The Gnus Newsreader

Table of Contents

Short Table of Contents

1 Don’t Panic

Welcome, gentle user, to the Gnus newsreader and email client! Gnus is unlike most clients, in part because of its endless configurability, in part because of its historical origins. Gnus is now a fully-featured email client, but it began life as a Usenet-style newsreader, and its genes are still newsreader genes. Thus it behaves a little differently than most mail clients.

The typical assumptions of a newsreader are:

  1. The server offers a potentially enormous number of newsgroups on a variety of subjects. The user may only be interested in some of those groups, and more interested in some than others.
  2. Many groups see a high volume of articles, and the user won’t want to read all of them. Mechanisms are needed for foregrounding interesting articles, and backgrounding uninteresting articles.
  3. Once a group has been scanned and dealt with by the user, it’s unlikely to be of further interest until new articles come in.

These assumptions lead to certain default Gnus behaviors:

  1. Not all interesting groups are equally interesting, thus groups have varying degrees of “subscribedness”, with different behavior depending on “how subscribed” a group is.
  2. There are many commands and tools for scoring and sorting articles, or otherwise sweeping them under the rug.
  3. Gnus will only show you groups with unread or ticked articles; groups with no new articles are hidden.
  4. When entering a group, only unread or ticked articles are shown, all other articles are hidden.

If this seems draconian, think of it as Automatic Inbox Zero. This is the way Gnus works by default. It is possible to make it work more like an email client (always showing read groups and read articles), but that takes some effort on the part of the user.

The brief introduction below should be enough to get you off the ground.

The Basics of Servers, Groups, and Articles

The fundamental building blocks of Gnus are servers, groups, and articles. Servers can be local or remote. Each server maintains a list of groups, and those groups contain articles. Because Gnus presents a unified interface to a wide variety of servers, the vocabulary doesn’t always quite line up (see Glossary, for a more complete glossary). Thus a local Maildir is referred to as a “server” (see Finding the News) the same as a Usenet or IMAP server is; “groups” (see Group Buffer) might mean an NNTP group, IMAP folder, or local mail directory; and an “article” (see Summary Buffer) might elsewhere be known as a message or an email. Gnus employs unified terms for all these things.

Servers fall into two general categories: “news-like”, meaning that the articles are part of a public archive and can’t be manipulated by the user; and “mail-like”, meaning that the articles are owned by the user, who can freely edit them, move them around, and delete them.

For news-like servers, which typically offer hundreds or thousands of groups, it’s important to be able to subscribe to a subset of those groups. For mail-like servers, the user is generally automatically subscribed to all groups (though IMAP, for example, also allows selective subscription). To change group subscription, enter the Server buffer (with ^) and press RET on the server in question. From here, Gnus provides commands to change or toggle your group subscriptions (see Browse Foreign Server).

A Gnus installation is basically just a list of one or more servers, plus the user’s subscribed groups from those servers, plus articles in those groups.

Servers can be added and configured in two places: in the user’s gnus.el startup file, using the gnus-select-method and gnus-secondary-select-methods options, or within Gnus itself using interactive commands in the Server buffer. See Finding the News, for details.

Fetching Mail

New mail has to come from somewhere. Some servers, such as NNTP or IMAP, are themselves responsible for fetching newly-arrived articles. Others, such as Maildir or mbox servers, only store articles and don’t fetch them from anywhere.

In the latter case, Gnus provides for mail sources: places where new mail is fetched from. A mail source might be a local spool, or a remote POP server, or some other source of incoming articles. Mail sources are usually configured globally, but can be specified per-group (see Mail Sources for more information).

See Scanning New Messages, for details on fetching new mail.

Viewing Mail

By default, Gnus’s Group buffer only displays groups with unread articles. It is always possible to display all the groups temporarily with L, and to configure Gnus to always display some groups (see Listing Groups).

See Selecting a Group, for how to enter a group, and see Summary Buffer for what to do once you’re there.

Sending Mail

New message composition can be initiated from the Group buffer (see Misc Group Stuff). If you’re in a Summary buffer, you can compose replies and forward emails in addition to starting new messages, see Summary Mail Commands, for details.

For information about what happens once you’ve started composing a message, see Composing Messages. For information on setting up SMTP servers in particular, see Mail Variables in Message manual.

2 Starting Gnus

If you haven’t used Emacs much before using Gnus, read Emacs for Heathens first.

If your system administrator has set things up properly, starting Gnus and reading news is extremely easy—you just type M-x gnus in your Emacs. If not, you should customize the variable gnus-select-method as described in Finding the News. For a minimal setup for posting should also customize the variables user-full-name and user-mail-address.

If you want to start Gnus in a different frame, you can use the command M-x gnus-other-frame instead.

If things do not go smoothly at startup, you have to twiddle some variables in your ~/.gnus.el file. This file is similar to ~/.emacs, but is read when Gnus starts.

If you puzzle at any terms used in this manual, please refer to the terminology section (see Terminology).

2.1 Finding the News

First of all, you should know that there is a special buffer called *Server* that lists all the servers Gnus knows about. You can press ^ from the Group buffer to see it. In the Server buffer, you can press RET on a defined server to see all the groups it serves (subscribed or not!). You can also add or delete servers, edit a foreign server’s definition, agentize or de-agentize a server, and do many other neat things. See Server Buffer. See Foreign Groups. See Agent Basics.

The gnus-select-method variable says where Gnus should look for news. This variable should be a list where the first element says how and the second element says where. This method is your native method. All groups not fetched with this method are secondary or foreign groups.

For instance, if the ‘news.somewhere.eduNNTP server is where you want to get your daily dosage of news from, you’d say:

(setq gnus-select-method '(nntp ""))

If you want to read directly from the local spool, say:

(setq gnus-select-method '(nnspool ""))

If you can use a local spool, you probably should, as it will almost certainly be much faster. But do not use the local spool if your server is running Leafnode (which is a simple, standalone private news server); in this case, use (nntp "localhost").

If this variable is not set, Gnus will take a look at the NNTPSERVER environment variable. If that variable isn’t set, Gnus will see whether gnus-nntpserver-file (/etc/nntpserver by default) has any opinions on the matter. If that fails as well, Gnus will try to use the machine running Emacs as an NNTP server. That’s a long shot, though.

However, if you use one NNTP server regularly and are just interested in a couple of groups from a different server, you would be better served by using the B command in the group buffer. It will let you have a look at what groups are available, and you can subscribe to any of the groups you want to. This also makes .newsrc maintenance much tidier. See Foreign Groups.

A slightly different approach to foreign groups is to set the gnus-secondary-select-methods variable. The select methods listed in this variable are in many ways just as native as the gnus-select-method server. They will also be queried for active files during startup (if that’s required), and new newsgroups that appear on these servers will be subscribed (or not) just as native groups are.

For instance, if you use the nnmbox back end to read your mail, you would typically set this variable to

(setq gnus-secondary-select-methods '((nnmbox "")))

2.2 The Server is Down

If the default server is down, Gnus will understandably have some problems starting. However, if you have some mail groups in addition to the news groups, you may want to start Gnus anyway.

Gnus, being the trusting sort of program, will ask whether to proceed without a native select method if that server can’t be contacted. This will happen whether the server doesn’t actually exist (i.e., you have given the wrong address) or the server has just momentarily taken ill for some reason or other. If you decide to continue and have no foreign groups, you’ll find it difficult to actually do anything in the group buffer. But, hey, that’s your problem. Blllrph!

If you know that the server is definitely down, or you just want to read your mail without bothering with the server at all, you can use the gnus-no-server command to start Gnus. That might come in handy if you’re in a hurry as well. This command will not attempt to contact your primary server—instead, it will just activate all groups on level 1 and 2. (You should preferably keep no native groups on those two levels.) Also see Group Levels.

2.3 Child Gnusae

You might want to run more than one Emacs with more than one Gnus at the same time. If you are using different .newsrc files (e.g., if you are using the two different Gnusae to read from two different servers), that is no problem whatsoever. You just do it.

The problem appears when you want to run two Gnusae that use the same .newsrc file.

To work around that problem some, we here at the Think-Tank at the Gnus Towers have come up with a new concept: Parents and children.

Anyway, you start one Gnus up the normal way with M-x gnus (or however you do it). Each subsequent child Gnusae should be started with M-x gnus-child. These children won’t save normal .newsrc files, but instead save child files that contain information only on what groups have been read in the child session. When a parent Gnus starts, it will read (and delete) these child files, incorporating all information from them. (The child files will be read in the sequence they were created, so the latest changes will have precedence.)

Information from the child files has, of course, precedence over the information in the normal (i.e., parent) .newsrc file.

If the .newsrc* files have not been saved in the parent when the child starts, you may be prompted as to whether to read an auto-save file. If you answer “yes”, the unsaved changes to the parent will be incorporated into the child. If you answer “no”, the child may see some messages as unread that have been read in the parent.

2.4 New Groups

If you are satisfied that you really never want to see any new groups, you can set gnus-check-new-newsgroups to nil. This will also save you some time at startup. Even if this variable is nil, you can always subscribe to the new groups just by pressing U in the group buffer (see Group Maintenance). This variable is ask-server by default. If you set this variable to always, then Gnus will query the back ends for new groups even when you do the g command (see Scanning New Messages).

2.4.1 Checking New Groups

Gnus normally determines whether a group is new or not by comparing the list of groups from the active file(s) with the lists of subscribed and dead groups. This isn’t a particularly fast method. If gnus-check-new-newsgroups is ask-server, Gnus will ask the server for new groups since the last time. This is both faster and cheaper. This also means that you can get rid of the list of killed groups (see Group Levels) altogether, so you may set gnus-save-killed-list to nil, which will save time both at startup, at exit, and all over. Saves disk space, too. Why isn’t this the default, then? Unfortunately, not all servers support this command.

I bet I know what you’re thinking now: How do I find out whether my server supports ask-server? No? Good, because I don’t have a fail-safe answer. I would suggest just setting this variable to ask-server and see whether any new groups appear within the next few days. If any do, then it works. If none do, then it doesn’t work. I could write a function to make Gnus guess whether the server supports ask-server, but it would just be a guess. So I won’t. You could telnet to the server and say HELP and see whether it lists ‘NEWGROUPS’ among the commands it understands. If it does, then it might work. (But there are servers that lists ‘NEWGROUPS’ without supporting the function properly.)

This variable can also be a list of select methods. If so, Gnus will issue an ask-server command to each of the select methods, and subscribe them (or not) using the normal methods. This might be handy if you are monitoring a few servers for new groups. A side effect is that startup will take much longer, so you can meditate while waiting. Use the mantra “dingnusdingnusdingnus” to achieve permanent bliss.

2.4.2 Subscription Methods

What Gnus does when it encounters a new group is determined by the gnus-subscribe-newsgroup-method variable.

This variable should contain a function. This function will be called with the name of the new group as the only parameter.

Some handy pre-fab functions are:


Make all new groups zombies (see Group Levels). This is the default. You can browse the zombies later (with A z) and either kill them all off properly (with S z), or subscribe to them (with u).


Subscribe all new groups in arbitrary order. This really means that all new groups will be added at “the top” of the group buffer.


Subscribe all new groups in alphabetical order.


Subscribe all new groups hierarchically. The difference between this function and gnus-subscribe-alphabetically is slight. gnus-subscribe-alphabetically will subscribe new groups in a strictly alphabetical fashion, while this function will enter groups into its hierarchy. So if you want to have the ‘rec’ hierarchy before the ‘comp’ hierarchy, this function will not mess that configuration up. Or something like that.


Subscribe new groups interactively. This means that Gnus will ask you about all new groups. The groups you choose to subscribe to will be subscribed hierarchically.


Kill all new groups.


Put the groups into the topic that has a matching subscribe topic parameter (see Topic Parameters). For instance, a subscribe topic parameter that looks like


will mean that all groups that match that regex will be subscribed under that topic.

If no topics match the groups, the groups will be subscribed in the top-level topic.

A closely related variable is gnus-subscribe-hierarchical-interactive. (That’s quite a mouthful.) If this variable is non-nil, Gnus will ask you in a hierarchical fashion whether to subscribe to new groups or not. Gnus will ask you for each sub-hierarchy whether you want to descend the hierarchy or not.

One common mistake is to set the variable a few paragraphs above (gnus-subscribe-newsgroup-method) to gnus-subscribe-hierarchical-interactive. This is an error. This will not work. This is ga-ga. So don’t do it.

2.4.3 Filtering New Groups

A nice and portable way to control which new newsgroups should be subscribed (or ignored) is to put an options line at the start of the .newsrc file. Here’s an example:

options -n !alt.all !rec.all sci.all

This line obviously belongs to a serious-minded intellectual scientific person (or she may just be plain old boring), because it says that all groups that have names beginning with ‘alt’ and ‘rec’ should be ignored, and all groups with names beginning with ‘sci’ should be subscribed. Gnus will not use the normal subscription method for subscribing these groups. gnus-subscribe-options-newsgroup-method is used instead. This variable defaults to gnus-subscribe-alphabetically.

The “options -n” format is very simplistic. The syntax above is all that is supports: you can force-subscribe hierarchies, or you can deny hierarchies, and that’s it.

If you don’t want to mess with your .newsrc file, you can just set the two variables gnus-options-subscribe and gnus-options-not-subscribe. These two variables do exactly the same as the .newsrcoptions -n’ trick. Both are regexps, and if the new group matches the former, it will be unconditionally subscribed, and if it matches the latter, it will be ignored.

Yet another variable that meddles here is gnus-auto-subscribed-groups. It works exactly like gnus-options-subscribe, and is therefore really superfluous, but I thought it would be nice to have two of these. This variable is more meant for setting some ground rules, while the other variable is used more for user fiddling. By default this variable makes all new groups that come from mail back ends (nnml, nnbabyl, nnfolder, nnmbox, nnmh, nnimap, and nnmaildir) subscribed. If you don’t like that, just set this variable to nil.

As if that wasn’t enough, gnus-auto-subscribed-categories also allows you to specify that new groups should be subscribed based on the category their select methods belong to. The default is ‘(mail post-mail)’, meaning that all new groups from mail-like backends should be subscribed automatically.

New groups that match these variables are subscribed using gnus-subscribe-options-newsgroup-method.

2.5 Changing Servers

Sometimes it is necessary to move from one NNTP server to another. This happens very rarely, but perhaps you change jobs, or one server is very flaky and you want to use another.

Changing the server is pretty easy, right? You just change gnus-select-method to point to the new server?


Article numbers are not (in any way) kept synchronized between different NNTP servers, and the only way Gnus keeps track of what articles you have read is by keeping track of article numbers. So when you change gnus-select-method, your .newsrc file becomes worthless.

You can use the M-x gnus-group-clear-data-on-native-groups command to clear out all data that you have on your native groups. Use with caution.

Clear the data from the current group only—nix out marks and the list of read articles (gnus-group-clear-data).

After changing servers, you must move the cache hierarchy away, since the cached articles will have wrong article numbers, which will affect which articles Gnus thinks are read. gnus-group-clear-data-on-native-groups will ask you if you want to have it done automatically; for gnus-group-clear-data, you can use M-x gnus-cache-move-cache (but beware, it will move the cache for all groups).

2.6 Startup Files

Most common Unix news readers use a shared startup file called .newsrc. This file contains all the information about what groups are subscribed, and which articles in these groups have been read.

Things got a bit more complicated with GNUS. In addition to keeping the .newsrc file updated, it also used a file called .newsrc.el for storing all the information that didn’t fit into the .newsrc file. (Actually, it also duplicated everything in the .newsrc file.) GNUS would read whichever one of these files was the most recently saved, which enabled people to swap between GNUS and other newsreaders.

That was kinda silly, so Gnus went one better: In addition to the .newsrc and .newsrc.el files, Gnus also has a file called .newsrc.eld. It will read whichever of these files that are most recent, but it will never write a .newsrc.el file. You should never delete the .newsrc.eld file—it contains much information not stored in the .newsrc file.

You can turn off writing the .newsrc file by setting gnus-save-newsrc-file to nil, which means you can delete the file and save some space, as well as exiting from Gnus faster. However, this will make it impossible to use other newsreaders than Gnus. But hey, who would want to, right? Similarly, setting gnus-read-newsrc-file to nil makes Gnus ignore the .newsrc file and any .newsrc-SERVER files, which can be convenient if you use a different news reader occasionally, and you want to read a different subset of the available groups with that news reader.

If gnus-save-killed-list (default t) is nil, Gnus will not save the list of killed groups to the startup file. This will save both time (when starting and quitting) and space (on disk). It will also mean that Gnus has no record of what groups are new or old, so the automatic new groups subscription methods become meaningless. You should always set gnus-check-new-newsgroups to nil or ask-server if you set this variable to nil (see New Groups). This variable can also be a regular expression. If that’s the case, remove all groups that do not match this regexp before saving. This can be useful in certain obscure situations that involve several servers where not all servers support ask-server.

The gnus-startup-file variable says where the startup files are. The default value is ~/.newsrc, with the Gnus (El Dingo) startup file being whatever that one is, with a ‘.eld’ appended. If you want to keep multiple numbered backups of this file, set gnus-backup-startup-file. It respects the same values as the version-control variable.

gnus-save-newsrc-hook is called before saving any of the newsrc files, while gnus-save-quick-newsrc-hook is called just before saving the .newsrc.eld file, and gnus-save-standard-newsrc-hook is called just before saving the .newsrc file. The latter two are commonly used to turn version control on or off. Version control is on by default when saving the startup files. If you want to turn backup creation off, say something like:

(defun turn-off-backup ()
  (set (make-local-variable 'backup-inhibited) t))

(add-hook 'gnus-save-quick-newsrc-hook 'turn-off-backup)
(add-hook 'gnus-save-standard-newsrc-hook 'turn-off-backup)

When Gnus starts, it will read the gnus-site-init-file (.../site-lisp/gnus-init by default) and gnus-init-file (~/.gnus by default) files. These are normal Emacs Lisp files and can be used to avoid cluttering your ~/.emacs and site-init files with Gnus stuff. Gnus will also check for files with the same names as these, but with .elc and .el suffixes. In other words, if you have set gnus-init-file to ~/.gnus, it will look for ~/.gnus.elc, ~/.gnus.el, and finally ~/.gnus (in this order). If Emacs was invoked with the -q or --no-init-file options (see Initial Options in The Emacs Manual), Gnus doesn’t read gnus-init-file.

2.7 Auto Save

Whenever you do something that changes the Gnus data (reading articles, catching up, killing/subscribing groups), the change is added to a special dribble buffer. This buffer is auto-saved the normal Emacs way. If your Emacs should crash before you have saved the .newsrc files, all changes you have made can be recovered from this file.

If Gnus detects this file at startup, it will ask the user whether to read it. The auto save file is deleted whenever the real startup file is saved.

If gnus-use-dribble-file is nil, Gnus won’t create and maintain a dribble buffer. The default is t.

Gnus will put the dribble file(s) in gnus-dribble-directory. If this variable is nil, which it is by default, Gnus will dribble into the directory where the .newsrc file is located. (This is normally the user’s home directory.) The dribble file will get the same file permissions as the .newsrc file.

If gnus-always-read-dribble-file is non-nil, Gnus will read the dribble file on startup without querying the user.

2.8 The Active File

When Gnus starts, or indeed whenever it tries to determine whether new articles have arrived, it reads the active file. This is a very large file that lists all the active groups and articles on the server.

Before examining the active file, Gnus deletes all lines that match the regexp gnus-ignored-newsgroups. This is done primarily to reject any groups with bogus names, but you can use this variable to make Gnus ignore hierarchies you aren’t ever interested in. However, this is not recommended. In fact, it’s highly discouraged. Instead, see New Groups for an overview of other variables that can be used instead.

The active file can be rather Huge, so if you have a slow network, you can set gnus-read-active-file to nil to prevent Gnus from reading the active file. This variable is some by default.

Gnus will try to make do by getting information just on the groups that you actually subscribe to.

Note that if you subscribe to lots and lots of groups, setting this variable to nil will probably make Gnus slower, not faster. At present, having this variable nil will slow Gnus down considerably, unless you read news over a 2400 baud modem.

This variable can also have the value some. Gnus will then attempt to read active info only on the subscribed groups. On some servers this is quite fast (on sparkling, brand new INN servers that support the LIST ACTIVE group command), on others this isn’t fast at all. In any case, some should be faster than nil, and is certainly faster than t over slow lines.

Some news servers (old versions of Leafnode and old versions of INN, for instance) do not support the LIST ACTIVE group. For these servers, nil is probably the most efficient value for this variable.

If this variable is nil, Gnus will ask for group info in total lock-step, which isn’t very fast. If it is some and you use an NNTP server, Gnus will pump out commands as fast as it can, and read all the replies in one swoop. This will normally result in better performance, but if the server does not support the aforementioned LIST ACTIVE group command, this isn’t very nice to the server.

If you think that starting up Gnus takes too long, try all the three different values for this variable and see what works best for you.

In any case, if you use some or nil, you should definitely kill all groups that you aren’t interested in to speed things up.

Note that this variable also affects active file retrieval from secondary select methods.

2.9 Startup Variables


A hook called as the first thing when Gnus is started.


A hook called as the first thing when Gnus is resumed after a suspend.


A hook run as the very last thing after starting up Gnus


A hook that is run as the very last thing after starting up Gnus successfully.


A hook that is run after reading the .newsrc file(s), but before generating the group buffer.


If non-nil, Gnus will check for and delete all bogus groups at startup. A bogus group is a group that you have in your .newsrc file, but doesn’t exist on the news server. Checking for bogus groups can take quite a while, so to save time and resources it’s best to leave this option off, and do the checking for bogus groups once in a while from the group buffer instead (see Group Maintenance).


If non-nil, the startup message won’t be displayed. That way, your boss might not notice as easily that you are reading news instead of doing your job. Note that this variable is used before ~/.gnus.el is loaded, so it should be set in .emacs instead.


Message displayed by Gnus when no groups are available.

3 Group Buffer

The group buffer lists all (or parts) of the available groups. It is the first buffer shown when Gnus starts, and will never be killed as long as Gnus is active.

3.1 Group Buffer Format

You can customize the Group Mode tool bar, see M-x customize-apropos RET gnus-group-tool-bar.

The tool bar icons are now (de)activated correctly depending on the cursor position. Therefore, moving around in the Group Buffer is slower. You can disable this via the variable gnus-group-update-tool-bar. Its default value depends on your Emacs version.

3.1.1 Group Line Specification

The default format of the group buffer is nice and dull, but you can make it as exciting and ugly as you feel like.

Here’s a couple of example group lines:

     25: news.announce.newusers
 *    0:

Quite simple, huh?

You can see that there are 25 unread articles in ‘news.announce.newusers’. There are no unread articles, but some ticked articles, in ‘’ (see that little asterisk at the beginning of the line?).

You can change that format to whatever you want by fiddling with the gnus-group-line-format variable. This variable works along the lines of a format specification, which is pretty much the same as a printf specifications, for those of you who use (feh!) C. See Formatting Variables.

%M%S%5y:%B%(%g%)\n’ is the value that produced those lines above.

There should always be a colon on the line; the cursor always moves to the colon after performing an operation. See Positioning Point. Nothing else is required—not even the group name. All displayed text is just window dressing, and is never examined by Gnus. Gnus stores all real information it needs using text properties.

(Note that if you make a really strange, wonderful, spreadsheet-like layout, everybody will believe you are hard at work with the accounting instead of wasting time reading news.)

Here’s a list of all available format characters:


An asterisk if the group only has marked articles.


Whether the group is subscribed.


Level of subscribedness.


Number of unread articles.


Number of dormant articles.


Number of ticked articles.


Number of read articles.


Number of unseen articles.


Estimated total number of articles. (This is really max-number minus min-number plus 1.)

Gnus uses this estimation because the NNTP protocol provides efficient access to max-number and min-number but getting the true unread message count is not possible efficiently. For hysterical raisins, even the mail back ends, where the true number of unread messages might be available efficiently, use the same limited interface. To remove this restriction from Gnus means that the back end interface has to be changed, which is not an easy job.

The nnml backend (see Mail Spool) has a feature called “group compaction” which circumvents this deficiency: the idea is to renumber all articles from 1, removing all gaps between numbers, hence getting a correct total count. Other backends may support this in the future. In order to keep your total article count relatively up to date, you might want to compact your groups (or even directly your server) from time to time. See Misc Group Stuff, See Server Commands.


Number of unread, unticked, non-dormant articles.


Number of ticked and dormant articles.


Full group name.


Group name.


Group comment (see Group Parameters) or group name if there is no comment element in the group parameters.


Newsgroup description. You need to read the group descriptions before these will appear, and to do that, you either have to set gnus-read-active-file or use the group buffer M-d command.


m’ if moderated.


(m)’ if moderated.


Select method.


If the summary buffer for the group is open or not.


Select from where.


A string that looks like ‘<%s:%n>’ if a foreign select method is used.


Indentation based on the level of the topic (see Group Topics).


Short (collapsed) group name. The gnus-group-uncollapsed-levels variable says how many levels to leave at the end of the group name. The default is 1—this will mean that group names like ‘gnu.emacs.gnus’ will be shortened to ‘g.e.gnus’.


%’ (gnus-new-mail-mark) if there has arrived new mail to the group lately.


#’ (gnus-process-mark) if the group is process marked.


A string that says when you last read the group (see Group Timestamp).


The disk space used by the articles fetched by both the cache and agent. The value is automatically scaled to bytes(B), kilobytes(K), megabytes(M), or gigabytes(G) to minimize the column width. A format of %7F is sufficient for a fixed-width column.


User defined specifier. The next character in the format string should be a letter. Gnus will call the function gnus-user-format-function-X’, where ‘X’ is the letter following ‘%u’. The function will be passed a single dummy parameter as argument. The function should return a string, which will be inserted into the buffer just like information from any other specifier.

All the “number-of” specs will be filled with an asterisk (‘*’) if no info is available—for instance, if it is a non-activated foreign group, or a bogus native group.

3.1.2 Group Mode Line Specification

The mode line can be changed by setting gnus-group-mode-line-format (see Mode Line Formatting). It doesn’t understand that many format specifiers:


The native news server.


The native select method.

3.1.3 Group Highlighting

Highlighting in the group buffer is controlled by the gnus-group-highlight variable. This is an alist with elements that look like (form . face). If form evaluates to something non-nil, the face will be used on the line.

Here’s an example value for this variable that might look nice if the background is dark:

(cond (window-system
       (setq custom-background-mode 'light)
       (defface my-group-face-1
         '((t (:foreground "Red" :bold t))) "First group face")
       (defface my-group-face-2
         '((t (:foreground "DarkSeaGreen4" :bold t)))
         "Second group face")
       (defface my-group-face-3
         '((t (:foreground "Green4" :bold t))) "Third group face")
       (defface my-group-face-4
         '((t (:foreground "SteelBlue" :bold t))) "Fourth group face")
       (defface my-group-face-5
         '((t (:foreground "Blue" :bold t))) "Fifth group face")))

(setq gnus-group-highlight
      '(((> unread 200) . my-group-face-1)
        ((and (< level 3) (zerop unread)) . my-group-face-2)
        ((< level 3) . my-group-face-3)
        ((zerop unread) . my-group-face-4)
        (t . my-group-face-5)))

Also see Faces and Fonts.

Variables that are dynamically bound when the forms are evaluated include:


The group name.


The number of unread articles in the group.


The select method.


Whether the group is a mail group.


The level of the group.


The score of the group.


The number of ticked articles in the group.


The total number of articles in the group. Or rather, max-number minus min-number plus one.


When using the topic minor mode, this variable is bound to the current topic being inserted.

When the forms are evaled, point is at the beginning of the line of the group in question, so you can use many of the normal Gnus functions for snarfing info on the group.

gnus-group-update-hook is called when a group line is changed. It will not be called when gnus-visual is nil.

3.2 Group Maneuvering

All movement commands understand the numeric prefix and will behave as expected, hopefully.


Go to the next group that has unread articles (gnus-group-next-unread-group).


Go to the previous group that has unread articles (gnus-group-prev-unread-group).


Go to the next group (gnus-group-next-group).


Go to the previous group (gnus-group-prev-group).


Go to the next unread group on the same (or lower) level (gnus-group-next-unread-group-same-level).


Go to the previous unread group on the same (or lower) level (gnus-group-prev-unread-group-same-level).

Three commands for jumping to groups:


Jump to a group (and make it visible if it isn’t already) (gnus-group-jump-to-group). Killed groups can be jumped to, just like living groups.


Jump to the unread group with the lowest level (gnus-group-best-unread-group).


Jump to the first group with unread articles (gnus-group-first-unread-group).

If gnus-group-goto-unread is nil, all the movement commands will move to the next group, not the next unread group. Even the commands that say they move to the next unread group. The default is t.

If gnus-summary-next-group-on-exit is t, when a summary is exited, the point in the group buffer is moved to the next unread group. Otherwise, the point is set to the group just exited. The default is t.

3.3 Selecting a Group


Select the current group, switch to the summary buffer and display the first unread article (gnus-group-read-group). If there are no unread articles in the group, or if you give a non-numerical prefix to this command, Gnus will offer to fetch all the old articles in this group from the server. If you give a numerical prefix n, n determines the number of articles Gnus will fetch. If n is positive, Gnus fetches the n newest articles, if n is negative, Gnus fetches the abs(n) oldest articles.

Thus, SPC enters the group normally, C-u SPC offers old articles, C-u 4 2 SPC fetches the 42 newest articles, and C-u - 4 2 SPC fetches the 42 oldest ones.

When you are in the group (in the Summary buffer), you can type M-g to fetch new articles, or C-u M-g to also show the old ones.


Select the current group and switch to the summary buffer (gnus-group-select-group). Takes the same arguments as gnus-group-read-group—the only difference is that this command does not display the first unread article automatically upon group entry.


This does the same as the command above, but tries to do it with the minimum amount of fuzz (gnus-group-quick-select-group). No scoring/killing will be performed, there will be no highlights and no expunging. This might be useful if you’re in a real hurry and have to enter some humongous group. If you give a 0 prefix to this command (i.e., 0 M-RET), Gnus won’t even generate the summary buffer, which is useful if you want to toggle threading before generating the summary buffer (see Summary Generation Commands).


This is yet one more command that does the same as the RET command, but this one does it without expunging and hiding dormants (gnus-group-visible-select-group).


Finally, this command selects the current group ephemerally without doing any processing of its contents (gnus-group-select-group-ephemerally). Even threading has been turned off. Everything you do in the group after selecting it in this manner will have no permanent effects.

The gnus-large-newsgroup variable says what Gnus should consider to be a big group. If it is nil, no groups are considered big. The default value is 200. If the group has more (unread and/or ticked) articles than this, Gnus will query the user before entering the group. The user can then specify how many articles should be fetched from the server. If the user specifies a negative number (−n), the n oldest articles will be fetched. If it is positive, the n articles that have arrived most recently will be fetched.

gnus-large-ephemeral-newsgroup is the same as gnus-large-newsgroup, but is only used for ephemeral newsgroups.

In groups in some news servers, there might be a big gap between a few very old articles that will never be expired and the recent ones. In such a case, the server will return the data like (1 . 30000000) for the LIST ACTIVE group command, for example. Even if there are actually only the articles 1–10 and 29999900–30000000, Gnus doesn’t know it at first and prepares for getting 30000000 articles. However, it will consume hundreds megabytes of memories and might make Emacs get stuck as the case may be. If you use such news servers, set the variable gnus-newsgroup-maximum-articles to a positive number. The value means that Gnus ignores articles other than this number of the latest ones in every group. For instance, the value 10000 makes Gnus get only the articles 29990001–30000000 (if the latest article number is 30000000 in a group). Note that setting this variable to a number might prevent you from reading very old articles. The default value of the variable gnus-newsgroup-maximum-articles is nil, which means Gnus never ignores old articles.

If gnus-auto-select-first is non-nil, select an article automatically when entering a group with the SPC command. Which article this is controlled by the gnus-auto-select-subject variable. Valid values for this variable are:


Place point on the subject line of the first unread article.


Place point on the subject line of the first article.


Place point on the subject line of the first unseen article.


Place point on the subject line of the first unseen article, and if there is no such article, place point on the subject line of the first unread article.


Place point on the subject line of the highest-scored unread article.

This variable can also be a function. In that case, that function will be called to place point on a subject line.

If you want to prevent automatic selection in some group (say, in a binary group with Huge articles) you can set the gnus-auto-select-first variable to nil in gnus-select-group-hook, which is called when a group is selected.

3.4 Subscription Commands

The following commands allow for managing your subscriptions in the Group buffer. If you want to subscribe to many groups, it’s probably more convenient to go to the Server Buffer, and choose the server there using RET or SPC. Then you’ll have the commands listed in Browse Foreign Server at hand.

S t

Toggle subscription to group under point (gnus-group-toggle-subscription-at-point).

S s

Prompt for group, and toggle its subscription. (gnus-group-toggle-subscription).

S k

Kill the current group (gnus-group-kill-group).

S y

Yank the last killed group (gnus-group-yank-group).

C-x C-t

Transpose two groups (gnus-group-transpose-groups). This isn’t really a subscription command, but you can use it instead of a kill-and-yank sequence sometimes.

S w

Kill all groups in the region (gnus-group-kill-region).

S z

Kill all zombie groups (gnus-group-kill-all-zombies).

S C-k

Kill all groups on a certain level (gnus-group-kill-level). These groups can’t be yanked back after killing, so this command should be used with some caution. The only time where this command comes in really handy is when you have a .newsrc with lots of unsubscribed groups that you want to get rid off. S C-k on level 7 will kill off all unsubscribed groups that do not have message numbers in the .newsrc file.

Also see Group Levels.

3.5 Group Data


Mark all unticked articles in this group as read (gnus-group-catchup-current). gnus-group-catchup-group-hook is called when catching up a group from the group buffer.


Mark all articles in this group, even the ticked ones, as read (gnus-group-catchup-current-all).


Clear the data from the current group—nix out marks and the list of read articles (gnus-group-clear-data).

M-x gnus-group-clear-data-on-native-groups

If you have switched from one NNTP server to another, all your marks and read ranges have become worthless. You can use this command to clear out all data that you have on your native groups. Use with caution.

3.6 Group Levels

All groups have a level of subscribedness. For instance, if a group is on level 2, it is more subscribed than a group on level 5. You can ask Gnus to just list groups on a given level or lower (see Listing Groups), or to just check for new articles in groups on a given level or lower (see Scanning New Messages).

Remember: The higher the level of the group, the less important it is.

S l

Set the level of the current group. If a numeric prefix is given, the next n groups will have their levels set. The user will be prompted for a level.

Gnus considers groups from levels 1 to gnus-level-subscribed (inclusive) (default 5) to be subscribed, gnus-level-subscribed (exclusive) and gnus-level-unsubscribed (inclusive) (default 7) to be unsubscribed, gnus-level-zombie to be zombies (walking dead) (default 8) and gnus-level-killed to be killed (completely dead) (default 9). Gnus treats subscribed and unsubscribed groups exactly the same, but zombie and killed groups store no information on what articles you have read, etc. This distinction between dead and living groups isn’t done because it is nice or clever, it is done purely for reasons of efficiency.

It is recommended that you keep all your mail groups (if any) on quite low levels (e.g., 1 or 2).

Maybe the following description of the default behavior of Gnus helps to understand what these levels are all about. By default, Gnus shows you subscribed nonempty groups, but by hitting L you can have it show empty subscribed groups and unsubscribed groups, too. Type l to go back to showing nonempty subscribed groups again. Thus, unsubscribed groups are hidden, in a way.

Zombie and killed groups are similar to unsubscribed groups in that they are hidden by default. But they are different from subscribed and unsubscribed groups in that Gnus doesn’t ask the news server for information (number of messages, number of unread messages) on zombie and killed groups. Normally, you use C-k to kill the groups you aren’t interested in. If most groups are killed, Gnus is faster.

Why does Gnus distinguish between zombie and killed groups? Well, when a new group arrives on the server, Gnus by default makes it a zombie group. This means that you are normally not bothered with new groups, but you can type A z to get a list of all new groups. Subscribe the ones you like and kill the ones you don’t want. (A k shows a list of killed groups.)

If you want to play with the level variables, you should show some care. Set them once, and don’t touch them ever again. Better yet, don’t touch them at all unless you know exactly what you’re doing.

Two closely related variables are gnus-level-default-subscribed (default 3) and gnus-level-default-unsubscribed (default 6), which are the levels that new groups will be put on if they are (un)subscribed. These two variables should, of course, be inside the relevant valid ranges.

If gnus-keep-same-level is non-nil, some movement commands will only move to groups of the same level (or lower). In particular, going from the last article in one group to the next group will go to the next group of the same level (or lower). This might be handy if you want to read the most important groups before you read the rest.

If this variable is best, Gnus will make the next newsgroup the one with the best level.

All groups with a level less than or equal to gnus-group-default-list-level will be listed in the group buffer by default. This variable can also be a function. In that case, that function will be called and the result will be used as value.

If gnus-group-list-inactive-groups is non-nil, non-active groups will be listed along with the unread groups. This variable is t by default. If it is nil, inactive groups won’t be listed.

If gnus-group-use-permanent-levels is non-nil, once you give a level prefix to g or l, all subsequent commands will use this level as the “work” level.

Gnus will normally just activate (i.e., query the server about) groups on level gnus-activate-level or less. If you don’t want to activate unsubscribed groups, for instance, you might set this variable to 5. The default is 6.

3.7 Group Score

You would normally keep important groups on high levels, but that scheme is somewhat restrictive. Don’t you wish you could have Gnus sort the group buffer according to how often you read groups, perhaps? Within reason?

This is what group score is for. You can have Gnus assign a score to each group through the mechanism described below. You can then sort the group buffer based on this score. Alternatively, you can sort on score and then level. (Taken together, the level and the score is called the rank of the group. A group that is on level 4 and has a score of 1 has a higher rank than a group on level 5 that has a score of 300. (The level is the most significant part and the score is the least significant part.))

If you want groups you read often to get higher scores than groups you read seldom you can add the gnus-summary-bubble-group function to the gnus-summary-exit-hook hook. This will result (after sorting) in a bubbling sort of action. If you want to see that in action after each summary exit, you can add gnus-group-sort-groups-by-rank or gnus-group-sort-groups-by-score to the same hook, but that will slow things down somewhat.

3.8 Marking Groups

If you want to perform some command on several groups, and they appear subsequently in the group buffer, you would normally just give a numerical prefix to the command. Most group commands will then do your bidding on those groups.

However, if the groups are not in sequential order, you can still perform a command on several groups. You simply mark the groups first with the process mark and then execute the command.

M m

Toggle the process mark for the current group (gnus-group-mark-group).
If gnus-process-mark-toggle is nil, set the process mark for the current group.

M u

Remove the process mark, if any, from the current group (gnus-group-unmark-group).


Remove the process mark from all groups (gnus-group-unmark-all-groups).

M w

Mark groups in region (gnus-group-mark-region).

M b

Mark all groups in the buffer (gnus-group-mark-buffer).

M r

Mark all groups that match some regular expression (gnus-group-mark-regexp).

Also see Process/Prefix.

If you want to execute some command on all groups that have been marked with the process mark, you can use the M-& (gnus-group-universal-argument) command. It will prompt you for the command to be executed.

3.9 Foreign Groups

If you recall how to subscribe to servers (see Finding the News) you will remember that gnus-secondary-select-methods and gnus-select-method let you write a definition in Emacs Lisp of what servers you want to see when you start up. The alternate approach is to use foreign servers and groups. “Foreign” here means they are not coming from the select methods. All foreign server configuration and subscriptions are stored only in the ~/.newsrc.eld file.

Below are some group mode commands for making and editing general foreign groups, as well as commands to ease the creation of a few special-purpose groups. All these commands insert the newly created groups under point—gnus-subscribe-newsgroup-method is not consulted.

Changes from the group editing commands are stored in ~/.newsrc.eld (gnus-startup-file). An alternative is the variable gnus-parameters, See Group Parameters.

G m

Make a new group (gnus-group-make-group). Gnus will prompt you for a name, a method and possibly an address. For an easier way to subscribe to NNTP groups (see Browse Foreign Server).


Make an ephemeral group (gnus-group-read-ephemeral-group). Gnus will prompt you for a name, a method and an address.

G r

Rename the current group to something else (gnus-group-rename-group). This is valid only on some groups—mail groups mostly. This command might very well be quite slow on some back ends.

G c

Customize the group parameters (gnus-group-customize).

G e

Enter a buffer where you can edit the select method of the current group (gnus-group-edit-group-method).

G p

Enter a buffer where you can edit the group parameters (gnus-group-edit-group-parameters).


Enter a buffer where you can edit the group info (gnus-group-edit-group).

G d

Make a directory group (see Directory Groups). You will be prompted for the directory’s name (gnus-group-make-directory-group).

G h

Make the Gnus help group (gnus-group-make-help-group).


Read an arbitrary directory as if it were a newsgroup with the nneething back end (gnus-group-enter-directory). See Anything Groups.

G f

Make a group based on some file or other (gnus-group-make-doc-group). If you give a prefix to this command, you will be prompted for a file name and a file type. Currently supported types are mbox, babyl, digest, news, rnews, mmdf, forward, rfc934, rfc822-forward, mime-parts, standard-digest, slack-digest, clari-briefs, nsmail, outlook, oe-dbx, and mailman. If you run this command without a prefix, Gnus will guess at the file type. See Document Groups.

G u

Create one of the groups mentioned in gnus-useful-groups (gnus-group-make-useful-group).

G w

Make an ephemeral group based on a web search (gnus-group-make-web-group). If you give a prefix to this command, make a solid group instead. You will be prompted for the search engine type and the search string. Valid search engine types include google and dejanews. See Web Searches.

If you use the google search engine, you can limit the search to a particular group by using a match string like ‘shaving group:alt.sysadmin.recovery’.


Make a group based on an RSS feed (gnus-group-make-rss-group). You will be prompted for an URL. See RSS.


This function will delete the current group (gnus-group-delete-group). If given a prefix, this function will actually delete all the articles in the group, and forcibly remove the group itself from the face of the Earth. Use a prefix only if you are absolutely sure of what you are doing. This command can’t be used on read-only groups (like nntp groups), though.


Make a new, fresh, empty nnvirtual group (gnus-group-make-empty-virtual). See Virtual Groups.

G v

Add the current group to an nnvirtual group (gnus-group-add-to-virtual). Uses the process/prefix convention.

See Select Methods, for more information on the various select methods.

If gnus-activate-foreign-newsgroups is a positive number, Gnus will check all foreign groups with this level or lower at startup. This might take quite a while, especially if you subscribe to lots of groups from different NNTP servers. Also see Group Levels; gnus-activate-level also affects activation of foreign newsgroups.

