Microsoft's Universal Windows Platform is not dead, but it has evolved over the years

What you need to know

  • New claims that UWP is dead mischaracterized what is really happening.
  • Microsoft did shift UWP app strategy over two years ago to focus on desktop.
  • Developers now have more ways to bring apps to Windows 10.
  • UWP is still the primary dev platform for future Windows experiences.

Updated November 11, 2019: We're re-upping this post in response to Microsoft emphasizing its plans for the evolution of UWP apps at Ignite 2019. The original story, initially published on June 7, 2019, follows.

It was just over two years ago when I wrote that Microsoft was pivoting away from Universal Windows Apps (UWA) based on the Universal Windows Platform (UWP), and instead focusing more on desktop-style applications for the Microsoft Store. The reasoning at the time was obvious: Windows Phone was on its last gasps, and without it, developers had one less target for UWA.

Fast forward to 2019, and Thurrott and The Verge are both running headlines that seemingly spin things. Both articles have truth to them, but there's also a more considerable misconception out there – mostly from non-developers – about UWP, Microsoft's announcements at Build this year, and what the future may hold.

Today, I want to straighten the record slightly, but first, let's define what we mean by UWP.

Nuance is important

Microsoft UWP or UWA?

The Universal Windows Platform goes back to Windows 8, and it was meant to usher in a new age of Microsoft and computing. Back then, the mantra was "three screens and a cloud" referring to Xbox, PCs, and phone.

Of course, what people mean when talking about apps like Microsoft News, Weather, Mail, and apps from third-party developers is Universal Windows Apps – apps that run across all three devices with few changes.

Conversely, the "universal platform" part in UWP refers to the shared APIs and resources that developers have access to when writing an app, not the app's hardware destination. This distinction is crucial as we'll see below.

There's a difference between the Univeral Windows Platform and Universal Apps

Microsoft has conflated these terms frequently, using UWP for shorthand. This mixing of terminology was especially true during the Windows 8 days when even desktop PCs featured UWAs as the primary app experience.

Tom Warren, from The Verge, said, "This [UWP] dream really started to fall apart after Windows Phone failed, but it's well and truly over now." I'd argue differently. The original fiasco goes back to Windows 8 and its failed tablet strategy where UWAs were supposed to shine. Once Microsoft rolled back the new Start Menu experience in Windows 8.1 – and ditching it entirely in Windows 10 – UWAs lost momentum in PC.

The decline and eventual loss of Windows Phone only made matters worse.

Undoing the damage

Microsoft's developer mistake(s)

The failure of Microsoft's Windows tablet and phone experiences is undoubtedly a significant contributor to the de-emphasis of UWP.

But one other reason — which Microsoft has been trying to rectify these last few years — was the insistence that developers convert all their "classic" Windows apps to UWA using UWP. This approach was all-or-nothing and driven heavily at Microsoft's Build developer events between 2013 and 2016.

Microsoft will meet developers wherever they are regardless of dev platform.

To be fair had Windows 8, PC tablets, and Windows Phone taken off UWP and UWA would be heralded as ahead of its time. Instead, it failed and with it the broad ambition of UWA. (Apple and Google are ironically running with the idea now.)

Developers bristled at the move. Mary Jo Foley of ZDNet got as much from Kevin Gallo, Corporate Vice President of the Windows Developer Platform, who said "we shouldn't have gone that way" noting the eventual Win32 and UWP divide that it caused.

Case in point, UWP could not match the power of 20+ years of Win32 development — it was too green. With missing APIs and features even if developers wanted to port a mapping app over, if the mapping API was incomplete, or lacked features they needed, there was little motivation to do so.

It's true that Microsoft did not make it easy for developers to port to UWP and create UWAs. But that started to change, and it was especially prevalent at Microsoft Build this year, where the company took a more conciliatory approach: we'll meet developers wherever they are.

Open up to devs

Just call 'em Windows Apps

From Build (2019) "State of the Union: The Windows Presentation Platform"

From Build (2019) "State of the Union: The Windows Presentation Platform"

This point brings us to where Microsoft is going with all these changes today. For Microsoft, the future isn't only about UWA, but about WPF, Win32, and WinRT too. Toss in support for XAML Islands, React Native, Electron, or being able to use Chrome's JavaScript engine for Progressive Web Apps (PWA) for good measure.

Under Microsoft CEO Satya Nadella, the company has been moving to an inclusive approach to the company's structure, services, and yes, app development. The company is not shying away from UWA or even UWP, but they are backing down from the "all or nothing" mantra. Instead, the company wants developers to use whatever tools they have to bring apps and games to Windows 10 and the Microsoft Store.

