Microsoft’s Universal Windows Platform Challenge

Last week, Epic Games co-founder Tim Sweeney wrote an opinion piece for the Guardian complaining about the Microsoft Windows Store and Universal Windows Platform (UWP). He said the company is trying to create a walled garden that’s bad for everyone except Microsoft. Executives at Microsoft were quick to dispute Sweeney’s argument. In short order, much of the tech press also sounded off. This week, Sweeney wrote a follow-up article in Venture Beat, striking a more conciliatory tone but still pushing for changes from Microsoft. All sides have valid points, but the discussion itself is fascinating as it succinctly reflects the challenges facing Microsoft. Broadly speaking, Microsoft is attempting to transition Windows from the complex and powerful operating system that dominated the world during the boom days of the PC industry to a more modern, secure, and restricted OS in a world where PCs are an important but minority device stakeholder. What’s perhaps the most interesting piece of all of this is how the company’s UWP strategy will impact its big bet on the future: HoloLens.

Understanding UWP

UWP is an extension of work Microsoft began with Windows 8 to enable developers to create modern, universal Windows apps. The idea was that, by unifying the code base between the traditional version of Windows used for desktops and notebooks with that of Windows Phone, the company could incentivize developers to create apps that would, with minimal work, run across all form factors. At the time, one could argue, Microsoft’s primary driver here was an attempt to close the widening app gap Windows Phone faced versus Apple’s iOS and Google’s Android. When it launched Windows 10, Microsoft further evolved the model, bringing forward a fully unified OS platform so that every device, from phones to tablets to PCs to even the Xbox will run the same apps, with developers able to easily tailor for the screen size and capabilities of the device in question.

In addition to enabling cross-device capabilities, UWP also brings forth a host of modern OS functions. The most notable being a new, more restrictive set of rules on what apps can and can’t do. Chief among these is security features that effectively sandbox UWP apps, limiting the impact they can have on the underlying OS as well as adjacent apps. In effect, UWP apps act more like apps running on modern mobile operating systems such as iOS. The difference here, however, is Windows 10 also still supports the traditional Win32 system, the underlying platform for Windows going back to Windows NT. This platform supports the legacy apps most of us use daily and is part of what made the PC so powerful, as it allowed programs deep access to the OS and adjacent apps. It also lets users download and install apps from anywhere. Alas, it’s also what makes the PC such a fertile ground for malware along with security and long-term performance issues.

By supporting both UWP and Win32 on Windows 10, Microsoft is trying to straddle two worlds. The traditional PC ecosystem with a long, rich tradition of powerful, backward compatible apps that can, at their worst, bring the OS to its knees, versus the modern mobile world where apps have a more limited impact on the underlying OS, creating a more secure, more stable experience. At present, the success of UWP is hard to gauge. I fully expect to hear more about the platform’s progress at Microsoft’s Build conference later this month.

Sweeney’s Diatribe

In his original piece, Sweeney calls UWP “a closed platform within a platform” that he posits is bad for consumers, developers, and publishers because Microsoft is launching some features exclusively in UWP it won’t enable for traditional Win32 applications. Further, he suggests the company’s long-term goal is to get everyone to develop only UWP apps they sell exclusively through the Microsoft-controlled (and monetized) Windows Store. He argues this is harmful to the industry in general and to developers who sell through their Web sites and that it effectively ends the open PC platform. The company rebutted this accusation, with Microsoft’s Kevin Gallo noting “The Universal Windows Platform is a fully open ecosystem, available to every developer that can be supported by any store.”

Sweeney acknowledges Microsoft’s points in his follow-up piece but suggests UWP is still less open than Win32 because developers on the new platform must become a Microsoft Registered Developer and must submit their apps to Microsoft for approval. If Microsoft accepts the app, it digitally signs it and returns it to the developer, who can then distribute it. He questions whether this is a truly open system, and suggests he would like to see a CEO-level commitment from Microsoft to keeping the PC and UWP open.

Frankly, I’m not sure he can expect such a commitment. As Microsoft moves Windows 10 toward the future, you have to imagine company will endeavor to exert ever-greater control over the applications that run on its platform and the resulting experience for users. And, while it is equally hard to imagine a Windows platform that doesn’t support legacy Win32 apps, it also seems counter-intuitive to assume already ancient legacy apps will run as they do today on all future versions of Windows. Many of these apps, while powerful, contribute to a poor long-term experience for the average user. Sure, the serious PC gamers Sweeney serves want their games to have deep access to the OS to drive better gaming experiences. But the vast majority of consumer and commercial PC users would likely choose the other option — a PC experience that feels more like the one they have on their mobile phones. Fewer crashes, fewer issues when installing or deleting apps, and an operating system that remains more stable over the lifetime of the product.

While Microsoft has addressed Sweeney’s earlier comments, it’s unlikely we’ll see the company specifically address his call for a long-term commitment to his version of an open UWP. But a closer look at another set of developer documentation from the company offers a glimpse of its future.

The Future with HoloLens

As noted previously, Universal Apps and the evolved UWP started life in part as an incentive to get developers creating apps to run on Windows phone, which had fallen far behind iOS and Android in mobile. Bluntly, Microsoft missed the broader transition to mobile. The company has no intention of missing the next big evolution of computing: Augmented Reality. This is evident in its announcements and pending developer launch of its impressive HoloLens product. (Actually, Microsoft calls HoloLens a Mixed Reality product, but that’s a column for another day.) A review of Microsoft’s developer information for HoloLens is telling. The very first sentence on the development overview page states: “Developing holographic apps uses the Universal Windows Platform. All holographic apps are Universal Windows Apps, and all Universal Windows apps can be made to run on Microsoft HoloLens.”