The following commands create ephemeral groups. They can be called not only from the Group buffer, but in any Gnus buffer.


Read an Emacs bug report in an ephemeral group. Gnus will prompt for multiple bug numbers. The default is the number at point. The URL template is specified in gnus-bug-group-download-format-alist.


Read a Debian bug report in an ephemeral group. Analog to gnus-read-ephemeral-emacs-bug-group.

Some of these command are also useful for article buttons, See Article Buttons.

Here is an example:

(require 'gnus-art)
 '("#\\([0-9]+\\)\\>" 1
   (string-match "\\<emacs\\>" (or gnus-newsgroup-name ""))
   gnus-read-ephemeral-emacs-bug-group 1))

3.10 Group Parameters

The group parameters store information local to a particular group.

Use the G p or the G c command to edit group parameters of a group. (G p presents you with a Lisp-based interface, G c presents you with a Customize-like interface. The latter helps avoid silly Lisp errors.) You might also be interested in reading about topic parameters (see Topic Parameters). Additionally, you can set group parameters via the gnus-parameters variable, see below.

Here’s an example group parameter list:

((to-address . "")
 (auto-expire . t))

We see that each element consists of a “dotted pair”—the thing before the dot is the key, while the thing after the dot is the value. All the parameters have this form except local variable specs, which are not dotted pairs, but proper lists.

Some parameters have correspondent customizable variables, each of which is an alist of regexps and values.

The following group parameters can be used:


Address used by when doing followups and new posts.

(to-address . "")

This is primarily useful in mail groups that represent closed mailing lists—mailing lists where it’s expected that everybody that writes to the mailing list is subscribed to it. Since using this parameter ensures that the mail only goes to the mailing list itself, it means that members won’t receive two copies of your followups.

Using to-address will actually work whether the group is foreign or not. Let’s say there’s a group on the server that is called ‘fa.4ad-l’. This is a real newsgroup, but the server has gotten the articles from a mail-to-news gateway. Posting directly to this group is therefore impossible—you have to send mail to the mailing list address instead.

See also gnus-parameter-to-address-alist.


Address used when doing a in that group.

(to-list . "")

It is totally ignored when doing a followup—except that if it is present in a news group, you’ll get mail group semantics when doing f.

If you do an a command in a mail group and you have neither a to-list group parameter nor a to-address group parameter, then a to-list group parameter will be added automatically upon sending the message if gnus-add-to-list is set to t.

If this variable is set, gnus-mailing-list-mode is turned on when entering summary buffer.


If this parameter is set to t, Gnus will consider the to-address and to-list parameters for this group as addresses of mailing lists you are subscribed to. Giving Gnus this information is (only) a first step in getting it to generate correct Mail-Followup-To headers for your posts to these lists. The second step is to put the following in your .gnus.el

(setq message-subscribed-address-functions

See Mailing Lists in The Message Manual, for a complete treatment of available MFT support.


If the group parameter list has the element (visible . t), that group will always be visible in the Group buffer, regardless of whether it has any unread articles.

This parameter cannot be set via gnus-parameters. See gnus-permanently-visible-groups as an alternative.


Elements like (broken-reply-to . t) signals that Reply-To headers in this group are to be ignored, and for the header to be hidden if reply-to is part of gnus-boring-article-headers. This can be useful if you’re reading a mailing list group where the listserv has inserted Reply-To headers that point back to the listserv itself. That is broken behavior. So there!


Elements like (to-group . "") means that all posts in that group will be sent to


If you have (newsgroup . t) in the group parameter list, Gnus will treat all responses as if they were responses to news articles. This can be useful if you have a mail group that’s really a mirror of a news group.


If (gcc-self . t) is present in the group parameter list, newly composed messages will be gccd to the current group. If (gcc-self . none) is present, no Gcc: header will be generated, if (gcc-self . "group") is present, this string will be inserted literally as a Gcc: header. It should be a group name. The gcc-self value may also be a list of strings and t, e.g., (gcc-self "group1" "group2" t) means to gcc the newly composed message into the groups "group1" and "group2", and into the current group. The gcc-self parameter takes precedence over any default Gcc rules as described later (see Archived Messages), with the exception for messages to resend.

Caveat: Adding (gcc-self . t) to the parameter list of nntp groups (or the like) isn’t valid. An nntp server doesn’t accept articles.


If the group parameter has an element that looks like (auto-expire . t), all articles read will be marked as expirable. For an alternative approach, see Expiring Mail.

See also gnus-auto-expirable-newsgroups.


If the group parameter has an element that looks like (total-expire . t), all read articles will be put through the expiry process, even if they are not marked as expirable. Use with caution. Unread, ticked and dormant articles are not eligible for expiry.

See also gnus-total-expirable-newsgroups.


If the group parameter has an element that looks like (expiry-wait . 10), this value will override any nnmail-expiry-wait and nnmail-expiry-wait-function settings (see Expiring Mail) when expiring expirable messages. The value can be either a number of days (not necessarily an integer), or one of the symbols never or immediate.


Where expired messages end up. This parameter overrides nnmail-expiry-target.


Elements that look like (score-file . "file") will make file into the current score file for the group in question. All interactive score entries will be put into this file.


Elements that look like (adapt-file . "file") will make file into the current adaptive file for the group in question. All adaptive score entries will be put into this file.


When unsubscribing from a mailing list you should never send the unsubscription notice to the mailing list itself. Instead, you’d send messages to the administrative address. This parameter allows you to put the admin address somewhere convenient.


Elements that look like (display . MODE) say which articles to display on entering the group. Valid values are:


Display all articles, both read and unread.

an integer

Display the last integer articles in the group. This is the same as entering the group with C-u integer.


Display the default visible articles, which normally includes unread and ticked articles.

an array

Display articles that satisfy a predicate.

Here are some examples:


Display only unread articles.

[not expire]

Display everything except expirable articles.

[and (not reply) (not expire)]

Display everything except expirable and articles you’ve already responded to.

The available operators are not, and and or. Predicates include tick, unsend, undownload, unread, dormant, expire, reply, killed, bookmark, score, save, cache, forward, and unseen.

The display parameter works by limiting the summary buffer to the subset specified. You can pop the limit by using the / w command (see Limiting).


Elements that look like (comment . "This is a comment") are arbitrary comments on the group. You can display comments in the group line (see Group Line Specification).


Elements that look like (charset . iso-8859-1) will make iso-8859-1 the default charset; that is, the charset that will be used for all articles that do not specify a charset.

See also gnus-group-charset-alist.


Elements that look like (ignored-charsets x-unknown iso-8859-1) will make iso-8859-1 and x-unknown ignored; that is, the default charset will be used for decoding articles.

See also gnus-group-ignored-charsets-alist.


You can store additional posting style information for this group here (see Posting Styles). The format is that of an entry in the gnus-posting-styles alist, except that there’s no regexp matching the group name (of course). Style elements in this group parameter will take precedence over the ones found in gnus-posting-styles.

For instance, if you want a funky name and signature in this group only, instead of hacking gnus-posting-styles, you could put something like this in the group parameters:

  (name "Funky Name")
  ("X-Message-SMTP-Method" "smtp 587")
  ("X-My-Header" "Funky Value")
  (signature "Funky Signature"))

If you’re using topics to organize your group buffer (see Group Topics), note that posting styles can also be set in the topics parameters. Posting styles in topic parameters apply to all groups in this topic. More precisely, the posting-style settings for a group result from the hierarchical merging of all posting-style entries in the parameters of this group and all the topics it belongs to.


If it is set, the value is used as the method for posting message instead of gnus-post-method.


If it is set, and the setting of mail-sources includes a group mail source (see Mail Sources), the value is a mail source for this group.


An item like (banner . regexp) causes any part of an article that matches the regular expression regexp to be stripped. Instead of regexp, you can also use the symbol signature which strips the last signature or any of the elements of the alist gnus-article-banner-alist.


This parameter contains a Sieve test that should match incoming mail that should be placed in this group. From this group parameter, a Sieve ‘IF’ control structure is generated, having the test as the condition and ‘fileinto "";’ as the body.

For example, if the ‘INBOX.list.sieve’ group has the (sieve address "sender" "") group parameter, when translating the group parameter into a Sieve script (see Sieve Commands) the following Sieve code is generated:

if address "sender" "" {
        fileinto "INBOX.list.sieve";

To generate tests for multiple email-addresses use a group parameter like (sieve address "sender" ("" "")). When generating a sieve script (see Sieve Commands) Sieve code like the following is generated:

if address "sender" ["", ""] {
        fileinto "INBOX.list.sieve";

You can also use regexp expansions in the rules:

(sieve header :regex "list-id" "<c++std-\\>")

See Sieve Commands, for commands and variables that might be of interest in relation to the sieve parameter.

The Sieve language is described in RFC 3028. See Top in Emacs Sieve.


If this parameter is set to t and nnmail-split-method is set to gnus-group-split, Gnus will match to-address, to-list, extra-aliases and split-regexp against the list split abbreviation. The split regexp is modified to match either a @ or a dot . in mail addresses to conform to RFC2919 List-ID.

See nnmail-split-abbrev-alist for the regular expression matching mailing-list headers.

See Group Mail Splitting, for details on how to automatically split on group parameters.

(agent parameters)

If the agent has been enabled, you can set any of its parameters to control the behavior of the agent in individual groups. See Agent Parameters in Category Syntax. Most users will choose to set agent parameters in either an agent category or group topic to minimize the configuration effort.

(variable form)

You can use the group parameters to set variables local to the group you are entering. If you want to turn threading off in ‘news.answers’, you could put (gnus-show-threads nil) in the group parameters of that group. gnus-show-threads will be made into a local variable in the summary buffer you enter, and the form nil will be evaled there.

Note that this feature sets the variable locally to the summary buffer if and only if variable has been bound as a variable. Otherwise, only evaluating the form will take place. So, you may want to bind the variable in advance using defvar or other if the result of the form needs to be set to it.

But some variables are evaluated in the article buffer, or in the message buffer (of a reply or followup or otherwise newly created message). As a workaround, it might help to add the variable in question to gnus-newsgroup-variables. See Various Summary Stuff. So if you want to set message-from-style via the group parameters, then you may need the following statement elsewhere in your ~/.gnus.el file:

(add-to-list 'gnus-newsgroup-variables 'message-from-style)

A use for this feature is to remove a mailing list identifier tag in the subject fields of articles. E.g., if the news group

has the tag ‘DOC-BOOK-APPS:’ in the subject of all articles, this tag can be removed from the article subjects in the summary buffer for the group by putting (gnus-list-identifiers "DOCBOOK-APPS:") into the group parameters for the group.

This can also be used as a group-specific hook function. If you want to hear a beep when you enter a group, you could put something like (dummy-variable (ding)) in the parameters of that group. If dummy-variable has been bound (see above), it will be set to the (meaningless) result of the (ding) form.

Alternatively, since the VARIABLE becomes local to the group, this pattern can be used to temporarily change a hook. For example, if the following is added to a group parameter

  (lambda nil (local-set-key "d" (local-key-binding "n"))))

when the group is entered, the ’d’ key will not mark the article as expired.

Group parameters can be set via the gnus-parameters variable too. But some variables, such as visible, have no effect (For this case see gnus-permanently-visible-groups as an alternative.). For example:

(setq gnus-parameters
         (gnus-show-threads nil)
         (gnus-use-scoring nil)
          "%U%R%z%I%(%[%d:%ub%-23,23f%]%) %s\n")
         (gcc-self . t)
         (display . all))

         (to-group . "\\1"))

         (gnus-use-scoring t))

         (total-expire . t)
         (broken-reply-to . t))))

All clauses that match the group name will be used, but the last setting “wins”. So if you have two clauses that both match the group name, and both set, say display, the last setting will override the first.

Parameters that are strings will be subjected to regexp substitution, as the to-group example shows.

By default, whether comparing the group name and one of those regexps specified in gnus-parameters is done in a case-sensitive manner or a case-insensitive manner depends on the value of case-fold-search at the time when the comparison is done. The value of case-fold-search is typically t; it means, for example, the element ("INBOX\\.FOO" (total-expire . t)) might be applied to both the ‘INBOX.FOO’ group and the ‘’ group. If you want to make those regexps always case-sensitive, set the value of the gnus-parameters-case-fold-search variable to nil. Otherwise, set it to t if you want to compare them always in a case-insensitive manner.

You can define different sorting to different groups via gnus-parameters. Here is an example to sort an NNTP group by reverse date to see the latest news at the top and an RSS group by subject. In this example, the first group is the Debian daily news group from The RSS group corresponds to the Debian weekly news RSS feed, See RSS.

    (gnus-show-threads nil)
    (gnus-article-sort-functions '((not gnus-article-sort-by-date)))
    (gnus-use-adaptive-scoring nil)
    (gnus-use-scoring nil))
    (gnus-show-threads nil)
    (gnus-article-sort-functions 'gnus-article-sort-by-subject)
    (gnus-use-adaptive-scoring nil)
    (gnus-use-scoring t)
    (gnus-score-find-score-files-function 'gnus-score-find-single)
    (gnus-summary-line-format "%U%R%z%d %I%(%[ %s %]%)\n"))))

3.11 Listing Groups

These commands all list various slices of the groups available.

A s

List all groups that have unread articles (gnus-group-list-groups). If the numeric prefix is used, this command will list only groups of level ARG and lower. By default, it only lists groups of level five (i.e., gnus-group-default-list-level) or lower (i.e., just subscribed groups).

A u

List all groups, whether they have unread articles or not (gnus-group-list-all-groups). If the numeric prefix is used, this command will list only groups of level ARG and lower. By default, it lists groups of level seven or lower (i.e., just subscribed and unsubscribed groups).

A l

List all unread groups on a specific level (gnus-group-list-level). If given a prefix, also list the groups with no unread articles.

A k

List all killed groups (gnus-group-list-killed). If given a prefix argument, really list all groups that are available, but aren’t currently (un)subscribed. This could entail reading the active file from the server.

A z

List all zombie groups (gnus-group-list-zombies).

A m

List all unread, subscribed groups with names that match a regexp (gnus-group-list-matching).


List groups that match a regexp (gnus-group-list-all-matching).


List absolutely all groups in the active file(s) of the server(s) you are connected to (gnus-group-list-active). This might very well take quite a while. It might actually be a better idea to do a A M to list all matching, and just give ‘.’ as the thing to match on. Also note that this command may list groups that don’t exist (yet)—these will be listed as if they were killed groups. Take the output with some grains of salt.

A a

List all groups that have names that match a regexp (gnus-group-apropos).

A d

List all groups that have names or descriptions that match a regexp (gnus-group-description-apropos).

A c

List all groups with cached articles (gnus-group-list-cached).

A ?

List all groups with dormant articles (gnus-group-list-dormant).

A !

List all groups with ticked articles (gnus-group-list-ticked).

A /

Further limit groups within the current selection (gnus-group-list-limit). If you’ve first limited to groups with dormant articles with A ?, you can then further limit with A / c, which will then limit to groups with cached articles, giving you the groups that have both dormant articles and cached articles.

A f

Flush groups from the current selection (gnus-group-list-flush).

A p

List groups plus the current selection (gnus-group-list-plus).

Groups that match the gnus-permanently-visible-groups regexp will always be shown, whether they have unread articles or not. You can also add the visible element to the group parameters in question to get the same effect.

Groups that have just ticked articles in it are normally listed in the group buffer. If gnus-list-groups-with-ticked-articles is nil, these groups will be treated just like totally empty groups. It is t by default.

3.12 Sorting Groups

The C-c C-s (gnus-group-sort-groups) command sorts the group buffer according to the function(s) given by the gnus-group-sort-function variable. Available sorting functions include:


Sort the group names alphabetically. This is the default.


Sort the group alphabetically on the real (unprefixed) group names.


Sort by group level.


Sort by group score. See Group Score.


Sort by group score and then the group level. The level and the score are, when taken together, the group’s rank. See Group Score.


Sort by number of unread articles.


Sort alphabetically on the select method.


Sort alphabetically on the Gnus server name.

gnus-group-sort-function can also be a list of sorting functions. In that case, the most significant sort key function must be the last one.

There are also a number of commands for sorting directly according to some sorting criteria:

G S a

Sort the group buffer alphabetically by group name (gnus-group-sort-groups-by-alphabet).

G S u

Sort the group buffer by the number of unread articles (gnus-group-sort-groups-by-unread).

G S l

Sort the group buffer by group level (gnus-group-sort-groups-by-level).

G S v

Sort the group buffer by group score (gnus-group-sort-groups-by-score). See Group Score.

G S r

Sort the group buffer by group rank (gnus-group-sort-groups-by-rank). See Group Score.

G S m

Sort the group buffer alphabetically by back end name

G S n

Sort the group buffer alphabetically by real (unprefixed) group name (gnus-group-sort-groups-by-real-name).

All the commands below obey the process/prefix convention (see Process/Prefix).

When given a symbolic prefix (see Symbolic Prefixes), all these commands will sort in reverse order.

You can also sort a subset of the groups:

G P a

Sort the groups alphabetically by group name (gnus-group-sort-selected-groups-by-alphabet).

G P u

Sort the groups by the number of unread articles (gnus-group-sort-selected-groups-by-unread).

G P l

Sort the groups by group level (gnus-group-sort-selected-groups-by-level).

G P v

Sort the groups by group score (gnus-group-sort-selected-groups-by-score). See Group Score.

G P r

Sort the groups by group rank (gnus-group-sort-selected-groups-by-rank). See Group Score.

G P m

Sort the groups alphabetically by back end name

G P n

Sort the groups alphabetically by real (unprefixed) group name (gnus-group-sort-selected-groups-by-real-name).

G P s

Sort the groups according to gnus-group-sort-function.

And finally, note that you can use C-k and C-y to manually move groups around.

3.13 Group Maintenance


Find bogus groups and delete them (gnus-group-check-bogus-groups).


Find new groups and process them (gnus-group-find-new-groups). With 1 C-u, use the ask-server method to query the server for new groups. With 2 C-u’s, use most complete method possible to query the server for new groups, and subscribe the new groups as zombies.

C-c C-x

Run all expirable articles in the current group through the expiry process (if any) (gnus-group-expire-articles). That is, delete all expirable articles in the group that have been around for a while. (see Expiring Mail).

C-c C-M-x

Run all expirable articles in all groups through the expiry process (gnus-group-expire-all-groups).

3.14 Browse Foreign Server


You will be queried for a select method and a server name. Gnus will then attempt to contact this server and let you browse the groups there (gnus-group-browse-foreign-server).

A new buffer with a list of available groups will appear. This buffer will use the gnus-browse-mode. This buffer looks a bit (well, a lot) like a normal group buffer.

Here’s a list of keystrokes available in the browse mode:


Go to the next group (gnus-group-next-group).


Go to the previous group (gnus-group-prev-group).


Enter the current group and display the first article (gnus-browse-read-group).


Enter the current group (gnus-browse-select-group).


Toggle subscription of the current group (gnus-browse-toggle-subscription). You can affect the way the new group is entered into the Group buffer using the variable gnus-browse-subscribe-newsgroup-method. See see Subscription Methods for available options.


Exit browse mode (gnus-browse-exit).


Describe the current group (gnus-browse-describe-group).


Describe browse mode briefly (well, there’s not much to describe, is there) (gnus-browse-describe-briefly).


This function will delete the current group (gnus-browse-delete-group). If given a prefix, this function will actually delete all the articles in the group, and forcibly remove the group itself from the face of the Earth. Use a prefix only if you are absolutely sure of what you are doing.

3.15 Exiting Gnus

Yes, Gnus is ex(c)iting.


Suspend Gnus (gnus-group-suspend). This doesn’t really exit Gnus, but it kills all buffers except the Group buffer. I’m not sure why this is a gain, but then who am I to judge?


Quit Gnus (gnus-group-exit).


Quit Gnus without saving the .newsrc files (gnus-group-quit). The dribble file will be saved, though (see Auto Save).

gnus-suspend-gnus-hook is called when you suspend Gnus and gnus-exit-gnus-hook is called when you quit Gnus, while gnus-after-exiting-gnus-hook is called as the final item when exiting Gnus.

3.16 Group Topics

If you read lots and lots of groups, it might be convenient to group them hierarchically according to topics. You put your Emacs groups over here, your sex groups over there, and the rest (what, two groups or so?) you put in some misc section that you never bother with anyway. You can even group the Emacs sex groups as a sub-topic to either the Emacs groups or the sex groups—or both! Go wild!

Here’s an example:

  Emacs -- I wuw it!
     3: comp.emacs
     2: alt.religion.emacs
    Naughty Emacs
     8: comp.binaries.fractals
    13: comp.sources.unix

To get this fab functionality you simply turn on (ooh!) the gnus-topic minor mode—type t in the group buffer. (This is a toggling command.)

Go ahead, just try it. I’ll still be here when you get back. La de dum… Nice tune, that… la la la… What, you’re back? Yes, and now press l. There. All your groups are now listed under ‘misc’. Doesn’t that make you feel all warm and fuzzy? Hot and bothered?

If you want this permanently enabled, you should add that minor mode to the hook for the group mode. Put the following line in your ~/.gnus.el file:

(add-hook 'gnus-group-mode-hook 'gnus-topic-mode)

3.16.1 Topic Commands

When the topic minor mode is turned on, a new T submap will be available. In addition, a few of the standard keys change their definitions slightly.

In general, the following kinds of operations are possible on topics. First of all, you want to create topics. Secondly, you want to put groups in topics and to move them around until you have an order you like. The third kind of operation is to show/hide parts of the whole shebang. You might want to hide a topic including its subtopics and groups, to get a better overview of the other groups.

Here is a list of the basic keys that you might need to set up topics the way you like.

T n

Prompt for a new topic name and create it (gnus-topic-create-topic).


“Indent” the current topic so that it becomes a sub-topic of the previous topic (gnus-topic-indent). If given a prefix, “un-indent” the topic instead.


“Un-indent” the current topic so that it becomes a sub-topic of the parent of its current parent (gnus-topic-unindent).

The following two keys can be used to move groups and topics around. They work like the well-known cut and paste. C-k is like cut and C-y is like paste. Of course, this being Emacs, we use the terms kill and yank rather than cut and paste.


Kill a group or topic (gnus-topic-kill-group). All groups in the topic will be removed along with the topic.


Yank the previously killed group or topic (gnus-topic-yank-group). Note that all topics will be yanked before all groups.

So, to move a topic to the beginning of the list of topics, just hit C-k on it. This is like the “cut” part of cut and paste. Then, move the cursor to the beginning of the buffer (just below the “Gnus” topic) and hit C-y. This is like the “paste” part of cut and paste. Like I said—E-Z.

You can use C-k and C-y on groups as well as on topics. So you can move topics around as well as groups.

After setting up the topics the way you like them, you might wish to hide a topic, or to show it again. That’s why we have the following key.


Either select a group or fold a topic (gnus-topic-select-group). When you perform this command on a group, you’ll enter the group, as usual. When done on a topic line, the topic will be folded (if it was visible) or unfolded (if it was folded already). So it’s basically a toggling command on topics. In addition, if you give a numerical prefix, group on that level (and lower) will be displayed.

Now for a list of other commands, in no particular order.

T m

Move the current group to some other topic (gnus-topic-move-group). This command uses the process/prefix convention (see Process/Prefix).

T j

Go to a topic (gnus-topic-jump-to-topic).

T c

Copy the current group to some other topic (gnus-topic-copy-group). This command uses the process/prefix convention (see Process/Prefix).

T h

Hide the current topic (gnus-topic-hide-topic). If given a prefix, hide the topic permanently.

T s

Show the current topic (gnus-topic-show-topic). If given a prefix, show the topic permanently.


Remove a group from the current topic (gnus-topic-remove-group). This command is mainly useful if you have the same group in several topics and wish to remove it from one of the topics. You may also remove a group from all topics, but in that case, Gnus will add it to the root topic the next time you start Gnus. In fact, all new groups (which, naturally, don’t belong to any topic) will show up in the root topic.

This command uses the process/prefix convention (see Process/Prefix).


Move all groups that match some regular expression to a topic (gnus-topic-move-matching).


Copy all groups that match some regular expression to a topic (gnus-topic-copy-matching).


Toggle hiding empty topics (gnus-topic-toggle-display-empty-topics).

T #

Toggle the process mark for all groups in the current topic (gnus-topic-mark-topic). This command works recursively on sub-topics unless given a prefix.
If gnus-process-mark-toggle is nil, set the process mark for the current topic.

T M-#

Remove the process mark from all groups in the current topic (gnus-topic-unmark-topic). This command works recursively on sub-topics unless given a prefix.

C-c C-x

Run all expirable articles in the current group or topic through the expiry process (if any) (gnus-topic-expire-articles). (see Expiring Mail).

T r

Rename a topic (gnus-topic-rename).


Delete an empty topic (gnus-topic-delete).


List all groups that Gnus knows about in a topics-ified way (gnus-topic-list-active).

T M-n

Go to the next topic (gnus-topic-goto-next-topic).

T M-p

Go to the previous topic (gnus-topic-goto-previous-topic).

G p

Edit the topic parameters (gnus-topic-edit-parameters). See Topic Parameters.

3.16.2 Topic Variables

The previous section told you how to tell Gnus which topics to display. This section explains how to tell Gnus what to display about each topic.

The topic lines themselves are created according to the gnus-topic-line-format variable (see Formatting Variables). Valid elements are:




Topic name.






Number of groups in the topic.


Number of groups in the topic and all its subtopics.


Number of unread articles in the topic.


Number of unread articles in the topic and all its subtopics.

Each sub-topic (and the groups in the sub-topics) will be indented with gnus-topic-indent-level times the topic level number of spaces. The default is 2.

gnus-topic-mode-hook is called in topic minor mode buffers.

The gnus-topic-display-empty-topics says whether to display even topics that have no unread articles in them. The default is t.

If gnus-topic-display-predicate is non-nil, it should be a function that says whether the topic is to be displayed or not. The function will be called with one parameter (the name of the topic) and should return non-nil is the topic is to be displayed.

For instance, if you don’t even want to be reminded that work exists outside of office hours, you can gather all the work-related groups into a topic called ‘"Work"’, and then say something like the following:

(setq gnus-topic-display-predicate
      (lambda (name)
        (or (not (equal name "Work"))
            (< 090000
               (string-to-number (format-time-string "%H%M%S"))

3.16.3 Topic Sorting

You can sort the groups in each topic individually with the following commands:

T S a

Sort the current topic alphabetically by group name (gnus-topic-sort-groups-by-alphabet).

T S u

Sort the current topic by the number of unread articles (gnus-topic-sort-groups-by-unread).

T S l

Sort the current topic by group level (gnus-topic-sort-groups-by-level).

T S v

Sort the current topic by group score (gnus-topic-sort-groups-by-score). See Group Score.

T S r

Sort the current topic by group rank (gnus-topic-sort-groups-by-rank). See Group Score.

T S m

Sort the current topic alphabetically by back end name (gnus-topic-sort-groups-by-method).

T S e

Sort the current topic alphabetically by server name (gnus-topic-sort-groups-by-server).

T S s

Sort the current topic according to the function(s) given by the gnus-group-sort-function variable (gnus-topic-sort-groups).

When given a prefix argument, all these commands will sort in reverse order. See Sorting Groups, for more information about group sorting.

3.16.4 Topic Topology

So, let’s have a look at an example group buffer:

  Emacs -- I wuw it!
     3: comp.emacs
     2: alt.religion.emacs
    Naughty Emacs
     8: comp.binaries.fractals
    13: comp.sources.unix

So, here we have one top-level topic (‘Gnus’), two topics under that, and one sub-topic under one of the sub-topics. (There is always just one (1) top-level topic). This topology can be expressed as follows:

(("Gnus" visible)
 (("Emacs -- I wuw it!" visible)
  (("Naughty Emacs" visible)))
 (("Misc" visible)))

This is in fact how the variable gnus-topic-topology would look for the display above. That variable is saved in the .newsrc.eld file, and shouldn’t be messed with manually—unless you really want to. Since this variable is read from the .newsrc.eld file, setting it in any other startup files will have no effect.

This topology shows what topics are sub-topics of what topics (right), and which topics are visible. Two settings are currently allowed—visible and invisible.

3.16.5 Topic Parameters

All groups in a topic will inherit group parameters from the parent (and ancestor) topic parameters. All valid group parameters are valid topic parameters (see Group Parameters). When the agent is enabled, all agent parameters (See Agent Parameters in Category Syntax) are also valid topic parameters.

In addition, the following parameters are only valid as topic parameters:


When subscribing new groups by topic (see Subscription Methods), the subscribe topic parameter says what groups go in what topic. Its value should be a regexp to match the groups that should go in that topic.


When subscribing new groups by topic (see the subscribe parameter), the group will be subscribed with the level specified in the subscribe-level instead of gnus-level-default-subscribed.

Group parameters (of course) override topic parameters, and topic parameters in sub-topics override topic parameters in super-topics. You know. Normal inheritance rules. (Rules is here a noun, not a verb, although you may feel free to disagree with me here.)

     3: comp.emacs
     2: alt.religion.emacs
     8: comp.binaries.fractals
    13: comp.sources.unix

The ‘Emacs’ topic has the topic parameter (score-file . "emacs.SCORE"); the ‘Relief’ topic has the topic parameter (score-file . "relief.SCORE"); and the ‘Misc’ topic has the topic parameter (score-file . "emacs.SCORE"). In addition,
alt.religion.emacs’ has the group parameter (score-file . "religion.SCORE").

Now, when you enter ‘’ in the ‘Relief’ topic, you will get the relief.SCORE home score file. If you enter the same group in the ‘Emacs’ topic, you’ll get the emacs.SCORE home score file. If you enter the group ‘alt.religion.emacs’, you’ll get the religion.SCORE home score file.

This seems rather simple and self-evident, doesn’t it? Well, yes. But there are some problems, especially with the total-expiry parameter. Say you have a mail group in two topics; one with total-expiry and one without. What happens when you do M-x gnus-expire-all-expirable-groups? Gnus has no way of telling which one of these topics you mean to expire articles from, so anything may happen. In fact, I hereby declare that it is undefined what happens. You just have to be careful if you do stuff like that.

3.17 Accessing groups of non-English names

There are some news servers that provide groups of which the names are expressed with their native languages in the world. For instance, in a certain news server there are some newsgroups of which the names are spelled in Chinese, where people are talking in Chinese. You can, of course, subscribe to such news groups using Gnus. Currently Gnus supports non-ASCII group names not only with the nntp back end but also with the nnml back end and the nnrss back end.

Every such group name is encoded by a certain charset in the server side (in an NNTP server its administrator determines the charset, but for groups in the other back ends it is determined by you). Gnus has to display the decoded ones for you in the group buffer and the article buffer, and needs to use the encoded ones when communicating with servers. However, Gnus doesn’t know what charset is used for each non-ASCII group name. The following two variables are just the ones for telling Gnus what charset should be used for each group:


An alist of select methods and charsets. The default value is nil. The names of groups in the server specified by that select method are all supposed to use the corresponding charset. For example:

(setq gnus-group-name-charset-method-alist
      '(((nntp "") . cn-gb-2312)))

Charsets specified for groups with this variable are preferred to the ones specified for the same groups with the gnus-group-name-charset-group-alist variable (see below).

A select method can be very long, like:

(nntp "gmane"
      (nntp-address "")
      (nntp-end-of-line "\n")
      (nntp-via-rlogin-command "ssh")
       ("-C" "-t" "-e" "none"))
      (nntp-via-address …))

In that case, you can truncate it into (nntp "gmane") in this variable. That is, it is enough to contain only the back end name and the server name.


An alist of regexp of group name and the charset for group names. ((".*" . utf-8)) is the default value if UTF-8 is supported, otherwise the default is nil. For example:

(setq gnus-group-name-charset-group-alist
      '(("\\.com\\.cn:" . cn-gb-2312)
        (".*" . utf-8)))

Note that this variable is ignored if the match is made with gnus-group-name-charset-method-alist.

Those two variables are used also to determine the charset for encoding and decoding non-ASCII group names that are in the back ends other than nntp. It means that it is you who determine it. If you do nothing, the charset used for group names in those back ends will all be utf-8 because of the last element of gnus-group-name-charset-group-alist.

There is one more important variable for non-ASCII group names:


The value of this variable should be a coding system or nil. The default is nil in Emacs.

The nnml back end, the nnrss back end, the agent, and the cache use non-ASCII group names in those files and directories. This variable overrides the value of file-name-coding-system which specifies the coding system used when encoding and decoding those file names and directory names.

Emacs uses the value of default-file-name-coding-system if file-name-coding-system is nil or it is bound to the value of nnmail-pathname-coding-system which is nil.

Normally the value of default-file-name-coding-system is initialized according to the locale, so you will need to do nothing if the value is suitable to encode and decode non-ASCII group names.

The value of this variable (or default-file-name-coding-system) does not necessarily need to be the same value that is determined by gnus-group-name-charset-method-alist and gnus-group-name-charset-group-alist.

If default-file-name-coding-system or this variable is initialized by default to iso-latin-1-unix for example, although you want to subscribe to the groups spelled in Chinese, that is the most typical case where you have to customize nnmail-pathname-coding-system. The utf-8-unix coding system is a good candidate for it. Otherwise, you may change the locale in your system so that default-file-name-coding-system or this variable may be initialized to an appropriate value.

Note that when you copy or move articles from a non-ASCII group to another group, the charset used to encode and decode group names should be the same in both groups. Otherwise the Newsgroups header will be displayed incorrectly in the article buffer.

3.18 Misc Group Stuff


The key v is reserved for users. You can bind it to some command or better use it as a prefix key. For example:

(define-key gnus-group-mode-map (kbd "v j d")
  (lambda ()
    (gnus-group-jump-to-group "nndraft:drafts")))

On keys reserved for users in Emacs and on key bindings in general See Keymaps in The Emacs Editor.


Enter the server buffer (gnus-group-enter-server-mode). See Server Buffer.


Start composing a message (a news by default) (gnus-group-post-news). If given a prefix, post to the group under the point. If the prefix is 1, prompt for a group to post to. Contrary to what the name of this function suggests, the prepared article might be a mail instead of a news, if a mail group is specified with the prefix argument. See Composing Messages.


Mail a message somewhere (gnus-group-mail). If given a prefix, use the posting style of the group under the point. If the prefix is 1, prompt for a group name to find the posting style. See Composing Messages.


Start composing a news (gnus-group-news). If given a prefix, post to the group under the point. If the prefix is 1, prompt for group to post to. See Composing Messages.

This function actually prepares a news even when using mail groups. This is useful for “posting” messages to mail groups without actually sending them over the network: they’re just saved directly to the group in question. The corresponding back end must have a request-post method for this to work though.

G z

Compact the group under point (gnus-group-compact-group). Currently implemented only in nnml (see Mail Spool). This removes gaps between article numbers, hence getting a correct total article count.

Variables for the group buffer:


is called after the group buffer has been created.


is called after the group buffer is generated. It may be used to modify the buffer in some strange, unnatural way.


is called as the very last thing after the group buffer has been generated. It may be used to move point around, for instance.


Groups matching this regexp will always be listed in the group buffer, whether they are empty or not.

3.18.1 Scanning New Messages


Check the server(s) for new articles. If the numerical prefix is used, this command will check only groups of level arg and lower (gnus-group-get-new-news). If given a non-numerical prefix, this command will force a total re-reading of the active file(s) from the back end(s).


Check whether new articles have arrived in the current group (gnus-group-get-new-news-this-group). gnus-goto-next-group-when-activating says whether this command is to move point to the next group or not. It is t by default.

C-c M-g

Activate absolutely all groups (gnus-activate-all-groups).


Restart Gnus (gnus-group-restart). This saves the .newsrc file(s), closes the connection to all servers, clears up all run-time Gnus variables, and then starts Gnus all over again.

gnus-get-new-news-hook is run just before checking for new news.

gnus-after-getting-new-news-hook is run after checking for new news.

3.18.2 Group Information

H d
C-c C-d

Describe the current group (gnus-group-describe-group). If given a prefix, force Gnus to re-read the description from the server.


Describe all groups (gnus-group-describe-all-groups). If given a prefix, force Gnus to re-read the description file from the server.

H v

Display current Gnus version numbers (gnus-version).


Give a very short help message (gnus-group-describe-briefly).

C-c C-i

Go to the Gnus info node (gnus-info-find-node).

3.18.3 Group Timestamp

It can be convenient to let Gnus keep track of when you last read a group. To set the ball rolling, you should add gnus-group-set-timestamp to gnus-select-group-hook:

(add-hook 'gnus-select-group-hook 'gnus-group-set-timestamp)

After doing this, each time you enter a group, it’ll be recorded.

This information can be displayed in various ways—the easiest is to use the ‘%d’ spec in the group line format:

(setq gnus-group-line-format
      "%M\%S\%p\%P\%5y: %(%-40,40g%) %d\n")

This will result in lines looking like:

*        0: mail.ding                                19961002T012943
         0: custom                                   19961002T012713

As you can see, the date is displayed in compact ISO 8601 format. This may be a bit too much, so to just display the date, you could say something like:

(setq gnus-group-line-format
      "%M\%S\%p\%P\%5y: %(%-40,40g%) %6,6~(cut 2)d\n")

If you would like greater control of the time format, you can use a user-defined format spec. Something like the following should do the trick:

(setq gnus-group-line-format
      "%M\%S\%p\%P\%5y: %(%-40,40g%) %ud\n")
(defun gnus-user-format-function-d (headers)
  (let ((time (gnus-group-timestamp gnus-tmp-group)))
    (if time
        (format-time-string "%b %d  %H:%M" time)

To see what variables are dynamically bound (like gnus-tmp-group), you have to look at the source code. The variable names aren’t guaranteed to be stable over Gnus versions, either.

3.18.4 File Commands


Re-read the init file (gnus-init-file, which defaults to ~/.gnus.el) (gnus-group-read-init-file).


Save the .newsrc.eld file (and .newsrc if wanted) (gnus-group-save-newsrc). If given a prefix, force saving the file(s) whether Gnus thinks it is necessary or not.

3.18.5 Sieve Commands

Sieve is a server-side mail filtering language. In Gnus you can use the sieve group parameter (see Group Parameters) to specify sieve rules that should apply to each group. Gnus provides two commands to translate all these group parameters into a proper Sieve script that can be transferred to the server somehow.

The generated Sieve script is placed in gnus-sieve-file (by default ~/.sieve). The Sieve code that Gnus generate is placed between two delimiters, gnus-sieve-region-start and gnus-sieve-region-end, so you may write additional Sieve code outside these delimiters that will not be removed the next time you regenerate the Sieve script.

The variable gnus-sieve-crosspost controls how the Sieve script is generated. If it is non-nil (the default) articles is placed in all groups that have matching rules, otherwise the article is only placed in the group with the first matching rule. For example, the group parameter ‘(sieve address "sender" "")’ will generate the following piece of Sieve code if gnus-sieve-crosspost is nil. (When gnus-sieve-crosspost is non-nil, it looks the same except that the line containing the call to stop is removed.)

if address "sender" "" {
        fileinto "INBOX.ding";

See Top in Emacs Sieve.

D g

Regenerate a Sieve script from the sieve group parameters and put you into the gnus-sieve-file without saving it.

D u

Regenerates the Gnus managed part of gnus-sieve-file using the sieve group parameters, save the file and upload it to the server using the sieveshell program.

4 Summary Buffer

A line for each article is displayed in the summary buffer. You can move around, read articles, post articles and reply to articles.

The most common way to a summary buffer is to select a group from the group buffer (see Selecting a Group).

You can have as many summary buffers open as you wish.

You can customize the Summary Mode tool bar, see M-x customize-apropos RET gnus-summary-tool-bar.

The key v is reserved for users. You can bind it to some command or better use it as a prefix key. For example:

(define-key gnus-summary-mode-map (kbd "v -") "LrS") ;; lower subthread

4.1 Summary Buffer Format

Gnus will use the value of the gnus-extract-address-components variable as a function for getting the name and address parts of a From header. Two pre-defined functions exist: gnus-extract-address-components, which is the default, quite fast, and too simplistic solution; and mail-extract-address-components, which works very nicely, but is slower. The default function will return the wrong answer in 5% of the cases. If this is unacceptable to you, use the other function instead:

(setq gnus-extract-address-components

gnus-summary-same-subject is a string indicating that the current article has the same subject as the previous. This string will be used with those specs that require it. The default is "".

4.1.1 Summary Buffer Lines

You can change the format of the lines in the summary buffer by changing the gnus-summary-line-format variable. It works along the same lines as a normal format string, with some extensions (see Formatting Variables).

There should always be a colon or a point position marker on the line; the cursor always moves to the point position marker or the colon after performing an operation. (Of course, Gnus wouldn’t be Gnus if it wasn’t possible to change this. Just write a new function gnus-goto-colon which does whatever you like with the cursor.) See Positioning Point.

The default string is ‘%U%R%z%I%(%[%4L: %-23,23f%]%) %s\n’.

The following format specification characters and extended format specification(s) are understood:


Article number.


Subject string. List identifiers stripped, gnus-list-identifiers. See Article Hiding.


Subject if the article is the root of the thread or the previous article had a different subject, gnus-summary-same-subject otherwise. (gnus-summary-same-subject defaults to "".)


Full From header.


The name (from the From header).


The name, To header or the Newsgroups header (see To From Newsgroups).


The name (from the From header). This differs from the n spec in that it uses the function designated by the gnus-extract-address-components variable, which is slower, but may be more thorough.


The address (from the From header). This works the same way as the a spec.


Number of lines in the article.


Retrieval Score Value (RSV) of the article; nil if not in an nnselect group.


Originating group name of the article; nil if not in an nnselect group.


Short form of the originating group name of the article; nil if not in an nnselect group.


Number of characters in the article. This specifier is not supported in some methods (like nnfolder).


Pretty-printed version of the number of characters in the article; for example, ‘1.2k’ or ‘0.4M’.


Indentation based on thread level (see Customizing Threading).


A complex trn-style thread tree, showing response-connecting trace lines. A thread could be drawn like this:

| +->
| | \->
| |   \->
| \->

You can customize the appearance with the following options. Note that it is possible to make the thread display look really neat by replacing the default ASCII characters with graphic line-drawing glyphs.


Used for the root of a thread. If nil, use subject instead. The default is ‘> ’.


Used for the false root of a thread (see Loose Threads). If nil, use subject instead. The default is ‘> ’.


Used for a thread with just one message. If nil, use subject instead. The default is ‘’.


Used for drawing a vertical line. The default is ‘| ’.


Used for indenting. The default is ‘ ’.