Sharing his opinion on the matter, Matt Velloso, Technical Advisor to the CEO at Microsoft, had this to say:

Why does it [UWP] need an identifier? The more the walls separating these different Windows development platforms crumble, the more developers needs can be met where they are, with their existing code, that can evolve, mix and use whatever works best. We want no cliffs. Pick whatever, evolve it whatever way... Life is too short. I don't get all the "this needs to beat that" and stuff.

The point by Velloso is crucial. Windows never really had a primary dev platform (Win32 was close) because of the open nature of the system. Developers have been porting apps, bringing new languages, emulating, sandboxing, scripting for years. It's what makes Windows great.

Nothing has changed regarding UWP, but how we talk about it changed years ago

With Microsoft bringing XAML Islands to Win32, or Fluent Design Language to iOS, Android, and the web, it's clear Microsoft wants to be the one-stop developer house for all developers. The days of app developers being platform-centric are fading away. The future is PWA, Electron, React, or JavaScript — things that UWP are already compatible with.

UWP targets: From "Building UWP apps for Multiple Devices" (Build, 2019)

UWP targets: From "Building UWP apps for Multiple Devices" (Build, 2019)

As announced at Build, these new tools let developers take existing code from legacy apps and by piecemeal, versus a complete rewrite, improve them using UWP. It gives developers the flexibility to adopt UWP at their own pace. As to why developers would want to do this, there are legitimate performance improvements in switching to partial-UWP as well as enhancements from XAML Islands (UI elements from UWP that can now work in Win32 applications).

Why UWP is not dead

Universal Windows Apps are still important

New "Legere" UWP app for Reddit released June, 2019

New "Legere" UWP app for Reddit released June, 2019

But are true Universal Windows Apps dead? No. In fact, they play a critical part in Microsoft's future computing experiences.

Moving away from desktop PCs to HoloLens 2, Surface Hub 2, Windows on ARM, IoT, and Windows Lite these systems are much more dependent on UWA. While Microsoft will let such devices run "classic" Win32 apps — emulated or virtualized — Win32 is by no means the primary dev platform for such systems (I'd argue Win32 is more is closer to maintenance mode than UWP).

That's where all this "UWP is dead" talk gets strange as Microsoft is betting huge on holographic computing and things like dual-screen PCs and lighter laptops as part of its future — that's UWP. The idea of people firing up Win32 Adobe PhotoShop in HoloLens or Surface Hub seems quite improbable. That's because Win32 apps are meant for desktop PCs with powerful x86 processors, not ARM, light computing, or holographic experiences.

The very basis of Windows Core OS and Windows Lite is built on UWP as the main app layer. Without UWP, Microsoft could only make legacy experiences, not new ones. There is no alternative to it.

TL;DR

Microsoft UWP and what it all means

HoloLens 2

HoloLens 2 (Image credit: Windows Central)

Since the failure of the Windows 8 tablet strategy followed by the demise of Windows Phone, Microsoft's vision of Universal Windows Apps (UWA) lost a lot of impetus. Combined with developer distaste to being forced to rewrite apps as UWA and non-matured developer tools Microsoft had begun shifting away from this approach starting back in early 2017.

As desktop and laptop PCs became the focus of attention again, Microsoft opened the Universal Windows Platform to cross-platform tools like Electron, PWA, JavaScript and older systems like Win32. With the announcements at Build 2019, Microsoft is now allowing developers to mix and match differing developer technologies with UWP to meet developers where they are now.

Pure cross-device Windows Apps still play an important role, especially for future, non-legacy experiences

So-called "pure" UWAs still have a vital role for Microsoft and developers. These apps serve as the primary app platform for Windows Core OS, Xbox, HoloLens 2, Surface Hub, and IoT. While these areas of computing still pale in scope to "classic" x86 PCs, Microsoft believes these newer systems will grow in importance, especially with the shift to cloud and ambient computing in the coming years.

PC manufacturers are also now using UWP to deliver customized app experiences, configuration tools, and drivers through the Microsoft Store, part of Microsoft's Universal Windows drivers program (opens in new tab).

Nothing has changed in the last year regarding UWP or UWA. Microsoft's failed strategy for tablet and phone is years old, but the company is adapting to the times. Looking forward towards next-gen computing experiences — versus legacy desktop PCs — UWP still plays a critical part serving as the primary (but not only) app platform for those systems.

UPW is not dead, it's merely one of many tools developers have to bring great apps and games to Windows devices.

