Reminder: Windows 7 really is the new XP

Back during the darkest days of Microsoft's GWX campaign, when they'd abandoned all pretense of believing in the quality of the product and offering Windows users a free upgrade, and instead started switching users' systems to Windows 10 no matter how many times they'd refused previously, it was already becoming clear that Microsoft had done lasting harm to their own brand, and to the relationship of trust and goodwill that they'd previously enjoyed with users of Windows 7.

I wasn't alone in referring to Microsoft's GWX fiasco as "upgrade-gate," or to point out the consequences with which Microsoft would have to deal for the next several years; pieces like this one, from Makeof.com, were pretty easily found at the time:

Steve Jobs famously said “people don’t know what they want until you show it to them.” Microsoft must think this is true for Windows 10. And so its developers keep finding new ways to trick Windows 7 and 8 users into upgrading because surely they will like Windows 10 once the see it. Or they’ll just surrender.

Personally, I do like Windows 10, but I also appreciate the reasons of those who oppose the upgrade. And I think what Microsoft has been doing is deeply disturbing and unethical. Microsoft acts as if its goal for 1 billion Windows 10 users supersedes the company’s responsibility for its existing Windows customers.

This reckless battle has unintended consequences, which not only hurt Microsoft’s customers, but also its business.

From loss of trust in the Windows; to users simply turning off Windows Update to avoid the hated GWX payloads; to actual monetary costs in the form of lost time, bandwidth, and productivity; reasons abounded why Microsoft's overly-aggressive GWX push was a bad idea. And while the worst of these for Microsoft, "Home Users Will Abandon Windows," hasn't yet come to pass, there's still no sign that consumers have forgiven Microsoft for the liberties, excesses, and borderline (or actual) abuses of GWX.

Microsoft's GWX push was of a piece with Terry Myerson's Windows-centric strategy, which Microsoft has since abandoned. Two years after GWX's failure, Myerson is no longer at Microsoft; his Windows and Devices Group no longer exists, its various teams having been redistributed across other business units which, according to Microsoft, are actually the future of the company. And Windows 10 is still not as popular as Windows 7... depending on who you ask, of course.

The fallout from GWX still hasn't stopped falling, either. Every month, Microsoft delivers updates for Windows 7, and every month, the description of those updates includes the same disclaimer: "does not include windows 10 upgrade functionality." That's still necessary, more than two years after GWX; that is truly epic levels of fail.

But it actually gets worse for Microsoft.

Do you remember when WX was supposedly on pace to surpass W7 by November?

NMS's end-of-January numbers are out, and once again, WX has managed modest gains at the expense of Windows XP, while W7 and W8.1 remain mostly unchanged. And, no, WX still hasn't caught up to its nine-year-old rival.

WX gained, of course, from 32.93% to 34.29% (+1.36); W7 ticked down, from 43.08% to 42.39% (-0.69); W8.1 ticked down slightly, from 5.71% to 5.56% (-0.15); and XP slid the most, from 5.18% to 4.05% (-1.13). Except for W8.1's, all of these results are above the ±0.5% "noise threshold," but WX's gains are not enough to encompass the losses of W7, W8.1, and XP. Some of those former Windows users are going elsewhere.

Where are they going, you ask? By the looks of it, Apple. Windows' overall market share slid from 88.51% to 87.79% (-0.72), while MacOS grew its overall market share from 9.02% to 9.95% (+0.93), propelled by MacOS X 10.13 (from 3.53% to 4.46%, +0.93). An overall decline in Windows' user base probably isn't something that Microsoft want to see; yes, WX gained more than a percentage point to start the year, but the fact that those gains are mostly coming at the expense of the 16½ year old XP, rather than the market-leading W7, can't be good news, either.

Is Windows 7 the new Windows XP?

Or is the situation actually much, much worse than that for Microsoft?

Well, according to this piece at Computerworld, it's looking like the latter:

In April 2014, when Microsoft rescinded Windows XP support, that version accounted for about 29% of all copies of Windows worldwide. Currently, the best-guess for Windows 7 at its end-of-support is a full 10 percentage points higher.

But that's not even the half of it.

In 2012, two years before XP shuffled off the support scene, Computerworld, using the same 12-month trend that produced a 39% user share for Windows 7 in January 2020, figured that XP would have a mere 17% in user share at retirement. The fact that Computerworld's back-of-the-envelope forecast for XP was off by 12 percentage points does not make for much confidence in the 39% mark for Windows 7 in 24 months.

In other words, today's calculation may be significantly lower than the reality of 2020.

Microsoft has at times resorted to some of the same name calling it used in 2012 for Windows XP when it has talked about the venerable Windows 7, which remains the stock enterprise OS. A year ago, for example, the head of Microsoft Germany said Windows 7 was "long outdated" and argued that the operating system did not meet "the high security requirements of IT departments."

Expect Microsoft to ramp up that line of attack as Jan. 14 [2020] approaches, when Microsoft is almost certain to remind customers that the Windows 7 drop-dead date is two years away and closing fast.

Worse yet, comparisons with Windows XP seem to be ignoring another important trend in personal computing: the  end of Moore's Law as personal computing's driving principle. When Windows 7 was released in 2009, Moore's Law was still driving the tech industry, with shrinking dies resulting in the shrinking size and increasing power of, e.g., smartphones, even as the PC power growth curve was already levelling off. Now, though, even smartphones' computing power has plateaued, and PC and laptop sales have been decreasing year-over-year for six years and counting.

All of which means that new PC sales will not drive Windows 10 adoption with individual consumers, in the way that it did for Windows 7. Even if large corporations and institutions adopt Windows 10 in a yooge way this year (something which isn't at all certain, IMHO), Windows 7 is clearly set up to hold its user base much, much longer than Windows XP did, with a dug-in user base that distrusts Microsoft's every utterance just on principle. Trash talk from Redmond is unlikely to yield any more positive results in two years' time than it's yielding now.

It's unclear what, if anything, Microsoft can do to earn back the trust and goodwill of those Windows 7 die-hards. GWX damaged that relationship pretty thoroughly, and their ongoing efforts to push Bing, Cortana, Edge, and other Microsoft products on consumers that simply aren't interested are probably not helping. Two years is a long time, but the clock is ticking; if Microsoft really want to win over Win7 users, then they need to start work now on some positive messaging that will appeal to these consumers, rather than continuing to rely on the same tired negative messaging which has clearly not been working, up to now.

Given Microsoft's track record (going all the way back to 2012, remember), I'd say that it's unlikely, but the Redmond team have managed to surprise me once or twice over the last couple of years, so I guess we'll see. Either way, though, I think that Microsoft's master plan to take over the computing world with Windows 10 is in something of a shambles, and people are clearly starting to notice.

