will be lost forever.

To kill af, saving all modified file-visiting buffers, type M-z (save-all-kill-af). We don't recommend using this, since you might accidentally save changes you didn't mean to, but you can use M-z to exit if you prefer.

The operating system usually listens for certain special characters whose meaning is to kill or suspend the program you are running. This operating system feature is turned off while you are in af. The meanings of C-z and C-x C-c as keys in af were inspired by the use of C-z and C-c on several operating systems as the characters for stopping or killing a program, but that is their only relationship with the operating system. You can customise these keys to run any commands of your choice (see section Customising Key Bindings).

5. Basic af Commands

We now give the basics of how to read, save, and delete messages, send mail to people and save your changes back to your mailbox.

5.1 Reading Messages

The most basic operation in af is to open the current message; to display it's contents via typeout (see section Typeout), or some other pager.

RET or C-o

Open the current message, using the pager named in the pager variable, usually `typeout' (open-message).

M-RET or C-M-o

Open the current message, using the pager named in the PAGER environment variable (page-message).

C-x C-o

Open the current message, displaying all the headers and the message body using the pager named in the pager variable. The body is not decoded, and multipart or partial messages are not processed in any way. (open-raw-message).

Normally, only mail headers not listed in the headers-not-displayed variable will be displayed when you open a message. A positive numeric argument makes af display all the message's headers. A negative argument makes af skip all headers, and only display the body of the message. See section Numeric Arguments, for more information on numeric arguments.

When you open a multipart message, the headers and each textual body part will be displayed to a separate section of typeout. Non-textual body parts will be displayed using an external program if possible, otherwise you will be asked whether to view the text of the body part, save it to a file, pipe it into a command, or skip it. If any body part is an attachment, then af will offer you the choice of viewing, saving or skipping that body part.

Normally, af will ask for confirmation before it displays image, audio, video or application body parts, even if it knows how to. The confirm-viewing variable controls this, listing the content types which af will ask for confirmation before displaying.

When you open a partial message, af will try to rebuild and display the entire original message. If af can't rebuild the message then it will warn you about the problem, and offer you the choice of viewing or skipping that part of the message.

If your external pager doesn't need af to pause after displaying the message; then you can make reading messages more convenient by setting the variable pause-after-paging-message to true.

5.2 Changing the Location of Point

To do anything useful with af, you have to know how to move point (see section Point). There are several keys which move point within a buffer.

C-n or n

Move down one line, vertically (next-line).

C-p or p

Move up one line, vertically (previous-line).

M-.

Move point to the vertical centre of the window. (move-to-window-line). Text does not move on the screen.

A numeric argument says which screen line to place point on. It counts screen lines down from the top of the window (zero for the top line). A negative argument counts lines from the bottom (-1 for the bottom line).

M-< or <

Move to the top of the buffer (beginning-of-buffer). With numeric argument n, move to n/10 of the way from the top. See section Numeric Arguments, for more information on numeric arguments.

M-> or >

Move to the end of the buffer (end-of-buffer).

M-g

Read a number n and move point to line number n (goto-line). Line 1 is the beginning of the buffer.

5.3 Scrolling

Since only part of a large buffer fits in the window, af tries to show the part that is likely to be interesting. The display control commands allow you to specify which part of the buffer you want to see.

C-l

Clear screen and redisplay, scrolling the selected window to center point vertically within it (recenter). A numeric argument n says to move point to screen line n.

C-v

Scroll forward (a windowful or a specified number of lines) (scroll-up).

M-v

Scroll backward (scroll-down).

The names of all scroll commands are based on the direction that the messages move in the window. Thus, the command to scroll forward is called scroll-up, since the messages move up.

When scrolling a windowful at a time, af leaves two lines that were visible before you scrolled still visible afterwards, so that you can retain the context you were in before you scrolled. The number of lines of overlap across a C-v or M-v is controlled by the variable next-screen-context-lines; by default, it is two.

5.4 Deleting Messages

C-k or k

Kill the current message, removing it from the folder (kill-line). See section Killing and Yanking.

M-x delete-message

Mark the current message as deleted. This doesn't actually delete the message until you save the buffer. At that point any messages marked as deleted will be killed, and not written to the folder.

M-x undelete-message

Remove the deleted marker from the current message; so that it will not be killed when you save the buffer.

5.5 Message and Position information

Here are commands to get information about messages, and your position in the buffer.

M-=

Print line number of point in the buffer (what-cursor-position).

C-x =

Print a short summary of some of the message's details in the echo area (message-info).

C-t ?

Print the message's tags in the echo area (message-tags). See section Tags.

5.6 Quitting

At any time in af except when a command is running, you can type C-g (keyboard-quit) to quit from what you are doing. If you have typed part of a command, or a numeric argument, then C-g will get rid of it. If there isn't a partial command to get rid of, but you are in the minibuffer or typeout then typing C-g will exit back to the mail buffer, aborting the command you were running.

This means, that you can always get back to the top level of af by typing C-g C-g. This is useful for aborting commands, or if you are unsure where you are in af.

5.7 Help

If you forget what a key does, you can find out with the Help character, which is C-h. Type C-h k followed by the key you want to know about; for example, C-h k C-n tells you all about what C-n does. C-h is a prefix key; C-h k is just one of its subcommands (the command describe-key). The other subcommands of C-h provide different kinds of help. Type C-h three times to get a description of all the help facilities. (see section Help).

5.8 Numeric Arguments

Any af command can be given a numeric argument (also called a prefix argument), although it may be ignored. Some commands interpret the argument as a repetition count. For example, C-n with an argument of ten moves down ten lines instead of one. With these commands, no argument is equivalent to an argument of one. Negative arguments tell most such commands to move or act in the opposite direction.

If your terminal keyboard has a META key, the easiest way to specify a numeric argument is to type digits and/or a minus sign while holding down the META key. For example,

M-5 C-n

would move down five lines. The characters Meta-1, Meta-2, and so on, as well as Meta--, do this because they are keys bound to commands (digit-argument and negative-argument) that are defined to contribute to an argument for the next command.

Another way of specifying an argument is to use the C-u (universal-argument) command followed by the digits of the argument. With C-u, you can type the argument digits without holding down modifier keys; C-u works on all terminals. To type a negative argument, type a minus sign after C-u. Just a minus sign without digits normally means -1.

C-u followed by a character which is neither a digit nor a minus sign has the special meaning of "multiply by four". It multiplies the argument for the next command by four. C-u twice multiplies it by sixteen. Thus, C-u C-u C-n moves down sixteen lines. This is a good way to move down "fast", since it moves about 2/3 of a screen in the usual size screen. C-u is also a handy way of providing an argument when you don't care about the value.

Many commands care only about whether there is an argument, and not about its value. Other commands care only about the sign of the argument. For example, the command RET (open-message) with no argument shows only some of the message's headers; with a positive argument, it shows all the headers, and with a negative argument it shows no headers at all. This may seem strange, but it is a convenient way of modifying the behaviour of a command.

We use the term "prefix argument" as well as "numeric argument" to emphasise that you type the argument before the command, and to distinguish these arguments from minibuffer arguments that come after the command.

6. Typeout

Typeout is the mode af uses to display large amounts of text such as mail messages or help information. When af uses typeout, the typeout window appears, covering the entire screen apart from the echo area. Once you exit typeout, the window disappears, "uncovering" the original windows.

If possible, each line of the text is displayed as a single line on the screen. If a line is too long to fit on a single screen line, then it will be displayed on several screen lines, with a `\' at the extreme right margin of all but the last of them. The `\' says that the following screen line is not really a distinct line in the text, but just the continuation of a line too long to fit the screen. This is called line wrapping, and the lines after the first are often referred to as continuation lines.

When you are viewing typeout, the position indicator in the mode line will often say `MOR'. This is because typeout does not wait to read the entire input before displaying the first page, and since af doesn't know how many lines there are in the text yet, it will give the position as `MOR' to indicate this.

Most of the time, typeout consists of a single section. But when typeout is displaying related outputs, such as the body parts of a multipart message, typeout will consist of several sections. Each section of typeout will be displayed separately.

Sometimes, the typeout window will appear, but you will still be able to carry on with some other operation (such as displaying the help options via C-h C-h C-h). In these cases the typeout window will disappear when the operation is completed.

Within typeout, only the cursor motion commands, the scrolling commands and the text searching commands can be used. To exit typeout, simply type C-g. The other af commands are irrelevant when you are just viewing text. Here is a brief summary of the commands available in typeout:

SPC

Scroll the text up to show the next page of text. If you are already at the end of the text, then exit typeout (typeout-scroll). This is a convenient way of paging through short amounts of text.

C-v

Scroll forward (a windowful, or a specified number of lines) (scroll-up).

M-v

Scroll backward (a windowful, or a specified number of lines) (scroll-down).

C-n or n

Scroll forward one line (next-line).

C-p or p

Scroll backward one line (previous-line).

C-] or ]

Move to the next typeout section (next-section).

C-[ or [

Move to the previous typeout section (previous-section).

C-l

Redraw the display (recenter).

C-s regex RET

Search forward through the text for a line matching regex (search-forward). See section Searching Mail Folders.

C-r regex RET

Search backward through the text for a line matching regex (search-forward). See section Searching Mail Folders.

7. The Minibuffer

The minibuffer is the facility used by af commands to read arguments more complicated than a single number. Minibuffer arguments can be file names, buffer names, af command names, address lists, and many other things, depending on the command reading the argument. You can use many editing commands in the minibuffer to edit the argument.

When the minibuffer is in use, it appears in the echo area, and the terminal's cursor moves there. The beginning of the minibuffer line displays a prompt which says what kind of input you should supply and how it will be used. Often this prompt is derived from the name of the command that the argument is for. The prompt normally ends with a colon.

Sometimes a default argument appears in parentheses after the colon; it too is part of the prompt. The default will be used as the argument value if you enter an empty argument (e.g., just type RET). For example, commands that read buffer names always show a default, which is the name of the buffer that will be used if you type just RET.

The simplest way to enter a minibuffer argument is to type the text you want, terminated by RET which exits the minibuffer. You can cancel the command that wants the argument, and get out of the minibuffer, by typing C-g.

Sometimes, a default response will already be in the minibuffer when you enter it. If it is completely wrong, then you can type C-a C-k to remove it. These default responses appear for convenience when you are prompted for a long value that you are likely to want to edit, rather then retype from scratch.

Most commands which use typeout allow you to redirect the output into a file, instead of viewing it. The major exception to this is RET (open-message), since there are several ways to save a message to a file. To redirect typeout to a file simply give the command a prefix argument of any value; you will be prompted for the file name using the minibuffer.

7.1 Echo Area Conflicts

Since the minibuffer uses the screen space of the echo area, it can conflict with other ways af customarily uses the echo area. Here is how af handles such conflicts:

• If a command gets an error while you are in the minibuffer, this does not cancel the minibuffer. However, the echo area is needed for the error message and therefore the minibuffer itself is hidden for a while. It comes back after a few seconds, or as soon as you type anything.
• If in the minibuffer you use a command whose purpose is to print a message in the echo area, such as C-x =, the message is printed normally, and the minibuffer is hidden for a while. It comes back after a few seconds, or as soon as you type anything.
• Echoing of keystrokes does not take place while the minibuffer is in use.

