Monday, January 8th, 2007, 3:57 pm
The True Market Share of Linux
My machine at an older office. It
ran Ubuntu Linux (see daily photolog)
OME would argue that the ‘market share’ of Linux has long ago exceeded 4%. Installed based, as opposed to market share, is difficult to gauge. People install Linux as they download or pass CD’s around. They do not buy Linux because Microsoft has a chokehold on OEM’s (AKA ‘Windows tax’, c/f 1, 2, 3). It’s a strategic, anti-competitive tactic.
Web statistics studies are still biased because they usually exclude Linux sites, they throw away “unknown” (often Linux with diverse http-header footprint/string), they ignore Squid, they don’t account for agent forgers (not just for MISE-only sites), and they neglect to account all the traffic that comes from Windows zombies (Windows/IE). In short, they cannot be relied on.
What is the true ‘market share’ of Linux then? There is no way of knowing. Software that is passed from hand tom hand or gets downloaded remains quiet and obscure. But the truth is out there. One just needs to sweep away disinformation which is enforced by companies that have plenty in stake. One of my Web sites, which boasts over 1,000 visits a day, indicates that Linux has a market share of over 40%. It’s not a typical and mainstream site. But will Web survey ever wish to include such a site in a survey? Who will ‘sponsor’ Freedom and encourage outing of the truth? Studies are typically backed and funded by commercial bodies which select their desired hypothesis and fit the conclusions to it, by carefully choosing the methods and data.
Take everything with a large barrel of salt. Just some thoughts to ponder…