Theory behind Reactor

[Jernej Jerin]

After reading through docs I have only encountered at the start of the docs about the theory behind the Reactor:

Being non-blocking matters—​especially when scaling message-passing becomes critical (10k msg/s, 100k msg/s 1M…​). There is some theory behind this (see Amdahl’s Law), but we get bored and distracted easily, so let’s first appeal to common sense.

Also there is a mention about using architectural patterns such as Reactive and Event Driven Architecture. It was also mentioned here that you use Proactor pattern most of the time. There is also the Reactive Streams specification. Is there anything else on the theory side that you used when developing the Reactor library?