But really, what is the point of Web browser?
Originally, it’s purpose was to format and properly display the documentation which used HTML to mark it up. The first browser – sir Tim’s World Wide Web – didn’t even support images, and tables weren’t introduced until years later.
It’s funny, really. The Web has brought a revolution, providing standards that allowed anyone to produce network-distributed applications with relatively little effort. Its standards are simple to understand and implement, and they require no tools apart from a text editor.
But when it comes to full-scale applications, the browser still leaves a lot to be desired. Its first disadvantage is that you still need to send the whole user interface along with the data, creating a large overhead. There are some advantages for browser-based applications – like no need to update all the client software – but even that is on one side solved by services like AppStore and on the other made moot by the constant upgrade of browsers.
In my opinion, the browser is going the way of the command-line interface. It will always be here (there is console even in Windows 7), but it will be used less and less, only for those who need to quickly set up an Internet-based application or as an entry point to a larger company.
However, we will see more and more native applications for all platforms which will present a nice interface to the user but will heavily rely on a Web-service powered communication with the main service. The best example for that are all the Twitter clients (which are according to some research used by 85% of all Twitter users), which have no other purpose but to serve as a front-end to Twitter. Many Web sites have developed native clients for their data – mostly for iPhone and other mobile devices, but increasingly there are ones for the desktop as well (and devices like iPad arguably start to erase the difference).