Linux vs. Windows Metrics -- Nothing Is Quite What It Seems To Be

In a reaction piece to the Linux Loop blog post titled “Linux Eee PC Far Faster Than Windows Version” Caitlyn Martin of the Linux Dev Center continues the debate over the criteria used.

The original piece had concluded Linux was faster based on three tasks - boot up, loading Firefox, and shutting down. Obviously this was not comprehensive and very inconclusive especially for users who are heavy users of certain apps such as MS Works or MS Office vis-a-vis OpenOffice. Hence the arguments.

Ms. Martin, while obviously a Linux supporters, also claims a bias against hype and BS. And she did right in clarifying that the numbers actually only compared Xandros on the Asus Eee PC and Windows XP on the same subportable. And more importantly, she reminded everyone not all Linux distros are created equal. Thus, the metrics mentioned are about one specific laptop, one specific distribution and version of Linux, and Windows XP only, nothing more.

As for the comparison themselves, shutdown/start up time is a straightforward and Xandros is clearly faster than XP. All the arguments dismissing this as unimportant doesn’t change the fact that Xandros is faster at this given task. On identical hardware running an identical application (e.g. Firefox), Xandros is also significantly faster than XP. Any comparison of dissimilar apps (IE vs FIrefox or Works vs. OpenOffice) is an apples and oranges comparison and tells us nothing about the speed of the underlying OS.

But what does this all prove? Not much, really. But just that. Xandros runs faster than Windows XP on the Asus Eee PC. :)

You can read the entire article here.