In-Depth

Software Crusades

When platform zealots attack, it can threaten everything from productivity to profits.

Battles are nothing new to the computing world. PCs vs. mainframes. OS/2 against Windows and NetWare squaring up against Windows NT. Zealots line up behind their favorites and take no quarter.

Sometimes a clear winner is declared. Other times, debate rages on, and on and on.

This routine is a natural, if not necessary, driver of technological change. For many end users, corporate executives and even IT professionals, the debate is personal. Call it software religion, platform bias, software zealotry—those who believe deeply can become so biased that they close their mind to all other options.

Today's top battles include Windows clients vs. Linux and the Mac, Windows servers versus Linux and Unix, Internet Explorer versus Firefox. Devotees chew up countless hours arguing for their favorites and badmouthing the ones they despise. There are even camps for programming languages, databases, and handhelds.

To the casual observer, software zealots are a mere nuisance—the coworker you avoid debating, the flamer in an online forum. But according to Redmond readers and industry experts, software religion has real world costs—often the wrong technology is chosen based purely on opinion, rather than logic, ROI and need.

One of the most frustrating aspects of encountering a software zealot is the deterioration of reasoned debate. But there are some counterzealotry strategies to try.

Departments Divided
Mike (last name withheld by request), a network engineer in Lincoln, Neb., knows firsthand the difficulties of working with a software zealot. In his case, the software zealot was his boss who had a self-professed hatred of Microsoft. "He made his distaste known for anything Microsoft-related … pretty much from the beginning," he says. "Microsoft did everything wrong. You could not argue with the guy."

Things came to a head when the manager decided to ditch Exchange, just because it was from Microsoft (see "Software Religion: Tales from the Trenches"). This was one of many conflicts and the manager's bias helped convince Mike to resign.

Because software was chosen for ideological, rather than logical reasons, end users ended up saddled with substandard systems, which makes IT look bad. "It all falls back to the end result. If your users aren't happy, that reflects badly on you," he explains.

"[Software religion] is a huge problem. It's good to be passionate about things, but when it reaches a point when you're blind to other alternatives, then there isn't a benefit."

Tom (not his real name), an analyst in Eugene, Ore., is currently in the situation Mike got himself out of: Working for an anti-Microsoft zealot.

The situation, while difficult, has an upside—he never would have implemented the open source e-mail server they're running if not for the manager's platform bias. "It's a great solution … we're very happy with it," he says.

Still, Tom struggles with the manager's zealotry on a regular basis, which has hurt the small company he works for because they spend too much time arguing over solutions instead of implementing them: "It's very counterproductive."

Although Mike and Tom favor Microsoft technology, they insist they are not zealots and try to keep an open mind. "I'll admit that I have my comfort zone," Mike explains. "I certainly like A … but that's not to say that I'm opposed to option B. If you go through life and career just believing one thing and being so closed minded, you're never going to grow [and] you're going to anger a lot of people," he explains.

When Software Religion Turns Critical

Whatever the reasons for software religion, end users are almost always the ones who suffer, argues Dr. Scot M. Silverstein, a doctor and professor at the College of Information Science and Technology at Drexel University in Philadelphia.

And in mission-critical environments like health care, choosing the wrong software can be dangerous. “You can kill people," Silverstein says.

Any mistake in healthcare IT projects—not just those relating to software religion—can lead to disaster. But when platform bias enters, it's more tragic because it is so
preventable, he says.

Silverstein cites case after case. There was the IT department that refused to change the hardware in an intensive care unit, even though the keyboards couldn't be properly disinfected, the fans in the computers—which were rigged to hang directly above patients' heads—could spread deadly infections, and the chips in the computers and the ICU equipment interfered with each other. Even though alternative hardware was available, the IT department wouldn't budge because it was “non-standard."
Sidebar
In another case, a transportation department was forced to rehire a number of drivers fired for drug and alcohol use after a labor negotiation. The department responsible for the drug testing schedule of the rehired drivers wanted to set up the testing schedule on their NT clients. The IT department insisted the scheduling application run off its mainframes because they didn't like NT, Silverstein says. The IT department failed to add a driver to the schedule. The driver later crashed while driving drunk, killing several passengers.