7.2 Inserting Text

To insert printing characters into the minibuffer, just type them. This inserts the characters you type into the buffer at the cursor (that is, at point; which has a different appearance in the minibuffer, but the same meaning). The cursor moves forward, and any text after the cursor moves forward too. If the text in the buffer is `foobar', with the cursor before the `b', then if you type xx, you get `fooxxbar', with the cursor still before the `b'. This all works because in the minibuffer all the printing characters run the command self-insert-command.

Direct insertion works for printing characters and SPC, but other characters act as editing commands and do not insert themselves. If you need to insert a control character or a character whose code is above 200 octal, you must quote it by typing the character C-q (quoted-insert) first. There are two ways to use C-q:

• C-q followed by any non-graphic character (even C-g) inserts that character.
• C-q followed by three octal digits inserts the character with the specified character code.

A numeric argument to C-q specifies how many copies of the quoted character should be inserted (see section Numeric Arguments).

When you have finished entering the text, simply type RET (newline) to accept the argument and exit the minibuffer.

7.3 Deleting Text

To delete a character you have inserted, use DEL (delete-backward-char). DEL deletes the character before the cursor (not the one that the cursor is on top of or under; that is the character after the cursor). The cursor and all characters after it move backwards. Therefore, if you type a printing character and then type DEL, they cancel out.

To delete the character that the cursor is under, use C-d (delete-char). This deletes the character, moving all the characters after the cursor backwards. The cursor is left in place. Therefore C-d is equivalent to C-f DEL.

To delete the text from the cursor to the end of the line, use C-k (kill-line). This kills the characters from the one the cursor is under to the end of the line, leaving the cursor at the end of the line (see section Killing and Yanking in the Minibuffer).

7.4 Editing the Text

To edit the text in the minibuffer, you need to be able to move around in the text. There are several commands available to do this:

C-a

Move to the beginning of the line (beginning-of-line).

C-e

Move to the end of the line (end-of-line).

C-f

Move forward one character (forward-char).

C-b

Move backward one character (backward-char).

C-t

Transpose the two characters on either side of point, moving point forward one character (transpose-chars).

7.5 Dealing with Words

The minibuffer includes several commands for dealing with words rather than characters. Often these take the usual command for moving point, but use a Meta key rather then a control key:

M-f

Move forward one word (forward-word).

M-b

Move backward one word (backward-word).

M-DEL

Delete the next word (backward-kill-word).

M-d

Delete the previous word (kill-word).

M-l

Convert the next word to lower case (downcase-word).

M-u

Convert the next word to upper case (upcase-word).

M-c

Capitalise the next word (capitalize-word).

7.6 Completion

For certain kinds of arguments, you can use completion to enter the argument value. Completion means that you type part of the argument, then af visibly fills in the rest, or as much as can be determined from the part you have typed.

When completion is available, certain keys--TAB, RET, and SPC--are rebound to complete the text present in the minibuffer into a longer string that it stands for, by matching it against a set of completion alternatives provided by the command reading the argument. ? is defined to display a list of possible completions of what you have inserted.

For example, when M-x uses the minibuffer to read the name of a command, it provides a list of all available af command names to complete against. The completion keys match the text in the minibuffer against all the command names, find any additional name characters implied by the ones already present in the minibuffer, and add those characters to the ones you have given. This is what makes it possible to type M-x del SPC m RET instead of M-x delete-message RET (for example).

Case is often significant in completion, because it is significant in many of the names that you can complete (buffer names and file names). Thus, `fo' often does not complete to `Foo'. Completion does ignore case distinctions for certain arguments in which case does not matter (such as af command or configuration variable names).

7.6.1 Completion Example

A concrete example may help here. If you type M-x co TAB, the TAB looks for alternatives (in this case, command names) that start with `co'. There are only two: copy-region-as-kill and copy-tagset-as-kill. These are the same as far as copy-, so the `co' in the minibuffer changes to `copy-'.

If you type TAB again immediately, there are multiple possibilities for the very next character--it could be `r' or `t'--so no more characters are added; instead, TAB just produces a beep. If you now type ?, af will produce a list of all possible completions to typeout; when you exit typeout, the minibuffer is still waiting for the next character.

If you go on to type r TAB, this TAB sees `copy-r'. The only command name starting this way is copy-region-as-kill, so completion fills in the rest of that. You now have `copy-region-as-kill' in the minibuffer after typing just co TAB r TAB. Note that TAB has this effect because in the minibuffer it is bound to the command minibuffer-complete when completion is available.

7.6.2 Completion Commands

Here is a list of the completion commands defined in the minibuffer when completion is available.

TAB

Complete the text in the minibuffer as much as possible (minibuffer-complete).

SPC

Complete the minibuffer text, but don't go beyond one word (minibuffer-complete-word).

RET

Submit the text in the minibuffer as the argument, possibly completing first as described below (minibuffer-complete-and-exit).

?

Print a list of all possible completions of the text in the minibuffer (minibuffer-list-completions).

SPC completes much like TAB, but never goes beyond the next hyphen or space. If you have `copy-r' in the minibuffer and type SPC, it finds that the completion is `copy-region-as-kill', but it stops completing after `region-'. This gives `copy-region-'. Another SPC at this point completes to `copy-region-as-'.

7.6.3 Strict Completion

There are three different ways that RET can work in completing the minibuffer, depending on how the argument will be used.

• Strict completion is used when it is meaningless to give any argument except one of the known alternatives. For example, when M-x reads the name of a command to run, it is meaningless to give anything but the name of an existing command or macro. In strict completion, RET refuses to exit if the text in the minibuffer does not complete to an exact match.
• Cautious completion is similar to strict completion, except that RET exits only if the text was an exact match already, not needing completion. If the text is not an exact match, RET does not exit, but it does complete the text. If it completes to an exact match, a second RET will exit.

  Cautious completion is used for reading file names for files that must already exist.

• Permissive completion is used when any string whatever is meaningful, and the list of completion alternatives is just a guide. For example, when C-x C-f reads the name of a folder to visit, any folder name is allowed, in case you want to create a folder. In permissive completion, RET takes the text in the minibuffer exactly as given, without completing it.

7.7 Minibuffer History

Every argument that you enter with the minibuffer is saved on a minibuffer history list so that you can use it again later in another argument. You can think of the minibuffer history as a buffer which you can move through displaying a line at a time.

C-p

Move to the next earlier argument string saved in the minibuffer history (previous-line).

C-n

Move to the next later argument string saved in the minibuffer history (next-line).

M-<

Move to the first argument string saved in the minibuffer history (beginning-of-buffer).

M->

Move to the last argument string saved in the minibuffer history (end-of-buffer). This will always move you to the text you were editing before using the minibuffer history.

C-r

Move to an earlier saved argument in the minibuffer history that matches the current text (search-backward).

C-s

Move to a later saved argument in the minibuffer history that matches the current text (search-forward)..

The simplest way to reuse the saved arguments in the history list is to move through the history list one element at a time. While in the minibuffer, type C-p to "move to" the previous minibuffer input, and use C-n to "move to" the next input. Similarly, you can use C-< and C-> to "move to" the first and last elements in the history list.

When you move to a new history line, any text you had typed is stored for you as the last history element, so you can return to it by typing M->. To use a history element as the argument, exit the minibuffer as usual with RET. You can also edit the text before you reuse it; this does not change the history element that you "moved" to, but your new argument does go at the end of the history list in its own right.

There are also commands to search forward or backward through the history. At the moment they search for a history element that starts with the text you have typed in the minibuffer, up to the position of the cursor; so typing `list- C-r' will search for a history element that begins with `list-'. If a matching history element is found then the cursor will stay in the same column, so that typing C-r again will search for another history element beginning with list-.

8. Composing and Sending Mail

Af allows you to compose and send mail as well as read it. There are several different commands to send mail, each of which sends mail in a slightly different way.

Af composes messages in three stages. First you are prompted for any details needed for the message, such as the addresses to send it to. Then af starts an editor, so that you can edit the message. Finally, you are prompted for what to do with the message (its disposition).

8.1 Commands for Sending Mail

There are seven commands for sending mail in af. These commands are all similar in use, but conceptually different.

M-m or M-s

Send a message to one or more people. This allows you to compose and send a mail message (send-mail). A numeric argument makes af compose a MIME message. Alternatively, you can attach files or compositions to the message later.

M-a

Prompts for a file name and then uses the contents of the file to compose a message to one or more people (send-file). You will not normally edit the text of the message. A numeric argument makes af prompt for the MIME Content-Type of the message to compose. If af detects that the file name indicates some content type other than plain text then it will default the Content-Type and prompt for the information anyway.

M-r

Reply to the current message. (reply-to-message). This lets you compose and send mail to whoever the current message is from.

C-M-r

Group reply to the current message. (group-reply-to-message). This lets you compose and send mail to whoever the message is from, and all the addresses listed in the `To:' header. Copies are sent to the addresses listed in the `Cc:' header.

M-f

Forwarding allows you to compose an annotated copy of the current message and send it to one or more people. (forward-message).

C-M-f

Allows you to compose a new message, and the current message will then be attached to the message you are sending (forward-message-as-attachment). A numeric argument makes af include the full headers of the message to attach. This is often useful when sending reports of unsolicited commercial email to the sender's postmaster.

C-x C-d

Allows you to compose a new message, and the messages in the region (see section The Mark and the Region) are attached to the outgoing message as a MIME digest (forward-region-as-digest). This is useful for forwarding several messages to someone in a convenient format.

C-t C-d

Allows you to compose a new message, and the messages in the tagset (see section Tags) are attached to the outgoing message as a MIME digest (forward-tagset-as-digest). This is useful for forwarding several messages to someone in a convenient format.

M-b

Bouncing means sending an exact copy of the current message to one or more addresses, making it appear that the copy you send them is from the person who sent the message to you (bounce-message). You will not normally edit the text of the message.

8.2 Initial Details for Composing Mail

Depending on how you are sending mail, and how you have configured af, you will be prompted for several things when you begin composing a mail message. Normally af will only prompt for `To:' and `Subject:', and possibly `Content-Type:', `Content-Disposition:', and `Content-Description:' if you give send-file an argument, but here are all the possible questions:

`To:'

The addresses you want send the mail to.

`Subject:'

A brief description of what the message is about.

`Cc:'

Any mail addresses to send additional copies of the message to.

`Bcc:'

Any mail addresses to send blind copies of the mail to. These addresses will not be listed in the headers of the message when it is sent.

`Content-Type:'

The MIME content type which describes the contents of the message, for example `text/plain', or `image/jpeg'. See section The Message Body.

`Content-Disposition:'

Can be set to `inline' or `attachment', indicating that the message contains an attachment which should not be viewed by default. The optional `filename' parameter (which is automatically set when you edit a file) indicates a suggested file name for the attachment. See section The Message Body.

