I use both Macs and Windows everyday and I’m happy with both – they serve different purposes for me. But some people can’t stand being shown how small a minority their favorite platform actually holds and are quick to abuse and insult when it’s pointed out to them. The article says nothing about one being better than the other, just the state of play in the world. You can disagree with the figures if you want to and you can show alternative figures if you have them and they can be discussed. But people like #84 above have no clue and they probably didn’t understand the article’s points at all.

There’s a lunatic fringe in every group.

As a teacher who uses Macs at school every day (I am in charge of over 40) and who uses Windows at home every day (XP I see no need to go to Vista yet); I feel that I have a fair picture of both sides of this debate. I would not consider buying a Mac for myself at this stage, I see no advantage in it.

I would prefer it if the OS i used was not an issue at all, which it is becoming with the internet. OS X is ok but is not perfect. Windows is in my experience at home just as reliable (using anti-virus software and being careful). Love your blog.

That said, Vista’s “overtaking” of Mac OS X isn’t exactly a fair comparison since there are far more previous Windows users who are simply upgrading to the new Version.

Compare new Windows users (not previous Windows users upgrading to Vista) to new Mac OS X users and I think you’ll find something very different.

http://www.brightcove.tv/title.jsp?title=741891990

OF COURSE IT IS — because those anal retentive dickheads in Cupertino put limericks in their bloody kernel code to beg people not to run the OS on non-Apple hardware. Because Steve Jobs has such a history of dicking over companies that want to do business with him, because despite Dell and most likely both HP and IBM’s Lenovo wanting to ship OS X they won’t BECAUSE STEVE JOBS DOESN’T LIKE IT AND STEVE JOBS CONTROLS EVERYTHING.

I’m just wondering what sleeping draught Steve Jobs gave to the Apple board and shareholders – in other words when the F they’re going to WAKE UP.

I use bootcamp to maintain my present winxp until all my applications are eventually replaced on the OSX.

For those poor souls that needed to get a new PC after Jan. 30 – most of us HATE vista. I had to buy an iMac just to do basic things that Vista was taking way to long to do. And in the end, many of us UPGRADED back to XP anyway.

I’m surprised you don’t see more people buying Macs after Vista came out, but besides that, this graph really isn’t anything interesting.

Apple’s primary markets exclude most manufacturers biggest markets.

As for 17% Apple does have > 17% of the US retail notebook market and the notebook market has become the largest part of the US retail computer market. They sold more laptops than Gateway! Dell and HP are the only companies to sell more laptops but if you remove the retail business sales numbers Apples share increases dramatically above 17%.

Year over year apple increased shipments by 33% which is 2.5 times the industry growth rate These numbers stand in direct disagreement with the measurements presented in the original article.

Apple doesn’t compete to sell more operating systems than Microsoft. They are competing to put more computers in to homes, primarily in the US, than Gateway, Dell, IBM/Levono, etc. The OS market is secondary for them.

Now, you can use bad statistics with bad sampling to draw the conclusion that Apple is loosing ground but the reality of 33% growth rates and growth rates in their primary sector of 2.5 times the industry can’t be denied. You can dislike Apples all you want but you can’t deny that more and more are being used every day.

Using 3.85 Mac and 96.15 “everything else” (used because the calculations are way more involved for three ways or more) the standard error of the difference of the percentages is different for visits than for users. For visits the standard error of the difference of the visits of 93.3% is 0.3848%.

On the other hand the standard error for the difference of the users would be 1.21%.

The writer is using metrics of how much a type of computer is used and trying to correlate it to how many of that computer there are. Apart from all the sources of bias my point is that we don’t even know what the actual statistic is and what the margin of error of the difference is. You can’t tell how many people own an item based o how much that item is and more importantly you can’t know how good the statistic is unless you know what N is. He’s using the number of “visits” to talk about the number of “users” and he has no way to present the number of total users in order to do an analysis with the numbers.

I’m not saying the results are wrong, I’m saying the results are meaningless because they have no basis in given values. His conclusion might be right but not for the reasons for which he’s concluded.

In conclusion despite working with percentages which gives you the same median we don’t know the margin of error to conclude that a particular value has really increased or decreased.

In addition the results come from the last 25,000 visits to some number of websites. This does not give a random sampling at all because for many large websites 25,000 visits happen in less than an hour. We don’t know what time of day the results were taken from. If the distribution of Macintosh users isn’t even across all timezones (and I’ve read it is not) then there’s a bias for the time of day the result was taken from. If the result was taken during work hours the results are going to be skewed towards Microsoft since most people would be using their computer at work instead of their home computer. If the results are taken from the MSDN website the results are going to be skewed.

In order to draw his conclusion the websites would have to be chosen at random. The sample would have to be long enough to consider weekend, week day and time of day usage and would have to identify unique computers, not unique visitors. It doesn’t so you can’t draw this conclusion.

The only conclusions you can draw are that the numbers are correct to some degree for the visits to specific websites at specific times and that’s a far cry from what the author of this article has stated.

Wrong, because you are still working with percentages. As with all statistics, the more your base of data is, the more accurate your stats will be.

The only skew would be collecting site data with a built-in bias… like seeing what % of visitors to Apple.com or Microsoft.com are using what browser or OS. But probably less so with Apple, since people using Windows at work may be checking there… so their work computer shows up in the stats, and thus actually makes the figures MORE accurate. Statistics are funny like that.

According to the stats, XP still holds 83% of the market… and ALL OTHER OPERATING SYSTEMS are in the remaining 17%… along with Windows Vista, Windows 2000, OSX, Windows 98, Teh Lunix, etc… and the slice of that slice of pie is ranked pretty much in that order.

As of today, only 3.74% of people tracked by W3Counter are using OS X… which means there are more web users using Windows 2000 than OS X (to say nothing of all the Windows 2000 Servers still out there).

I’m willing to bet that if we had the stats on all computers used in the world, Windows NT 4.0 would be beating out OS X as well.