Authentication: REPORT_IS_LOGGED_IN v.s. LOG_IN, & more

[hmaurer]

Hello hello,

I was wondering what the thought process behind having separate events for REPORT_IS_LOGGED_IN and LOG_IN was (and similarly for REPORT_IS_LOGGED_OUT and LOG_OUT)? There's a difference in semantics but it seems subtle.

In addition:

• Should the machine be extended to respond to REPORT_IS_LOGGED_OUT in the loggedIn state, so that it could be sent when an authentication token expires?
• Should the machine be extended with a loggingIn state? (in which case the difference between LOG_IN and REPORT_IS_LOGGED_IN would become much starker).

[mattpocock]

https://www.loom.com/share/d683ebd657664886a92902b82c7dfdf9