This question kind of relates to my question for lecture 1. I did a lot of work with azure messages & events over the summer, and I know that in higher-level software development for data streaming, there are distributed event platforms like Apache Kafka that have internal logic to preserve message sequencing, avoid duplicates, etc. -- do we have an equivalent low-level platform for this implementation? or are these kinds of messages more self-contained than those sent to the cloud, and clashes/backlogs are impossible?

Using the message passing model instead of the shared address space model also offers significant security benefits. With the shared address space, an attack on 1 thread can affect other threads, but by isolating the address spaces and forcing all communication through messages, an attack on one thread doesn't affect another (unless there is an exploit with the messaging system). This is why chrome uses many different processes - processes with higher likely hood of exploits restrict their own permissions so an exploit in the renderer can't cause massive damage to the system.