`Content-Description:'

A brief description of the contents of the message body. See section The Message Body.

`Signature file'

Which signature file (see section Signatures) do you want to include at the end of the message.

`Copy message (y/n)'

If you are replying to a message, you can copy the text of of the original message into the initial text of the reply. This is called quoting.

8.3 Editing the Mail Message

When you first edit a message you will normally see either a blank file, or the textual body parts of the original message with each line prefixed with `> '. You can then use your editor as usual to compose the text of the message. When you save the message, your `.signature' file will be appended to the message, separated by a line containing `-- '.

When you edit a message for the second time (or if you have configured af to do so), the headers of the message are inserted into the file you edit, separated by a blank line, and your signature will be visible at the end of the file. You can edit the headers with your editor, and af will process the changes when you exit the editor; reporting any problems to typeout. You can create headers by adding them into the header section of the file, or delete them by deleting them from the file.

You should be very careful when editing a message's headers. If you accidentally insert a blank line before the headers, then af will not find them when you exit the editor, and will therefore think that you have deleted them all. This can normally be fixed by editing the file again, and removing the offending text from the start of the file.

If a message is a multipart MIME message, then instead of using an editor to edit the message, the various body parts of the message will be displayed in compose mode. There you can add or delete body parts, or edit existing ones. See section Non-Text Messages, for more details.

8.4 Options After Editing the Message

When you finish editing a message and exit the editor, af will display a summary of the message's headers and prompt you with

Send, Edit, Attachments, Check spelling, List or Forget?

It is a very good idea to check the headers that af displays, to make sure that you are sending the message to the people you think you are, that the subject is appropriate, and so on. Your options at this point are:

`Send'

Confirm the message, and give it to the system for delivery.

`Edit'

Edit the message again, this time inserting the headers and any signature into the text for you to edit.

`Attachments'

Enter compose mode on the message you're composing, allowing you to attach files or create new body parts. See section Non-Text Messages, for more details on compose mode.

`Check spelling'

Run an interactive spelling checker on the body of the message (but not the headers). This option may not be available on systems where there isn't an interactive spelling checker installed.

`List'

List the message to typeout, exactly as it will be given to the system for delivery. This can be useful for checking a message, as it is much quicker than starting up an editor.

`Forget'

Cancel the message and exit mail composition. There is no way to recover a message that you have "forgotten".

8.5 Composing MIME Mail.

Normally, electronic mail is composed in the `us-ascii' character set, which doesn't contain any accented or international characters. Af allows you to compose MIME messages in character sets other than `us-ascii', and to encode these messages so that they are not damaged in transit.

8.5.1 Internationalised Mail

The default `us-ascii' character set contains only the characters used in American English. If you want to include international characters such as accents, Hebrew, Cyrillic or so on, you will need to make af compose messages in another character set, for example the `iso-8859-1' character set that covers Western European languages, or the `iso-8859-8' character set for Hebrew text.

The `Content-Type:' header defines how the message is to be treated. For text messages it is usually set to `text/plain; charset=us-ascii'. If you change the value to `text/plain; charset=iso-8859-1', then the message should be interpreted as containing characters in the `iso-8859-1' character set.

The default-charset variable controls which character set you compose mail messages in by default. For example, to compose messages in the `iso-8859-1' character set, you would set the default-charset variable to `iso-8859-1'. Once you have done this, any messages you compose will be marked as containing characters from that character set. You can override this default by editing the message headers.

The person who installed af at your site may have already configured it to compose messages in the correct character set. In this case you don't need to do anything to make af handle internationalised messages correctly.

Even if you compose messages in an international character set by default, af will still mark them as `us-ascii' if they don't contain any international characters. So you should rarely, if ever, need to change the character set of a message yourself.

8.5.2 Encoding Mail

Unfortunately, the international mail network is often unable to handle messages including international characters properly. To avoid this, af allows you to encode the message, converting any international characters into special sequences of ascii characters. Unfortunately, this means that the message will not be properly readable if the recipient doesn't have a MIME-capable mail reader. Alternatively, you can choose not to encode your message, and risk it being damaged in transit. Which choice is better is entirely dependent on your local situation.

The default-text-encoding variable controls how international messages should be encoded. The possible values are:

`7bit'

The message consists only of us-ascii characters.

`8bit'

The message contains international characters, but will not be encoded.

`quoted-printable'

The message will be encoded with the MIME `quoted-printable' encoding. This is usually the best encoding for internationalised text messages.

`base64'

The message will be encoded with the MIME `base64' encoding. This encoding is not normally a good choice for text, unless the text consists mostly of international characters.

The person who installed af at your site may have already configured it to encode internationalised messages in the best way. In this case you don't need to do anything to make af encode internationalised messages correctly.

8.5.3 International Headers

Af handles international characters in headers by encoding them in a special form, to avoid damage in transit. This is done automatically whenever af encounters an international character in a header.

Sometimes you will encounter comments like `(**iso-8859-8**)' in a header containing international characters. If you are viewing the header, this means that the header contains international characters in a character set that isn't listed in the viewable-charsets variable, and is a warning that af can't correctly display the characters, what you see is a "best effort".

If you are editing the header, then such a comment means that the character set is not the one you are composing the message in. In this case the af will preserve the character set by assuming that any international characters after the comment contain characters in the specified character set. You can change this by adding another comment, such as `(**iso-8859-1**)' to force the character set back to your preference.

If you often need to type international characters into mail, the command minibuffer-set-iso-keys may be useful. This command sets the allow-meta-bindings variable (see section Keymaps), and then binds all the meta characters found in the `iso-8859-*' character sets to insert themselves when typed in the minibuffer. This is often useful when composing headers in languages other than English.

8.5.4 Non-Text Messages

There are two ways to compose non-text messages with af. You can use the send-file command, and af will ask you for the content type, disposition, and description if it recognises that the name of the file indicates that it isn't textual (see section Mime Configuration). Or you can use an argument to send-file to force af to ask you for the content type, disposition, and description, but that requires you to know the correct content-type. Af will encode the message with the `base64' encoding if required.

The second way is to supply an argument to the send-mail command, or select the Attachments option from the menu presented after editing the text of a message. In either case you will enter compose mode, where you can attach files or other newly-created data to the message.

Compose mode looks very much like Mail mode, except that each line displays one body part of the message. Each body part can be of any type. The commands for moving around the screen, and viewing and deleting messages that you use in mail mode all work the same way in compose mode.

You can use two commands to add a body part in compose mode. One is a or M-a (attach-file). This prompts for a file name, a Content-Type (which may be defaulted from the file name), and a Content-Description, a textual description of the new body part. The contents of the file are encoded if required, and then used to create a new body part, which is inserted into the compose buffer before point.

The second command is c or M-c (compose-body-part). This prompts for a Content-Type, Disposition, and Description, and then uses an external program to create the body-part. The system needs to be properly configured to create MIME mail, or af won't be able to find out which program to use, and will give an error message, unless the Content-Type is textual, in which case it will use an editor to create the content.

In a similar way you can use e or M-e (edit-body-part) to edit the contents of a body part. Again, if the system is not configured to do this then af will give an error message for body part's that aren't textual.

You can edit the description of a body part using d or M-d (edit-description). You will be presented with the description of the body part in the minibuffer for editing.

You can also edit the top-level headers of the composition using h or M-h (edit-headers). This presents the headers of the message in an editor, just as when creating a non-MIME message. Any text other than headers in the file will be discarded, and a warning given.

When you've finished attaching files or composing body parts, then use C-M-c or q (save-composition-and-exit). This will exit compose mode, creating a new message from the body parts. You will be prompted for the disposition of the message, just as for any other message you've edited.

If you want to abandon your changes in compose mode, then use C-M-x (abandon-composition-and-exit). You will be asked if you really want to lose all your changes, and if you confirm, then compose mode will exit, ignoring all your changes.

If you entered compose mode by supplying an argument to the send-mail command, then the entire command will be aborted. Otherwise you will be prompted for the disposition of the message normally.

8.6 Variables Related to Composing Mail

There are a number of configuration variables related to composing and sending mail. Here is a summary of each variable and its effects:

addresses-to-ignore

Contains mail addresses which are to be ignored in group replies, typically your mail addresses on this and other machines. Default is unset.

ask-bcc

If set to true, then af will prompt for addresses to send blind copies to whenever you send mail. Default is false.

ask-cc

If set to true, then af will prompt for addresses to send copies to whenever you send mail. Default is false.

auto-fold-headers

If set to false, then af will not automagically fold long header lines onto several lines. Default is true.

compose-line-format

Defines the format of each of the body part lines in compose mode, the the same way as header-line-format defines the format of header lines in mail mode. Defaults to %a %t %k %S.

copy-on-reply

If set to false, af will never quote the text of the original message in the initial text of a reply. If set to true af will always quote the original message. If set to ask then you will be asked each time you reply to a message. Defaults to ask.

copy-preface

If set, the value is inserted before any text copied into the body of a replying or forwarding message. Any conversion characters in the value are expanded to their full value. A good value for this variable might be `On %D, %o wrote:'. Default is unset.

copy-prefix

The prefix for lines of text copied into an outgoing message when replying or forwarding a message. Defaults to `> '.

edit-initial-headers

It set to true, headers will be placed into the file to edit the first time you edit a message. When set to accept, headers will not be displayed for editing, but any valid headers typed at the start of the message will be accepted. Defaults to false.

edit-initial-signature

When set to true, any signature will be placed into the editor the first time you edit a message. Defaults to false.

edit-reply-address

When set to true, allows editing of the destination address when replying to mail. It is intended for use by people at sites with inadequate mail configurations. Defaults to false.

editor

The program to use to edit messages. Defaults to the value of the VISUAL or EDITOR environment variable.

forward-subject

If set, the value is used as a default for the subject of a forwarded or resent message. Any conversion characters in the value are expanded to their full value. A good value for this variable might be `%s (Forwarded message from %o)'. Default is unset.

headers-to-copy

A colon separated list of header names which should be included in text copied into an outgoing message when replying to or forwarding a message. Default is unset (no headers are copied).

multiple-reply-warning

If set to true, then af will prompt for confirmation if you try to reply to a message that you have already replied to. Defaults to false.

organization

If set, the the value is inserted into an `Organization:' header in the outgoing message. Defaults to the value of the ORGANIZATION environment variable.

outbound-folder

Gives the name of a folder in which to store a copy of all messages sent with af. The messages are stored with headers (except for a `Message-ID:') in the named folder, immediately after being sent. Default is unset (don't save messages).

outbound-threshold

Gives the maximum number of lines in a message body which should be silently saved to the outbound-folder. Af will prompt for confirmation that a message longer than this limit is to be saved. If set to 0 then all outbound messages will be saved regardless of length. Default is 100 lines.

preserve-cc-in-group-reply

If set to true, any Cc: header on a message is duplicated in an outgoing group-reply to that message, so that recipients of a carbon-copy will also receive the reply. Default is true.

real-name

Your real name, for inclusion in `From:' headers. Defaults to the value of the NAME environment variable, or your details stored in the file `/etc/passwd'.

reply-address

Contains a mail address which forms a `Reply-To:' header in all messages you send, so that replies to those messages will be sent to the address you specify. Default is unset.

signature-file

The name of your signature file. If the value begins with `ask:', then you will be prompted for the signature file to use, with any value after the `ask:' being used as a default. Defaults to `.signature' in your home directory.

signature-separator

A string which will be printed on a line between the body of an outgoing message and any automatically-included signature. If not set, then no separator line is printed. Defaults to `--'.

spell-checker

The program to use to interactively spell-check the body of a message before you send it. Defaults to `ispell -x' if ispell is installed on your system, or unset otherwise.

