The Soul of Windows

After reading several replies to Mark Smith's article "The Soul of Windows," I can't resist formulating my own theory about the reason Microsoft changed the marketing focus from systems administrators to top executives. This is what I think: As many of us agree, Microsoft didn't take market leadership because they offer better products. Was Windows 3.0 better than Macintosh? No. Was Windows NT Server better than Novell? No. The Microsoft products succeeded simply because they were less expensive.

Microsoft's target used to be systems administrators. Microsoft was trying to convince administrators that its programs were good enough to replace the market leaders but at much less cost for the enterprise*so, we could make the appropriate justification (i.e., the Return on Investment—ROI—argument) to the finance group. Let's face it: We don't have any decision power in the enterprise. Or do we? Can any admins implement any high-tech project (e.g., Storage Area Network—SAN, wireless) without the approval of the chief financial guy? Our job is to evaluate, recommend, and, if approved, implement. Microsoft understood the buying process very well and still does. Now that the company has taken the enterprise environment, it doesn't need the systems administrators any more to justify its programs. Now it can go directly to that financial guy. The reason Microsoft is scared by Linux and the open-source initiative is not that the programs could be better than Microsoft's but because implementing and maintaining them could be much less expensive than using Microsoft products.

Until one company demonstrates that it can live with open-source applications alone, Microsoft's dominance won't be threatened. But Microsoft doesn't want to wait for that day. What does that mean for systems administrators? Well, in the same way we brought Microsoft to the enterprise, we can get rid of Microsoft by justifying the cost of using alternatives to Microsoft products. I have to admit that this task is more difficult than it was in the early 1990s. But it's not impossible. Install a small network in your home, test the applications, extrapolate the results, and calculate the cost. You might be surprised!

Mark Smith's "The Soul of Windows" (January 2003, http://www.winnetmag.com, InstantDoc 27392) was informative, but I would like to present an additional view. I wasn't a part of the initial Windows "club" when it first started, and I've been a network administrator for only 4 years. Microsoft isn't a club or a hobby anymore. It's a company. Why is this so wrong? The company was successful and made money. Most of the magazine's readers have jobs because someone somewhere made a profit and hired people to help their company make a profit.

Some of the readers responding to this article are looking for a warm and fuzzy club feeling from Microsoft. But considering the company's size, I think that's an unrealistic and possibly unfair expectation. I think Microsoft makes an honest effort to serve the customer and be helpful. I've found that Microsoft's technicians and customer representatives have great attitudes, and they're concerned about equipping me to help the people I support.

Many new IT employees in the market aren't looking for a club but for employment. I read Windows & .NET Magazine for helpful instruction and advice, not to be in a community. For most of us, it's how we get a job. It's not about being in a club. So let's just get to work.

I agree with Mark Smith's assertion that Microsoft has abandoned IT administrators. The first version of Windows NT that I booted up was the October 1992 beta version of NT 3.1. At the time, I was researching alternatives to Windows 3.1, which required frequent reboots, and I wasn't looking forward to the warmed-over upgrade that Microsoft named Windows 95. NT has been a good choice for my organization, and I think it has contributed to making our organization productive. At the time that I recommended our organization standardize on NT, I expected the licensing cost to drop after Microsoft succeeded in migrating home users to an NT-based version of Windows. I based my expectations on the industrywide trend for software costs to fall over time. That has never happened with Windows. I am distressed to see licensing costs rise as Microsoft gains a larger share of the business enterprise. In addition, Microsoft's onerous licensing terms practically force regular upgrades according to Microsoft's schedule, paying for Windows twice on every desktop. Microsoft has deliberately confusing licensing terms, sales reps who feign ignorance of licensing details (I speak from experience), and (worst of all) Windows Product Activation (WPA). I feel betrayed by Microsoft. I am so philosophically opposed to WPA that I plan to load Linux on my next home computer.

I've been a Microsoft support engineer since the release of Windows/286 and Windows NT 3.1, and I've subscribed to Windows & .NET Magazine since the first issue. I'm also a former Microsoft employee. In the last year, I've mostly returned to the UNIX environment where I began my computing career 12 years ago. Although I've felt slighted by Microsoft for the last several years, I made my decision to depart from Windows mainly because, as I told my supervisors, Microsoft doesn't understand computer security. The Klez and Bugbear worms are the latest examples of that lack of understanding.

Regarding the Windows community, in the past, Microsoft cultivated the community by holding technical presentations in major cities and giving attendees CD-ROMs of Microsoft OSs and applications. I got my certifications not by attending expensive classes but by setting up small networks at home and living with the code so that I could learn how to deploy and support it long before I installed it in the office. I passed the certification tests by living in an NT environment at home first and later at work. I still have the NT 4.0 three­CD-ROM pack that Microsoft gave to those who attended the IT conference when NT 4.0 was released.

Since then, Microsoft has stopped building the IT community that got NT into corporate offices. The company stopped listening to IT people and developed global products that have features that most offices never use and sometimes make the products poorly suited to small to midsized shops. For example, the integration of Microsoft Exchange Server with Active Directory (AD) prevents you from using a single Exchange Server system as a bridgehead to route mail to multiple Internet domains. If you're supporting several 100-user email domains, this "feature" makes Exchange 2000 Server an expensive choice. Exchange Server 5.5 was better suited to small to midsized offices, and I supported Exchange 5.5 at Nortel Networks, where we had 87,000 users at one point.

I know that Microsoft, like any other software developer, makes most of its money by licensing software to businesses. I have no doubt that it's easier to ensure license compliance in corporations, which use license-metering applications to ensure compliance, than in small businesses that are more concerned with surviving to fight another day. But small businesses—not global corporations—made Microsoft successful. It was Microsoft's determination to scale Windows and Windows applications to global corporation sizes that exposed the security flaws in the Microsoft platform and the scalability limitations in SQL Server. Because of these problems, most Windows administrators still choose to scale out, rather than scale up, and use someone else's (read that non-Microsoft) tools to secure the network perimeter.

Most of my contracts currently involve integrating UNIX, particularly Linux, into already existing Windows networks. However, one client was redeveloping its point-of-sales application to run on Red Hat Linux 7.0, offering a non-Microsoft alternative that's easier to support than Windows 2000 or NT in a highly distributed environment (convenience store chains). Linux looks poised to eat up the low end of the Microsoft customer base just as Windows did to Sun Microsystems, IBM, and Hewlett-Packard (HP) several years ago. The Linux community very much reminds me of the NT community of the NT 4.0 days. Has Microsoft lost its community? Yes, but it doesn't seem to care. The company seems to think that CEOs, not IT managers, decide whose software gets installed. This top-down marketing scheme is one more slap in the face of those who really do implement the technology that runs businesses, daring us to try to go to some other platform. OK, Microsoft, we will.

Regarding Mark Smith's Fast Forward: "The Soul of Windows," congratulations on saying what needed to be said. The author voiced the feelings of everyone involved with the trials and tribulations of Microsoft. I hope the company will sit up and take notice. In the meantime, I will continue to juggle my time between Windows NT and Linux, waiting to see who will come out on top.