Instead of Return value:

Similar to map/unmap, this needs to follow the actual
widget tree. Without this, several test cases in
the templates test fail.

These were added to the header by mistake in a recent commit.
They have no implementation, so dropping them won't cause
any harm.

GtkPrintUnixDialog was leaking a GFile.
Spotted by Christian Persch in
https://bugzilla.gnome.org/show_bug.cgi?id=724631

And document the fact that the popover will get destroyed if
a NULL relative-to is given on a parented popover, if no extra
references are kept.

For gtk_popover_new*(), a NULL relative-to will leave the widget
as a floating object, to be sunk by a later call to
gtk_widget_set_relative_to().

https://bugzilla.gnome.org/show_bug.cgi?id=724407

This gives a critical warning since that function expects a
valid widget.

https://bugzilla.gnome.org/show_bug.cgi?id=724407

Use a weak reference to notice when parent_scrollable is
going away.

https://bugzilla.gnome.org/show_bug.cgi?id=724392

This error was causing children to not get allocated when
the center widget is the first one in the list of the box's
children.

It makes sense to allow this, and gtk_action_bar_set_center_widget
already assumes that it can pass NULL to this function.

The recursion in map needs to follow the actual physical
widget tree, otherwise we violate invariants. The generic
container map implementation uses gtk_container_forall to
operate on the children, and thus is not suitable for
containers where the children are inside some internal
container.

We need to always iterate over non-internal children.

This reverts commit 0ff0fdfe.

This is needed to set a padding and draw a border since GtkBox does
not honour css padding yet.

https://bugzilla.gnome.org/show_bug.cgi?id=724693

This broke recently when the window dragging code was changed
to require motion events to be propagated as well.

https://bugzilla.gnome.org/show_bug.cgi?id=724539

This makes them draggable, and is the right thing to do.

We should only eat button release events when the label is
actually selectable, since the comment indicates that we
want to eat the release events belonging to press events
that triggered a selection. This fixes problems with actions
on parent widgets that are triggered by button release,
as seen in this bug:

https://bugzilla.gnome.org/show_bug.cgi?id=724541

GtkActionBar and GtkSearchBar also use templates,
add them here.

Suprisingly, this bug has been there for a very long time.
I'm fixing it now because we now use a custom search entry
in the app chooser dialog, and this is causing the templates
cleanup test to fail.

This was also causing the templates test to fail.

With nested internal children, this interferes with proper
cleanup, and makes the templates test fail.