8.7 Defining and Using Mail Aliases

Af will let you set mail aliases. These are short mnemonic names which stand for mail addresses or groups of mail addresses. Like many other mail programs, af expands aliases when they occur in the `To:', `From:', `Cc:', `Bcc:', and `Reply-to:' headers.

The easiest way to define a mail alias within af is with C-x C-a (set-alias). Af will prompt for the alias and then the real name and the full address of the alias, defaulted from the From: header of the current message. If the alias already exists, af will prompt for confirmation that you want to change the alias. (2)

C-x C-a sets the alias internally; but it also writes the alias to the file `.afalias' in your home directory. This file is read whenever af starts up, so your aliases will be available next time you run af. You can also edit your `.afalias' file directly; each line should either be blank, be a comment beginning with `;', or look like:

alias:Real Name:addresses

alias is the name of the alias; Real Name is the real name of the person or group the alias represents, and addresses stands for one or more mail addresses for alias to expand into. Separate multiple addresses with spaces or (preferably) commas. You can make an address list span more than one line by starting the second and subsequent lines with a space or tab, in the same way as mail headers are split over several lines of text.

For instance, to make afbug stand for af-bug@csv.warwick.ac.uk put in these lines:

; Set up the af-bug mailing list as an alias
afbug:Af Bug Mailing List:af-bug@csv.warwick.ac.uk

Af expands aliases as soon as you finish editing the mail file, so you can check that the alias expanded properly by looking at the list of headers that af shows you when it prompts for the disposition of the mail.

9. Running Commands by Name

The af commands that are used often or that must be quick to type are bound to keys--short sequences of characters--for convenient use. Other af commands that do not need to be brief are not bound to keys; to run them, you must refer to them by name.

A command name is, by convention, made up of one or more words, separated by hyphens; for example, af-version or list-commands. The use of English words makes the command name easier to remember than a key made up of obscure characters, even though it is more characters to type.

The way to run a command by name is to start with M-x, type the command name, and finish it with RET. M-x uses the minibuffer to read the command name. RET exits the minibuffer and runs the command. The string `M-x' appears at the beginning of the minibuffer as a prompt to remind you to enter the name of a command to be run. See section The Minibuffer, for full information on the features of the minibuffer.

Note that next-line is the same command that you invoke with the key C-n. You can run any af command by name using M-x, whether or not any keys are bound to it.

If you type C-g while the command name is being read, you cancel the M-x command and get out of the minibuffer, ending up at top level.

To pass a numeric argument to the command you are invoking with M-x, specify the numeric argument before the M-x. M-x passes the argument along to the command it runs. The argument value appears in the prompt while the command name is being read.

Normally, when describing a command that is run by name, we omit the RET that is needed to terminate the name. Thus we might speak of M-x af-version rather than M-x af-version RET. We mention the RET only when there is a need to emphasise its presence, such as when we show the command together with following arguments.

M-x works by running the command execute-extended-command, which is responsible for reading the name of another command and invoking it.

10. Help

Af provides extensive help features accessible through the command C-h (help-command). C-h is a prefix key that is used only for documentation-printing commands. The characters that you can type after C-h are called help options. One help option is C-h; that is how you ask for help about using C-h. To cancel, type C-g.

C-h C-h displays a list of the possible help options. Typing C-h again displays a longer list of the options, each with a brief description.

10.1 Documentation for a Key

The most basic C-h options are C-h c (describe-key-briefly) and C-h k (describe-key). C-h c key prints in the echo area the name of the command that key is bound to. For example, C-h c C-n prints `next-line'. Since command names are chosen to describe what the commands do, this is a good way to get a very brief description of what key does.

C-h k key is similar but gives more information: it displays the documentation string of the command as well as its name. This is too big for the echo area, so typeout is used for the display.

10.2 Help by Command, Mode or Variable Name

C-h f (describe-function) reads the name of an af command, keyboard macro or afl function using the minibuffer, then displays that function's documentation string to typeout. For example,

C-h f delete-message RET

displays the documentation of delete-message. This is the only way to get the documentation of a command that is not bound to any key (one which you would normally run using M-x).

C-h v (describe-variable) is like C-h f but describes configuration variables instead of commands and functions (see section Variables).

C-h m (describe-mode) is also like C-h f but describes major and minor modes rather than commands and functions (see section Major Modes).

10.3 Help for Lists of Things

C-h b (list-bindings) displays the current key bindings active in af to typeout. The global bindings are shown first, and then those for each of the major modes. This be quite a long list, but might be useful as a wall chart.

C-h d (list-commands) lists all of af's commands, with their bindings, to typeout. Bindings which only take effect in mail mode are listed in square brackets (eg `[M-s]'). Bindings for typeout mode only are listed in braces (eg `{C-f}'). Bindings that are minibuffer-specific are listed in parentheses (eg `(SPC)').

C-h C-f lists all of the afl functions to typeout. Only those functions which are not also commands are listed. This is quite a short list at present (see section Afl).

C-h C-k list-kbd-macros lists all of the named keyboard macros, with their values, to typeout (see section Keyboard Macros).

C-h C-v lists all of the af configuration variables, with their values, to typeout (see section Variables).

C-h C-a (list-aliases) lists all the mail aliases you have defined to typeout, along with their values (see section Defining and Using Mail Aliases).

10.4 Apropos

A more sophisticated sort of question to ask is, "What is relevant for working with messages?" To ask this question, type C-h a message RET, which displays a list of all command, function, keyboard macro and variable names that contain `message', including open-message, print-message, and so on. The bindings for each command is listed with that command, in the same format as list-bindings (see section Help for Lists of Things). Similarly the values of keyboard macros and variables are listed. The a in C-h a stands for `Apropos'; C-h a runs the command command-apropos.

Because C-h a looks only for things whose names contain the string which you specify, you must use ingenuity in choosing the string. If you are looking for commands for reading messages and C-h a read-message RET doesn't reveal any, don't give up. Try just read, or just message. Be persistent. Also note that you can use a regular expression as the argument, for more flexibility (see section Syntax of Regular Expressions).

Here is a set of arguments to give to C-h a that covers many classes of af commands, since there are strong conventions for naming the standard af commands. By giving you a feel for the naming conventions, this set should also serve to aid you in developing a technique for picking apropos strings.

char, line, word, region, list, buffer, window, file, mode, beginning, end, forward, backward, next, previous, up, down, search, goto, kill, delete, mark, insert, yank, case, set, what, view, describe, default, print, page, message, send.

10.5 Other Help Commands

C-h i (info) runs the Info program, which is used for browsing through structured documentation files. The entire af manual is available within Info. Type h after entering Info to run a tutorial on using Info.

If something surprising happens, and you are not sure what commands you typed, use C-h l (view-lossage). C-h l prints the last 100 command characters you typed in. If you see commands that you don't know, you can use C-h c to find out what they do.

Af has numerous major modes, each of which redefines keys and alters the behaviour of some commands. C-h j (describe-major-mode) prints documentation on the current major mode to typeout.

The other C-h options display assorted useful information. C-h n (view-af-news) displays information on recent changes to af. C-h C-w (display-no-warranty) displays details on the complete absence of warranty for af. C-h C-c (describe-copying) displays the conditions you must obey in distributing copies of af.

11. More Ways to Handle Messages

You can use af to handle messages in more sophisticated ways than by simply displaying them; You can save messages to folders, print a hardcopy of them, pipe them into commands, and more.

Most commands that deal with a single message also have similar commands which deal with a set of messages, selected either by using the mark and region (see section The Mark and the Region), or with tags (see section Tags). These similar commands will be covered later.

11.1 Saving Messages

Af allows you to save messages by copying them to folders. One way to save a message to a folder is to use M-+ (save-message). Af will prompt you for the folder name, offering a default based on the sender's mail address (see section Folder Handling). The message will be appended to the folder you specify, so any messages already in the folder will be left unchanged.

When the message is saved to a folder, the original copy is not deleted. an `S' will appear in the system tags of the message, so that you know you have saved it.

M-+ can take a numeric argument, in which case it only saves the body of the message, not the headers. Obviously the file containing a saved copy of the message body shouldn't be read with af; it isn't a mail folder. Saving the body of a message can be useful when the body is input for some other program (such as a patch file).

When you save message bodies, any MIME encoding is removed from the body parts. Partial messages are rebuilt before they are saved. Af will only save a rebuilt message once, no matter how many parts of the message are included in the messages to save. This is convenient when you are saving a set of message bodies.

11.2 Printing Messages

Af allows you to print a hardcopy of a message to a spooler. The program to use to print hard copies of mail is defined by the print-command variable for textual body parts, or the mime-printer variable for non-textual body parts.

To print a hardcopy of a message, use C-p (print-message). Af will usually prompt you to confirm that you want a hardcopy of the message; this feature can be turned off by setting the variable confirm-print to false. After you print the message, a `P' will appear in the system tags of the message, to remind you that you have printed it.

Normally, the mail headers listed in the headers-not-displayed variable will not be printed when you print a message. A positive numeric argument makes af print all the message's headers, while a negative argument makes af skip all headers, and only print the body of the message.

When you print messages, any MIME encoding is removed from the body parts. Partial messages are rebuilt before they are printed. Af will only print a rebuilt message once, no matter how many parts of the message are included in the messages to print. This is convenient when you are printing a set of message bodies.

If you print a non-text message, af will use an external program to print any non-text body parts. For this reason, non-textual body parts may not appear in the correct sequence with textual body parts. If you print a non-text body part, and af doesn't know how to handle it, it will assume that you know what you are doing, and print it anyway. The results may well be less than ideal if you print an image or audio body part.

11.3 Piping Messages

Af allows you to pipe a message into the standard input of a system command. To do this use M-| (pipe-message). Af will prompt for the command to pipe the message into, and then execute the command. Any output from the command will be displayed on the screen. Once the command has completed, af will wait for you to press a key before continuing; giving you a chance to read any output from the command.

Normally, the mail headers listed in the headers-not-displayed variable will not be piped into the command. A positive numeric argument makes af pass all the message's headers to the command. A negative argument makes af skip all headers, and only pass the body of the message to the command.

When you pipe messages into a command, any MIME encoding is removed from the body parts. Partial messages are rebuilt before they are piped into the command. Af will only pipe a rebuilt message once, no matter how many parts of the message are included in the messages to pipe. This is convenient when you are piping a set of messages into a command

11.4 Other Message-Handling Commands

C-x j

Create a new buffer containing the body parts of the current message. (open-body-parts).

