Bill Veghte, vice president of Microsoft's Windows Server group, spoke with Computerworld's Carol Sliwa about the competition the company's newly launched Windows Server 2003 faces and about his thinking on future releases. Excerpts from the interview follow:
Q: Do you consider Linux to be the chief competition to Windows Server 2003?
On the one hand, NT 4 is a competitive framework that I look at. Solaris is, AIX is, and Linux is.
Q: Do you view them all equally?
Windows Server has three core roles: application platform, IT infrastructure and information worker productivity, (for) large enterprise, medium-size business and small business. For each one of those, I think of the competition as different.
Where Linux is getting traction is, "I've got a Unix app and I'm not doing any feature development on it. I want the cost efficiencies of x86. And I don't want to have to recompile the app in any significant way, etc. Linux looks interesting." (If) I've got a simple Web server that's just serving up static HTML, Linux looks interesting. Those would be examples of app workloads predominantly in the enterprise.
Q: Microsoft will ship some new functionality in the coming months as add-ons for Windows Server 2003. What do you think should be free, and what should be separate, for-fee products?
The philosophy that I've had is anything that has been delivered in essence out of the box in Windows Server to date, that we're effectively improving on, is available for any licensed Windows Server 2003 customer.
An example of that would be Windows SharePoint Services. I think of Windows SharePoint Services as the next-generation file server. It would stand to reason that if I think of Windows SharePoint Services as a file server, and Windows servers are naturally file servers, that would be free for any Windows Server 2003 customer.
Q: Is there any chance you will revisit the decision about Exchange Server 2000 not running on Windows Server 2003?
Nope. Not on that one. That was a decision we made based on some hard engineering realities and what we believe was the best way to solve some customer feedback.
Q: Is this the pattern customers can expect for future versions of Exchange -- that they will run only on the newest operating system?
No. Boy, that statement I would hate to be true.
It was surprising just over a month ago when Brian Valentine, senior vice president of the Windows division, told us there might be a Longhorn server release, after Microsoft had said last fall that Longhorn would be a client-only release. I was shocked, too, I promise you.
Q: Is there any new line of thinking on that topic?
Yeah, I'll walk you through it, because . . . I've given Brian lots of feedback.
Here's how I think about that server release framework and what I'm trying to do for customers. When customers think about Microsoft, they want two things. They want a regular supply of innovations that are solving their problems in an integrated, comprehensive fashion. At the same time, they don't want it every 12 months. They want it at less frequent intervals. I sort of like the three-year time horizon, frankly . . where it's going to make good business sense for customers.
So I look at this and say, There's a set of things that I want to slipstream out in the marketplace for the customers for whom it makes sense and then do a big-bang server release every couple years.
The question that the company's been wrestling with is, Are there a set of slipstreams that we can do to the server to make those more relevant? Or are we doing a big-bang release? Frankly, we've had lots and lots of discussions on it. Now that (the launch is) done, it is my highest priority over the next couple months to get that locked in.