Browser War? No way.

Roshan reckons that the browser war isn't over and that Mozilla has a fighting chance. Well, I'd like nothing better, if for no other reason than I actually wrote a very tiny part of the core Mozilla browser, but I've formed a very different view because of some work I've been doing recently. (I was going to write this at some point anyway, Roshan, but I thought you'd like another trackback!)

I'm writing a massively distributed application at the moment, entirely on top of the Microsoft .NET Framework. The Framework is a sandbox where bytecode is interpreted very quickly and efficiently at runtime (JIT compilation), just like the much older and more established Java technology. I think Microsoft have cleverly avoided making too much of the comparison with Java, instead presenting the Framework as an improved Windows API. But in reality, the only real difference between the two is that Java specifies the language, API and the bytecode, whereas .NET is specifically designed to allow a wide range of programming languages. (Java bytecode can of course be generated from non-Java code, though.)

So far the application areas have been quite different, with Java being widely adopted to provide safe applications -- applets -- within web pages, and also for general applications. .NET has been used primarily for business apps so far and is only really getting started -- it hasn't been bundled with Windows yet but is going to come with every copy of the upcoming Longhorn release.

Two things I've learnt recently: (a) you can write applets using the .NET Framework that can run within web pages on every Windows PC just like Java applets, and (b) Java doesn't come with Windows XP any more, due to legal agreements between Sun and Microsoft.

So the Framework does everything Java does, but the Framework is bundled with future releases of Windows and Java is not. The Framework will win, in my view, it's inevitable.

What does this mean for the brower wars? Well, it means that Windows Forms controls are going to make their way rapidly from apps onto web pages that the general public use. It means that rich web sites with really cool interactivity will only run on PC's that support the .NET Framework. And although there is a very serious independent attempt to support it on non-Windows PC's, Microsoft can't possibly allow them to succeed because it would undermine them not just in the browser market but in their core OS market too. They'll keep 'upgrading' to keep Mono one step behind, and they'll probably find some other clever strategy to prevent Linux being used to run Framework apps on business desktops.

So my view is that the Web is going to be killed as websites switch from pure HTML thin client technology to what's being called the 'smart client' approach which has all the advantages of a thin client (e.g. no-installation access to a company's services) plus all those of a thick client on top (e.g. rich windowed-application style interactivity). It's great for profitability of the companies that make the most of it -- and that's why I'm concentrating on it -- but it is very sad for the future of the Internet. Mono is the only hope.