C-x 4 j

Create a new buffer in the other window containing the body parts of the current message. (open-body-parts-other window).

C-x C-e

Edit the text of the current message.

If a message contains one or more body parts, or an encapsulated message, then af can create a buffer containing each body part as a separate message. Similarly, af can rebuild the original message from a single part if all the parts are present in the current buffer. Or if the message is an old, pre-MIME, mail digest, then af is still capable of extracting the parts of the message.

To extract or rebuild the parts of a message body use C-x j (open-body-parts) or C-x 4 j (open-body-parts-other-window). This will create a new buffer containing the extracted or rebuilt messages, and switch to that buffer. The buffer will not be visiting any folder, and so will not be saved to a folder when you exit af.

The new buffer will be in Body Parts minor mode, which means that the body-part-line-format variable will be used to display each message in the buffer, rather than the header-line-format. This allows you to show different details for body parts to those you show for messages.

Af will allow you to edit a message in a folder with C-x C-e (edit-message). Af will save the message to a temporary file and execute an editor to edit the file. When you finish editing, af will read back the temporary file, and update its internal copy of the message. When you save the buffer, the modified message will be written to the folder.

With an argument, C-x C-e will only pass the body of the message to the editor, and will preserve the message header unchanged.

11.5 Running Shell Commands from Af

Af has commands for running inferior shell processes, displaying the result to typeout or letting the inferior process take control of the terminal.

M-! cmd RET

Run the shell command line cmd and display the output to typeout (shell-command-to-typeout).

C-x ! cmd RET

Start an inferior shell to run the command line cmd. The inferior shell will take control of the terminal until it exits. (shell-command)

M-x shell

Run an inferior shell interactively under af. The inferior shell will take control of the terminal until it exits (shell).

12. The Mark and the Region

Many af commands operate on an arbitrary contiguous part of the current buffer. To specify the text for such a command to operate on, you set the mark at one end of it, and move point to the other end. The text between point and the mark is called the region.

You can move point or the mark to adjust the boundaries of the region. It doesn't matter which one is set first chronologically, or which one comes earlier in the buffer. Once the mark has been set, it remains where you put it until you set it again at another place. Each af buffer has its own mark, so that when you return to a buffer that had been selected previously, it has the same mark it had before.

Many commands that insert messages (or text in the minibuffer), such as C-y (yank) and C-x i (insert-file), position point and the mark at opposite ends of the inserted text, so that the region contains the text just inserted.

12.1 Setting the Mark

There are two commands to set the mark in a buffer. These commands work identically in mail buffers, typeout, and the minibuffer.

C-SPC or C-@ or M-SPC

Set the mark where point is (set-mark-command).

C-x C-x

Interchange mark and point (exchange-point-and-mark).

For example, suppose you wish to save part of a mail buffer to a folder, using the C-x + (save-region) command, which operates on the messages in the region. You can first go to the beginning of the messages to be saved, type C-SPC to put the mark there, move to the end, and then type C-x +. Or, you can set the mark at the end of the text, move to the beginning, and then type C-x +.

The most common way to set the mark is with the C-SPC command (set-mark-command). This sets the mark where point is. Then you can move point away, leaving the mark behind.

Af doesn't show you where the mark is located. You have to remember. The usual solution to this problem is to set the mark and then use it soon, before you forget where it is. Alternatively, you can see where the mark is with the command C-x C-x (exchange-point-and-mark) which puts the mark where point was and point where the mark was. The extent of the region is unchanged, but point is now at the previous position of the mark.

C-x C-x is also useful when you are satisfied with the position of point but want to move the mark; do C-x C-x to put point at that end of the region, and then move it. A second use of C-x C-x, if necessary, puts the mark at the new position with point back at its original position.

There is no such character as C-SPC in ASCII; when you type SPC while holding down CTRL, what you get on most ordinary terminals is the character C-@. This key is actually bound to set-mark-command. But unless you are unlucky enough to have a terminal where typing C-SPC does not produce C-@, you might as well think of this character as C-SPC. If you are on a terminal where C-SPC doesn't work, you could use M-SPC instead; it is also bound to set-mark-command.

12.2 Operating on the Region

Once you have a region in a mail buffer, here are some of the commands that operate on the region. Note that these commands all have the word region in their names.

C-w

Kill the messages in the region (kill-region). See section Killing and Yanking.

M-w

Copy the region into the kill buffer as if it had been killed (copy-region-as-kill). See section Killing and Yanking.

C-x + folder RET

Save the messages in the region into folder (save-region). This command handles the argument in the same way as M-+ (save-message). See section Saving Messages.

C-x p

Print a hardcopy of all the messages in the region (print-region). This command handles the argument in the same way as M-p (print-message). See section Printing Messages.

C-x | command RET

Pipe the messages in the region into a single instance of the shell command command (pipe-region). Again, this command handles the argument in the same way as M-| (pipe-message) See section Piping Messages.

12.3 Using the Region in Typeout

Typeout is only used to display text, and not to edit or delete anything. For that reason, there are no commands which operate on the region available in typeout. It is still possible to set the mark in the typeout buffer; but the only thing you can do with it is to use C-x C-x (exchange-point-and-mark) to move back to the mark.

12.4 Using the Region in the Minibuffer

The minibuffer supports the mark and region, but only a subset of the commands which are available in a mail buffer to process the region. Since the minibuffer deals with text rather than messages, the region delimits a block of text in the minibuffer.

Just like any other buffer, the minibuffer has its own mark. The mark is not preserved between uses of the minibuffer, or when you move between minibuffer history entries, so you will usually need to set the mark in the minibuffer before you use it.

There are two commands which operate on the region in the minibuffer:

C-w

Delete the text in the region (kill-region). See section Killing and Yanking in the Minibuffer.

M-w

Copy the text in the region into the kill buffer as if it had been killed (copy-region-as-kill). See section Killing and Yanking in the Minibuffer.

13. Tags

The region allows you to conveniently process blocks of messages, but what if you want to handle an arbitrary set of messages? Af uses tags to let you mark a set of messages and process them quickly.

Tags are only available in mail buffers; they wouldn't make any sense in the minibuffer or the mail buffer.

13.1 Tag Concepts

A tag is a single-character mark which is set on a message. A message can have any number of tags set on it. There are two types of tags: system tags which are automatically set by af for various reasons, and user tags which you can set and unset yourself. You cannot set or remove system tags, but you can use them in exactly the same ways as you can user tags.

There are many commands which operate on tagged messages; and you can use either single tags or a combination of them to specify a set of messages to operate on. Tags are central to advanced use of af; they provide a way for you to easily set up a set of messages, which you can then work with in several ways.

13.1.1 System Tags

The system tags are all upper-case letters, which give information on the status of messages. The possible system tags are:

`D'

The message has been marked for deletion with M-x delete-message.

`E'

Af found an error in the message's headers when the folder was visited. This is not normally critical, but you may not be able to reply to the message; or it may even (rarely) indicate a corrupted folder.

`F'

You have forwarded or bounced the message (see section Composing and Sending Mail).

`M'

The message is in MIME format; and isn't text that you can display on your terminal. When you read the message, af will display each body part separately if there are more that one, and will use external commands to display any non-textual body parts. If af doesn't know how to display a given body part then it will ask you what to do.

`T'

The message is in MIME format, and is textual, but is either not plain text, or is in a character set not listed in the viewable-charsets variable. When you read the message then it will be displayed as usual, which may not be ideal.

`N'

The message is new; the message has arrived since last time you visited the folder and hasn't yet been read.

`O'

The message is old but unread. The message was present last time you visited the folder, but you haven't read it yet.

`P'

A hardcopy of the message has been printed to a spooler.

`R'

You have replied to the message (see section Composing and Sending Mail).

`S'

The message has been saved to a folder.

13.1.2 User Tags

The user tags you can use are the lower case letters, and `+'. `+' is sometimes called the default tag, since it is the default whenever you are asked about tags.

Normally, the user tags are considered transient, and are lost when you exit af. If you would like to keep some or all of the tags across mail reading sessions, then you can set the variable persistent-tags to the tags you would like to be preserved. For example, setting the value of persistent-tags to `aeiou', would mean that the user tags `a', `e', `i', `o', and `u' will be preserved whenever you save a folder, but the other tags will be lost. You cannot make the default tag (`+') persistent.

13.1.3 Tag Lists

Whenever you are setting or removing tags, af will expect you to enter a tag list. Tag lists consist of one or more user tags, which may be separated by spaces. For example, `abcd' and `a b c d', both specify a list of four tags, `a', `b', `c', and `d'.

Whenever you are prompted for a tag list, you can just hit return to use the default tag (`+').

13.1.4 Tag Expressions

When you are using tags to define a set of messages to apply some command to, you will be expected to enter a tag expression. That tag expression is then compared to the tags of each message in the buffer, and the command will operate on the message if the expression is true for that message. We often refer to the messages which match a tag expression as the tagset.

Tag expressions are made up of sub-expressions, which may be system or user tags. You can use operators to combine subexpressions, and build arbitrarily complex expressions. Any tag will evaluate to true if the message being checked has that tag set, false otherwise. The possible operators are, in decreasing precedence:

`( … )'

Override any precedence, and force the contents of the parentheses to be evaluated as a unit. `(a)' is equivalent to `a'.

`!'

Logical not; true if the expression it precedes is false. `!a' is true if `a' is not set.

`&'

Logical and; true if both sides of the expression are true. `a & b' is true if both `a' and `b' are set.

`^'

Logical exclusive-or; true if only one side of the expression is true. `a ^ b' is true if either `a' or `b' is set, but not both.

`|'

Logical or; true if either side of the expression is true. `a | b' is true if either `a' or `b' is set, or if both are set.

Here are some examples of tag expressions, with their meanings.

a & b | c & d

True if `a' and `b' are both set, or `c' and `d' are both set.

a & (b | c) & d

True if `a', `d', and either `b' or `c' are set.

a & !b

True if `a' is set and `b' is not set.

a & !(b | c)

True if `a' is set and neither of `b' or `c' are set.

13.2 Setting and Removing Tags

C-t t tags RET

Set tags on the current message (tag-message).

C-t u tags RET

Remove tags from the current message (untag-message).

C-t r tags RET

Remove tags from all the messages in the buffer (remove-tag).

M-x tag-thread tags RET

To tag the current message, use C-t t. You will be prompted for the tags to set; simply type in the tag list to set. The tags in the list will be added to those already set on the message.

To remove one or more tags from the current message use C-t u. You will be prompted for the tags to remove. The tags you specify will be removed from the message's tags.

To remove one or more tags from all the messages in the buffer, use C-t r. You will be prompted for the tags to remove. The tags you specify will be removed from all the messages in the buffer.

The experimental command M-x tag-thread tags RET tags all the messages in a thread of conversation, a set of messages which are all sent as replies to other related messages. (This is similar to the concept of threading found in most news readers.) This command works, but at the moment there are too many mail composers which don't include references to messages when they generate a reply, so that the threads are often broken by replies which af can't detect as a part of the thread. It is probably best not to use this command for the moment, unless you can be sure that all the participants are using a mail composer that does include references to messages in replies.

