I’m going to make a bold prediction here: I think that in 10 years (or shorter), Windows will either disappear, or use the Linux kernel by turning into a sort-of Linux distribution (not like Ubuntu or Debian, but more like Android).

You might wonder why I think that. After all, Windows dominates the desktop and laptop market, having roughly 87% of the market. Meanwhile, Linux has less than 3%. Why would this change in 10 years? Surely Microsoft is gonna continue developing Windows and keep making tons of money off of it. Right?

Well, I don’t think so. And to understand why, I first have to explain why Windows is popular.

Why is Windows popular?

Imagine computers as being on a scale, from smaller and less powerful to bigger and more powerful. Windows dominates the devices around the middle, ranging from laptops to desktops. Everything else is dominated by Linux - from routers to smartphones, from servers to supercomputers. Even Microsoft uses Linux to run their cloud services. But if Linux runs all this stuff, why is it so unpopular on the desktop?

Linux Torvalds - the creator of Linux - was asked this question. His answer is basically that Windows comes pre-installed on these devices, because consumers don’t want to install their own operating system. They just want to take it home, turn it on, and start working on it. It’s why Android is so successful - not because millions of people download and install an ISO every day, but because it comes pre-installed with the phone. Windows is successful for the same reason - because it comes with desktops and laptops.

So why would that change?

The problem with Windows

Windows isn’t any better than Linux. I’d even argue that it’s worse: Linux has a central, rebootless update system that upgrades not just your operating system, but also all your software. Everything from the kernel to your music player gets updated in the background, while your computer is running. And the next time you turn it on, you immediately get to use the latest software. You also install all your software from a sort-of app store, and you can customise which app stores you want to use. The updates are also not forced on you: you get to decide when to update, and what to update.

Windows on the other hand, while it finally got an app store after “10” versions, you still have to install most of your software from all kinds of different websites. And each piece of software either has a built-in update mechanism, or not, and you have to update it manually. Meanwhile Windows updates itself during shutdown and startup, which depending on how much it has to do and how powerful your computer is, can take a short time or a really long time.

These issues aren’t new, however. They existed from the very beginning, and it didn’t stop people from using Windows. I’d argue that there is only one thing good about Windows: everyone’s using it. And because of that, all the programs are written for Windows. All the computer manufacturers install Windows on their devices. And everyone knows how to use Windows. This makes it near impossible to de-throne Windows as the king of the operating systems, however there are 2 reasons why I think it might happen: portability, and development cost.

Portability means the ability to run on other systems. Windows needs a certain amount of RAM, a certain type of CPU, and this and that. Linux however, doesn’t really have any hardware requirements. It can run on your toaster - and probably does, if you have a smart toaster. This is because Linux is open source, and so are most of the applications written for it. So if a new architecture was made, Linux could be easily run on it by just making a compiler for the architecture and suddenly you’ll have most Linux software running. Windows on the other hand, can only really be used on the x86 architecture. Sure, there is an ARM version of Windows called Windows RT, but we all know what happened to that. And it happened for a good reason: almost all Windows applications are designed for x86, making it hard for Windows to change architectures. This could become a big problem very quickly, because of Apple’s new ARM-based chips. These are ridiculously fast yet consume a lot less power, and I’m certain that ARM-based processors are the future of computing. Linux has existed on ARM since forever, not just because of Android, which runs on ARM, but also because of many single-board computers like the Raspberry Pi. In fact, Linux exists for pretty much every architecture, due to it being open-source. Windows however, is closed source, and so are most of the apps written for it. Because the main selling point of Windows is that it has the most applications, creating a new version of Windows that runs much faster is pointless if people can’t use the apps they’re used to.

The other issue I see is development cost. Because Windows needs to keep adding new features while making sure to not break old ones for applications, the cost of developing the operating system goes up a lot. Meanwhile, Windows is constantly becoming cheaper and cheaper, to the point where nowadays it’s given away almost for free (remember that 1 year when you could upgrade to Windows 10 free of charge? You can still do that, even though the program ended 4 years ago). Of course, organisations still have to buy licenses in bulk, and manufacturers who want to put Windows on their devices also buy a lot of licenses, however, Microsoft is increasingly making money from other sources, like Azure, ads in Windows, the Microsoft Store, and selling Surface devices. This gives Microsoft an incentive to either abandon Windows, or to cut development costs for it, which I think is the more likely option. And an excellent way to cut development costs would be to switch kernels, and thus remove the need, or at least cut down on the need for kernel developers.

WSL

A lot of people seem to think Windows Subsystem for Linux hurts Linux, because it means that people will be able to run Linux apps easily on Windows without switching operating systems. I actually believe the exact opposite - it provides an easy way to try out Linux for anyone who’s curious. People don’t just start using Linux out of nowhere. Most Linux users - including me - start out with having a dedicated machine, physical or virtual, running Linux. As they spend more and more time getting to know Linux better, they’ll use it more and more, and slowly integrate it more and more into their workflow. Once someone realises that they spend more time in Linux than in Windows, they’ll switch operating systems. This is because Linux runs Linux applications the best, similarly how Windows runs Windows applications the best. This will always be true, no matter how hard each system tries to emulate the other. I suspect most people don’t use Wine in their workflow, and instead choose to run Linux-native applications, because those simply run better. The same will probably be true for WSL, meaning that most people will either rarely use WSL, or use it so much that they’ll just switch over to Linux. Anyone who’s in between will only be so temporarily.

Some people however (like Richard Stallman) have accused Microsoft of trying to embrace, extend and extinguish Linux. This is Microsoft’s internal name for a tactic that has the following 3 steps:

  1. Embrace: Create a piece of software that’s similar or even compatible with a competitor’s software.

  2. Extend: Implement some features that are not supported by the competing software, and make it proprietary to their own product.

  3. Extinguish: Once the improvements to the product become widely used, the competition that doesn’t or cannot support these new features will go out of business.

I am, however, not worried that this will happen. Many corporations want and do try to control Linux, however, none of them will be able to. This is because the main benefit of Linux is that all the code is developed in the open - companies like Intel, IBM, Samsung, Google, AMD, ARM, Facebook - they all have their own agenda, features that they want, and having these features made in an open way allows anyone - from the biggest tech companies to random people - to work on it. If Microsoft tried to put proprietary features into Linux, it’d take away it’s biggest feature, the only thing that makes Linux good and successful.

This is also why Linus Torvalds - the creator of Linux - isn’t worried about Microsoft taking over Linux either. Many companies have tried and failed, and Microsoft wouldn’t be the first, not by a long shot. However, I think Microsoft is aware of this. Their contributions and interest in Linux seems genuine, and it looks like any other company that has joined the Linux community.

On another note, the name Windows Subsystem for Linux is weird, because it’s a Linux system running on Windows, meaning that Linux is the subsystem, not Windows. I don’t know if it’s just a poor naming choice, or perhaps foreshadowing - that one day Windows will be running under the Linux kernel.

Closing thoughts

This is my end of year prediction for what’s to come. I will certainly be looking back at this in 2030. Maybe I’ll eat my own words, maybe not, but one thing is certain: the moment Microsoft ports Office over to Linux will be the first nail in the coffin for Windows. I don’t know when will that happen - if at all - but Microsoft has already made one Office app available on Linux: Teams. Sure, it’s not Word, or Outlook, or even SharePoint, but it could be the start of something bigger. We’ll just have to see.