Used for a leaf with brothers. The default is ‘+-> ’.


Used for a leaf without brothers. The default is ‘\->


Nothing if the article is a root and lots of spaces if it isn’t (it pushes everything after it off the screen).


Opening bracket, which is normally ‘[’, but can also be ‘<’ for adopted articles (see Customizing Threading). This can be customized using following settings:


Opening bracket for normal (non-adopted) articles. The default is ‘[’.


Opening bracket for adopted articles. The default is ‘<’.


Closing bracket, which is normally ‘]’, but can also be ‘>’ for adopted articles. This can be customized using following settings:


Closing bracket for normal (non-adopted) articles. The default is ‘]’.


Closing bracket for adopted articles. The default is ‘>’.


One space for each thread level.


Twenty minus thread level spaces.


Unread. See Read Articles.


This misleadingly named specifier is the secondary mark. This mark will say whether the article has been replied to, has been cached, or has been saved. See Other Marks.


Score as a number (see Scoring).


Zcore, ‘+’ if above the default level and ‘-’ if below the default level. If the difference between gnus-summary-default-score and the score is less than gnus-summary-zcore-fuzz, this spec will not be used.


Total thread score.






The Date in DD-MMM format.


The Date in YYYYMMDDTHHMMSS format.






Number of articles in the current sub-thread. Using this spec will slow down summary buffer generation somewhat.


An ‘=’ (gnus-not-empty-thread-mark) will be displayed if the article has any children.


The line number.


Download mark.


Desired cursor position (instead of after first colon).


Age sensitive date format. Various date format is defined in gnus-user-date-format-alist.


User defined specifier. The next character in the format string should be a letter. Gnus will call the function gnus-user-format-function-x, where x is the letter following ‘%u’. The function will be passed the current header as argument. The function should return a string, which will be inserted into the summary just like information from any other summary specifier.

Text between ‘%(’ and ‘%)’ will be highlighted with gnus-mouse-face when the mouse point is placed inside the area. There can only be one such area.

The ‘%U’ (status), ‘%R’ (replied) and ‘%z’ (zcore) specs have to be handled with care. For reasons of efficiency, Gnus will compute what column these characters will end up in, and “hard-code” that. This means that it is invalid to have these specs after a variable-length spec. Well, you might not be arrested, but your summary buffer will look strange, which is bad enough.

The smart choice is to have these specs as far to the left as possible. (Isn’t that the case with everything, though? But I digress.)

This restriction may disappear in later versions of Gnus.

4.1.2 To From Newsgroups

In some groups (particularly in archive groups), the From header isn’t very interesting, since all the articles there are written by you. To display the information in the To or Newsgroups headers instead, you need to decide three things: What information to gather; where to display it; and when to display it.

  1. The reading of extra header information is controlled by the gnus-extra-headers. This is a list of header symbols. For instance:
    (setq gnus-extra-headers
          '(To Newsgroups X-Newsreader))

    This will result in Gnus trying to obtain these three headers, and storing it in header structures for later easy retrieval.

  2. The value of these extra headers can be accessed via the gnus-extra-header function. Here’s a format line spec that will access the X-Newsreader header:
    "%~(form (gnus-extra-header 'X-Newsreader))@"
  3. The gnus-ignored-from-addresses variable says when the ‘%f’ summary line spec returns the To, Newsreader or From header. The variable may be a regexp or a predicate function. If this matches the contents of the From header, the value of the To or Newsreader headers are used instead.

    To distinguish regular articles from those where the From field has been swapped, a string is prefixed to the To or Newsgroups header in the summary line. By default the string is ‘-> ’ for To and ‘=> ’ for Newsgroups, you can customize these strings with gnus-summary-to-prefix and gnus-summary-newsgroup-prefix.

A related variable is nnmail-extra-headers, which controls when to include extra headers when generating overview (NOV) files. If you have old overview files, you should regenerate them after changing this variable, by entering the server buffer using ^, and then g on the appropriate mail server (e.g., nnml) to cause regeneration.

You also have to instruct Gnus to display the data by changing the %n spec to the %f spec in the gnus-summary-line-format variable.

In summary, you’d typically put something like the following in ~/.gnus.el:

(setq gnus-extra-headers
      '(To Newsgroups))
(setq nnmail-extra-headers gnus-extra-headers)
(setq gnus-summary-line-format
      "%U%R%z%I%(%[%4L: %-23,23f%]%) %s\n")
(setq gnus-ignored-from-addresses
      "Your Name Here")

(The values listed above are the default values in Gnus. Alter them to fit your needs.)

A note for news server administrators, or for users who wish to try to convince their news server administrator to provide some additional support:

The above is mostly useful for mail groups, where you have control over the NOV files that are created. However, if you can persuade your nntp admin to add (in the usual implementation, notably INN):


to the end of her overview.fmt file, then you can use that just as you would the extra headers from the mail groups. Otherwise, you have to disable fetching headers with ‘XOVER’:

(setq nntp-nov-is-evil t
      gnus-nov-is-evil t)

Be aware, though, that this will make entering an NNTP group much, much slower, so this is not recommended.

One particular scenario in which it can be desirable to not use ‘XOVER’ is for nnvirtual groups in order to support limiting by extra headers (e.g., by the newsgroup of its component groups). Because group parameters are not inherited, a separate select method for the component groups with the appropriate nov-is-evil set as a method variable is required.

4.1.3 Summary Buffer Mode Line

You can also change the format of the summary mode bar (see Mode Line Formatting). Set gnus-summary-mode-line-format to whatever you like. The default is ‘Gnus: %%b [%A] %Z’.

Here are the elements you can play with:


Group name.


Unprefixed group name.


Current article number.


Current article score.


Gnus version.


Number of unread articles in this group.


Number of unread articles in this group that aren’t displayed in the summary buffer.


A string with the number of unread and unselected articles represented either as ‘<%U(+%e) more>’ if there are both unread and unselected articles, and just as ‘<%U more>’ if there are just unread articles and no unselected ones.


Shortish group name. For instance, ‘rec.arts.anime’ will be shortened to ‘r.a.anime’.


Subject of the current article.


User-defined spec (see User-Defined Specs).


Name of the current score file (see Scoring).


Number of dormant articles (see Unread Articles).


Number of ticked articles (see Unread Articles).


Number of articles that have been marked as read in this session.


Number of articles expunged by the score files.

4.1.4 Summary Highlighting


This hook is run after selecting an article. It is meant to be used for highlighting the article in some way. It is not run if gnus-visual is nil.


This hook is called when a summary line is changed. It is not run if gnus-visual is nil.


This is the face (or font as some people call it) used to highlight the current article in the summary buffer.


Summary lines are highlighted according to this variable, which is a list where the elements are of the format (form . face). If you would, for instance, like ticked articles to be italic and high-scored articles to be bold, you could set this variable to something like

(((eq mark gnus-ticked-mark) . italic)
 ((> score default) . bold))

As you may have guessed, if form returns a non-nil value, face will be applied to the line.

4.2 Summary Maneuvering

All the straight movement commands understand the numeric prefix and behave pretty much as you’d expect.

None of these commands select articles.

G M-n

Go to the next summary line of an unread article (gnus-summary-next-unread-subject).

G M-p

Go to the previous summary line of an unread article (gnus-summary-prev-unread-subject).

G g

Ask for an article number and then go to the summary line of that article without displaying the article (gnus-summary-goto-subject).

If Gnus asks you to press a key to confirm going to the next group, you can use the C-n and C-p keys to move around the group buffer, searching for the next group to read without actually returning to the group buffer.

Variables related to summary movement:


If you issue one of the movement commands (like n) and there are no more unread articles after the current one, Gnus will offer to go to the next group. If this variable is t and the next group is empty, Gnus will exit summary mode and return to the group buffer. If this variable is neither t nor nil, Gnus will select the next group with unread articles. As a special case, if this variable is quietly, Gnus will select the next group without asking for confirmation. If this variable is almost-quietly, the same will happen only if you are located on the last article in the group. Finally, if this variable is slightly-quietly, the Z n command will go to the next group without confirmation. Also see Group Levels.


If non-nil, all the movement commands will try to go to the next article with the same subject as the current. (Same here might mean roughly equal. See gnus-summary-gather-subject-limit for details (see Customizing Threading).) If there are no more articles with the same subject, go to the first unread article.

This variable is not particularly useful if you use a threaded display.


Control whether to select the next/previous article when paging (with commands like SPC or DEL). If non-nil, select the next article when reaching the end of the article (or the previous article when paging backwards).

If nil, don’t do anything at the end/start of the articles.


If non-nil, all the “unread” movement commands will not proceed to the next (or previous) article if the current article is unread. Instead, they will choose the current article.


If non-nil, Gnus will keep the point in the summary buffer centered at all times. This makes things quite tidy, but if you have a slow network connection, or simply do not like this un-Emacsism, you can set this variable to nil to get the normal Emacs scrolling action. This will also inhibit horizontal re-centering of the summary buffer, which might make it more inconvenient to read extremely long threads.

This variable can also be a number. In that case, center the window at the given number of lines from the top.


If non-nil, don’t go to the next article when hitting SPC, and you’re at the end of the article.

4.3 Choosing Articles

4.3.1 Choosing Commands

None of the following movement commands understand the numeric prefix, and they all select and display an article.

If you want to fetch new articles or redisplay the group, see Exiting the Summary Buffer.


Select the current article, or, if that one’s read already, the next unread article (gnus-summary-next-page).

If you have an article window open already and you press SPC again, the article will be scrolled. This lets you conveniently SPC through an entire newsgroup. See Scrolling the Article.

G n

Go to next unread article (gnus-summary-next-unread-article).

G p

Go to previous unread article (gnus-summary-prev-unread-article).


Go to the next article (gnus-summary-next-article).


Go to the previous article (gnus-summary-prev-article).

G u

Go to the next unseen article (gnus-summary-next-unseen-article).


Go to the previous unseen article (gnus-summary-prev-unseen-article).

G C-n

Go to the next article with the same subject (gnus-summary-next-same-subject).

G C-p

Go to the previous article with the same subject (gnus-summary-prev-same-subject).

G f

Go to the first unread article (gnus-summary-first-unread-article).

G b

Go to the unread article with the highest score (gnus-summary-best-unread-article). If given a prefix argument, go to the first unread article that has a score over the default score.

G l

Go to the previous article read (gnus-summary-goto-last-article).

G o

Pop an article off the summary history and go to this article (gnus-summary-pop-article). This command differs from the command above in that you can pop as many previous articles off the history as you like, while l toggles the two last read articles. For a somewhat related issue (if you use these commands a lot), see Article Backlog.

G j

Ask for an article number or Message-ID, and then go to that article (gnus-summary-goto-article).

4.3.2 Choosing Variables

Some variables relevant for moving and selecting articles:


All the movement commands will try to go to the previous (or next) article, even if that article isn’t displayed in the Summary buffer if this variable is non-nil. Gnus will then fetch the article from the server and display it in the article buffer.


This hook is called whenever an article is selected. The default is nil. If you would like each article to be saved in the Agent as you read it, putting gnus-agent-fetch-selected-article on this hook will do so.


This hook is called whenever an article is selected. It is intended to be used for marking articles as read. The default value is gnus-summary-mark-read-and-unread-as-read, and will change the mark of almost any article you read to gnus-read-mark. The only articles not affected by this function are ticked, dormant, and expirable articles. If you’d instead like to just have unread articles marked as read, you can use gnus-summary-mark-unread-as-read instead. It will leave marks like gnus-low-score-mark, gnus-del-mark (and so on) alone.

4.4 Scrolling the Article


Pressing SPC will scroll the current article forward one page, or, if you have come to the end of the current article, will choose the next article (gnus-summary-next-page).

If gnus-article-skip-boring is non-nil and the rest of the article consists only of citations and signature, then it will be skipped; the next article will be shown instead. You can customize what is considered uninteresting with gnus-article-boring-faces. You can manually view the article’s pages, no matter how boring, using C-M-v.


Scroll the current article back one page (gnus-summary-prev-page).


Scroll the current article one line forward (gnus-summary-scroll-up).


Scroll the current article one line backward (gnus-summary-scroll-down).

A g

(Re)fetch the current article (gnus-summary-show-article). If given a prefix, show a completely “raw” article, just the way it came from the server. If given a prefix twice (i.e., C-u C-u g'), fetch the current article, but don’t run any of the article treatment functions.

If given a numerical prefix, you can do semi-manual charset stuff. C-u 0 g cn-gb-2312 RET will decode the message as if it were encoded in the cn-gb-2312 charset. If you have

(setq gnus-summary-show-article-charset-alist
      '((1 . cn-gb-2312)
        (2 . big5)))

then you can say C-u 1 g to get the same effect.

A <

Scroll to the beginning of the article (gnus-summary-beginning-of-article).

A >

Scroll to the end of the article (gnus-summary-end-of-article).

A s

Perform an isearch in the article buffer (gnus-summary-isearch-article).


Select the article buffer (gnus-summary-select-article-buffer).

4.5 Reply, Followup and Post

4.5.1 Summary Mail Commands

Commands for composing a mail message:

S r

Mail a reply to the author of the current article (gnus-summary-reply).


Mail a reply to the author of the current article and include the original message (gnus-summary-reply-with-original). This command uses the process/prefix convention.

S w

Mail a wide reply to the author of the current article (gnus-summary-wide-reply). A wide reply is a reply that goes out to all people listed in the To, From (or Reply-To) and Cc headers. If Mail-Followup-To is present, that’s used instead.


Mail a wide reply to the current article and include the original message (gnus-summary-wide-reply-with-original). This command uses the process/prefix convention, but only uses the headers from the first article to determine the recipients.


When replying to a message from a mailing list, send a reply to that message to the mailing list, and include the original message (gnus-summary-reply-to-list-with-original).

S v

Mail a very wide reply to the author of the current article (gnus-summary-wide-reply). A very wide reply is a reply that goes out to all people listed in the To, From (or Reply-To) and Cc headers in all the process/prefixed articles. This command uses the process/prefix convention.


Mail a very wide reply to the author of the current article and include the original message (gnus-summary-very-wide-reply-with-original). This command uses the process/prefix convention.

S B r

Mail a reply to the author of the current article but ignore the Reply-To field (gnus-summary-reply-broken-reply-to). If you need this because a mailing list incorrectly sets a Reply-To header pointing to the list, you probably want to set the broken-reply-to group parameter instead, so things will work correctly. See Group Parameters.


Mail a reply to the author of the current article and include the original message but ignore the Reply-To field (gnus-summary-reply-broken-reply-to-with-original).

S o m
C-c C-f

Forward the current article to some other person (gnus-summary-mail-forward). If no prefix is given, the message is forwarded according to the value of (message-forward-as-mime) and (message-forward-show-mml); if the prefix is 1, decode the message and forward directly inline; if the prefix is 2, forward message as an rfc822 MIME section; if the prefix is 3, decode message and forward as an rfc822 MIME section; if the prefix is 4, forward message directly inline; otherwise, the message is forwarded as no prefix given but use the flipped value of (message-forward-as-mime). By default, the forwarded message is inlined into the mail.

S m

Prepare a mail (gnus-summary-mail-other-window). By default, use the posting style of the current group. If given a prefix, disable that. If the prefix is 1, prompt for a group name to find the posting style.

S i

Prepare a news (gnus-summary-news-other-window). By default, post to the current group. If given a prefix, disable that. If the prefix is 1, prompt for a group to post to.

This function actually prepares a news even when using mail groups. This is useful for “posting” messages to mail groups without actually sending them over the network: they’re just saved directly to the group in question. The corresponding back end must have a request-post method for this to work though.

S D b

If you have sent a mail, but the mail was bounced back to you for some reason (wrong address, transient failure), you can use this command to resend that bounced mail (gnus-summary-resend-bounced-mail). You will be popped into a mail buffer where you can edit the headers before sending the mail off again. If you give a prefix to this command, and the bounced mail is a reply to some other mail, Gnus will try to fetch that mail and display it for easy perusal of its headers. This might very well fail, though.

S D r

Not to be confused with the previous command, gnus-summary-resend-message will prompt you for an address to send the current message off to, and then send it to that place. The headers of the message won’t be altered—but lots of headers that say Resent-To, Resent-From and so on will be added. This means that you actually send a mail to someone that has a To header that (probably) points to yourself. This will confuse people. So, natcherly you’ll only do that if you’re really eVIl.

This command is mainly used if you have several accounts and want to ship a mail to a different account of yours. (If you’re both root and postmaster and get a mail for postmaster to the root account, you may want to resend it to postmaster. Ordnung muss sein!

This command understands the process/prefix convention (see Process/Prefix).

S D e

Like the previous command, but will allow you to edit the message as if it were a new message before resending.

S O m

Digest the current series (see Decoding Articles) and forward the result using mail (gnus-uu-digest-mail-forward). This command uses the process/prefix convention (see Process/Prefix).

S M-c

Send a complaint about excessive crossposting to the author of the current article (gnus-summary-mail-crosspost-complaint).

This command is provided as a way to fight back against the current crossposting pandemic that’s sweeping Usenet. It will compose a reply using the gnus-crosspost-complaint variable as a preamble. This command understands the process/prefix convention (see Process/Prefix) and will prompt you before sending each mail.

Also See Header Commands in The Message Manual, for more information.

4.5.2 Summary Post Commands

Commands for posting a news article:

S p

Prepare for posting an article (gnus-summary-post-news). By default, post to the current group. If given a prefix, disable that. If the prefix is 1, prompt for another group instead.

S f

Post a followup to the current article (gnus-summary-followup).


Post a followup to the current article and include the original message (gnus-summary-followup-with-original). This command uses the process/prefix convention.

S n

Post a followup to the current article via news, even if you got the message through mail (gnus-summary-followup-to-mail).


Post a followup to the current article via news, even if you got the message through mail and include the original message (gnus-summary-followup-to-mail-with-original). This command uses the process/prefix convention.

S o p

Forward the current article to a newsgroup (gnus-summary-post-forward). If no prefix is given, the message is forwarded according to the value of (message-forward-as-mime) and (message-forward-show-mml); if the prefix is 1, decode the message and forward directly inline; if the prefix is 2, forward message as an rfc822 MIME section; if the prefix is 3, decode message and forward as an rfc822 MIME section; if the prefix is 4, forward message directly inline; otherwise, the message is forwarded as no prefix given but use the flipped value of (message-forward-as-mime). By default, the message is decoded and forwarded as an rfc822 MIME section.

S O p

Digest the current series and forward the result to a newsgroup (gnus-uu-digest-post-forward). This command uses the process/prefix convention.

S u

Uuencode a file, split it into parts, and post it as a series (gnus-uu-post-news). (see Uuencoding and Posting).

Also See Header Commands in The Message Manual, for more information.

4.5.3 Summary Message Commands

S y

Yank the current article into an already existing Message composition buffer (gnus-summary-yank-message). This command prompts for what message buffer you want to yank into, and understands the process/prefix convention (see Process/Prefix).


Attach the current article into an already existing Message composition buffer (gnus-summary-attach-message). If no such buffer exists, a new one is created. This command prompts for what message buffer you want to yank into, and understands the process/prefix convention (see Process/Prefix).

4.5.4 Canceling Articles

Have you ever written something, and then decided that you really, really, really wish you hadn’t posted that?

Well, you can’t cancel mail, but you can cancel posts.

Find the article you wish to cancel (you can only cancel your own articles, so don’t try any funny stuff). Then press C or S c (gnus-summary-cancel-article). Your article will be canceled—machines all over the world will be deleting your article. This command uses the process/prefix convention (see Process/Prefix).

Be aware, however, that not all sites honor cancels, so your article may live on here and there, while most sites will delete the article in question.

Gnus will use the “current” select method when canceling. If you want to use the standard posting method, use the ‘a’ symbolic prefix (see Symbolic Prefixes).

Gnus ensures that only you can cancel your own messages using a Cancel-Lock header (see Canceling News in Message Manual).

If you discover that you have made some mistakes and want to do some corrections, you can post a superseding article that will replace your original article.

Go to the original article and press S s (gnus-summary-supersede-article). You will be put in a buffer where you can edit the article all you want before sending it off the usual way.

The same goes for superseding as for canceling, only more so: Some sites do not honor superseding. On those sites, it will appear that you have posted almost the same article twice.

If you have just posted the article, and change your mind right away, there is a trick you can use to cancel/supersede the article without waiting for the article to appear on your site first. You simply return to the post buffer (which is called *sent ...*). There you will find the article you just posted, with all the headers intact. Change the Message-ID header to a Cancel or Supersedes header by substituting one of those words for the word Message-ID. Then just press C-c C-c to send the article as you would do normally. The previous article will be canceled/superseded.

Just remember, kids: There is no ’c’ in ’supersede’.

4.6 Delayed Articles

Sometimes, you might wish to delay the sending of a message. For example, you might wish to arrange for a message to turn up just in time to remind your about the birthday of your Significant Other. For this, there is the gnus-delay package. Setup is simple:


Normally, to send a message you use the C-c C-c command from Message mode. To delay a message, use C-c C-j (gnus-delay-article) instead. This will ask you for how long the message should be delayed. Possible answers are:

  • A time span. Consists of an integer and a letter. For example, 42d means to delay for 42 days. Available letters are m (minutes), h (hours), d (days), w (weeks), M (months) and Y (years).
  • A specific date. Looks like YYYY-MM-DD. The message will be delayed until that day, at a specific time (eight o’clock by default). See also gnus-delay-default-hour.
  • A specific time of day. Given in hh:mm format, 24h, no am/pm stuff. The deadline will be at that time today, except if that time has already passed, then it’s at the given time tomorrow. So if it’s ten o’clock in the morning and you specify 11:15, then the deadline is one hour and fifteen minutes hence. But if you specify 9:20, that means a time tomorrow.

The action of the gnus-delay-article command is influenced by a couple of variables:


When you specify a specific date, the message will be due on that hour on the given date. Possible values are integers 0 through 23.


This is a string and gives the default delay. It can be of any of the formats described above.


Delayed articles will be kept in this group on the drafts server until they are due. You probably don’t need to change this. The default value is "delayed".


The deadline for each article will be stored in a header. This variable is a string and gives the header name. You probably don’t need to change this. The default value is "X-Gnus-Delayed".

The way delaying works is like this: when you use the gnus-delay-article command, you give a certain delay. Gnus calculates the deadline of the message and stores it in the X-Gnus-Delayed header and puts the message in the nndraft:delayed group.

And whenever you get new news, Gnus looks through the group for articles which are due and sends them. It uses the gnus-delay-send-queue function for this. By default, this function is added to the hook gnus-get-new-news-hook. But of course, you can change this. Maybe you want to use the demon to send drafts? Just tell the demon to execute the gnus-delay-send-queue function.


By default, this function installs gnus-delay-send-queue in gnus-get-new-news-hook. But it accepts the optional second argument no-check. If it is non-nil, gnus-get-new-news-hook is not changed. The optional first argument is ignored.

For example, (gnus-delay-initialize nil t) means to do nothing. Presumably, you want to use the demon for sending due delayed articles. Just don’t forget to set that up :-)

When delaying an article with C-c C-j, Message mode will automatically add a "Date" header with the current time. In many cases you probably want the "Date" header to reflect the time the message is sent instead. To do this, you have to delete Date from message-draft-headers.

4.7 Marking Articles

There are several marks you can set on an article.

You have marks that decide the readedness (whoo, neato-keano neologism ohoy!) of the article. Alphabetic marks generally mean read, while non-alphabetic characters generally mean unread.

In addition, you also have marks that do not affect readedness.

4.7.1 Unread Articles

The following marks mark articles as (kinda) unread, in one form or other.


Marked as ticked (gnus-ticked-mark).

Ticked articles are articles that will remain visible always. If you see an article that you find interesting, or you want to put off reading it, or replying to it, until sometime later, you’d typically tick it. However, articles can be expired (from news servers by the news server software, Gnus itself never expires ticked messages), so if you want to keep an article forever, you’ll have to make it persistent (see Persistent Articles).


Marked as dormant (gnus-dormant-mark).

Dormant articles will only appear in the summary buffer if there are followups to it. If you want to see them even if they don’t have followups, you can use the / D command (see Limiting). Otherwise (except for the visibility issue), they are just like ticked messages.


Marked as unread (gnus-unread-mark).

Unread articles are articles that haven’t been read at all yet.

4.7.2 Read Articles

All the following marks mark articles as read.


These are articles that the user has marked as read with the d command manually, more or less (gnus-del-mark).


Articles that have actually been read (gnus-read-mark).


Articles that were marked as read in previous sessions and are now old (gnus-ancient-mark).


Marked as killed (gnus-killed-mark).


Marked as killed by kill files (gnus-kill-file-mark).


Marked as read by having too low a score (gnus-low-score-mark).


Marked as read by a catchup (gnus-catchup-mark).


Canceled article (gnus-canceled-mark)


Sparsely reffed article (gnus-sparse-mark). See Customizing Threading.


Article marked as read by duplicate suppression (gnus-duplicate-mark). See Duplicate Suppression.

All these marks just mean that the article is marked as read, really. They are interpreted differently when doing adaptive scoring, though.

One more special mark, though:


Marked as expirable (gnus-expirable-mark).

Marking articles as expirable (or have them marked as such automatically) doesn’t make much sense in normal groups—a user doesn’t control expiring of news articles, but in mail groups, for instance, articles marked as expirable can be deleted by Gnus at any time.

4.7.3 Other Marks

There are some marks that have nothing to do with whether the article is read or not.

  • You can set a bookmark in the current article. Say you are reading a long thesis on cats’ urinary tracts, and have to go home for dinner before you’ve finished reading the thesis. You can then set a bookmark in the article, and Gnus will jump to this bookmark the next time it encounters the article. See Setting Marks.
  • All articles that you have replied to or made a followup to (i.e., have answered) will be marked with an ‘A’ in the second column (gnus-replied-mark).
  • All articles that you have forwarded will be marked with an ‘F’ in the second column (gnus-forwarded-mark).
  • Articles stored in the article cache will be marked with an ‘*’ in the second column (gnus-cached-mark). See Article Caching.
  • Articles “saved” (in some manner or other; not necessarily religiously) are marked with an ‘S’ in the second column (gnus-saved-mark).
  • Articles that haven’t been seen before in Gnus by the user are marked with a ‘.’ in the second column (gnus-unseen-mark).
  • When using the Gnus agent (see Agent Basics), articles may be downloaded for unplugged (offline) viewing. If you are using the ‘%O’ spec, these articles get the ‘+’ mark in that spec. (The variable gnus-downloaded-mark controls which character to use.)
  • When using the Gnus agent (see Agent Basics), some articles might not have been downloaded. Such articles cannot be viewed while you are unplugged (offline). If you are using the ‘%O’ spec, these articles get the ‘-’ mark in that spec. (The variable gnus-undownloaded-mark controls which character to use.)
  • The Gnus agent (see Agent Basics) downloads some articles automatically, but it is also possible to explicitly mark articles for download, even if they would not be downloaded automatically. Such explicitly-marked articles get the ‘%’ mark in the first column. (The variable gnus-downloadable-mark controls which character to use.)
  • If the ‘%e’ spec is used, the presence of threads or not will be marked with gnus-not-empty-thread-mark and gnus-empty-thread-mark in the third column, respectively.
  • Finally we have the process mark (gnus-process-mark). A variety of commands react to the presence of the process mark. For instance, X u (gnus-uu-decode-uu) will uudecode and view all articles that have been marked with the process mark. Articles marked with the process mark have a ‘#’ in the second column.

You might have noticed that most of these “non-readedness” marks appear in the second column by default. So if you have a cached, saved, replied article that you have process-marked, what will that look like?

Nothing much. The precedence rules go as follows: process -> cache -> replied -> saved. So if the article is in the cache and is replied, you’ll only see the cache mark and not the replied mark.

4.7.4 Setting Marks

All the marking commands understand the numeric prefix.

M c

Clear all readedness-marks from the current article (gnus-summary-clear-mark-forward). In other words, mark the article as unread.

M t

Tick the current article (gnus-summary-tick-article-forward). See Article Caching.

M ?

Mark the current article as dormant (gnus-summary-mark-as-dormant). See Article Caching.

M d

Mark the current article as read (gnus-summary-mark-as-read-forward).


Mark the current article as read and move point to the previous line (gnus-summary-mark-as-read-backward).

M k

Mark all articles that have the same subject as the current one as read, and then select the next unread article (gnus-summary-kill-same-subject-and-select).


Mark all articles that have the same subject as the current one as read (gnus-summary-kill-same-subject).


Mark all unread articles as read (gnus-summary-catchup).

M C-c

Mark all articles in the group as read—even the ticked and dormant articles (gnus-summary-catchup-all).


Catchup the current group to point (before the point) (gnus-summary-catchup-to-here).

M h

Catchup the current group from point (after the point) (gnus-summary-catchup-from-here).


Mark all articles between point and mark as read (gnus-summary-mark-region-as-read).

M V k

Kill all articles with scores below the default score (or below the numeric prefix) (gnus-summary-kill-below).

M e

Mark the current article as expirable (gnus-summary-mark-as-expirable).

M b

Set a bookmark in the current article (gnus-summary-set-bookmark).


Remove the bookmark from the current article (gnus-summary-remove-bookmark).

M V c

Clear all marks from articles with scores over the default score (or over the numeric prefix) (gnus-summary-clear-above).

M V u

Tick all articles with scores over the default score (or over the numeric prefix) (gnus-summary-tick-above).

M V m

Prompt for a mark, and mark all articles with scores over the default score (or over the numeric prefix) with this mark (gnus-summary-clear-above).

The gnus-summary-goto-unread variable controls what action should be taken after setting a mark. If non-nil, point will move to the next/previous unread article. If nil, point will just move one line up or down. As a special case, if this variable is never, all the marking commands as well as other commands (like SPC) will move to the next article, whether it is unread or not. The default is t.

4.7.5 Generic Marking Commands

Some people would like the command that ticks an article (!) to go to the next article. Others would like it to go to the next unread article. Yet others would like it to stay on the current article. And even though I haven’t heard of anybody wanting it to go to the previous (unread) article, I’m sure there are people that want that as well.

Multiply these five behaviors with five different marking commands, and you get a potentially complex set of variable to control what each command should do.

To sidestep that mess, Gnus provides commands that do all these different things. They can be found on the M M map in the summary buffer. Type M M C-h to see them all—there are too many of them to list in this manual.

While you can use these commands directly, most users would prefer altering the summary mode keymap. For instance, if you would like the ! command to go to the next article instead of the next unread article, you could say something like:

(add-hook 'gnus-summary-mode-hook 'my-alter-summary-map)
(defun my-alter-summary-map ()
  (local-set-key "!" 'gnus-summary-put-mark-as-ticked-next))


(defun my-alter-summary-map ()
  (local-set-key "!" "MM!n"))

4.7.6 Setting Process Marks

Process marks are displayed as # in the summary buffer, and are used for marking articles in such a way that other commands will process these articles. For instance, if you process mark four articles and then use the * command, Gnus will enter these four articles into the cache. For more information, see Process/Prefix.

M P p

Toggle the process mark for the current article (gnus-summary-mark-as-processable).
If gnus-process-mark-toggle is nil, set the process mark for the current article.

M P u

Remove the process mark, if any, from the current article (gnus-summary-unmark-as-processable).


Remove the process mark from all articles (gnus-summary-unmark-all-processable).

M P i

Invert the list of process marked articles (gnus-uu-invert-processable).


Mark articles that have a Subject header that matches a regular expression (gnus-uu-mark-by-regexp).


Unmark articles that have a Subject header that matches a regular expression (gnus-uu-unmark-by-regexp).

M P r

Mark articles in region (gnus-uu-mark-region).

M P g

Unmark articles in region (gnus-uu-unmark-region).

M P t

Mark all articles in the current (sub)thread (gnus-uu-mark-thread).


Unmark all articles in the current (sub)thread (gnus-uu-unmark-thread).

M P v

Mark all articles that have a score above the prefix argument (gnus-uu-mark-over).

M P s

Mark all articles in the current series (gnus-uu-mark-series).


Mark all series that have already had some articles marked (gnus-uu-mark-sparse).

M P a

Mark all articles in series order (gnus-uu-mark-all).

M P b

Mark all articles in the buffer in the order they appear (gnus-uu-mark-buffer).

M P k

Push the current process mark set onto the stack and unmark all articles (gnus-summary-kill-process-mark).

M P y

Pop the previous process mark set from the stack and restore it (gnus-summary-yank-process-mark).

M P w

Push the current process mark set onto the stack (gnus-summary-save-process-mark).

Also see the & command in Searching for Articles, for how to set process marks based on article body contents.

4.8 Limiting

It can be convenient to limit the summary buffer to just show some subset of the articles currently in the group. The effect most limit commands have is to remove a few (or many) articles from the summary buffer.

Limiting commands work on subsets of the articles already fetched from the servers. These commands don’t query the server for additional articles.

/ /
/ s

Limit the summary buffer to articles that match some subject (gnus-summary-limit-to-subject). If given a prefix, exclude matching articles.

/ a

Limit the summary buffer to articles that match some author (gnus-summary-limit-to-author). If given a prefix, exclude matching articles.

/ R

Limit the summary buffer to articles that match some recipient (gnus-summary-limit-to-recipient). If given a prefix, exclude matching articles.

/ A

Limit the summary buffer to articles in which contents of From, To or Cc header match a given address (gnus-summary-limit-to-address). If given a prefix, exclude matching articles.

/ S

Limit the summary buffer to articles that aren’t part of any displayed threads (gnus-summary-limit-to-singletons). If given a prefix, limit to articles that are part of displayed threads.

/ x

Limit the summary buffer to articles that match one of the “extra” headers (see To From Newsgroups) (gnus-summary-limit-to-extra). If given a prefix, exclude matching articles.

/ u

Limit the summary buffer to articles not marked as read (gnus-summary-limit-to-unread). If given a prefix, limit the buffer to articles strictly unread. This means that ticked and dormant articles will also be excluded.

/ m

Ask for a mark and then limit to all articles that have been marked with that mark (gnus-summary-limit-to-marks).

/ t

Ask for a number and then limit the summary buffer to articles older than (or equal to) that number of days (gnus-summary-limit-to-age). If given a prefix, limit to articles younger than that number of days.

/ n

With prefix ‘n’, limit the summary buffer to the next ‘n’ articles. If not given a prefix, use the process marked articles instead. (gnus-summary-limit-to-articles).

/ w

Pop the previous limit off the stack and restore it (gnus-summary-pop-limit). If given a prefix, pop all limits off the stack.

/ .

Limit the summary buffer to the unseen articles (gnus-summary-limit-to-unseen).

/ v

Limit the summary buffer to articles that have a score at or above some score (gnus-summary-limit-to-score). If given a prefix, below some score.

/ p

Limit the summary buffer to articles that satisfy the display group parameter predicate (gnus-summary-limit-to-display-predicate). See Group Parameters, for more on this predicate.

/ r

Limit the summary buffer to replied articles (gnus-summary-limit-to-replied). If given a prefix, exclude replied articles.

/ E

Include all expunged articles in the limit (gnus-summary-limit-include-expunged).

/ D

Include all dormant articles in the limit (gnus-summary-limit-include-dormant).

/ *

Include all cached articles in the limit (gnus-summary-limit-include-cached).

/ d

Exclude all dormant articles from the limit (gnus-summary-limit-exclude-dormant).

/ M

Exclude all marked articles (gnus-summary-limit-exclude-marks).

/ T

Include all the articles in the current thread in the limit.

/ c

Exclude all dormant articles that have no children from the limit

/ C

Mark all excluded unread articles as read (gnus-summary-limit-mark-excluded-as-read). If given a prefix, also mark excluded ticked and dormant articles as read.

/ b

Limit the summary buffer to articles that have bodies that match a certain regexp (gnus-summary-limit-to-bodies). If given a prefix, reverse the limit. This command is quite slow since it requires selecting each article to find the matches.

/ h

Like the previous command, only limit to headers instead (gnus-summary-limit-to-headers).

The following commands aren’t limiting commands, but use the / prefix as well.

/ N

Insert all new articles in the summary buffer. It scans for new emails if back-end-get-new-mail is non-nil.

/ o

Insert all old articles in the summary buffer. If given a numbered prefix, fetch this number of articles.

4.9 Threading

Gnus threads articles by default. To thread is to put responses to articles directly after the articles they respond to—in a hierarchical fashion.

Threading is done by looking at the References headers of the articles. In a perfect world, this would be enough to build pretty trees, but unfortunately, the References header is often broken or simply missing. Weird news propagation exacerbates the problem, so one has to employ other heuristics to get pleasing results. A plethora of approaches exists, as detailed in horrible detail in Customizing Threading.

First, a quick overview of the concepts:


The top-most article in a thread; the first article in the thread.


A tree-like article structure.


A small(er) section of this tree-like structure.

loose threads

Threads often lose their roots due to article expiry, or due to the root already having been read in a previous session, and not displayed in the summary buffer. We then typically have many sub-threads that really belong to one thread, but are without connecting roots. These are called loose threads.

thread gathering

An attempt to gather loose threads into bigger threads.

sparse threads

A thread where the missing articles have been “guessed” at, and are displayed as empty lines in the summary buffer.

4.9.1 Customizing Threading Loose Threads


If non-nil, Gnus will gather all loose subtrees into one big tree and create a dummy root at the top. (Wait a minute. Root at the top? Yup.) Loose subtrees occur when the real root has expired, or you’ve read or killed the root in a previous session.

When there is no real root of a thread, Gnus will have to fudge something. This variable says what fudging method Gnus should use. There are four possible values:


Gnus will make the first of the orphaned articles the parent. This parent will adopt all the other articles. The adopted articles will be marked as such by pointy brackets (‘<>’) instead of the standard square brackets (‘[]’). This is the default method.


Gnus will create a dummy summary line that will pretend to be the parent. This dummy line does not correspond to any real article, so selecting it will just select the first real article after the dummy article. gnus-summary-dummy-line-format is used to specify the format of the dummy roots. It accepts only one format spec: ‘S’, which is the subject of the article. See Formatting Variables. If you want all threads to have a dummy root, even the non-gathered ones, set gnus-summary-make-false-root-always to t.


Gnus won’t actually make any article the parent, but simply leave the subject field of all orphans except the first empty. (Actually, it will use gnus-summary-same-subject as the subject (see Summary Buffer Format).)


Don’t make any article parent at all. Just gather the threads and display them after one another.


Don’t gather loose threads.


Loose threads are gathered by comparing subjects of articles. If this variable is nil, Gnus requires an exact match between the subjects of the loose threads before gathering them into one big super-thread. This might be too strict a requirement, what with the presence of stupid newsreaders that chop off long subject lines. If you think so, set this variable to, say, 20 to require that only the first 20 characters of the subjects have to match. If you set this variable to a really low number, you’ll find that Gnus will gather everything in sight into one thread, which isn’t very helpful.

If you set this variable to the special value fuzzy, Gnus will use a fuzzy string comparison algorithm on the subjects (see Fuzzy Matching).


This can either be a regular expression or list of regular expressions that match strings that will be removed from subjects if fuzzy subject simplification is used.


If you set gnus-summary-gather-subject-limit to something as low as 10, you might consider setting this variable to something sensible:

(setq gnus-simplify-ignored-prefixes
       (regexp-opt '("looking"
                     "wanted" "followup" "summary" "summary of"
                     "help" "query" "problem" "question"
                     "answer" "reference" "announce"
                     "How can I" "How to" "Comparison of"
                     ;; ...
       "\\)\\s *\\("
       (regexp-opt '("for" "for reference" "with" "about"))
       "\\)?\\]?:?[ \t]*"))

All words that match this regexp will be removed before comparing two subjects.


If non-nil, this variable overrides gnus-summary-gather-subject-limit. This variable should be a list of functions to apply to the Subject string iteratively to arrive at the simplified version of the string.

Useful functions to put in this list include:


Strip the leading ‘Re:’.


Simplify fuzzily.


Remove excessive whitespace.


Remove all whitespace.

You may also write your own functions, of course.


Since loose thread gathering is done on subjects only, that might lead to many false hits, especially with certain common subjects like ‘’ and ‘(none)’. To make the situation slightly better, you can use the regexp gnus-summary-gather-exclude-subject to say what subjects should be excluded from the gathering process.
The default is ‘^ *$\\|^(none)$’.


Gnus gathers threads by looking at Subject headers. This means that totally unrelated articles may end up in the same “thread”, which is confusing. An alternate approach is to look at all the Message-IDs in all the References headers to find matches. This will ensure that no gathered threads ever include unrelated articles, but it also means that people who have posted with broken newsreaders won’t be gathered properly. The choice is yours—plague or cholera:


This function is the default gathering function and looks at Subjects exclusively.


This function looks at References headers exclusively.

If you want to test gathering by References, you could say something like:

(setq gnus-summary-thread-gathering-function
      'gnus-gather-threads-by-references) Filling In Threads


If non-nil, Gnus will attempt to build old threads by fetching more old headers—headers to articles marked as read. If you would like to display as few summary lines as possible, but still connect as many loose threads as possible, you should set this variable to some or a number. If you set it to a number, no more than that number of extra old headers will be fetched. In either case, fetching old headers only works if the back end you are using carries overview files—this would normally be nntp, nnspool, nnml, and nnmaildir. Also remember that if the root of the thread has been expired by the server, there’s not much Gnus can do about that.

This variable can also be set to invisible. This won’t have any visible effects, but is useful if you use the A T command a lot (see Finding the Parent).

The server has to support NOV for any of this to work.

This feature can seriously impact performance it ignores all locally cached header entries. Setting it to t for groups for a server that doesn’t expire articles (such as, leads to very slow summary generation.


Same as gnus-fetch-old-headers, but only used for ephemeral newsgroups.


Fetching old headers can be slow. A low-rent similar effect can be gotten by setting this variable to some. Gnus will then look at the complete References headers of all articles and try to string together articles that belong in the same thread. This will leave gaps in the threading display where Gnus guesses that an article is missing from the thread. (These gaps appear like normal summary lines. If you select a gap, Gnus will try to fetch the article in question.) If this variable is t, Gnus will display all these “gaps” without regard for whether they are useful for completing the thread or not. Finally, if this variable is more, Gnus won’t cut off sparse leaf nodes that don’t lead anywhere. This variable is nil by default.


This is a rather obscure variable that few will find useful. It’s intended for those non-news newsgroups where the back end has to fetch quite a lot to present the summary buffer, and where it’s impossible to go back to parents of articles. This is mostly the case in the web-based groups.

If you don’t use those, then it’s safe to leave this as the default nil. If you want to use this variable, it should be a regexp that matches the group name, or t for all groups. More Threading


If this variable is nil, no threading will be done, and all of the rest of the variables here will have no effect. Turning threading off will speed group selection up a bit, but it is sure to make reading slower and more awkward.


