task.js

• What? Cooperative concurrency for ES6
• Why? Who says JavaScript I/O has to be ugly?
• Where? http://taskjs.org
• When? As soon as your JS engine supports generators!
• How? http://taskjs.org

task.js provides an automatic task scheduler along with a library of first-class, synchronizable events, making it easy to do I/O without callbacks.

With task.js you can write non-blocking I/O in a synchronous style, even with error handling:

spawn(function*() {
    try {
        var [foo, bar] = yield join(read("foo.json"),
                                    read("bar.json")).timeout(1000);
        render(foo);
        render(bar);
    } catch (e) {
        console.log("read failed: " + e);
    }
});

Compared with callbacks:

var foo, bar;
var tid = setTimeout(function() { failure(new Error("timed out")) }, 1000);

var xhr1 = makeXHR("foo.json",
                   function(txt) { foo = txt; success() },
                   function(err) { failure() });
var xhr2 = makeXHR("bar.json",
                   function(txt) { bar = txt; success() },
                   function(e) { failure(e) });

function success() {
    if (typeof foo === "string" && typeof bar === "string") {
        cancelTimeout(tid);
        xhr1 = xhr2 = null;
        render(foo);
        render(bar);
    }
}
    
function failure(e) {
    xhr1 && xhr1.abort();
    xhr1 = null;
    xhr2 && xhr2.abort();
    xhr2 = null;
    console.log("read failed: " + e);
}

...tasks can be a lot simpler and cleaner. And unlike pre-emptive threads, yield always makes it clear where tasks block.

Contributing

Currently the best way to contribute is to hang out on IRC: the channel is #task.js on irc.mozilla.org. Or you can always send me email (my Github nick at mozilla.com). And I'm always happy to accept pull requests!

If you're looking for interesting things to work on, check out the issue tracker.