NMS make major changes, and finally explain themselves.

I'm quite late posting today's OS market share numbers, in part because I honestly wasn't expecting much to change from last month. Well, was I ever wrong on that, because NetMarketShare have changed just about everything... and, more importantly, they've finally explained why:

Welcome to the new NetMarketShare.com

---

We completely rearchitected the product to provide more accurate and accessible data. Here's what's new:

Elimination of botnet data

---

As bot traffic across the web has risen dramatically, it has been a challenge to detect and remove it from our dataset. This is a critical issue since bots can cause significant skewing of data. In particular, we have seen situations where traffic from certain large countries is almost completely bot traffic. In other countries, ad fraudsters generate traffic that spoofs certain technologies in order to generate high-value clicks. Or, they heavily favor a particular browser or platform.

The primary focus of this release was to build detection methods to eliminate this traffic. We rewrote the entire collection and aggregation infrastucture to address this issue.

Please note: This dataset is separate from and replaces the legacy data. Subscribers that need legacy data please contact us for access.

The "bot traffic" explanation is... plausible, actually. Bots are a huge online problem, and for a business whose business is measuring online traffic, it makes sense that large-volume bot activity would seriously mess with their data, and their analyses.

For the tracking of OS usage numbers, of course, this basically renders every previously posted result (i.e. the "legacy data") completely irrelevant. Therefore, I will let it go, and start over fresh, much as NMS have themselves. It's important to keep some sense of the history here, through, so... let's see what NMS have on tap for history.

Annoyingly, the new table which accompanies this graph doesn't break out each month's numbers in a nice grid, but rather shows only the average for the period. Each month's numbers are only available as mouse-over tooltips, which is worse for at-a-glance comparisons but probably better for NMS's business model (which is, after all, to sell the numbers to subscribers). Also, the new NMS data doesn't go all the way back to Windows 10's launch, so we can only look at the range from May of 2016 through the end of November, 2017.

Even so, the botnet-less trend lines look rather different than the botnet-filled versions. Mousing over the data points on that graph, we can learn that Windows 7 currently holds 43.12% of the market, basically unchanged from last month's 43.05%, while Windows 10 holds 31.95% of the market, an uptick from last month's 28.86%. Interestingly, Windows 10's growth has come mostly at the expense of Windows XP, which declined from 7.64% to 5.73%.

Generic "Linux" only shows up as 1.65% of the market, down slightly from last month's 1.85%, although that doesn't include Ubuntu (which is also Linux) or Chrome OS (which is also Linux), etc. That kind of fragmentation is part of the problem that Linux has in attracting new users, though, so that seems fair enough, and at least NMS is now listing Chrome OS separately as a desktop OS, which will allow us to track its progress over time - potentially very interesting. Combining all the various versions of each OS together gives a better overall picture of how Linux-based OSes compare to Windows:

Exactly why botnets would be inflating Linux' OS market share numbers, rather than Windows', is something that I'd like to see explained in more detail at some point. Linux is famously more secure than Windows, generally, so it would make far more sense for botnet traffic to be inflating Windows' market share results, instead. I'll be keeping an eye out for a better explanation there.

In the meantime, though, we have a new, and hopefully more reliable, baseline for watching the progress of the OS market from this point on. It'll be interesting to see what happens. I'm still not expecting to see much movement for the next few months, for the reasons previously stated, but I wasn't expecting NMS to revamp their entire business in a month, either, so who the fuck knows? If nothing else, it should be interesting to see what happens.

NetMarketShare numbers revisited.

NetMarketShare has revised their earlier posted numbers for OS market share, and these ones look a little bit more like what I expected.

First, the market share by version:

(Click to enlarge)

This version of the August stats shows Windows 7 losing, and Windows 10 gaining, both by less than ±0.5% - not enough to be very significant, which is what I expected for August. Windows 8.1 and XP also drop slightly, but again by less than ±0.5%.

The overall picture is, as one would expect, equally boring:

(Click to enlarge)

Windows still loses overall market share to Linux, but it's a smaller swing, still putting Windows just above the 90% mark (90.70% overall), and Linux at only 3.37%, which is more reasonable. MacOS actually drops slightly in this version, but by less than ±0.5% - not enough to be significant. Interestingly, while Linux's number in the overall matched Other's version column in the early morning version of these numbers, they're different here, which makes me wonder what the rest of Other's 7.84% market share consists of. Google Chrome, perhaps?

So, what does this all mean? As I said with the early morning's, more-extreme version of these numbers, it's hard to say. The tech media machine is doing all they can to help Microsoft out by hyping the upcoming Fall Creators Update, so maybe this is just a temporary blip; it's certainly not the seismic shift that the early morning version of the stats looked like it might be. But it's still another month of glacial adoption rates for Microsoft's flagship product, combined with signs that their customers may be abandoning Windows entirely - Windows' overall market share is still down, and Linux is still up. If next month's numbers repeat that trend, then Microsoft could be in serious trouble.

For now, though, it appears to be business as usual, with a complicit tech media machine mostly ignoring this month's market share statistics altogether, instead focusing on the upcoming Fall Creators Update. Wayne Williams at BetaNews appears to be the only other person I could find who's covered this at all, so far:

According to NetMarketShare, in July Windows 10 grew its usage share by 0.83 percentage points -- its largest increase in three months. [...] With a new major update to the operating system right around the corner, you might expect Windows 10 to have grown its share by a similar figure in August, but no. It's back to the glacial growth we usually see for the new OS.

That glacial pace is, I think, the problem here. Honestly, it's hard to blame the tech media for their apathy, when Windows 10's adoption rate keeps on being a story in which almost nothing has happened from one month to the next, for a year now. The Fall Creators Update is a non-story, too, but at least it's a new non-story.

And that's a shame, because spending a little extra time on this story, looking at the state of play over the last year, is actually rather fascinating.

Windows 7 loses market share at last... to Linux!

OK, I'll admit it: my predictions regarding the end-of-August market share numbers were dead wrong.

I'd been confidently predicting that not much of anything would change for the end of August, in much the same way that nothing changed at the end of July, or June, or May, and so on, but it looks like Microsoft may have finally convinced Windows 7 users to switch operating systems. NetMarketShare have put up their end-of-August numbers OS usage statistics, and Windows 7's share of the market has contracted by 1.90%, which is considerably more than the ±0.5% "noise threshold" (my guess at NetMarketShare's error bars).

