The biggest challenge they face is convincing their NT 4.0 base to move up to Windows 2000 or Windows Server 2003. It's proving to be a daunting task. A lot of customers are looking at Linux despite all of [Microsoft's] pleading and prodding.
It's a huge issue for Microsoft because it's a large amount of their customer base. [Microsoft says that somewhere between 60% and 70% of its installed base is still on NT]. These people are at a crossroads.
There are alternatives to NT 4.0. Linux continues to be a serious challenge to them in terms of the marketplace and rhetoric. More people are looking, but the deployment will only happen in niches. It's a longer challenge, but it has started.
How close is Microsoft's road map to the realities of customer adoption?
According to Microsoft's road map, there's not another big release for three or four years. At the same time, Microsoft has technology to be competitive. One of Microsoft's biggest struggles will be packaging and releasing updates to its operating system. There are new features for Active Directory; how will they release that?
Microsoft is trying to find a way to issue more product releases between big releases. Historically, this is not their strength. Will they be free? For a charge? It's critical for them to have these supplemental releases between releases or else they won't remain competitive.
How about customers already on Windows 2000? Do you expect them to upgrade?
People on Windows 2000 are saying 'why should I bother [upgrading to Windows Server 2003]?' Microsoft hasn't made a case for why customers should upgrade. The most improvements come from Active Directory and enabling new hardware. There is increasing interest in running mixed workload applications. Some people are looking at [Windows] 2003 as a workload manager.
Windows 2003 has the potential to change how servers are deployed. Today, its one application per server, but workload manager and virtual server will let people deploy more applications per server.
Are there any areas where you see Novell re-emerging as a significant competitor to Microsoft?
As a [network operating system,] Novell holds its own, but I don't see them as a rising competitor to the Windows operating system. They have some features in [6.5,] but it's not enough to move the market back in their favor. They're not as much competition as Linux.
But on the directory side, they have the meta-directory and identity management solution. Novell competes with Microsoft on that front.
Microsoft recently made public a schedule for product life cycles. Do you think they are fair?
They are fair compared with what some other vendors are doing. The only big gun against people's heads is the NT 4.0 end of life.
Microsoft, to its credit, has rolled out life cycles for all its products. Customers can see when the products they use will be made obsolete. Moving forward, it will be easier for [customers] to come up with transition strategies.
Will new software releases mean customers must spend more on hardware?
Windows Server 2003 has formal support for Itanium, so larger databases are possible with Windows. There is support for IBM's xSeries 440 and, shortly, support for AMD 64[-bit processors]. But enterprises don't have to go to those platforms. The price-performance of current servers is so good there is not a lot of incentive for customers to buy into these next-generation products.
Windows Server 2003 won't push people to buy new hardware unless you are still on NT 4.0. But, if you do have to buy new hardware, you will find there are more choices than ever before.
When does Gartner expect customers to start spending for significant projects again?
We think 2003 is a hold year, with flat spending. We may see increased budgets in 2004.
Meet John Enck, and hear his keynote address, "Understanding Microsoft's road map," at Enterprise Windows Decisions, May 14-16 in Chicago