There is also a command to set tags on messages which match search criteria. See section Tagging Matching Messages.

13.3 Using Tags

Once you have tagged messages in the region, you can use single tags or tag expressions to operate on them. Here are some of the ways in which you can operate on tagged messages:

C-t C-k tag-expr RET

Kill the messages in the tagset defined by tag-expr (kill-tagset). See section Killing and Yanking.

C-t w tag-expr RET

Copy the messages in the tagset into the kill buffer as if they had been killed (copy-tagset-as-kill). See section Killing and Yanking.

C-t + tag-expr RET folder RET

Save the messages in the tagset into folder (save-tagset). This command handles the argument in the same way as M-+ (save-message). See section Saving Messages.

C-t p tag-expr RET

Print a hardcopy of all the messages in the tagset (print-tagset). This command handles the argument in the same way as M-p (print-message). See section Printing Messages.

C-t | tag-expr RET folder RET

Pipe the messages in the tagset into a single instance of the shell command command (pipe-tagset). Again, this command handles the argument in the same way as M-| (pipe-message) See section Piping Messages.

14. Killing and Yanking

Killing means erasing messages (or text in the minibuffer) and copying them into the kill ring, from which it can be retrieved by yanking it. Some systems use the terms "cutting" and "pasting" for these operations.

The commonest way of moving or copying messages within af is to kill them, and later yank them elsewhere one or more times. This is very safe because af remembers several recent kills, not just the last one. It is versatile, because the many commands for killing messages can also be used for moving them.

Af has only one kill ring for all mail buffers, so you can kill messages in one buffer and yank them in another buffer. The minibuffer has a separate kill ring for storing text.

All commands which delete messages from the buffer save it in the kill ring (even delete-message when you save the folder) so that you can move or copy it to other parts of the buffer. These commands are known as kill commands.

14.1 Killing Messages

There are two ways to add messages to the kill ring; killing and copying. Killing means deleting messages and adding them to the kill ring, while copying means copying messages to the kill ring without deleting them.

There are commands to kill a single message (see section Deleting Messages), commands to kill or copy the region (see section Operating on the Region), and commands to kill or copy a tagset (see section Using Tags).

14.2 Yanking Messages

All killed messages are recorded in the kill ring, a list of blocks of messages that have been killed. There is only one kill ring, shared by all mail buffers, so you can kill messages in one buffer and yank them in another buffer. This is the usual way to move messages from one folder to another (See section Saving Messages, for another way).

The command C-y (yank) reinserts the most recently killed messages. It leaves the cursor at the end of the text. It sets the mark at the beginning of the text (see section The Mark and the Region).

C-y can take a numeric argument, in which case it select an earlier kill to yank, rather than the last (see section Yanking Earlier Kills).

14.3 Appending Kills

Normally, each kill command pushes a new entry onto the kill ring. However, two or more kill commands in a row combine their messages into a single entry, so that a single C-y yanks all the messages as a unit, just as they were before they were killed.

Thus, if you want to yank messages as a unit, you need not kill all of them with one command; you can keep killing line after line, until you have killed them all, and you can still get them all back at once.

Commands that kill forward from point add the messages onto the end of the previous killed messages. Commands that kill backward from point add them onto the beginning. This way, any sequence of mixed forward and backward kill commands puts all the killed messages into one entry without rearrangement. Numeric arguments do not break the sequence of appending kills.

14.4 Yanking Earlier Kills

To recover killed messages that are no longer the most recent kill, use the M-y command (yank-pop). It takes the messages previously yanked and replaces them with the messages from an earlier kill. So, to recover the messages of the next-to-the-last kill, first use C-y to yank the last kill, and then use M-y to replace it with the previous kill. M-y is allowed only after a C-y or another M-y.

You can understand M-y in terms of a "last yank" pointer which points at an entry in the kill ring. Each time you kill, the "last yank" pointer moves to the newly made entry at the front of the ring. C-y yanks the entry which the "last yank" pointer points to. M-y moves the "last yank" pointer to a different entry, and the text in the buffer changes to match. Enough M-y commands can move the pointer to any entry in the ring, so you can get any entry into the buffer. Eventually the pointer reaches the end of the ring; the next M-y moves it to the first entry again.

M-y moves the "last yank" pointer around the ring, but it does not change the order of the entries in the ring, which always runs from the most recent kill at the front to the oldest one still remembered.

M-y can take a numeric argument, which tells it how many entries to advance the "last yank" pointer by. A negative argument moves the pointer toward the front of the ring; from the front of the ring, it moves "around" to the last entry and continues forward from there.

Once the messages you are looking for are brought into the buffer, you can stop doing M-y commands and it will stay there. They are a copy of the kill ring entry, so altering them in the buffer does not change what's in the ring. As long as no new killing is done, the "last yank" pointer remains at the same place in the kill ring, so repeating C-y will yank another copy of the same previous kill.

If you know how many M-y commands it would take to find the messages you want, you can yank those messages in one step using C-y with a numeric argument. C-y with an argument restores the text the specified number of entries back in the kill ring. Thus, C-u 2 C-y gets the next to the last block of killed messages. It is equivalent to C-y M-y. C-y with a numeric argument starts counting from the "last yank" pointer, and sets the "last yank" pointer to the entry that it yanks.

The length of the kill ring is controlled by the variable kill-ring-max; no more than that many blocks of killed messages are saved.

14.5 Killing and Yanking in the Minibuffer

The minibuffer is used for editing text rather than handling mail, so it has it's own kill ring. Any command which kills more than one character at a time (the word and line killing commands) will store the killed text in the minibuffer kill ring. You can yank the killed text back in exactly the same way as killed messages using C-y.

The minibuffer kill commands will append successive kills to the kill ring, just like the mail buffer kill commands do. If you use M-d (delete-word) three times, then the three words you kill will all be stored in a single kill ring entry; and a single yank will insert them into the text at point.

For simplicity, the minibuffer kill ring only has one entry, so you can't use M-y in the minibuffer. If many people find this a problem, then it may change at some point in the future.

15. Searching Mail Folders

Af has search commands, which allow you to search through the messages in a buffer and find those which match a regular expression. You can also search for messages which match a tag expression.

15.1 Searching for Regular Expressions

C-s regex RET

Search for regex (search-forward).

C-r regex RET

Search backward for regex (search-backward).

To do a search on a buffer (whether typeout or a mail buffer), use C-s or C-r. Af will prompt you for the regular expression to search for, and then the search takes place. If no messages match the regular expression then the search will fail with an error.

A second search immediately after the first, will not match the current message, so repeated searches will move through all the messages which match the regular expression. To make this more convenient, the search expression is defaulted to the last one you entered.

The search commands with a numeric argument will only search the headers of the messages. This is often convenient when (for example) looking for messages which are from a particular person.

With a negative numeric argument the search commands will only search the bodies of the messages. This can be useful when (for example) looking for messages which mention your machine's host name (which is included in the headers of all messages).

15.2 Searching for Tagged Messages

C-t C-s tagexpr RET

Search for messages matching tagexpr (tag-search-forward).

C-t C-r tagexpr RET

Search backward for messages matching tagexpr (tag-search-backward). 

To search for a message matching a tag expression use C-t C-s or C-t C-r. Af will prompt you for the tag expression (see section Tag Expressions) to search for; and then the search takes place. If no messages match the tag expression then the search will fail with an error.

Just as with regular expression searches, a second search immediately after the first, will not match the current message, so repeated searches will move through all the messages which match the tag expression. To make this more convenient, the search expression is defaulted to the last one you entered.

15.3 Tagging Matching Messages

As well as just searching for a regular expression and moving point to the first matching message, af can tag all the messages which match a regular expression. To do this use C-t s (search-and-tag). You will be prompted for the regular expression to search for, and the tags to set on the matching messages (see section Setting and Removing Tags). Once the search has finished, af will report how many messages were tagged.

With a numeric argument this command will only search the headers of the messages. This is often convenient when (for example) looking for messages which are from a particular person or mailing list.

15.4 Syntax of Regular Expressions

Regular expressions have a syntax in which a few characters are special constructs and the rest are ordinary. An ordinary character is a simple regular expression which matches that same character and nothing else. The special characters are `$', `^', `.', `*', `+', `?', `[', `]' and `\'. Any other character appearing in a regular expression is ordinary, unless a `\' precedes it.

For example, `f' is not a special character, so it is ordinary, and therefore `f' is a regular expression that matches the string `f' and no other string. (It does not match the string `ff'.) Likewise, `o' is a regular expression that matches only `o'. (When case distinctions are being ignored, these regular expressions also match `F' and `O', but we consider this a generalization of "the same string", rather than an exception.)

Any two regular expressions a and b can be concatenated. The result is a regular expression which matches a string if a matches some amount of the beginning of that string and b matches the rest of the string.

As a simple example, we can concatenate the regular expressions `f' and `o' to get the regular expression `fo', which matches only the string `fo'. Still trivial. To do something nontrivial, you need to use one of the special characters. Here is a list of them.

. (Full stop)

is a special character that matches any single character except a newline. Using concatenation, we can make regular expressions like `a.b' which matches any three-character string which begins with `a' and ends with `b'.

*

is not a construct by itself; it is a postfix operator, which means to match the preceding regular expression repetitively as many times as possible. Thus, `o*' matches any number of `o's (including no `o's).

`*' always applies to the smallest possible preceding expression. Thus, `fo*' has a repeating `o', not a repeating `fo'. It matches `f', `fo', `foo', and so on.

+

is a postfix character, similar to `*' except that it must match the preceding expression at least once. So, for example, `ca+r' matches the strings `car' and `caaaar' but not the string `cr', whereas `ca*r' matches all three strings.

?

is a postfix character, similar to `*' except that it can match the preceding expression either once or not at all. For example, `ca?r' matches `car' or `cr'; nothing else.

[ … ]

is a character set, which begins with `[' and is terminated by `]'. In the simplest case, the characters between the two brackets are what this set can match.

Thus, `[ad]' matches either one `a' or one `d', and `[ad]*' matches any string composed of just `a's and `d's (including the empty string), from which it follows that `c[ad]*r' matches `cr', `car', `cdr', `caddaar', etc.

You can also include character ranges in a character set, by writing two characters with a `-' between them. Thus, `[a-z]' matches any lower-case letter. Ranges may be intermixed freely with individual characters, as in `[a-z$%.]', which matches any lower case letter or `$', `%' or `.'.

Note that the usual regex special characters are not special inside a character set. A completely different set of special characters exists inside character sets: `]', `-' and `^'.

To include a `]' in a character set, you must make it the first character. For example, `[]a]' matches `]' or `a'. To include a `-', write `-' as the first or last character of the set. Thus, `[]-]' matches both `]' and `-'.

To include `^', make it other than the first character in the set.

[^ … ]

`[^' begins a complemented character set, which matches any character except the ones specified. Thus, `[^a-z0-9A-Z]' matches all characters except letters and digits.

