George Belotsky has an interesting and forward-looking article at OnLamp about moving Linux to a web-based operating system. This approach isn’t particularly novel — there are some web-based operating systems in development — but Belotsky envisions moving a lot of what comprises a normal distro into web-based services like Google Apps.
One area where I agree with him is that the post-PC era is upon us and mobile devices will reign. Where I disagree is that Web 2.0 is going to be up to the task of fulfilling demands of individuals and businesses. As he correctly notes, many of the services now available are too insecure (and too unreliable) to be acceptable replacements for traditional software.
I also think browsers already define the desktop experience and will continue to become more important. Belotsky notes that there is talk in the Mozilla development community about extending XUL beyond the browser and onto the desktop (or mobile device desktop). They’re not alone. Sun has tried this before with Java without much success on desktops — but they’re increasingly important in mobile phone technology. I think Sun will continue to dominate as they increasingly open their code and put out products like JavaFX, the release of which I wrote about earlier today.
Belotsky also covers other existing tools that would allow for a hybrid web-based OS to be built atop Linux as a local client. The local client would feature the kernel and compilation of applications run locally and used to access those shared on the web, where applications could be used via the browser.
I agree with Belotsky that providers will be able to offer myriad set-ups to consumers. My concerns, though, remain focused on the security of such arrangements. Web-based services will have to both guarantee availability and provide adequate protection to their users — that’s something that will rely on a Catch-22 between consumers’ wanting secure web-based services and their willingness to actually pay companies enough to make it more secure (the false comfort of the status quo, while not ideal, is where most people will elect to remain because people are resistant to change). The Internet is already a hostile place when it comes to issues like data security and privacy. Moving everything to the web will only invite more invasions of privacy, so companies offering software-as-service will have to move quickly to establish credibility.