reactiveX vs Async

Joseph L. Casale jcasale at
Wed Jan 11 21:12:29 EST 2017

> Thanks Joseph
> Trouble is there is stew of technologies/languages…
> (meta)-stewed with more abstract concepts, eg push vs pull, Enumerable-Observable
> duality, continuous vs discrete time
> The last causing its own share of confusion with “functional reactive programming” (FRP) meaning sometimes the one and sometimes the other:

Yeah I don't claim to be an expert, I have only just started using the syntax
in my own projects whereas I almost always used the classical thread model
with sync primitives and message passing.

> As for 'Overlook the “async/await can be used without significant buy-in” '
> I believe Ive seen people with considerable experience/understanding of the area
> take the opposite view, usually along the lines: “Once you start going non-blocking,
> you have to non-block all the way”

Right, as I mentioned overlook the pedantic argument just for the concept.
While you *can* do as I mentioned rather easily, it's not a useful or scalable
approach. Their suggestion is correct in reality but I only mentioned the minimal
buy-in it to illustrate that existing code *could* adopt the pattern trivially
whereas Rx is not simply two keywords, it's an entire framework.

Ultimately they accomplish the same thing, that is asynchronous execution
but use two different patterns. I'll paraphrase another source: "Rx adds a
convenient way to add callbacks and manage execution." The async/await
pattern is closer to the metal, it's not dealing with a sequence of observable
objects, it is however manipulating execution of the explicit block of code where
its applied.


More information about the Python-list mailing list