`^' is not special in a character set unless it is the first character. The character following the `^' is treated as if it were first (`-' and `]' are not special there).

^

is a special character that matches the empty string, but only at the beginning of a line in the text being matched. Otherwise it fails to match anything. Thus, `^foo' matches a `foo' which occurs at the beginning of a line.

$

is similar to `^' but matches only at the end of a line. Thus, `xx*$' matches a string of one `x' or more at the end of a line.

\

has two functions: it quotes the special characters (including `\'), and it introduces additional special constructs.

Because `\' quotes special characters, `\$' is a regular expression which matches only `$', and `\[' is a regular expression which matches only `[', etc.

For the most part, `\' followed by any character matches only that character. However, there are several exceptions: two-character sequences starting with `\' which have special meanings. The second character in the sequence is always an ordinary character on its own. Here is a table of ``\'' constructs.

\{ … \}

is a postfix construct, similar to `*' except that it allows you to specify the number of times the preceding expression must be matched. So, for example, `ca\{3\}r' will match only the string `caaar'.

If you add a comma after the number of times the expression must be matched, then the expression must be matched at least as many times as you specified. So `ca\{2,\}r' will match the strings `caar', `caaar', `caaaar', and so on.

You can also add a maximum value after the comma, to specify a range of values. So `ca\{1,3\}r' will match only the strings `car', `caar' and `caaar'.

\|

specifies an alternative. Two regular expressions a and b with `\|' in between form an expression that matches anything that either a or b matches.

Thus, `foo\|bar' matches either `foo' or `bar' but no other string.

`\|' applies to the largest possible surrounding expressions. Only a surrounding `\( … \)' grouping can limit the scope of `\|'.

\( … \)

is a grouping construct that serves three purposes:

1. To enclose a set of `\|' alternatives for other operations. Thus, `\(foo\|bar\)x' matches either `foox' or `barx'.
2. To enclose a complicated expression for the postfix operators `*', `+' and `?' to operate on. Thus, `ba\(na\)*' matches `bananana', etc., with any (zero or more) number of `na' strings.
3. To mark a matched substring for later reference with `\N'.

\N

When you use `\( … \)' in an expression, you can look for another match for the exact same text that was matched inside the `\( … \)'. The two-character sequence `\N' will match the same text as was matched by the Nth use of `\( … \)'. The first nine uses are remembered, and are assigned the numbers `1' to `9'. So `\1' matches the text that was matched by the first use of `\( … \)'.

For example, `\([a-z]\)\1' matches any two consecutive lower case characters. The `\([a-z]\)' matches any lower case character, while the `\1' must match the same character.

If a use of `\( … \)' matches more than once, which often happens if it is followed by `*' or `+', only the last match is stored for use with `\N'.

Here is a moderately complicated regex, which you might use to find messages from the af-bug or af-user mailing lists.

^From:.*af-\(bug\|user\)@csv.warwick.ac.uk

15.5 Searching and Case

Searches in af normally ignore the case of the text they are searching through. Thus, if you specify searching for `foo', then `Foo' and `foo' are also considered a match. Regular expressions, and in particular character sets, are included: `[ab]' would match `a' or `A' or `b' or `B'.

If you set the variable case-fold-search to false, then all letters must match exactly, including case.

16. Sorting Messages

Af provides commands for sorting some or all of the messages in a mail buffer. Each command prompts for a sort order, and then sorts the selected messages into that order.

M-x sort-buffer order RET

Sort all of the messages in the buffer into order order.

M-x sort-region order RET

Sort the messages in the region into order order.

M-x sort-tagset order RET

Sort the messages in the tagset into order order. The messages in the tagset are sorted correctly, but messages which are not in the tagset are not moved at all.

The sort orders which af understands are:

(reverse-)address

The address of the sender of the message, sorted case-independent. Any full name or routing information in the address is ignored.

(reverse-)date

The date the message was sent (or received if the date sent isn't known), from oldest to newest (newest to oldest). The variable show-dates-in-local-time indicates whether dates should be compared as they are, or converted to your local time before they are compared (the default).

(reverse-)lines

The number of lines in the message, from smallest to largest (largest to smallest).

(reverse-)mailbox

The (reverse) order of the messages in the folder when it was read. Messages which were yanked into the buffer after it was read may appear anywhere in the buffer.

(reverse-)sender

The full name of the sender of the message, sorted case-independent.

(reverse-)status

The system tags of the message (see section System Tags). Messages with no system tags are sorted after those that have them.

(reverse-)subject

The subject of the message, sorted alphabetically and case-independent. The prefix 'Re: ' is handled specially in subject lines, so that replies to a message will sort immediately after (before) the original message.

(reverse-)tags

The user tags of the message (see section User Tags). Messages with no user tags are sorted after those that have them.

When you sort, messages which sort identically will remain in their original order, so sorting a buffer by date and then subject will result in the messages being in date order within each subject. This is often a good approximation to the "threading" that news readers such as trn do, and is handy for reading mailing lists.

Af does not sort your folders by default when it reads them. If you have a preference for reading buffers in a certain order, then you can set the configuration variable initial-buffer-sort to the sort ordering you prefer.

If you are familiar with other mail readers it is worth noting that af, unlike most other mail readers, saves the messages back to the folders in whatever order you sort them into. This can often be convenient when managing large folders.

17. Narrowing

Narrowing means focusing in on some portion of the buffer, making the rest temporarily inaccessible. The portion which you can still get to is called the accessible part. Canceling the narrowing, which makes the entire buffer once again accessible, is called widening. The amount of narrowing in effect in a buffer at any time is called the buffer's restriction.

Narrowing can make it easier to concentrate on a set of messages by eliminating clutter. It can also be used to restrict the range of operation of a search command or repeating keyboard macro.

C-x n

Narrow down to between point and mark (narrow-to-region).

C-t n expression RET

Narrow to the tagset defined by tag-expression (narrow-to-tagset).

C-x w

Widen to make the entire buffer accessible again (widen).

When you have narrowed down to a part of the buffer, that part appears to be all there is. You can't see the rest, you can't move into it (motion commands won't go outside the accessible part), you can't change it in any way. However, it is not gone, and if you save the folder all the inaccessible messages will be saved. The word `Narrow' appears in the mode line whenever narrowing is in effect.

One narrowing command is C-x n (narrow-to-region). It sets the current buffer's restrictions so that the messages in the region remain accessible but all messages before or after the region are inaccessible. Point and mark do not change.

Another narrowing command is C-t n (narrow-to-tagset). It prompts for a tag expression, and then set the current buffer's restrictions so that the messages in the tagset remain accessible but all the other messages are inaccessible.

The way to cancel narrowing is to widen with C-x w (widen). This makes all the messages in the buffer accessible again.

Narrowing can be very useful if you have a set of messages that you want to deal with independently of any others; such as all the messages from a mailing list, or from a certain user. Use C-t s (search-and-tag) to tag the messages you are interested in, and then use C-t n to narrow the buffer to only the tagged messages. When you are done with the messages, use C-x w to widen to buffer.

Another use for narrowing is when you are using tags to kill messages. You can use C-t n to narrow to the messages you intend to kill, and then conveniently scan the messages to make sure that you are happy to kill them all; using C-t u to untag any messages that you decide you want to keep. Finally you can either kill the tagset with C-t C-k (kill-tagset), or set the mark and point at opposite ends of the buffer and use C-w (kill-region). When you have finished, use C-x w to widen the buffer.

18. Folder Handling

The operating system stores data permanently in named files. We refer to files which contain mail messages as folders. In general, the mail messages you process with af come from a folder which will need to be updated to reflect any changes you make.

Unlike most mail readers, af can read in and process several folders simultaneously, which makes it convenient to store messages in several folders rather than leaving them all in your incoming mailbox. Af can actually be used quite conveniently as a small database by using folders to store related messages.

By default, af reads in your incoming mailbox when it starts. You can also use command-line arguments to tell af to read a different folder (see section Command Line Arguments). Or you can tell an existing af to read another folder and prepare a buffer containing a copy of the folder's messages. This is called visiting the folder. Af commands apply to the messages in the buffer; that is, to the copy inside af. Your changes appear in the folder itself only when you save the buffer back into the folder.

18.1 File Names

Most af commands that operate on a file require you to specify the file name. (Saving and reverting are exceptions; the buffer knows which file name to use for them.) You enter the file name using the minibuffer (see section The Minibuffer). Completion is available, to make it easier to specify long file names (see section Completion).

Many people prefer to keep all their mail folders in one directory; any folders saved from Usenet news (news folders) in another, and so on. To make this kind of operation more convenient, af interprets several characters specially if they appear at the start of the file name.

`~user'

user's home directory, or your home directory if user is not given. For example `~/foo' is the file `foo' in your home directory, while `~malc/foo' is the file `foo' in malc's home directory.

`%user'

user's incoming mailbox, or your incoming mailbox if user is not given. So `%' is the your incoming mailbox, while `%malc' is malc's incoming mailbox. Note that you won't usually have the necessary permission to read other people's incoming mailboxes.

`+'

Your folder directory, as specified by the variable folder. This is either extracted from the environment variable FOLDER, or defaulted to `~/Mail'. So `+af' will usually expand to `~/Mail/af'. You will need to create the directory yourself if you plan to use it; af won't create it for you.

`='

Your news folder directory, where saved news articles are stored, as specified by the variable news-folder. This is either extracted from the environment variable SAVEDIR, or defaulted to `~/News'. So `=Comp.mail.headers' will usually expand to `~/News/Comp.mail.headers'. Normally your news-reader will have created this directory for you.

`\'

A leading `\' in a file name is ignored, but it prevents any special meaning being taken from the next character in the filename. To start a file name with a `\', use `\\'. For example `\+af' is the file `+af' in the current directory.

Af has a current directory, sometimes referred to as the working directory, which is normally the directory you were in when you ran af. When you enter a folder name without a directory or a special prefix, the current directory is used. If you specify a directory in a relative fashion, with a name that does not start with either a special prefix or a slash, it is interpreted with respect to the current directory.

For example, if the current directory is `/home/malc/src/af' and you type just `foo', which does not specify a directory, it is short for `/home/malc/src/af/foo'. `../../.afrc' would stand for `/home/malc/.afrc'. `new/foo' would stand for the file name `/home/malc/src/af/new/foo'.

The command M-x pwd prints af's current directory, and the command M-x cd sets it (to a value read using the minibuffer). Af's current directory can only be changed by using the cd command.

18.2 Reading Folders Over a Network

Af supports a mail protocol called POP3, which allows you to read a folder over a network. This allows people to read mail on a small machine which doesn't have a mail system itself; you simply connect to the POP server where your mail is stored, and read your mail over the network.

Your incoming mailbox may live on a POP server, or you may have a separate POP3 mailbox somewhere. To access a POP3 folder use a filename beginning in a colon. Two forms of POP3 folder name are acceptable: `:server' or `:user@server'. The first form implies that the user name to use on server is the same one that you are currently logged in as; the second explicitly specifies the user name to use to connect to the POP server.

