Data Centers

Until techs--not executives--make the technology choices, Windows will dominate

IT workers and executives will likely not see eye to eye on technology


ScOrp is TechRepublic's first opinion writer. His ideas and suggestions are his own and do not represent those of TechRepublic or CNET Networks.

Recently, I had an IM conversation with a Net Admin buddy of mine. This guy is in the process of switching from a mostly Novell-based operation to a totally Microsoft environment, migrating eDirectory to Active Directory and the like. You couldn't really label him happy about that. (In fact—and this IS a fact—people equally familiar with both systems tend to favor Novell by a large margin. That is, IF you call "100 percent" a mere "tendency.")

At any rate, the conversation involved a long-term favorite topic; namely, that Novell should develop and market a desktop OS to compete with Windows as part of its Server NOS and Directory Services offerings. Even in a complete NetWare server shop, all the end users ever see is Microsoft/Windows products. Most of them don't even know the name "Novell," and this puts a terrific crimp in Novell's sales at both ends. Face the facts—you buy what you know, whether you're a home user or a corporate VP with the main say-so over buying networking and application software. It has long been the contention of this writer and many associates that Novell is severely handicapped in the server market by virtue of having no presence out in the real world.

With the Novell acquisition of Ximian, this may change. Or not. Probably, it will depend on the price of the desktop OS (purchased freestanding, or included in the server/services package) and the price and structure of licensing server connections. (The truth is, Microsoft has made a LOT of enemies with its latest licensing policies.) Handled right, this move could save Novell from its slide into obscurity.

Here's an excerpt from my conversation with the Net Admin.
Admin: “The question is, what are the installs that the average users are using? Not DNS or Web servers on UNIX—I'm talking Joe Beercan. Or the office manager that's screaming for MS office to be installed back on his box, ‘because he knows how to use it.’ That's where the war is, not with us admin types.”
sc0rp: “The war's about money, and when Linux admins are cheaper....”
Admin: “Ha, they throw pens and t-shirts at us to buy our collective silence. The price of one decent admin is $50-60K. The price of the 40 secretaries in the company is $36K apiece. And the admins matter WHY????”
sc0rp: “Because they have the e-mail records. Don't you read BOFH?”
Admin: “Point stands.”

There are a couple of points to take from this:

First, always be careful what you discuss with a Web writer; you may well see it in print later.

Second, the bottom line is that it really is all about marketing. Some software vendors seem to take the attitude that their product stands on its own obvious merits. So much so, apparently, that nobody really needs to be told about the product; the customers will take it on faith based upon past performance, try it out, and like it. Other vendors simply follow the money—and wind up getting more of that money for themselves. Who's right?

Well, after all, who's winning?

The admin's plight
Let's just say—completely hypothetically of course—that there are two main "systems" available for use. They both do more or less the same thing, fill the same need, occupy the same marketing niche, and for the purpose of this discussion, cost about the same. One system's vendor spends more time on the drawing boards and has a clearly better product; the other vendor has one that does the basics well enough but places huge efforts into marketing and advertising.

Now, if anyone ever asked those technophiles who took the time to "taste-test" and compare both systems, the answer was close to unanimous—and NOT reflected in the sales statistics. Why? Because like it or not, the technophiles comprise a very small sector of the total market. Both the vast majority of the rank-and-file consumers and the top decision-makers don't really care so much about the quality of the product in isolation from other factors. They look at TCO and ROI—total cost of ownership and return on investment. Those terms themselves suggest that there's much more to the decision-making process than how well the system actually works. If it can be made to work adequately, understood/used by the average person to enhance the company's profitability, and has a history of vendor support, then, sorry, mediocre is going to be good enough.

This is why an admin can quote the story of the old NetWare server that got inadvertently, literally lost—drywalled over for four years but still working—with no effect on the listeners. That's not as important to the non-geek rest of the market.

This is why an admin can point to the difference in downtime and maintenance costs between Exchange and GroupWise (mostly because of e-mail-propagated virus issues, which almost do not exist on GroupWise) and get ignored. Everybody's used to using Outlook or Outlook Express at home and won't switch. That's especially bad when it's a VP or CEO that won't switch.

However, it's never that simple. It's not just a conflict between the hyper-informed techie and the hyper-propagandized masses or the wined-and-dined executive buyer.

The admin's vision
Heard this one? "Most network problems can be solved by eliminating a few users."

How many admins and techs see users as a complete bother? They break things. They ask stupid questions and take up your time ("I got an error message!" "What did it say?" "Ummm... I don't know..."). They open (gasp!) e-mails with .pif attachments. They always "want" things. Never grateful, they don't care about the late nights they cause you. They don't know how to do anything. They forget their passwords. (Or they write their passwords on sticky notes on the monitor.) They leave databases open and go home. They download the dreaded "Hotbar." And the world ends because, "The network is down!!!"

Yes, "L/Users," what good are they?

Well, for one thing, they pay your salary.

Like it or not, no board meeting ever starts with the CEO proudly proclaiming that "the network only had five minutes of downtime this quarter!" That's not going to be an issue. The issue is going to be company profitability, and that's what users take care of—and the admin/tech's job is to help that happen. That's all. The IT staff is merely in support of the moneymaking activity; the IT department makes no money itself. The object of a well-run network is to assist those "pathetic" users in their production efforts. The object of a company is not to have a wonderful network. The object is to produce whatever the company produces, assisted by a network and group of applications that everybody can use effectively with minimal training.

Of course, it helps a lot if people have the same Web browser, e-mail client, and office suite at work that they have at home—and the same desktop OS (or close) that they have at home. For now and in the foreseeable future, these will be Microsoft products, and the techie who swears by Opera and OpenOffice run on Debian should be prepared to be pretty much alone in his preference—even if he is right. Most users aren't going to care. They don't know much, but what they know, they know—and it works for them.

But you'd think the suits would understand that NetWare is more secure or that Linux is "free," wouldn't you?

"Nothing is impossible—for the guy who doesn't actually have to do it"
If Dilbert didn't say that one, he should have. It's a corporate truism, and it's here to stay.

Couple that with another truism that “suits only listen to other suits.” An MBA trumps an MCSE any day. Another factor is understanding the suits' 100,000-foot view of things. That's where they live, even if it seems superficial to a detail-oriented tech.

Take this study by IDG that purports that Windows is cheaper to administer than Linux. Some of the holes are obvious, such as the fact that Microsoft funded the study. It's difficult to expect objectivity there somehow. An opposite view that claims that Linux costs 40 percent less to run than Windows seems to have attracted less attention, but another study suggesting that Windows is 68 percent cheaper than Linux is simply a case of relying on "headline-ism." People see, "Windows ‘cheaper than Linux!’" and, unless they read it and see that it relates only to embedded systems, merely make a mental note and move on. Oh—Microsoft funded that study too.

If only the world were run by techies.

Editor's Picks

Free Newsletters, In your Inbox