Time for a change

This is the last installment of my column that will appear in InfoWorld. After today's print edition and the coming online publication of my final May 5 column, I'm converting Window Manager to an e-mail newsletter. To continue receiving my findings on graphical operating systems, you will need to visit www.briansbuzz.com and subscribe to my new publication, Brian's Buzz on Windows.

When I started the Window Manager column in 1991, the computing world was on the verge of a major shift. It wasn't at all clear that the molasseslike Windows 3.0 — the first version of Microsoft's "pretty" interface that companies could really consider rolling out — would ever replace the faster, simpler, character-mode DOS.

In hindsight, we all know that the transition was an enormous success for the upstart that became the world's largest software company. And now, 12 years later, Microsoft is preparing to throw the world an even bigger curve ball.

I'm talking about Windows circa 2005, code-named Longhorn. I don't usually write much about vaporware because I think readers prefer to learn about products they can actually buy and use today. But I'll make an exception in this case. The next Windows will be a big, big change.

One huge difference is that Microsoft plans to introduce an object file system known as WinFS (Windows Future Storage). This data store will have full database functionality built on SQL Server enhancements code-named Yukon.

The benefit will be that different applications will be capable of accessing data in the same way. Searches you perform on a company name, for example, will find not just documents but e-mail messages, images, media clips, etc.

The downside is that most existing Windows applications will be incompatible with Longhorn and will have to be rewritten. As a result, Microsoft has already committed itself to overhauling Office, Outlook, and its other major products.

This may sound like a wrenching switchover. But I believe many enterprises will make the shift — the same way they dropped DOS and went to Windows — if Microsoft sweetens the deal with real security features. A hardened, secure-by-design Windows that didn't need weekly patching and was truly resistant to viruses would look very attractive to lots of exhausted IT pros.

At the same time, Longhorn may enhance Microsoft's already strong lock-in factor. If the new product won't run Windows apps, it may be renamed so that it wouldn't even be Windows. Microsoft may then be free to embrace and extend without any old antitrust rulings applying.

Because migrating to Longhorn will mean buying or building a lot of new software anyway, many companies may take a hard look at an alternative: switching to open source. Whether you consider this prospect to be aggravating or exhilarating, it's a debate you won't be able to avoid.

I hope to guide you through all this and to unveil hot topics as they arise, such as this week's launch of Windows Server 2003. To keep in touch, use the simple sign-up form at my site.