I have been using this code locally for years now, and it turned out to
be quite useful in many cases, so I thought it is about time to submit
it to the repository.
TreeSelect displays your workspaces or actions in a Tree-like format.
You can select the desired workspace/action with the cursor or hjkl keys.
This module is fully configurable and very useful if you like to have a
lot of workspaces.
Please see the Documentation provided by 'XMonad.Actions.TreeSelect'.
Run a ManageHook from handleEventHook when a window property changes.
You would use this to match e.g. browser windows whose title is not
"final" until after the on-load hooks of the loaded document complete.
A layout modifer that is meant to work with the BinarySpacePartition
layout. By removing windows from the window set and adding them back
at a later time, you can move windows to a different branch of the BSP
tree.
This adds three layouts: Spiral is a reimplementation of X.L.Spiral.spiral
with a (to me) more intuitive splitting policy. Dwindle is similar but pushes
the smaller windows into a corner rather than into the center. Squeeze just
stacks windows vertically or horizontally using geometrically decreasing sizes.
Checks for floating windows within the layout area and finds a maximum area
rectangle within that does not overlap with any of the floating windows.
This rectangle is used for all non-floating windows.
This new functionality introduced problems with the recommended configuration
of one of my other modules (X.A.FloatSnap.) A new and more reliable method of
distinguishing between clicks and drags where therefore introduced in the new
module X.A.AfterDrag.
This does not break any prior use of FloatSnap, but will require changes in
configuration if used together with AvoidFloats. (This is mentioned in the
docs for AvoidFloats and I recommend using the new configuration method even if
AvoidFloats is not in use.)
The module implements a proper way of finding out whether the window is
remote or local.
Just checking for a hostname and WM_CLIENT_MACHINE being equal is often
not enough because the hostname is a changing subject (without any
established notification mechanisms), and thus WM_CLIENT_MACHINE and the
hostname can diverge even for a local window.
This module solves the problem. As soon as there is a new window created,
we check the hostname and WM_CLIENT_MACHINE, and then we cache the result
into the XMONAD_REMOTE property.
Notice that XMonad itself does not know anything about hostnames, nor does
it have any dependency on Network.* modules. For this module it is not a
problem: you can provide a mean to get the hostname through your config
file (see usage). Or, if you don't like the hassle of handling dynamic
hostnames (suppose your hostname never changes), it is also fine: this
module will fallback to using environment variables.
This module implements a special kind of layout modifier, which when
applied to a layout, causes xmonad to stop all non-visible processes. In a
way, this is a sledge-hammer for applications that drain power. For
example, given a web browser on a stoppable workspace, once the workspace
is hidden the web browser will be stopped.
Note that the stopped application won't be able to communicate with X11
clipboard. For this, the module actually stops applications after a
certain delay, giving a chance for a user to complete copy-paste sequence.
By default, the delay equals to 15 seconds, it is configurable via
'Stoppable' constructor.
The stoppable modifier prepends a mark (by default equals to "Stoppable")
to the layout description (alternatively, you can choose your own mark and
use it with 'Stoppable' constructor). The stoppable layout (identified by
a mark) spans to multiple workspaces, letting you to create groups of
stoppable workspaces that only stop processes when none of the workspaces
are visible, and conversely, unfreezing all processes even if one of the
stoppable workspaces are visible.
To stop the process we use signals, which works for most cases. For
processes that tinker with signal handling (debuggers), another
(Linux-centric) approach may be used. See
https://www.kernel.org/doc/Documentation/cgroups/freezer-subsystem.txt
Note that the use of RebindableSyntax is because of the need to vary the
layoutHook type throughout the config. The alternative, using the existential
Layout type, was rejected because it required TemplateHaskell in order to look
nice, and TemplateHaskell is not portable.
I've tried to make a version of (>>) that also worked on normal monads, but
have had no luck as of yet. Maybe some intrepid soul can add it later.
This module provides 3 <XMonad.Prompt> to ease passwords manipulation (generate, read, remove):
- one to lookup passwords in the password-storage.
- one to generate a password for a given password label that the user inputs.
- one to delete a stored password for a given password label that the user inputs.
All those prompts benefit from the completion system provided by the module <XMonad.Prompt>.
The password store is setuped through an environment variable PASSWORD_STORE_DIR.
If this is set, use the content of the variable.
Otherwise, the password store is located on user's home @$HOME\/.password-store@.
Source:
- The password storage implementation is <http://git.zx2c4.com/password-store the password-store cli>.
- Inspired from <http://babushk.in/posts/combining-xmonad-and-pass.html>
Initial support for the Mate desktop environment (http://mate-desktop.org).
Based on existing Gnome 2 support, since Mate is a maintained fork of
Gnome 2.
A set of hooks, and convenience combinators, to help with ManageHook debugging.
Ordinary users may well want to use debugManageHookOn in normal configs,
specifying a key sequence which can be pressed before running a command in
order to capture debug information just for that command's main window.
This is especially useful when trying to diagnose issues such as programs
that do not play well with SpawnOn, or ManageHook matching on 'title' when
the program does not set the window title until after it is mapped.
This makes a breakage due to missing patches in core a bit more obvious.
Previously you would have a build failure regarding some missing identifiers
(def re-exported by XMonad from Data.Default), while after applying this patch
it will be clear that xmonad-core needs to be updated.
This module provides a way to link certain workspaces in a multihead setup.
That way, when switching to the first one the other heads display the linked
workspaces.
