Anne Zelenka is sceptical about RIAs:

The idea is that we need more rich interactivity from our browser apps than they give us. But is this just developer fantasy, or does it represent a real end user need?

It’s a great question. I believe it’s fair to say that the all the interest in RIA, sparked by Flash and enflamed by Silverlight, is still more hype than real-world usage (especially Silverlight, still in Alpha for the .NET version).

There are multiple issues here. In particular:

• Will we see HTML/CSS/JavaScript (call it AJAX if you like) gradually giving way to browser-hosted apps running in plug-ins (Flash, Silverlight, Java)?
• Will we see a new breed of internet-delivered, zero-install desktop apps that will diminish our dependence on web browsers?  

I have few doubts about the first of these. Ease of development, flexibility and predictability of design, performance benefits of JIT compilers, convergence between internet and broadcasting, richer content enabled by ubiquitous broadband, to name some of them. 

The second is more contentious. But I think it will happen. There is room for debate about what constitutes a “real end user need”; but if you rephrase that as “real end user benefits” then it makes more sense. The main reasons are offline use and better integration with local OS services.

A while back a web app sceptic (I forget who) described the browser to me. “I call it Window”, he said. His point still holds. There is no need to do all your work within a browser box.