Notes on Configuration and Preferences

Pine in Function Key Mode

The standard Pine uses alphabetic keys for most commands, and control keys in the composer. Despite possible appearances, the current bindings are the result of much discussion and thought. All the commands in the composer are single control characters. This keeps things very neat and simple for users. Two character commands in the composer are a possibility, but we're trying to avoid them because of the added complexity for the user.

Pine can also operate in a function-key mode. To go into this mode invoke pine -k or (on some UNIX systems) pinef. On a UNIX system, you can link or copy the pine executable to pinef to install pinef. Alternatively, users and systems administrators can set the use-function-keys feature in the personal or system-wide Pine configuration file. The command menus at the bottom of the screen will show F1-F12 instead of the alphabetic commands. In addition, the help screens will be written in terms of function keys and not alphabetic keys.

One of the results of using Pine in function-key mode is that users can only choose from twelve commands at any given time. In alphabetic-key mode, a user can press a key for a command (say, q to quit) and that command can be fulfilled. In function-key mode, the command must be visible on the bottom key-menu in order to be used. There are some screens where 34 commands are operational; function-key users can get to all of them, just not all at once.

Domain Settings

Pine uses the default domain for a few different tasks. First, it is tacked onto the user-id for outgoing email. Second, it is tacked onto all "local" (unqualified) addresses in the "To:" or "Cc:" fields of messages being composed (unless they are found in the address book). The domain name is also used to generate message-id lines for each outgoing message and to allow Pine to check if an address is that of the current Pine user.

Pine determines the domain name according to whichever of these it finds. The list here is in decreasing order of precedence.

  1. Value of the variable user-domain in the system fixed configuration file
  2. Value of the variable user-domain in the personal configuration file
  3. Value of the variable user-domain is the system-wide configuration file
  4. Value from an external database (DNS, /etc/hosts, NIS) as modified by a system fixed configuration file if use-domain-name-only set to "yes"
  5. Value from an external database (DNS, /etc/hosts, NIS) as modified by a personal configuration file if use-domain-name-only set to "yes"
  6. Value from an external database (DNS, /etc/hosts, NIS) as modified by a system configuration file if use-domain-name-only set to "yes"
  7. Unmodified value (host name) from an external database

The easiest way for this system to work is for PC-Pine users and UNIX Pine system administrators to set the user-domain variable. The variable use-only-domain-name is helpful if your site supports/requires hostless addressing, but for some reason you don't want to use the user-domain variable.

A new feature in 3.90 is called user-lookup-even-if-domain-mismatch. This will cause the personal name field to be looked up from the password file even if the domain of an address isn't a substring of the local host name. See the online help in the Setup/Config screen for full information.

Syntax for Collections

In many environments, it is quite common to have collections of archived mail on various hosts around the network. Using the folder collections facility in Pine, access to these archives is just as simple as access to folders on Pine's local disk.

"Collection" is the word we use in Pine to describe a set of folders. A collection corresponds loosely to a "directory" containing mail folders. Folders within a defined collection can be manipulated (opened, saved-to, etc) using just their simple name. Any number of folder collections can be defined, and pine will adjust its menus and prompts to help navigate them.

The way collections are defined in Pine is with the folder-collections variable in the Pine configuration file. Folder-collections takes a list of one or more collections, each (optionally) preceded by a user-defined logical name (label). Once collections are defined, Pine adjusts its menus and behavior to allow choosing files by their simple name within the collection. Collections are always defined in the configuration file; there is no time that Pine will ever ask a question which requires a user to input a collection specifier. This might change in the future if/when the Goto command is extended to allow jumping to a collection/directory as well as an individual folder.

Consider the following:

   folder-collections=	Local-Mail	C:\MAIL\[],
			Remote-Mail	{}mail/[]

The example shows two collections defined (a comma separated list; newlines in the list are OK if there's one or more spaces before the next entry), one local and one remote. Each collection is a space-delimited pair of elements-first an optional logical-name and second the collection specifier. The logical-name can have spaces if it has quotes around it (but keeping the logical name short and descriptive works best). Pine will use the logical-name (if provided) to reference all folders in the collection, so the user never has to see the ugliness of the collection specifier.