Bias often comes from wanting to keep IT jobs simple. “For IT people the maintenance and costs [can be less] by sticking to one platform, but the result [is] a shift of a work burden over to actual line personnel," he says. And in mission critical areas, such as cardiac units and ICUs, that can put lives at risk. “IT people need to realize that patients don't get sick and doctors don't toil in hospitals so that IT people can have easy-to-maintain computing."

— B.N.

Enter the Manager
If it's hard to work in a divided shop, it can be twice as hard to manage one. Dewey Landrum, a security manager for a large consulting company in Austin, Texas, runs into software religion all the time. "We have the hardcore Unix admins who swear their flavor of Unix is best, and we have Linux folks who are the same way," he explains.

That platform bias used to wreak havoc on his company's consulting projects. "It really did influence the way that things were done [here] because … the experts were religious about what they wanted to do, and since they were the architects, they got to make those decisions."

The situation improved dramatically after the company started an architecture review board. Now, for every project that comes up, the committee develops several platform options, outlining the pros and cons, then lets the (consulting) client make the final call. "A lot of these clients don't have the technical skill to run Unix, or maybe they're a Unix shop and they don't have the technical skills to run Windows," he explains. "You have to look at what's best for the client."

Derek (not his real name), an IT manager from London, also struggles with zealots on his staff. For instance, when his department was looking at SharePoint Portal Server for its intranet, one of the Unix zealots on his team fought against it tooth and nail. "It can be very difficult trying to achieve a balanced discussion if there is one person who is so focused on a particular approach that he cannot even consider the possibility that there are valid alternatives," he explains.

Some staffers think Derek is a Microsoft zealot. "I used to be a Unix admin, so some might see me as going to the 'dark side,' but I don't think so," the manager explains, adding that he's tried to combat the impression by researching alternatives thoroughly and spending more time explaining his decisions. "It concerns me that our working relationship is not as good as it could or should be."

Bob Perrone, a network administrator in Nazareth, Pa., has also been accused of being a Microsoft zealot because his infrastructure relies heavily on Microsoft. He doesn't exactly refute the accusation. "I am a Microsoft zealot from the point of, it works," he says.

Despite his Microsoft leanings, Perrone is open to alternatives—in fact, he's seriously considering implementing the next release of OpenOffice. "Everything I've read about it makes me want to give it a whirl."

Landrum, the security manager from Austin, says his background in everything from Linux to Unix to mainframes to Windows helps keep him above the software religion fray. "After a while you see the strengths and weaknesses of each one, and you try to utilize the right component for whatever you're designing at the time," he explains.

Even so, he says that IT managers can be particularly vulnerable to software religion, as they get more and more detached from day-to-day technology work. "[You get] stuck at that last version or that last thing you were really good at," he explains. "You reference everything back to that, so maybe you're referencing it so much [you're] no longer listening to anybody else."

He says IT managers not only need to be able to keep an open mind about alternatives, but rely on their IT staff in making platform and software decisions. "You have to turn it over to somebody else. It's a matter of hiring the right people and trusting them."

Regular IT staffers can fall prey too. According to Andrew Forman, an application developer in Iowa City, being hard-headed can be the kiss of death. "It can be career suicide to be stuck in one rut," he says. "If you focus too much on any one area and don't evolve with the industry, you'll get left in the dust."

Forman sees an upside to software religion. "When someone is really that zealous with whatever they're working with, they can dig in and become an expert because they're so enthralled. That can be a benefit if you can figure out how to extract the knowledge from them without getting a diatribe," he says with a laugh.

But zealotry is more often a barrier, and even if it doesn't cost you your job, it can marginalize you, according to the readers we talked with. "I've seen people lose credibility because everything out of their mouth is a particular flavor of a product," Landrum says. "It gets to the point so where people shy away from that person because they already know what the guy is going to say."