I doubt that Microsoft will be celebrating, though, because Windows 7's loss did not turn into a gain for Windows 10 -- Microsoft's current flagship product declined, too, by 3.83%, giving back months of gains. Windows 8.1 dropped, too, by 0.65% (again, over the ±0.5% threshold). In fact, the only version of Windows that gained in usage share appears to be Windows XP, which only ticked up by a negligible 0.67%, basically offsetting Windows 8.1's losses.

What does it all add up to? Well, it adds up to an overall 5.15% loss of market share for Windows. Yes, you're reading that correctly: after sitting comfortably at about a 91.5% share of the market for months, Windows overall share of the market has just dropped to 86.3%.

(Click to enlarge)

 Broken out by version, the trend looks like this:

(Click to enlarge)

Just to be perfectly clear: I did not see this coming.

Speaking of Microsoft's anti-consumer bullshit...

We're nearing the end of yet another month, and I guess Windows 10's adoption rate has not improved since the end of July, because Microsoft are scare-mongering again.

From NeoWin:

For the past few months, Microsoft has been touting the security features available to customers using Windows 10, especially with the increase in ransomware attacks. It even vowed to raise the bar for security with the upcoming Fall Creators Update.

Now, the company has published yet another blog post persuading customers to upgrade to the latest version of Windows 10, citing increasing security threats.

Microsoft's blog post apparently includes a handy flowchart of Windows 10's "constantly evolving threat mitigation," just in case you'd missed the point from the text. Apparently, Fear ≈ Windows 10, in Microsoft-logic. The problem for Microsoft is that they've been doing this for months already, without any noticeable result. They're the corporation that cried "wolf." It's all so transparently self-serving that nobody's really listening anymore.

We'll know for sure in another few days, but I'm going on record and predicting yet another month of essentially flat OS market share trendlines, with Windows 10 gaining no more than 1%, all of which will come at the expense of Windows XP and 8/8.1, rather than Windows 7. Place your bets!

Still pretty damn flat

After months of scaremongering about WannaCry and other potential ransomware attacks, Windows 10 has managed to actually tick up to just over 27% usage share, according to NetMarketShare:


"Other," a category which includes Linux, is now the third most popular OS in the market, ahead of Windows XP and Windows 8.1, and MacOS X seems to have stopped losing users to the Windows ecosystem. Microsoft's flagship product, meanwhile, is still not thriving. Adding Windows 7's loss to Windows 10's gain gives a swing of 0.96%; at that rate, assuming it stays constant, it will take eleven more months for Windows 10 to reach parity with Windows 7. That's parity by next fall, even with Microsoft giving Windows 10 away for free, to anyone willing to switch.

Some of the lickspittles in the tech media are, of course, presenting this as some sort of seismic shift in Windows 10's favour.

Today in Windows 10...

We're definitely into the summer doldrums, with very little of anything happening, but here are two tidbits that caught my eye today.

First up, from Mike Wheatley at siliconANGLE:

Although Microsoft Corp. probably deserves a ten out of ten for the efforts its made to make Windows 10 more enterprise friendly, it turns out the operating system is only slightly more popular than the aging Windows XP among business users.

That’s the main takeaway from a new survey on Windows 10 adoption by information technology network Spiceworks Inc. The survey found that Windows XP, which is by now most likely riddled with security holes as it no longer receives regular updates, is still being run on 11 percent of business PCs, down from 14 percent in March. That compares to just 13 percent of business PCs running Windows 10.

“Despite the gains in Windows 10 penetration, the absolute share of computers running the OS remains relatively low,” said Peter Tsai, senior technology analyst at Spiceworks.

Both operating systems remain a long way behind Microsoft’s legendary Windows 7 operating system, which is running on 68 percent of all enterprise PCs, Spiceworks found. However, Windows 10 does at least come second with its 13 percent adoption rate, and its overall share has improved markedly in recent months – back in March, it was only running on nine percent of business computers, Spiceworks said.

Ouch.

Bear in mind that Windows 10 is the future of Microsoft, on which they've basically bet the farm at this point; with general usage languishing at or below the 26% mark for months, Microsoft was counting on business adoption to make their Windows 10 strategy work. If these numbers are to be believe, then that isn't happening yet, at least not in any big way; the fact that Windows 10 is only finally outperforming the nearly sixteen year old Windows XP, after everything that Microsoft has done, must surely speak to some significant underlying issues. I doubt that the recent revelation that older hardware may not be supported by future Windows 10 builds will help, either.

I'll admit, I am looking forward to the end-of-July NetMarketShare stats, but I'm not expecting Microsoft to make any changes to their approach this year. Thousands of annual layoffs and stagnant market growth for their flagship product notwithstanding, the markets seem generally bullish on MSFT these days, which gives Satya Nadella all the cover he needs to keep right on failing.

That's all for tidbit one, though. Tidbit two is getting a lot more coverage, like this piece in The Reg:

Happy Canada Day!

It's July the 1st, and you know what that means... time to check in on NetMarketShare, and see what June hath wrought, as far as Windows 10's market share is concerned. Spoiler Alert: Not much has changed.

When last we checked, the OS landscape looked like this:

One month later, the picture now looks like this:

And the trend lines look like this:

Yep... still flat.

Windows 7 did dip slightly, from 49.46% to 49.04%, but that doesn't reverse all of its gains from the previous month, and is starting to look like just noise in the data - Win7 has been oscillating back and forth like this for months, now, but hasn't dipped below 47.01% in the past year; 49.04% is almost exactly where it was at the end of June 2016 (49.05%), so it would seem that (a) Windows 7's user base really is well and truly dug in, and (b) that they comprise just under half the OS market overall, and about 54.99% of the total Windows market.

Windows 10, meanwhile, managed a feeble 0.02% gain (from 26.78% to 26.80%), putting it pretty much exactly where it was the previous month. Other gainers for June included Windows XP (from 5.66% to 6.94%), which didn't quite manage to get back to 7% but came pretty close, and Linux (from 1.99% to 2.36%). Windows 8 (-0.22%), MacOS 10.12 (-0.10%), and "Other" (-0.58%), posted losses, but "Other" was the only category to show a change of more than ±0.5% (the "noise" threshold), and all of them are looking more like fluctuations than trends at this point.

The "Other" category, BTW, appears to consist mostly of older MacOS versions, and has been steadily declining for the last year, from 13.66% a year ago to only 7.33% now, a period over which MacOS has generally lost market share to Windows' combined versions (with Windows gaining 1.72%, and MacOS losing 1.75%, since last July). Windows' dominance on the desktop is still secure.

That's about it, though, as far as good news for Microsoft. Their hopes for the future have been pinned entirely to Windows 10, and Windows 10 is completely stagnant, for yet another month. The much-predicted Enterprise migration to 10 has failed to materialize yet again, and another month of scare-mongering by Microsoft's apologists about malware vulnerability in XP and 7 has resulted in XP recovering some of its previous month's losses while 7 stands pat. In fact, Windows 7 is looking more and more like the next XP, and with the Universal Windows Platform looking less and less viable as each of these stagnant months limps past, I think it's safe to say that Microsoft's Windows 10 strategy isn't working.

