Cables Pretty much always fibre, with the very last bit copper (be it a home user on DSL, but that’s changing, or the ethernet cable to a system. Very few servers historically directly had fibre, but it is becoming a lot more common – due to bandwidth) For long distances – travels better than electrical signals. Ground issues For short distances – electrical isolation Modems – essentially media converters. Copper -> Copper Fibre -> Copper Wireless -> Copper Routing Each point on the network only knows its neighbours – one of which is the next point to talk to for a destination – more than likely (but usually not for Tier 1/2) the one that is a form of “default”, for all the things that point doesn’t know about. Who controls the internet? Nobody really but there is ICANN. Also the Tier 1 bandwidth providers (Cogent, HE etc) Designed to be redundant so there’s not supposed to be one point of failure – however with the massive infrastructure that the internet now is, there are definitely some choke points. if something happened they’d get dealt with fairly quickly though. There are some countries where a single entity owns the majority of the telecommunications networks, so that makes it quite easy for them to “turn off” the internet such as Syria in 2013 IPv4 (netmasks are fun to explain) addresses – 0-255, 8 bits x 4 = 32 bits 4 billion addresses IPv6 You could give every grain of sand on 340 billion earth sized planets an address and you still wouldn’t run out (240 decillion addresses left) IPv1-IPv3 existed in theory but were scrapped due to design flaws IPv5 was an experimental internet made in the 1970s for streaming audio and video DNS “Telephone book of the internet” hierarchical. -> “.” -> “com.” -> “foo.com.” Google = 202.124.127.208 (for me) but not for everyone. Addresses change depending on where you are. Not like adding country code but actually a different number. This is due to geography – speed of light is still a thing, latency, send someone to a closer copy of a site/host/service/etc.