Software Religion: Tales from the Trenches

"Upper management at my previous employer decided that it was time to migrate from our NT4/Exchange 5.5 environment.

Our IT manager repeatedly made known his distaste for anything bearing the Microsoft logo, and was suddenly a Linux guru. So he suggested the company roll that direction. He put together pricing options—always, of course, passing on only the most expensive quotes for upgrading to Windows (Server) and Exchange 2000. Through all this price manipulation and Microsoft bashing, it kept being brought up that Linux had the amazing low price of 'free.' I kept opposing the suggestion on the grounds that there was a fully qualified staff in place to fix pretty much any problems that popped up on the Microsoft platform, and we'd lose that ability with a switch to Linux. Not ignoring the initial benefits, I begged to have the potential long-term consequences considered.

I was ignored.

Not only did we have to outsource the initial setup, but stuff never did run right, and we had more server troubles than we did when on the Microsoft platform. Further, we went through three different mail servers, the first two being 'free' and the third option costing more than Exchange ever would have, and our users were constantly unhappy.

Three years later, I have moved on. I still talk to people out there, and had to chuckle when I learned they recently switched again and are now happily running Windows Server 2003 and Exchange 2003."
— Mike, network engineer,
Lincoln, Nebraska
sidebar
"I once worked as a consultant at a (nameless) European site where the IT manager was a graduate of the local university, where Unix ruled. He was an utter Unix bigot. As a result, the IT staff had to fight every requirement through, and only when it absolutely couldn't be done with Unix would he grudgingly allow Windows. This in an active trading environment where busy and short-tempered dealers wanted Excel and Office, Windows and all the other stuff they knew and loved. The company had an NT4 setup with NT4 clients and the IT guys needed to move to Active Directory.

Long discussions followed, with me doing my best to sell the concept. Eventually we started talking about the LDAP aspects of the AD and he lit up. Leaning to me he asked, 'Well, if it's a standard LDAP database, we can get the schema and implement it on some Sun Unix boxes!'

Of course, Microsoft was no more likely to provide him with an open copy of the design of its Active Directory database than it would be to publish the source code of Office on the Internet. And he simply didn't grasp the huge range of tools that Microsoft provides to manage AD, equivalents for which would have to be found in the open source community or be written for his implementation.

This was one of several times in my consulting career where I was glad to do what I could then leave with a smile. I still don't know if they ever got AD."
— Richard Sargeant, consultant,
Wollerau, Switzerland

The 'Airplane Magazine' Syndrome
According to Landrum, even the best IT managers won't be able to keep their department open to all alternatives without top management's support. But what if top management is the one pushing the zealotry, like when the CEO reads an article about some hot new technology and suddenly wants to covert the whole company?

Forman says that top-down platform bias at the large telecom he used to work for cost the company millions.

One day upper management "discovered" Java. "While they weren't quite sure what Java was, they were sure it was cool, and they wanted to fund that, so any project that had the word Java in the name got funding," he explains.

Of course, it wasn't long before anyone who had a project they wanted funded would throw "Java" into the project name, even if it had nothing to do with the language.

"In some ways, it did help get people to think of different languages to program in, which isn't a bad thing," he says, "but the amazing ignorance of the executives … was a real problem."

Another time Forman wanted to get a Linux file server up quickly for a project, but management kept rejecting the request because Linux wasn't an approved platform. "They would say, 'Well, the vendor says you can do this on Solaris,'" he comments, "and I would respond, 'Yes, but that will cost us $50,000 and take an extra month.'"

Then one day, Forman got his server approved. "What really changed is that the execs started seeing Linux more in their trade journals, and suddenly it was like 'Weren't we doing something with that?'"

While many readers report the "airplane magazine" phenomenon, Joe Erickson, president of the software company Spreadsheet Gear LLC, says that the opposite is just as bad—companies that can't move off their current software due to bias.

