Open source studies

Analysts are paid to be objective. Admittedly, they have an abstract, from-orbit type of view on the industry, and most don’t even have technical backgrounds, but they do their best work when making fair judgement calls, above all else.

At ITAC Ontario’s seminar entitled “Open Source & Total Cost of Ownership”, Laura Koetzle of Forrester Research presented a well-balanced if comical view of her March 2004 study called “Is Linux more secure than Windows?” Self-deprecating at times, she listed several flaws the study had. For example, when pointing out that Windows received an eyebrow-raising security score of “100.00%”, she tacitly admitted that it is MS policy to say nothing about a security exploit until a patch – or a virus – appears.

“Feel free to throw muffins at me,” she jokingly said to the mostly IT administrator audience.

While there is great controversy in the methodology (read RedHat, SUSE, Debian and Mandrake’s counterpoints), her data rings true. She pointedly told IT staffers everywhere, regardless of what OS they have on their plate, to “eat your security vegetables”.

I maintained an open mind when Laura Didio, Yankee Group analyst, took the stand to present her TCO study discussed here on Quantified earlier. Sadly, she chose her time to lavish us with snippets of gossip on various executive clients, and how they supposedly confided her with their tales of Linux migration woe.

She even brought up Ballmer’s infamous Singapore strong-arming where he claimed a study concluded Linux violated hundreds of patents, an assertion that has been soundly debunked by the study’s author, Dan Ravicher.

I called her on the fact she curiously neglected to tell the audience that the surveys were distributed by Sunbelt Software, a Microsoft Certified Gold Partner, to its Windows customers. I was puzzled by this methodology, since it seemed like it was like walking into the Grand Ol’ Opry and asking who there liked to listen to the Bangles.

I tried to press on with questions, but the host got up and amicably suggested we continue the discussion in coffee break. We both smiled and nodded.

As Ms. Didio sat down, an IBM executive that was sitting behind me whispered, “Thank you for that!” and added, “She was incredibly biased.”

Two minutes later, I saw Ms. Didio walking out the door. Pity.

Analysts are paid to be objective. They do their best work when making fair judgement calls, above all else. When they stop becoming objective, they become useless.