Palmy, I have some very grave doubts about the validity of the stats they
collect, as they say:
quote
We collect data from the browsers of site visitors to our exclusive on-demand network of HitsLink Analytics and SharePost clients.
end quote
I know lots of people who have security settings on their systems and browsers that kill the ability of web sites to collect this data, heck the settings are recommended by a lot of law enforcement organisations as part of Internet security. Also, I'm sure there are large numbers of people who do NOT visit any of the sites they use to collect data. All of which lowers the validity of the data collected.
As best as I can tell from the archives for the last few years they show very little growth in Linux desktops, yet we know a number of large government agencies in Europe have switched from Windows to Linux and have seen a significant growth in the use of Ubuntu Linux and its derivation in recent years as well - yet no change in the overall percentage by these people.
If the majority of their data collection is from USA based retail sites then I would expect to see results that reflect the US IT market sales trends. I would also expect them to be like those stats on Linux desktop usage based on retail sales of systems preloaded with Linux, a process that represents a one digit fraction of actual new Linux systems that bears no relationship to the numbers of Linux downloads.
It's highly possible these stats are representative of the major US sites, but I doubt they are really that representative of the world wide stats.
