As the delivery date for Windows NT 5.0 - oops, Windows 2000 - gets later, Microsoft seems to be daring corporate customers to look at other server OS platforms.
In the past, many customers didn't see a viable alternative - Unix was too fragmented, Sun Solaris didn't talk to NT, and Linux was too new to consider. But Microsoft's seemingly lackadaisical pace should quicken rapidly after two Unix-related announcements last week from Sun Microsystems and IBM.
Sun announced the latest version of its server OS, Sun Solaris 7.0, and IBM, SCO, and Intel said they intend to deliver a 64-bit Intel- optimised Unix version called Monterey. These two announcements give corporate customers the option that their users and Web sites have been demanding.
It's quite ironic that Microsoft would be so flagrant about its OS slippage. For many IT departments, the challenge associated with integrating Web demands with existing client/server, legacy, and enterprise resource planning systems has forced the departments to push more processing to the server.
For many, NT, er, Windows 2000, would be a natural choice. But chronic delays are defeating that potentially strategic move.
To give context for "chronic", am I the only one who remembers that "Cairo" was initially announced for an early 1997 delivery? Now that Microsoft has publicly said Cairo's Beta 3 won't be available until the first quarter of 1999, this means a summer delivery - and that Microsoft may be too late to the game.
The teaming of IBM, SCO, and Intel to deliver Monterey presents a big challenge for Microsoft. Their plans call for this OS to be delivered well ahead of a 64-bit version of Windows 2000. Intel's presence creates some real challenges for Microsoft's apparent dominance on the Intel platform. Adding to the challenge is the fact that SCO and IBM are already shipping the 32-bit version today, and they plan to have an enhanced 32-bit version available by summer 1999. By bringing some of IBM's AIX features to the Intel architecture, SCO stands to offer a serious threat on the low end, especially because many hardware vendors stand ready to support this open offering on their products. But Sun continues to be the largest threat to Microsoft's planned server dominance. Although Sun has always been a threat with its high-end offerings, the features included in Solaris 7.0 surprised many - especially Microsoft. With enhanced NT interoperability in Solaris 7.0, corporations will be able to swap out their NT 4.x servers and replace them with Solaris 7.0 servers. The client PCs won't even see the server swap.
The easier-to-use administration tools will mean that support personnel won't need Unix jocks to manage the system. Even though Sun has been slower than other vendors, such as HP and IBM, to embrace the high-end 64-bit environment, combining its Web applications reputation with NT compatibility puts Sun back as a leading high-end contender.
Adding even more insult to Microsoft's delay is the discussion that Cisco may release its Solaris version of Active Directory early in 1999. That's right: Windows-oriented environments could get the key part of Microsoft's Windows 2000 offering on a Sun box many months before Microsoft releases its version. Because Cisco has exclusive rights to develop Active Directory on Solaris, it seems Microsoft can't do much to slow down this possibility. But with Microsoft, I won't rule anything out.
Although it would be too easy to take pot shots at the Windows NT naming changes, it is clear that Microsoft has bigger challenges than names. It is imperative that the company establish itself on the server, but delays are hurting its strategy.