Small wonder, then, that we're seeing rumours of a possible name change starting to make the rounds, but I have the feeling that Microsoft will need more of a plan B than that. So far, though, there's no sign of one.

UPDATE:
It looks like I slept better than normal last night, because betanews' Wayne Williams beat me to the punch:

According to NetMarketShare’s figures, Windows 10’s share of the desktop operating system market remains pretty uninspiring, with growth much slower than you’d expect.

In fact in a year, the new OS has grown by just over 5 percent. In comparison, Windows 7 grew by 2 percent in the same time frame.

Of course, this is usage share, rather than market share, but in some ways that’s more important as it shows the percentage of people actually using the different operating systems.

[...]

Will July be any kinder to Windows 10? We’ll find out in a month’s time.

We don't actually need to wait a month for the answer that question; unless Microsoft change course significantly, July is going to play out just like June did, and as May did before that, and April before that, and so on, and so on, and so on. Windows 10 is the only available choice for new PCs, so it's market share (and usage share) will continue to creep upward, but Microsoft really need for Windows 10's upward movement to happen faster than a creeping pace - that's the whole point of giving it away for free, after all, something that they're still doing, in spite of having officially ended the free giveaway almost a year ago.

Better slowly. than not at all...

From the notes on Windows 7 update KB2952664 [emphasis added]:

This update performs diagnostics on the Windows systems that participate in the Windows Customer Experience Improvement Program. The diagnostics evaluate the compatibility status of the Windows ecosystem, and help Microsoft to ensure application and device compatibility for all updates to Windows. There is no GWX or upgrade functionality contained in this update.

Well, well, well. It seems that they do learn, after all. Slowly, to be sure, and at some great cost, but they do seem to be finally getting the message.

Other good news? This month's security updates also include further patches for the exploit employed by WannaCry, and were released to Windows XP concurrently with Windows 7, 8, and 8.1, and for free, rather than after several months' delay of trying to sell expensive extended support packages to XP customers. So I guess they learned that lesson, too.

It's nice to see tacit acknowledgement from Microsoft that they and their Windows 7 users don't still have the same relationship of trust and good will that existed just two years ago, but it's also a clear sign of how seriously that relationship has deteriorated. How badly must Microsoft have fucked this up, that a note promising a clean update, i.e. with no upgrade bullshit, is even necessary? How many Windows 7 users do you think will avoid installing these updates anyway, just to be "safe?" How many of them still have Windows Update turned off completely, thanks to Microsoft's GWX abuses?

Microsoft have a long, long way to go, yet, to get back into their customers' good graces. Considering that they haven't yet actually apologized for all the bullshit they've pulled in the last couple of years, it's fair to say that you haven't even really started to make their way back. They do seem to be thinking about it, though; I just hope they learn the real lesson of their XBO-X failure, though, and start working to earn redemption before it's too late, and not after.

"WannaCry" ported to Windows 10

Remember when WannaCry was making the rounds, and Microsoft's apologists were taking advantage of that to scare reluctant WinXP and Win7 users into switching to Windows 10? Well, it turns out that Windows 10 may not be as safe as all that, after all, because the same exploit that WannaCry... exploits also works on Windows 10. Oops!

From threat post:

The NSA’s EternalBlue exploit has been ported to Windows 10 by white hats, meaning that every unpatched version of the Microsoft operating system back to Windows XP—and likely earlier—can be affected by one of the most powerful attacks ever made public.

Researchers at RiskSense, among the first to analyze EternalBlue, its DoublePulsar backdoor payload, and the NSA’s Fuzzbunch platform (think: Metasploit), said they would not release the source code for the Windows 10 port for some time, if ever. The proof of concept has been in the works since the ShadowBrokers’ April leak of Equation Group offensive hacking tools targeting Windows XP and Windows 7, as well as the development of a Metasploit module based on EternalBlue released two days after the WannaCry attacks. The best defense against EternalBlue, researchers maintain, is to apply the MS17-010 update provided in March by Microsoft.

So, it seems that the only advantages that Windows 10 provided were: (a) that not enough users had switched to Windows 10 for black hat hackers to bother targeting it, and (b) Windows 10 users have no control over Windows Update, which means that they were updated in spite of themselves...which is something of a mixed blessing.

This is part of the problem that Microsoft have when pitching Windows 10's alleged security superiority over Windows 7. Redmond had spent years working to convince Windows users that their OS was every bit as safe as any other OS on the market; Linux and MacOS may have been targeted by fewer attacks and exploits, but that was just because Windows was so much more popular than they were. Well, guess what? That worked. And now Windows 7 users aren't buying it, when Microsoft try to convince them that safe-as-houses Windows 7 is full of fatal flaws.

That's the catch-22, the cleft stick in which Microsoft find themselves. If Microsoft were telling the truth before, then Windows 7 is basically as safe as any other OS, and users have no reason to believe Microsoft's recent scare-mongering. On the other hand, if Microsoft were lying before, then Windows has never been safe, and users would always have been better served by switching to Linux, because why would Microsoft be any more trustworthy on the subject now, when they have a vested material interest in lying to us? Having spent two years eroding users' trust with their abusing GWX shenanigans, intrusive "privacy" policies, and monopolistic bullshit, Microsoft's customers simply aren't willing to listen as they WannaCry wolf.

GG, Microsoft. Good job.

Windows XP not only didn't spread WannaCry - it couldn't

It turns out that most of the WannaCry story that everybody thought they knew is actually wrong, and Microsoft's motives for patching Windows XP to defend against the malware attack may be even murkier than was previously reported.

From Computerworld:

Rather than take aim at Windows XP, WannaCry targeted Windows 7 and Windows Server 2008, Kaspersky's data showed. [...]

The reason for XP's absence from the WannaCry count was simple. "WannaCry itself did not support Windows XP," [Costin Raiu, director of Kaspersky Lab's global research and analysis team] said, noting that the exploit neither focused on XP or reliably worked on the 2001 operating system. Individual machines could be infected -- the researchers and testers who put WannaCry on Windows XP systems likely ran it manually -- but the worm-like attack code would not spread from an XP PC, and in some cases, executing the exploit crashed the computer.

That put Microsoft's decision to issue a security patch for Windows XP in a different light. [...] Computerworld, like many other publications, assumed Microsoft released patches for Windows XP and Server 2003 because it believed older -- and unprotected -- systems were instrumental in spreading WannaCry.

