Darn - they saw me rolling back those laptops to XP last week
{eom}
Windows 10 looks to be doing alright, according to measurements taken by StatCounter. We usually consider desktop OS market share data from StatCounter and Netmarketshare in the first week of each month, but the former analyst also offers weekly data. With the debut of Windows 10 just 11 days behind us, a look at weekly data …
If you can actually find laptop hardware that works under XP these days you're doing pretty well.
On another note, the uptick in XP can probably be classed as margin-of-error stuff. It will be interesting to see if W10 increases from now draw more from 8/8.1 or 7 though.
"If you can actually find laptop hardware that works under XP these days you're doing pretty well."
.
Lenovo is shifting a lot of refurnished laptops at the moment (T420, X220) with Sandy Bridge 3+GHz turbo, for around $200 on NewEgg and Ebay. I bought a few, and they work fine with the XP drivers Lenovo supplies. :)
.
But they are *noisy*. I relegated my T420s (i7-2640m) to beta-testing work for this reason, three years of a howling fan (even if the fan is brand new) were enough. Good machines on all other criteria, though. If you can live with a throttled CPU they are fine laptops.
It will be interesting to see if W10 increases from now draw more from 8/8.1 or 7 though.
Having just upgraded my work laptop from 8.1 to 10 this morning, I'd say that we'll be upgrading the 2 Win 8 machines, but the Win 7 ones will probably stay where they are. I don't think it's a got anything extra that's shiny to make it worth the change in UI. Being a small business, we don't have IT, there's only me. Anything I can't do we pay outsiders for, like hosted Exchange and CRM.
The last sad user with Vista will be stuck with it until the laptop dies. I'm amazed it's lasted this long. Hmmm. I'd forgotten how old that was, I might suggest our first ever upgrade of a laptop not due to it falling apart - they never seem to live that long.
Looking at Statcounter, Windows market share continues to shrink. All Win10 has managed so far is eating into 7 and 8, where it's free for consumers. That's like putting a "new formula" toothpaste on shelf besides the old one, yet still losing customers. Bad news, if news at all.
> Looking at Statcounter, I see Linux flat-lined as always as a desktop OS
"""StatCounter measures market share by counting the prevalence of operating systems hitting web servers"""
That is not quite true. Statcounter works by having Javascript built into web pages. This code sends data from the client to the statcounter servers.
Web servers that don't include the Javascript aren't counted. Clients that block Javascript, or have other security or privacy plug-ins, simply aren't recorded.
So the stats are skewed to those who are unconcerned about privacy and who mainly visit self-selected sites that collect data. Of course Windows users are over represented and Linux is under represented.
I suspect that XP is also under represented because XP users are likely to stay mostly offline due to scares about being unsupported and vulnerable.
Your point is valid, but it doesn't mean that the figures advanced by Statcounter have no meaning.
Over time, even if it is only counting web clients that execute the code, those webclients change and that change is representative of the global environment.
Your typical public survey only needs 2,000 responders to be considered representative of an entire nation. Statcounter gets way more than 2,000 individual data items, so its representativity should be all the more valid.
Not at all. The typical public survey is regarded as valid for precisely the reason that it is thought to be completely at random.* Given the proffered objection is continuous and can never be measured by your the survey, it doesn't matter who large the sample, because of the non-random bias, it can never be regarded as representative of the entire computer population.
*Indeed this is actually one of the current points of contention amongst surveyors as the advent of the cell phone may be inducing biases into traditional techniques.
If this is the case then no wonder opinion polls are such shite. If I chose to ask questions outside the Institute of Directors one week, then outside the Benefits Office the following week, I feel there would appear to be a big apparent swing in opinion.
Exaggerated example, but if one were to poll only traffic with a Safari User Agent being reported, and ignore all other traffic then the world would be full of Apple products.
> so its representativity should be all the more valid.
You failed to notice that these are self-selected web sites and self-selected (by default) clients. Thus the representativity [sic] is invalid.
It may be that it has some useful data for many Windows users moving to different Windows versions as this is unlikely to change their browsing habits or configuration. However the original poster was attempting to show something about a completely different set of users with different operating systems. These are likely to visit different sites than Windows users visit and will use different browsers.
For example Linux users are far less likely to visit Microsoft servers or Windows news sites. They are far more likely to visit sites that do not use statcounter.
"Of course Windows users are over represented and Linux is under represented."
Any evidence of that? I would have though lots more Windows users would be running advert and script blockers due to the much larger number of advert nasties and browser toolbars targeted at the more prevalent platform...
> I would have though lots more Windows users would be running advert and script blockers due to the much larger number of advert nasties and browser toolbars targeted at the more prevalent platform...
It may be true that they _should_ run ad, flash and script blockers because they are more likely to have security issues. But without actual statistics your "though" [sic] is not useful.
This post has been deleted by its author
Yes, there's like 4 billion linux desktops in the world which just never show up in the stats. Remember, this is at least the seventh year of linux on the desktop. It must be running on pretty much all devices by now, and any stats that show otherwise are clearly incorrect.
Looking at Statcounter, I see Linux flat-lined as always as a desktop OS. In 11 days Win 10 has managed to double its market share.
Can I add to this discussion with "who cares?" I think everyone knows that the reasons behind the various company market shares are a long and involved discussion stretching back decades. Far as I'm concerned, I don't really care if people use the things I use (I have Windows 8 installed, after all) as long as they do things the way I want - much like this lovely lovely Kubuntu installation I'm typing from.
A while back I was having a similar argument with some guy who claimed that nothing Linux invented deserved credit because those things hadn't improved its desktop market share any. Putting aside Linux's prominence basically everywhere else, that really makes no difference as to quality. All versions of OSX combined have lower market share than Windows 8. Does that make OSX crap? (OSX is rubbish for a host of completely different reasons!)
