@Chris:
I fixed another bug with the same behavior. Look at commit 469 in the trunk. box was first passed with the right signal, but then the signal was respawing with an empty box. Testing the pointer isn't null fix it and still enable keyboard navigation.
@Chris:
I fixed another bug with the same behavior. Look at commit 469 in the trunk. box was first passed with the right signal, but then the signal was respawing with an empty box. Testing the pointer isn't null fix it and still enable keyboard navigation.