I am thinking about this because Ive just read Mary Jo Foleys story about "Bigtop," the code name for a future grid-computing operating system that my friend Craig Mundie is supposed to be working on. Thats just like him, starting a new project and not telling me about it. (Like he ever has.)
When Mary Jo told me she was writing the piece, my first IMd response was, "Due when? In 2020?"
It wouldnt surprise me that Microsoft was planning out that far. The company even shows some evidence of it in its Center for Information Work, an Epcot-like exhibit at its Redmond conference center.
There, you will find things like an applicationless desktop environment that sometimes seems to think for itself. Sounds like 2015 if you ask me.
One challenge is that Microsoft cant just start working to create such an environment; it also must create a strategy to get us from here to there. Which do you think is more difficult?
With its huge installed base and its fascination with backward compatibility, Microsoft always moves slowly. Add the problems its having even building a future—lets use Longhorn as an example—and the process slows almost to a halt. Having to build in the security features necessary before computers should be allowed to play a much larger part in our lives sometimes makes progress seem to move backwards.
Microsofts present looks less bright to me now than at any time in the companys history. Some of that, of course, is a reflection of a maturing industry, and another factor is that as organizations get bigger, they tend to slow down. At the same time, Microsoft has wiped out most of the competition that could force it to innovate more rapidly.
I am tempted to think of Linux and open source as just a sideshow, even a distraction, on the road to world of more useful and user-friendly computing. Sure, Linux is cheap, but does it really expand the application set? Do computers do more things now that Linux is around, or do they merely do them less expensively?
Now, theres nothing wrong with lower cost, and a case can be made that increasing computer access is today more important than increasing computer power. But Im not the one to make that case, and anyway, Microsofts future involves computers that are both more powerful and easier to use.
But as Ive already said, the challenge is getting from here to there. And while Microsoft is often criticized for not producing useful roadmaps, Im wondering if the reason for that is the lack of any roads. What if there simply isnt a direct route from here to there?
I sincerely hope that somewhere buried deep in Microsoft Research are people locked away, working in secret on operating systems that dont have to be backward-compatible. Sure, it would be nice, but maybe Microsofts acquisition of Connectix will generate a Windows emulator we could use.
This new OS needs to be more secure, easier to use and more productive than what were using today. It needs to close the gap between desktop, user and applications as well as between clients, servers and the world at large. Basically, Im talking about the operating system and architecture wed build today if we were starting from scratch.
Heck, maybe this OS would even use a Unix kernel, as Apple has chosen. I dont think you have to reinvent the wheel just to reinvent computing.
Id like to think "Bigtop" is a reflection of Microsofts desire to do something really new and different for its customers. Its not as new and different, perhaps, as Id like. But for Microsoft, the concept of grid computing represents a significant change in thinking.
Even larger changes will be necessary if Microsoft really is going to take us from today to tomorrow anytime soon.