| Age | Commit message (Collapse) | Author |
|---|
|
When persistent is set to True then xmobar will always be mapped
(revealed) and never be hidden.
The flag is checked in eventLoop and operation to map/unmap windows is
not carried out if persistence is desired.
|
|
The window became hidden although the toggling behaviour was set to
False for a particular pipe. This fixes this behaviour and hides the
window only if the configuration option is set to True.
|
|
This solves a problem when there is only one pipe in place. With a
default value of "" and only one pipe with a timeout the value is
overwritten with "" after the timeout. To prevent this from happening a
TMVar is used which will never be filled if there is only one pipe.
|
|
Using the trigger method activity on a pipe can now cause the window to
appear (reveal) and disappear again after a given timeout. The timeout
for hiding the window is the same as for restoring the pipes content.
The timeout value is given in tenth of seconds.
|
|
Realign methods, remove unnecessary imports and remove clutter
|
|
Toggling is based is based on the current window status. If unmapped
then reveal else hide.
Sync is necessary or delays might occur.
The functions are called from the event loop when the according signal
is received
When mapping (revealing) the window again we need to set the struts
property again. The easiest way to do this is to call repositionWin.
However, repositionWin needs access to the Config structure which is
available in eventLoop.
Because decomposition wouldn't be easy and I don't want to pass Config
down to showWindow (which would need to return the new Rectangle then)
this is done here.
|
|
This is necessary for setting up the signal callback (trigger) from the
Plugin interface. As another benefit it is now possible to implement the
lowerOnStart config option properly by simply sending a Hide signal in
startLoop.
|
|
Also: realign methods to look pretty again.
|
|
Also make them {Read,Show}able which can be useful for printf debugging
and does not hurt otherwise.
|
|
This is necessary to make SignalType available for other modules without
import loops. This also decoupels the modules and their functionality a
bit more so this is generally a cleaner solution.
|
|
These functions are about creation, positioning and property setting of
the xmobar window. An own module does them justice and eases the task of
adding functions for revealing/hiding and toggling the window.
|
|
This plugin allows to display data from multiple pipes.
New data will always overwrite the currently displayed data.
However, if a timeout is specified, the previous content is restored.
Configuration works like this:
BufferedPipeReader <Alias> [ ( Timeout, "/path/to/fifo/pipe" ), (..), .. ]
If Timeout is set to 0 then the content is persistent, i.e. it will be
reset to any previous value, it will itself become the previous value.
If Timeout is set to a negative value the earth will stop spinning, so
don't do it.
|
|
We're using now the recommended statvfs interface, instead of the
obsolete statfs64. Moreover, we compute correctly the used space.
|
|
Conflicts:
xmobar.cabal
|
|
|
|
Now all values are returned as 'Value' wrapped entries.
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
|
|
This allows monitors to define update times other than time-based.
|
|
Fixes #36
|
|
Since a kernel upgrade about two months ago, DiskU just blows up my
xmobar apparently at random.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Apparently, temp1_label does not follow the same naming pattern as the
rest of label files.
|
|
|
|
|
|
|
|
|
|
https://github.com/mathstuf/xmobar into mathstuf-dev/use-x11-xrandr-support
|
|
dev/add-incremental-config-option
|
|
Computing used space from total and available (rather than free) space.
|
|
As per Joey Hess advice in
http://code.google.com/p/xmobar/issues/detail?id=28#c15
|
|
|
|
|
|
Using the -C option, *additional* commands can be given to Xmobar. This
allows a common configuration to be used and then differentiated in just
the -C calls.
|
|
|
|
Just as the Date plugin used the old-time package which would honor all
the available flags, so too did the DateZone plugin.
|
|
The Date plugin used the old-time package to format times which does not
honor all escape sequences and optional flags (e.g. the - (hyphen) that
should not pad the field).
|
|
|
|
Hat tip David U.
|
|
We were getting confused by devices appearing in mtab by their
symbolic path (such as an entry in /dev/disk/by-uuid and the like).
|
|
|
|
|
|
Must have been a glitch in my kernel version (?)
|
|
|
|
|