The collection specifier can be thought of as an extended IMAP format (see the "Remote Folders" section for a description of IMAP format names). Basically, a pair of square-brackets are placed in the fully qualified IMAP path where the simple folder name (the part without the host name and path) would appear. Like IMAP, the path can be either fully qualified (i.e., with a leading '/') or relative to your home directory.

An advanced feature of this notation is that a pattern within the square brackets allows the user to define a collection to be a subset of a directory. For example, a collection defined with the specifier:

	M-Mail		C:MAIL/[m*]
will provide a view in the folder lister of all folders in the PC's "C:MAIL" directory that start with the letter 'm' (case insensitive under DOS, of course). Further, the wildcard matching will honor characters trailing the '*' in the pattern.

From within Pine, the FOLDER LIST display will be adjusted to allow browsing of the folders in any defined collection. Even more, you'll notice in the Goto and Save commands a pair of sub-commands to rotate through the list of logical collection names, so only a simple name need be input in order to operate on a folder in any collection.

The first collection specified in the folder-collections has special significance. That folder is the "default collection for saves". In cases where the user does not specify which collection should be used to save a message, the default collection for saves will be used. Also, if the default-fcc is a relative file name, then it is relative to the default collection for saves.

The notion of collections encompasses both email folders and news reading. The variable news-collections uses nearly the same format as folder-collections. Newsgroups can be defined for convenient access via either IMAP or NNTP. There are advantages and disadvantages to both access methods. In the IMAP case, your news environment state is maintained on the server and, thus, will be seen by any client. The downside is that, at the moment, you must have an account on the server. In the NNTP case, server access is mostly anonymous and no state/accounting need be maintained on it. The downside is that each client, for now, must individually maintain news environment state.

An example pinerc entry might be:

     news-collections=	Remote-State	*{}[],
			Local-State	*{}[]
Note that each news collection must be preceded by a '*' to indicate non-mail access. Only newsgroups to which you are subscribed are included in the collection.

The pattern matching facility can be applied so as to define a news collection which is a subset of all the newsgroups you subscribe to. For example, this could be a valid collection:

			Newsfeed-News	*{}[clari.*]

Collection handling is a tough problem to solve in a general way, and the explanation of the syntax is a bit ugly. The upside is, hopefully, that for a little complexity in the Pine configuration file you get simple management of multiple folders in diverse locations.

Syntax for Remote Folders

Remote folders are distinguished from local folders by a leading host name bracketed by '{' and '}'. The path and folder name immediately following the closing bracket, '}', is interpreted by the IMAP server and is in a form compatible with that server (i.e., path delimiters and naming syntax relative to that server).

Typically, a folder name without any path description is understood to reside in the user's "home directory" (i.e., in some way the user's personal, writable file area), as are incomplete path designations. However, the IMAP specification does not require that unqualified folder names live in one's home directory, so some IMAP servers may require fully qualified names. An example of a remote folder specification would be,

This example simply specifies a folder named ``saved-messages'' on the imap server ``'', in the ``mail'' subdirectory of the user's home directory. Easy isn't it?

To confuse things a bit, qualifiers are permitted within the brackets following the host name. These qualifiers consist of a slash, '/' character followed by a keyword or keyword and value equality, and have the effect of modifying how the connection is made to the host specified. An example of such a specification might be,

Another example might be,

Both of these examples illustrate a different qualifier. The first, specifying ``anonymous'' access to the IMAP server on ``''. The second is interesting in that it specifies an altogether different access method: access via the Network News Transport Protocol (NNTP). Both examples bring to light one remaining subtlety. The leading ``*'' tells pine to treat the remote folder as a Bulletin-Board (i.e., typically a shared, read-only resource) and to adjusts its behavior accordingly.

Sorting a Folder

The mail index may be sorted by subject, size, sender, date, or arrival order. Each sort order can also be reversed. The $ command will prompt the user for the sort order. The sort order can also be specified on the command line with the -sort flag or (equivalently) with the sort-key variable in the .pinerc file. When a user changes folders, the sort order will go back to the original sort order. The command line (-sort) or configuration file sort specification (sort-key) changes the original sort order.

When a folder is sorted and new mail arrives in the folder it will be inserted in its properly sorted place. This can be a little odd when the folder is sorted by something like the subject. It can also be a little slow if you are viewing a large, sorted INBOX, since the INBOX will have to be re-sorted whenever new mail arrives.

The sorts are all independent of case and ignore leading or trailing white space. There are actually two forms of subject sort. One called "Subject" and the other called "OrderedSubj". They both ignore "Re:" at the beginning and "(fwd)" at the end of the subjects. Subject sorts all the subjects alphabetically. OrderedSubj sorts by subjects alphabetically, groups messages with the same subject (pseudo-threads), then sorts the groups by the date of the first message of the group. The sort by sender sorts by the userid, not the full name. The arrival sort is basically no sort at all and the date sort depends on the format of the date. Some dates are in strange formats and are unparsable. The time zone is also taken into account.

Sorting large mail folders can be very slow since it requires fetching all the headers of the mail messages. With UNIX Pine, only the first sort is slow since Pine keeps a copy of all the headers. One exception is sorting in reverse arrival order. This is fast because no headers have to be examined. Pine will show progress as it is sorting.

Alternate Editor

In the Pine composer you can use any text editor, such as vi or emacs, for composing the message text. The addresses and subject still must be edited using the standard Pine composer. If you include the feature enable-alternate-editor-cmd in your .pinerc you can type ^_ while in the body of the message in the composer and be prompted for the editor. If you also set the editor variable in your .pinerc then ^_ will invoke the configured editor when you type it.

Turning on the feature enable-alternate-editor-implicitly will automatically invoke the editor you have defined with the editor variable whenever you enter the body of a message you are composing. For example, when you move out of the last header line and into the body of the message, the alternate editor will be automatically invoked.

We know that many people would like to use the alternate editor to edit the mail header as well. We considered several designs for this and didn't come up with one that we liked and that was easy to implement. One of the main problems is that you lose access to the address book.

Signatures and Signature Placement

If the file ~/.signature (UNIX) or <PINERCdirectory>\PINE.SIG (PC) exists, it will be included in all outgoing messages. It is included before composition starts so that the user has a chance to edit it out if he or she likes. The file name for the signature can be changed by setting the signature-file variable in the .pinerc. There is no way to have Pine include different signatures in different outgoing messages automatically. You can do this by hand, however, by having multiple signature files (.sig1, .sig2, .sig3, etc) and choosing to include (^R in the composer) the correct one for the message being sent.

Pine's default behavior encourages a user to put his or her contribution before the inclusion of the original text of the message being forwarded or replied to, This is contrary to some conventions, but makes the conversation more readable when a long original message is included in a reply for context. The reader doesn't have to scroll through the original text that he or she has probably already seen to find the new text. If the reader wishes to see the old message(s), the reader can scroll further into the message. Users who prefer to add their input at the end of a message should set the signature-at-bottom feature in the feature-list. The signature will then be appended to the end of the message after any included text. This feature applies when replying, not when forwarding.

Feature List Variable

Pine used to have feature levels for users with different amounts of experience. We found that this was too restrictive. Pine now has a feature-list instead. Each user may pick and choose which features they would like enabled (simple to do in the Setup/Config screen). There is a short on-line help explaining the effect of each of the features in the Setup/Config screen. When the cursor is highlighting a feature, the "?" command will show the help text for that feature. Features don't have values, they are just turned on or off. They are all off by default.

The feature-list variable is different from all other configuration variables in that its value is additive. That is, the system-wide configuration file can have some features turned on by default. The user can select other features in their personal configuration file and those features will be added to the set of features turned on in the system-wide configuration file. (With all other configuration variables, the user's values replace the system-wide values.) Likewise, additional features may be set on the command-line with the argument "-feature-list=". These will be added to the others.

The treatment of feature-list in the system-wide fixed configuration file is also different from other variables. The system management can fix the value of individual features by placing them in the fixed configuration file. Users will not be able to alter those features, but will still be able to set the other non-restricted features the way they like.

Because feature-list is additive, there is a way to turn features off as well as on. Prepending the prefix "no-" to any feature sets it to off. This is useful for over-riding the system-wide default in the personal configuration file or for over-riding the system-wide default or the personal configuration value on the command line. For example, if the system-wide default configuration has the quit-without-confirm feature set, the user can over-ride that (and turn it off) by including no-quit-without-confirm in the personal configuration file or by giving the command line argument -feature-list=no-quit-without-confirm. More features (options) will no doubt continue to be added.

