Why Windows 2000 Faces an Uphill Battle for Adoption

When I took over the helm here at ENT, Windows NT 5.0 was the next great thing that was going to happen. Instead, it now looks like Windows NT, which as you now know is being renamed Windows 2000, may face an uphill battle to gain industry dominance.

Granted, Windows 2000 Server, Advanced Server and Datacenter Server editions are a pretty significant jump over NT 4.0. By most accounts, Windows 2000 will ultimately have between 30 million and 40 million lines of code, up from the 10 million lines that NT 4.0 has. Taking a pessimistic view, that means that two-thirds of the operating system code is brand new.

Microsoft has adopted a new tactic for promoting Windows 2000 -- one of lowering expectation levels. At a keynote speech delivered at Networld+Interop last month, Microsoft president Steve Ballmer stated, "Clearly, we still have a lot of issues to resolve, especially for customers who have high expectations from working with mainframes and well-honed Unix systems."

I thought Microsoft clearly positioned its operating system as more than a match for Unix. Wasn’t that one of the key points 18 months ago at Scalability Day? Besides, this new approach really seems out of character for Ballmer in the first place, considering the aggressive, take-no-prisoners approach for which he is well known.

Microsoft says that it is beginning to reset its measurement of expectation levels closer in line with the expectation levels that enterprise users really have. While this is an important step in the right direction, it also highlights the barriers that still stand in front of Windows 2000 before it will be accepted in large companies for true mission critical work. These factors will slow the adoption of Windows 2000 considerably from what it could have been had Microsoft dealt with some of these scalability and reliability issues with Windows NT 4.0.

Nevertheless, there’s little doubt that your BackOffice server box will be a good candidate to upgrade to Windows 2000. However, timing being what it is, some IT managers will probably hold back on introducing a new operation system into their environments only months before the Year 2000 date change takes place.

But others will implement it before 2000, and Windows 2000 will probably move into mainstream use – on the server side – as soon as the first Service Pack is released. If Windows 2000 is released in the early part of 1999, that could happen by the third quarter.

Where Windows 2000 will see the least resistance is on the power users and professional developer’s desktops, and for organizations needing good security. On the other hand, this is hardly a definition for reaching the masses.

Microsoft’s bread and butter are the personal productivity business PC user and, ultimately, the home user. This clearly will be a tough nut for Microsoft to crack with Windows 2000. It is clear that Windows 2000 has the potential to take us to nirvana and true zero admin for Windows. But at what cost?

The stated minimum resource requirements for Windows 200 Professional are a 166 MHz Pentium along with 32 MB of memory and 300 MB of disk space. Realistically, this configuration won’t serve as anything more than the most basic starting point -- a minimum configuration for the operating system to work, not to work well. You’ll probably need to double the memory and disk requirements to actually run some real applications.

Here’s the problem: Unless there’s a compelling reason to make an investment to upgrade the majority of your existing desktop systems, Windows 2000 won’t be practical. Likewise for TCO, ZAW and nirvana. At ENT’s parent company, we’ve got a collection of machines ranging from 75 MHz Dell deskside Pentiums, HP Vectra 133 MHz Pentiums, and a handful of laptops sporting processors as powerful as 266 MHz Pentium Pros.

I am one of the lucky few to have one of those 266 MHz laptops. What do I use it for? Mostly for running Word, Internet Explorer and, about 5 percent of the time, Excel and PowerPoint. I doubt we’ll be replacing all of our machines just so we can have Windows 2000 on our desktops to run Word and Internet Explorer.

Are you going to upgrade all your PCs just so you can roll out Windows 2000 -- or will you replace them? Or will you just leave a perfectly acceptable operating system and application set -- albeit, one that’s at times difficult to manage -- in place for a few more years.

Windows 2000 will ultimately succeed, without a doubt. Microsoft has proved that time and again it is willing to stick with its plans to conquer a market, and to stick with it even if that objective takes years to accomplish. Given the hype and excitement and literally years of drumming up interest, the actual adoption of the product can’t be anything but anti-climactic.