No New Version of Windows Can Turn Back the Clock, and Microsoft Doesn't Want To

Later this week, Microsoft is set to unveil the latest version of Windows, Windows 11. There's been a fair bit of chatter about the new OS version in the past 10 days, thanks in part to the leak of an early version of the operating system (exactly how early is not known).

Over at Axios(Opens in a new window), longtime tech journalist Ina Fried argues that Microsoft's desktop OS refresh "can't come soon enough." Her argument is straightforward and accurate: Microsoft has been losing ground to other operating systems for years. When Windows 10 launched in 2015, Microsoft still held ~85 percent of the total PC market. Today, it's down to 73.5 percent.

Microsoft earns far more money from non-Windows and Windows-adjacent businesses today than it did ten years ago, but as Fried points out, a full third of its revenue still comes from Xbox, Windows, and Surface, before counting Office. Given this fact, the company's rapidly declining market share is arguably a cause for concern.

There's a joke about Windows 10 with some bearing on this issue. You may have seen the image before:

The fact that we laugh at this image -- and the accuracy of its feedback -- is more telling than it might seem.

Once upon a time, people absolutely did give each other OS recommendations. When Windows 95 launched PC owners lined up around buildings and across parking lots, the way people would to buy an iPhone 12-14 years later. If you were to go digging around in archival copies of PC Magazine, PC World, or Byte, you'd find a lot of articles on topics like "Which PC and/or operating system is right for you?" from the early 1980s through the early 1990s. Even after WinTel buried every competitor but Apple, the significant software and hardware compatibility differences between Windows NT and Windows 9x kept these articles pertinent and topical.

Microsoft helped bring the era of general-purpose, consumer-oriented OS recommendations to a close when it launched Windows XP in 2001. Few computing transitions have been more praised. Windows XP was far more stable than Win 9x had ever been and unifying its consumer and professional product families gave Microsoft reason to improve its software compatibility and gaming performance relative to Win 2K. It was absolutely the right move for Microsoft to make. It was what Microsoft's customers wanted. But once Microsoft became the default choice for more than 90 percent of the market, the only question left to answer was whether the new version of Windows sucked more or less than the previous one.

I don't doubt that Microsoft would love to return to the days when PC owners wrapped themselves around buildings for the chance to buy an operating system, but there's no realistic chance of it doing so. The reason people lined up around the block to buy Windows 95 is that it promised a quantum leap forward for PCs could do and how they were organized. The Windows desktop, taskbar, Start Menu, and File Explorer all debuted (or debuted in something like their current forms) for the first time with Windows 95. Task Manager would come along a year later, in 1996.

When Microsoft positioned Vista as a dramatic reinvention of desktop computing, users hated it. They praised Windows 7, which was seen as a modest, welcome course correction. Microsoft tried to reinvent the PC UI again with Windows 8. Users hated it. I won't claim that Windows 10 got the same sunny reception Windows 7 did, but its UI was generally hailed as an improvement -- and, again, a return to form. Customers have embraced Surface as a brand, but they hated Microsoft's first attempt to build ARM devices. When people bought a Windows device, they expected the ability to run Windows x86 applications. Surface RT couldn't do that, and frustrated customers returned the device in droves. The characteristics that defined Windows in its early days have become something of a straitjacket in the modern era.

Microsoft Isn't Looking Backwards

I've never been shy about criticizing Microsoft, but the company has done a remarkable job in building a business on cloud computing in a way it never did in mobile. Under Ballmer, Microsoft built Xbox up from nothing while largely ignoring its own ownership of PC gaming. Windows got better versions of DirectX and games became more stable once the Windows GPU driver model improved, but there was no practical benefit to being both an Xbox gamer and a PC gamer. The two ecosystems were entirely separate.

Today, services such as Game Pass and game streaming between Xbox and PC offer benefits whether you play on an Xbox, a PC, or both. Microsoft's xCloud service extends this idea to tablets and smartphones running non-Microsoft operating systems. Office is available for iOS and Android. Windows 10 and 11 will run on ARM hardware, with x86 support (64-bit and 32-bit both) provided via emulation.

Windows rose to dominate the computing industry at a time when it was sold almost exclusively for a single type of CPU (NT4 on Alpha is an exception, don't @ me). Today, Windows runs on ARM and x86 CPUs alike. If RISC-V continues gaining prominence, we'll undoubtedly see a version of Windows for it as well. In the late 1990s and early 2000s, Microsoft was known for its hostility to open-source software. Today, Windows 10 ships with an integrated Linux kernel.

In the 1990s, Microsoft sold Windows on the promise that each new version of the OS was sufficiently transformational to justify whatever headache it took to get it running. Today, Microsoft markets its products as enabling you to access what you want to access, or doing what you need to do, with little to no headache at all. Feel like some AAA gaming on a smartphone? xCloud has you covered. Want to stream games from your Xbox to your PC? No problem. Need to access Office remotely? How about cloud storage integrated directly into Windows? Need to run Linux without the hassle of rebooting or prefer an ARM CPU? Microsoft has you covered.

I genuinely don't know what features Microsoft will unveil on Thursday. Even if Windows 11 has some bells and whistles under the hood that we haven't heard of, however, it's not going to be the fundamentally transformative product it would need to be to reverse the market share gains that macOS and Chromebooks have collectively made.

It's not clear how much this matters. The idea that Windows users must learn Windows in childhood in order to become lifelong OS users is widely believed, but no longer necessarily true. When I was a child, I computed on a Commodore 64, a TRS-80, an x86 PC running DOS, and a Windows 3.1 PC all within a few years of each other, and with overlap between them. This was not uncommon if your school had a hodgepodge of out-of-date computers. Today, a child will almost certainly be exposed to at least two, possibly three operating systems out of the five major ecosystems available today (Windows, macOS, iOS, Android, and Linux) before they hit third grade.

The only reason for Microsoft to have made so many fundamental changes to its business in a relatively short period of time is because the company reads the tea leaves the same way Axios does. Microsoft's marketing team will probably issue some remarks about how Windows 11 is the most powerful, transformative, competitive OS ever to be released, but it's a pro forma insistence at this point. Microsoft isn't trying to recapture the glory days when it bestrode the PC market like a titan and ruled supreme over more than 90 percent of desktop and laptop shipments. It's far more focused on creating cloud services that encourage people to subscribe to Microsoft platforms and services.

I don't think Microsoft is betting it can bring customers back with a new, amazing version of Windows. I think Microsoft is betting it can build an ecosystem that attracts and retains customers, whether they happen to run Windows or not.

• Microsoft Would Like You to Stop Downloading Leaked Windows 11
• Windows 11 Boosts Hybrid x86 CPU Performance Over Windows 10
• We’ve Tested Windows 11 Ahead of Next Week’s Launch