Paul Flessner, former in-the-trenches IT guy and now senior vice president for server applications at Microsoft, wants to run the data center. His opening punch comes in November with the release of SQL Server 2005, Visual Studio 2005 and the beta for BizTalk 2006. Flessner recently sat down with Network World Senior Editor John Fontana to talk about what lies ahead for Microsoft.
What impact will Visual Studio, SQL Server and BizTalk have on the enterprise?
We want to declare victory on the largest mission-critical apps in a company. We are tired of being pigeonholed into the Tier 2, Tier 3 apps. I want to run the enterprise. We think 2005 from both a developer Visual Studio perspective and a SQL Server perspective and also BizTalk, those are the releases that give us the credentials to take on most mission-critical, trading-floor applications, or power plant or distribution [applications].
Where is the sweet spot for the enterprise with business intelligence (BI) and the business activity monitoring (BAM) technologies you have talked about with SQL Server?
BI has come into its own. If you look at the revenue growth numbers for BI, they are off the chart. The relationship with your customers has changed. It used to be they came into the store, you could touch them, you could watch. Today, it is virtual in a lot of ways.
It is online. You need to be able to see trends and identify them, and the way to do that is through your data. Having integrated BI is compelling and we have invested a lot in it. Having the same thing for business process is compelling as well. BAM will give you the ability to look into your business processes as they occur. And that is a technology that today you have to handcraft. Data used to be this thing locked up that IT controlled. Data is on the loose now.
Where are you going to build RFID into the platform?
We'd love for the reader manufacturers to pick up Windows CE. Some will, some won't, but RFID is really about being able to grab the data off the wire from the reader. Right now that is a wild market. Over time we think standards will develop and you will get standard protocols. We'll enable a plug-and-play device driver architecture and allow you to have a driver for any reader. Basically, you can hook Windows up to RFID, RFID-enabled Windows, whether it is a desktop or a server, where you can take that data, get access to it through an easy object model and manipulate it and do anything you want with it.
What relationship will SQL Server and WinFS have?
WinFS is an important milestone; it really is focused on structured and unstructured data. It will allow an unprecedented ability to have data that resides on your desktop, the ability to get access to that data, to program against that data and to use that data both on the client and the server. Probably the most telling way to look at it from a high level is that WinFS is built to be a file system and think of SQL Server focused today on line-of-business applications, really running on the server.
Why should IT be in tune to this Common Engineering Criteria, for which you've announced the 2006 additions, and how it is built into all new versions of your servers?
Any feature that I put into any product has a fleeting life in terms of competitive advantage. The open source community can copy any feature and they do today. But customers want to run their app, they don't want to think about the infrastructure underneath. They want the infrastructure to integrate seamlessly, repair itself, manage itself. That is what the Common Engineering Criteria is. Even though it has humble beginnings, you are looking at a multi-decade journey for us. But it is causing a behavioral change.
What makes the Windows Server System start to look more like a whole than it has in the past?
It will take time. Integrated diagnostics where I can really understand what is going on, that is really compelling. Integrated serviceability, Windows Update, common setup and installation technology, best practice analyzer, a common management [user interface]. There will be a set of things that over time we will go for.
What sort of impact has the Longhorn delay had on this, wrapping all this server and developer technology together?
We have had some very ambitious plans and we have pared them back some. Longhorn will still be very compelling. There is still a lot of cool [user interface] experience, search, a lot of what you'll see at [Microsoft's Professional Developer's Conference in August] has been talked about, but the demos are compelling. That is the Longhorn show.
How do you classify the Linux challenge now?
It is maturing. It seems, from a database guy perspective, it is settling out. They are finding their niche. We now know how to compete with them.