If non-nil, all threads will be hidden when the summary buffer is generated.

This can also be a predicate specifier (see Predicate Specifiers). Available predicates are gnus-article-unread-p and gnus-article-unseen-p.

Here’s an example:

(setq gnus-thread-hide-subtree
      '(or gnus-article-unread-p

(It’s a pretty nonsensical example, since all unseen articles are also unread, but you get my drift.)


All threads that have a total score (as defined by gnus-thread-score-function) less than this number will be expunged. This variable is nil by default, which means that no threads are expunged.


if you kill a thread and this variable is non-nil, the subtree will be hidden.


Sometimes somebody changes the subject in the middle of a thread. If this variable is non-nil, which is the default, the subject change is ignored. If it is nil, a change in the subject will result in a new thread.


This is a number that says how much each sub-thread should be indented. The default is 4.


Sometimes, particularly with mailing lists, the order in which mails arrive locally is not necessarily the same as the order in which they arrived on the mailing list. Consequently, when sorting sub-threads using the default gnus-thread-sort-by-number, responses can end up appearing before the article to which they are responding to. Setting this variable to an alternate value (e.g., gnus-thread-sort-by-date), in a group’s parameters or in an appropriate hook (e.g., gnus-summary-generate-hook) can produce a more logical sub-thread ordering in such instances. Low-Level Threading


Hook run before parsing any headers.


If non-nil, this function will be called to allow alteration of article header structures. The function is called with one parameter, the article header vector, which it may alter in any way. For instance, if you have a mail-to-news gateway which alters the Message-IDs in systematic ways (by adding prefixes and such), you can use this variable to un-scramble the Message-IDs so that they are more meaningful. Here’s one example:

(setq gnus-alter-header-function 'my-alter-message-id)

(defun my-alter-message-id (header)
  (let ((id (mail-header-id header)))
    (when (string-match
           "\\(<[^<>@]*\\)\\.?cygnus\\..*@\\([^<>@]*>\\)" id)
       (concat (match-string 1 id) "@" (match-string 2 id))

4.9.2 Thread Commands

T k

Mark all articles in the current (sub-)thread as read (gnus-summary-kill-thread). If the prefix argument is positive, remove all marks instead. If the prefix argument is negative, tick articles instead.

T l

Lower the score of the current (sub-)thread (gnus-summary-lower-thread).

T i

Increase the score of the current (sub-)thread (gnus-summary-raise-thread).

T #

Set the process mark on the current (sub-)thread (gnus-uu-mark-thread).

T M-#

Remove the process mark from the current (sub-)thread (gnus-uu-unmark-thread).


Toggle threading (gnus-summary-toggle-threads).

T s

Expose the (sub-)thread hidden under the current article, if any

T h

Hide the current (sub-)thread (gnus-summary-hide-thread).


Expose all hidden threads (gnus-summary-show-all-threads).


Hide all threads (gnus-summary-hide-all-threads).

T t

Re-thread the current article’s thread (gnus-summary-rethread-current). This works even when the summary buffer is otherwise unthreaded.

T ^

Make the current article the child of the marked (or previous) article (gnus-summary-reparent-thread).

T M-^

Make the current article the parent of the marked articles (gnus-summary-reparent-children).

The following commands are thread movement commands. They all understand the numeric prefix.

T n

Go to the next thread (gnus-summary-next-thread).

T p

Go to the previous thread (gnus-summary-prev-thread).

T d

Descend the thread (gnus-summary-down-thread).

T u

Ascend the thread (gnus-summary-up-thread).

T o

Go to the top of the thread (gnus-summary-top-thread).

If you ignore subject while threading, you’ll naturally end up with threads that have several different subjects in them. If you then issue a command like T k (gnus-summary-kill-thread) you might not wish to kill the entire thread, but just those parts of the thread that have the same subject as the current article. If you like this idea, you can fiddle with gnus-thread-operation-ignore-subject. If it is non-nil (which it is by default), subjects will be ignored when doing thread commands. If this variable is nil, articles in the same thread with different subjects will not be included in the operation in question. If this variable is fuzzy, only articles that have subjects fuzzily equal will be included (see Fuzzy Matching).

4.10 Sorting the Summary Buffer

If you are using a threaded summary display, you can sort the threads by setting gnus-thread-sort-functions, which can be either a single function, a list of functions, or a list containing functions and (not some-function) elements.

By default, sorting is done on article numbers. Ready-made sorting predicate functions include gnus-thread-sort-by-number, gnus-thread-sort-by-author, gnus-thread-sort-by-recipient, gnus-thread-sort-by-subject, gnus-thread-sort-by-date, gnus-thread-sort-by-score, gnus-thread-sort-by-most-recent-number, gnus-thread-sort-by-most-recent-date, gnus-thread-sort-by-newsgroups and gnus-thread-sort-by-random and gnus-thread-sort-by-total-score.

Each function takes two threads and returns non-nil if the first thread should be sorted before the other. Note that sorting really is normally done by looking only at the roots of each thread. Exceptions to this rule are gnus-thread-sort-by-most-recent-number and gnus-thread-sort-by-most-recent-date.

If you use more than one function, the primary sort key should be the last function in the list. You should probably always include gnus-thread-sort-by-number in the list of sorting functions—preferably first. This will ensure that threads that are equal with respect to the other sort criteria will be displayed in ascending article order.

If you would like to sort by reverse score, then by subject, and finally by number, you could do something like:

(setq gnus-thread-sort-functions
        (not gnus-thread-sort-by-total-score)))

The threads that have highest score will be displayed first in the summary buffer. When threads have the same score, they will be sorted alphabetically. The threads that have the same score and the same subject will be sorted by number, which is (normally) the sequence in which the articles arrived.

If you want to sort by score and then reverse arrival order, you could say something like:

(setq gnus-thread-sort-functions
      '((not gnus-thread-sort-by-number)

By default, threads including their subthreads are sorted according to the value of gnus-thread-sort-functions. By customizing gnus-subthread-sort-functions you can define a custom sorting order for subthreads. This allows for example to sort threads from high score to low score in the summary buffer, but to have subthreads still sorted chronologically from old to new without taking their score into account.

The function in the gnus-thread-score-function variable (default +) is used for calculating the total score of a thread. Useful functions might be max, min, or squared means, or whatever tickles your fancy.

If you are using an unthreaded display for some strange reason or other, you have to fiddle with the gnus-article-sort-functions variable. It is very similar to the gnus-thread-sort-functions, except that it uses slightly different functions for article comparison. Available sorting predicate functions are gnus-article-sort-by-number, gnus-article-sort-by-author, gnus-article-sort-by-subject, gnus-article-sort-by-date, gnus-article-sort-by-newsgroups, gnus-article-sort-by-random, and gnus-article-sort-by-score.

If you want to sort an unthreaded summary display by subject, you could say something like:

(setq gnus-article-sort-functions

You can define group specific sorting via gnus-parameters, See Group Parameters.

4.11 Asynchronous Article Fetching

If you read your news from an NNTP server that’s far away, the network latencies may make reading articles a chore. You have to wait for a while after pressing n to go to the next article before the article appears. Why can’t Gnus just go ahead and fetch the article while you are reading the previous one? Why not, indeed.

First, some caveats. There are some pitfalls to using asynchronous article fetching, especially the way Gnus does it.

Let’s say you are reading article 1, which is short, and article 2 is quite long, and you are not interested in reading that. Gnus does not know this, so it goes ahead and fetches article 2. You decide to read article 3, but since Gnus is in the process of fetching article 2, the connection is blocked.

To avoid these situations, Gnus will open two (count ’em two) connections to the server. Some people may think this isn’t a very nice thing to do, but I don’t see any real alternatives. Setting up that extra connection takes some time, so Gnus startup will be slower.

Gnus will fetch more articles than you will read. This will mean that the link between your machine and the NNTP server will become more loaded than if you didn’t use article pre-fetch. The server itself will also become more loaded—both with the extra article requests, and the extra connection.

Ok, so now you know that you shouldn’t really use this thing… unless you really want to.

Here’s how: Set gnus-asynchronous to t. The rest should happen automatically.

You can control how many articles are to be pre-fetched by setting gnus-use-article-prefetch. This is 30 by default, which means that when you read an article in the group, the back end will pre-fetch the next 30 articles. If this variable is t, the back end will pre-fetch all the articles it can without bound. If it is nil, no pre-fetching will be done.

There are probably some articles that you don’t want to pre-fetch—read articles, for instance. The gnus-async-prefetch-article-p variable controls whether an article is to be pre-fetched. This function should return non-nil when the article in question is to be pre-fetched. The default is gnus-async-unread-p, which returns nil on read articles. The function is called with an article data structure as the only parameter.

If, for instance, you wish to pre-fetch only unread articles shorter than 100 lines, you could say something like:

(defun my-async-short-unread-p (data)
  "Return non-nil for short, unread articles."
  (and (gnus-data-unread-p data)
       (< (mail-header-lines (gnus-data-header data))

(setq gnus-async-prefetch-article-p 'my-async-short-unread-p)

These functions will be called many, many times, so they should preferably be short and sweet to avoid slowing down Gnus too much. It’s probably a good idea to byte-compile things like this.

After an article has been prefetched, this gnus-async-post-fetch-function will be called. The buffer will be narrowed to the region of the article that was fetched. A useful value would be gnus-html-prefetch-images, which will prefetch and store images referenced in the article, so that you don’t have to wait for them to be fetched when you read the article. This is useful for HTML messages that have external images.

Articles have to be removed from the asynch buffer sooner or later. The gnus-prefetched-article-deletion-strategy says when to remove articles. This is a list that may contain the following elements:


Remove articles when they are read.


Remove articles when exiting the group.

The default value is (read exit).

4.12 Article Caching

If you have an extremely slow NNTP connection, you may consider turning article caching on. Each article will then be stored locally under your home directory. As you may surmise, this could potentially use huge amounts of disk space, as well as eat up all your inodes so fast it will make your head swim. In vodka.

Used carefully, though, it could be just an easier way to save articles.

To turn caching on, set gnus-use-cache to t. By default, all articles ticked or marked as dormant will then be copied over to your local cache (gnus-cache-directory). Whether this cache is flat or hierarchical is controlled by the gnus-use-long-file-name variable, as usual.

When re-selecting a ticked or dormant article, it will be fetched from the cache instead of from the server. As articles in your cache will never expire, this might serve as a method of saving articles while still keeping them where they belong. Just mark all articles you want to save as dormant, and don’t worry.

When an article is marked as read, is it removed from the cache.

The entering/removal of articles from the cache is controlled by the gnus-cache-enter-articles and gnus-cache-remove-articles variables. Both are lists of symbols. The first is (ticked dormant) by default, meaning that ticked and dormant articles will be put in the cache. The latter is (read) by default, meaning that articles marked as read are removed from the cache. Possibly symbols in these two lists are ticked, dormant, unread and read.

So where does the massive article-fetching and storing come into the picture? The gnus-jog-cache command will go through all subscribed newsgroups, request all unread articles, score them, and store them in the cache. You should only ever, ever ever ever, use this command if 1) your connection to the NNTP server is really, really, really slow and 2) you have a really, really, really huge disk. Seriously. One way to cut down on the number of articles downloaded is to score unwanted articles down and have them marked as read. They will not then be downloaded by this command.

It is likely that you do not want caching on all groups. For instance, if your nnml mail is located under your home directory, it makes no sense to cache it somewhere else under your home directory. Unless you feel that it’s neat to use twice as much space.

To limit the caching, you could set gnus-cacheable-groups to a regexp of groups to cache, ‘^nntp’ for instance, or set the gnus-uncacheable-groups regexp to ‘^nnml’, for instance. Both variables are nil by default. If a group matches both variables, the group is not cached.

The cache stores information on what articles it contains in its active file (gnus-cache-active-file). If this file (or any other parts of the cache) becomes all messed up for some reason or other, Gnus offers two functions that will try to set things right. M-x gnus-cache-generate-nov-databases will (re)build all the NOV files, and gnus-cache-generate-active will (re)generate the active file.

gnus-cache-move-cache will move your whole gnus-cache-directory to some other location. You get asked to where, isn’t that cool?

4.13 Persistent Articles

Closely related to article caching, we have persistent articles. In fact, it’s just a different way of looking at caching, and much more useful in my opinion.

Say you’re reading a newsgroup, and you happen on to some valuable gem that you want to keep and treasure forever. You’d normally just save it (using one of the many saving commands) in some file. The problem with that is that it’s just, well, yucky. Ideally you’d prefer just having the article remain in the group where you found it forever; untouched by the expiry going on at the news server.

This is what a persistent article is—an article that just won’t be deleted. It’s implemented using the normal cache functions, but you use two explicit commands for managing persistent articles:


Make the current article persistent (gnus-cache-enter-article).


Remove the current article from the persistent articles (gnus-cache-remove-article). This will normally delete the article.

Both these commands understand the process/prefix convention.

To avoid having all ticked articles (and stuff) entered into the cache, you should set gnus-use-cache to passive if you’re just interested in persistent articles:

(setq gnus-use-cache 'passive)

4.14 Sticky Articles

When you select an article the current article buffer will be reused according to the value of the variable gnus-single-article-buffer. If its value is non-nil (the default) all articles reuse the same article buffer. Else each group has its own article buffer.

This implies that it’s not possible to have more than one article buffer in a group at a time. But sometimes you might want to display all the latest emails from your mother, your father, your aunt, your uncle and your 17 cousins to coordinate the next Christmas party.

That’s where sticky articles come in handy. A sticky article buffer basically is a normal article buffer, but it won’t be reused when you select another article. You can make an article sticky with:


Make the current article sticky. If a prefix arg is given, ask for a name for this sticky article buffer.

To close a sticky article buffer you can use these commands:


Puts this sticky article buffer at the end of the list of all buffers.


Kills this sticky article buffer.

To kill all sticky article buffers you can use:

Function: gnus-kill-sticky-article-buffers ARG

Kill all sticky article buffers. If a prefix ARG is given, ask for confirmation.

4.15 Article Backlog

If you have a slow connection, but the idea of using caching seems unappealing to you (and it is, really), you can help the situation some by switching on the backlog. This is where Gnus will buffer already read articles so that it doesn’t have to re-fetch articles you’ve already read. This only helps if you are in the habit of re-selecting articles you’ve recently read, of course. If you never do that, turning the backlog on will slow Gnus down a little bit, and increase memory usage some.

If you set gnus-keep-backlog to a number n, Gnus will store at most n old articles in a buffer for later re-fetching. If this variable is non-nil and is not a number, Gnus will store all read articles, which means that your Emacs will grow without bound before exploding and taking your machine down with you. I put that in there just to keep y’all on your toes.

The default value is 20.

4.16 Saving Articles

Gnus can save articles in a number of ways. Below is the documentation for saving articles in a fairly straight-forward fashion (i.e., little processing of the article is done before it is saved). For a different approach (uudecoding, unsharing) you should use gnus-uu (see Decoding Articles).

For the commands listed here, the target is a file. A directory name (ending in ‘/’) causes the target to be a file under that directory. If you want to save to a group, see the B c (gnus-summary-copy-article) command (see Mail Group Commands).

If gnus-save-all-headers is non-nil, Gnus will not delete unwanted headers before saving the article.

If the preceding variable is nil, all headers that match the gnus-saved-headers regexp will be kept, while the rest will be deleted before saving.

O o

Save the current article using the default article saver (gnus-summary-save-article).

O m

Save the current article in a Unix mail box (mbox) file (gnus-summary-save-article-mail).

O r

Save the current article in Rmail format (gnus-summary-save-article-rmail). This is mbox since Emacs 23, Babyl in older versions.

O f

Save the current article in plain file format (gnus-summary-save-article-file).


Write the current article in plain file format, overwriting any previous file contents (gnus-summary-write-article-file).

O b

Save the current article body in plain file format (gnus-summary-save-article-body-file).

O h

Save the current article in mh folder format (gnus-summary-save-article-folder).

O v

Save the current article in a VM folder (gnus-summary-save-article-vm).

O p

Save the current article in a pipe. Uhm, like, what I mean is—Pipe the current article to a process (gnus-summary-pipe-output). If given a symbolic prefix (see Symbolic Prefixes), include the complete headers in the piped output. The symbolic prefix r is special; it lets this command pipe a raw article including all headers. The gnus-summary-pipe-output-default-command variable can be set to a string containing the default command and options (default nil).


Save the current article into muttprint. That is, print it using the external program Muttprint. The program name and options to use is controlled by the variable gnus-summary-muttprint-program. (gnus-summary-muttprint).

All these commands use the process/prefix convention (see Process/Prefix). If you save bunches of articles using these functions, you might get tired of being prompted for files to save each and every article in. The prompting action is controlled by the gnus-prompt-before-saving variable, which is always by default, giving you that excessive prompting action you know and loathe. If you set this variable to t instead, you’ll be prompted just once for each series of articles you save. If you like to really have Gnus do all your thinking for you, you can even set this variable to nil, which means that you will never be prompted for files to save articles in. Gnus will simply save all the articles in the default files.

You can customize the gnus-default-article-saver variable to make Gnus do what you want it to. You can use any of the eight ready-made functions below, or you can create your own.


This is the default format, that used by the Rmail package. Since Emacs 23, Rmail uses standard mbox format. Before this, it used the Babyl format. Accordingly, this command writes mbox format since Emacs 23, unless appending to an existing Babyl file. In older versions of Emacs, it always uses Babyl format. Uses the function in the gnus-rmail-save-name variable to get a file name to save the article in. The default is gnus-plain-save-name.


Save in a Unix mail (mbox) file. Uses the function in the gnus-mail-save-name variable to get a file name to save the article in. The default is gnus-plain-save-name.


Append the article straight to an ordinary file. Uses the function in the gnus-file-save-name variable to get a file name to save the article in. The default is gnus-numeric-save-name.


Write the article straight to an ordinary file. The file is overwritten if it exists. Uses the function in the gnus-file-save-name variable to get a file name to save the article in. The default is gnus-numeric-save-name.


Append the article body to an ordinary file. Uses the function in the gnus-file-save-name variable to get a file name to save the article in. The default is gnus-numeric-save-name.


Write the article body straight to an ordinary file. The file is overwritten if it exists. Uses the function in the gnus-file-save-name variable to get a file name to save the article in. The default is gnus-numeric-save-name.


Save the article to an MH folder using rcvstore from the MH library. Uses the function in the gnus-folder-save-name variable to get a file name to save the article in. The default is gnus-folder-save-name, but you can also use gnus-Folder-save-name, which creates capitalized names.


Save the article in a VM folder. You have to have the VM mail reader to use this setting.


Pipe the article to a shell command. This function takes optional two arguments COMMAND and RAW. Valid values for COMMAND include:

  • a string
    The executable command name and possibly arguments.
  • nil
    You will be prompted for the command in the minibuffer.
  • the symbol default
    It will be replaced with the command which the variable gnus-summary-pipe-output-default-command holds or the command last used for saving.

Non-nil value for RAW overrides :decode and :headers properties (see below) and the raw article including all headers will be piped.

The symbol of each function may have the following properties:


The value non-nil means save decoded articles. This is meaningful only with gnus-summary-save-in-file, gnus-summary-save-body-in-file, gnus-summary-write-to-file, gnus-summary-write-body-to-file, and gnus-summary-save-in-pipe.


The value specifies an alternative function which appends, not overwrites, articles to a file. This implies that when saving many articles at a time, gnus-prompt-before-saving is bound to t and all articles are saved in a single file. This is meaningful only with gnus-summary-write-to-file and gnus-summary-write-body-to-file.


The value specifies the symbol of a variable of which the value specifies headers to be saved. If it is omitted, gnus-save-all-headers and gnus-saved-headers control what headers should be saved.

All of these functions, except for the last one, will save the article in the gnus-article-save-directory, which is initialized from the SAVEDIR environment variable. This is ~/News/ by default.

As you can see above, the functions use different functions to find a suitable name of a file to save the article in. Below is a list of available functions that generate names:


File names like ~/News/Alt.andrea-dworkin/45.


File names like ~/News/alt.andrea-dworkin/45.


File names like ~/News/Alt.andrea-dworkin.


File names like ~/News/alt.andrea-dworkin.


File names like ~/News/larsi.

You can have Gnus suggest where to save articles by plonking a regexp into the gnus-split-methods alist. For instance, if you would like to save articles related to Gnus in the file gnus-stuff, and articles related to VM in vm-stuff, you could set this variable to something like:

(("^Subject:.*gnus\\|^Newsgroups:.*gnus" "gnus-stuff")
 ("^Subject:.*vm\\|^Xref:.*vm" "vm-stuff")
 (my-choosing-function "../other-dir/my-stuff")
 ((equal gnus-newsgroup-name "mail.misc") "mail-stuff"))

We see that this is a list where each element is a list that has two elements—the match and the file. The match can either be a string (in which case it is used as a regexp to match on the article head); it can be a symbol (which will be called as a function with the group name as a parameter); or it can be a list (which will be evaled). If any of these actions have a non-nil result, the file will be used as a default prompt. In addition, the result of the operation itself will be used if the function or form called returns a string or a list of strings.

You basically end up with a list of file names that might be used when saving the current article. (All “matches” will be used.) You will then be prompted for what you really want to use as a name, with file name completion over the results from applying this variable.

This variable is ((gnus-article-archive-name)) by default, which means that Gnus will look at the articles it saves for an Archive-name line and use that as a suggestion for the file name.

Here’s an example function to clean up file names somewhat. If you have lots of mail groups called things like ‘nnml:mail.whatever’, you may want to chop off the beginning of these group names before creating the file name to save to. The following will do just that:

(defun my-save-name (group)
  (when (string-match "^nnml:mail." group)
    (substring group (match-end 0))))

(setq gnus-split-methods

Finally, you have the gnus-use-long-file-name variable. If it is nil, all the preceding functions will replace all periods (‘.’) in the group names with slashes (‘/’)—which means that the functions will generate hierarchies of directories instead of having all the files in the top level directory (~/News/alt/andrea-dworkin instead of ~/News/alt.andrea-dworkin.) This variable is t by default on most systems. However, for historical reasons, this is nil on Xenix and usg-unix-v machines by default.

This function also affects kill and score file names. If this variable is a list, and the list contains the element not-score, long file names will not be used for score files, if it contains the element not-save, long file names will not be used for saving, and if it contains the element not-kill, long file names will not be used for kill files.

If you’d like to save articles in a hierarchy that looks something like a spool, you could

(setq gnus-use-long-file-name '(not-save)) ; to get a hierarchy
(setq gnus-default-article-saver
      'gnus-summary-save-in-file)          ; no encoding

Then just save with o. You’d then read this hierarchy with ephemeral nneething groups—G D in the group buffer, and the top level directory as the argument (~/News/). Then just walk around to the groups/directories with nneething.

4.17 Decoding Articles

Sometime users post articles (or series of articles) that have been encoded in some way or other. Gnus can decode them for you.

All these functions use the process/prefix convention (see Process/Prefix) for finding out what articles to work on, with the extension that a “single article” means “a single series”. Gnus can find out by itself what articles belong to a series, decode all the articles and unpack/view/save the resulting file(s).

Gnus guesses what articles are in the series according to the following simplish rule: The subjects must be (nearly) identical, except for the last two numbers of the line. (Spaces are largely ignored, however.)

For example: If you choose a subject called ‘cat.gif (2/3)’, Gnus will find all the articles that match the regexp ‘^cat.gif ([0-9]+/[0-9]+).*$’.

Subjects that are non-standard, like ‘cat.gif (2/3) Part 6 of a series’, will not be properly recognized by any of the automatic viewing commands, and you have to mark the articles manually with #.

4.17.1 Uuencoded Articles

X u

Uudecodes the current series (gnus-uu-decode-uu).


Uudecodes and saves the current series (gnus-uu-decode-uu-and-save).

X v u

Uudecodes and views the current series (gnus-uu-decode-uu-view).

X v U

Uudecodes, views and saves the current series (gnus-uu-decode-uu-and-save-view).

Remember that these all react to the presence of articles marked with the process mark. If, for instance, you’d like to decode and save an entire newsgroup, you’d typically do M P a (gnus-uu-mark-all) and then X U (gnus-uu-decode-uu-and-save).

All this is very much different from how gnus-uu worked with GNUS 4.1, where you had explicit keystrokes for everything under the sun. This version of gnus-uu generally assumes that you mark articles in some way (see Setting Process Marks) and then press X u.

Note: When trying to decode articles that have names matching gnus-uu-notify-files, which is hard-coded to ‘[Cc][Ii][Nn][Dd][Yy][0-9]+.\\(gif\\|jpg\\)’, gnus-uu will automatically post an article on ‘comp.unix.wizards’ saying that you have just viewed the file in question. This feature can’t be turned off.

4.17.2 Shell Archives

Shell archives (“shar files”) used to be a popular way to distribute sources, but it isn’t used all that much today. In any case, we have some commands to deal with these:

X s

Unshars the current series (gnus-uu-decode-unshar).


Unshars and saves the current series (gnus-uu-decode-unshar-and-save).

X v s

Unshars and views the current series (gnus-uu-decode-unshar-view).

X v S

Unshars, views and saves the current series (gnus-uu-decode-unshar-and-save-view).

4.17.3 PostScript Files

X p

Unpack the current PostScript series (gnus-uu-decode-postscript).


Unpack and save the current PostScript series (gnus-uu-decode-postscript-and-save).

X v p

View the current PostScript series (gnus-uu-decode-postscript-view).

X v P

View and save the current PostScript series (gnus-uu-decode-postscript-and-save-view).

4.17.4 Other Files

X o

Save the current series (gnus-uu-decode-save).

X b

Unbinhex the current series (gnus-uu-decode-binhex). This doesn’t really work yet.


yEnc-decode the current series and save it (gnus-uu-decode-yenc).

4.17.5 Decoding Variables

Adjective, not verb. Rule Variables

Gnus uses rule variables to decide how to view a file. All these variables are of the form

      (list '(regexp1 command2)
            '(regexp2 command2)

This variable is consulted first when viewing files. If you wish to use, for instance, sox to convert an .au sound file, you could say something like:

(setq gnus-uu-user-view-rules
      (list '("\\\\.au$" "sox %s -t .aiff > /dev/audio")))

This variable is consulted if Gnus couldn’t make any matches from the user and default view rules.


This variable can be used to say what commands should be used to unpack archives. Other Decode Variables


All functions in this list will be called right after each file has been successfully decoded—so that you can move or view files right away, and don’t have to wait for all files to be decoded before you can do anything. Ready-made functions you can put in this list are:


View the file.


Move the file (if you’re using a saving function.)


Specifies what to do if unusual situations arise during decoding. If nil, be as conservative as possible. If t, ignore things that didn’t work, and overwrite existing files. Otherwise, ask each time.


Files with name matching this regular expression won’t be viewed.


Files with a MIME type matching this variable won’t be viewed. Note that Gnus tries to guess what type the file is based on the name. gnus-uu is not a MIME package (yet), so this is slightly kludgy.


Where gnus-uu does its work.


Non-nil means that gnus-uu won’t peek inside archives looking for files to display.


Non-nil means that the user will always be asked to save a file after viewing it.


Non-nil means that gnus-uu will ignore the default viewing rules.


Non-nil means that gnus-uu will ignore the default archive unpacking commands.


Non-nil means that gnus-uu will strip all carriage returns from articles.


Non-nil means that gnus-uu will mark unsuccessfully decoded articles as unread.


Non-nil means that gnus-uu will try to fix uuencoded files that have had trailing spaces deleted.


Hook run before sending a message to uudecode.


Non-nil means that gnus-uu, when asked to save without decoding, will save in digests. If this variable is nil, gnus-uu will just save everything in a file without any embellishments. The digesting almost conforms to RFC 1153—no easy way to specify any meaningful volume and issue numbers were found, so I simply dropped them. Uuencoding and Posting


Non-nil means that gnus-uu will ask for a file to encode before you compose the article. If this variable is t, you can either include an encoded file with C-c C-i or have one included for you when you post the article.


Maximum length of an article. The encoded file will be split into how many articles it takes to post the entire file.


Non-nil means that gnus-uu will post the encoded file in a thread. This may not be smart, as no other decoder I have seen is able to follow threads when collecting uuencoded articles. (Well, I have seen one package that does that—gnus-uu, but somehow, I don’t think that counts…) Default is nil.


Non-nil means that the description will be posted in a separate article. The first article will typically be numbered (0/x). If this variable is nil, the description the user enters will be included at the beginning of the first article, which will be numbered (1/x). Default is t.

4.17.6 Viewing Files

After decoding, if the file is some sort of archive, Gnus will attempt to unpack the archive and see if any of the files in the archive can be viewed. For instance, if you have a gzipped tar file pics.tar.gz containing the files pic1.jpg and pic2.gif, Gnus will uncompress and de-tar the main file, and then view the two pictures. This unpacking process is recursive, so if the archive contains archives of archives, it’ll all be unpacked.

Finally, Gnus will normally insert a pseudo-article for each extracted file into the summary buffer. If you go to these “articles”, you will be prompted for a command to run (usually Gnus will make a suggestion), and then the command will be run.

If gnus-view-pseudo-asynchronously is nil, Emacs will wait until the viewing is done before proceeding.

If gnus-view-pseudos is automatic, Gnus will not insert the pseudo-articles into the summary buffer, but view them immediately. If this variable is not-confirm, the user won’t even be asked for a confirmation before viewing is done.

If gnus-view-pseudos-separately is non-nil, one pseudo-article will be created for each file to be viewed. If nil, all files that use the same viewing command will be given as a list of parameters to that command.

If gnus-insert-pseudo-articles is non-nil, insert pseudo-articles when decoding. It is t by default.

So; there you are, reading your pseudo-articles in your virtual newsgroup from the virtual server; and you think: Why isn’t anything real anymore? How did we get here?

4.18 Article Treatment

Reading through this huge manual, you may have quite forgotten that the object of newsreaders is to actually, like, read what people have written. Reading articles. Unfortunately, people are quite bad at writing, so there are tons of functions and variables to make reading these articles easier.

4.18.1 Article Highlighting

Not only do you want your article buffer to look like fruit salad, but you want it to look like technicolor fruit salad.

W H a

Do much highlighting of the current article (gnus-article-highlight). This function highlights header, cited text, the signature, and adds buttons to the body and the head.

W H h

Highlight the headers (gnus-article-highlight-headers). The highlighting will be done according to the gnus-header-face-alist variable, which is a list where each element has the form (regexp name content). regexp is a regular expression for matching the header, name is the face used for highlighting the header name (see Faces and Fonts) and content is the face for highlighting the header value. The first match made will be used. Note that regexp shouldn’t have ‘^’ prepended—Gnus will add one.

W H c

Highlight cited text (gnus-article-highlight-citation).

Some variables to customize the citation highlights:


If the article size in bytes is bigger than this variable (which is 25000 by default), no citation highlighting will be performed.


Maximum possible length for a citation prefix (default 20).


List of faces used for highlighting citations (see Faces and Fonts). When there are citations from multiple articles in the same message, Gnus will try to give each citation from each article its own face. This should make it easier to see who wrote what.


Regexp matching normal Supercite attribution lines.


Regexp matching mangled Supercite attribution lines.


Minimum number of identical prefixes we have to see before we believe that it’s a citation.


Regexp matching the beginning of an attribution line.


Regexp matching the end of an attribution line.


Face used for attribution lines. It is merged with the face for the cited text belonging to the attribution.


If non-nil, no citation highlighting will be performed on lines beginning with ‘>From ’. Those lines may have been quoted by MTAs in order not to mix up with the envelope From line. The default value is t.

W H s

Highlight the signature (gnus-article-highlight-signature). Everything after gnus-signature-separator (see Article Signature) in an article will be considered a signature and will be highlighted with gnus-signature-face, which is italic by default.

See Customizing Articles, for how to highlight articles automatically.

4.18.2 Article Fontifying

People commonly add emphasis to words in news articles by writing things like ‘_this_’ or ‘*this*’ or ‘/this/’. Gnus can make this look nicer by running the article through the W e (gnus-article-emphasize) command.

How the emphasis is computed is controlled by the gnus-emphasis-alist variable. This is an alist where the first element is a regular expression to be matched. The second is a number that says what regular expression grouping is used to find the entire emphasized word. The third is a number that says what regexp grouping should be displayed and highlighted. (The text between these two groupings will be hidden.) The fourth is the face used for highlighting.

(setq gnus-emphasis-alist
      '(("_\\(\\w+\\)_" 0 1 gnus-emphasis-underline)
        ("\\*\\(\\w+\\)\\*" 0 1 gnus-emphasis-bold)))

By default, there are seven rules, and they use the following faces: gnus-emphasis-bold, gnus-emphasis-italic, gnus-emphasis-underline, gnus-emphasis-bold-italic, gnus-emphasis-underline-italic, gnus-emphasis-underline-bold, and gnus-emphasis-underline-bold-italic.

If you want to change these faces, you can either use M-x customize, or you can use copy-face. For instance, if you want to make gnus-emphasis-italic use a red face instead, you could say something like:

(copy-face 'red 'gnus-emphasis-italic)

If you want to highlight arbitrary words, you can use the gnus-group-highlight-words-alist variable, which uses the same syntax as gnus-emphasis-alist. The highlight-words group parameter (see Group Parameters) can also be used.

See Customizing Articles, for how to fontize articles automatically.

4.18.3 Article Hiding

Or rather, hiding certain things in each article. There usually is much too much cruft in most articles.

W W a

Do quite a lot of hiding on the article buffer (gnus-article-hide). In particular, this function will hide headers, PGP, cited text and the signature.

W W h

Hide headers (gnus-article-hide-headers). See Hiding Headers.

W W b

Hide headers that aren’t particularly interesting (gnus-article-hide-boring-headers). See Hiding Headers.

W W s

Hide signature (gnus-article-hide-signature). See Article Signature.

W W l

Strip list identifiers specified in gnus-list-identifiers. These are strings some mailing list servers add to the beginning of all Subject headers—for example, ‘[zebra 4711]’. Any leading ‘Re: ’ is skipped before stripping. gnus-list-identifiers may not contain \\(..\\).


A regular expression that matches list identifiers to be removed from subject. This can also be a list of regular expressions.


Hide PEM (privacy enhanced messages) cruft (gnus-article-hide-pem).


Strip the banner specified by the banner group parameter (gnus-article-strip-banner). This is mainly used to hide those annoying banners and/or signatures that some mailing lists and moderated groups adds to all the messages. The way to use this function is to add the banner group parameter (see Group Parameters) to the group you want banners stripped from. The parameter either be a string, which will be interpreted as a regular expression matching text to be removed, or the symbol signature, meaning that the (last) signature should be removed, or other symbol, meaning that the corresponding regular expression in gnus-article-banner-alist is used.

For instance:

(setq gnus-article-banner-alist
      ((googleGroups .

Regardless of a group, you can hide things like advertisements only when the sender of an article has a certain mail address specified in gnus-article-address-banner-alist.


Alist of mail addresses and banners. Each element has the form (address . banner), where address is a regexp matching a mail address in the From header, banner is one of a symbol signature, an item in gnus-article-banner-alist, a regexp and nil. If address matches author’s mail address, it will remove things like advertisements. For example, if a sender has the mail address ‘’ and there is a banner something like ‘Do You Yoo-hoo!?’ in all articles he sends, you can use the following element to remove them:

("@yoo-hoo\\.co\\.jp\\'" .
 "\n_+\nDo You Yoo-hoo!\\?\n.*\n.*\n")
W W c

Hide citation (gnus-article-hide-citation). Some variables for customizing the hiding:


Gnus adds buttons to show where the cited text has been hidden, and to allow toggle hiding the text. The format of the variable is specified by these format-like variable (see Formatting Variables). These specs are valid:


Starting point of the hidden text.


Ending point of the hidden text.


Number of characters in the hidden region.


Number of lines of hidden text.


The number of lines at the beginning of the cited text to leave shown. This can also be a cons cell with the number of lines at the top and bottom of the text, respectively, to remain visible.

W W C-c

Hide citation (gnus-article-hide-citation-maybe) depending on the following two variables:


If the cited text is of a bigger percentage than this variable (default 50), hide the cited text.


The cited text must have at least this length (default 10) before it is hidden.


Hide cited text in articles that aren’t roots (gnus-article-hide-citation-in-followups). This isn’t very useful as an interactive command, but might be a handy function to stick have happen automatically (see Customizing Articles).

All these “hiding” commands are toggles, but if you give a negative prefix to these commands, they will show what they have previously hidden. If you give a positive prefix, they will always hide.

Also see Article Highlighting for further variables for citation customization.

See Customizing Articles, for how to hide article elements automatically.

4.18.4 Article Washing

We call this “article washing” for a really good reason. Namely, the A key was taken, so we had to use the W key instead.

Washing is defined by us as “changing something from something to something else”, but normally results in something looking better. Cleaner, perhaps.

See Customizing Articles, if you want to change how Gnus displays articles by default.

C-u g

This is not really washing, it’s sort of the opposite of washing. If you type this, you see the article exactly as it exists on disk or on the server.


Force redisplaying of the current article (gnus-summary-show-article). This is also not really washing. If you type this, you see the article without any previously applied interactive Washing functions but with all default treatments (see Customizing Articles).

W l

Remove page breaks from the current article (gnus-summary-stop-page-breaking). See Misc Article, for page delimiters.

W r

Do a Caesar rotate (rot13) on the article buffer (gnus-summary-caesar-message). Unreadable articles that tell you to read them with Caesar rotate or rot13. (Typically offensive jokes and such.)

It’s commonly called “rot13” because each letter is rotated 13 positions in the alphabet, e.g., ‘B’ (letter #2) -> ‘O’ (letter #15). It is sometimes referred to as “Caesar rotate” because Caesar is rumored to have employed this form of, uh, somewhat weak encryption.

W m

Morse decode the article buffer (gnus-summary-morse-message).

W i

Decode IDNA encoded domain names in the current articles. IDNA encoded domain names looks like ‘xn--bar’. If a string remain unencoded after running invoking this, it is likely an invalid IDNA string (‘xn--bar’ is invalid). You must have GNU Libidn ( installed for this command to work.

W t

Toggle whether to display all headers in the article buffer (gnus-summary-toggle-header).

W v

Toggle whether to display all headers in the article buffer permanently (gnus-summary-verbose-headers).

W o

Treat overstrike (gnus-article-treat-overstrike).

W d

Treat “Microsoft smartquotes” according to gnus-article-smartquotes-map (gnus-article-treat-smartquotes). Note that this function guesses whether a character is a smartquote or not, so it should only be used interactively.

Smartquotes are Microsoft’s unilateral extension to the character map in an attempt to provide more quoting characters. If you see something like \222 or \264 where you’re expecting some kind of apostrophe or quotation mark, then try this wash.


Translate many non-ASCII characters into their ASCII equivalents (gnus-article-treat-non-ascii). This is mostly useful if you’re on a terminal that has a limited font and doesn’t show accented characters, “advanced” punctuation, and the like. For instance, ‘»’ is translated into ‘>>’, and so on.

W Y f

Full deuglify of broken Outlook (Express) articles: Treat \"smartquotes\", unwrap lines, repair attribution and rearrange citation (gnus-article-outlook-deuglify-article).

W Y u

Unwrap lines that appear to be wrapped citation lines. You can control what lines will be unwrapped by frobbing gnus-outlook-deuglify-unwrap-min and gnus-outlook-deuglify-unwrap-max, indicating the minimum and maximum length of an unwrapped citation line. (gnus-article-outlook-unwrap-lines).

W Y a

Repair a broken attribution line.

W Y c

Repair broken citations by rearranging the text. (gnus-article-outlook-rearrange-citation).

W w

Do word wrap (gnus-article-fill-cited-article).

You can give the command a numerical prefix to specify the width to use when filling.


Fill long lines (gnus-article-fill-long-lines).

You can give the command a numerical prefix to specify the width to use when filling.


Capitalize the first word in each sentence (gnus-article-capitalize-sentences).

W c

Translate CRLF pairs (i.e., ‘^M’s on the end of the lines) into LF (this takes care of DOS line endings), and then translate any remaining CRs into LF (this takes care of Mac line endings) (gnus-article-remove-cr).

W q

Treat quoted-printable (gnus-article-de-quoted-unreadable). Quoted-Printable is one common MIME encoding employed when sending non-ASCII (i.e., 8-bit) articles. It typically makes strings like ‘déjà vu’ look like ‘d=E9j=E0 vu’, which doesn’t look very readable to me. Note that this is usually done automatically by Gnus if the message in question has a Content-Transfer-Encoding header that says that this encoding has been done. If a prefix is given, a charset will be asked for.

W 6

Treat base64 (gnus-article-de-base64-unreadable). Base64 is one common MIME encoding employed when sending non-ASCII (i.e., 8-bit) articles. Note that this is usually done automatically by Gnus if the message in question has a Content-Transfer-Encoding header that says that this encoding has been done. If a prefix is given, a charset will be asked for.


Treat HZ or HZP (gnus-article-decode-HZ). HZ (or HZP) is one common encoding employed when sending Chinese articles. It typically makes strings look like ‘~{<:Ky2;S{#,NpJ)l6HK!#~}’.


Translate ANSI SGR control sequences into overlays or extents (gnus-article-treat-ansi-sequences). ANSI sequences are used in some Chinese hierarchies for highlighting.

W u

Remove newlines from within URLs. Some mailers insert newlines into outgoing email messages to keep lines short. This reformatting can split long URLs onto multiple lines. Repair those URLs by removing the newlines (gnus-article-unsplit-urls).

W h

Treat HTML (gnus-article-wash-html). Note that this is usually done automatically by Gnus if the message in question has a Content-Type header that says that the message is HTML.

If a prefix is given, a charset will be asked for. If it is a number, the charset defined in gnus-summary-show-article-charset-alist (see Scrolling the Article) will be used.

The default is to use the function specified by mm-text-html-renderer (see Display Customization in The Emacs MIME Manual) to convert the HTML. Pre-defined functions you can use include:


Use Gnus simple html renderer.


Use Gnus rendered based on w3m.


Use emacs-w3m.


Use w3m.


Use Links.


Use Lynx.


Toggle proportional fonts for HTML articles. This temporarily changes the shr-use-fonts variable in the current article buffer.

W b

Add clickable buttons to the article (gnus-article-add-buttons). See Article Buttons.


Add clickable buttons to the article headers (gnus-article-add-buttons-to-head).

