<aside> ✅ For common shit I never understood….
</aside>
https://ethresear.ch/t/clarification-on-how-calldata-persists-on-the-blockchain-and-how-optimistic-rollups-use-it/8136
https://www.reddit.com/r/ethereumnoobies/comments/sdkza2/what_is_calldata_in_relation_to_rollups/i89xdct/?context=8&depth=9
Calldata = History. That is all.
This is a question where I never understood how consensus practically works. In the real world of distributed machines, they are all distanced, and due to how networking performs, there is always latency reaching to other nodes. Therefore, if a group of validators are agreeing on the ordering of a chain, some outcast node following another consensus group may be choosing a different ordering, so this causes reorgs. TLDR; When describing the merge in theory, it is explaining as if there is 1 consensus. In reality, there can be many, since all machines are not communicating together.
But at that time, the outcast can enter the new consensus, and able to vote on the canonical chain.
Each of the peers in the network behaves both as a server and as a client.
Originally, NATs block p2p communication bc it hides global IPs of machines behind NATs address.
Therefore, you must configure your router manually to add a port mapping to local IP address.
Every node has a specific ID that is unique, uses sha3(public key)