Daniel Rubino
Editor-in-chief

Daniel Rubino is the Editor-in-chief of Windows Central, head reviewer, podcast co-host, and analyst. He has been here covering Microsoft since 2007 when this site was called WMExperts (and later Windows Phone Central). His interests include Windows, Microsoft Surface, laptops, next-gen computing, and for some reason, watches. Before all this tech stuff, he worked on a Ph.D. in linguistics and ran the projectors at movie theaters, which has done absolutely nothing for his career.

164 Comments
  • Funny. I read the article and I get that, *technically,* it's not dead. But nevertheless, because "Windows Core OS, Xbox, HoloLens 2, Surface Hub, and IoT" don't *really* exist yet, in the sense that I've never seen one, and may never become widely available, I can't help but think, "Yup. Dead."
  • Gears of War 5, Minecraft and other Xbox games were made in Windows UWP. Other software like the new Mail/Calendar were made in UWP too. Xamarin uses UWP to build apps that could be uploaded to the Windows Store. Saying the platform is dead is LOL Worthy.
  • Gears 5 is "confirmed" for Steam. If it can run on Steam, it's not a UWP app of the kind we've all come to know and love (or, rather, almost no one knew and even fewer loved). As you note, a few OS applications are UWP, but their most of their "real" versions (e.g., Outlook for Mail/Calendar) still can't be found in the store. Saying the platform isn't dead when MS won't eat their own dogfood is LOL Worthy.
  • Windows 10 is coded on this “universal thought”. Almost all of Microsoft’s inbox apps is builded on UWP, and most shells is builded in favour of UWP. Please explain to me, why UWP is dead, if UWP is in every corner of Windows.
  • Pls send me the store link of the teams uwp app
  • Because "P" stands for the platform. If there are no 3rd party apps it is not a platform.
  • There are 3rd party apps, so technically it is a platform lol.
  • That's not necessarily so. Unity is on both win32 and UWP. Some code bases can be compiled with minimal changes with different outputs. UWA can be technically side-loaded, they just generally aren't. Just because it's on steam, doesn't mean it doesn't exist as a UWA. If gears is a unity based game, it could be both. As a UWP, it would be instantly compatible with xbox, and PC at the same time. Porting it to steam would be just an extra platform, like having a playstation version.
  • Pretty sure Gears 5 is not on Unity, which is still not really ready for AAA console or PC games. Unreal is more likely as an engine. (yeah. just confirmed. Gears 5 uses Unreal 4)
  • Actually it is unless you think of AAA games of only fps games or such in contrary to big budget / marketing costs or blockbusters (e.g. Hearthstone is made with Unity).
    It is also technically ready just not preferred in most situations compared to in-house made engines or Unreal for games with top graphics budget. Unity games with impressive content or graphics: Pillars of Eternity I/II, Cities: Skylines, Kerbal Space Program, Homeworld: Deserts of Kharak etc.
  • You say you read the article but your comments seem to imply otherwise. You're not actually arguing against it, you're just saying "in my view" it's dead.
  • From the article itself: "But are true Universal Windows Apps dead? No. In fact, they play a critical part in Microsoft's future computing experiences." Key word: Future. It is *possible* that these scenarios will bring "true" UWP apps back, but for now, they're dangerously close to vaporware, and even the products that technically exist, such as HoloLens, are so niche as to be within rounding error of nonexistence. If every one of these "true" UWP devices disappeared tomorrow, it wouldn't show up as a blip on MS' balance sheets. And let's be clear here: I don't see *any* new device running Windows *ever* becoming part of mainstream computing again. So yeah, it's dead.
  • The problem with that is that Unity can compile to both UWP and win32 without changes to code. Unity will generate the appropriate calls depending on what compiler you're targeting. Also Unreal is touted to support UWP as well now, they have a new SDK in preview that you can download that targets UWP and win32 just like how it does on Unity.
  • Code in Unity is written in C# When you compile there is an option to select the target platform. UWP, Win32 etc...
  • Xbox games are not Windows UWP. Xamarin has mostly dropped the support for UWP as of now. You can still use UWP with Xamarin, but supporting UWP is completely removed from the user interface when creating the new project. The fact that Microsoft's Xamarin doesn't want to bother to have just a checkbox to target UWP is more than telling.
  • You haven't seen an Xbox? And HoloLens 2 and Surface Hub are too far along in development switch over to Win32. I think IoT is out now, but maybe not. Of course Xbox is the most main stream of these devices, but that will change a lot in the next 6 month as HoloLens and Surface Hub are released.
  • Meh, XBox is probably only 60 million or so users. Hololens and Surface Hub will take years to equal XBox (if ever). Combined the three are no more than a modest market relative to things like Win32.