Raiu thought different. "I think Microsoft was worried about the possibility of someone leveraging this exploit," Raiu argued. "Their fear was that it could be theoretically possible to repurpose the exploit to attack Windows XP."

It wasn't a surprise that WannaCry's backers had primarily pointed the attack at Windows 7. "They focused on the most-widespread platform," said Raiu.

According to analytics vendor Net Applications, approximately 53% of all Windows personal computer ran Windows 7 last month. That was nearly double the share of the newer Windows 10, which clocked in at 29%, and more than eight times that of Windows XP's 8%. Cyber criminals typically aim attacks at the most popular operating systems and versions within each OS, a logical practice when profit is paramount. That's especially true of extortion rackets like WannaCry's payload, which encrypts files and then demands a ransom payment to decrypt those hijacked files.

It's hard to say whether this makes Microsoft's decision to shake down Windows XP customers for more "custom" support contracts before finally patching the vulnerability for free look slightly less shitty, or even more so. After all, if WannaCry couldn't even affect machines running Windows XP in its extant form, then Microsoft were essentially shaking down customers like the UK hospital system for "protection" against a threat that actually posed more of a threat to their Windows 7/Server 2008 machines than it did to their Windows XP/Server 2003 PCs. The fact that less harm may have resulted from the delay than was previously believed mitigates the shittiness somewhat... but only somewhat.

Microsoft apologists used headlines that blamed Windows XP for the spread of the malware to blame the victims, telling them to just switch to Windows 10, already, and the same apologists are predictably using this latest news to argue that Windows 7 users should do the same. It's an argument that conveniently ignores the simple fact that Windows 10 was no more the target of WannaCry than Windows XP was, for the simple reason that big, rich corporations and other large institutions haven't yet adopted the latest iteration of Microsoft's OS. Whether the WannaCry outbreak will drive people towards Windows 10 or not, remains to be seen; with most of the early headlines blaming XP for the outbreak, many Windows 7 users may already have lost interest, especially since Windows 7 has already been patched to defend against the WannaCry exploit.

Can I call "backsies" on that?

A couple of days ago, I was praising Microsoft for patching Windows XP to protect users of that old OS against the WannaCry ransomware that was spreading like wildfire through organizations like the NHS. I even said that it was better that they did it late, than that they not do it at all, and praised them for not exploiting the situation to shake down WinXP users for more money, or to push them to switch to Windows 10, either of which would have been more in keeping with their pattern of behaviour over the last couple of years.

Today, however, I'm taking all of that back. Because it turns out that Microsoft had the XP WannaCry patch ready to go months ago, held it back while they shook down their customers for more money, and only finally released it for free once the unpatched vulnerability started taking down hospitals.

From Tech Times:

Microsoft, which called out the NSA and other government agencies for their role in the creation and launch of WannaCry, may itself have been part of why the ransomware was able to cause so much chaos.
As the world attempts to recover from the damage caused by WannaCry, a new report claims that Microsoft could have helped prevent its spread, but decided not to do so.

According to a report by the Financial Times, Microsoft held back a free update that would have patched up the vulnerability that WannaCry used to compromise computers running on the old Windows XP operating system.

The report claims that Microsoft delayed the rollout of the patch because it initially wanted payments of up to $1,000 per Windows XP computer for "custom" support.

Microsoft has struggled to push users and corporations to upgrade from older versions of the Windows operating system to the latest and secure Windows 10, even if the company had already stopped the support for versions such as Windows XP. The significant number of users who have not yet upgraded to Windows 10 were highly vulnerable to WannaCry when it started its worldwide rampage last week.

Microsoft still continues to provide support for governments and organizations, but in exchange for hefty payments. While the company offers special deals for the first year, the high costs have forced entities such as the National Health Service of the United Kingdom to discontinue receiving support.

The National Health Service turned out to be one of the biggest victims of WannaCry, as it spread across 150 countries and infecting about 200,000 computers.

That is so much bullshit, in one tidy package. The fact that Microsoft had the sheer gall to be complaining about spy agencies' stockpiling of these vulnerabilities, when they themselves were using the same vulnerabilities to shake the UK's hospital system down for an amount of cash that they damn well knew the NHS didn't have to spend, is reprehensible. Microsoft's blatant greed, and their wilful disregard for the consequences to innocent bystanders when their broken shit took down the UK's hospital system, all feels like something that should be actionable. If there isn't already a law against this, there should be.

Good job, Microsoft! You've managed to take the one halfway-decent thing you've done in the last two years, and turn it into bullshit. Of all the egregiously anti-consumer shit you've pulled in the last two years, this is literally the worst. Fuck you all.

And fuck the tech writers, too, who keep trying to blame the victims for having been victims here. And, yes, that includes Tech Times, who end their article with this chestnut:

However, the victims of WannaCry may also blame themselves for remaining unprotected against the ransomware attack. Many users and corporations could have prevented having their systems locked by the ransomware by upgrading their operating systems and installing the necessary updates, instead of subscribing to the theory of "if it's not broke, don't fix it."

According to Microsoft, it prefers for users and enterprise customers to upgrade to Windows 10 instead of having to pay for support for older versions of the operating system. It can be argued that Microsoft should have released the patch to fix the vulnerability that WannaCry exploited in Windows XP, but perhaps it would have been better off if customers were not on Windows XP in the first place.

There are reasons why the publicly-funded NHS hasn't replaced all of its fully-functional Windows XP machines with expensive new PCs, you dicks, and the hospital-specific software they're running may not even be compatible with newer versions of Windows. The fact that you'd even think to blame the victims for this, after it's been revealed that Microsoft actually tried to cash in on WannaCry by extorting money from the UK hospital system, is beyond the pale. 

The NHS's patients (also victims of WannaCry) are not at fault, here, and the NHS certainly doesn't bear any weight of culpability comparable to that of the actors who exploited this vulnerability for financial gain. That burden falls entirely on two sets of shoulders: those of the black hats who shipped this ransomware in the first place, and those of Microsoft, who tried to exploit the occasion to squeeze some extra money out of the UK's fucking hospital system. Fuck anyone who says otherwise, and fuck Microsoft, too.

Fuck.

Doing the right thing, because it's the right thing to do.

Microsoft has pulled a lot of anti-consumer bullshit over the last couple of years... like when they literally broke Windows Update for users of older versions of the OS that were running them on new PCs, and responded to the outcry by recommending that we just all just embrace Windows 10, already. The fact that users had to fix that for themselves, and did, does not in any way excuse that bit of bullshit, and that's really just the tip of the iceberg of bullshit that Microsoft's shovelled at consumers in the last couple of years. Suffice it to say that the bullshit is neither forgotten nor forgiven, and that occasions to actually praise the Redmond team have been pretty few and far between.