SMTP Servers

It is sometimes desireable to set smtp-server=localhost instead of setting sendmail-path to overcome the inability to negotiate ESMTP options when sendmail is invoked with the -t option. Sendmail can also be subject to unacceptable delays due to slow DNS lookups and other problems.

It is sometimes desireable to configure an SMTP server on a port other than the default port 25. This may be used to provide an alternate service that is optimized for a particular environment or provides different features from the port 25 server. An example would be a program that negotiates ESMTP options and queues a message, but does not attempt to deliver messages. This would avoid delays frequently encountered when invoking sendmail directly.

A typical configuration would consist of

MIME.Types file

Pine's MIME-TYPE support is based on code contributed by Hans Drexler <>. Pine assigns MIME Content-Types types according to file name extensions found in a system wide files /usr/local/lib/mime.types and /etc/mime.types, and a user specific ~/.mime.types file.

In DOS and OS/2, Pine looks in the same directory as the PINERC file and the same dir as PINE.EXE. This is similar to the UNIX situation with personal config info coming before potentially shared config data. An alternate search path can be specified by setting mimetype-search-path variable in the user or system-wide configuration or by setting the MIMETYPES environment variable.

These files specify file extensions that will be connected to a mime type. Lines beginning with a '#' character are treated as comments and ignored. All other lines are treated as a mime type definition. The first word is a type/subtype specification. All following words are file extensions belonging to that type/subtype. Words are separated by whitespace characters. If a file extension occurs more than once, then the first definition determines the file type and subtype.

Additional Notes on PC-Pine

Below are a few odds and ends worth mentioning about PC-Pine. They have to do with DOS-specific behavior that is either necessary or useful (and sometimes both!).

As PC-Pine runs in an environment with limited access control, accounting or auditing, an additional line is automatically inserted into the header of mail messages generated by PC-Pine:

	X-Sender: <userid>@<>

By popular demand of system administrators, PC-Pine has been modified to prevent sending messages until the user has successfully logged into a remote mail server. Even though PC-Pine cannot prevent users from changing the apparent identity of the sender of a message, the IMAP server login name host name included in the X-Sender line provide some level of traceability by the recipient. However, this should not be considered a rigorous form of authentication. It is extremely lightweight, and is not a replacement for true authentication.

Hand in hand with authentication and accounting is user information. Since PC-Pine has no user database to consult for user-id, personal-name, etc., necessary information must be provided by the user/installer before PC-Pine can properly construct the "From" address required for outbound messages. PC-Pine will, by default, prompt for the requisite pieces as they are needed. This information corresponds to the PINERC variables user-id, personal-name, user-domain, and smtp-server.

The user is then asked whether or not this information should automatically be saved to the PINERC. This is useful behavior in general, but can lead to problems in a lab or other shared environment. Hence, these prompts and automatic saving of configuration can be turned off on an entry by entry basis by setting any of the above values in the PINERC to the null string (i.e., a pair of double quotes). This means that the user will be prompted for the information once during each pine session, and no opportunity to save them in the PINERC will be offered.

Along similar lines, a feature allowing automatic login to the imap-server containing the user's INBOX has also been requested. This feature is not enabled by default, but requires the existence of the file named PINE.PWD in the same directory as the PINERC. Even with the existence of this file, the user must still acknowledge a prompt before the password is saved to the file. If PC-Pine is configured to access several different IMAP servers, each password entered will be kept (associated with the corresponding host name) in memory during the current session, and optionally, in the PINE.PWD file for use in subsequent sessions.

WARNING! Use this feature with caution! It effectively makes the user's mail no more secure than the physical security of the machine running PC-Pine. What's more, while the password is cloaked by a mild (some might say, feeble) encryption scheme, it is nonetheless sitting in a file on the PC's disk and subject to cracking by anyone with access to it. BEWARE!

Another feature of DOS is the lack of standard scratch area for temporary files. During the course of a session, PC-Pine may require numerous temporary files (large message texts, various caches, etc.). Where to create them can be a problem, particularly when running under certain network operating systems. PC-Pine observes the TMP and TEMP environment variables, and creates temporary files in the directory specified by either. In their absence, PC-Pine creates these files in the root of the current working drive.