  • Between xbox and PC, Microsoft platforms make up the majority of sit down gaming. UWP bridges the two, unlike win32. This is like saying iOS is irrelevant, when every web coder that exists, codes for safari and chrome. Not to mention that coding once for the two gaming platforms saves major developer dollars. There are some issues with that, like the need to get games in other PC gaming stores, but for example unity is crossplatform. It's easier than porting to say playstation. Less total effort for more total reach.
  • The death knell for any Microsoft project/service/platform is Daniel Rubino declaring it's not dead.
  • IOT, Xbox, Hub, Hololens don't exist yet?
    Is that why 95% of Fortune 500 are working with MS? Well... those new-gen AI cashiers or Kiosks in Japan are Win10.IOT.
    The automatic canteen system in Japanese enterprises are powered by Win1.IOT.
    Starbucks's coffee machine and kiosk are Win10.IOT.
    Walmart, Walgreen, Samsung, Singapore's smart city project are Win10.IOT.
    Then there's car, water / electricity meter, surveillance, drone, etc.
  • Xbox has been around for years, HoloLens 2 is shipping, Surface Hub is behind schedule I think but is still being worked on, and IoT id don't know the status. But yes, customers have them (some are just not available for sale yet).
  • Surface Hub 2 has already been released. I got a quote for two of them last week for my client. They told me they have a 3 week delivery lead time.
  • I said this on Reddit. Apple has made the smartphone a standard for improving hardware and software. Every year they have a keynote at WWDC and talk about all the things they have improved. They go through category after category: here’s the update to iMessage, here’s the update to Photos, here’s the updates to Siri, and here’s dark mode. I think the last time I remember Microsoft talking about improving the Photos app was two years ago. Apple has a coordinated plan. It really makes developers want to write code for the platform. Meanwhile, both Microsoft and Google seem to present whatever they have been randomly working on. And sometimes poorly, like migrating everything over to Settings and it still not being completed after four years or putting together dark mode and leaving several UI elements out of the theme. So what I'm saying Microsoft should do is have teams working on each app and present it altogether at the Surface Event in October or Build 2020. They need to make a phone. I remember when Windows 8 had just came out and Windows Phone was around. I thought to myself "Microsoft is getting to the same point as Apple, they are just sort of funneling in same direction." When UWP became an coordinated effort between desktops and phones with Windows 10 I felt like this is the right direction.
  • Well said!! Your points are exactly right. Whatever else you may think of Apple, they are FAR more disciplined in developing their various OS's than either MS or Google.
  • Good points, but I personally feel Microsoft is going the right route at this time. They need to focus on services that make developing for all devices, even Apple and Google. Azure is their future. And this platform is getting better and better at a rapid rate even if it is hard to see from a consumer view, but it is having a big impact across mobile, home and corporate. There is no room for another phone OS, but for services that enhance all devices there is no end to the growth possibilities.
  • Sadly, both iPhone and Android are horrible, and, worse, Microsoft's play on Android is anemic at best. As a consumer, I don't really care how you nuts-and-bolts an app. I just don't. I do care about whether it provides all the functionality I need without stupid amount of "well, you need to customize this or that" garbage. I want it to WORK. I can't get that anywhere now.
  • What apps can't you get working on Android or iOS? Your criticism is vague.
  • How is it anemic, if anything it's made me integrate tighter into Microsoft. The apps I use are: ToDo - great for shared family lists and personal reminders
    Office 365 - Word, Excel, PP, One Drive
    One Drive - I have all of my personal files, and phone photos backed onto it
    Authenticator - I use for google and microsoft
    Skype for Buisness - For Work
    Sift Key - I prefer the simple swift key logic From a singular point of view the most apps I have from one developer is Microsoft on my android phone Also one more thing (you microsoft hater), android has a far superior file system compared to windows phone. The ability to map network drives, one drive, google drive you can do in a singular app. Just like in windows explorer So stop hating so much and get over yourself
  • And that's not even getting into the app nuts-and-bolts, Project Rome SDK, that allows any Android developer build "Continue on Windows" and "Continue on Android" experiences.
  • I want them to make a better consumer product and I think it will be easier to make a phone OS in the future. The internet is becoming more and more designed around mobile. Azure is good for businesses, but they could focus on their consumer efforts too. Wouldn't be good for Micrsooft to come back to mobile just to encourage the PWA platform? Even though that is what is being adopted anyway?
  • At this point, no. They are moving their entire corporate customer base to Azure, one service at a time. That is far too lucrative to expend resources in the phone space. When OneCore is ready for prime time, I can see other OEMs making Windows 10 phones again, perhaps.