So when news broke earlier in the week about the massive "Wana Decrypt0r" ransomware attack, which was taking down hospitals in the UK and spreading like wildfire, I wasn't expecting Microsoft to offer much help to users of Windows XP. WinXP hasn't been supported by Microsoft for years, after all, and the fact that lots of hospitals still use it hadn't been enough to change Microsoft's mind about that before now; most articles that I read on the subject also took for granted that WinXP users were basically screwed, and needed to upgrade their PCs to something that could run Windows 10.

Microsoft, however, either decided that (a) the optics of of patching every other version of Windows against Wana Decrypt0r but leaving hospitals vulnerable were seriously sub-optimal, or (b) that the life-and-death realities of patching every other version of Windows against Wana Decrypt0r but leaving hospitals vulnerable were too awful to think about, or (c) both. Whatever the thinking was, though, they issued patch for Windows XP today that fixes the weakness that this ransomware was exploiting.

From bleepingcomputer:

Following the massive Wana Decrypt0r ransomware outbreak from yesterday afternoon, Microsoft has released an out-of-bound patch for older operating systems to protect them against Wana Decrypt0r's self-spreading mechanism.

The operating systems are Windows XP, Windows 8, and Windows Server 2003. These are old operating systems that Microsoft stopped supporting years before and did not receive a fix for the SMBv1 exploit that the Wana Decrypt0r ransomware used yesterday as a self-spreading mechanism.

[...]

Microsoft had released a fix for that exploit a month before, in March, in security bulletin MS17-010. That security bulletin only included fixes for Windows Vista, Windows 7, Windows 8.1, Windows 10, Windows Server 2008, Windows Server 2012, and Windows Server 2016.

As the SMBv1 is a protocol that comes built-in with all Windows versions, the computers which did not receive MS17-010 remained vulnerable to exploitation via Wana Decrypt0r's self-spreading package.

"Given the potential impact to customers and their businesses, we made the decision to make the Security Update for platforms in custom support only, Windows XP, Windows 8, and Windows Server 2003, broadly available for download," Microsoft said in a statement. "This decision was made based on an assessment of this situation, with the principle of protecting our customer ecosystem overall, firmly in mind."

Researchers believe that Wana Decrypt0r [...] infected over 141,000 computers [...] While unconfirmed, many believe older Windows XP and Windows Server versions were the bulk of the infections pool, as they had no way to protect themselves.

The customer ecosystem here, remember, disproportionately involves hospitals, and other essential institutions that are still using Windows XP because their publicly-funded budgets can't afford to upgrade all of their PCs. It would have been great if they'd patched those older OS versions last month, of course, or at least before so much damage was done, but better late than never. And I mean that sincerely, considering how many vulnerable PCs and servers are out there, it really is better that they did this now, than not at all.

Good job, Microsoft. You've done a good thing today, and one that nobody expected you to do. Now we just have to convince you to make this a habit...

No, Win7 users of Google's Chrome browser do not need to switch to Win10

Users of Windows XP or Windows Vista, however, may want to give Firefox another try.

From Digitaltrends:

Many reasons exist to upgrade to Windows 10, and for Windows XP and Vista users — which according to some data represent a bit more than 10 percent of all PC users — perhaps the biggest reason is for the night-and-day differences in support and security that Windows 10 provides. Google just offered another reason update to Windows, specifically that Gmail will reduce support for Windows XP and Vista, as Google announced on the G Suite blog.

While those users will still be able to access their Gmail messages, they will be doing so with the much less robust HTML version as early as December. The Windows version is actually a secondary cause of the reduction in functionality. More specifically, Google will be shifting all users running Chrome Browser v53 or below and it just so happens that the latest Chrome version supported on Windows XP and Vista is v49.

[...]

Google’s specific statement regarding the reduced functionality is as follows: “Gmail will continue to function on Chrome Browser v53 and below through the end of the year. Users who remain on Chrome v53 and below could be redirected to the basic HTML version of Gmail as early as Dec 2017.”

As a Windows 7 user who's currently running Chrome v56, I can say categorically that this does not affect users of Win7 and up (unlike some reports that you may have seen), so 47.2% of Windows users will not be affected by this. Chrome, however, currently runs on 57.9% of all PCs, so there's definitely some significant portion of PC's Chrome user base that will be affected by this. It will be interesting to see whether those users stick with Chrome and accept the reduced Gmail functionality, or switch to Firefox to get a better experience.

The odds of them dumping Gmail, naturally, I rate at basically zero. Switching your email address is such a pain in the ass to do, I can't imagine that anyone will bother with it, especially since the main cloud-based email alternative is probably Hotmail, which Microsoft is presumably keen to marry to both Bing and Edge.

Windows 10's market share creeps upwards

In spite of Microsoft's desperate scare-mongering, and even-more-desperate tactic of resuming the free giveaway of Windows 10, their new OS has managed to tick upwards by slightly less than 1%, according to new stats from NetMarketShare.

Just to recap, these were last month's stats:

And these are this month's: 

Windows 8.1 remained unchanged, while Windows 7 ticked downwards; Windows XP ticked upwards very slightly, but that's probably just noise in NMS's data, rather than actual gains. Interestingly, MacOS and Linux are both up a bit, although it remains to be seen whether those are real gains that will persist into next month, or just transient blips.

So, what does it all mean?

With Windows 7 and 8 no longer available, their share of the OS market should be going down steadily, even if only from attrition, i.e. as old PCs wear out and get replaced with new ones. Windows 8.1 was showing steep decline until this month, but seems to have stabilized; Windows 7, which was the target of Microsoft's scare-mongering last month, seems to have declined slightly, but only slightly, and still holds 47% of the global desktop/laptop OS market.

Perhaps most tellingly, Windows 10's market share did not increase as much as Windows 7's decreased, with Linux and MacOS also making gains. It would seem that some people were indeed scared away from Windows 7... straight into the arms of Microsoft's competitors.

Overall, though, the OS market place didn't change much. Again.

This is a problem for Microsoft, who desperately need for the market place to swing away from Windows 7 and towards their new OS. Every month that goes by with it not happening weakens the narrative that MS are trying to promote: that Windows 10 is a runaway smash hit, and due to take over from Windows 7 and 8.1 any day now. Enterprise customers, in particular, don't seem to be in any hurry to adopt Windows 10, or to buy new PCs, and individual users seem to have dug in with their old OS. Windows 7 is looking more and more like the next Windows XP, something that Microsoft was desperate to avoid.

