slide

Microsoft windows is dead, long live lite os

Ned Bellavance
8 min read

Cover

On last week’s Buffer Overflow we were talking about this very strange blog post from Microsoft about a “Modern OS” and what it should include. It’s obvious that Microsoft has something big brewing over in Redmond. In the build-up to \build, rumors were flying fast and free that Microsoft was going to unveil their new Lite OS and provide a roadmap for development. That did not happen, and the blog post from Computex seems to indicate that while Microsoft is definitely developing something that is not Windows, they aren’t quite ready to share with the world what that “Modern OS” actually is. The more I think about it, I believe that Microsoft might be poised to come barreling back into the mobile market from a totally unexpected direction.

Let’s review the past few years shall we? Microsoft has increasingly become a services company. That is the entirety of their focus. Every move they make is about increasing customer consumption of their services, including Azure, Office 365, Xbox Live, GitHub and more. The main reason they killed off Windows Phone was the simple fact that there was no financial incentive to compete against iOS and Android. Microsoft wants people who are using iOS and Android to consume their services on those platforms, and therefore their focus turned towards making their services integrate better with those devices. It also means that Microsoft doesn’t especially care about Windows on the consumer side. Sure they make some money from selling Windows licenses to consumers, but the lion’s share of that revenue comes from corporations. I think you’ll see all Windows 10 development focus squarely on enterprise friendly features, while the consumer side becomes more and more anemic. Consumers for their part are flocking to mobile devices and Chromebooks. That’s where we are today when it comes to client operating systems. Microsoft just isn’t that interested in developing Windows 10 for the consumer. And why would they be? Windows 10 has 20 years of cruft tied to its neck, making it a fragile beast at the best of times. Sure, Microsoft has done a lot to stabilize the platform, but given the chance to design a new operating system from the ground up, Microsoft would never build something that functioned like Windows 10. And that is especially true from an internal structure side.

So let’s assume that Windows 10 development is all about the enterprise customer at this point. Microsoft is left with a hole where their consumer OS could be. Given what I just said about Microsoft just wanting to make services better on iOS and Android, why would Microsoft develop a competitor to those operating systems? I’ve got three reasons:

  1. Microsoft is trying to drive hardware innovation
  2. Controlling the stack leads to better outcomes
  3. There is a long game, and Microsoft is playing it

Let me expand on these a little bit.

Microsoft is trying to drive hardware innovation

One of the key themes of Computex was the need for better devices and more innovation. Intel came out with Project Athena to try and drive hardware manufactures to develop new and better hardware. Some of the OEMs, like Asus and Lenovo, had some pretty odd designs with different screen types and placement. We’re in a bit of a slump when it comes to consumers purchasing new mobile devices and laptops, and innovation is one way to drive sales and demand. Microsoft has been trying to do a similar thing with its Surface line of devices. In the same vein as Google’s Pixel line of devices, Microsoft is trying to get OEMs to up their game by showing a path forward. It seems to have worked, but now there is a new problem. These innovative devices need an operating system and applications that can take full advantage of unique form factors. Foldable screens, heads-up displays, ancillary screens in strange locations, and even stranger form factors require an operating system that is just as flexible with it’s shell and UI. Is that OS Windows? I’ll pause for a moment while you mop up whatever just shot out of your nose.

If Microsoft wants to stay relevant on new platforms and form factors, it has to be able to support them. I think that support includes building an operating systems because of point two.

Controlling the stack leads to better outcomes

Microsoft could certainly wait for iOS and Android to get in the game and support all these wacky new devices. Then they can rewrite their applications to try and take advantage of the features exposed on iOS and Android, and hope that things work really well. That’s not a great bet, and certainly not one I would take. Despite their best efforts, many Microsoft products still deliver a subpar experience on iOS and Android and even MacOS. The best experience is usually on a Windows device, because Microsoft controls the whole stack. It’s the same reason that Apple builds the hardware, operating system, and apps for many of its services. Controlling the entire stack leads to a better experience. Better experiences lead to happy consumers that want to use more of your services. And what is Microsoft all about? Ding, ding, ding! Selling services.

Windows is not going to be the platform to make all this magic happen. Instead, Microsoft has been drastically evolving their approach. For instance, Edge has adopted Chromium for its rendering engine, and that has made it orders of magnitude faster. Most of our applications run in a browser these days, and having a first-class browser that you own and manage should be a priority for Microsoft. So should first class support of WebAssembly to power native applications running in a browser.

Microsoft is ditching the current Windows-only .Net framework for a cross-platform implementation called .Net Core. The current version of .Net Core is 3.0, and going forward Microsoft is going to drop the Core label and make .Net Core the only .Net Framework for the future. Microsoft has also announced that the next version of Windows will ship with a Linux kernel that can run the Windows Subsystem for Linux v2 in a super-lightweight VM that doesn’t require Hyper-V. The future of applications is containers, and that’s not just for server side deployments. In the Computex blog post they talked about isolating the hardware and operating system from the applications that run on it, in a similar vein to how iOS handles things. Containers can do that handily, and you can package up a Windows legacy app in a container if you need to by using the technology being developed by Droplet Computing.

Most of the device innovation is going to happen on the consumer side of the house, enterprises tend to move more slowly. But Microsoft needs to get ahead of this trend now, which brings me to point three.

There is a long game, and Microsoft is playing it

When it comes to making money off of an operating system, enterprise money is where it’s at. Microsoft already has Windows 10 as a subscription as part of the Microsoft 365 package. But if you look to the horizon, there is trouble brewing. Chromebooks, tablets, and mobile phones are steadily being adopted by organizations as a primary device. As new form factors and device types are adopted by the consumer space, they will also slowly trickle into the enterprise. While Windows 10 may not die off entirely, a steady decline seems inevitable, and Microsoft needs to head that off at the pass. Hence they are putting money into develop a completely new operating system that meets the current and future needs of consumers and enterprises alike.

Building a successful and fully feature operating system isn’t going to happen overnight. Adoption of it won’t happen overnight either. But if Microsoft can gain critical mass with a new operating system in the consumer space over the next five years, they will be poised for the adoption of their new operating system into the enterprise.

I think it’s also clear that we are in the infancy of device form factors. The smartphone as we know it is less than 15 years old. We’ve pretty much maxed out what a slab of flat glass can do for us at this point, and I think that the next killer device will come along in the next five years. As tied as we are to our phones, they still deliver a sub-optimal experience when it comes to interacting with the rest of the world. As technology, hardware, and software progress we are going to reach a tipping point where new and better devices are possible and we have the software to make them viable. Don’t forget, the Newton existed back in the 80s, but the technology and software wasn’t ready yet to support the vision. Windows Phone may forever be dead, but Microsoft is betting that the next big device (whether it’s an AR headset, overlay contacts, or a neural implant) isn’t all that far away, and this time they don’t intend to miss it.