De-Fanging the Zealots


1. Communicate
2. Turn the tables
3. Focus on the bottom line
4. Balance the power
5. Pick your battles
6. Don't aim for a mind meld
7. Keep your mind open

"Once an individual or a group chooses a platform, whatever it is, you have a vested interest as far as reputation and maybe financial interest in proving that that was a correct decision," he says. "Most people do not enjoy change. This leads some to make passionate arguments against a change of platforms, which winds up looking like zealotry for their current platform."

This can happen when companies don't look ahead. "In corporate America, it can be so much of a focus to get from one day to the next that you just stick with what you've been doing, and you don't have the time to look at what advancements are out there," he comments.

When it comes to zealots on the corporate level, Forman recommends keeping technology discussions focused on business issues. "A lot of VPs are far enough from the technology that they don't understand the minutia," he explains. "Saying, 'Hey, this will be easier to maintain in the long run, this will be something that won't need so many hours of work'—that can really help."

Erickson also recommends that IT managers consider approaching corporate directly about any top-down bias. "Honesty and transparency are always the best policy," he says, "even if they might cause some temporary discomfort."

Countering the Zealots
Some zealots will never be appeased, but don't give up until you've evaluated or tried all these parrying strategies.

1. Communicate. Communication is key to battling software religion at all levels.

When facing end users, Mike from Nebraska encourages involving them in IT decisions as much as possible, not just to combat their biases, but to make sure IT decisions aren't seen as dogmatic: "They may not get the answer they want, but if there's open communication channels, at least their views are hopefully getting considered."

2. Turn the tables. This communication strategy is for managers facing zealotry from below on upcoming technology decisions. If a zealot sounds too much like a salesperson for the vendor whose software he's promoting, appeal to the zealot's expertise in that platform to provide a deep understanding of the hidden costs, integration headaches and other potential blindsides specific to your organization of taking his chosen approach. Your zealot may not be happy with the choice you make eventually, but your decision will be the stronger for it and he will be more deeply invested in the end result.

3. Focus on the bottom line. Keep discussions focused on the bottom line. "[Then] it's not really what product you're going to," explains Tom from Oregon, "but the cost to get your solution finished."

4. Balance the power. Creating a formal process, such as Landrum's architecture committee, to approve major technology decisions takes power way from individual zealots, helps everyone feel involved and can stem bias. But eventually, you'll still need to deal one-on-one with the zealot, and getting through takes more than just technology know-how.

5. Pick your battles. No matter how bad it gets, don't resort to arguing—especially when there's no way you can win. Sometimes, you have to let go, says Perrone: "As long as their religion is not interfering with the operation of the enterprise as a whole, deal with it professionally—no shouting required—and leave it up to your adversary to end up looking foolish."

6. Don't aim for a mind meld. Don't be surprised if your efforts come up short of satisfying the zealot. Many readers have looked for answers for years, with no success.

"For a zealot of any stripe, be it software or cars or religion, it's not about the product, it's about the person," comments Ernie Lowell, an IT director in Stockbridge, Mass. "You can't tell them the solution is better because they won't hear you … You've got to find why they are this way and go with that."

"No matter what you do, these people will come back and say, 'You know, you could have done it cheaper if you used my product,' or 'It would have been more secure if you used my product,'" comments Landrum. "You can show them a hundred examples of why other systems are valid, just as valid as the one that they love—you'll never win the argument."

7. Keep your mind open. Zealots can keep you from becoming a zealot yourself. "[They] have a vested interest in pointing out the deficiencies of competing software," says Forman.

As hard as it is to listen to someone who won't listen back, the effort can be worthwhile. "By closely examining the competing software, you avoid being trapped by software religion [yourself]," comments Perrone. "When you are sure of your convictions and can logically show the positive differences without turning into just another zealot, you can at least feel confident in your decision."

And, he says, you never know: "You may find that their suggestion is better than yours."

Featured

comments powered by Disqus

Subscribe on YouTube