I was playing around with firewalls and
iptables. When I blocked any traffic except HTTP and DNS I was able to establish a connection to
wget google.com. But I wasn't able to open google.com via Firefox.
After a little research I found out that Firefox uses the
loopback device to communicate with itself (Firefox loopback connection). After I added ACCEPT rules for INPUT and OUTPUT traffic for the loop device, I could open any website via Firefox.
My question is: What exactly does Firefox need the loopback device for?