Is actually better to list out the main LinkedList APIs we want to build as base requirements. And then talk about the details of how to achieve those with tradeOff considerations. For lamport timestamp based hybrid approach, can you describe how the application service (list access service) knows the node_id? Is the nodeId related to the data node or the list service itself?
"billions of nodes" "the value is an int" That's just 4 bytes. 1 billion times that's only 4GB. Even taking into account the "s" on the billions, its still only a few GB, which can easily be handled by a single machine.
It could be better I believe. You have great knowledge but it's hard to follow this video, you go back and forth between question and the different approach, I'm not sure at the end what I should learn and not, it's a bit confusing.
What is a celebrity problem?
Great content. I think one thing you can improve is to limit filler words. You use "erhm", "uhmm" way too often
Excalidraw links? How to get diagrams?
@eduardopotter3278