Stop page execution like the alert() function

Yes it's possible, i did inaccurate and not very well tested demo which does this.

Main concept:

1. in this example we have method Login.Try() which is executing Login.Proceed() method. Login.Proceed() makes AJAX query and we would like to wait for its execution, but don't want bind any handlers (just wait for it as window.alert() does)
2. instead of direct function's execution Login.Proceed, we use async() and await() methods (like in C#)
3. when we need to "pause" script and wait for something, we stop method execution using throw, and parse caller's function, to run its second part when waited (async) method has completed its execution.

What was left out of scope:

1. Clean code
2. Test solution for different browsers
3. Save/Restore local variables
4. Not works for loops.

Demo:

<script src="http://code.jquery.com/jquery-1.9.1.min.js"></script>

<script>

Login.Try(); // START!! START!! START!!

var Login = {
    Url: "http://xxxx",
    Try: async(this, function (T) {

        console.log('before login');

        //var success = call(this, Login.Proceed); // normal call
        var success = await(this, Login.Proceed);  // that we want!

        console.log('after login');
        console.log('success ' + success);

    }),

    Proceed: function (callback) {
        console.log('before ajax');
        $.ajax({
            url: this.Url,
            context: document.body
        }).done(function () {
            console.log('after ajax');
            callback("role=admin");
        });
    }
}


function async(T, method){
   console.log('before async create');
   return function () { return method.apply(T); };
   console.log('after async create');
};

function await(T, method) {
    var fn = arguments.callee.caller.toString();
    var pos = fn.indexOf('await(');
    var allBeforeAwait = fn.substring(0, pos);

    var pos1 = fn.indexOf('await(');
    pos1 = fn.indexOf(',', pos1) + 1;
    var pos2 = fn.indexOf(')', pos1);
    var cc = fn.substring(pos1, pos2);


    pos = allBeforeAwait.lastIndexOf(';');
    var allBeforeCall = allBeforeAwait.substring(0, pos + 1) + "}";
    var callResult = allBeforeAwait.substring(pos + 1);

    var result = 10;
    var allAfterCall = "("+fn.substring(0, fn.indexOf(")")) + ",V){" + callResult + "V;";
    pos = fn.indexOf(')', pos) + 2;
    allAfterCall = allAfterCall + fn.substring(pos)+")";

    //uncomment to see function's parts after split
    //console.debug(allBeforeCall);
    //console.debug(cc);
    //console.debug(allAfterCall);

    method.apply(T, [function (value) {
        console.log('ajax response ' + value);
        eval(allAfterCall).apply(T, [T, value]);
    } ]);

    throw "";
};

</script>

Hope this demo will inspire you with some ideas.

Also, you can take a look on http://blogs.msdn.com/b/rbuckton/archive/2011/08/15/promise-js-2-0-promise-framework-for-javascript.aspx

You can use following code to pause execution for long period of time.

function PauseExcecution() {
//Do some stuff
sleep(1000);
//Do some stuff...

}

function sleep(milliseconds) {
  var start = new Date().getTime();
  for (var i = 0; i < 1e7; i++) {
    if ((new Date().getTime() - start) > milliseconds){
      break;
    }
  }
}

You can't. Only the special built-ins can do that. For a while there was the showModalDialog special built-in that let you specify a URI for the content and thus customize it, but it was never widely supported and is now deprecated even by browsers that once supported it.

Instead, make your current alerting function that uses the div accept a callback for when the alert is closed (or return a promise that's settled when it's closed), to allow you to continue processing.

So for instance, if your code used to use alert and work like this:

function foo() {
    var x;

    x = doSomething();
    alert("Alert! Alert!");
    doSomethingAfterTheAlertIsCleared(x);
    doAnotherThingAfterward();
}

...you'd change it to:

function foo() {
    var x;

    x = doSomething();
    fakeAlert("Alert! Alert!", function() {
        doSomethingAfterTheAlertIsCleared(x);
        doAnotherThingAfterward();
    });
}

Note that now all the code that followed the alert is in a function, whose reference we pass into the fakeAlert. The foo function returns while the fake alert is still showing, but eventually the user dismisses the fake alert and our callback gets called. Note that our callback code has access to the locals in the call to foo that we were processing, because our callback is a closure (don't worry if that's a fairly new and/or mysterious term, closures are not complicated).

Of course, if the only thing following the alert is a single function call that doesn't take any arguments, we could just pass that function reference directly. E.g., this:

function foo() {
    doSomething();
    alert("Alert! Alert!");
    doSomethingAfterTheAlertIsCleared();
}

becomes:

function foo() {
    doSomething();
    fakeAlert("Alert! Alert!", doSomethingAfterTheAlertIsCleared);
}

(Note that there are no () after doSomethingAfterTheAlertIsCleared -- we're referring to the function object, not calling the function; fakeAlert will call it.)

In case you're not sure how fakeAlert would call the callback, it would be within the event handler for the user "closing" the alert div, and you just call the argument for the callback just like you do with any other reference to a function. So if fakeAlert receives it as callback, you call it by saying callback();.