xmonad-contrib-bluetilebranch- Third party extensions for xmonad





Utility functions for simulating independent sets of workspaces on each screen (like dwm's workspace model), using internal tags to distinguish workspaces associated with each screen.



You can use this module with the following in your ~/.xmonad/xmonad.hs:

 import XMonad.Layout.IndependentScreens

You can define your workspaces by calling withScreens:

 myConfig = defaultConfig { workspaces = withScreens 2 ["web", "email", "irc"] }

This will create "physical" workspaces with distinct internal names for each (screen, virtual workspace) pair.

Then edit any keybindings that use the list of workspaces or refer to specific workspace names. In the default configuration, only the keybindings for changing workspace do this:

 [((m .|. modm, k), windows $ f i)
     | (i, k) <- zip (XMonad.workspaces conf) [xK_1 .. xK_9]
     , (f, m) <- [(W.greedyView, 0), (W.shift, shiftMask)]]

This should change to

 [((m .|. modm, k), windows $ onCurrentScreen f i)
     | (i, k) <- zip (workspaces' conf) [xK_1 .. xK_9]
     , (f, m) <- [(W.greedyView, 0), (W.shift, shiftMask)]]

In particular, the analogue of XMonad.workspaces is workspaces', and you can use onCurrentScreen to convert functions of virtual workspaces to functions of physical workspaces, which work by marshalling the virtual workspace name and the currently focused screen into a physical workspace name.



:: ScreenId

The number of screens to make workspaces for

-> [VirtualWorkspace]

The desired virtual workspace names

-> [PhysicalWorkspace]

A list of all internal physical workspace names

marshallPP :: ScreenId -> PP -> PPSource

This turns a naive pretty-printer into one that is aware of the independent screens. That is, you can write your pretty printer to behave the way you want on virtual workspaces; this function will convert that pretty-printer into one that first filters out physical workspaces on other screens, then converts all the physical workspaces on this screen to their virtual names.

For example, if you have handles hLeft and hRight for bars on the left and right screens, respectively, and pp is a pretty-printer function that takes a handle, you could write

 logHook = let log screen handle = dynamicLogWithPP . marshallPP screen . pp $ handle
           in log 0 hLeft >> log 1 hRight

countScreens :: (MonadIO m, Integral i) => m iSource

In case you don't know statically how many screens there will be, you can call this in main before starting xmonad. For example, part of my config reads

 main = do
   nScreens <- countScreens
   xmonad $ defaultConfig {
     workspaces = withScreens nScreens (workspaces defaultConfig),

Converting between virtual and physical workspaces

You shouldn't need to use the functions below very much. They are used internally. However, in some cases, they may be useful, and so are exported just in case. In general, the "marshall" functions convert the convenient form (like "web") you would like to use in your configuration file to the inconvenient form (like "2_web") that xmonad uses internally. Similarly, the "unmarshall" functions convert in the other direction.