Maybe thats why one of the not-so-bad ideas how to go about distributed systems came from a guy who was trained as a physicist and used to complain to his fellow programmers that “a lot of systems actually break the laws of physics”[1]?
> In distributed systems there is no real shared state (imagine one machine in the USA another in Sweden) where is the shared state? In the middle of the Atlantic? - shared state breaks laws of physics. State changes are propagated at the speed of light - we always know how things were at a remote site not how they are now. What we know is what they last told us. If you make a software abstraction that ignores this fact you’ll be in trouble.[2]
[1]: “The Mess We’re In”, 2014 https://www.youtube.com/watch?v=lKXe3HUG2l4 [2]: https://news.ycombinator.com/item?id=19708900