Inherently, UI IS state. Anytime you have a user interface, you literally are writing software to manage state of some kind. Take the TV remote... the user cont

Inherently, UI IS state. Anytime you have a user interface, you literally are writing software to manage state of some kind. Take the TV remote... the user controls the state of the TV, namely the channels, and properties like volume. The remote is an interface to change this state. Microservices is like trying to solve the problem of a really, really complicate TV with crazy channels and multiple volume properties by creating 10 different isolated controllers. It won't work because there exists a shared state, yet each controller is aiming to not share any state while still having a scaleable system. This is of course only possible in simpler UI's, because the more complex the UI, the more state will need to be shared, unless you want to throw away DRY principles and duplicate logic and styles all over what is seemingly the same app.