Once af has opened a connection to the POP server, you are prompted for a password. The password won't be echoed (for security reasons), and it won't be stored in the minibuffer history either. If you don't know your password then you will need to check with your system administrator to find out what it is. The password you use to log in is often worth trying.

Once af has read the POP folder (which may like some time; network connections are often slow), it displays the buffer as normal. However, the buffer containing the folder will have the POP3 minor mode set. This disallows several af features, notably killing and yanking messages in the buffer. The reason for this is that the POP3 protocol was designed to work well with existing mail readers, and it doesn't have the functionality to cope with af's method of folder management. (3) One way of working around this is to use a pending folder (see section Pending Folders).

18.3 Visiting Files

C-x C-f

Visit a folder (find-file).

C-x C-r

Visit a folder for viewing, without allowing changes to it (find-file-read-only).

C-x C-v

Visit a different folder instead of the one visited last (find-alternate-file).

C-x 4 C-f

Visit a file, in another window (find-file-other-window).

Visiting a folder means reading its contents into an af buffer so you can process them. Af makes a new buffer for each folder that you visit. We say that this buffer is visiting the folder that it was created to hold. Af constructs the buffer name from the folder name by throwing away the directory, keeping just the name proper. For example, a file named `/home/malc/Mail/af' would get a buffer named `af'. If there is already a buffer with that name, a unique name is constructed by appending `<2>', `<3>', or so on, using the lowest number that makes a name that is not already in use.

Each window's mode line shows the name of the buffer that is being displayed in that window, so you can always tell what buffer you are processing.

The changes you make with af commands are made in the af buffer. They do not take effect in the folder that you visited, or any place permanent, until you save the buffer. Saving the buffer means that af writes the current contents of the buffer into its visited folder. See section Saving Files.

If a buffer contains changes that have not been saved, we say the buffer is modified. This is important because it implies that some changes will be lost if the buffer is not saved. The mode line displays two stars near the left margin to indicate that the buffer is modified.

If the status of one or more messages in a buffer has changed (perhaps they were new before you read the folder), but no other changes have been made, then we say that the buffer is status modified. This is not so important as being modified, but some information will be lost if the buffer is not changed. The mode line displays two plus signs near the left margin to indicate that the buffer is status modified.

To visit a folder, use the command C-x C-f (find-file). Follow the command with the name of the folder you wish to visit, terminated by a RET.

The folder name is read using the minibuffer (see section The Minibuffer), with defaulting and completion in the standard manner (see section File Names). While in the minibuffer, you can abort C-x C-f by typing C-g.

If the specified folder does not exist and could not be created, or cannot be read, then you get an error, with an error message displayed in the echo area.

While af is reading the folder, it reports on how many messages it has read, usually after every five messages read. You can control how often af updates the count by setting the variable message-count-update. Once the folder has been read, af will print a message like this:

(Read 15 messages)

This confirms that af has successfully read the folder into a buffer. The message may also sometimes look like this:

(Read 15 messages; including 1 with bad headers)

This means that the folder was read successfully, but some of the messages in the folder had mail headers that af couldn't interpret properly. This is not usually critical, but some people prefer to know about errors in their mailboxes.

If the variable first-unread-message is set to true; then af will move point to the first message in the buffer that you haven't read, if there are any. Otherwise you will begin processing the buffer with point on the first message.

If you visit a file that is already in af, C-x C-f does not make another copy. It selects the existing buffer containing that file.

What if you want to create a new folder? Just visit it. Af prints `(New File)' in the echo area, but in other respects behaves as if you had visited an existing empty file. If you make any changes and save them, the file is created.

If you visit a file that the operating system won't let you modify, Af makes the buffer read-only, so that you won't go ahead and make changes that you'll have trouble saving afterward. You can make the buffer writable with C-x C-q (toggle-read-only).

Occasionally you might want to visit a folder as read-only in order to protect yourself from entering changes accidentally; do so by visiting the file with the command C-x C-r (find-file-read-only).

If you accidentally visit the wrong folder (perhaps you mistyped the file name), use the C-x C-v command (find-alternate-file) to visit the file you really wanted. C-x C-v is similar to C-x C-f, but it kills the current buffer (after first offering to save it if it is modified).

C-x 4 f (find-file-other-window) is like C-x C-f except that the buffer containing the specified file is selected in another window. The window that was selected before C-x 4 f continues to show the same buffer it was already showing. If this command is used when only one window is being displayed, that window is split in two, with one window showing the same buffer as before, and the other one showing the newly requested file. See section Multiple Windows.

18.4 Pending Folders

If you are reading mail via POP3, you may prefer to store your mail on your local machine, rather than on a POP server. You can do this conveniently with af by using a pending folder, a folder which contains two folders. The simplest way of doing this is to set the variable pending-folder to the name of the folder where all your incoming mail should be stored; say `+mailbox'. You must set the pending-folder variable in your `.afrc' file, or it will have no affect (see section Startup Files).

Once the pending-folder variable is set, it changes the behaviour of af when you start without specifying folders to read or addresses to send mail to. Instead of reading your incoming mailbox into its own buffer, it reads the folder specified in the pending-folder variable, and then appends the contents of your incoming mailbox to the buffer.

The buffer is considered to contain both the pending folder and your incoming mailbox; it will be resynchronised from both, and visiting either folder will switch to the buffer displaying the pending folder. When you save the pending folder your incoming mailbox will be cleared, ready for any new mail. If you don't save the buffer, both folders are left untouched.

One advantage of pending folders is that they are normal disk folders. If your incoming mailbox is read via POP3, using a pending folder will allow you to avoid the limitations of buffers in POP3 mode, and use all of af's features in the buffer. Once you've set it up, using a default pending folder is transparent enough that you'll probably never notice the difference.

You can also read a folder into an existing buffer, making the buffer pending on the folder you read. To do this use the command M-x read-pending-file. You will be prompted for the name of the file to make the buffer pending on. This has exactly the same effect as using the pending-folder variable.

18.5 Resynchronizing Buffers

As we have discussed, buffers are copies of the contents of a folder. So what can we do to handle new mail arriving in the folder after we have read it? The solution to this is resynchronizing the buffer, so that any new messages in the file on disk are incorporated into the buffer.

When a buffer is resynchronised, any new messages are appended to the buffer, and a message will appear in the echo area to let you know that new messages were read into the buffer.

Af will check for new messages whenever you try to save a buffer to disk. If there are new messages then the buffer is resynchronised, and the save will fail.

Af also checks for new mail in each folder you are visiting on a regular basis. If any buffers are out of date, then they are resynchronised. How often af does this is controlled by the resync-time variable, normally it is done every 600 seconds (10 minutes).

To manually check for new messages in a buffer's folder, use C-x r (resync-buffer). If the buffer needs to be resynchronised, then point will be set on the first new message.

18.6 Saving Files

Saving a buffer in af means writing its contents back into the folder that was visited in the buffer.

C-x C-s

Save the current buffer in its visited file (save-buffer).

C-x s

Save any or all buffers in their visited files (save-some-buffers).

M-x save-all-buffers

Save all modified buffers in their visited files.

M-~

Forget that the current buffer has been changed (not-modified).

C-x C-w

Save the current buffer in a specified file (write-file).

When you wish to save the folder and make your changes permanent, type C-x C-s (save-buffer). After saving is finished, C-x C-s displays a message like this:

(Wrote 12 messages)

If the selected buffer is not modified or status modified (no changes have been made in it since the buffer was created or last saved), saving is not really done, because it would have no effect. Instead, C-x C-s displays a message like this in the echo area:

(No changes need to be written)

The command C-x s (save-some-buffers) offers to save any or all modified buffers. It asks you what to do with each buffer. These questions are also asked if you exit af by typing C-x C-c. The buffer is automatically resynchronised before being saved, and if any new messages are found the save will be abandoned.

The command M-x save-all-buffers saves all modified buffers. This can be dangerous if you have several buffers, since you may not want to save one of them, so be careful!

If you have changed a buffer but you do not want to save the changes, you should take some action to prevent it. Otherwise, each time you use C-x s or C-x C-c, you are liable to save this buffer by mistake. One thing you can do is type M-~ (not-modified), which clears out the indication that the buffer is modified. If you do this, none of the save commands will believe that the buffer needs to be saved. (`~' is often used as a mathematical symbol for `not'; thus M-~ is `not', metafied.) Alternatively, you can cancel all the changes made since the file was visited or saved, by reading the text from the file again. This is called reverting. See section Reverting a Buffer.

If you wish to mark the buffer as visiting a different folder and save it right away, use C-x C-w (write-file). It prompts for the new folder, marks the buffer as visiting that folder, and saves the buffer. If the folder already exists, af will ask for confirmation that it is ok to overwrite the file with the buffer.

18.7 Reverting a Buffer

If you have made extensive changes to a folder and then change your mind about them, you can get rid of them by reading in the previous version of the folder. To do this, use M-x revert-buffer, which operates on the current buffer. Since reverting a buffer unintentionally could lose a lot of work, you must confirm this command with yes.

Reverting marks the buffer as "not modified" until another change is made.

18.8 Inserting a Folder

Sometimes you may want to merge two folders. An easy way to do this is to use C-x i (insert-file), which inserts the contents of the specified folder into the buffer at point, leaving mark at the start of the inserted contents and point after them. The buffer will not be visiting the folder you inserted, and so will be saved to its original folder.

19. Using Multiple Buffers

The messages you are processing in af reside in an object called a buffer. Each time you visit a folder, a buffer is created to hold the messages in the folder.

At any time, one and only one buffer is selected. It is also called the current buffer. Often we say that a command operates on "the buffer" as if there were only one; but really this means that the command operates on the selected buffer (most commands do).

When af has multiple windows, each window has a chosen buffer which is displayed there, but at any time only one of the windows is selected and its chosen buffer is the selected buffer. Each window's mode line displays the name of the buffer that the window is displaying (see section Multiple Windows).

Each buffer has a name, which can be of any length, and you can select any buffer by giving its name. Most buffers are made by visiting files, and their names are derived from the files' names. But you can also create an empty buffer with any name you want. A newly started af has a buffer named `*scratch*' which can be used as a temporary place to store messages. The distinction between upper and lower case matters in buffer names.

Each buffer records individually what folder it is visiting, whether it is modified, and what major mode and minor modes are in effect in it (see section Major Modes).

19.1 Creating and Selecting Buffers

C-x b buffer RET

Select or create a buffer named buffer (switch-to-buffer).

C-x 4 b buffer RET

Similar, but select buffer in another window (switch-to-buffer-other-window).

To select the buffer named bufname, type C-x b bufname RET. This runs the command switch-to-buffer with argument bufname. You can use completion on an abbreviation for the buffer name you want (see section Completion). An empty argument to C-x b specifies the most recently selected buffer that is not displayed in any window.

Most buffers are created by visiting files, but you can also create a buffer explicitly by typing C-x b bufname RET. This makes a new, empty buffer which is not visiting any file, and selects it. If you try and save a buffer created in this way, you are asked for the file name to use.

Note that C-x C-f, and any other command for visiting a file, can also be used to switch to an existing file-visiting buffer. See section Visiting Files.