Windows 7 and 8.1 market share surge, XP falls behind OS X

Re: Well as usually those statistics must be taken with a _lot_ of salt

They represents the market shares of those people who don't block Javascript from non-trustworthy sources.

We'll lets be honest here, that is the majority.

Re: Now students, please use "free" and "Microsoft" in the same sentence.

While agree with most of what you said for the home market...

Powershell.

Re: Now students, please use "free" and "Microsoft" in the same sentence.

" slowest horse in the cavalry, as the old joke goes "

. . and in that vein, it's the second mouse that gets the cheese.

Re: Now students, please use "free" and "Microsoft" in the same sentence.

Really?

How about:

Not having to run your OS on top of DOS?

64 bit memory addressing.

128 GB limitation of FAT disk image.

No security on a FAT file system

And, whatever you may say about the stability and security of Windows XP,Vista,8/8.1, they're light years ahead of the stability of Windows 95.

Re: Now students, please use "free" and "Microsoft" in the same sentence.

"It's really hard for me to think of a single crucial feature since Windows 95"

Yes, the Win 9x architecture was famously an impenetrable bastion of security and stability.

If you wish to add a 'technical' argument to not getting win 10, then it would help if you didn't back it up with technically illiterate reasoning.

Re: Now students, please use "free" and "Microsoft" in the same sentence.

"If I had the choice, I'd probably be running a completely debugged version of Windows 95"

Don't like Cortana? Mayber you'd also be happier driving a Cortina?

Re: Error Margins

The statistical flaws in this monthly meta-analysis (more a "poll of polls" and ask Peter Kellner about how reliable those are) are almost without limit. Variation with the margin of error is probably the least.

The main ones:

• no description of the sample set – see Christian's post
• failure to corroborate with El Reg's own data
• failure to corroborate with source like Akamai's non-JS data. This would involve real work, not just copy & paste.
• failure to account for a general shift to mobile affecting not just the sample size, but what's left in it
• failure to account for the number of working days in any particular month (work days favour Windows and IE in general)

Re: Error Margins

Margin of error

Estimation of uncertainty

Precision of measurement

Any statistical report (claiming accuracy) that falls to give some acknowledgement of these types of issues in its methodology is flawed.

Also remember by definition it's looking at machines connected to the internet. Anyone connecting an unsupported and unpatchable XP box to the net wants their head examined.

Same tired old meme. Don't you realise Linux users stopped talking about the year of Linux on the desktop years ago. We want to keep the riff-raff out.

Re: A bit of a false comparison

"You can't really compare OSX as a single entity against multiple versions of Windows - OSX comes in new versions too at reasonably predictable intervals. I know the upgrading of it is a lot nicer, and most Apple users shell out for new versions fairly quickly as (a) they can generally afford to and (b) the upgrades tend to be trouble free and add something worthwhile, even if they are smaller in scope than new major versions of Windows."

You make a good point about the versions, but OSX upgrades have been free for some time now.

And I'm yet to see an upgrade yet that doesn't break some third party app or utility.

Re: A bit of a false comparison

"You can't really compare OSX as a single entity against multiple versions of Windows"

Well no, you can't. It's also not really right to lump every different Linux distro into one category either.

But the point is more that the number of users on each individual version of desktop Linux or OS X is meaninglessly small in a market that remains so totally and utterly dominated by Microsoft. Individual versions of unpopular Microsoft OSes have greater market share than the combined total of all versions of OS X. The most popular MS OS has nearly 60% market share alone. You could quite easily split the graph into 'Windows 7', 'Windows 8', 'Windows XP' and 'Other', and find that Other still struggles to come in third; if you lump all versions of windows together then MS are more dominant in the space even than android is in smartphones - and with less fragmentation.

Now I'm sure that pointing out this objective truth will get me a huge number of downvotes from Apple fans and Penguins, but really everything other than Microsoft remains effectively irrelevant in the desktop space. That's not a statement on relative quality - I like Linux myself, and even though I despise OS X I do see the benefit in that locked-down fisher-price interface for small children or attempts to communicate with chimpanzees and dolphins (so anyone in high level management). But the simple truth is, what we're seeing is the emergence of a world that wants Windows on its PC, Linux on its web server, and Android on its phone.

No-one really cares about OS X, and even Apple's enormous marketing clout from the successes of the past ten years hasn't really impacted on that. The trumpeting of all versions of OS X combined reaching parity with a 14-year old OS that's been out of support for over a year is frankly embarrassing. Meanwhile iOS is rapidly declining to the same small market share as its desktop cousin; Apple's approach to mobile is repeating pretty much the same strategic mistake it made in the desktop space back in the 1980s, surrendering market share for short-term profits; and you can see just how well that turned out from these very graphs.

Re: Don't want 10

I'm planning on something similar.

I have 3 Windows devices.

* A current (ish) home built desktop, which is my main rig. (Gaming, Office etc).

* A functional, but retired Laptop. (Replaced by tablets and a Chromebook over the last couple of years or so).

* And an old Acer Aspire R3610, that used to be my Media PC under the TV (replaced by an Intel NUC last year).

All three have Win 7 64 bit, full on the Desktop, OEM on the other two.

The Desktop is the only one to pop up the "Win 10" notification so far, but I will not be updating that machine any time soon.

I will update the Laptop though, and maybe the Aspire, so that I'll at least get some familiarity with the new OS, and to see what it might screw up on my local setup (Nas drive access etc) (Things like the Live login etc worry me).

Edit: Typo.

Re: Don't want 10

>I have resurrected a Win 8.1 install

Remember if that was originally a Win8 install, you will need to keep you Win8 install media. Because the Win8 key won't work with a Win8.1 install disk, only with Win8 and upgrade via WuP to 8.1.

Hence I expect that the 'Free' Win10 update will be similar, it will only install over a previously activated OS and won't accept a licence key for 7/8/8.1. Hence if you every need to reinstall... it may be simpler to buy a retail copy when MS discount it to a few pounds just as they did with Win8, which could be had for £14.99...

Re: The way Nadella's Microsoft is going..

If Microsoft were smart (a dubious proposition, given past history), device drivers that work on Win 7 or Win 8 will also work with Win 10. But that would mean a device driver model that remains unchanged (Gee, like Linux!). But it seems that there is always some device driver tweak that renders your trusty {printer, scanner, all-in-one, web cam, other} device inoperative with the new release. Why? So the hardware manufacturers can sell you a replacement for something that works perfectly. Fix it when it's not broken, they say.