A few regressions were introduced in the last version:
- *optional* module was not correctly used
- interface selection was performed before mac spoofing, resulting in a disabled interface if spoofing used
- typos