Thinking About Windows in 2020 (Premium)

I’ve been writing about Windows for over 25 years, and while it is still central for me, it has never been less relevant or less well-cared-for by its maker than it is right now. So, yes, I have questions. Lots of questions. And complaints. Lots and lots of complaints.

But first, let’s examine the facts that led us to this time of great uncertainty and see how reality shapes our perception of what can—and most likely will—happen to Windows in 2020.

The first signs of trouble arrived in the early 2000s, when Microsoft was besieged by antitrust action in the United States, Europe, and elsewhere. Many cite the resulting “lost decade” as evidence of Microsoft’s decline, when a demoralized software giant sat by as a new generation of Big Tech—Amazon, Apple, Google, and even Facebook—rose to collectively take its place atop the personal computing industry.

That view is correct, but looking at Windows specifically, it was the broken promises of Longhorn, which was both years late and completely neutered by the time it arrived as Windows Vista, which triggered the great decline. This was when Microsoft’s developer base simply stopped embracing every new SDK and framework that the firm foisted on them. It’s when they stopped listening.

Related to this, the personal computing market went mobile in the 2000s and 2010s, with the back-to-back releases of iPod, iPhone, iPad, and the many copycats of each that appeared in their collective wakes. Today, both Android and iOS command markets that are larger than Windows, in Android’s case, many times larger.

Microsoft never responded adequately to these threats. There was little it could do to generate excitement for new application development on the Windows desktop as long ago as the release of Windows Vista. And Steven Sinofsky’s foolhardy push into mobile was just as doomed thanks to his terrible decision-making: Instead of embracing the Windows phone application model for Windows 8, he simply created something that was similar to it, resulting in two incompatible platforms. We’re still reeling from that idiocy. In fact, it was the death blow to Microsoft’s mobile ambitions.

Not coincidentally, this month will mark a terrible new milestone: The PC industry has no doubt experienced 8 straight years of decline—the final numbers aren’t in quite yet—and is regardless now under two-thirds the size it was at its peak in 2011. The last time the PC industry sold so few PCs was 2005, when we were still breathlessly waiting for Microsoft to complete Longhorn and change the world.

Microsoft’s inability to extend Windows to new markets over these years is telling. Windows phone eventually failed, thanks to Sinofsky. Microsoft was forced to use Linux to create Azure Sphere because Windows doesn’t scale well enough to work on truly small IoT devices. Windows 10 S/S mode, yet another attempt to wash Windows of its legacy Win32 desktop underpinnings, failed. And Windows 10 on ARM is both flailing and failing, and will no doubt be consigned to the dustbin of history alongside Windows RT, Zune, Windows Media Center, and Cortana soon enough; we already have Intel-based PCs with 20+ hours of battery life, so what’s the point of ARM again?

Tied to all of this, too, is the abject failure of Windows as a Service (WaaS), an oxymoronic belief that Windows, a mess of gigabytes of code spanning decades of time, can somehow be updated, yes, like an online service. How silly is WaaS? Microsoft has tried to ship two major new versions of Windows each year since 2015, and has failed every single year; its mobile competitors—Android and iOS—are only upgraded “only” once each year, and Apple, in particular, has failed to do that successfully twice in the past three years alone. Worse, online services are updated one time, in the cloud, making the process easier and more reliable. But there are almost 1 billion Windows 10 PCs in the world, each with its own configuration, making such a process difficult and error-prone.

And before anyone tries to defend WaaS, I’ll just point to 2019, when Microsoft shipped exactly one major Windows 10 upgrade, and it took several long months to do so because the previous two were so buggy and it wanted to finally manage the process manually. The second major upgrade was really just a normal monthly cumulative update that was disguised as a real feature update. And the cheering that occurred after that success is particularly telling. The only thing we can really celebrate here is that Microsoft finally figured out a system that sort of works. And it is undermining that learning by threatening to go back to the old system for 2020.

