2007-11-09  Richard Hughes  <[email protected]>

       * src/gpm-button.c: (gpm_button_filter_x_events),
       (gpm_button_grab_keystring), (gpm_button_button_key):
       Properly process the GdkXEvent so we don't hide the GtkStatusIcon
       when we compile with --enable-xevents.
       The bug of icon not appearing happened because addition of
       notification-area-applet generates an X Event for notification-manager
       ownership change, and the bug in gpm_button_filter_x_events didn't allow
       gtktrayicon-x11.c to see the event due to the propagation stopping from
       GDK_FILTER_REMOVE being returned on all non-keypresses.
       This is a slightly modied patch from Mart Raudsepp, many thanks.
       Fixes #413360

2007-11-02  Richard Hughes  <[email protected]>

       * src/gpm-cell-array.c: (phone_device_removed_cb):
       Check to see if we have a gnome-phone-manager device before we unref
       it - I've a sneaking suspicion gnome-phone-manager is telling us
       Remove before Added at session start, although we certainly shouldn't
       crash out in this case. Hopefully fixes #484516

2007-10-16  Richard Hughes  <[email protected]>

       * src/gpm-manager.c: (idle_do_sleep):
       Always pass an error into gpm_control_ functions, else we can't detect
       if the method failed or if it just timed out.
       Patch from David Zeuthen, many thanks.

2007-10-13  Richard Hughes  <[email protected]>

       * src/gpm-manager.c: (gpm_manager_action_suspend),
       (gpm_manager_action_hibernate), (manager_policy_do),
       (gpm_manager_suspend), (gpm_manager_hibernate),
       (gpm_manager_reboot), (gpm_manager_shutdown), (idle_do_sleep),
       (idle_changed_cb):
       Check for inhibit in more places. Needed for gnome-packagekit work.

2007-10-05  Richard Hughes  <[email protected]>

       * src/gpm-manager.c: (gpm_manager_action_suspend),
       (gpm_manager_action_hibernate), (idle_do_sleep):
       Apply a patch from David Zeuthen:
       OK, so it turns out we're getting a D-Bus timeout. Actually g-p-m got
       codepaths to handle this but the call site passes NULL for the
       GError so we never get to examine this and properly handle timeouts.
       This patch takes care of that. I've tested and it works. Fixes rh:312761

2007-10-05  Richard Hughes  <[email protected]>

       * libhal-glib/libhal-gpower.c: (hal_gpower_suspend),
       (hal_gpower_pm_method_void), (hal_gpower_enable_power_save):
       * src/gpm-brightness-lcd.c: (gpm_brightness_lcd_get_hw),
       (gpm_brightness_lcd_set_hw):
       Use the correct DBUS return types now HAL enforces the return type.
       Based on a patch from David Zeuthen. Many thanks.

2007-11-02  Richard Hughes  <[email protected]>

       * configure.in:
       Post release version bump to 2.20.1 - better late than never.