Meanwhile, every couple of weeks sees a new headline about how Microsoft is adding more bullshit to Windows 10, from taskbar pop-up ads for their own (broken) Chrome extensions, to returning to the practice of reinstalling bloatware on Home users' PCs, even after users have uninstalled it. Their steps in the direction of finally recognizing users' privacy concerns as a valid thing, while welcome, haven't gone nearly far enough yet, and their practice of forcing updates on Home users who serve as unpaid product testers for MS's Pro & up licence-holders shows no sign of ending anytime soon.

Windows 10 is a shit product, full of bullshit "features" that no sane user wants any part of, and users are continuing to refuse the switch, especially given that switching clearly costs more than just the purchase price. It doesn't sell itself on its merits, something which Microsoft has tacitly admitted with their scare-mongering and renewed giveaway efforts; as long as that remains true, I don't expect the overall picture to change much, any time soon.

The week's bad and ugly in Windows 10 (and the good in other OSes).

I'd mentioned MS's new bit of advertising bullshit in passing yesterday, but its badness really deserves a closer look. Fortunately, PC Mag have us covered:

Microsoft really wants you to use its software products as well as running Windows 10, and that includes the Edge browser. But it can't stop you choosing to use an alternative web browser. However, if you opt to use Chrome, then expect to start seeing adverts right on your Windows desktop.

As Myce reports, if you have Chrome installed and the icon present on the Windows Taskbar, chances are you're going to start seeing a pop-up advert appear [...] suggesting you install Microsoft's Personal Shopping Assistant Chrome extension. Microsoft touts it as "Your smart shopping cart across the web."

Opting to install the extension results in Microsoft monitoring which products you've searched for and viewed while using Chrome, and then offering to compare those products to find the best price. There's also alerts when prices change, and the ability to track products across all your devices. Of course, Microsoft will make money if you opt to purchase any products using the Assistant.

[...]

Reviews of the extension, as you've probably guessed, are not great. Users refer to it as "spam" and "adware," and there's multiple calls for Microsoft to stop advertising it on Windows 10. "The assistant is poorly designed and implemented" and "Spammed by Microsoft to install. Doesn't really do much to assist in my shopping" sum up the kind of feedback Microsoft is generating.

This is a move that I've come to regard as classic Microsoft: rather than making something actually good, that people would actually want to use, MS have banged out a piece of crap that even they have no confidence in, and are now abusing their control over the OS to push it on users in the most aggressive and intrusive way possible. It's bullshit, of exactly the kind that inspired nearly half of all PC users to stay with Windows 7 rather than be subjected to this sort of nonsense.

Worse yet, it's coming at exactly the time when MS are also fucking up their sales pitch to desperately-needed Enterprise customers. From Digital Journal:

This week, Microsoft publicly claimed its own Windows 7 operating system is "long-outdated" and based on insecure security models. The statement has been criticised by security experts as the OS is still three years away from the end of support.

Microsoft posted a lengthy piece to its German newsroom in which it said an "early goodbye" to Windows 7. It suggested the aging platform cannot support the "increased security requirements" of modern devices and internet users, advising customers upgrade to Windows 10 for increased protection.

[...]

Its post has been criticised by security experts though. According to people in the industry, there is no reason to migrate away from Windows 7 today, interpreting the company's claims as an effort to increase Windows 10's adoption. Some experts suggested Windows 10 is actually less secure than Windows 7.

German Windows expert Günter Born published a response to Microsoft's article on his blog that explains the company's flawed logic. Born cited a recent study from Carnegie Mellon's CERT division that presents in great detail evidence showing Windows 10 is less secure than Windows 7 in some situations.

Windows 7 with Microsoft's EMET security software offers better protection than Windows 10 without, according to the study. In its enterprise marketing materials, Microsoft implies Windows 10 does not need EMET. Last year, it announced EMET will be discontinued on July 31, 2018, to the protest of many users.

[...]

Born also noted that Windows 10 isn't what many business customers want. Microsoft's focus on attracting new users and appealing to consumers has led to feature bloat that isn't desirable on enterprise systems.

"Machines with Windows XP and then Windows 7 have been a solid foundation for my SoHo business," said Born. "Windows 10 isn’t what I need as a SoHo business user. It’s focused on things Microsoft’s marketing identified as 'good for the company’s revenue.'"

Again, a now-classic Microsoft move, obviously prioritizing their own interests over their users' to the point of actually making Windows 10 less attractive as a platform, all at exactly the same moment when their own interests rely heavily on Windows users migrating to the Win10 en masse. GG, Microsoft. GG.

Maybe that's why I'm starting to see articles like this one from Paul Thurrott:

With Microsoft looking at new ways of monetizing Windows 10—in many cases, with ever-aggressive advertising in the product itself—I’m getting a lot of questions about alternatives. And while the case for moving off Windows 10 on PCs is not clear, I feel the frustration too.

In fact, I spend a lot more time than you may realize exploring those alternatives. This week alone, I’ve done work on macOS, using my MacBook Air, and I’ve installed the latest versions of Ubuntu and Mint Linux. In other recent weeks, I’ve spent time with a surprisingly high-quality Acer Chromebook as well.

I do this with no sense of joy. And to be clear I still find Windows 10 to be the obvious winner when I evaluate what it is that I’m looking for personally. And that’s true regardless of my job: Were I to suddenly hit the lottery, I’d keep using Windows 10 myself. I mean, I’d probably buy a really expensive Surface PC. But I would stick with Windows 10.

This isn’t the case on phone, of course, nor is it true on tablets. On these form factors, Windows 10 is either unusable (phone) or pointless because of the lack of ecosystem support (tablet). Phones and tablets, though, are more clear-cut: You should choose iPhone or Android on phone (I prefer iPhone), and iPad if you want a tablet. (I am not aware of a single decent Android tablet, which I still find curious.)

His list, oddly, places the wildly unpopular (and no longer available) Windows 8.1 at the very top, followed at #2 by the much more popular Windows 7; MacOS sits at #3, which probably isn't much of an option for most users given that you'd need to buy an overpriced Mac PC to get it; the much more affordable Chromebook clocks in at #4. 

The most affordable option, Linux, is last on Thurrott's list at #5. Linux will run on any PC you own, of course, and is completely free, but it doesn't natively run Windows apps, and its bewildering assortment of distributions/versions, and its learning curve, are clearly hurting it. Linux developers clearly need to start doing a better job of working together to "sell" their platform to new users, but Mint looks like one of the simplest available options. 

Thurrott's list isn't interesting as much for its recommendations, however, as for the fact that it exists at all. Thurrott likes Windows 10, remember, something that he mentions more than once in the article, but has been fielding so many inquiries from readers wanting off the Microsoft hamster wheel that he felt compelled to compile a list of recommendations. When even Windows 10's fans are talking about alternatives to the OS, it's probably not a good sign for Redmond.