W p

Verify a signed control message (gnus-article-verify-x-pgp-sig). Control messages such as newgroup and checkgroups are usually signed by the hierarchy maintainer. You need to add the PGP public key of the maintainer to your keyring to verify the message.1

W s

Verify a signed (PGP, PGP/MIME or S/MIME) message (gnus-summary-force-verify-and-decrypt). See Security.

W a

Strip headers like the X-No-Archive header from the beginning of article bodies (gnus-article-strip-headers-in-body).

W E l

Remove all blank lines from the beginning of the article (gnus-article-strip-leading-blank-lines).

W E m

Replace all blank lines with empty lines and then all multiple empty lines with a single empty line. (gnus-article-strip-multiple-blank-lines).

W E t

Remove all blank lines at the end of the article (gnus-article-remove-trailing-blank-lines).

W E a

Do all the three commands above (gnus-article-strip-blank-lines).


Remove all blank lines (gnus-article-strip-all-blank-lines).

W E s

Remove all white space from the beginning of all lines of the article body (gnus-article-strip-leading-space).

W E e

Remove all white space from the end of all lines of the article body (gnus-article-strip-trailing-space).

See Customizing Articles, for how to wash articles automatically.

4.18.5 Article Header

These commands perform various transformations of article header.

W G u

Unfold folded header lines (gnus-article-treat-unfold-headers).

W G n

Fold the Newsgroups and Followup-To headers (gnus-article-treat-fold-newsgroups).

W G f

Fold all the message headers (gnus-article-treat-fold-headers).

W E w

Remove excessive whitespace from all headers (gnus-article-remove-leading-whitespace).

4.18.6 Article Buttons

People often include references to other stuff in articles, and it would be nice if Gnus could just fetch whatever it is that people talk about with the minimum of fuzz when you hit RET or use the middle mouse button on these references.

Gnus adds buttons to certain standard references by default: Well-formed URLs, mail addresses, Message-IDs, Info links, man pages and Emacs or Gnus related references. This is controlled by two variables, one that handles article bodies and one that handles article heads:


This is an alist where each entry has this form:

(regexp button-par use-p function data-par)

All text that match this regular expression (case insensitive) will be considered an external reference. Here’s a typical regexp that matches embedded URLs: ‘<URL:\\([^\n\r>]*\\)>’. This can also be a variable containing a regexp, useful variables to use include gnus-button-url-regexp and gnus-button-mid-or-mail-regexp.


Gnus has to know which parts of the matches is to be highlighted. This is a number that says what sub-expression of the regexp is to be highlighted. If you want it all highlighted, you use 0 here.


This form will be evaled, and if the result is non-nil, this is considered a match. This is useful if you want extra sifting to avoid false matches. Often variables named gnus-button-*-level are used here, See Article button levels, but any other form may be used too.


This function will be called when you click on this button.


As with button-par, this is a sub-expression number, but this one says which part of the match is to be sent as data to function.

So the full entry for buttonizing URLs is then

("<URL:\\([^\n\r>]*\\)>" 0 t gnus-button-url 1)

This is just like the other alist, except that it is applied to the article head only, and that each entry has an additional element that is used to say what headers to apply the buttonize coding to:

(header regexp button-par use-p function data-par)

header is a regular expression.

4.18.7 Article button levels

The higher the value of the variables gnus-button-*-level, the more buttons will appear. If the level is zero, no corresponding buttons are displayed. With the default value (which is 5) you should already see quite a lot of buttons. With higher levels, you will see more buttons, but you may also get more false positives. To avoid them, you can set the variables gnus-button-*-level local to specific groups (see Group Parameters). Here’s an example for the variable gnus-parameters:

;; increase gnus-button-*-level in some groups:
(setq gnus-parameters
      '(("\\<\\(emacs\\|gnus\\)\\>" (gnus-button-emacs-level 10))
        ("\\<unix\\>"               (gnus-button-man-level 10))
        ("\\<tex\\>"                (gnus-button-tex-level 10))))

Controls the display of references to message IDs, mail addresses and news URLs. Related variables and functions include gnus-button-url-regexp, browse-url, and browse-url-browser-function.


Controls the display of Emacs or Gnus references. Related functions are gnus-button-handle-custom, gnus-button-handle-describe-function, gnus-button-handle-describe-variable, gnus-button-handle-symbol, gnus-button-handle-describe-key, gnus-button-handle-apropos, gnus-button-handle-apropos-command, gnus-button-handle-apropos-variable, gnus-button-handle-apropos-documentation, and gnus-button-handle-library.


Controls the display of references to (Unix) man pages. See gnus-button-man-handler.


Controls the display of message IDs, mail addresses and news URLs. Related variables and functions include gnus-button-mid-or-mail-regexp, gnus-button-prefer-mid-or-mail, gnus-button-mid-or-mail-heuristic, and gnus-button-mid-or-mail-heuristic-alist.

4.18.8 Article Date

The date is most likely generated in some obscure timezone you’ve never heard of, so it’s quite nice to be able to find out what the time was when the article was sent.

W T u

Display the date in UT (aka. GMT, aka ZULU) (gnus-article-date-ut).

W T i

Display the date in international format, aka. ISO 8601 (gnus-article-date-iso8601).

W T l

Display the date in the local timezone (gnus-article-date-local).

W T p

Display the date in a format that’s easily pronounceable in English (gnus-article-date-english).

W T s

Display the date using a user-defined format (gnus-article-date-user). The format is specified by the gnus-article-time-format variable, and is a string that’s passed to format-time-string. See the documentation of that variable for a list of possible format specs.

W T e

Say how much time has elapsed between the article was posted and now (gnus-article-date-lapsed). It looks something like:

Date: 6 weeks, 4 days, 1 hour, 3 minutes, 8 seconds ago

To make this line updated continually, set the gnus-article-update-date-headers variable to the frequency in seconds (the default is nil).

W T o

Display the original date (gnus-article-date-original). This can be useful if you normally use some other conversion function and are worried that it might be doing something totally wrong. Say, claiming that the article was posted in 1854. Although something like that is totally impossible. Don’t you trust me? *titter*

See Customizing Articles, for how to display the date in your preferred format automatically.

4.18.9 Article Display

These commands add various frivolous display gimmicks to the article buffer in Emacs versions that support them.

X-Face headers are small black-and-white images supplied by the message headers (see X-Face).

Face headers are small colored images supplied by the message headers (see Face).

Smileys are those little ‘:-)’ symbols that people like to litter their messages with (see Smileys).

Picons, on the other hand, reside on your own system, and Gnus will try to match the headers to what you have (see Picons).

Gravatars reside on-line and are fetched from (see Gravatars).

All these functions are toggles—if the elements already exist, they’ll be removed.

W D x

Display an X-Face in the From header. (gnus-article-display-x-face).

W D d

Display a Face in the From header. (gnus-article-display-face).

W D s

Display smileys (gnus-treat-smiley).

W D f

Piconify the From header (gnus-treat-from-picon).

W D m

Piconify all mail headers (i.e., Cc, To) (gnus-treat-mail-picon).

W D n

Piconify all news headers (i.e., Newsgroups and Followup-To) (gnus-treat-newsgroups-picon).

W D g

Gravatarify the From header (gnus-treat-from-gravatar).

W D h

Gravatarify all mail headers (i.e., Cc, To) (gnus-treat-from-gravatar).

W D e

Some symbols have both a non-emoji presentation and an emoji presentation. This command will make Gnus choose the emoji presentation (gnus-article-emojize-symbols).


Remove all images from the article buffer (gnus-article-remove-images).


If you’re reading an HTML article rendered with gnus-article-html, then you can insert any blocked images in the buffer with this command. (gnus-html-show-images).

4.18.10 Article Signature

Each article is divided into two parts—the head and the body. The body can be divided into a signature part and a text part. The variable that says what is to be considered a signature is gnus-signature-separator. This is normally the standard ‘^-- $’ as mandated by RFC 5536. However, many people use non-standard signature separators, so this variable can also be a list of regular expressions to be tested, one by one. (Searches are done from the end of the body towards the beginning.) One likely value is:

(setq gnus-signature-separator
      '("^-- $"         ; The standard
        "^-- *$"        ; A common mangling
        "^-------*$"    ; Many people just use a looong
                        ; line of dashes.  Shame!
        "^ *--------*$" ; Double-shame!
        "^________*$"   ; Underscores are also popular
        "^========*$")) ; Pervert!

The more permissive you are, the more likely it is that you’ll get false positives.

gnus-signature-limit provides a limit to what is considered a signature when displaying articles.

  1. If it is an integer, no signature may be longer (in characters) than that integer.
  2. If it is a floating point number, no signature may be longer (in lines) than that number.
  3. If it is a function, the function will be called without any parameters, and if it returns nil, there is no signature in the buffer.
  4. If it is a string, it will be used as a regexp. If it matches, the text in question is not a signature.

This variable can also be a list where the elements may be of the types listed above. Here’s an example:

(setq gnus-signature-limit
      '(200.0 "^---*Forwarded article"))

This means that if there are more than 200 lines after the signature separator, or the text after the signature separator is matched by the regular expression ‘^---*Forwarded article’, then it isn’t a signature after all.

4.18.11 Article Miscellanea

A t

Translate the article from one language to another (gnus-article-babel).

4.19 MIME Commands

The following commands all understand the numerical prefix. For instance, 3 K v means “view the third MIME part”.

K v

View the MIME part.

K o

Save the MIME part.


Prompt for a file name, then save the MIME part and strip it from the article. The stripped MIME object will be referred via the message/external-body MIME type.

K r

Replace the MIME part with an external body.

K d

Delete the MIME part and add some information about the removed part.

K c

Copy the MIME part.

K e

View the MIME part externally.

K i

View the MIME part internally.

K |

Pipe the MIME part to an external command.

The rest of these MIME commands do not use the numerical prefix in the same manner:


View ‘text/html’ parts of the current article with a WWW browser. Inline images embedded in a message using the cid scheme, as they are generally considered to be safe, will be processed properly. The message header is added to the beginning of every HTML part unless the prefix argument is given.

Warning: Spammers use links to images (using the http scheme) in HTML articles to verify whether you have read the message. As this command passes the HTML content to the browser without eliminating these “web bugs” you should only use it for mails from trusted senders.

This command creates temporary files to pass HTML contents including images if any to the browser, and deletes them when exiting the group (if you want).

K b

Make all the MIME parts have buttons in front of them. This is mostly useful if you wish to save (or perform other actions) on inlined parts.

W M h

Display MIME part buttons in the end of the header of an article (gnus-mime-buttonize-attachments-in-header). This command toggles the display. Note that buttons to be added to the header are only the ones that aren’t inlined in the body. If you want those buttons always to be displayed, set gnus-mime-display-attachment-buttons-in-header to non-nil. The default is t. To change the appearance of buttons, customize gnus-header-face-alist.

K m

Some multipart messages are transmitted with missing or faulty headers. This command will attempt to “repair” these messages so that they can be viewed in a more pleasant manner (gnus-summary-repair-multipart).

X m

Save all parts matching a MIME type to a directory (gnus-summary-save-parts). Understands the process/prefix convention (see Process/Prefix).


Toggle the buttonized display of the article buffer (gnus-summary-toggle-display-buttonized).

W M w

Decode RFC 2047-encoded words in the article headers (gnus-article-decode-mime-words).

W M c

Decode encoded article bodies as well as charsets (gnus-article-decode-charset).

This command looks in the Content-Type header to determine the charset. If there is no such header in the article, you can give it a prefix, which will prompt for the charset to decode as. In regional groups where people post using some common encoding (but do not include MIME headers), you can set the charset group/topic parameter to the required charset (see Group Parameters).

W M v

View all the MIME parts in the current article (gnus-mime-view-all-parts).

Relevant variables:


This is a list of regexps. MIME types that match a regexp from this list will be completely ignored by Gnus. The default value is nil.

To have all Vcards be ignored, you’d say something like this:

(setq gnus-ignored-mime-types

If non-nil, Gnus won’t require the ‘MIME-Version’ header before interpreting the message as a MIME message. This helps when reading messages from certain broken mail user agents. The default is t.


There are other, non-MIME encoding methods used. The most common is ‘uuencode’, but yEncode is also getting to be popular. If this variable is non-nil, Gnus will look in message bodies to see if it finds these encodings, and if so, it’ll run them through the Gnus MIME machinery. The default is t. Only single-part yEnc encoded attachments can be decoded. There’s no support for encoding in Gnus.


This is a list of regexps. MIME types that match a regexp from this list won’t have MIME buttons inserted unless they aren’t displayed or this variable is overridden by gnus-buttonized-mime-types. The default value is (".*/.*"). This variable is only used when gnus-inhibit-mime-unbuttonizing is nil.


This is a list of regexps. MIME types that match a regexp from this list will have MIME buttons inserted unless they aren’t displayed. This variable overrides gnus-unbuttonized-mime-types. The default value is nil. This variable is only used when gnus-inhibit-mime-unbuttonizing is nil.

E.g., to see security buttons but no other buttons, you could set this variable to ("multipart/signed") and leave gnus-unbuttonized-mime-types at the default value.

You could also add "multipart/alternative" to this list to display radio buttons that allow you to choose one of two media types those mails include. See also mm-discouraged-alternatives (see Display Customization in The Emacs MIME Manual).


If this is non-nil, then all MIME parts get buttons. The default value is nil.


For each MIME part, this function will be called with the MIME handle as the parameter. The function is meant to be used to allow users to gather information from the article (e.g., add Vcard info to the bbdb database) or to do actions based on parts (e.g., automatically save all jpegs into some directory).

Here’s an example function the does the latter:

(defun my-save-all-jpeg-parts (handle)
  (when (equal (car (mm-handle-type handle)) "image/jpeg")
      (insert (mm-get-part handle))
      (write-region (point-min) (point-max)
                    (read-file-name "Save jpeg to: ")))))
(setq gnus-article-mime-part-function

Alist of MIME multipart types and functions to handle them.


Display "multipart/alternative" parts as "multipart/mixed".


Display "multipart/related" parts as "multipart/mixed".

If displaying ‘text/html’ is discouraged, see mm-discouraged-alternatives, images or other material inside a "multipart/related" part might be overlooked when this variable is nil. Display Customization in Emacs-Mime Manual.


Display "multipart" parts as "multipart/mixed". If t, it overrides nil values of gnus-mime-display-multipart-alternative-as-mixed and gnus-mime-display-multipart-related-as-mixed.


List of functions used for rewriting file names of MIME parts. Each function takes a file name as input and returns a file name.

Ready-made functions include
mm-file-name-delete-whitespace, mm-file-name-trim-whitespace, mm-file-name-collapse-whitespace, and mm-file-name-replace-whitespace. The later uses the value of the variable mm-file-name-replace-whitespace to replace each whitespace character in a file name with that string; default value is "_" (a single underscore).

The standard functions capitalize, downcase, upcase, and upcase-initials may be useful, too.

Everybody knows that whitespace characters in file names are evil, except those who don’t know. If you receive lots of attachments from such unenlightened users, you can make live easier by adding

(setq mm-file-name-rewrite-functions

to your ~/.gnus.el file.

4.20 Charsets

People use different charsets, and we have MIME to let us know what charsets they use. Or rather, we wish we had. Many people use newsreaders and mailers that do not understand or use MIME, and just send out messages without saying what character sets they use. To help a bit with this, some local news hierarchies have policies that say what character set is the default. For instance, the ‘fj’ hierarchy uses iso-2022-jp.

This knowledge is encoded in the gnus-group-charset-alist variable, which is an alist of regexps (use the first item to match full group names) and default charsets to be used when reading these groups.

In addition, some people do use soi-disant MIME-aware agents that aren’t. These blithely mark messages as being in iso-8859-1 even if they really are in koi-8. To help here, the gnus-newsgroup-ignored-charsets variable can be used. The charsets that are listed here will be ignored. The variable can be set on a group-by-group basis using the group parameters (see Group Parameters). The default value is (unknown-8bit x-unknown), which includes values some agents insist on having in there.

When posting, gnus-group-posting-charset-alist is used to determine which charsets should not be encoded using the MIME encodings. For instance, some hierarchies discourage using quoted-printable header encoding.

This variable is an alist of regexps and permitted unencoded charsets for posting. Each element of the alist has the form (test header body-list), where:


is either a regular expression matching the newsgroup header or a variable to query,


is the charset which may be left unencoded in the header (nil means encode all charsets),


is a list of charsets which may be encoded using 8bit content-transfer encoding in the body, or one of the special values nil (always encode using quoted-printable) or t (always use 8bit).

See Encoding Customization in The Emacs MIME Manual, for additional variables that control which MIME charsets are used when sending messages.

Other charset tricks that may be useful, although not Gnus-specific:

If there are several MIME charsets that encode the same Emacs charset, you can choose what charset to use by saying the following:

(put-charset-property 'cyrillic-iso8859-5
                      'preferred-coding-system 'koi8-r)

This means that Russian will be encoded using koi8-r instead of the default iso-8859-5 MIME charset.

If you want to read messages in koi8-u, you can cheat and say

(define-coding-system-alias 'koi8-u 'koi8-r)

This will almost do the right thing.

And finally, to read charsets like windows-1251, you can say something like

(codepage-setup 1251)
(define-coding-system-alias 'windows-1251 'cp1251)

4.21 Article Commands


Generate and print a PostScript image of the article buffer (gnus-summary-print-article). gnus-ps-print-hook will be run just before printing the buffer. An alternative way to print article is to use Muttprint (see Saving Articles).


If <backend>-fetch-partial-articles is non-nil, Gnus will fetch partial articles, if the backend it fetches them from supports it. Currently only nnimap does. If you’re looking at a partial article, and want to see the complete article instead, then the A C command (gnus-summary-show-complete-article) will do so.

A w

Scan the article buffer for links, and offer them to the user for browsing with browse-url. With a prefix argument, browse with browse-url-secondary-browser-function instead.

4.22 Summary Sorting

You can have the summary buffer sorted in various ways, even though I can’t really see why you’d want that.

C-c C-s C-n

Sort by article number (gnus-summary-sort-by-number).

C-c C-s C-m C-n

Sort by most recent article number (gnus-summary-sort-by-most-recent-number).

C-c C-s C-a

Sort by author (gnus-summary-sort-by-author).

C-c C-s C-t

Sort by recipient (gnus-summary-sort-by-recipient).

C-c C-s C-s

Sort by subject (gnus-summary-sort-by-subject).

C-c C-s C-d

Sort by date (gnus-summary-sort-by-date).

C-c C-s C-m C-d

Sort by most recent date (gnus-summary-sort-by-most-recent-date).

C-c C-s C-l

Sort by lines (gnus-summary-sort-by-lines).

C-c C-s C-c

Sort by article length (gnus-summary-sort-by-chars).

C-c C-s C-m C-m

Sort by article “readedness” marks (gnus-summary-sort-by-marks).

C-c C-s C-i

Sort by score (gnus-summary-sort-by-score).

C-c C-s C-u

Sort by newsgroups (gnus-summary-sort-by-newsgroups).

C-c C-s C-x

Prompts for extra header to sort by (gnus-summary-sort-by-extra). An error will be raised if no sort functions for the header are defined.

C-c C-s C-r

Randomize (gnus-summary-sort-by-random).

C-c C-s C-o

Sort using the default sorting method (gnus-summary-sort-by-original).

These functions will work both when you use threading and when you don’t use threading. In the latter case, all summary lines will be sorted, line by line. In the former case, sorting will be done on a root-by-root basis, which might not be what you were looking for. To toggle whether to use threading, type T T (see Thread Commands).

If a prefix argument if given, the sort order is reversed.

4.23 Finding the Parent


If you’d like to read the parent of the current article, and it is not displayed in the summary buffer, you might still be able to. That is, if the current group is fetched by NNTP, the parent hasn’t expired and the References in the current article are not mangled, you can just press ^ or A r (gnus-summary-refer-parent-article). If everything goes well, you’ll get the parent. If the parent is already displayed in the summary buffer, point will just move to this article.

If given a positive numerical prefix, fetch that many articles back into the ancestry. If given a negative numerical prefix, fetch just that ancestor. So if you say 3 ^, Gnus will fetch the parent, the grandparent and the great-grandparent of the current article. If you say -3 ^, Gnus will only fetch the great-grandparent of the current article.

A R (Summary)

Fetch all articles mentioned in the References header of the article (gnus-summary-refer-references).

A T (Summary)

Display the full thread where the current article appears (gnus-summary-refer-thread). By default this command looks for articles only in the current group. Some backends (currently only nnimap) know how to find articles in the thread directly. In other cases each header in the current group must be fetched and examined, so it usually takes a while. If you do it often, you may consider setting gnus-fetch-old-headers to invisible (see Filling In Threads). This won’t have any visible effects normally, but it’ll make this command work a whole lot faster. Of course, it’ll make group entry somewhat slow.

If gnus-refer-thread-use-search is non-nil then those backends that know how to find threads directly will search not just in the current group but all groups on the same server.

The gnus-refer-thread-limit variable says how many old (i.e., articles before the first displayed in the current group) headers to fetch when doing this command. The default is 200. If t, all the available headers will be fetched. This variable can be overridden by giving the A T command a numerical prefix.

In most cases gnus-refer-thread adds any articles it finds to the current summary buffer. (When gnus-refer-thread-use-search is true and the initial referral starts from a summary buffer for a non-virtual group this may not be possible. In this case a new summary buffer is created holding a virtual group with the result of the thread search.) If gnus-refer-thread-limit-to-thread is non-nil then the summary buffer will be limited to articles in the thread.

M-^ (Summary)

You can also ask Gnus for an arbitrary article, no matter what group it belongs to. M-^ (gnus-summary-refer-article) will ask you for a Message-ID, which is one of those long, hard-to-read thingies that look something like ‘<38o6up$>’. You have to get it all exactly right. No fuzzy searches, I’m afraid.

Gnus looks for the Message-ID in the headers that have already been fetched, but also tries all the select methods specified by gnus-refer-article-method if it is not found.

If the group you are reading is located on a back end that does not support fetching by Message-ID very well (like nnspool), you can set gnus-refer-article-method to an NNTP method. It would, perhaps, be best if the NNTP server you consult is the one updating the spool you are reading from, but that’s not really necessary.

It can also be a list of select methods, as well as the special symbol current, which means to use the current select method. If it is a list, Gnus will try all the methods in the list until it finds a match.

Here’s an example setting that will first try the current method, and then ask Google if that fails:

(setq gnus-refer-article-method
        (nnweb "google" (nnweb-type google))))

Most of the mail back ends support fetching by Message-ID, but do not do a particularly excellent job at it. That is, nnmbox, nnbabyl, nnmaildir, nnml, are able to locate articles from any groups, while nnfolder, and nnimap are only able to locate articles that have been posted to the current group. nnmh does not support this at all.

Fortunately, the special nnregistry back end is able to locate articles in any groups, regardless of their back end (see fetching by Message-ID using the registry).

4.24 Alternative Approaches

Different people like to read news using different methods. This being Gnus, we offer a small selection of minor modes for the summary buffers.

4.24.1 Pick and Read

Some newsreaders (like nn and, uhm, Netnews on VM/CMS) use a two-phased reading interface. The user first marks in a summary buffer the articles she wants to read. Then she starts reading the articles with just an article buffer displayed.

Gnus provides a summary buffer minor mode that allows this—gnus-pick-mode. This basically means that a few process mark commands become one-keystroke commands to allow easy marking, and it provides one additional command for switching to the summary buffer.

Here are the available keystrokes when using pick mode:


Pick the article or thread on the current line or unpick it if is already picked (gnus-pick-article-or-thread). If the variable gnus-thread-hide-subtree is true, then this key selects the entire thread when used at the first article of the thread. Otherwise, it selects just the article. If given a numerical prefix, go to that thread or article and pick it. (The line number is normally displayed at the beginning of the summary pick lines.) If gnus-process-mark-toggle is nil, this key will pick an article or thread.


Scroll the summary buffer up one page (gnus-pick-next-page). If at the end of the buffer, start reading the picked articles.


Unpick the thread or article (gnus-pick-unmark-article-or-thread). If the variable gnus-thread-hide-subtree is true, then this key unpicks the thread if used at the first article of the thread. Otherwise it unpicks just the article. You can give this key a numerical prefix to unpick the thread or article at that line.


Start reading the picked articles (gnus-pick-start-reading). If given a prefix, mark all unpicked articles as read first. If gnus-pick-display-summary is non-nil, the summary buffer will still be visible when you are reading.

All the normal summary mode commands are still available in the pick-mode, with the exception of u. However ! is available which is mapped to the same function gnus-summary-tick-article-forward.

If this sounds like a good idea to you, you could say:

(add-hook 'gnus-summary-mode-hook 'gnus-pick-mode)

gnus-pick-mode-hook is run in pick minor mode buffers.

If gnus-mark-unpicked-articles-as-read is non-nil, mark all unpicked articles as read. The default is nil.

The summary line format in pick mode is slightly different from the standard format. At the beginning of each line the line number is displayed. The pick mode line format is controlled by the gnus-summary-pick-line-format variable (see Formatting Variables). It accepts the same format specs that gnus-summary-line-format does (see Summary Buffer Lines).

4.24.2 Binary Groups

If you spend much time in binary groups, you may grow tired of hitting X u, n, RET all the time. M-x gnus-binary-mode is a minor mode for summary buffers that makes all ordinary Gnus article selection functions uudecode series of articles and display the result instead of just displaying the articles the normal way.

The only way, in fact, to see the actual articles is the g command, when you have turned on this mode (gnus-binary-show-article).

gnus-binary-mode-hook is called in binary minor mode buffers.

4.25 Tree Display

If you don’t like the normal Gnus summary display, you might try setting gnus-use-trees to t. This will create (by default) an additional tree buffer. You can execute all summary mode commands in the tree buffer.

There are a few variables to customize the tree display, of course:


A hook called in all tree mode buffers.


A format string for the mode bar in the tree mode buffers (see Mode Line Formatting). The default is ‘Gnus: %%b %S %Z’. For a list of valid specs, see Summary Buffer Mode Line.


Face used for highlighting the selected article in the tree buffer. The default is modeline.


A format string for the tree nodes. The name is a bit of a misnomer, though—it doesn’t define a line, but just the node. The default value is ‘%(%[%3,3n%]%)’, which displays the first three characters of the name of the poster. It is vital that all nodes are of the same length, so you must use ‘%4,4n’-like specifiers.

Valid specs are:


The name of the poster.


The From header.


The number of the article.


The opening bracket.


The closing bracket.


The subject.

See Formatting Variables.

Variables related to the display are:


This is used for differentiating between “real” articles and “sparse” articles. The format is

((real-open . real-close)
 (sparse-open . sparse-close)
 (dummy-open . dummy-close))

and the default is ((?[ . ?]) (?( . ?)) (?{ . ?}) (?< . ?>)).


This is a list that contains the characters used for connecting parent nodes to their children. The default is (?- ?\\ ?|).


If this variable is non-nil, Gnus will try to keep the tree buffer as small as possible to allow more room for the other Gnus windows. If this variable is a number, the tree buffer will never be higher than that number. The default is t. Note that if you have several windows displayed side-by-side in a frame and the tree buffer is one of these, minimizing the tree window will also resize all other windows displayed next to it.

You may also wish to add the following hook to keep the window minimized at all times:

(add-hook 'gnus-configure-windows-hook

The function that actually generates the thread tree. Two predefined functions are available: gnus-generate-horizontal-tree and gnus-generate-vertical-tree (which is the default).

Here’s an example from a horizontal tree buffer:

     |      \[Jan]
     |      \[odd]-[Eri]
     |      \(***)-[Eri]
     |            \[odd]-[Paa]

Here’s the same thread displayed in a vertical tree buffer:

(***)                         [Bjo] [Gun] [Gun]
  |--\-----\-----\                          |
[odd] [Jan] [odd] (***)                   [Jor]
  |           |     |--\
[Gun]       [Eri] [Eri] [odd]

If you’re using horizontal trees, it might be nice to display the trees side-by-side with the summary buffer. You could add something like the following to your ~/.gnus.el file:

(setq gnus-use-trees t
      gnus-generate-tree-function 'gnus-generate-horizontal-tree
      gnus-tree-minimize-window nil)
   (vertical 1.0
             (horizontal 0.25
                         (summary 0.75 point)
                         (tree 1.0))
             (article 1.0))))

See Window Layout.

4.26 Mail Group Commands

Some commands only make sense in mail groups. If these commands are invalid in the current group, they will raise a hell and let you know.

All these commands (except the expiry and edit commands) use the process/prefix convention (see Process/Prefix).

B e

Run all expirable articles in the current group through the expiry process (gnus-summary-expire-articles). That is, delete all expirable articles in the group that have been around for a while. (see Expiring Mail).

B C-M-e

Delete all the expirable articles in the group (gnus-summary-expire-articles-now). This means that all articles eligible for expiry in the current group will disappear forever into that big /dev/null in the sky.


Delete the mail article. This is “delete” as in “delete it from your disk forever and ever, never to return again.” Use with caution. (gnus-summary-delete-article).

B m

Move the article from one mail group to another (gnus-summary-move-article). Marks will be preserved if gnus-preserve-marks is non-nil (which is the default).

B c

Copy the article from one group (mail group or not) to a mail group (gnus-summary-copy-article). Marks will be preserved if gnus-preserve-marks is non-nil (which is the default).


Crosspost the current article to some other group (gnus-summary-crosspost-article). This will create a new copy of the article in the other group, and the Xref headers of the article will be properly updated.

B i

Import an arbitrary file into the current mail newsgroup (gnus-summary-import-article). You will be prompted for a file name, a From header and a Subject header.


Create an empty article in the current mail newsgroups (gnus-summary-create-article). You will be prompted for a From header and a Subject header.

B r

Respool the mail article (gnus-summary-respool-article). gnus-summary-respool-default-method will be used as the default select method when respooling. This variable is nil by default, which means that the current group select method will be used instead. Marks will be preserved if gnus-preserve-marks is non-nil (which is the default).

B w

Edit the current article (gnus-summary-edit-article). To finish editing and make the changes permanent, type C-c C-c (gnus-summary-edit-article-done). If you give a prefix to the C-c C-c command, Gnus won’t re-highlight the article.

B q

If you want to re-spool an article, you might be curious as to what group the article will end up in before you do the re-spooling. This command will tell you (gnus-summary-respool-query).

B t

Similarly, this command will display all fancy splitting patterns used when respooling, if any (gnus-summary-respool-trace).

B p

Some people have a tendency to send you “courtesy” copies when they follow up to articles you have posted. These usually have a Newsgroups header in them, but not always. This command (gnus-summary-article-posted-p) will try to fetch the current article from your news server (or rather, from gnus-refer-article-method or gnus-select-method) and will report back whether it found the article or not. Even if it says that it didn’t find the article, it may have been posted anyway—mail propagation is much faster than news propagation, and the news copy may just not have arrived yet.


Encrypt the body of an article (gnus-article-encrypt-body). The body is encrypted with the encryption protocol specified by the variable gnus-article-encrypt-protocol.

If you move (or copy) articles regularly, you might wish to have Gnus suggest where to put the articles. gnus-move-split-methods is a variable that uses the same syntax as gnus-split-methods (see Saving Articles). You may customize that variable to create suggestions you find reasonable. (Note that gnus-move-split-methods uses group names where gnus-split-methods uses file names.)

(setq gnus-move-split-methods
      '(("^From:.*Lars Magne" "nnml:junk")
        ("^Subject:.*gnus" "nnfolder:important")
        (".*" "nnml:misc")))

4.27 Various Summary Stuff


If non-nil, show and update the summary buffer as it’s being built. If t, update the buffer after every line is inserted. If the value is an integer, n, update the display every n lines. The default is nil.


If non-nil, display an arrow in the fringe to indicate the current article.


This hook is called when creating a summary mode buffer.


This is called as the last thing before doing the threading and the generation of the summary buffer. It’s quite convenient for customizing the threading variables based on what data the newsgroup has. This hook is called from the summary buffer after most summary buffer variables have been set.


It is called after the summary buffer has been generated. You might use it to, for instance, highlight lines or modify the look of the buffer in some other ungodly manner. I don’t care.


A hook called as the very last thing after the summary buffer has been generated.


When Gnus discovers two articles that have the same Message-ID, it has to do something drastic. No articles are allowed to have the same Message-ID, but this may happen when reading mail from some sources. Gnus allows you to customize what happens with this variable. If it is nil (which is the default), Gnus will rename the Message-ID (for display purposes only) and display the article as any other article. If this variable is t, it won’t display the article—it’ll be as if it never existed.


This function, which takes two parameters (the group name and the list of articles to be selected), is called to allow the user to alter the list of articles to be selected.

For instance, the following function adds the list of cached articles to the list in one particular group:

(defun my-add-cached-articles (group articles)
  (if (string= group "")
      (append gnus-newsgroup-cached articles)

A list of newsgroup (summary buffer) local variables, or cons of variables and their default expressions to be evalled (when the default values are not nil), that should be made global while the summary buffer is active.

Note: The default expressions will be evaluated (using function eval) before assignment to the local variable rather than just assigned to it. If the default expression is the symbol global, that symbol will not be evaluated but the global value of the local variable will be used instead.

These variables can be used to set variables in the group parameters while still allowing them to affect operations done in other buffers. For example:

(setq gnus-newsgroup-variables
        (gnus-visible-headers .

Also see Group Parameters.

4.27.1 Summary Group Information

H d

Give a brief description of the current group (gnus-summary-describe-group). If given a prefix, force rereading the description from the server.

H h

Give an extremely brief description of the most important summary keystrokes (gnus-summary-describe-briefly).

H i

Go to the Gnus info node (gnus-info-find-node).

4.27.2 Searching for Articles

M-s M-s

Search through all subsequent (raw) articles for a regexp (gnus-summary-search-article-forward).

M-s M-r

Search through all previous (raw) articles for a regexp (gnus-summary-search-article-backward).


Repeat the previous search forwards (gnus-summary-repeat-search-article-forward).


Repeat the previous search backwards (gnus-summary-repeat-search-article-backward).


This command will prompt you for a header, a regular expression to match on this field, and a command to be executed if the match is made (gnus-summary-execute-command). If the header is an empty string, the match is done on the entire article. If given a prefix, search backward instead.

For instance, & RET some.*string RET # will put the process mark on all articles that have heads or bodies that match ‘some.*string’.


Perform any operation on all articles that have been marked with the process mark (gnus-summary-universal-argument).

4.27.3 Summary Generation Commands

Y g

Regenerate the current summary buffer (gnus-summary-prepare).

Y c

Pull all cached articles (for the current group) into the summary buffer (gnus-summary-insert-cached-articles).

Y d

Pull all dormant articles (for the current group) into the summary buffer (gnus-summary-insert-dormant-articles).

Y t

Pull all ticked articles (for the current group) into the summary buffer (gnus-summary-insert-ticked-articles).

4.27.4 Really Various Summary Commands


If the current article is a collection of other articles (for instance, a digest), you might use this command to enter a group based on that article (gnus-summary-enter-digest-group). Gnus will try to guess what article type is currently displayed unless you give a prefix to this command, which forces a “digest” interpretation. Basically, whenever you see a message that is a collection of other messages of some format, you C-d and read these messages in a more convenient fashion.

The variable gnus-auto-select-on-ephemeral-exit controls what article should be selected after exiting a digest group. Valid values include:


Select the next article.


Select the next unread article.


Move the cursor to the next article. This is the default.


Move the cursor to the next unread article.

If it has any other value or there is no next (unread) article, the article selected before entering to the digest group will appear.


This command is very similar to the one above, but lets you gather several documents into one biiig group (gnus-summary-read-document). It does this by opening several nndoc groups for each document, and then opening an nnvirtual group on top of these nndoc groups. This command understands the process/prefix convention (see Process/Prefix).


Toggle truncation of summary lines (gnus-summary-toggle-truncation). This will probably confuse the line centering function in the summary buffer, so it’s not a good idea to have truncation switched off while reading articles.


Expand the summary buffer window (gnus-summary-expand-window). If given a prefix, force an article window configuration.


Edit the group parameters (see Group Parameters) of the current group (gnus-summary-edit-parameters).


Customize the group parameters (see Group Parameters) of the current group (gnus-summary-customize-parameters).

4.28 Exiting the Summary Buffer

Exiting from the summary buffer will normally update all info on the group and return you to the group buffer.


Exit the current group and update all information on the group (gnus-summary-exit). gnus-summary-prepare-exit-hook is called before doing much of the exiting, which calls gnus-summary-expire-articles by default. gnus-summary-exit-hook is called after finishing the exit process. gnus-group-no-more-groups-hook is run when returning to group mode having no more (unread) groups.


Exit the current group without updating any information on the group (gnus-summary-exit-no-update).

Z c

Mark all unticked articles in the group as read and then exit (gnus-summary-catchup-and-exit).


Mark all articles, even the ticked ones, as read and then exit (gnus-summary-catchup-all-and-exit).

Z n

Mark all articles as read and go to the next group (gnus-summary-catchup-and-goto-next-group).

Z p

Mark all articles as read and go to the previous group (gnus-summary-catchup-and-goto-prev-group).

C-x C-s

Exit this group, and then enter it again (gnus-summary-reselect-current-group). If given a prefix, select all articles, both read and unread.


Exit the group, check for new articles in the group, and select the group (gnus-summary-rescan-group). If given a prefix, select all articles, both read and unread.


Exit the group and go to the next group (gnus-summary-next-group).


Exit the group and go to the previous group (gnus-summary-prev-group).

Z s

Save the current number of read/marked articles in the dribble buffer and then save the dribble buffer (gnus-summary-save-newsrc). If given a prefix, also save the .newsrc file(s). Using this command will make exit without updating (the Q command) worthless.

gnus-exit-group-hook is called when you exit the current group with an “updating” exit. For instance Q (gnus-summary-exit-no-update) does not call this hook.

If you’re in the habit of exiting groups, and then changing your mind about it, you might set gnus-kill-summary-on-exit to nil. If you do that, Gnus won’t kill the summary buffer when you exit it. (Quelle surprise!) Instead it will change the name of the buffer to something like *Dead Summary ... * and install a minor mode called gnus-dead-summary-mode. Now, if you switch back to this buffer, you’ll find that all keys are mapped to a function called gnus-summary-wake-up-the-dead. So tapping any keys in a dead summary buffer will result in a live, normal summary buffer.

There will never be more than one dead summary buffer at any one time.

The data on the current group will be updated (which articles you have read, which articles you have replied to, etc.) when you exit the summary buffer. If the gnus-use-cross-reference variable is t (which is the default), articles that are cross-referenced to this group and are marked as read, will also be marked as read in the other subscribed groups they were cross-posted to. If this variable is neither nil nor t, the article will be marked as read in both subscribed and unsubscribed groups (see Crosspost Handling).

4.29 Crosspost Handling

Marking cross-posted articles as read ensures that you’ll never have to read the same article more than once. Unless, of course, somebody has posted it to several groups separately. Posting the same article to several groups (not cross-posting) is called spamming, and you are by law required to send nasty-grams to anyone who perpetrates such a heinous crime.

Remember: Cross-posting is kinda ok, but posting the same article separately to several groups is not. Massive cross-posting (aka. velveeta) is to be avoided at all costs, and you can even use the gnus-summary-mail-crosspost-complaint command to complain about excessive crossposting (see Summary Mail Commands).

One thing that may cause Gnus to not do the cross-posting thing correctly is if you use an NNTP server that supports XOVER (which is very nice, because it speeds things up considerably) which does not include the Xref header in its NOV lines. This is Evil, but all too common, alas, alack. Gnus tries to Do The Right Thing even with XOVER by registering the Xref lines of all articles you actually read, but if you kill the articles, or just mark them as read without reading them, Gnus will not get a chance to snoop the Xref lines out of these articles, and will be unable to use the cross reference mechanism.

To check whether your NNTP server includes the Xref header in its overview files, try ‘telnet your.nntp.server nntp’, ‘MODE READER’ on inn servers, and then say ‘LIST overview.fmt’. This may not work, but if it does, and the last line you get does not read ‘Xref:full’, then you should shout and whine at your news admin until she includes the Xref header in the overview files.

If you want Gnus to get the Xrefs right all the time, you have to set nntp-nov-is-evil to t, which slows things down considerably. Also see Slow/Expensive Connection.

C’est la vie.

For an alternative approach, see Duplicate Suppression.

4.30 Duplicate Suppression

By default, Gnus tries to make sure that you don’t have to read the same article more than once by utilizing the crossposting mechanism (see Crosspost Handling). However, that simple and efficient approach may not work satisfactory for some users for various reasons.

  1. The NNTP server may fail to generate the Xref header. This is evil and not very common.
  2. The NNTP server may fail to include the Xref header in the .overview data bases. This is evil and all too common, alas.
  3. You may be reading the same group (or several related groups) from different NNTP servers.
  4. You may be getting mail that duplicates articles posted to groups.

I’m sure there are other situations where Xref handling fails as well, but these four are the most common situations.

If, and only if, Xref handling fails for you, then you may consider switching on duplicate suppression. If you do so, Gnus will remember the Message-IDs of all articles you have read or otherwise marked as read, and then, as if by magic, mark them as read all subsequent times you see them—in all groups. Using this mechanism is quite likely to be somewhat inefficient, but not overly so. It’s certainly preferable to reading the same articles more than once.

Duplicate suppression is not a very subtle instrument. It’s more like a sledge hammer than anything else. It works in a very simple fashion—if you have marked an article as read, it adds this Message-ID to a cache. The next time it sees this Message-ID, it will mark the article as read with the ‘M’ mark. It doesn’t care what group it saw the article in.


If non-nil, suppress duplicates.


If non-nil, save the list of duplicates to a file. This will make startup and shutdown take longer, so the default is nil. However, this means that only duplicate articles read in a single Gnus session are suppressed.


This variable says how many Message-IDs to keep in the duplicate suppression list. The default is 10000.


The name of the file to store the duplicate suppression list in. The default is ~/News/suppression.

If you have a tendency to stop and start Gnus often, setting gnus-save-duplicate-list to t is probably a good idea. If you leave Gnus running for weeks on end, you may have it nil. On the other hand, saving the list makes startup and shutdown much slower, so that means that if you stop and start Gnus often, you should set gnus-save-duplicate-list to nil. Uhm. I’ll leave this up to you to figure out, I think.

4.31 Security

Gnus is able to verify signed messages or decrypt encrypted messages. The formats that are supported are PGP, PGP/MIME and S/MIME, however you need some external programs to get things to work:

  1. To handle PGP and PGP/MIME messages, you have to install an OpenPGP implementation such as GnuPG. The Lisp interface to GnuPG included with Emacs is called EasyPG (see EasyPG in EasyPG Assistant user’s manual), but Mailcrypt is also supported.
  2. To handle S/MIME message, you need to install OpenSSL. OpenSSL 0.9.6 or newer is recommended.

