That aside, the web itself is not ready for the next stage of development, where we move into more stateless applications. Google Docs/Spreadsheets is a good start, and AJAX is a great tool, but JavaScript was never designed for pushing this kind of information, and current browsers aren’t really suited to data intensive AJAX functions either. Before advanced web applications and advanced services that actually utilize the client as more than something of a smart terminal and utilize its data processing power appropriately can be made, the frameworks used to deliver all this over the web need to be updated with smart data delivery in mind.

Presently, I’m in favour of using client side applications that handle external data themselves, in whatever proprietary method the developers felt was best suited, because currently that is the best and most efficient way, and currently there is no universal framework that can cater to everyone’s needs. I’m sure one will come about in time, but it’s some way off yet.

Now, moving on to the real point made by the author. I completely agree with Vik in that we’ve spent the last 10 years ramping up our CPUs, VGUs (and other components) in a bid to facilitate demanding applications. The reality is that the general consumer does not typically need 3GHz or 256mB video RAM.

Sure, we have a few savvy consumers (let’s exclude the techies because, well, we can all find a way to use up 100% CPU or RAM if the challenge arises, even from Bash or Cmd) who not only manage to rotate photos but also encode videos and watch DVDs, thus successfully managing to utilise a PC like a VCR.

Most applications today are still plain old windows with static buttons and drop-down menus, and their expected output is not much more spectacular than a well crafted colourful Excel spreadsheet filled with sorted grids and meaningful numbers.

Operating system developers are of course pushing development of visually rich environments but this is happening way too slowly and probably for the wrong reasons (flashy ads are not why one should develop a good looking OS). Try convincing a small company (say 10 PCs) or even a savvy home user to upgrade to Vista. With XP SP2, free Windows Media Encoder and a bunch of other tools that they’ve learnt how to use – the users today have all they need (or think they need). XP has had more its bugs fixed by now, Vista has over 100 patches to be zipped into SP1 – and users feel that grown stability of their existing OS. In contrast, an OS change = painful setting backups, document backups, reset application preferences, reinstalling of applications and most of all the fear of having to learn new ways of doing things.

What we need is a near stateless operating system, and near stateless applications. That’s right, the OS should be an interface between the program you want to run, and the hardware that you have. What kind of an interface is it, if it’s irreplaceable? If you say this is technically impossible, I say – look at VMWare, Zen, Virtual PC. An OS, as far as the user is concerned is the location of his menus, and the colour of the login screen – not much else. We should be able to write applications today that run on any PC, full stop.

From a technical system administration standpoint this does present new challenges, but also new solutions to existing problems. Let’s not go into this field at this time, not because it’s not worthy, but because this post is not about this. I will be happy to discuss the implications of a stateless OS on large-scale system administration in a separate blog/discussion.

Coming back to the original point, yes, these applications will need to take care of visual presentation and other bells and whistles themselves. But so they should – each application should try to be more accessible, friendlier, and more intuitive. Why should the application developer stop at the thought that “the OS will take care of all that stuff”?

Only then will we have programs that obey our speech, follow our hand/glove (Minority Report reference), and present results (whatever they are) the best possible way we would want them to be presented, considering the available hardware.

Too difficult? Well, it is happening on the Web. This is where ’semantic web design’ originated. What does it mean? It means “present the data in the basic form, but if the environment supports JS – use it; if it supports animations – use them; if it supports asynchronous data calls – use them; if it supports fancy visuals driven by Flash – use them!” But even here we feel the lack of interaction and compatibility – why is it that we still need to use Flash to let the user see a pretty font; why is it that web developers have to write hacks for compatibility with the browser?

It is time. Time for Microsoft to include the Square Root button in the Scientific mode of Windows Calculator. Time for Linux, an environment with so much freedom and potential, to have more menus and options that the Windows Control Panel. Finally, it is time for Apple, the now permanent runner-up of Windows, to contribute to development of a stateless OS for just one reason – so more people can plug-in MacOS on their PCs (DVD, memory stick, a wireless download – who cares how – make it happen – invent).

That’s a pretty misleading title and even the title you suggest afterwards is misleading. I think you should have made a post about the possibilities of using every resource the computer has to improve web applications.