News Stay informed about the latest enterprise technology news and product updates.

Microsoft admits Linux research is met with skepticism

A Redmond executive says IT decision makers don't have to believe the numbers in Microsoft-backed studies to get something out of them.

While IT administrators and managers generally turn up their noses at vendor-sponsored research, they may actually get something out of the latest Microsoft-backed study -- and it has nothing to do with whether Windows is better than Linux.

This week, Microsoft released the results of yet another study. This time it was from VeriTest, a software-testing division of Lionbridge Technologies Inc., in Waltham, Mass., which was paid by Microsoft to compare the reliability of Windows Server 2003 against Red Hat Enterprise Linux AS 3.0.

Many IT managers say they refuse to take seriously any sponsored research, and insist that their choice of an operating system platform has far more to do with their individual situation and the applications they are using, which often makes a pure comparison between two platforms unhelpful.

"What we really need is the infrastructure that can support

More on Linux v. Windows
Linux lags Windows in new security report
both platforms," said John McGlinchey, a senior consultant at a major pharmaceutical company. "When you hear that Microsoft paid for the study, it will always make you wonder. It's just like all the tobacco companies paying for studies that said smoking wasn't bad for your health."

Ernie Coldwell, a network analyst at MAHLE Industries Inc., a Morristown, Tenn., supplier to the automotive industry, suspects that senior executives who are not IT experts may take some of the materials at face value. "I would put more weight on this if it came from someplace independent, like a university," Coldwell said.

A starting point, not a clincher

But Microsoft's Ryan Gavin, a director of platform strategy, said what he hopes the studies do is give IT executives some data to help frame a discussion, even if they disregard the specific numbers contained within the materials.

"Customers are looking for ways to think through and evaluate criteria," Gavin said. "You don't have to believe what the numbers are."

For example, the VeriTest study released this week compared reliability and manageability of a production environment built around Windows Server 2003 compared with a similar environment built around Red Hat Enterprise Linux 3.0. There were test environments with an infrastructure server, an e-mail server and a file/print server.

True benchmarks are not available

Gavin said customers can use the study to get a sense of how to view what is a reliable system. There are no standard benchmarks on what defines reliability. "It's not kernel reliability, as there is little difference between the two platforms," he said.

Gavin said Microsoft's "Get the Facts" campaign, which has essentially spawned a slew of Microsoft-sponsored research that pits Windows against Linux on a variety of levels, has succeeded in at least one area. The first studies were about total cost of ownership, and many customers believed that Linux was free, which it isn't, he said. "You don't hear a lot of dialog around free Linux anymore," he said.

Note: This article originally appeared on

Dig Deeper on Alternative operating system security

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.