The variables that control security functionality on reading/composing messages include:


Option of verifying signed parts. never, not verify; always, always verify; known, only verify known protocols. Otherwise, ask user.


Option of decrypting encrypted parts. never, no decryption; always, always decrypt; known, only decrypt known protocols. Otherwise, ask user.


Option of creating signed parts. nil, use default signing keys; guided, ask user to select signing keys from the menu.


Option of creating encrypted parts. nil, use the first public-key matching the ‘From:’ header as the recipient; guided, ask user to select recipient keys from the menu.


Symbol indicating elisp interface to OpenPGP implementation for PGP messages. The default is epg, but mailcrypt is also supported although deprecated. By default, Gnus uses the first available interface in this order.


Symbol indicating elisp interface to OpenPGP implementation for PGP/MIME messages. The default is epg, but mailcrypt is also supported although deprecated. By default, Gnus uses the first available interface in this order.

By default the buttons that display security information are not shown, because they clutter reading the actual e-mail. You can type K b manually to display the information. Use the gnus-buttonized-mime-types and gnus-unbuttonized-mime-types variables to control this permanently. MIME Commands for further details, and hints on how to customize these variables to always display security information.

Snarfing OpenPGP keys (i.e., importing keys from articles into your key ring) is not supported explicitly through a menu item or command, rather Gnus do detect and label keys as ‘application/pgp-keys’, allowing you to specify whatever action you think is appropriate through the usual MIME infrastructure. You can use a ~/.mailcap entry (see mailcap in The Emacs MIME Manual) such as the following to import keys using GNU Privacy Guard when you click on the MIME button (see Using MIME).

application/pgp-keys; gpg --import --interactive --verbose; needsterminal

This happens to also be the default action defined in mailcap-mime-data.

More information on how to set things for sending outgoing signed and encrypted messages up can be found in the message manual (see Security in Message Manual).

4.32 Mailing List

Gnus understands some mailing list fields of RFC 2369. To enable it, add a to-list group parameter (see Group Parameters), possibly using A M (gnus-mailing-list-insinuate) in the summary buffer.

That enables the following commands to the summary buffer:

C-c C-n h

Send a message to fetch mailing list help, if List-Help field exists.

C-c C-n s

Send a message to subscribe the mailing list, if List-Subscribe field exists.

C-c C-n u

Send a message to unsubscribe the mailing list, if List-Unsubscribe field exists.

C-c C-n p

Post to the mailing list, if List-Post field exists.

C-c C-n o

Send a message to the mailing list owner, if List-Owner field exists.

C-c C-n a

Browse the mailing list archive, if List-Archive field exists.

5 Article Buffer

The articles are displayed in the article buffer, of which there is only one. All the summary buffers share the same article buffer unless you tell Gnus otherwise.

5.1 Hiding Headers

The top section of each article is the head. (The rest is the body, but you may have guessed that already.)

There is a lot of useful information in the head: the name of the person who wrote the article, the date it was written and the subject of the article. That’s well and nice, but there’s also lots of information most people do not want to see—what systems the article has passed through before reaching you, the Message-ID, the References, etc. ad nauseam—and you’ll probably want to get rid of some of those lines. If you want to keep all those lines in the article buffer, you can set gnus-show-all-headers to t.

Gnus provides you with two variables for sifting headers:


If this variable is non-nil, it should be a regular expression that says what headers you wish to keep in the article buffer. All headers that do not match this variable will be hidden.

For instance, if you only want to see the name of the person who wrote the article and the subject, you’d say:

(setq gnus-visible-headers "^From:\\|^Subject:")

This variable can also be a list of regexps to match headers to remain visible.


This variable is the reverse of gnus-visible-headers. If this variable is set (and gnus-visible-headers is nil), it should be a regular expression that matches all lines that you want to hide. All lines that do not match this variable will remain visible.

For instance, if you just want to get rid of the References line and the Xref line, you might say:

(setq gnus-ignored-headers "^References:\\|^Xref:")

This variable can also be a list of regexps to match headers to be removed.

Note that if gnus-visible-headers is non-nil, this variable will have no effect.

Gnus can also sort the headers for you. (It does this by default.) You can control the sorting by setting the gnus-sorted-header-list variable. It is a list of regular expressions that says in what order the headers are to be displayed.

For instance, if you want the name of the author of the article first, and then the subject, you might say something like:

(setq gnus-sorted-header-list '("^From:" "^Subject:"))

Any headers that are to remain visible, but are not listed in this variable, will be displayed in random order after all the headers listed in this variable.

You can hide further boring headers by setting gnus-treat-hide-boring-headers to head. What this function does depends on the gnus-boring-article-headers variable. It’s a list, but this list doesn’t actually contain header names. Instead it lists various boring conditions that Gnus can check and remove from sight.

These conditions are:


Remove all empty headers.


Remove the Followup-To header if it is identical to the Newsgroups header.


Remove the Reply-To header if it lists the same addresses as the From header, or if the broken-reply-to group parameter is set.


Remove the Newsgroups header if it only contains the current group name.


Remove the To header if it only contains the address identical to the current group’s to-address parameter.


Remove the To header if it only contains the address identical to the current group’s to-list parameter.


Remove the Cc header if it only contains the address identical to the current group’s to-list parameter.


Remove the Date header if the article is less than three days old.


Remove the To and/or Cc header if it is very long.


Remove all To and/or Cc headers if there are more than one.

To include these three elements, you could say something like:

(setq gnus-boring-article-headers
      '(empty followup-to reply-to))

This is also the default value for this variable.

5.2 Using MIME

Mime is a standard for waving your hands through the air, aimlessly, while people stand around yawning.

MIME, however, is a standard for encoding your articles, aimlessly, while all newsreaders die of fear.

MIME may specify what character set the article uses, the encoding of the characters, and it also makes it possible to embed pictures and other naughty stuff in innocent-looking articles.

Gnus pushes MIME articles through gnus-display-mime-function to display the MIME parts. This is gnus-display-mime by default, which creates a bundle of clickable buttons that can be used to display, save and manipulate the MIME objects.

The following commands are available when you have placed point over a MIME button:

RET (Article)
BUTTON-2 (Article)

Toggle displaying of the MIME object (gnus-article-press-button). If built-in viewers can not display the object, Gnus resorts to external viewers in the mailcap files. If a viewer has the ‘copiousoutput’ specification, the object is displayed inline.

M-RET (Article)
v (Article)

Prompt for a method, and then view the MIME object using this method (gnus-mime-view-part).

t (Article)

View the MIME object as if it were a different MIME media type (gnus-mime-view-part-as-type).

C (Article)

Prompt for a charset, and then view the MIME object using this charset (gnus-mime-view-part-as-charset).

o (Article)

Prompt for a file name, and then save the MIME object (gnus-mime-save-part).

C-o (Article)

Prompt for a file name, then save the MIME object and strip it from the article. Then proceed to article editing, where a reasonable suggestion is being made on how the altered article should look like. The stripped MIME object will be referred via the message/external-body MIME type. (gnus-mime-save-part-and-strip).

r (Article)

Prompt for a file name, replace the MIME object with an external body referring to the file via the message/external-body MIME type. (gnus-mime-replace-part).

d (Article)

Delete the MIME object from the article and replace it with some information about the removed MIME object (gnus-mime-delete-part).

c (Article)

Copy the MIME object to a fresh buffer and display this buffer (gnus-mime-copy-part). If given a prefix, copy the raw contents without decoding. If given a numerical prefix, you can do semi-manual charset stuff (see gnus-summary-show-article-charset-alist in Scrolling the Article). Compressed files like .gz and .bz2 are automatically decompressed if auto-compression-mode is enabled (see Accessing Compressed Files in The Emacs Editor).

p (Article)

Print the MIME object (gnus-mime-print-part). This command respects the ‘print=’ specifications in the .mailcap file.

i (Article)

Insert the contents of the MIME object into the buffer (gnus-mime-inline-part) as ‘text/plain’. If given a prefix, insert the raw contents without decoding. If given a numerical prefix, you can do semi-manual charset stuff (see gnus-summary-show-article-charset-alist in Scrolling the Article). Compressed files like .gz and .bz2 are automatically decompressed depending on jka-compr regardless of auto-compression-mode (see Accessing Compressed Files in The Emacs Editor).

E (Article)

View the MIME object with an internal viewer. If no internal viewer is available, use an external viewer (gnus-mime-view-part-internally).

e (Article)

View the MIME object with an external viewer. (gnus-mime-view-part-externally).

| (Article)

Output the MIME object to a process (gnus-mime-pipe-part).

. (Article)

Interactively run an action on the MIME object (gnus-mime-action-on-part).

Gnus will display some MIME objects automatically. The way Gnus determines which parts to do this with is described in the Emacs MIME manual.

It might be best to just use the toggling functions from the article buffer to avoid getting nasty surprises. (For instance, you enter the group ‘alt.sing-a-long’ and, before you know it, MIME has decoded the sound file in the article and some horrible sing-a-long song comes screaming out your speakers, and you can’t find the volume button, because there isn’t one, and people are starting to look at you, and you try to stop the program, but you can’t, and you can’t find the program to control the volume, and everybody else in the room suddenly decides to look at you disdainfully, and you’ll feel rather stupid.)

Any similarity to real events and people is purely coincidental. Ahem.

Also see MIME Commands.

5.3 HTML

Gnus can display HTML articles nicely formatted in the article buffer. There are many methods for doing that, but two of them are kind of default methods.

If your Emacs copy has been built with libxml2 support, then Gnus uses Emacs’ built-in, plain elisp Simple HTML Renderer shr 2 which is also used by Emacs’ browser EWW (see EWW in The Emacs Manual).

If your Emacs copy lacks libxml2 support but you have w3m installed on your system, Gnus uses that to render HTML mail and display the results in the article buffer (gnus-w3m).

For a complete overview, consult See Display Customization in The Emacs MIME Manual. This section only describes the default method.


If set to shr, Gnus uses its own simple HTML renderer. If set to gnus-w3m, it uses w3m.


External images that have URLs that match this regexp won’t be fetched and displayed. For instance, to block all URLs that have the string “ads” in them, do the following:

(setq gnus-blocked-images "ads")

This can also be a function to be evaluated. If so, it will be called with the group name as the parameter. The default value is gnus-block-private-groups, which will return ‘"."’ for anything that isn’t a newsgroup. This means that no external images will be fetched as a result of reading mail, so that nobody can use web bugs (and the like) to track whether you’ve read email.

If you have specific private groups that you want to have treated as if they were public groups, you can add the name of that group to the gnus-global-groups list.

Also see Misc Article for gnus-inhibit-images.


The width to use when rendering HTML. The default is 70.


How big pictures displayed are in relation to the window they’re in. A value of 0.7 (the default) means that they are allowed to take up 70% of the width and height of the window. If they are larger than this, and Emacs supports it, then the images will be rescaled down to fit these criteria.


If non-nil, display the cursor in the article buffer even when the article buffer isn’t the current buffer.

To use this, make sure that you have w3m and curl installed. If you have, then Gnus should display HTML automatically.

5.4 Customizing Articles

A slew of functions for customizing how the articles are to look like exist. You can call these functions interactively (see Article Washing), or you can have them called automatically when you select the articles.

To have them called automatically, you should set the corresponding “treatment” variable. For instance, to have headers hidden, you’d set gnus-treat-hide-headers. Below is a list of variables that can be set, but first we discuss the values these variables can have.

Note: Some values, while valid, make little sense. Check the list below for sensible values.

  1. nil: Don’t do this treatment.
  2. t: Do this treatment on all body parts.
  3. head: Do the treatment on the headers.
  4. first: Do this treatment on the first body part.
  5. last: Do this treatment on the last body part.
  6. An integer: Do this treatment on all body parts that have a length less than this number.
  7. A list of strings: Do this treatment on all body parts that are in articles that are read in groups that have names that match one of the regexps in the list.
  8. A list where the first element is not a string:

    The list is evaluated recursively. The first element of the list is a predicate. The following predicates are recognized: or, and, not and typep. Here’s an example:

    (or last
        (typep "text/x-vcard"))
  9. A function: the function is called with no arguments and should return nil or non-nil. The current article is available in the buffer named by gnus-article-buffer.

You may have noticed that the word part is used here. This refers to the fact that some messages are MIME multipart articles that may be divided into several parts. Articles that are not multiparts are considered to contain just a single part.

Are the treatments applied to all sorts of multipart parts? Yes, if you want to, but by default, only ‘text/plain’ parts are given the treatment. This is controlled by the gnus-article-treat-types variable, which is a list of regular expressions that are matched to the type of the part. This variable is ignored if the value of the controlling variable is a predicate list, as described above.

The following treatment options are available. The easiest way to customize this is to examine the gnus-article-treat customization group. Values in parenthesis are suggested sensible values. Others are possible but those listed are probably sufficient for most people.

gnus-treat-buttonize (t, integer)
gnus-treat-buttonize-head (head)

See Article Buttons.

gnus-treat-capitalize-sentences (t, integer)
gnus-treat-overstrike (t, integer)
gnus-treat-strip-cr (t, integer)
gnus-treat-emojize-symbols (t, integer)
gnus-treat-strip-headers-in-body (t, integer)
gnus-treat-strip-leading-blank-lines (t, first, integer)
gnus-treat-strip-multiple-blank-lines (t, integer)
gnus-treat-strip-pem (t, last, integer)
gnus-treat-strip-trailing-blank-lines (t, last, integer)
gnus-treat-unsplit-urls (t, integer)
gnus-treat-wash-html (t, integer)

See Article Washing.

gnus-treat-date (head)

This will transform/add date headers according to the gnus-article-date-headers variable. This is a list of Date headers to display. The formats available are:


Universal time, aka GMT, aka ZULU.


The user’s local time zone.


A semi-readable English sentence.


The time elapsed since the message was posted.


Both the original date header and a (shortened) elapsed time.


Both the time in the user’s local time zone a (shortened) elapsed time.


The original date header.


ISO8601 format, i.e., “2010-11-23T22:05:21”.


A format done according to the gnus-article-time-format variable.

See Article Date.

gnus-treat-from-picon (head)
gnus-treat-mail-picon (head)
gnus-treat-newsgroups-picon (head)

See Picons.

gnus-treat-from-gravatar (head)
gnus-treat-mail-gravatar (head)

See Gravatars.

gnus-treat-display-smileys (t, integer)
gnus-treat-body-boundary (head)

Adds a delimiter between header and body, the string used as delimiter is controlled by gnus-body-boundary-delimiter.

See Smileys.

gnus-treat-display-x-face (head)

See X-Face.

gnus-treat-display-face (head)

See Face.

gnus-treat-emphasize (t, head, integer)
gnus-treat-fill-article (t, integer)
gnus-treat-fill-long-lines (t, integer)
gnus-treat-hide-boring-headers (head)
gnus-treat-hide-citation (t, integer)
gnus-treat-hide-citation-maybe (t, integer)
gnus-treat-hide-headers (head)
gnus-treat-hide-signature (t, last)
gnus-treat-strip-banner (t, last)
gnus-treat-strip-list-identifiers (head)

See Article Hiding.

gnus-treat-highlight-citation (t, integer)
gnus-treat-highlight-headers (head)
gnus-treat-highlight-signature (t, last, integer)

See Article Highlighting.

gnus-treat-ansi-sequences (t)
gnus-treat-x-pgp-sig (head)
gnus-treat-unfold-headers (head)
gnus-treat-fold-headers (head)
gnus-treat-fold-newsgroups (head)
gnus-treat-leading-whitespace (head)

See Article Header.

You can, of course, write your own functions to be called from gnus-part-display-hook. The functions are called narrowed to the part, and you can do anything you like, pretty much. There is no information that you have to keep in the buffer—you can change everything.

5.5 Article Keymap

Most of the keystrokes in the summary buffer can also be used in the article buffer. They should behave as if you typed them in the summary buffer, which means that you don’t actually have to have a summary buffer displayed while reading. You can do it all from the article buffer.

The key v is reserved for users. You can bind it to some command or better use it as a prefix key.

A few additional keystrokes are available:


Scroll forwards one page (gnus-article-next-page). This is exactly the same as h SPC h.


Scroll backwards one page (gnus-article-prev-page). This is exactly the same as h DEL h.

C-c ^

If point is in the neighborhood of a Message-ID and you press C-c ^, Gnus will try to get that article from the server (gnus-article-refer-article).

C-c C-m

Send a reply to the address near point (gnus-article-mail). If given a prefix, include the mail.


Reconfigure the buffers so that the summary buffer becomes visible (gnus-article-show-summary).


Give a very brief description of the available keystrokes (gnus-article-describe-briefly).


Go to the next button, if any (gnus-article-next-button). This only makes sense if you have buttonizing turned on.


Go to the previous button, if any (gnus-article-prev-button).


Send a reply to the current article and yank the current article (gnus-article-reply-with-original). If the region is active, only yank the text in the region.


Send a wide reply to the current article and yank the current article (gnus-article-wide-reply-with-original). If the region is active, only yank the text in the region.


Send a followup to the current article and yank the current article (gnus-article-followup-with-original). If the region is active, only yank the text in the region.

5.6 Misc Article


If non-nil, use the same article buffer for all the groups. (This is the default.) If nil, each group will have its own article buffer.


If non-nil, selecting the article buffer with the h command will “widen” the article window to take the entire frame.


Hook used to decode MIME articles. The default value is (article-decode-charset article-decode-encoded-words)


This hook is called right after the article has been inserted into the article buffer. It is mainly intended for functions that do something depending on the contents; it should probably not be used for changing the contents of the article buffer.


Hook called in article mode buffers.


Syntax table used in article buffers. It is initialized from text-mode-syntax-table.


If non-nil, allow scrolling the article buffer even when there no more new text to scroll in. The default is nil.


This variable is a format string along the same lines as gnus-summary-mode-line-format (see Summary Buffer Mode Line). It accepts the same format specifications as that variable, with two extensions:


The wash status of the article. This is a short string with one character for each possible article wash operation that may have been performed. The characters and their meaning:


Displayed when cited text may be hidden in the article buffer.


Displayed when headers are hidden in the article buffer.


Displayed when article is digitally signed or encrypted, and Gnus has hidden the security headers. (N.B. does not tell anything about security status, i.e., good or bad signature.)


Displayed when the signature has been hidden in the Article buffer.


Displayed when Gnus has treated overstrike characters in the article buffer.


Displayed when Gnus has treated emphasized strings in the article buffer.


The number of MIME parts in the article.


Controls whether page breaking is to take place. If this variable is non-nil, the articles will be divided into pages whenever a page delimiter appears in the article. If this variable is nil, paging will not be done.


This is the delimiter mentioned above. By default, it is ‘^L’ (formfeed).


This variable controls whether Gnus performs IDNA decoding of internationalized domain names inside ‘From’, ‘To’ and ‘Cc’ headers. See IDNA in The Message Manual, for how to compose such messages. This requires GNU Libidn, and this variable is only enabled if you have installed it.


If this is non-nil, inhibit displaying of images inline in the article body. It is effective to images that are in articles as MIME parts, and images in HTML articles rendered when mm-text-html-renderer (see Display Customization in The Emacs MIME Manual) is shr or gnus-w3m.

6 Composing Messages

All commands for posting and mailing will put you in a message buffer where you can edit the article all you like, before you send the article by pressing C-c C-c. See Overview in Message Manual. Where the message will be posted/mailed to depends on your setup (see Posting Server).

Also see Canceling Articles for information on how to remove articles you shouldn’t have posted.

6.1 Mail

Variables for customizing outgoing mail:


List of regexps to match headers included in digested messages. The headers will be included in the sequence they are matched. If nil include all headers.


If non-nil, add a to-list group parameter to mail groups that have none when you do a a.


If non-nil, Gnus will ask you for a confirmation when you are about to reply to news articles by mail. If it is nil, nothing interferes in what you want to do. This can also be a function receiving the group name as the only parameter which should return non-nil if a confirmation is needed, or a regular expression matching group names, where confirmation should be asked for.

If you find yourself never wanting to reply to mail, but occasionally press R anyway, this variable might be for you.


If non-nil, Gnus also requests confirmation according to gnus-confirm-mail-reply-to-news when replying to mail. This is useful for treating mailing lists like newsgroups.

6.2 Posting Server

When you press those magical C-c C-c keys to ship off your latest (extremely intelligent, of course) article, where does it go?

Thank you for asking. I hate you.

It can be quite complicated.

When posting news, Message usually invokes message-send-news (see News Variables in Message Manual). Normally, Gnus will post using the same select method as you’re reading from (which might be convenient if you’re reading lots of groups from different private servers). However. If the server you’re reading from doesn’t allow posting, just reading, you probably want to use some other server to post your (extremely intelligent and fabulously interesting) articles. You can then set the gnus-post-method to some other method:

(setq gnus-post-method '(nnspool ""))

Now, if you’ve done this, and then this server rejects your article, or this server is down, what do you do then? To override this variable you can use a non-zero prefix to the C-c C-c command to force using the “current” server, to get back the default behavior, for posting.

If you give a zero prefix (i.e., C-u 0 C-c C-c) to that command, Gnus will prompt you for what method to use for posting.

You can also set gnus-post-method to a list of select methods. If that’s the case, Gnus will always prompt you for what method to use for posting.

Finally, if you want to always post using the native select method, you can set this variable to native.

When sending mail, Message invokes the function specified by the variable message-send-mail-function. Gnus tries to set it to a value suitable for your system. See Mail Variables in Message manual, for more information.

6.3 POP before SMTP

Does your ISP use POP-before-SMTP authentication? This authentication method simply requires you to contact the POP server before sending email. To do that, put the following lines in your ~/.gnus.el file:

(add-hook 'message-send-mail-hook 'mail-source-touch-pop)

The mail-source-touch-pop function does POP authentication according to the value of mail-sources without fetching mails, just before sending a mail. See Mail Sources.

If you have two or more POP mail servers set in mail-sources, you may want to specify one of them to mail-source-primary-source as the POP mail server to be used for the POP-before-SMTP authentication. If it is your primary POP mail server (i.e., you are fetching mails mainly from that server), you can set it permanently as follows:

(setq mail-source-primary-source
      '(pop :server "pop3.mail.server"
            :password "secret"))

Otherwise, bind it dynamically only when performing the POP-before-SMTP authentication as follows:

(add-hook 'message-send-mail-hook
          (lambda ()
            (let ((mail-source-primary-source
                   '(pop :server "pop3.mail.server"
                         :password "secret")))

6.4 Mail and Post

Here’s a list of variables relevant to both mailing and posting:


If your news server offers groups that are really mailing lists gatewayed to the NNTP server, you can read those groups without problems, but you can’t post/followup to them without some difficulty. One solution is to add a to-address to the group parameters (see Group Parameters). An easier thing to do is set the gnus-mailing-list-groups to a regexp that matches the groups that really are mailing lists. Then, at least, followups to the mailing lists will work most of the time. Posting to these groups (a) is still a pain, though.


This variable controls which information should be exposed in the User-Agent header. It can be a list of symbols or a string. Valid symbols are gnus (show Gnus version) and emacs (show Emacs version). In addition to the Emacs version, you can add config (show system configuration) or type (show system type). If you set it to a string, be sure to use a valid format, see RFC 2616.

You may want to do spell-checking on messages that you send out. Or, if you don’t want to spell-check by hand, you could add automatic spell-checking via the ispell package:

(add-hook 'message-send-hook 'ispell-message)

If you want to change the ispell dictionary based on what group you’re in, you could say something like the following:

(add-hook 'gnus-select-group-hook
          (lambda ()
               "^de\\." (gnus-group-real-name gnus-newsgroup-name))
              (ispell-change-dictionary "deutsch"))
              (ispell-change-dictionary "english")))))

Modify to suit your needs.

If gnus-message-highlight-citation is t, different levels of citations are highlighted like in Gnus article buffers also in message mode buffers.

6.5 Archived Messages

Gnus provides a few different methods for storing the mail and news you send. The default method is to use the archive virtual server to store the messages. If you want to disable this completely, the gnus-message-archive-group variable should be nil. The default is "sent.%Y-%m", which gives you one archive group per month.

For archiving interesting messages in a group you read, see the B c (gnus-summary-copy-article) command (see Mail Group Commands).

gnus-message-archive-method says what virtual server Gnus is to use to store sent messages. The default is "archive", and when actually being used it is expanded into:

(nnfolder "archive"
          (nnfolder-directory   "~/Mail/archive")
          (nnfolder-active-file "~/Mail/archive/active")
          (nnfolder-get-new-mail nil)
          (nnfolder-inhibit-expiry t))

Note: a server like this is saved in the ~/.newsrc.eld file first so that it may be used as a real method of the server which is named "archive" (that is, for the case where gnus-message-archive-method is set to "archive") ever since. If it once has been saved, it will never be updated by default even if you change the value of gnus-message-archive-method afterward. Therefore, the server "archive" doesn’t necessarily mean the nnfolder server like this at all times. If you want the saved method to reflect always the value of gnus-message-archive-method, set the gnus-update-message-archive-method variable to a non-nil value. The default value of this variable is nil.

You can, however, use any mail select method (nnml, nnmbox, etc.). nnfolder is a quite likable select method for doing this sort of thing, though. If you don’t like the default directory chosen, you could say something like:

(setq gnus-message-archive-method
      '(nnfolder "archive"
                 (nnfolder-inhibit-expiry t)
                 (nnfolder-active-file "~/News/sent-mail/active")
                 (nnfolder-directory "~/News/sent-mail/")))

Gnus will insert Gcc headers in all outgoing messages that point to one or more group(s) on that server. Which group to use is determined by the gnus-message-archive-group variable.

This variable can be used to do the following:

a string

Messages will be saved in that group.

Note that you can include a select method in the group name, then the message will not be stored in the select method given by gnus-message-archive-method, but in the select method specified by the group name, instead. Suppose gnus-message-archive-method has the default value shown above. Then setting gnus-message-archive-group to "foo" means that outgoing messages are stored in ‘nnfolder+archive:foo’, but if you use the value "nnml:foo", then outgoing messages will be stored in ‘nnml:foo’.

a list of strings

Messages will be saved in all those groups.

an alist of regexps, functions and forms

When a key “matches”, the result is used.


No message archiving will take place.

Let’s illustrate:

Just saving to a single group called ‘MisK’:

(setq gnus-message-archive-group "MisK")

Saving to two groups, ‘MisK’ and ‘safe’:

(setq gnus-message-archive-group '("MisK" "safe"))

Save to different groups based on what group you are in:

(setq gnus-message-archive-group
      '(("^alt" "sent-to-alt")
        ("mail" "sent-to-mail")
        (".*" "sent-to-misc")))

More complex stuff:

(setq gnus-message-archive-group
      '((if (message-news-p)

How about storing all news messages in one file, but storing all mail messages in one file per month:

(setq gnus-message-archive-group
      '((if (message-news-p)
          (concat "mail." (format-time-string "%Y-%m")))))

Now, when you send a message off, it will be stored in the appropriate group. (If you want to disable storing for just one particular message, you can just remove the Gcc header that has been inserted.) The archive group will appear in the group buffer the next time you start Gnus, or the next time you press F in the group buffer. You can enter it and read the articles in it just like you’d read any other group. If the group gets really big and annoying, you can simply rename if (using G r in the group buffer) to something nice—‘misc-mail-september-1995’, or whatever. New messages will continue to be stored in the old (now empty) group.


If non-nil, automatically mark Gcc articles as read.


If nil, attach files as normal parts in Gcc copies; if a regexp and matches the Gcc group name, attach files as external parts; if it is all, attach local files as external parts; if it is other non-nil, the behavior is the same as all, but it may be changed in the future.


Like the gcc-self group parameter, applied only for unmodified messages that gnus-summary-resend-message (see Summary Mail Commands) resends. Non-nil value of this variable takes precedence over any existing Gcc header.

If this is none, no Gcc copy will be made. If this is t, messages resent will be Gcc copied to the current group. If this is a string, it specifies a group to which resent messages will be Gcc copied. If this is nil, Gcc will be done according to existing Gcc header(s), if any. If this is no-gcc-self, that is the default, resent messages will be Gcc copied to groups that existing Gcc header specifies, except for the current group.


These hooks are run before/after encoding the message body of the Gcc copy of a sent message. The current buffer (when the hook is run) contains the message including the message header. Changes made to the message will only affect the Gcc copy, but not the original message. You can use these hooks to edit the copy (and influence subsequent transformations), e.g., remove MML secure tags (see Signing and encrypting).

6.6 Posting Styles

All them variables, they make my head swim.

So what if you want a different Organization and signature based on what groups you post to? And you post both from your home machine and your work machine, and you want different From lines, and so on?

One way to do stuff like that is to write clever hooks that change the variables you need to have changed. That’s a bit boring, so somebody came up with the bright idea of letting the user specify these things in a handy alist. Here’s an example of a gnus-posting-styles variable:

  (signature "Peace and happiness")
  (organization "What me?"))
  (signature "Death to everybody"))
  (organization "Emacs is it")))

As you might surmise from this example, this alist consists of several styles. Each style will be applicable if the first element “matches”, in some form or other. The entire alist will be iterated over, from the beginning towards the end, and each match will be applied, which means that attributes in later styles that match override the same attributes in earlier matching styles. So ‘comp.programming.literate’ will have the ‘Death to everybody’ signature and the ‘What me?Organization header.

The first element in each style is called the match. If it’s a string, then Gnus will try to regexp match it against the group name. If it is the form (header match regexp), then Gnus will look in the original article for a header whose name is match and compare that regexp. match and regexp are strings. (The original article is the one you are replying or following up to. If you are not composing a reply or a followup, then there is nothing to match against.) If the match is a function symbol, that function will be called with no arguments. If it’s a variable symbol, then the variable will be referenced. If it’s a list, then that list will be evaled. In any case, if this returns a non-nil value, then the style is said to match.

Each style may contain an arbitrary amount of attributes. Each attribute consists of a (name value) pair. In addition, you can also use the (name :file value) form or the (name :value value) form. Where :file signifies value represents a file name and its contents should be used as the attribute value, :value signifies value does not represent a file name explicitly. The attribute name can be one of:

  • signature
  • signature-file
  • x-face-file
  • address, overriding user-mail-address
  • name, overriding (user-full-name)
  • body

Note that the signature-file attribute honors the variable message-signature-directory.

The attribute name can also be a string or a symbol. In that case, this will be used as a header name, and the value will be inserted in the headers of the article; if the value is nil, the header name will be removed. If the attribute name is eval, the form is evaluated, and the result is thrown away.

The attribute value can be a string, a function with zero arguments (the return value will be used), a variable (its value will be used) or a list (it will be evaled and the return value will be used). The functions and sexps are called/evaled in the message buffer that is being set up.

In the case of a string value, if the match is a regular expression, or if it takes the form (header match regexp), a ‘gnus-match-substitute-replacement’ is proceed on the value to replace the positional parameters ‘\n’ by the corresponding parenthetical matches (see Replacing the Text that Matched in The Emacs Lisp Reference Manual.)

If you wish to check whether the message you are about to compose is meant to be a news article or a mail message, you can check the values of the message-news-p and message-mail-p functions.

So here’s a new example:

(setq gnus-posting-styles
         (signature-file "~/.signature")
         (name "User Name")
         (x-face-file "~/.xface")
         (x-url (getenv "WWW_HOME"))
         (organization "People's Front Against MWM"))
         (signature my-funny-signature-randomizer))
        ((equal (system-name) "gnarly")  ;; A form
         (signature my-quote-randomizer))
        (message-news-p        ;; A function symbol
         (signature my-news-signature))
        (window-system         ;; A value symbol
         ("X-Window-System" (format "%s" window-system)))
        ;; If I’m replying to Larsi, set the Organization header.
        ((header "from" "larsi.*org")
         (Organization "Somewhere, Inc."))
        ;; Reply to a message from the same subaddress the message
        ;; was sent to.
        ((header "x-original-to" "me\\(\\+.+\\)")
         (address "me\\"))
        ((posting-from-work-p) ;; A user defined function
         (signature-file "~/.work-signature")
         (address "")
         (body "You are fired.\n\nSincerely, your boss.")
         ("X-Message-SMTP-Method" "smtp 587")
         (organization "Important Work, Inc"))
         (From (with-current-buffer gnus-article-buffer
                 (message-fetch-field "to"))))
         (signature-file "~/.mail-signature"))))

The ‘nnml:.*’ rule means that you use the To address as the From address in all your outgoing replies, which might be handy if you fill many roles. You may also use message-alternative-emails instead. See Message Headers in Message Manual.

Of particular interest in the “work-mail” style is the ‘X-Message-SMTP-Method’ header. It specifies how to send the outgoing email. You may want to sent certain emails through certain SMTP servers due to company policies, for instance. See Message Variables in Message Manual.

6.7 Drafts

If you are writing a message (mail or news) and suddenly remember that you have a steak in the oven (or some pesto in the food processor, you craaazy vegetarians), you’ll probably wish there was a method to save the message you are writing so that you can continue editing it some other day, and send it when you feel its finished.

Well, don’t worry about it. Whenever you start composing a message of some sort using the Gnus mail and post commands, the buffer you get will automatically associate to an article in a special draft group. If you save the buffer the normal way (C-x C-s, for instance), the article will be saved there. (Auto-save files also go to the draft group.)

The draft group is a special group (which is implemented as an nndraft group, if you absolutely have to know) called ‘nndraft:drafts’. The variable nndraft-directory says where nndraft is to store its files. What makes this group special is that you can’t tick any articles in it or mark any articles as read—all articles in the group are permanently unread.

If the group doesn’t exist, it will be created and you’ll be subscribed to it. The only way to make it disappear from the Group buffer is to unsubscribe it. The special properties of the draft group comes from a group property (see Group Parameters), and if lost the group behaves like any other group. This means the commands below will not be available. To restore the special properties of the group, the simplest way is to kill the group, using C-k, and restart Gnus. The group is automatically created again with the correct parameters. The content of the group is not lost.

When you want to continue editing the article, you simply enter the draft group and push D e (gnus-draft-edit-message) to do that. You will be placed in a buffer where you left off.

Rejected articles will also be put in this draft group (see Rejected Articles).

If you have lots of rejected messages you want to post (or mail) without doing further editing, you can use the D s command (gnus-draft-send-message). This command understands the process/prefix convention (see Process/Prefix). The D S command (gnus-draft-send-all-messages) will ship off all messages in the buffer.

If you have some messages that you wish not to send, you can use the D t (gnus-draft-toggle-sending) command to mark the message as unsendable. This is a toggling command.

Finally, if you want to delete a draft, use the normal B DEL command (see Mail Group Commands).

6.8 Rejected Articles

Sometimes a news server will reject an article. Perhaps the server doesn’t like your face. Perhaps it just feels miserable. Perhaps there be demons. Perhaps you have included too much cited text. Perhaps the disk is full. Perhaps the server is down.

These situations are, of course, totally beyond the control of Gnus. (Gnus, of course, loves the way you look, always feels great, has angels fluttering around inside of it, doesn’t care about how much cited text you include, never runs full and never goes down.) So Gnus saves these articles until some later time when the server feels better.

The rejected articles will automatically be put in a special draft group (see Drafts). When the server comes back up again, you’d then typically enter that group and send all the articles off.

6.9 Signing and encrypting

Gnus can digitally sign and encrypt your messages, using vanilla PGP format or PGP/MIME or S/MIME. For decoding such messages, see the mm-verify-option and mm-decrypt-option options (see Security).

Often, you would like to sign replies to people who send you signed messages. Even more often, you might want to encrypt messages which are in reply to encrypted messages. Gnus offers gnus-message-replysign to enable the former, and gnus-message-replyencrypt for the latter. In addition, setting gnus-message-replysignencrypted (on by default) will sign automatically encrypted messages.

Instructing MML to perform security operations on a MIME part is done using the C-c C-m s key map for signing and the C-c C-m c key map for encryption, as follows.

C-c C-m s s

Digitally sign current message using S/MIME.

C-c C-m s o

Digitally sign current message using PGP.

C-c C-m s p

Digitally sign current message using PGP/MIME.

C-c C-m c s

Digitally encrypt current message using S/MIME.

C-c C-m c o

Digitally encrypt current message using PGP.

C-c C-m c p

Digitally encrypt current message using PGP/MIME.

C-c C-m C-n

Remove security related MML tags from message.

See Security in Message Manual, for more information.

7 Select Methods

A foreign group is a group not read by the usual (or default) means. It could be, for instance, a group from a different NNTP server, it could be a virtual group, or it could be your own personal mail group.

A foreign group (or any group, really) is specified by a name and a select method. To take the latter first, a select method is a list where the first element says what back end to use (e.g., nntp, nnspool, nnml) and the second element is the server name. There may be additional elements in the select method, where the value may have special meaning for the back end in question.

One could say that a select method defines a virtual server—so we do just that (see Server Buffer).

The name of the group is the name the back end will recognize the group as.

For instance, the group ‘soc.motss’ on the NNTP server ‘’ will have the name ‘soc.motss’ and select method (nntp ""). Gnus will call this group ‘’, even though the nntp back end just knows this group as ‘soc.motss’.

The different methods all have their peculiarities, of course.

7.1 Server Buffer

Traditionally, a server is a machine or a piece of software that one connects to, and then requests information from. Gnus does not connect directly to any real servers, but does all transactions through one back end or other. But that’s just putting one layer more between the actual media and Gnus, so we might just as well say that each back end represents a virtual server.

For instance, the nntp back end may be used to connect to several different actual NNTP servers, or, perhaps, to many different ports on the same actual NNTP server. You tell Gnus which back end to use, and what parameters to set by specifying a select method.

These select method specifications can sometimes become quite complicated—say, for instance, that you want to read from the NNTP server ‘’ on port number 13, which hangs if queried for NOV headers and has a buggy select. Ahem. Anyway, if you had to specify that for each group that used this server, that would be too much work, so Gnus offers a way of naming select methods, which is what you do in the server buffer.

To enter the server buffer, use the ^ (gnus-group-enter-server-mode) command in the group buffer.

gnus-server-mode-hook is run when creating the server buffer.

7.1.1 Server Buffer Format

You can change the look of the server buffer lines by changing the gnus-server-line-format variable. This is a format-like variable, with some simple extensions:


How the news is fetched—the back end name.


The name of this server.


Where the news is to be fetched from—the address.


The opened/closed/denied status of the server.


Whether this server is agentized.

The mode line can also be customized by using the gnus-server-mode-line-format variable (see Mode Line Formatting). The following specs are understood:


Server name.


Server method.

Also see Formatting Variables.

7.1.2 Server Commands

The following key bindings are available in the server buffer. Be aware that some of the commands will only work on servers that you’ve added through this interface (with a), not with servers you’ve defined in your init files.


The key v is reserved for users. You can bind it to some command or better use it as a prefix key.


Add a new server (gnus-server-add-server).


Edit a server (gnus-server-edit-server).


Show the definition of a server (gnus-server-show-server).


Browse the current server (gnus-server-read-server).


Return to the group buffer (gnus-server-exit).


Kill the current server (gnus-server-kill-server).


Yank the previously killed server (gnus-server-yank-server).


Copy the current server (gnus-server-copy-server).


List all servers (gnus-server-list-servers).


Request that the server scan its sources for new articles (gnus-server-scan-server). This is mainly sensible with mail servers.


Request that the server regenerate all its data structures (gnus-server-regenerate-server). This can be useful if you have a mail back end that has gotten out of sync.


Compact all groups in the server under point (gnus-server-compact-server). Currently implemented only in nnml (see Mail Spool). This removes gaps between article numbers, hence getting a correct total article count.

Some more commands for closing, disabling, and re-opening servers are listed in Unavailable Servers.

7.1.3 Example Methods

Most select methods are pretty simple and self-explanatory:

(nntp "")

Reading directly from the spool is even simpler:

(nnspool "")

As you can see, the first element in a select method is the name of the back end, and the second is the address, or name, if you will.

After these two elements, there may be an arbitrary number of (variable form) pairs.

To go back to the first example—imagine that you want to read from port 15 on that machine. This is what the select method should look like then:

(nntp "" (nntp-port-number 15))

You should read the documentation to each back end to find out what variables are relevant, but here’s an nnmh example:

nnmh is a mail back end that reads a spool-like structure. Say you have two structures that you wish to access: One is your private mail spool, and the other is a public one. Here’s the possible spec for your private mail:

(nnmh "private" (nnmh-directory "~/private/mail/"))

(This server is then called ‘private’, but you may have guessed that.)

Here’s the method for a public spool:

(nnmh "public"
      (nnmh-directory "/usr/information/spool/")
      (nnmh-get-new-mail nil))

If you are behind a firewall and only have access to the NNTP server from the firewall machine, you can instruct Gnus to rlogin on the firewall machine and connect with netcat from there to the NNTP server. Doing this can be rather fiddly, but your virtual server definition should probably look something like this:

(nntp "firewall"
      (nntp-open-connection-function nntp-open-via-rlogin-and-netcat)
      (nntp-via-address "the.firewall.machine")
      (nntp-address ""))

If you want to use the wonderful ssh program to provide a compressed connection over the modem line, you could add the following configuration to the example above:

      (nntp-via-rlogin-command "ssh")

See also nntp-via-rlogin-command-switches. Here’s an example for an indirect connection:

(setq gnus-select-method
      '(nntp "indirect"
             (nntp-address "news.server.example")
             (nntp-via-user-name "intermediate_user_name")
             (nntp-via-address "")
             (nntp-via-rlogin-command "ssh")
             (nntp-via-rlogin-command-switches ("-C"))
             (nntp-open-connection-function nntp-open-via-rlogin-and-netcat)))

This means that you have to have set up ssh-agent correctly to provide automatic authorization, of course.

If you’re behind a firewall, but have direct access to the outside world through a wrapper command like "runsocks", you could open a socksified netcat connection to the news server as follows:

(nntp "outside"
      (nntp-pre-command "runsocks")
      (nntp-open-connection-function nntp-open-netcat-stream)
      (nntp-address ""))

7.1.4 Creating a Virtual Server

If you’re saving lots of articles in the cache by using persistent articles, you may want to create a virtual server to read the cache.

First you need to add a new server. The a command does that. It would probably be best to use nnml to read the cache. You could also use nnspool or nnmh, though.

Type a nnml RET cache RET.

You should now have a brand new nnml virtual server called ‘cache’. You now need to edit it to have the right definitions. Type e to edit the server. You’ll be entered into a buffer that will contain the following:

(nnml "cache")

Change that to:

(nnml "cache"
         (nnml-directory "~/News/cache/")
         (nnml-active-file "~/News/cache/active"))

Type C-c C-c to return to the server buffer. If you now press RET over this virtual server, you should be entered into a browse buffer, and you should be able to enter any of the groups displayed.

