Bugs and problems in omnibook module code
=========================================

* You can use /proc/apm emulation for battery status monitoring only.
* Volume Control buttons do not generate scancodes on XE3GC style models.
  Scancode emulation needs intensive testing to be reliable.
* Volume Control buttons on machine (not on docking station) do
  not generate scancodes on OB500 style models. It is unhandled yet.
* Sometimes (I assume when the embedded controller is busy) the procfs
  interface displays fake values.
* Setting the LCD brightness on HP OmniBook XE3 GF via /proc/omnibook/lcd
  is working if you press one of the brightness control keys once after
  writing the value into /proc/omnibook/lcd.
* Undocking does not work reliable, sometimes it does not
  undock the supported machines.
* After reenabling the touchpad via /proc/omnibook/touchpad
  you may restart X / gpm.
* DMI code in kernel available at boot time only so I need
  to reuse it to identify the machine in an independent module.
* Early boot time workarounds could not be added to a module
  because of it loaded later.

Problems with OmniBook machines running Linux
=============================================

* If you use APM there are a lot of mysterious hangs at suspend/resume on
  most models. The OmniBooks use ACPI not APM but ACPI for Linux does not
  provide reliable power management functions yet. Use swsusp instead.
* On XE3 GF models the touchpad can freeze the laptop after disabling and
  reenabling it with the disable button and doing a tap action when neither
  gpm nor X is running. Please report if you can reproduce it.
