Jave the next GAE language? That's probably half right.

Chris Herron pointed me to the article where Michael Podrazik suggests that Java will be the next language for the Google App Engine runtime.

I think he's half right.

By that I think if there's any Java in the next GAE language, it will be JavaScript.

Why is that?

It's pretty clear that Google has a lot of experience in house with JavaScript. The GWT runtime is entirely in JavaScript. They have their own XSLT processor in JavaScript (for browsers that don't support XSLT natively). Also, they have their Rhino on Rails project, which is a Ruby on Rails port to JavaScript.

Next, JavaScript fits nicely in to the existing GAE infrastructure. It can be run just like Python is now. Also, there are several OSS JavaScript interpreters available to be used, of varying quality. The new runtime for FireFox 3 based on Adobes ActionScript is one, also the recently announced SquirrelFish runtime from WebKit could be used.

The GAE API would fit well in to a JavaScript world, with less "square peg round hole" work that using Java would entail.

JavaScript, with its push to JavaScript 2.0 is rapidly growing up. It's always been an elegant language with its prototype inheritance scheme (some would argue it's a blight on the planet, but that's more a paradigm complaint I think). The 2.0 changes will make it a bit more mainstream, make it faster, and even more powerful. So JavaScript is powerful today, but getting even moreso. The tooling surrounding it is getting better as well.

Finally, there are a bazillion web developers who are becoming, whether they like it or not, conversational in JavaScript. Before there was a clean separation between the client side and server side developers. Client side did HTML and CSS, while server side did scripting and logic.

But with the modern browsers having powerful JavaScript engines, and UI demands requiring fancier client side scripting for effects etc., not to mention Ajax, the client side developer has had the world of scripting and programming logic thrust upon them.

Some take to it well and become adept at leveraging JavaScript and its powers. Others simply cut and paste their way to success using the multitude of examples on the web. Either way, whether novice or expert, the client side developer is learning the fundamentals of programming and the nuances of the runtime through JavaScript.

If the client side developer were able to leverage that JavaScript knowledge on the server side, that empowers them even more.

JavaScript has had a mixed history on the server side. Netscapes server has supported server side JavaScript since forever, but obviously when someone thinks about the server, JavaScript is far from their mind. It has almost no mindshare.

Yet, we have, for example, the Phobos project which is a JavaScript back end, as well as the previously mentioned Rhino on Rails internal Google project. These are recent, though, without a lot of public history.

Now, to be fair, these are both Java systems operating as the host for a JavaScript based system. But there's no reason they have to be Java. The major browsers certainly don't use a Java runtime for their JavaScript systems, they use a C/C++ implementation.

With a C/C++ implementation, Google could readily launch a JavaScript runtime for their GAE that would fit quite well with their current infrastructure. Also, since there's very little momentum on the JavaScript server side, there's no real competition. No "why can't it operate like Project X". This gives Google even more freedom in terms of shaping the runtime the way they think it should be done.

So, I think that if there is any Java in GAE future, the near term will be in name only, with JavaScript.

You heard it here first.