This is most definitely correct .. until you want to scale beyond a few gigabytes of memory and then with a distributed cluster many synchronization issues don’

This is most definitely correct .. until you want to scale beyond a few gigabytes of memory and then with a distributed cluster many synchronization issues don’t seem that trivial anymore.