You already know the punchline: Microsoft hired Security Innovation to compare Windows against Linux. Here’s what the author says distinguishes this study from other Microsoft-funded results:

What differentiates the methodology presented here from component-based studies is that it considers the reliability of a system holistically. A system is a collection of smaller parts, each of which can be benchmarked and studied in isolation but assessing the true “reliability” of that system means understanding both how those components interact and their evolution over time. The methodology presented here does this by simulating evolving solution requirements. Specifically, it allows us to compare two systems by adapting those systems to fulfill new business requirements over time and measuring solution failures and downtime.

To accomplish the fait accompli, ZDNet’s Paul Murphy says Security Innovation pitted a nearly 100% Microsoft e-commerce solution against a polyglot of Linux products shipped on SUSE disks. Indeed, it appears that techs loaded most everything they found in the SUSE distribution. Result? Managing Windows was easier, cheaper and more secure. Duh.

Of course if you were building a Linux e-commerce server, you would likely download a pre-fitted set of programs like, on the low end, NetMAX E-Commerce Powerpack 5. On the high end, you might match Window’s total e-commerce solution to IBM’s WebSphere Commerce. Comparing apples to apples would produce a more valid study.

You might also simplify maintenance, debugging and security, by spinning off select pieces of your Linux e-commerce suite to standalone hardware. Who knows? You may spread that work out across seven or eight cheap boxes, so each performs only one task. Try that with Windows and you’ll be buried under a mountain of licensing costs… and probably lose your job to boot.

Background: