I was recently talking with a friend about the value of the traditional operating system in the age of the internet. I think the OS will be marginalized, but there’s still a lot of money to be squeezed out of that cash cow. Below is a quote from a letter written by J. Scott Kasten to the Department of Justice in 2002.
Although Microsoft has managed to keep an artificial floor on the value of their operating system products through monopolistic practices, even they realized that the inevitable pressures to marginalize the operating system would become too great for even them to bear. Thus they planned its obsolescence. The new target development platform of choice is going to be the .NET infrastructure. Ancient PC's had a BIOS containing the BASIC programming language/operating system that was permanently embedded in their ROM memory. As full fledged disk based operating systems came about, they marginalized the BIOS. None of the BIOS products these days has a built in programming language. It's only roll is to pull the disk based operating system in off disk now. It has no real apparent value to the end user of the system that rarely even notices the brief BIOS messages that flash by as the system boots up. No one programs to that interface anymore. Microsoft is trying to do the same thing to their own Windows operating system and replace it with .NET. Windows will become little more than a fancy video display driver. No one will program to it anymore. The .NET infrastructure will be the actual target for most future software development.
Silverlight 1.1 does allow cross-platform and cross-browser development, so the statement above seems to have correctly predicted, at least in part, some of Microsoft’s strategic direction. But I have to wonder about the more recent clamoring over all of the new Web “operating systems” (Web OS).
I took a whirlwind tour of Ghost, one of the new Web OSes, and it’s impressive. I like the idea of a virtual operating environment that I can access from any machine with an internet connection. But, why haven’t the heavy hitters like Microsoft and Google released their own Web OS? There’s been rumor of a Google Web OS for a long time, but I so far, no beta product. Both MS and Google have a most of the pieces and parts required for a Web OS. I don’t think the release of a Web OS by Microsoft would cannibalize its OS sales in the near term. If a Web OS is so compelling, why don’t we see something coming out of those companies who are most likely to succeed?
Back in the 90’s, when object-oriented databases where all the rage, Microsoft stayed the course with their relational databases. Looking back, that was a good strategic decision. Maybe a Web OS just isn’t necessary or terribly useful when you can pick and choose applications and services via the browser. Here’s a statement I kind of agree with:
I still don't fully get the whole Web operating system concept. Why run an OS inside a browser when your browser is running in an OS to begin with?
I understand the need to virtualize and to run your own applications with your own data on any machine, but why do we have to be beholden to yet another “operating system”? Why can’t we just mash-up the stuff we need? In fact, I think we can, but it still takes a little more work and time than the average user would like to invest, and it may currently be beyond the skill level of most users. (Popfly anyone?) If so, maybe the Web OS is just a good way for smart people to spend their time so that they can’t create something that’ll really threaten the big guys. I dunno, just wondering.