How easy is it to fake asynchronicity? [closed] - asynchronous

Question

Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 8 years ago.
Improve this question
Clearly I don't understand the big deal about "asynchronous" environments (such as NodeJS) versus "synchronous" ones.
Let's say you're trapped in a synchronous environment. Can't your main loop just say:
while(1) {
events << check_for_stuff_from_the_outside_world();
for e in events {e.process()}
}
What's wrong with doing that, how is that not an asynchronous environment, how are asynchronous environments different?

Answer 1

Yes, this is more or less what Node.js does, except that instead of check_for_stuff_from_the_outside_world(), it should really be check_for_stuff_from_the_outside_world_plus_follow_on_stuff_from_previous_events(); and all of your events must also be written in such a way that, instead of completing their processing, they simply do a chunk of their work and then call register_stuff_for_follow_up(follow_on_event). In other words, you actually have to write all of your code to interact with this event framework; it can't be done "transparently", with only the main loop having to worry about it.
That's a big part of why Node.js is JavaScript; most languages have pre-existing standard libraries (for I/O and so on) that aren't built on top of asynchronous frameworks. JavaScript is relatively unusual in expecting each hosting environment to supply a library that's appropriate for its own purposes (e.g., the "standard library" of browser JS might have almost nothing in common with the "standard library" of a command-line JS environment such as SpiderMonkey), which gave Node.js the flexibility to design libraries that worked together with its event loop.

Answer 2

Take a look at the example on the Wikipedia page:
https://en.wikipedia.org/wiki/Nodejs#Examples
Notice how the code is really focused on the functionality of the server - what it should do. Node.js basically says, "give me a funciton for what you want to do when stuff arrives from the network, and we'll call it when stuff arrives from the network" so you're relieved of having to write all the code to deal with managing network connections, etc.
If you've ever written network code by hand, you know that you end up writing the same stuff over and over again, but it's also non-trivial code (in both size and complexity) if you're trying to make it professional quality, robust, highly performant, and scalable... (This is the hidden complexity of check_for_stuff_from_the_outside_world() that everyone keeps refering to.) So Node.js takes the responsibility for doing all of that for you (including hadling the HTTP protocol, if you're using HTTP) and you only need to write your server logic.
So it's not that asynchronous is better, per se. It just hapens to be the natural model to fit the functionality they're providing.
You'll see the asynchronous model come up in a lot of other places too: event-based programming (which is used in a lot of GUI stuff), RPC servers (e.g., Thrift), REST servers, just to name a few... and of course, asynchronous I/O. ;)