In a couple of weeks, we'll get the first OS Market Share numbers of the new year. Thanks to Microsoft's hamfisted ineptitude, I'm expecting them to look almost exactly the same as the last numbers of last year: with Windows 7 still wildly popular among PC users, and Windows 10 growing very little, if at all. Satya Nadella's team truly need to re-think their approach, here, if only for Microsoft's own sake, and there are some hints of awareness on their part about how badly the ham-fistedness is hurting them.

Still, I'm predicting very little actual change in their tactics until closer to the end of the year, when Windows 10 still won't have taken off with Enterprise users in a way that even Nadella can't ignore anymore. Whether Nadella is actually capable of overseeing that kind of a tactical, and possibly even strategic, shift, and whether Microsoft's board of directors will be willing to let him try, are yet to be seen. Nadella's bet heavily on Windows 10, and it's not a stretch to say that his own success is heavily tied to the success of the platform; if the platform fails to thrive, Nadella's days as head of Microsoft may be numbered.

It was bound to happen eventually

With Windows 7 & 8 no longer available for sale, and Windows 10 now the only option available for buyers of new Windows PCs, Microsoft's new OS was going to see its market share start to increase this month. The only question was how much. Well, we now know the answer to that question: not that much.

This was October...

... and this is November.

From InfoWorld:

Windows usage reports for November are in, and the numbers are sobering. There's no call to start building a Windows 10-sized coffin, but it’s definitely time to stow the confetti and air horns. And if you're expecting to see a billion Win10 devices, you should be looking for an alternate universe.

Windows 10 adoption continues to disappoint, with NetMarketshare saying that OS accounted for 23.72 percent of all desktop OS usage in November, compared to 22.59 percent in October (see screenshot). The slight uptick stands in marked contrast to the Win10 ratchet slope that ended in August—the month when free upgrades to Win10 and the “Get Windows 10” campaign ceased.

Since January, Win10 usage has increased by nearly 12 percent, according to NetMarketshare. Over that same period, Win7 share has decreased by 5.3 percent, Win 8.1 by 2.4 percent, and XP by 2.8 percent. Windows 7 continues to dominate the desktop, with more than 47 percent of measured November usage—higher than it was in July.

Interestingly, while Windows XP may be down compared to January, it actually regained some ground over the past month, ticking up from 8.24% to 8.63%; Linux was up, too, from 2.18% to 2.31%; and MacOS gained as well, from 6.43% for all versions combined to 6.74%. 

But the big number, the one that everyone is watching, has to be Windows 10, which has finally crossed the 23% boundary for the first time since flirting with it (at 22.99%) in August. And all it took was to remove its two most popular competitors, Win7 and Win8, from the market. Microsoft, for the win! Pay no attention to the fact that Edge is still losing ground to Chrome and Firefox, on the browser side of things.

Again, from InfoWorld:

Browsers were another story, as Microsoft continues to drop precipitously.

According to NetMarketshare (screenshot), Chrome and Firefox use was up slightly in November, while the combination of IE and Edge share fell from 28.39 percent in October to 26.87 percent in November. Individually, both IE and Edge usage declined.

I guess the hard-sell tactics aren't working, eh?

So, yes, by eliminating Windows 10's most popular competitors from the market, Microsoft have finally managed to see some growth in Windows 10's share of the desktop OS market, which is significantly better than they'd been doing for the last couple of months. None of the earlier versions of Windows, including Windows XP, will be going anywhere anytime soon, though, and with Windows' other competitors are both showing signs of renewed life, Redmond should still be very nervous about the slow pace of Windows 10 adoption.

UPDATE: The Reg has a different, and rather interesting, take on the numbers.

This is something that hadn't occurred to me. From The Register:

As we've seen before, Windows 10 use spikes at the same time Windows 7 usage dips. Trust us: this is a weekend effect. Over at the right-hand edge of the chart you can see the impact of the Thanksgiving holiday, on which it looks like Americans really did spend time with family rather than online. The next day we see a smaller-than-usual weekday bounce for Windows 7 and then a normal weekend surge for Windows 10.

Windows 7's weekday decline is also obvious in the graph.

So, not only is Windows 10's growth spurt pretty marginal, it might be ephemeral, too? Suddenly, I can't wait for January 1st, so that we can test that hypothesis.

Windows 10 lost market share to Windows 7 in September

I was expecting Windows 10's update problems and privacy issues to have an impact on its market share. With adoption by home users slowing after the end of Microsoft's free "upgrade" offer, and with adoption by business of Pro and Enterprise versions also reported to be lacklustre, I was expecting Win10 to gain basically nothing in terms of market share; maybe there would be a slight uptick for Win10 as the market shares of Win8 and XP continue to decline, but I expected that Win7 would continue to be #1, with a market share number that was basically unchanged from August.

Well, NetMarketShare's latest desktop OS market share numbers are now posted, and Win8 and WinXP both showed the expected slight downticks, while Win7 remained the #1 desktop OS, but that's the end of the list of my met expectations. Because Win10's market share actually ticked down in September, from 22.99% to 22.53%, while Win7's share rose from 47.25% to 48.27%. Linux also gained a bit, rising from 2.11% to 2.23%.

For the record, this was August: 

And this is September: 

BetaNews' reaction to this was... priceless, really:

WTF? Windows 10 now actually losing market share

By Wayne Williams

It was expected, that once Windows 10 stopped being free, upgrades would slow significantly. That turned out not to be the case last month when NetMarketShare’s usage figures showed it, rather surprisingly, to be business as usual. Growth in August was no different from growth in previous months, although I speculated it might have been buoyed by sales of new back to school PCs.

In September though, according to NetMarketShare, Windows 10 didn’t just show slower growth, it actually went into reverse gear and lost usage share. Yes, you read that right.

According to the figures, Windows 10 went from 22.99 percent globally, to 22.53 percent, a drop of 0.46 percentage points. It’s important to remember that NetMarketShare measures usage (people actually using the operating system, rather than having it installed), and that isn’t a precise science. Even so, Windows 10 losing share is a big surprise. When Windows 8.x did it two years ago, it came after months of dwindling growth. Here, NetMarketShare is showing us a healthy growing operating system coming to a dramatic and sudden stop, and then actually rolling backwards a bit. Are the figures to be believed?

Well, while it wouldn’t be the first time that NetMarketShare has released usage numbers and then revised them a few days later, rival usage share monitoring firm StatCounter has similar findings.

Remember earlier this week, when Microsoft announced that Windows 10 was now running installed on over 400 million devices, a claim which I described as lying with statistics? There's a reason why I reacted that way.