7.1.5 Server Variables

One sticky point when defining variables (both on back ends and in Emacs in general) is that some variables are typically initialized from other variables when the definition of the variables is being loaded. If you change the “base” variable after the variables have been loaded, you won’t change the “derived” variables.

This typically affects directory and file variables. For instance, nnml-directory is ~/Mail/ by default, and all nnml directory variables are initialized from that variable, so nnml-active-file will be ~/Mail/active. If you define a new virtual nnml server, it will not suffice to set just nnml-directory—you have to explicitly set all the file variables to be what you want them to be. For a complete list of variables for each back end, see each back end’s section later in this manual, but here’s an example nnml definition:

(nnml "public"
      (nnml-directory "~/my-mail/")
      (nnml-active-file "~/my-mail/active")
      (nnml-newsgroups-file "~/my-mail/newsgroups"))

Server variables are often called server parameters.

7.1.6 Servers and Methods

Wherever you would normally use a select method (e.g., gnus-secondary-select-method, in the group select method, when browsing a foreign server) you can use a virtual server name instead. This could potentially save lots of typing. And it’s nice all over.

7.1.7 Unavailable Servers

If a server seems to be unreachable, Gnus will mark that server as denied. That means that any subsequent attempt to make contact with that server will just be ignored. “It can’t be opened,” Gnus will tell you, without making the least effort to see whether that is actually the case or not.

That might seem quite naughty, but it does make sense most of the time. Let’s say you have 10 groups subscribed to on server ‘’. This server is located somewhere quite far away from you and the machine is quite slow, so it takes 1 minute just to find out that it refuses connection to you today. If Gnus were to attempt to do that 10 times, you’d be quite annoyed, so Gnus won’t attempt to do that. Once it has gotten a single “connection refused”, it will regard that server as “down”.

So, what happens if the machine was only feeling unwell temporarily? How do you test to see whether the machine has come up again?

You jump to the server buffer (see Server Buffer) and poke it with the following commands:


Try to establish connection to the server on the current line (gnus-server-open-server).


Close the connection (if any) to the server (gnus-server-close-server).


Mark the current server as unreachable (gnus-server-deny-server). This will effectively disable the server.


Open the connections to all servers in the buffer (gnus-server-open-all-servers).


Close the connections to all servers in the buffer (gnus-server-close-all-servers).


Remove all marks to whether Gnus was denied connection from any servers (gnus-server-remove-denials).


Copy a server and give it a new name (gnus-server-copy-server). This can be useful if you have a complex method definition, and want to use the same definition towards a different (physical) server.


Set server status to offline (gnus-server-offline-server).

7.2 Getting News

A newsreader is normally used for reading news. Gnus currently provides only two methods of getting news—it can read from an NNTP server, or it can read from a local spool.

7.2.1 NNTP

Subscribing to a foreign group from an NNTP server is rather easy. You just specify nntp as method and the address of the NNTP server as the, uhm, address.

If the NNTP server is located at a non-standard port, setting the third element of the select method to this port number should allow you to connect to the right port. You’ll have to edit the group info for that (see Foreign Groups).

The name of the foreign group can be the same as a native group. In fact, you can subscribe to the same group from as many different servers you feel like. There will be no name collisions.

The following variables can be used to create a virtual nntp server:


is run after a connection has been made. It can be used to send commands to the NNTP server after it has been contacted. By default it sends the command MODE READER to the server with the nntp-send-mode-reader function. This function should always be present in this hook.


This function will be used to send ‘AUTHINFO’ to the NNTP server. The default function is nntp-send-authinfo, which looks through your ~/.authinfo for applicable entries. If none are found, it will prompt you for a login name and a password. The format of the ~/.authinfo file is (almost) the same as the ftp ~/.netrc file, which is defined in the ftp manual page, but here are the salient facts:

  1. The file contains one or more line, each of which define one server.
  2. Each line may contain an arbitrary number of token/value pairs.

    The valid tokens include ‘machine’, ‘login’, ‘password’, ‘default’. In addition Gnus introduces two new tokens, not present in the original .netrc/ftp syntax, namely ‘port’ and ‘force’. (This is the only way the .authinfo file format deviates from the .netrc file format.) ‘port’ is used to indicate what port on the server the credentials apply to and ‘force’ is explained below.

Here’s an example file:

machine login larsi password geheimnis
machine login larsi force yes

The token/value pairs may appear in any order; ‘machine’ doesn’t have to be first, for instance.

In this example, both login name and password have been supplied for the former server, while the latter has only the login name listed, and the user will be prompted for the password. The latter also has the ‘force’ tag, which means that the authinfo will be sent to the nntp server upon connection; the default (i.e., when there is not ‘force’ tag) is to not send authinfo to the nntp server until the nntp server asks for it.

You can also add ‘default’ lines that will apply to all servers that don’t have matching ‘machine’ lines.

default force yes

This will force sending ‘AUTHINFO’ commands to all servers not previously mentioned.

Remember to not leave the ~/.authinfo file world-readable.


This is a list of regexps to match on server types and actions to be taken when matches are made. For instance, if you want Gnus to beep every time you connect to innd, you could say something like:

(setq nntp-server-action-alist
      '(("innd" (ding))))

You probably don’t want to do that, though.

The default value is

'(("nntpd 1\\.5\\.11t"
   (remove-hook 'nntp-server-opened-hook

This ensures that Gnus doesn’t send the MODE READER command to nntpd 1.5.11t, since that command chokes that server, I’ve been told.


If the NNTP server doesn’t support NOV headers, this back end will collect headers by sending a series of head commands. To speed things up, the back end sends lots of these commands without waiting for reply, and then reads all the replies. This is controlled by the nntp-maximum-request variable, and is 400 by default. If your network is buggy, you should set this to 1.


If you have lots of foreign nntp groups that you connect to regularly, you’re sure to have problems with NNTP servers not responding properly, or being too loaded to reply within reasonable time. This is can lead to awkward problems, which can be helped somewhat by setting nntp-connection-timeout. This is an integer that says how many seconds the nntp back end should wait for a connection before giving up. If it is nil, which is the default, no timeouts are done.


If the NNTP server does not support NOV, you could set this variable to t, but nntp usually checks automatically whether NOV can be used.


List of strings used as commands to fetch NOV lines from a server. The default value of this variable is ("XOVER" "XOVERVIEW").


nntp normally sends just one big request for NOV lines to the server. The server responds with one huge list of lines. However, if you have read articles 2–5000 in the group, and only want to read article 1 and 5001, that means that nntp will fetch 4999 NOV lines that you will not need. This variable says how big a gap between two consecutive articles is allowed to be before the XOVER request is split into several request. Note that if your network is fast, setting this variable to a really small number means that fetching will probably be slower. If this variable is nil, nntp will never split requests. The default is 5.


When Gnus refers to an article having the Message-ID that a user specifies or having the Message-ID of the parent article of the current one (see Finding the Parent), Gnus sends a HEAD command to the NNTP server to know where it is, and the server returns the data containing the pairs of a group and an article number in the Xref header. Gnus normally uses the article number to refer to the article if the data shows that that article is in the current group, while it uses the Message-ID otherwise. However, some news servers, e.g., ones running Diablo, run multiple engines having the same articles but article numbers are not kept synchronized between them. In that case, the article number that appears in the Xref header varies by which engine is chosen, so you cannot refer to the parent article that is in the current group, for instance. If you connect to such a server, set this variable to a non-nil value, and Gnus never uses article numbers. For example:

(setq gnus-select-method
      '(nntp "newszilla"
             (nntp-address "")
             (nntp-xref-number-is-evil t)

The default value of this server variable is nil.


A hook run before attempting to connect to an NNTP server.


If non-nil, nntp will log all commands it sends to the NNTP server (along with a timestamp) in the *nntp-log* buffer. This is useful if you are debugging a Gnus/NNTP connection that doesn’t seem to work.


It is possible to customize how the connection to the nntp server will be opened. If you specify an nntp-open-connection-function parameter, Gnus will use that function to establish the connection. Seven pre-made functions are supplied. These functions can be grouped in two categories: direct connection functions (four pre-made), and indirect ones (three pre-made).


Non-nil means the nntp server never echoes commands. It is reported that some nntps server doesn’t echo commands. So, you may want to set this to non-nil in the method for such a server setting nntp-open-connection-function to nntp-open-ssl-stream for example. The default value is nil. Note that the nntp-open-connection-functions-never-echo-commands variable overrides the nil value of this variable.


List of functions that never echo commands. Add or set a function which you set to nntp-open-connection-function to this list if it does not echo commands. Note that a non-nil value of the nntp-never-echoes-commands variable overrides this variable. The default value is (nntp-open-network-stream).


A hook run just before posting an article. If there is no Message-ID header in the article and the news server provides the recommended ID, it will be added to the article before running this hook. It is useful to make Cancel-Lock headers even if you inhibit Gnus to add a Message-ID header, you could say:

(add-hook 'nntp-prepare-post-hook 'canlock-insert-header)

Note that not all servers support the recommended ID. This works for INN versions 2.3.0 and later, for instance.


If nil, then always use ‘GROUP’ instead of ‘LIST ACTIVE’. This is usually slower, but on misconfigured servers that don’t update their active files often, this can help. Direct Functions

These functions are called direct because they open a direct connection between your machine and the NNTP server. The behavior of these functions is also affected by commonly understood variables (see Common Variables).


This is the default, and simply connects to some port or other on the remote system. If both Emacs and the server supports it, the connection will be upgraded to an encrypted STARTTLS connection automatically.


The same as the above, but don’t do automatic STARTTLS upgrades.


Opens a connection to a server over a secure channel. To use this you must have GnuTLS installed. You then define a server as follows:

;; "nntps" is port 563 and is predefined in our /etc/services
;; however, ‘gnutls-cli -p’ doesn’t like named ports.
(nntp ""
      (nntp-open-connection-function nntp-open-tls-stream)
      (nntp-port-number 563)
      (nntp-address ""))

Opens a connection to a server over a secure channel. To use this you must have OpenSSL installed. You then define a server as follows:

;; "snews" is port 563 and is predefined in our /etc/services
;; however, ‘openssl s_client -port’ doesn’t like named ports.
(nntp ""
      (nntp-open-connection-function nntp-open-ssl-stream)
      (nntp-port-number 563)
      (nntp-address ""))

Opens a connection to an NNTP server using the netcat program. You might wonder why this function exists, since we have the default nntp-open-network-stream which would do the job. (One of) the reason(s) is that if you are behind a firewall but have direct connections to the outside world thanks to a command wrapper like runsocks, you can use it like this:

(nntp "socksified"
      (nntp-pre-command "runsocks")
      (nntp-open-connection-function nntp-open-netcat-stream)
      (nntp-address ""))

With the default method, you would need to wrap your whole Emacs session, which is not a good idea.


Like nntp-open-netcat-stream, but uses telnet rather than netcat. telnet is a bit less robust because of things like line-end-conversion, but sometimes netcat is simply not available. The previous example would turn into:

(nntp "socksified"
      (nntp-pre-command "runsocks")
      (nntp-open-connection-function nntp-open-telnet-stream)
      (nntp-address "")
      (nntp-end-of-line "\n")) Indirect Functions

These functions are called indirect because they connect to an intermediate host before actually connecting to the NNTP server. All of these functions and related variables are also said to belong to the “via” family of connection: they’re all prefixed with “via” to make things cleaner. The behavior of these functions is also affected by commonly understood variables (see Common Variables).


Does an ‘rlogin’ on a remote system, and then uses netcat to connect to the real NNTP server from there. This is useful for instance if you need to connect to a firewall machine first.

nntp-open-via-rlogin-and-netcat-specific variables:


Command used to log in on the intermediate host. The default is ‘rsh’, but ‘ssh’ is a popular alternative.


List of strings to be used as the switches to nntp-via-rlogin-command. The default is nil. If you use ‘ssh’ for nntp-via-rlogin-command, you may set this to ‘("-C")’ in order to compress all data connections.


Does essentially the same, but uses telnet instead of ‘netcat’ to connect to the real NNTP server from the intermediate host. telnet is a bit less robust because of things like line-end-conversion, but sometimes netcat is simply not available.

nntp-open-via-rlogin-and-telnet-specific variables:


Command used to connect to the real NNTP server from the intermediate host. The default is ‘telnet’.


List of strings to be used as the switches to the nntp-telnet-command command. The default is ("-8").


Command used to log in on the intermediate host. The default is ‘rsh’, but ‘ssh’ is a popular alternative.


List of strings to be used as the switches to nntp-via-rlogin-command. If you use ‘ssh’, you may need to set this to ‘("-t" "-e" "none")’ or ‘("-C" "-t" "-e" "none")’ if the telnet command requires a pseudo-tty allocation on an intermediate host. The default is nil.

Note that you may want to change the value for nntp-end-of-line to ‘\n’ (see Common Variables).


Does essentially the same, but uses ‘telnet’ instead of ‘rlogin’ to connect to the intermediate host.

nntp-open-via-telnet-and-telnet-specific variables:


Command used to telnet the intermediate host. The default is ‘telnet’.


List of strings to be used as the switches to the nntp-via-telnet-command command. The default is ‘("-8")’.


Password to use when logging in on the intermediate host.


If non-nil, the intermediate telnet session (client and server both) will support the ENVIRON option and not prompt for login name. This works for Solaris telnet, for instance.


Regexp matching the shell prompt on the intermediate host. The default is ‘bash\\|\$ *\r?$\\|> *\r?’.

Note that you may want to change the value for nntp-end-of-line to ‘\n’ (see Common Variables).

Here are some additional variables that are understood by all the above functions:


User name to use when connecting to the intermediate host.


Address of the intermediate host to connect to. Common Variables

The following variables affect the behavior of all, or several of the pre-made connection functions. When not specified, all functions are affected (the values of the following variables will be used as the default if each virtual nntp server doesn’t specify those server variables individually).


A command wrapper to use when connecting through a non native connection function (all except nntp-open-network-stream, nntp-open-tls-stream, and nntp-open-ssl-stream). This is where you would put a ‘SOCKS’ wrapper for instance.


The address of the NNTP server.


Port number to connect to the NNTP server. The default is ‘nntp’. If you use NNTP over TLS/SSL, you may want to use integer ports rather than named ports (i.e., use ‘563’ instead of ‘snews’ or ‘nntps’), because external TLS/SSL tools may not work with named ports.


String to use as end-of-line marker when talking to the NNTP server. This is ‘\r\n’ by default, but should be ‘\n’ when using a non native telnet connection function.


Command to use when connecting to the NNTP server through ‘netcat’. This is not for an intermediate host. This is just for the real NNTP server. The default is ‘nc’.


A list of switches to pass to nntp-netcat-command. The default is ‘()’.

7.2.2 News Spool

Subscribing to a foreign group from the local spool is extremely easy, and might be useful, for instance, to speed up reading groups that contain very big articles—‘’, for instance.

Anyway, you just specify nnspool as the method and "" (or anything else) as the address.

If you have access to a local spool, you should probably use that as the native select method (see Finding the News). It is normally faster than using an nntp select method, but might not be. It depends. You just have to try to find out what’s best at your site.


Program used to post an article.


Parameters given to the inews program when posting an article.


Where nnspool looks for the articles. This is normally /usr/spool/news/.


Where nnspool will look for NOV files. This is normally


Where the news lib dir is (/usr/lib/news/ by default).


The name of the active file.


The name of the group descriptions file.


The name of the news history file.


The name of the active date file.


If non-nil, nnspool won’t try to use any NOV files that it finds.


If non-nil, which is the default, use sed to get the relevant portion from the overview file. If nil, nnspool will load the entire file into a buffer and process it there.

7.3 Using IMAP

The most popular mail backend is probably nnimap, which provides access to IMAP servers. IMAP servers store mail remotely, so the client doesn’t store anything locally. This means that it’s a convenient choice when you’re reading your mail from different locations, or with different user agents.

7.3.1 Connecting to an IMAP Server

Connecting to an IMAP can be very easy. Type B in the group buffer, or (if your primary interest is reading email), say something like:

(setq gnus-select-method
      '(nnimap ""))

You’ll be prompted for a user name and password. If you grow tired of that, then add the following to your ~/.authinfo file:

machine login <username> password <password> port imap

That should basically be it for most users.

7.3.2 Customizing the IMAP Connection

Here’s an example method that’s more complex:

(nnimap ""
        (nnimap-inbox "INBOX")
        (nnimap-split-methods default)
        (nnimap-expunge t)
        (nnimap-stream ssl))

The address of the server, like ‘’.


Username to use for authentication to the IMAP server. This corresponds to the value of the ‘login’ token in your ~/.authinfo file. Set this variable if you want to access multiple accounts from the same IMAP server.


If the server uses a non-standard port, that can be specified here. A typical port would be "imap" or "imaps".


How nnimap should connect to the server. Possible values are:


This is the default, and this first tries the ssl setting, and then tries the network setting.


This uses standard TLS/SSL connections.


Non-encrypted and unsafe straight socket connection, but will upgrade to encrypted STARTTLS if both Emacs and the server supports it.


Encrypted STARTTLS over the normal IMAP port.


If you need to tunnel via other systems to connect to the server, you can use this option, and customize nnimap-shell-program to be what you need.


Non-encrypted and unsafe straight socket connection. STARTTLS will not be used even if it is available.


Some IMAP servers allow anonymous logins. In that case, this should be set to anonymous. If this variable isn’t set, the normal login methods will be used. If you wish to specify a specific login method to be used, you can set this variable to either login (the traditional IMAP login method), plain, cram-md5 or xoauth2. (The latter method requires using the oauth2.el library.)


When to expunge deleted messages. If never, deleted articles are marked with the IMAP \\Delete flag but not automatically expunged. If immediately, deleted articles are immediately expunged (this requires the server to support the UID EXPUNGE command). If on-exit, deleted articles are flagged, and all flagged articles are expunged when the group is closed.

For backwards compatibility, this variable may also be set to t or nil. If the server supports UID EXPUNGE, both t and nil are equivalent to immediately. If the server does not support UID EXPUNGE, nil is equivalent to never, while t will immediately expunge all articles that are currently flagged as deleted (i.e., potentially not only the article that was just deleted).


Virtually all IMAP server support fast streaming of data. If you have problems connecting to the server, try setting this to nil.


If non-nil, fetch partial articles from the server. If set to a string, then it’s interpreted as a regexp, and parts that have matching types will be fetched. For instance, ‘"text/"’ will fetch all textual parts, while leaving the rest on the server.


If non-nil, record all IMAP commands in the ‘"*imap log*"’ buffer.


If non-nil, omit the IMAP namespace prefix in nnimap group names. If your IMAP mailboxes are called something like ‘INBOX’ and ‘INBOX.Lists.emacs’, but you’d like the nnimap group names to be ‘INBOX’ and ‘Lists.emacs’, you should enable this option.


By default, nnimap will send occasional ‘NOOP’ (keepalive) commands to the server, to keep the connection alive. This option governs how often that happens. It is a cons of two integers, representing seconds: first how often to run the keepalive check, and the second how many seconds of user inactivity are required to actually send the command. The default, (900 . 300), means run the check every fifteen minutes and, if the user has been inactive for five minutes, send the keepalive command. Set to nil to disable keepalive commands altogether.

7.3.3 Client-Side IMAP Splitting

Many people prefer to do the sorting/splitting of mail into their mail boxes on the IMAP server. That way they don’t have to download the mail they’re not all that interested in.

If you do want to do client-side mail splitting, then the following variables are relevant:


This is the IMAP mail box that will be scanned for new mail. This can also be a list of mail box names.


Uses the same syntax as nnmail-split-methods (see Splitting Mail), except the symbol default, which means that it should use the value of the nnmail-split-methods variable.


Uses the same syntax as nnmail-split-fancy.


List of flag symbols to ignore when doing splitting. That is, articles that have these flags won’t be considered when splitting. The default is ‘(%Deleted %Seen)’.

By default, the nnimap back end only retrieves the message headers; the option nnimap-split-download-body (which is a regular customization option, not a server variable) tells it to retrieve the message bodies as well. We don’t set this by default because it will slow IMAP down, and that is not an appropriate decision to make on behalf of the user.

Here’s a complete example nnimap backend with a client-side “fancy” splitting method:

(nnimap ""
        (nnimap-inbox "INBOX")
         (| ("MailScanner-SpamCheck" "spam" "spam.detected")
            (to "" "foo")

7.3.4 Support for IMAP Extensions

If you’re using Google’s Gmail, you may want to see your Gmail labels when reading your mail. Gnus can give you this information if you ask for ‘X-GM-LABELS’ in the variable gnus-extra-headers. For example:

(setq gnus-extra-headers
      '(To Newsgroups X-GM-LABELS))

This will result in Gnus storing your labels in message header structures for later use. The content is always a parenthesized (possible empty) list.

7.4 Getting Mail

Reading mail with a newsreader—isn’t that just plain WeIrD? But of course.

7.4.1 Mail in a Newsreader

If you are used to traditional mail readers, but have decided to switch to reading mail with Gnus, you may find yourself experiencing something of a culture shock.

Gnus does not behave like traditional mail readers. If you want to make it behave that way, you can, but it’s an uphill battle.

Gnus, by default, handles all its groups using the same approach. This approach is very newsreaderly—you enter a group, see the new/unread messages, and when you read the messages, they get marked as read, and you don’t see them any more. (Unless you explicitly ask for them.)

In particular, you do not do anything explicitly to delete messages.

Does this mean that all the messages that have been marked as read are deleted? How awful!

But, no, it means that old messages are expired according to some scheme or other. For news messages, the expire process is controlled by the news administrator; for mail, the expire process is controlled by you. The expire process for mail is covered in depth in Expiring Mail.

What many Gnus users find, after using it a while for both news and mail, is that the transport mechanism has very little to do with how they want to treat a message.

Many people subscribe to several mailing lists. These are transported via SMTP, and are therefore mail. But we might go for weeks without answering, or even reading these messages very carefully. We may not need to save them because if we should need to read one again, they are archived somewhere else.

Some people have local news groups which have only a handful of readers. These are transported via NNTP, and are therefore news. But we may need to read and answer a large fraction of the messages very carefully in order to do our work. And there may not be an archive, so we may need to save the interesting messages the same way we would personal mail.

The important distinction turns out to be not the transport mechanism, but other factors such as how interested we are in the subject matter, or how easy it is to retrieve the message if we need to read it again.

Gnus provides many options for sorting mail into “groups” which behave like newsgroups, and for treating each group (whether mail or news) differently.

Some users never get comfortable using the Gnus (ahem) paradigm and wish that Gnus should grow up and be a male, er, mail reader. It is possible to whip Gnus into a more mailreaderly being, but, as said before, it’s not easy. People who prefer proper mail readers should try VM instead, which is an excellent, and proper, mail reader.

I don’t mean to scare anybody off, but I want to make it clear that you may be required to learn a new way of thinking about messages. After you’ve been subjected to The Gnus Way, you will come to love it. I can guarantee it. (At least the guy who sold me the Emacs Subliminal Brain-Washing Functions that I’ve put into Gnus did guarantee it. You Will Be Assimilated. You Love Gnus. You Love The Gnus Mail Way. You Do.)

7.4.2 Getting Started Reading Mail

It’s quite easy to use Gnus to read your new mail. You just plonk the mail back end of your choice into gnus-secondary-select-methods, and things will happen automatically.

For instance, if you want to use nnml (which is a “one file per mail” back end), you could put the following in your ~/.gnus.el file:

(setq gnus-secondary-select-methods '((nnml "")))

Now, the next time you start Gnus, this back end will be queried for new articles, and it will move all the messages in your spool file to its directory, which is ~/Mail/ by default. The new group that will be created (‘mail.misc’) will be subscribed, and you can read it like any other group.

You will probably want to split the mail into several groups, though:

(setq nnmail-split-methods
      '(("junk" "^From:.*Lars Ingebrigtsen")
        ("crazy" "^Subject:.*die\\|^Organization:.*flabby")
        ("other" "")))

This will result in three new nnml mail groups being created: ‘nnml:junk’, ‘nnml:crazy’, and ‘nnml:other’. All the mail that doesn’t fit into the first two groups will be placed in the last group.

This should be sufficient for reading mail with Gnus. You might want to give the other sections in this part of the manual a perusal, though. Especially see Choosing a Mail Back End and see Expiring Mail.

7.4.3 Splitting Mail

The nnmail-split-methods variable says how the incoming mail is to be split into groups.

(setq nnmail-split-methods
  '(("mail.junk" "^From:.*Lars Ingebrigtsen")
    ("mail.crazy" "^Subject:.*die\\|^Organization:.*flabby")
    ("mail.other" "")))

This variable is a list of lists, where the first element of each of these lists is the name of the mail group (they do not have to be called something beginning with ‘mail’, by the way), and the second element is a regular expression used on the header of each mail to determine if it belongs in this mail group. The first string may contain ‘\\1’ forms, like the ones used by replace-match to insert sub-expressions from the matched text. For instance:

("list.\\1" "From:.* \\(.*\\)")

In that case, nnmail-split-lowercase-expanded controls whether the inserted text should be made lowercase. See Fancy Mail Splitting.

The second element can also be a function. In that case, it will be called narrowed to the headers with the first element of the rule as the argument. It should return a non-nil value if it thinks that the mail belongs in that group.

The last of these groups should always be a general one, and the regular expression should always be ‘""’ so that it matches any mails that haven’t been matched by any of the other regexps. (These rules are processed from the beginning of the alist toward the end.

If multiple rules match (excluding the general ‘""’ group), mail is crossposted to all these groups. However, if nnmail-crosspost is set to nil, the first rule to make a match will “win”.

If no rule matched, the mail will end up in the ‘bogus’ group. When new groups are created by splitting mail, you may want to run gnus-group-find-new-groups to see the new groups. This also applies to the ‘bogus’ group.

If you like to tinker with this yourself, you can set this variable to a function of your choice. This function will be called without any arguments in a buffer narrowed to the headers of an incoming mail message. The function should return a list of group names that it thinks should carry this mail message.

This variable can also be a fancy split method. For the syntax, see Fancy Mail Splitting.

Note that the mail back ends are free to maul the poor, innocent, incoming headers all they want to. They all add Lines headers; some add X-Gnus-Group headers; most rename the Unix mbox FromSPC line to something else.

The mail back ends all support cross-posting. If several regexps match, the mail will be “cross-posted” to all those groups. nnmail-crosspost says whether to use this mechanism or not. Note that no articles are crossposted to the general (‘""’) group.

nnmh and nnml makes crossposts by creating hard links to the crossposted articles. However, not all file systems support hard links. If that’s the case for you, set nnmail-crosspost-link-function to copy-file. (This variable is add-name-to-file by default.)

If you wish to see where the previous mail split put the messages, you can use the M-x nnmail-split-history command. If you wish to see where re-spooling messages would put the messages, you can use gnus-summary-respool-trace and related commands (see Mail Group Commands).

Header lines longer than the value of nnmail-split-header-length-limit are excluded from the split function.

By default, splitting does not decode headers, so you can not match on non-ASCII strings. But it is useful if you want to match articles based on the raw header data. To enable it, set the nnmail-mail-splitting-decodes variable to a non-nil value. In addition, the value of the nnmail-mail-splitting-charset variable is used for decoding non-MIME encoded string when nnmail-mail-splitting-decodes is non-nil. The default value is nil which means not to decode non-MIME encoded string. A suitable value for you will be undecided or be the charset used normally in mails you are interested in.

By default, splitting is performed on all incoming messages. If you specify a directory entry for the variable mail-sources (see Mail Source Specifiers), however, then splitting does not happen by default. You can set the variable nnmail-resplit-incoming to a non-nil value to make splitting happen even in this case. (This variable has no effect on other kinds of entries.)

Gnus gives you all the opportunity you could possibly want for shooting yourself in the foot. Let’s say you create a group that will contain all the mail you get from your boss. And then you accidentally unsubscribe from the group. Gnus will still put all the mail from your boss in the unsubscribed group, and so, when your boss mails you “Have that report ready by Monday or you’re fired!”, you’ll never see it and, come Tuesday, you’ll still believe that you’re gainfully employed while you really should be out collecting empty bottles to save up for next month’s rent money.

7.4.4 Mail Sources

Mail can be gotten from many different sources—the mail spool, from a POP mail server, from a procmail directory, or from a maildir, for instance. Mail Source Specifiers

You tell Gnus how to fetch mail by setting mail-sources (see Fetching Mail) to a mail source specifier.

Here’s an example:

(pop :server "" :user "myname")

As can be observed, a mail source specifier is a list where the first element is a mail source type, followed by an arbitrary number of keywords. Keywords that are not explicitly specified are given default values.

The mail-sources is global for all mail groups. You can specify an additional mail source for a particular group by including the group mail specifier in mail-sources, and setting a mail-source group parameter (see Group Parameters) specifying a single mail source. When this is used, mail-sources is typically just ((group)); the mail-source parameter for a group might look like this:

(mail-source . (file :path "home/user/spools/foo.spool"))

This means that the group’s (and only this group’s) messages will be fetched from the spool file ‘/user/spools/foo.spool’.

The following mail source types are available:


Get mail from a single file; typically from the mail spool.



The file name. Defaults to the value of the MAIL environment variable or the value of rmail-spool-directory (usually something like /usr/mail/spool/user-name).


Script run before/after fetching mail.

An example file mail source:

(file :path "/usr/spool/mail/user-name")

Or using the default file name:


If the mail spool file is not located on the local machine, it’s best to use POP or IMAP or the like to fetch the mail. You can not use ange-ftp file names here—it has no way to lock the mail spool while moving the mail.

If it’s impossible to set up a proper server, you can use ssh instead.

(setq mail-sources
      '((file :prescript "ssh host bin/getmail >%t")))

The ‘getmail’ script would look something like the following:

#  getmail - move mail from spool to stdout

rm -f $TMP; $MOVEMAIL $MAIL $TMP >/dev/null && cat $TMP

Alter this script to fit the ‘movemail’ and temporary file you want to use.


Get mail from several files in a directory. This is typically used when you have procmail split the incoming mail into several files. That is, there is a one-to-one correspondence between files in that directory and groups, so that mail from the file will be put in the group (You can change the suffix to be used instead of .spool.) Setting nnmail-scan-directory-mail-source-once to non-nil forces Gnus to scan the mail source only once. This is particularly useful if you want to scan mail groups at a specified level.

There is also the variable nnmail-resplit-incoming, if you set that to a non-nil value, then the normal splitting process is applied to all the files from the directory, Splitting Mail.



The name of the directory where the files are. There is no default value.


Only files ending with this suffix are used. The default is ‘.spool’.


Only files that have this predicate return non-nil are returned. The default is identity. This is used as an additional filter—only files that have the right suffix and satisfy this predicate are considered.


Script run before/after fetching mail.

An example directory mail source:

(directory :path "/home/user-name/procmail-dir/"
           :suffix ".prcml")

Get mail from a POP server.



The name of the POP server. The default is taken from the MAILHOST environment variable.


The port number of the POP server. This can be a number (e.g., ‘:port 1234’) or a string (e.g., ‘:port "pop3"’). If it is a string, it should be a service name as listed in /etc/services on Unix systems. The default is ‘"pop3"’. On some systems you might need to specify it as ‘"pop-3"’ instead.


The user name to give to the POP server. The default is the login name.


The password to give to the POP server. If not specified, the user is prompted.


The program to use to fetch mail from the POP server. This should be a format-like string. Here’s an example:

fetchmail %u@%s -P %p %t

The valid format specifier characters are:


The name of the file the mail is to be moved to. This must always be included in this string.


The name of the server.


The port number of the server.


The user name to use.


The password to use.

The values used for these specs are taken from the values you give the corresponding keywords.


A script to be run before fetching the mail. The syntax is the same as the :program keyword. This can also be a function to be run.

One popular way to use this is to set up an SSH tunnel to access the POP server. Here’s an example:

(pop :server ""
     :port 1234
     :user "foo"
     :password "secret"
     "nohup ssh -f -L 1234:pop.server:110 sleep 3600 &")

A script to be run after fetching the mail. The syntax is the same as the :program keyword. This can also be a function to be run.


The function to use to fetch mail from the POP server. The function is called with one parameter—the name of the file where the mail should be moved to.


This can be either the symbol password or the symbol apop and says what authentication scheme to use. The default is password.


Non-nil if the mail is to be left on the POP server after fetching. Only the built-in pop3-movemail program (the default) supports this keyword.

If this is a number, leave mails on the server for this many days since you first checked new mails. In that case, mails once fetched will never be fetched again by the UIDL control. If this is nil (the default), mails will be deleted on the server right after fetching. If this is neither nil nor a number, all mails will be left on the server, and you will end up getting the same mails again and again.

The pop3-uidl-file variable specifies the file to which the UIDL data are locally stored. The default value is ~/.pop3-uidl.

Note that POP servers maintain no state information between sessions, so what the client believes is there and what is actually there may not match up. If they do not, then you may get duplicate mails or the whole thing can fall apart and leave you with a corrupt mailbox.

If the :program and :function keywords aren’t specified, pop3-movemail will be used.

Here are some examples for getting mail from a POP server.

Fetch from the default POP server, using the default user name, and default fetcher:


Fetch from a named server with a named user and password:

(pop :server "my.pop.server"
     :user "user-name" :password "secret")

Leave mails on the server for 14 days:

(pop :server "my.pop.server"
     :user "user-name" :password "secret"
     :leave 14)

Use ‘movemail’ to move the mail:

(pop :program "movemail po:%u %t %p")

Get mail from a maildir. This is a type of mailbox that is supported by at least qmail and postfix, where each file in a special directory contains exactly one mail.



The name of the directory where the mails are stored. The default is taken from the MAILDIR environment variable or ~/Maildir/.


The subdirectories of the Maildir. The default is ‘("new" "cur")’.

You can also get mails from remote hosts (because maildirs don’t suffer from locking problems).

Two example maildir mail sources:

(maildir :path "/home/user-name/Maildir/"
         :subdirs ("cur" "new"))
(maildir :path "/"
         :subdirs ("new"))

Get mail from an IMAP server. If you don’t want to use IMAP as intended, as a network mail reading protocol (i.e., with nnimap), for some reason or other, Gnus lets you treat it similar to a POP server and fetches articles from a given IMAP mailbox. See Using IMAP, for more information.



The name of the IMAP server. The default is taken from the MAILHOST environment variable.


The port number of the IMAP server. The default is ‘143’, or ‘993’ for TLS/SSL connections.


The user name to give to the IMAP server. The default is the login name.


The password to give to the IMAP server. If not specified, the user is prompted.


What stream to use for connecting to the server, this is one of the symbols in imap-stream-alist. Right now, this means ‘gssapi’, ‘kerberos4’, ‘starttls’, ‘tls’, ‘ssl’, ‘shell’ or the default ‘network’.


Which authenticator to use for authenticating to the server, this is one of the symbols in imap-authenticator-alist. Right now, this means ‘gssapi’, ‘kerberos4’, ‘digest-md5’, ‘cram-md5’, ‘anonymous’ or the default ‘login’.


When using the ‘shell’ :stream, the contents of this variable is mapped into the imap-shell-program variable. This should be a format-like string (or list of strings). Here’s an example:

ssh %s imapd

Make sure nothing is interfering with the output of the program, e.g., don’t forget to redirect the error output to the void. The valid format specifier characters are:


The name of the server.


User name from imap-default-user.


The port number of the server.

The values used for these specs are taken from the values you give the corresponding keywords.


The name of the mailbox to get mail from. The default is ‘INBOX’ which normally is the mailbox which receives incoming mail. Instead of a single mailbox, this can be a list of mailboxes to fetch mail from.


The predicate used to find articles to fetch. The default, ‘UNSEEN UNDELETED’, is probably the best choice for most people, but if you sometimes peek in your mailbox with a IMAP client and mark some articles as read (or; SEEN) you might want to set this to ‘1:*’. Then all articles in the mailbox is fetched, no matter what. For a complete list of predicates, see RFC 2060 section 6.4.4.


How to flag fetched articles on the server, the default ‘\Deleted’ will mark them as deleted, an alternative would be ‘\Seen’ which would simply mark them as read. These are the two most likely choices, but more flags are defined in RFC 2060 section 2.3.2.


If non-nil, don’t remove all articles marked as deleted in the mailbox after finishing the fetch.

An example IMAP mail source:

(imap :server ""
      :stream kerberos4
      :fetchflag "\\Seen")

Get the actual mail source from the mail-source group parameter, See Group Parameters.

Common Keywords

Common keywords can be used in any type of mail source.



If non-nil, fetch the mail even when Gnus is unplugged. If you use directory source to get mail, you can specify it as in this example:

(setq mail-sources
      '((directory :path "/home/pavel/.Spool/"
                   :suffix ""
                   :plugged t)))

Gnus will then fetch your mail even when you are unplugged. This is useful when you use local mail and news. Function Interface

Some of the above keywords specify a Lisp function to be executed. For each keyword :foo, the Lisp variable foo is bound to the value of the keyword while the function is executing. For example, consider the following mail-source setting:

(setq mail-sources '((pop :user "jrl"
                          :server "pophost" :function fetchfunc)))

While the function fetchfunc is executing, the symbol user is bound to "jrl", and the symbol server is bound to "pophost". The symbols port, password, program, prescript, postscript, function, and authentication are also bound (to their default values).

See above for a list of keywords for each type of mail source. Mail Source Customization

The following is a list of variables that influence how the mail is fetched. You would normally not need to set or change any of these variables.


File where mail will be stored while processing it. The default is


If non-nil, delete incoming files after handling them. If t, delete the files immediately, if nil, never delete any files. If a positive number, delete files older than number of days (the deletion will only happen when receiving new mail). You may also set mail-source-delete-incoming to nil and call mail-source-delete-old-incoming from a hook or interactively.


If non-nil, ask for confirmation before deleting old incoming files. This variable only applies when mail-source-delete-incoming is a positive number.


Directory where incoming mail source files (if any) will be stored. The default is ~/Mail/. At present, the only thing this is used for is to say where the incoming files will be stored if the variable mail-source-delete-incoming is nil or a number.


Prefix for file name for storing incoming mail. The default is Incoming, in which case files will end up with names like Incoming30630D_ or Incoming298602ZD. This is really only relevant if mail-source-delete-incoming is nil or a number.


All new mail files will get this file mode. The default is #o600.


If non-nil, name of program for fetching new mail. If nil, movemail in exec-directory. Fetching Mail

The way to actually tell Gnus where to get new mail from is to set mail-sources to a list of mail source specifiers (see Mail Source Specifiers).

If this variable is nil, the mail back ends will never attempt to fetch mail by themselves.

If you want to fetch mail both from your local spool as well as a POP mail server, you’d say something like:

(setq mail-sources
        (pop :server "pop3.mail.server"
             :password "secret")))

Or, if you don’t want to use any of the keyword defaults:

(setq mail-sources
      '((file :path "/var/spool/mail/user-name")
        (pop :server "pop3.mail.server"
             :user "user-name"
             :port "pop3"
             :password "secret")))

When you use a mail back end, Gnus will slurp all your mail from your inbox and plonk it down in your home directory. Gnus doesn’t move any mail if you’re not using a mail back end—you have to do a lot of magic invocations first. At the time when you have finished drawing the pentagram, lightened the candles, and sacrificed the goat, you really shouldn’t be too surprised when Gnus moves your mail.

7.4.5 Mail Back End Variables

These variables are (for the most part) pertinent to all the various mail back ends.


The mail back ends all call this hook after reading new mail. You can use this hook to notify any mail watch programs, if you want to.


Hook run in the buffer where the mail headers of each message is kept just before the splitting based on these headers is done. The hook is free to modify the buffer contents in any way it sees fit—the buffer is discarded after the splitting has been done, and no changes performed in the buffer will show up in any files. gnus-article-decode-encoded-words is one likely function to add to this hook.


These are two useful hooks executed when treating new incoming mail—nnmail-pre-get-new-mail-hook (is called just before starting to handle the new mail) and nnmail-post-get-new-mail-hook (is called when the mail handling is done). Here’s and example of using these two hooks to change the default file modes the new mail files get:

(add-hook 'nnmail-pre-get-new-mail-hook
          (lambda () (set-default-file-modes #o700)))

(add-hook 'nnmail-post-get-new-mail-hook
          (lambda () (set-default-file-modes #o775)))

If non-nil, the mail back ends will use long file and directory names. Groups like ‘mail.misc’ will end up in directories (assuming use of nnml back end) or files (assuming use of nnfolder back end) like mail.misc. If it is nil, the same group will end up in mail/misc.


Function called to delete files. It is delete-file by default.


If non-nil, put the Message-IDs of articles imported into the back end (via Gcc, for instance) into the mail duplication discovery cache. The default is nil.


This can be a regular expression or a list of regular expressions. Group names that match any of the regular expressions will never be recorded in the Message-ID cache.

This can be useful, for example, when using Fancy Splitting (see Fancy Mail Splitting) together with the function nnmail-split-fancy-with-parent.

7.4.6 Fancy Mail Splitting

If the rather simple, standard method for specifying how to split mail doesn’t allow you to do what you want, you can set nnmail-split-methods to nnmail-split-fancy. Then you can play with the nnmail-split-fancy variable.

Let’s look at an example value of this variable first:

;; Messages from the mailer daemon are not crossposted to any of
;; the ordinary groups.  Warnings are put in a separate group
;; from real errors.
(| ("from" mail (| ("subject" "warn.*" "mail.warning")
   ;; Non-error messages are crossposted to all relevant
   ;; groups, but we don’t crosspost between the group for the
   ;; (ding) list and the group for other (ding) related mail.
   (& (| (any "ding@ifi\\.uio\\.no" "ding.list")
         ("subject" "ding" "ding.misc"))
      ;; Other mailing lists…
      (any "procmail@informatik\\.rwth-aachen\\.de" "procmail.list")
      (any "SmartList@informatik\\.rwth-aachen\\.de" "SmartList.list")
      ;; Both lists below have the same suffix, so prevent
      ;; cross-posting to mkpkg.list of messages posted only to
      ;; the bugs- list, but allow cross-posting when the
      ;; message was really cross-posted.
      (any "bugs-mypackage@somewhere" "mypkg.bugs")
      (any "mypackage@somewhere" - "bugs-mypackage" "mypkg.list")
      ;; People…
      (any "larsi@ifi\\.uio\\.no" "people.Lars_Magne_Ingebrigtsen"))
   ;; Unmatched mail goes to the catch all group.

This variable has the format of a split. A split is a (possibly) recursive structure where each split may contain other splits. Here are the possible split syntaxes:


If the split is a string, that will be taken as a group name. Normal regexp match expansion will be done. See below for examples.

