I just watched a nice Tuts+ tutorial about server.on. It's an eventlistener, in that when the server gets a certain event (specified by the first argument), it calls a callback function (specified by the second argument).

If you are listening for a request event, be sure to include res.end(); at the conclusion of your callback function to keep the client from hanging on, endlessly waiting for a request.

.on is a method that the http module inherits from the EventEmitter class.

Here's the sample server I made using the Tuts+ tutorial:

var http = require('http');

var server = http.createServer();

server.on('request', function(req, res){
  console.log('got a request');
  res.end();
})
server.listen(8000);

You can use the node documentation to see what events are emitted for different objects. See the bullets that start with "Event:" on this page, for example: https://nodejs.org/api/http.html

Closures

To include a closure, pass in your function as an argument but wrap it in parentheses. That way it'll call the first function and close over, then call the code again and run the internal function you included in the code.

Example:

var events = require('events');
var em = new events.EventEmitter();

// sets up the event listener
em.on('timedEvent', function(interval){
  console.log('Interval: ' + interval);
});

// sets up the event emitter
setInterval(function(){
  var i = 0;
  return function(){   //first function layer frozen
    i++;
    em.emit('timedEvent', i);
    }
  })(), 10000)    //note the () calls the first arg

I assumed the function setInterval already exists natively in the environment and just takes a function then an interval as arguments and calls the function every interval. A quick Google-search confirmed my hunch.

So why use a closure here instead of a normal function? I think it's to freeze our i variable into existence in a new, safe namespace.