This adds a "Make Default" button in gnome-display-properties, which is used
to set up a systemwide default RANDR configuration. For users which don't have
a personal ~/.config/monitors.xml, then that systemwide configuration will
be used instead.
There is a PolicyKit helper program, gnome-display-properties-install-systemwide,
which is used to create the file
/etc/gnome-settings-daemon/xrandr/monitors.xml
in a secure fashion.
'passwd' returns an error when the new and the old passwords only differ in case. If we don't handle this (unlikely) case, users-admin hangs. Assimilate this error to "passwords are too similar".
into it and on the next run all fields are empty again.
The GTK+ 2.19.2 release notes say:
* Bugs fixed:
...
591085 GtkBuilder object ID bounded to GtkWidget "name" property
Apparently gtk_widget_get_name() was never supposed to return the ID of
the widget, so about_me_focus_out() worked only by mistake until GTK+
fixed the "bug".
right thing by setting "me = NULL" afterwards, but "me" in this case is
not the global variable, but the local one which shadows it. So the
global "me" remains non-NULL and when later about_me_focus_out() is
called, it cannot know it should return early.
It only makes sense for about_me_destroy() to operate on the global
"me", so there's no point in passing it as a parameter.
This button will use a PolicyKit helper to save the display configuration
to a systemwide file.
Signed-off-by: Federico Mena Quintero <federico@novell.com>
This feature was lost in 2.28, and people miss it. See bug 609809.
Note that this adds one new string, 'Span', which appears in the 'Style'
combobox in the background tab of the appearance capplet.
Now that the label with instructions is gone, let's have
a tooltip with better instructions.
Signed-off-by: Federico Mena Quintero <federico@novell.com>
Cross-with-arrows means 'you can move me', while hand means 'you can click me'
in this day and age. Pointed out by Ka-Hing Cheung <kahing@gmail.com>
Signed-off-by: Federico Mena Quintero <federico@novell.com>
The new window fits in 640x480 without any adjustments to the size of the
FooScrollArea, so we don't need to adjust that size anymore.
Signed-off-by: Federico Mena Quintero <federico@novell.com>
This should make it more obvious that outputs are in fact draggable
in the FooScrollArea, and it will let us remove the ugly label
with instructions for how to drag the monitors.
Signed-off-by: Federico Mena Quintero <federico@novell.com>
We move the per-monitor's widgets to the right of the FooScrollArea.
This makes the dialog fit even in 640x480 displays, and makes it more
obvious that the monitor's controls actually refer to the selected
monitor.
Signed-off-by: Federico Mena Quintero <federico@novell.com>
We simply hack the FooScrollArea for the monitor display to be shorter.
We can get fancier in the future by really computing the size based
on the monitor into which the window eventually gets mapped.
Patch by Takashi Iwai <tiwai@novell.com> - https://bugzilla.novell.com/show_bug.cgi?id=564579
Signed-off-by: Federico Mena Quintero <federico@novell.com>
When current password is wrong, passwd outputs on Ubuntu 9.10
passwd: Authentication token manipulation error
This doesn't correspond to what we're expecting. Add "error" to the list of words we look for to avoid waiting forever.
When pam-cracklib is used, using an existing word prompts 'passwd' to output
WEAK PASSWORD: it's based on a dictionary word
Detect "dictionary" too and consider this as meaning "Password is too simple".
The gnome-at-session autostart desktop file launches a shell script, which in
turn calls gconftool-2, just to find out that it's disabled, and exit. This
wastes unnecessary cycles on starting the desktop.
Add an AutostartCondition to the .desktop file which only runs when a11y is
enabled.
https://bugzilla.gnome.org/show_bug.cgi?id=608138
Apple touchpads have a single physical button and are thus capable of
working even if tapping is disabled. Don't force it on for these touchpads.
This patch simply changes the check for physical buttons from the previous
"check for LMR be present" to "check only for L".
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
This makes it even more obvious that you don't need to turn on
individual outputs for Mirror Screens to work properly.
It also means that we don't support "combined" setups like two mirrored
screens and one extended screen. It seems that the majority of people only care about a few cases:
- Laptop plus projector or monitor showing the same thing.
- Computer plus secondary monitor(s) with an extended desktop.
- Laptop plus docking station with external monitor only (built-in LCD turned off).
For more exotic combinations, people can use the xrandr(1) tool.
Signed-off-by: Federico Mena Quintero <federico@novell.com>
Previously it was always sensitive, even if there were not enough connected outputs
for mirroring, or if those those outputs didn't support the same resolution for
mirroring.
Signed-off-by: Federico Mena Quintero <federico@novell.com>