(field value [- restrict […] ] split [invert-partial])

The split can be a list containing at least three elements. If the first element field (a regexp matching a header) contains value (also a regexp) then store the message as specified by split.

If restrict (yet another regexp) matches some string after field and before the end of the matched value, the split is ignored. If none of the restrict clauses match, split is processed.

The last element invert-partial is optional. If it is non-nil, the match-partial-words behavior controlled by the variable nnmail-split-fancy-match-partial-words (see below) is be inverted. (New in Gnus 5.10.7)

(| split …)

If the split is a list, and the first element is | (vertical bar), then process each split until one of them matches. A split is said to match if it will cause the mail message to be stored in one or more groups.

(& split …)

If the split is a list, and the first element is &, then process all splits in the list.


If the split is the symbol junk, then don’t save (i.e., delete) this message. Use with extreme caution.

(: function arg1 arg2 …)

If the split is a list, and the first element is ‘:’, then the second element will be called as a function with args given as arguments. The function should return a split.

For instance, the following function could be used to split based on the body of the messages:

(defun split-on-body ()
      (goto-char (point-min))
      (when (re-search-forward "Some.*string" nil t)

The buffer is narrowed to the header of the message in question when function is run. That’s why (widen) needs to be called after save-excursion and save-restriction in the example above. Also note that with the nnimap backend, message bodies will not be downloaded by default. You need to set nnimap-split-download-body to t to do that (see Client-Side IMAP Splitting).

(! func split)

If the split is a list, and the first element is !, then split will be processed, and func will be called as a function with the result of split as argument. func should return a split.


If the split is nil, it is ignored.

In these splits, field must match a complete field name.

Normally, value in these splits must match a complete word according to the fundamental mode syntax table. In other words, all value’s will be implicitly surrounded by \<...\> markers, which are word delimiters. Therefore, if you use the following split, for example,

(any "joe" "joemail")

messages sent from ‘’ will normally not be filed in ‘joemail’. If you want to alter this behavior, you can use any of the following three ways:

  1. You can set the nnmail-split-fancy-match-partial-words variable to non-nil in order to ignore word boundaries and instead the match becomes more like a grep. This variable controls whether partial words are matched during fancy splitting. The default value is nil.

    Note that it influences all value’s in your split rules.

  2. value beginning with .* ignores word boundaries in front of a word. Similarly, if value ends with .*, word boundaries in the rear of a word will be ignored. For example, the value "@example\\.com" does not match ‘’ but ".*@example\\.com" does.
  3. You can set the invert-partial flag in your split rules of the ‘(field value …)’ types, aforementioned in this section. If the flag is set, word boundaries on both sides of a word are ignored even if nnmail-split-fancy-match-partial-words is nil. Contrarily, if the flag is set, word boundaries are not ignored even if nnmail-split-fancy-match-partial-words is non-nil. (New in Gnus 5.10.7)

field and value can also be Lisp symbols, in that case they are expanded as specified by the variable nnmail-split-abbrev-alist. This is an alist of cons cells, where the CAR of a cell contains the key, and the CDR contains the associated value. Predefined entries in nnmail-split-abbrev-alist include:


Matches the ‘From’, ‘Sender’ and ‘Resent-From’ fields.


Matches the ‘To’, ‘Cc’, ‘Apparently-To’, ‘Resent-To’ and ‘Resent-Cc’ fields.


Is the union of the from and to entries.


Matches the ‘List-ID’, ‘List-Post’, ‘X-Mailing-List’, ‘X-BeenThere’ and ‘X-Loop’ fields.

nnmail-split-fancy-syntax-table is the syntax table in effect when all this splitting is performed.

If you want to have Gnus create groups dynamically based on some information in the headers (i.e., do replace-match-like substitutions in the group names), you can say things like:

(any "debian-\\b\\(\\w+\\)" "mail.debian.\\1")

In this example, messages sent to ‘’ will be filed in ‘’.

If the string contains the element ‘\\&’, then the previously matched string will be substituted. Similarly, the elements ‘\\1’ up to ‘\\9’ will be substituted with the text matched by the groupings 1 through 9.

Where nnmail-split-lowercase-expanded controls whether the lowercase of the matched string should be used for the substitution. Setting it as non-nil is useful to avoid the creation of multiple groups when users send to an address using different case (i.e., mailing-list@domain vs Mailing-List@Domain). The default value is t.

nnmail-split-fancy-with-parent is a function which allows you to split followups into the same groups their parents are in. Sometimes you can’t make splitting rules for all your mail. For example, your boss might send you personal mail regarding different projects you are working on, and as you can’t tell your boss to put a distinguishing string into the subject line, you have to resort to manually moving the messages into the right group. With this function, you only have to do it once per thread.

To use this feature, you have to set nnmail-treat-duplicates and nnmail-cache-accepted-message-ids to a non-nil value. And then you can include nnmail-split-fancy-with-parent using the colon feature, like so:

(setq nnmail-treat-duplicates 'warn     ; or delete
      nnmail-cache-accepted-message-ids t
      '(| (: nnmail-split-fancy-with-parent)
          ;; other splits go here

This feature works as follows: when nnmail-treat-duplicates is non-nil, Gnus records the message id of every message it sees in the file specified by the variable nnmail-message-id-cache-file, together with the group it is in (the group is omitted for non-mail messages). When mail splitting is invoked, the function nnmail-split-fancy-with-parent then looks at the References (and In-Reply-To) header of each message to split and searches the file specified by nnmail-message-id-cache-file for the message ids. When it has found a parent, it returns the corresponding group name unless the group name matches the regexp nnmail-split-fancy-with-parent-ignore-groups. It is recommended that you set nnmail-message-id-cache-length to a somewhat higher number than the default so that the message ids are still in the cache. (A value of 5000 appears to create a file some 300 kBytes in size.) When nnmail-cache-accepted-message-ids is non-nil, Gnus also records the message ids of moved articles, so that the followup messages goes into the new group.

Also see the variable nnmail-cache-ignore-groups if you don’t want certain groups to be recorded in the cache. For example, if all outgoing messages are written to an “outgoing” group, you could set nnmail-cache-ignore-groups to match that group name. Otherwise, answers to all your messages would end up in the “outgoing” group.

If nnmail-debug-splitting is non-nil, the mail splitting code will log all splitting decisions to the ‘*nnmail split*’ buffer.

7.4.7 Group Mail Splitting

If you subscribe to dozens of mailing lists but you don’t want to maintain mail splitting rules manually, group mail splitting is for you. You just have to set to-list and/or to-address in group parameters or group customization and set nnmail-split-methods to gnus-group-split. This splitting function will scan all groups for those parameters and split mail accordingly, i.e., messages posted from or to the addresses specified in the parameters to-list or to-address of a mail group will be stored in that group.

Sometimes, mailing lists have multiple addresses, and you may want mail splitting to recognize them all: just set the extra-aliases group parameter to the list of additional addresses and it’s done. If you’d rather use a regular expression, set split-regexp.

All these parameters in a group will be used to create an nnmail-split-fancy split, in which the field is ‘any’, the value is a single regular expression that matches to-list, to-address, all of extra-aliases and all matches of split-regexp, and the split is the name of the group. restricts are also supported: just set the split-exclude parameter to a list of regular expressions.

If you can’t get the right split to be generated using all these parameters, or you just need something fancier, you can set the parameter split-spec to an nnmail-split-fancy split. In this case, all other aforementioned parameters will be ignored by gnus-group-split. In particular, split-spec may be set to nil, in which case the group will be ignored by gnus-group-split.

gnus-group-split will do cross-posting on all groups that match, by defining a single & fancy split containing one split for each group. If a message doesn’t match any split, it will be stored in the group named in gnus-group-split-default-catch-all-group, unless some group has split-spec set to catch-all, in which case that group is used as the catch-all group. Even though this variable is often used just to name a group, it may also be set to an arbitrarily complex fancy split (after all, a group name is a fancy split), and this may be useful to split mail that doesn’t go to any mailing list to personal mail folders. Note that this fancy split is added as the last element of a | split list that also contains a & split with the rules extracted from group parameters.

It’s time for an example. Assume the following group parameters have been defined:
((to-address . "")
 (split-regexp . ".*@femail\\.com"))
((to-list . "")
 (extra-aliases "foo@localhost" "foo-redist@home")
 (split-exclude "bugs-foo" "rambling-foo")
 (admin-address . ""))
((split-spec . catch-all))

Setting nnmail-split-methods to gnus-group-split will behave as if nnmail-split-fancy had been selected and variable nnmail-split-fancy had been set as follows:

(| (& (any "\\(bar@femail\\.com\\|.*@femail\\.com\\)" "")
      (any "\\(foo@nowhere\\.gov\\|foo@localhost\\|foo-redist@home\\)"
           - "bugs-foo" - "rambling-foo" ""))

If you’d rather not use group splitting for all your mail groups, you may use it for only some of them, by using nnmail-split-fancy splits like this:

(: gnus-group-split-fancy groups no-crosspost catch-all)

groups may be a regular expression or a list of group names whose parameters will be scanned to generate the output split. no-crosspost can be used to disable cross-posting; in this case, a single | split will be output. catch-all is the fall back fancy split, used like gnus-group-split-default-catch-all-group. If catch-all is nil, or if split-regexp matches the empty string in any selected group, no catch-all split will be issued. Otherwise, if some group has split-spec set to catch-all, this group will override the value of the catch-all argument.

Unfortunately, scanning all groups and their parameters can be quite slow, especially considering that it has to be done for every message. But don’t despair! The function gnus-group-split-setup can be used to enable gnus-group-split in a much more efficient way. It sets nnmail-split-methods to nnmail-split-fancy and sets nnmail-split-fancy to the split produced by gnus-group-split-fancy. Thus, the group parameters are only scanned once, no matter how many messages are split.

However, if you change group parameters, you’d have to update nnmail-split-fancy manually. You can do it by running gnus-group-split-update. If you’d rather have it updated automatically, just tell gnus-group-split-setup to do it for you. For example, add to your ~/.gnus.el:

(gnus-group-split-setup auto-update catch-all)

If auto-update is non-nil, gnus-group-split-update will be added to gnus-get-top-new-news-hook, so you won’t ever have to worry about updating nnmail-split-fancy again. If you don’t omit catch-all (it’s optional, equivalent to nil), gnus-group-split-default-catch-all-group will be set to its value.

Because you may want to change nnmail-split-fancy after it is set by gnus-group-split-update, this function will run gnus-group-split-updated-hook just before finishing.

7.4.8 Incorporating Old Mail

Most people have lots of old mail stored in various file formats. If you have set up Gnus to read mail using one of the spiffy Gnus mail back ends, you’ll probably wish to have that old mail incorporated into your mail groups.

Doing so can be quite easy.

To take an example: You’re reading mail using nnml (see Mail Spool), and have set nnmail-split-methods to a satisfactory value (see Splitting Mail). You have an old Unix mbox file filled with important, but old, mail. You want to move it into your nnml groups.

Here’s how:

  1. Go to the group buffer.
  2. Type G f and give the file name to the mbox file when prompted to create an nndoc group from the mbox file (see Foreign Groups).
  3. Type SPC to enter the newly created group.
  4. Type M P b to process-mark all articles in this group’s buffer (see Setting Process Marks).
  5. Type B r to respool all the process-marked articles, and answer ‘nnml’ when prompted (see Mail Group Commands).

All the mail messages in the mbox file will now also be spread out over all your nnml groups. Try entering them and check whether things have gone without a glitch. If things look ok, you may consider deleting the mbox file, but I wouldn’t do that unless I was absolutely sure that all the mail has ended up where it should be.

Respooling is also a handy thing to do if you’re switching from one mail back end to another. Just respool all the mail in the old mail groups using the new mail back end.

7.4.9 Expiring Mail

Traditional mail readers have a tendency to remove mail articles when you mark them as read, in some way. Gnus takes a fundamentally different approach to mail reading.

Gnus basically considers mail just to be news that has been received in a rather peculiar manner. It does not think that it has the power to actually change the mail, or delete any mail messages. If you enter a mail group, and mark articles as “read”, or kill them in some other fashion, the mail articles will still exist on the system. I repeat: Gnus will not delete your old, read mail. Unless you ask it to, of course.

To make Gnus get rid of your unwanted mail, you have to mark the articles as expirable. (With the default key bindings, this means that you have to type E.) This does not mean that the articles will disappear right away, however. In general, a mail article will be deleted from your system if, 1) it is marked as expirable, AND 2) it is more than one week old. If you do not mark an article as expirable, it will remain on your system until hell freezes over. This bears repeating one more time, with some spurious capitalizations: IF you do NOT mark articles as EXPIRABLE, Gnus will NEVER delete those ARTICLES.

You do not have to mark articles as expirable by hand. Gnus provides two features, called “auto-expire” and “total-expire”, that can help you with this. In a nutshell, “auto-expire” means that Gnus hits E for you when you select an article. And “total-expire” means that Gnus considers all articles as expirable that are read. So, in addition to the articles marked ‘E’, also the articles marked ‘r’, ‘R’, ‘O’, ‘K’, ‘Y’ (and so on) are considered expirable. gnus-auto-expirable-marks has the full list of these marks.

When should either auto-expire or total-expire be used? Most people who are subscribed to mailing lists split each list into its own group and then turn on auto-expire or total-expire for those groups. (See Splitting Mail, for more information on splitting each list into its own group.)

Which one is better, auto-expire or total-expire? It’s not easy to answer. Generally speaking, auto-expire is probably faster. Another advantage of auto-expire is that you get more marks to work with: for the articles that are supposed to stick around, you can still choose between tick and dormant and read marks. But with total-expire, you only have dormant and ticked to choose from. The advantage of total-expire is that it works well with adaptive scoring (see Adaptive Scoring). Auto-expire works with normal scoring but not with adaptive scoring.

Groups that match the regular expression gnus-auto-expirable-newsgroups will have all articles that you read marked as expirable automatically. All articles marked as expirable have an ‘E’ in the first column in the summary buffer.

By default, if you have auto expiry switched on, Gnus will mark all the articles you read as expirable, no matter if they were read or unread before. To avoid having articles marked as read marked as expirable automatically, you can put something like the following in your ~/.gnus.el file:

(remove-hook 'gnus-mark-article-hook
(add-hook 'gnus-mark-article-hook 'gnus-summary-mark-unread-as-read)

Note that making a group auto-expirable doesn’t mean that all read articles are expired—only the articles marked as expirable will be expired. Also note that using the d command won’t make articles expirable—only semi-automatic marking of articles as read will mark the articles as expirable in auto-expirable groups.

Let’s say you subscribe to a couple of mailing lists, and you want the articles you have read to disappear after a while:

(setq gnus-auto-expirable-newsgroups

Another way to have auto-expiry happen is to have the element auto-expire in the group parameters of the group.

If you use adaptive scoring (see Adaptive Scoring) and auto-expiring, you’ll have problems. Auto-expiring and adaptive scoring don’t really mix very well.

The nnmail-expiry-wait variable supplies the default time an expirable article has to live. The value of this variable can be either a number of days (not necessarily an integer), or one of the symbols immediate or never, meaning an article is immediately or never expirable, respectively.

Gnus starts counting days from when the message arrived, not from when it was sent. The default is seven days.

The nnmail-expiry-wait-function variable lets you fine-tune how long articles are to live, based on what group they are in. When set to a function, its returned value, if non-nil, overrides that of nnmail-expiry-wait. Otherwise, the value of nnmail-expiry-wait is used instead.

For example, let’s say you want to have a one month expiry period in the ‘mail.private’ group, a one day expiry period in the ‘mail.junk’ group, and a six day expiry period everywhere else. This can be achieved as follows:

(setq nnmail-expiry-wait-function
      (lambda (group)
        (cond ((string= group "mail.private")
              ((string= group "mail.junk")
              ((string= group "important")

The group names this function is fed are “unadorned” group names—no ‘nnml:’ prefixes and the like.

As an alternative to the variables nnmail-expiry-wait or nnmail-expiry-wait-function, you can also use the expiry-wait group parameter to selectively change the expiry period (see Group Parameters).

The normal action taken when expiring articles is to delete them. However, in some circumstances it might make more sense to move them to other groups instead of deleting them. The variable nnmail-expiry-target (and the expiry-target group parameter) controls this. The variable supplies a default value for all groups, which can be overridden for specific groups by the group parameter. default value is delete, but this can also be a string (which should be the name of the group the message should be moved to), or a function (which will be called in a buffer narrowed to the message in question, and with the name of the group being moved from as its parameter) which should return a target—either a group name or delete.

Here’s an example for specifying a group name:

(setq nnmail-expiry-target "nnml:expired")

Gnus provides a function nnmail-fancy-expiry-target which will expire mail to groups according to the variable nnmail-fancy-expiry-targets. Here’s an example:

 (setq nnmail-expiry-target 'nnmail-fancy-expiry-target
       '((to-from "boss" "nnfolder:Work")
         ("subject" "IMPORTANT" "nnfolder:IMPORTANT.%Y.%b")
         ("from" ".*" "nnfolder:Archive-%Y")))

With this setup, any mail that has IMPORTANT in its Subject header and was sent in the year YYYY and month MMM, will get expired to the group nnfolder:IMPORTANT.YYYY.MMM. If its From or To header contains the string boss, it will get expired to nnfolder:Work. All other mail will get expired to nnfolder:Archive-YYYY.

If nnmail-keep-last-article is non-nil, Gnus will never expire the final article in a mail newsgroup. This is to make life easier for procmail users.

By the way: That line up there, about Gnus never expiring non-expirable articles, is a lie. If you put total-expire in the group parameters, articles will not be marked as expirable, but all read articles will be put through the expiry process. Use with extreme caution. Even more dangerous is the gnus-total-expirable-newsgroups variable. All groups that match this regexp will have all read articles put through the expiry process, which means that all old mail articles in the groups in question will be deleted after a while. Use with extreme caution, and don’t come crying to me when you discover that the regexp you used matched the wrong group and all your important mail has disappeared. Be a man! Or a woman! Whatever you feel more comfortable with! So there!

Most people make most of their mail groups total-expirable, though.

If gnus-inhibit-user-auto-expire is non-nil, user marking commands will not mark an article as expirable, even if the group has auto-expire turned on.

The expirable marks of articles will be removed when copying or moving them to a group in which auto-expire is not turned on. This is for preventing articles from being expired unintentionally. On the other hand, to a group that has turned auto-expire on, the expirable marks of articles that are copied or moved will not be changed by default. I.e., when copying or moving to such a group, articles that were expirable will be left expirable and ones that were not expirable will not be marked as expirable. So, even though in auto-expire groups, some articles will never get expired (unless you read them again). If you don’t side with that behavior that unexpirable articles may be mixed into auto-expire groups, you can set gnus-mark-copied-or-moved-articles-as-expirable to a non-nil value. In that case, articles that have been read will be marked as expirable automatically when being copied or moved to a group that has auto-expire turned on. The default value is nil.

7.4.10 Washing Mail

Mailers and list servers are notorious for doing all sorts of really, really stupid things with mail. “Hey, RFC 822 doesn’t explicitly prohibit us from adding the string wE aRe ElItE!!!!!1!! to the end of all lines passing through our server, so let’s do that!!!!1!” Yes, but RFC 822 and its successors weren’t designed to be read by morons. Things that were considered to be self-evident were not discussed. So. Here we are.

Case in point: The German version of Microsoft Exchange adds ‘AW: ’ to the subjects of replies instead of ‘Re: ’. I could pretend to be shocked and dismayed by this, but I haven’t got the energy. It is to laugh.

Gnus provides a plethora of functions for washing articles while displaying them, but it might be nicer to do the filtering before storing the mail to disk. For that purpose, we have three hooks and various functions that can be put in these hooks.


This hook is called before doing anything with the mail and is meant for grand, sweeping gestures. It is called in a buffer that contains all the new, incoming mail. Functions to be used include:


Remove trailing carriage returns from each line. This is default on Emacs running on MS machines.


This hook is called narrowed to each header. It can be used when cleaning up the headers. Functions that can be used include:


Clear leading white space that “helpful” listservs have added to the headers to make them look nice. Aaah.

(Note that this function works on both the header and the body of all messages, so it is a potentially dangerous function to use (if a body of a message contains something that looks like a header line). So rather than fix the bug, it is of course the right solution to make it into a feature by documenting it.)


Some list servers add an identifier—for example, ‘(idm)’—to the beginning of all Subject headers. I’m sure that’s nice for people who use stone age mail readers. This function will remove strings that match the nnmail-list-identifiers regexp, which can also be a list of regexp. nnmail-list-identifiers may not contain \\(..\\).

For instance, if you want to remove the ‘(idm)’ and the ‘nagnagnag’ identifiers:

(setq nnmail-list-identifiers
      '("(idm)" "nagnagnag"))

This can also be done non-destructively with gnus-list-identifiers, See Article Hiding.


Translate all ‘TAB’ characters into ‘SPC’ characters.


Some mail user agents (e.g., Eudora and Pegasus) produce broken References headers, but correct In-Reply-To headers. This function will get rid of the References header if the headers contain a line matching the regular expression nnmail-broken-references-mailers.


This hook is called narrowed to each message. Functions to be used include:


Decode Quoted Readable encoding.

7.4.11 Duplicates

If you are a member of a couple of mailing lists, you will sometimes receive two copies of the same mail. This can be quite annoying, so nnmail checks for and treats any duplicates it might find. To do this, it keeps a cache of old Message-IDs: nnmail-message-id-cache-file, which is ~/.nnmail-cache by default. The approximate maximum number of Message-IDs stored there is controlled by the nnmail-message-id-cache-length variable, which is 1000 by default. (So 1000 Message-IDs will be stored.) If all this sounds scary to you, you can set nnmail-treat-duplicates to warn (which is what it is by default), and nnmail won’t delete duplicate mails. Instead it will insert a warning into the head of the mail saying that it thinks that this is a duplicate of a different message.

This variable can also be a function. If that’s the case, the function will be called from a buffer narrowed to the message in question with the Message-ID as a parameter. The function must return either nil, warn, or delete.

You can turn this feature off completely by setting the variable to nil.

If you want all the duplicate mails to be put into a special duplicates group, you could do that using the normal mail split methods:

(setq nnmail-split-fancy
      '(| ;; Messages duplicates go to a separate group.
        ("gnus-warning" "duplicat\\(e\\|ion\\) of message" "duplicate")
        ;; Message from daemons, postmaster, and the like to another.
        (any mail "mail.misc")
        ;; Other rules.
        [...] ))

Or something like:

(setq nnmail-split-methods
      '(("duplicates" "^Gnus-Warning:.*duplicate")
        ;; Other rules.

Here’s a neat feature: If you know that the recipient reads her mail with Gnus, and that she has nnmail-treat-duplicates set to delete, you can send her as many insults as you like, just by using a Message-ID of a mail that you know that she’s already received. Think of all the fun! She’ll never see any of it! Whee!

7.4.12 Not Reading Mail

If you start using any of the mail back ends, they have the annoying habit of assuming that you want to read mail with them. This might not be unreasonable, but it might not be what you want.

If you set mail-sources to nil, none of the back ends will ever attempt to read incoming mail, which should help.

This might be too much, if, for instance, you are reading mail quite happily with nnml and just want to peek at some old (pre-Emacs 23) Rmail file you have stashed away with nnbabyl. All back ends have variables called back-end-get-new-mail. If you want to disable the nnbabyl mail reading, you edit the virtual server for the group to have a setting where nnbabyl-get-new-mail to nil.

All the mail back ends will call nn*-prepare-save-mail-hook narrowed to the article to be saved before saving it when reading incoming mail.

7.4.13 Choosing a Mail Back End

Gnus will read the mail spool when you activate a mail group. The mail file is first copied to your home directory. What happens after that depends on what format you want to store your mail in.

There are six different mail back ends in the standard Gnus, and more back ends are available separately. The mail back end most people use (because it is possibly the fastest) is nnml (see Mail Spool). Unix Mail Box

The nnmbox back end will use the standard Un*x mbox file to store mail. nnmbox will add extra headers to each mail article to say which group it belongs in.

Virtual server settings:


The name of the mail box in the user’s home directory. Default is ~/mbox.


The name of the active file for the mail box. Default is ~/.mbox-active.


If non-nil, nnmbox will read incoming mail and split it into groups. Default is t. Babyl

The nnbabyl back end will use a Babyl mail box to store mail. nnbabyl will add extra headers to each mail article to say which group it belongs in.

Virtual server settings:


The name of the Babyl file. The default is ~/RMAIL


The name of the active file for the Babyl file. The default is ~/.rmail-active


If non-nil, nnbabyl will read incoming mail. Default is t Mail Spool

The nnml spool mail format isn’t compatible with any other known format. It should be used with some caution.

If you use this back end, Gnus will split all incoming mail into files, one file for each mail, and put the articles into the corresponding directories under the directory specified by the nnml-directory variable. The default value is ~/Mail/.

You do not have to create any directories beforehand; Gnus will take care of all that.

If you have a strict limit as to how many files you are allowed to store in your account, you should not use this back end. As each mail gets its own file, you might very well occupy thousands of inodes within a few weeks. If this is no problem for you, and it isn’t a problem for you having your friendly systems administrator walking around, madly, shouting “Who is eating all my inodes?! Who? Who!?!”, then you should know that this is probably the fastest format to use. You do not have to trudge through a big mbox file just to read your new mail.

nnml is probably the slowest back end when it comes to article splitting. It has to create lots of files, and it also generates NOV databases for the incoming mails. This makes it possibly the fastest back end when it comes to reading mail.

Virtual server settings:


All nnml directories will be placed under this directory. The default is the value of message-directory (whose default value is ~/Mail).


The active file for the nnml server. The default is ~/Mail/active.


The nnml group descriptions file. See Newsgroups File Format. The default is ~/Mail/newsgroups.


If non-nil, nnml will read incoming mail. The default is t.


If non-nil, this back end will ignore any NOV files. The default is nil.


The name of the NOV files. The default is .overview.


Hook run narrowed to an article before saving.


If non-nil, nnml will allow using compressed message files. This requires auto-compression-mode to be enabled (see Compressed Files in The Emacs Manual). If the value of nnml-use-compressed-files is a string, it is used as the file extension specifying the compression program. You can set it to ‘.bz2’ if your Emacs supports it. A value of t is equivalent to ‘.gz’.


Default size threshold for compressed message files. Message files with bodies larger than that many characters will be automatically compressed if nnml-use-compressed-files is non-nil.

If your nnml groups and NOV files get totally out of whack, you can do a complete update by typing M-x nnml-generate-nov-databases. This command will trawl through the entire nnml hierarchy, looking at each and every article, so it might take a while to complete. A better interface to this functionality can be found in the server buffer (see Server Commands). MH Spool

nnmh is just like nnml, except that is doesn’t generate NOV databases and it doesn’t keep an active file or marks file. This makes nnmh a much slower back end than nnml, but it also makes it easier to write procmail scripts for.

Virtual server settings:


All nnmh directories will be located under this directory. The default is the value of message-directory (whose default is ~/Mail)


If non-nil, nnmh will read incoming mail. The default is t.


If non-nil, nnmh will go to ridiculous lengths to make sure that the articles in the folder are actually what Gnus thinks they are. It will check date stamps and stat everything in sight, so setting this to t will mean a serious slow-down. If you never use anything but Gnus to read the nnmh articles, you do not have to set this variable to t. The default is nil. Maildir

nnmaildir stores mail in the maildir format, with each maildir corresponding to a group in Gnus. This format is documented here: nnmaildir also stores extra information in the .nnmaildir/ directory within a maildir.

Maildir format was designed to allow concurrent deliveries and reading, without needing locks. With other back ends, you would have your mail delivered to a spool of some kind, and then you would configure Gnus to split mail from that spool into your groups. You can still do that with nnmaildir, but the more common configuration is to have your mail delivered directly to the maildirs that appear as group in Gnus.

nnmaildir is designed to be perfectly reliable: C-g will never corrupt its data in memory, and SIGKILL will never corrupt its data in the filesystem.

nnmaildir stores article marks and NOV data in each maildir. So you can copy a whole maildir from one Gnus setup to another, and you will keep your marks.

Virtual server settings:


For each of your nnmaildir servers (it’s very unlikely that you’d need more than one), you need to create a directory and populate it with maildirs or symlinks to maildirs (and nothing else; do not choose a directory already used for other purposes). Each maildir will be represented in Gnus as a newsgroup on that server; the filename of the symlink will be the name of the group. Any filenames in the directory starting with ‘.’ are ignored. The directory is scanned when you first start Gnus, and each time you type g in the group buffer; if any maildirs have been removed or added, nnmaildir notices at these times.

The value of the directory parameter should be a Lisp form which is processed by eval and expand-file-name to get the path of the directory for this server. The form is evaled only when the server is opened; the resulting string is used until the server is closed. (If you don’t know about forms and eval, don’t worry—a simple string will work.) This parameter is not optional; you must specify it. I don’t recommend using "~/Mail" or a subdirectory of it; several other parts of Gnus use that directory by default for various things, and may get confused if nnmaildir uses it too. "~/.nnmaildir" is a typical value.


This should be a Lisp form which is processed by eval and expand-file-name. The form is evaled only when the server is opened; the resulting string is used until the server is closed.

When you create a group on an nnmaildir server, the maildir is created with target-prefix prepended to its name, and a symlink pointing to that maildir is created, named with the plain group name. So if directory is "~/.nnmaildir" and target-prefix is "../maildirs/", then when you create the group foo, nnmaildir will create ~/.nnmaildir/../maildirs/foo as a maildir, and will create ~/.nnmaildir/foo as a symlink pointing to ../maildirs/foo.

You can set target-prefix to a string without any slashes to create both maildirs and symlinks in the same directory; in this case, any maildirs found in directory whose names start with target-prefix will not be listed as groups (but the symlinks pointing to them will be).

As a special case, if target-prefix is "" (the default), then when you create a group, the maildir will be created in directory without a corresponding symlink. Beware that you cannot use gnus-group-delete-group on such groups without the force argument.


This should be a function with the same interface as directory-files (such as directory-files itself). It is used to scan the server’s directory for maildirs. This parameter is optional; the default is nnheader-directory-files-safe if nnheader-directory-files-is-safe is nil, and directory-files otherwise. (nnheader-directory-files-is-safe is checked only once when the server is opened; if you want to check it each time the directory is scanned, you’ll have to provide your own function that does that.)


If non-nil, then after scanning for new mail in the group maildirs themselves as usual, this server will also incorporate mail the conventional Gnus way, from mail-sources according to nnmail-split-methods or nnmail-split-fancy. The default value is nil.

Do not use the same maildir both in mail-sources and as an nnmaildir group. The results might happen to be useful, but that would be by chance, not by design, and the results might be different in the future. Group parameters

nnmaildir uses several group parameters. It’s safe to ignore all this; the default behavior for nnmaildir is the same as the default behavior for other mail back ends: articles are deleted after one week, etc. Except for the expiry parameters, all this functionality is unique to nnmaildir, so you can ignore it if you’re just trying to duplicate the behavior you already have with another back end.

If the value of any of these parameters is a vector, the first element is evaluated as a Lisp form and the result is used, rather than the original value. If the value is not a vector, the value itself is evaluated as a Lisp form. (This is why these parameters use names different from those of other, similar parameters supported by other back ends: they have different, though similar, meanings.) (For numbers, strings, nil, and t, you can ignore the eval business again; for other values, remember to use an extra quote and wrap the value in a vector when appropriate.)


An integer specifying the minimum age, in seconds, of an article before it will be expired, or the symbol never to specify that articles should never be expired. If this parameter is not set, nnmaildir falls back to the usual nnmail-expiry-wait(-function) variables (the expiry-wait group parameter overrides nnmail-expiry-wait and makes nnmail-expiry-wait-function ineffective). If you wanted a value of 3 days, you could use something like [(* 3 24 60 60)]; nnmaildir will evaluate the form and use the result. An article’s age is measured starting from the article file’s modification time. Normally, this is the same as the article’s delivery time, but editing an article makes it younger. Moving an article (other than via expiry) may also make an article younger.


If this is set to a string such as a full Gnus group name, like


and if it is not the name of the same group that the parameter belongs to, then articles will be moved to the specified group during expiry before being deleted. If this is set to an nnmaildir group, the article will be just as old in the destination group as it was in the source group. So be careful with expire-age in the destination group. If this is set to the name of the same group that the parameter belongs to, then the article is not expired at all. If you use the vector form, the first element is evaluated once for each article. So that form can refer to nnmaildir-article-file-name, etc., to decide where to put the article. Even if this parameter is not set, nnmaildir does not fall back to the expiry-target group parameter or the nnmail-expiry-target variable.


If this is set to t, nnmaildir will treat the articles in this maildir as read-only. This means: articles are not renamed from new/ into cur/; articles are only found in new/, not cur/; articles are never deleted; articles cannot be edited. new/ is expected to be a symlink to the new/ directory of another maildir—e.g., a system-wide mailbox containing a mailing list of common interest. Everything in the maildir outside new/ is not treated as read-only, so for a shared mailbox, you do still need to set up your own maildir (or have write permission to the shared mailbox); your maildir just won’t contain extra copies of the articles.


A function with the same interface as directory-files. It is used to scan the directories in the maildir corresponding to this group to find articles. The default is the function specified by the server’s directory-files parameter.


If non-nil, nnmaildir will always count the lines of an article, rather than use the Lines: header field. If nil, the header field will be used if present.


A list of mark symbols, such as ['(read expire)]. Whenever Gnus asks nnmaildir for article marks, nnmaildir will say that all articles have these marks, regardless of whether the marks stored in the filesystem say so. This is a proof-of-concept feature that will probably be removed eventually; it ought to be done in Gnus proper, or abandoned if it’s not worthwhile.


A list of mark symbols, such as ['(tick expire)]. Whenever Gnus asks nnmaildir for article marks, nnmaildir will say that no articles have these marks, regardless of whether the marks stored in the filesystem say so. never-marks overrides always-marks. This is a proof-of-concept feature that will probably be removed eventually; it ought to be done in Gnus proper, or abandoned if it’s not worthwhile.


An integer specifying the size of the NOV memory cache. To speed things up, nnmaildir keeps NOV data in memory for a limited number of articles in each group. (This is probably not worthwhile, and will probably be removed in the future.) This parameter’s value is noticed only the first time a group is seen after the server is opened—i.e., when you first start Gnus, typically. The NOV cache is never resized until the server is closed and reopened. The default is an estimate of the number of articles that would be displayed in the summary buffer: a count of articles that are either marked with tick or not marked with read, plus a little extra. Article identification

Articles are stored in the cur/ subdirectory of each maildir. Each article file is named like uniq:info, where uniq contains no colons. nnmaildir ignores, but preserves, the :info part. (Other maildir readers typically use this part of the filename to store marks.) The uniq part uniquely identifies the article, and is used in various places in the .nnmaildir/ subdirectory of the maildir to store information about the corresponding article. The full pathname of an article is available in the variable nnmaildir-article-file-name after you request the article in the summary buffer. NOV data

An article identified by uniq has its NOV data (used to generate lines in the summary buffer) stored in .nnmaildir/nov/uniq. There is no nnmaildir-generate-nov-databases function. (There isn’t much need for it—an article’s NOV data is updated automatically when the article or nnmail-extra-headers has changed.) You can force nnmaildir to regenerate the NOV data for a single article simply by deleting the corresponding NOV file, but beware: this will also cause nnmaildir to assign a new article number for this article, which may cause trouble with seen marks, the Agent, and the cache. Article marks

An article identified by uniq is considered to have the mark flag when the file .nnmaildir/marks/flag/uniq exists. When Gnus asks nnmaildir for a group’s marks, nnmaildir looks for such files and reports the set of marks it finds. When Gnus asks nnmaildir to store a new set of marks, nnmaildir creates and deletes the corresponding files as needed. (Actually, rather than create a new file for each mark, it just creates hard links to .nnmaildir/markfile, to save inodes.)

You can invent new marks by creating a new directory in .nnmaildir/marks/. You can tar up a maildir and remove it from your server, untar it later, and keep your marks. You can add and remove marks yourself by creating and deleting mark files. If you do this while Gnus is running and your nnmaildir server is open, it’s best to exit all summary buffers for nnmaildir groups and type s in the group buffer first, and to type g or M-g in the group buffer afterwards. Otherwise, Gnus might not pick up the changes, and might undo them. Mail Folders

nnfolder is a back end for storing each mail group in a separate file. Each file is in the standard Un*x mbox format. nnfolder will add extra headers to keep track of article numbers and arrival dates.

Virtual server settings:


All the nnfolder mail boxes will be stored under this directory. The default is the value of message-directory (whose default is ~/Mail)


The name of the active file. The default is ~/Mail/active.


The name of the group descriptions file. See Newsgroups File Format. The default is ~/Mail/newsgroups


If non-nil, nnfolder will read incoming mail. The default is t


Hook run before saving the folders. Note that Emacs does the normal backup renaming of files even with the nnfolder buffers. If you wish to switch this off, you could say something like the following in your .emacs file:

(defun turn-off-backup ()
  (set (make-local-variable 'backup-inhibited) t))

(add-hook 'nnfolder-save-buffer-hook 'turn-off-backup)

Hook run in a buffer narrowed to the message that is to be deleted. This function can be used to copy the message to somewhere else, or to extract some information from it before removing it.


If non-nil, this back end will ignore any NOV files. The default is nil.


The extension for NOV files. The default is .nov.


The directory where the NOV files should be stored. If nil, nnfolder-directory is used.

If you have lots of nnfolder-like files you’d like to read with nnfolder, you can use the M-x nnfolder-generate-active-file command to make nnfolder aware of all likely files in nnfolder-directory. This only works if you use long file names, though. Comparing Mail Back Ends

First, just for terminology, the back end is the common word for a low-level access method—a transport, if you will, by which something is acquired. The sense is that one’s mail has to come from somewhere, and so selection of a suitable back end is required in order to get that mail within spitting distance of Gnus.

The same concept exists for Usenet itself: Though access to articles is typically done by NNTP these days, once upon a midnight dreary, everyone in the world got at Usenet by running a reader on the machine where the articles lay (the machine which today we call an NNTP server), and access was by the reader stepping into the articles’ directory spool area directly. One can still select between either the nntp or nnspool back ends, to select between these methods, if one happens actually to live on the server (or can see its spool directly, anyway, via NFS).

The goal in selecting a mail back end is to pick one which simultaneously represents a suitable way of dealing with the original format plus leaving mail in a form that is convenient to use in the future. Here are some high and low points on each:


UNIX systems have historically had a single, very common, and well-defined format. All messages arrive in a single spool file, and they are delineated by a line whose regular expression matches ‘^From_’. (My notational use of ‘_’ is to indicate a space, to make it clear in this instance that this is not the RFC-specified ‘From:’ header.) Because Emacs and therefore Gnus emanate historically from the Unix environment, it is simplest if one does not mess a great deal with the original mailbox format, so if one chooses this back end, Gnus’ primary activity in getting mail from the real spool area to Gnus’ preferred directory is simply to copy it, with no (appreciable) format change in the process. It is the “dumbest” way to move mail into availability in the Gnus environment. This makes it fast to move into place, but slow to parse, when Gnus has to look at what’s where.


Once upon a time, there was the DEC-10 and DEC-20, running operating systems called TOPS and related things, and the usual (only?) mail reading environment was a thing called Babyl. I don’t know what format was used for mail landing on the system, but Babyl had its own internal format to which mail was converted, primarily involving creating a spool-file-like entity with a scheme for inserting Babyl-specific headers and status bits above the top of each message in the file. Rmail was Emacs’s first mail reader, it was written by Richard Stallman, and Stallman came out of that TOPS/Babyl environment, so he wrote Rmail to understand the mail files folks already had in existence. Gnus (and VM, for that matter) continue to support this format because it’s perceived as having some good qualities in those mailer-specific headers/status bits stuff. Rmail itself still exists as well, of course, and is still maintained within Emacs. Since Emacs 23, it uses standard mbox format rather than Babyl.

Both of the above forms leave your mail in a single file on your file system, and they must parse that entire file each time you take a look at your mail.


nnml is the back end which smells the most as though you were actually operating with an nnspool-accessed Usenet system. (In fact, I believe nnml actually derived from nnspool code, lo these years ago.) One’s mail is taken from the original spool file, and is then cut up into individual message files, 1:1. It maintains a Usenet-style active file (analogous to what one finds in an INN- or CNews-based news system in (for instance) /var/lib/news/active, or what is returned via the ‘NNTP LIST’ verb) and also creates overview files for efficient group entry, as has been defined for NNTP servers for some years now. It is slower in mail-splitting, due to the creation of lots of files, updates to the nnml active file, and additions to overview files on a per-message basis, but it is extremely fast on access because of what amounts to the indexing support provided by the active file and overviews.

nnml costs inodes in a big way; that is, it soaks up the resource which defines available places in the file system to put new files. Sysadmins take a dim view of heavy inode occupation within tight, shared file systems. But if you live on a personal machine where the file system is your own and space is not at a premium, nnml wins big.

It is also problematic using this back end if you are living in a FAT16-based Windows world, since much space will be wasted on all these tiny files.


The Rand MH mail-reading system has been around UNIX systems for a very long time; it operates by splitting one’s spool file of messages into individual files, but with little or no indexing support—nnmh is considered to be semantically equivalent to “nnml without active file or overviews”. This is arguably the worst choice, because one gets the slowness of individual file creation married to the slowness of access parsing when learning what’s new in one’s groups.


Basically the effect of nnfolder is nnmbox (the first method described above) on a per-group basis. That is, nnmbox itself puts all one’s mail in one file; nnfolder provides a little bit of optimization to this so that each of one’s mail groups has a Unix mail box file. It’s faster than nnmbox because each group can be parsed separately, and still provides the simple Unix mail box format requiring minimal effort in moving the mail around. In addition, it maintains an “active” file making it much faster for Gnus to figure out how many messages there are in each separate group.

If you have groups that are expected to have a massive amount of messages, nnfolder is not the best choice, but if you receive only a moderate amount of mail, nnfolder is probably the most friendly mail back end all over.


For configuring expiry and other things, nnmaildir uses incompatible group parameters, slightly different from those of other mail back ends.

nnmaildir is largely similar to nnml, with some notable differences. Each message is stored in a separate file, but the filename is unrelated to the article number in Gnus. nnmaildir also stores the equivalent of nnml’s overview files in one file per article, so it uses about twice as many inodes as nnml. (Use df -i to see how plentiful your inode supply is.) If this slows you down or takes up very much space, a non-block-structured file system.

Since maildirs don’t require locking for delivery, the maildirs you use as groups can also be the maildirs your mail is directly delivered to. This means you can skip Gnus’ mail spl