Slashdot Mirror
Windows 7 Share Grows At XP's Expense
CWmike writes "Microsoft's Windows ran to stay in place last month as Window 7's market share gains made up for the largest-ever declines in Windows XP and Vista, data released today by Web metrics firm Net Applications showed. By these numbers, Windows 7's gains were primarily at the expense of Windows XP. For each copy of Vista replaced by Windows 7 during November, more than six copies of XP were swapped out. Meanwhile, Apple's Mac OS X lost share during November... betcha Ballmer is having an extra giddy time with that news. Linux came up a winner last month, returning to the 1% share mark for the first time since July. Linux's all-time high in Net Applications' rankings was May 2009, when it nearly reached 1.2%."
There are no error bars. This is a straight dump from their collected information on web traffic.
Anyone who mistakes this information for a statistical evaluation of actual market share by physical units or even actual market share by "web presence" is misusing the data.
They may well try to make a fairly representative sampling based on diversifying websites they collect data from, but day-to-day, let alone month-to-month, variation makes this data at best a rough approximation of the actual market. But they're not claiming that this data is a reflective snapshot of any actual market--they leave that to lazy journalists. Instead, what their statistics track are trends over time using a consistent methodology. It's a clue about the state of the actual market, but nothing more. Only lazy journalists would take a single month's reported numbers and make a claim about actual market share.
Their numbers are accurate to several decimal places--they have an exact count of the "survey respondents"--the over 100 million reporting machines each month. Where there is insufficient data is making a projection from that sample to the actual market (but again, the data can't realistically be used for that). Linux's NetApp share has bounced up and down a distance of 0.1% since the middle of the year. This probably has nothing to do with Linux's actual market share changing and more to do with variations in browsing habits and which sites are recorded.
"Vista runs on anything 7 runs on"
Not so. I have installed 7 on the very same hardware that Vista barfed on. I've never seen the BOSD on 7 that Vista threw up frequently. More, 7 runs beautifully in a VM, while Vista is something of a dog. As I said, 7 runs decently on a 1 ghz machine with one gig of RAM. Vista will not.
"Windows is like the faint smell of piss in a subway: it's there, and there's nothing you can do about it." - Charlie Br
Active directory is not only used to authenticate users, where it's value is derived from is the ability to organized your entire computer network into organizational units and apply custom policies to each of those OUs. Think of this as having a *nix repository for every OU and in this repo there is a custom script to modify /etc to apply the correct policy to all your machines. Now add the ability to do all this with a couple mouse clicks. This is active directory.