John Detlefs Programmer

Crazy rambling about observables

Here’s some random nonsense about my thinking that I want to share/refine while I’m still learning about Observables.

Consider the term Behavior to be an output from a program that is expected 1..N times. This output can have many different side-effects but should occur in a sequential, predictable manner. Observables allow for predictable creation of behavior, and it seems that in a creation of an observable each side-effect should come from the result of a pure function.

The act of subscribing to an observable triggers behavior. If an observable consists entirely of pure functions, this maintains ensures the benefits of pure functional programming.

This also seems to tie into a deeper idea from topology. Or at least a pattern; covering spaces are an entirely different subject from observables. But this picture seems to tie into the idea of an observable too well. covering space

Consider the behavior $B$, we lift it to the sandwich of the functions we want to create the desired behavior. Upon subscription we flatten back to the behavior $B$.


$B$ = {$f_1$, $f_2$, $f_3$}

And thats all I have, thanks, please refine my thinking in the comments!