And then there’s the Insider Program, which like Windows itself is now sans a leader of any kind and is stumbling along in the dark, spewing out inaccurate and incomplete communications about how the program even works anymore. It’s embarrassing and unprofessional, and it is perhaps the clearest sign of all that Microsoft doesn’t care one whit about Windows, despite it generating over $10 billion per quarter last year. It boggles my mind.

So here we are, in January 2020, with a new year to ponder. Granted, there is some excitement, mostly (OK, solely) in the form of Windows 10X, which Microsoft has inaccurately publicized as being designed for a new generation of dual-screen devices. (It is, in fact, just the latest attempt to simplify Windows generally, in line after Windows RT and Windows 10 S/S mode.)

And yes, like an idiot, I’m somewhat excited by Windows 10X, of course. The residual feelings of 20 years ago come flooding back each time, despite the intervening decades of defeat. But you have to wonder. Why would Windows 10X succeed where its predecessors failed? Why would dual-screen PCs do a thing to slow, halt, or even reverse the PC market sales slide?

And the answer is, they won’t.

Windows 10X, like the dual-screen PCs on which it will run at first (before moving to more traditional form factors), isn’t designed to usher in a mythical resurgence in the PC market. It’s designed to slow the fall. That’s it.

Put simply, I don’t personally believe that Windows 10X will have much of an impact. Because, hell, even the retirement of Windows 7, the most successful Windows version of all time, didn’t have much of an impact on PC sales this past year. Because Microsoft embracing a device form factor that others have already innovated has never panned out, not once. Because PCs are what PCs are, and the market for this kind of device has reached its new and smaller plateau. Because that continued shrinking of this market in the long run, despite what may happen in any one quarter, is the new normal.

You can see the problem by looking at the Windows 10X developer story: There isn’t one. Interested developers, and you have to think they’re queuing up by the tens to port their HoloLens and Surface Hub apps to a new form factor, have to email Microsoft—yes, really—to indicate their interest. There is no online hub or documentation of any kind. Folks, this system is almost certainly complete, internally at Microsoft, right now. And they’re still not talking to developers in any meaningful way. It’s almost like the Insider Program is in charge.

Yes, I know. This is all very negative, even for me. Sadly, I feel it’s accurate.

That said, it’s not all bad, assuming you can get past the notion that Windows will suddenly surge to any kind of relevance in the years ahead. As I’ve written in the past, I like that Windows is now all about productivity, about getting work done, as less important and more entertainment-driven experiences have long-since moved to other device types. And despite the in-box advertising, the bundled crapware, and the telemetry we can’t disable, it’s still the best version of Windows ever, and it is much more interesting and useful than macOS, Linux, or Chrome OS. If it wasn’t, I’d have moved along years ago myself.

The PC industry, too, continues to spit out ever more unique and innovative form factors, as evidenced most recently by this week’s CES announcements. Apple users, meanwhile, have few choices, but at least they’re really expensive and most of the keyboards don’t work. Linux users can just hope and pray that everything works correctly on whatever Windows PC it is they really want. And Chrome OS fans just get the PC industry’s leftovers for the most part, usually in the form of lackluster configurations. If you’re in the market for a new PC, there has literally never been a better selection.

That said, things could be better.

That in-box advertising, bundled crapware, and telemetry we can’t disable should be removed from Windows 10. Microsoft should own up to its WaaS mistakes and move the platform to a more reasonable schedule, with the 2019 experience being a baseline. The Windows Insider program needs to be shaken, stirred, and upended, and it needs to commit to both transparency and clear communication. And my God, Microsoft. $10+ billion in revenues per quarter: Why doesn’t this important product have direct representation on Satya Nadella’s senior leadership team? That’s a kick to the sensitive bits.

And that, really, is the problem. From the top (SLT) to the bottom (Insider Program) of Microsoft, Windows just doesn’t get the respect or attention it deserves. And that is what needs to change in 2020, before it’s too late.