  • The reason I bring it up is they seem to be for open sourcing and having services platform agnostic. I think the best thing the OEMs could do is start making tablets and phones. If they could, maybe one of them should buy a phone company. It would nice if they came out with phones or tablets with Android and/or Windows Lite.
  • Samsung might give it a crack. They have a long history of trying to get out from under google, and a fair relationship with MS.
  • Azure is good for everything. It is hard to see from a consumer view but it is everywhere right now, helping make all kinds of projects better on the consumer and business side.
  • The platform agnostic pitch is really getting stream as well, people just don't see it since it's mostly behind the scenes. Case in point .Net core is the most loved and wanted framework on stackoverflow 2019, React Native (React Native desktop) get's native access to Windows APIs via UWP without having to use Electron (this is a good thing, those apps in electron are resource hogs). Azure is used everywhere and is language and platform agnostic, you can use any language and any OS (Linux, Windows, etc...). Their dev tools are the most popular today VSCode, Typescript. Consumers may not see this, but slowly they're getting attention for developers that they didn't have in the past.
  • I think they will come back to a phone when Windows Core is done because they want to use a true cross platform OS so they can add features and updates in sync.
    Having said that they shouldn't make a phone unless they could somehow have a container like or converter for Android apps because cause Xamarin, while an option, is not the answer for building a good app ecosystem.
  • Why buy a first generation Microsoft phone to run Android apps? Android is mature, fully featured, and has a decade of work behind it. Whatever Microsoft launches in a couple years will be at the ground floor. It is hopeless. The only thing they can do now is catch the next wave.
  • Actually, the biggest problem Microsoft had with phones and getting developers to make apps for them was not having an adequate OS and APIs designed for phones. Well, they are finally getting there with both the OS and API's. They we next work on getting an apps library for HoloLens 2, Surface Hub, Xbox 2, and other devices. The next step will be smaller mobile devices, and then finally phones.
  • Developers not making apps is a symptom of poor sales. You are putting the cart before the horse. Platforms never launch with app ecosystems. Android didn't have apps either when it launched or even when Windows Phone launched. People actually liked and bought Android phones. Windows, not so much.
  • Android phones are boring slabs, so are iPhones and Windows phones resurrects wouldn't be any different.
    They are all boring slabs.
    If Microsoft wants to go in on mobile computing, phones are not the answer.
    Whatever comes next might be and having a solid platform framework for that in place ahead of time is probably a big advantage. Will it pay off? Who knows.
    One thing we do know, however, and that is that smartphones will only comtinute to be boring slabs.
  • Microsoft already missed "whatever comes next", and so did Apple for that matter. It's ambient computing, and Amazon owns that space, with Google coming in second.
  • No, their biggest problem was whenever developers were starting to get comfortable with the mobile OS Microsoft up and scrapped it and released a new one which then developers needed to modify their programs to be compatible with.
  • Maybe you're here because you're using a first generation Android phone?
  • Nope. I have a One Plus 6T. Runs all Android apps without issue. What advantage could a Windows phone that ran a limited subset of Android apps bring me?
  • Liar. You're a 5 year old low end Android device user, and you can't hang with the real Android fans over there on AC because you're embarrassed. What's worse is you come here to Troll to make yourself feel better about your life, but it never could work. Do you even vist Android Central?
    Nope. SMDH get a life
  • Still can't make a valid point, just say "Live Tiles are the best". Only thing you have is personal attacks.
  • "The only thing they can do now is catch the next wave." Phones are not waves. They should make a phone.
  • Microsoft will never catch-up with phones. Way too late and there is no discernible way for them to innovate sufficiently enough to sway the market.
  • There's not much to catch up on. They simply start where everyone else is because other companies are figuring this out too. With Windows Lite, they are still going to have to build outward. The same way they would have to with Windows 10. The same way Google and Apple do. It's not that they're that much behind as much as "thinking" about how far it is. Microsoft in some ways has made gains. They experimented with Windows 8 and they have a platform with UWP. They just need to give it more attention.
  • "The only thing they can do now is catch the next wave" That I agree with. Smartphones are post-peak. As an investment area, it's over. Dual screens, AI, AR, VR, IoT, whatever else, if there is a new 'wave', it's somewhere in there.