Win10 seems to be hard to sell, I got some desktop notifications those days: "Your chance to secure a free copy", "Get it now while it's free", ... It's hardly selling, so they are selling hard.
Maybe their customers got scared by the price (if you aren't the customer, you are the product).
Windows 10 is a no brainer update from 8 or 8.1 and it's a reasonable upgrade from 7. It's got its issues, but its actually pretty good. At the least it eliminates most of the stupid bits of 8 and 8.1 and without the stupid bits, 8 and 8.1 were solid upgrades from 7. I won't bother to talk about Vista and XP as if you're still using either of those you're insane.
You're not the product because you're still going to be buying Windows. The payoff for Microsoft is pushing more home users into current versions of Windows to accomplish the following:
1) Push enterprise users to update their deployed Operating Systems to ones their employees are more familiar with faster than they might otherwise.
2) so that Microsoft can spend less money on backward compatibility.
Essentially expect that after the 2016 releases the Office products won't be supported on out of date Microsoft OS versions.
Windows 10 is a no brainer update from 8 or 8.1
Unless, surprisingly, you have a tablet, where many are complaining they took all the tabletly goodness out of the OS in favour of the desktop.
Also Edge is unfinished (and needs some settings and maybe a full screen mode!!!), Cortana feels like more of a proof of concept than working software and Notifications sometimes works, sometimes doesn't.
Plus, I've already had to roll back two peoples computers, due to 'issues'...
"Unless, surprisingly, you have a tablet, where many are complaining they took all the tabletly goodness out of the OS in favour of the desktop."
If people are complaining because of this then they may be complaining based on false information. I say this because tablet mode is still there and easily enabled in Windows 10, you can set it to load into it at boot as well I believe (I still load into desktop mode).
Hell on the Surface tablet you can set it to autoswitch to tablet mode when you remove the keyboard and put it back into desktop mode when you re-attach it.
Its hard to tell with situations such as this if complaints are coming from actual users or just people who dislike the company, its becoming the equivalent of the "console wars" situation gaming has, where its actually hard to filter out the fact from the fiction.
Windows 10 is a no brainer NON update for anyone who values his privacy and does not consider feeding the Borg with his private life data.
I'll be sticking with 7 until the hardware doesn't accept it any more, then I'll just have to move to Linux I guess. But Microsoft is no longer on my buddy list.
I am NOT trying to start a flame war, just giving my 2p worth...
Mom's laptop has (had) Windows 7, won't be going to Win 10. Yesterday I installed dual boot Linux Mint 17.2 (now the default) and will see how she does. (She can always boot back to Win 7 if she gets in a bind). If after a month or so she's okay with Linux, Win 7 is no more, but it will never be Win10.
From a technical point of view, Win 10 seems to be fine, I actually find it easy and a pleasure to use.
From a privacy, corporate overload point of view: Fuck you Microsoft. I do not want a computer that I can not control, I do not want to be tracked, I do not want to have my every key stroke logged. I do not want a walled garden. (There's a Dr. Suse rhyme in there somewhere). I will do everything in my power to personally avoid using this product on any machine that I come into contact with.
I don't trust disabling any settings until someone reports back with wireshark logs that with the settings off, no data is sent to MS (or any other corporate) servers. Until/When/If that happens I will not use Windows 10 on any machine that has external access to the Internet. I am forced to use it for Work.
Sorry MS, you've gone wayyyyyy tooooo far for me...
Oh yeah, forgot, I added a bunch of Solitare games (aisleriot) that she doesn't have to pay money to remove the adverts from while playing.. What a load of crap.
"I don't trust disabling any settings until someone reports back with wireshark logs that with the settings off..."
Even then, how can you be sure they won't silently be "accidentally" re-activated by some update or service pack at some arbitrary point(s) in the future? Are we all to keep visiting everyone's machines to check the myriad hiding places for all these settings... or set up startup jobs to re-write the registry settings every boot and hope no new "pwn me" settings get slipped in?
It's untenable insanity.
"Even then, how can you be sure they won't silently be "accidentally" re-activated by some update or service pack at some arbitrary point(s) in the future."
You can't, Windows 7 had an update which made it send more telemetry data. A second update silently changed the CEIP setting to on to reflect this, but even turning it back off again you're not sure any more because trust has been lost,
> You're not the product because you're still going to be buying Windows
Paying does not stop you being the product.
For example Sky TV still shows ads. They sell you viewing the channel to advertisers.
Even with Windows, if you pay for it or not, Microsoft will sell your searches or store purchases.
You poor fool! All I can say is that I hope you posted this from your smartphone...
If not, nemesis will hit you with multiple issues from now on. It's a bit like saying out loud that you need the printer to hurry up, as this document is for a meeting in 5 minutes. The speech recognition in the printer OS will detect this, and it'll go into a cleaning cycle, dump all its ink if it knows you're out of cartridges, or suddenly become incompatible with your PC.
Who doesn't, if even only occasionally, play Solitaire?
So how are the happy upgraders going to react when they discover that they have to visit the Windows Store and download and install it? And how happy are they likely to be, when they discover that “free” means “ad-supported?” Finally, how happy are they likely to be, when they discover that, to get rid of the ads, they have to subscribe to the ad-free version, at a cost of $1.95 per month or $9.95 per year?
Then there is the add-on DVD player, free for now, and $14.95 later, and even if the PC has a Blue Ray drive, the DVD player will only play standard DVDs.
At some point, the happy upgraders are going to feel cheated by their “free upgrade.” Welcome to Windows as a Service. (Service is what bulls do to cows.)
Windows 7, a solid OS, with support until January 14, 2020, will be calling them home.