Just another WordPress.com weblog

Now that Google wave has been out for a while and all the hype has died down, just what was Google’s strategy with wave? It was a cool developer side project that was pushed to the forefront prematurely – why was that?

Hype aside, we have to admit that the wave team screwed up a bit. They built something that sounds nice but is slow and not usable – in doing so, they built a Microsoft product rather than the next big thing. However, I don’t think they screwed up too much. They acquired Etherpad relatively early, who did what the original wave team intended to do, only much better. With a fast, usable interface and a comet/polling server architecture that is actually responsive, wave will probably get the response that etherpad has had all along – wow, this is pretty awesome.

But Wave is much more than that. I think Google pushed it out prematurely because they believe that is where the next generation of websites is coming from. Wave is the first Web 3.0 app, and it centers around comet/server-side pushes.

First, lets look back at a particular side of Web 2.0 – AJAX. When Web 2.0 was first coined, arguments about AJAX’s usefulness flew all over the web. However, it wasn’t until good AJAX apps came out that it became something substantial rather than just a cool technique. Arguably, Gmail was one of those apps. It set the standard for javascript and ajax use successfully on the web.

Now after the kool-aid has worn off, its easy for a developer to see when, where and how you would use AJAX in your website. Its no longer the next new thing, but the standard.

Comet is a technique that allows server pushes, and it will be the main technology behind web 3.0. Its actually already in use in chat programs all across major social websites, but it has yet to propagate outside that realm. Etherpad is one of the awesome exceptions.

The reason that comet isn’t used right now (or even considered) by many is because the technology isn’t there. Etherpad is built on appjet, the company’s own technology (which they took offline – a shame, seeing the awesome things they’ve built on top of it). Ruby has eventmachine and python has twisted but neither of those has a newbie-friendly framework to work with (yet). Php is hopelessly lost and javascript’s server side development facilities and its event driven frameworks (nodejs) are still far too premature. But the whole comet ecosystem is building up fast. In 6-9 months, I think comet frameworks will be the next thing that first-mover developers play around with. And when the first good app (or set of apps) that uses comet for non-chat purposes comes out, the ecoystem will hit critical mass and will start to take off.

Google wave is already one of those apps.

Now, I’m not predicting that comet will take over the web as we know it – there were silly AJAX supporters who claimed AJAX everything, and that was just incorrect. I’m predicting that comet will open up a new web ecosystem, which is substantial and different enough from the current web ecosystem that is deserves a ’3.0′ tag. A lot of failures will happen initially: people experimenting with comet on everything causing mass confusion. But when people find what comet is good for (maybe pushing ads, dynamic data updates, embedded chats in social mediums, nextgen browser games, etc.), comet will be taken seriously as a new technology.

Personally I have my bets on javascript becoming the next big thing (in programming languages). It needs to get moving on its server side development efforts (go commonjs, go!) but its almost the perfect language for event driven websites.