I fully agree, and I'd like to add that this kind of fusion is already happening all over the place, blurring the line between what we used to call "thin" and "

I fully agree, and I'd like to add that this kind of fusion is already happening all over the place, blurring the line between what we used to call "thin" and "thick" clients. For instance, a browser running huge amounts of JavaScript (e.g. ExtJS and stuff) can't really be called "thin" client anymore. On the other hand, a slick native app that does little more than presenting nicely some weather data can't really be called a "thick" client. Also, a huge "native" Java application that depends on the web and has auto-upgrade mechanisms is in many aspects very similar to huge JavaScript applications that are cached by the Browser. Any application that needs to share data with other instances will need some server and some client part. The only question is where to put which part of the application. Is the client merely a displaying tool like VNC? Or does the client actually run application code, and the server is just some kind of database? Which parts of the data is processed server-side and which client side? Or, is the server part decentralized and the clients communicate via P2P? Maybe the application is even a combination of all those strategies, using in each corner of the application the most appropriate strategy. Given all those possibilities and the wide range of succesful strategies that can be observed in the wild, drawing a sharp line between "thin" or "thick" client is already impossible today.