Microsoft obviously sees HoloLens as crucial future technology for the company. It clearly states holographic apps developed for HoloLens will be UWP. That’s not to say the device won’t run Win32 apps but it seems clear the company expects developers to create all new, holographic apps on the UWP. I’ll let you draw your own conclusions about Microsoft’s stance on the future of app development for Windows from there.

So does that mean Microsoft will abandon support for Win32 on traditional PCs? Certainly not. But it’s hard to imagine the company maintaining the status quo indefinitely. I suspect at some point we’ll see such support evolve into something different from what it is today. While this will undoubtedly cause a fair amount of consternation among a large subset of existing users, it seems an inevitable conclusion to Microsoft’s current course.

Published by

Tom Mainelli

Tom Mainelli has covered the technology industry since 1995. He manages IDC's Devices and Displays group, which covers a broad range of hardware categories including PCs, tablets, smartphones, thin clients, displays, and wearables. He works closely with tech companies, industry contacts, and other analysts to provide in-depth insight and analysis on the always-evolving market of endpoint devices and their related services. In addition to overseeing the collection of historical shipment data and the forecasting of shipment trends in cooperation with IDC's Tracker organization, he also heads up numerous primary research initiatives at IDC. Chief among them is the fielding and analysis of IDC's influential, multi-country Consumer and Commercial PC, Tablet, and Smartphone Buyer Surveys. Mainelli is also driving new research at IDC around the technologies of augmented and virtual reality.

629 thoughts on “Microsoft’s Universal Windows Platform Challenge”

  1. Interesting, thank you.

    I think there are a lot of intermingled issues:

    1- Walled garden vs fenced garden. Only iOS is truly walled-in, Android, MacOS, and Windows allow apps (including UWP apps) to be sideloaded or downloaded from other stores aside from their respective official appstores. As long as that ability persists (and there are no more reasons to believe it is going away than there are to believe iOS will allow sideloading anytime soon), “fenced” is probably the right approach. Stay safe with the blue pill, or take the red pill. I also don’t think the 30% cut on app sales is much of an issue, though it might be on media sales.

    2- UWP vs Win32. It’s clear MS has every reason to push UWP: it gets them apps for their Mobile platform, but also Mobile apps for their desktops (Clash of Clans on a PC ! w/o resorting to AMI duOS’s Android VM and handing over the sales cut -10s or 100s of millions ?- to Google !), and probably better security and easier maintenance (modern/streamlined APIs, inherently cross-platform…). Also, probably more developers, developers, developers since the concepts and paradigms are much closer to iOS/Android (plus the iOS porting tool and Xamarin).
    I think the HoloLens decision is only the more visible part of MS’s decision to go UWP-first, with Win32 kept for legacy reasons but intentionally no longer moving forward, or with a significant lag. I’m sure more features will follow the same UWP-first pattern, I’ve actually been told that’s already the case but I can’t remember specifics. Resolution-independence maybe ?

    3- Consumer vs Corporate. Consumers will probably appreciate the automatic, centralized updates that come with an appstore. Corp, not so much, since they need to validate updates and patches.

    1. I’d love to write for UWP. But there is no way on god’s green earth i’d get the go ahead to port. To be quite clear, nobody wants to pay the Microsoft Tax, particularly if it takes redevelopment dollars to do so. At the same time, the software we sell is only really applicable to PC’s. Its cute you can write it to be able to use on other devices, but there isn’t a single use case where a customer would WANT to use it on other devices. Its that way for the majority of existing desktop software. You simply do not gain anything, while hosing 20 percent of your sales to Microsoft.

      So what we’re basically seeing is that companies will write new apps for UWP that can use it and actually have a reason to work cross-device, and simply ignore the platform for existing desktop applications(of which there are a *TON* of companies that Microsoft would like to convert to UWP that would rather die first). So UWP will be the same as all versions of their apps… uptake will continue to be very, very slow.

      Btw… who uses the Microsoft Store? The apps I do see in there are usually very poorly built fly-by-night apps. If I really need something, I find a desktop version from a reputable company.

      1. You’re very right about the MS appstore being a cesspool. Installed one twitter client from there, got hacked. I understand MS want to seem to have a lot of apps, but when over half of those “apps” are useless at best, malware quite often, and most interesting/good software is missing, apps totals mean nothing.

        UWP apps don’t have to carry the MS tax, you can distribute them on your site.

        And I’m sure lots of people would play Clash on their PC if they could ^^ Not true for any app though, indeed.

        1. Thanks for the clarification on install options. Now someone needs to make a .exe that runs a .appx just so we don’t have to field complaints about wth an appx is(and if you know the business software world.. this would SOOO happen) :P.

          Also, the other thing to keep in mind is that UWP and appx are windows 10 exclusive, no? When it comes to business software, not having a version for Windows 7 and 8 would be cutting off *at least* half your sales. That would pretty much be worse than the Microsoft Tax you’re now avoiding.

      2. You’re definitely right about companies not using UWP for existing WPF or Win32 software. Microsoft’s strategy is to get as many people writing UWP apps as possible while also pushing the security and mobile benefits so that at some point in the future the UWP talent pool will be large enough that companies will start to look at it seriously.

        I’m hoping that they’re also going to make writing cross platform (OS) possible on UWP via Xamarin, so companies will see a bigger monetary advantage to its fairly unified code model. Of course that won’t mean a single exe.

        But that’s gonna take a long while.

        1. A very long while. I just don’t really see them getting a core base of actual professional UWP Developers until there is actual money in it. They’ll just get niche vendors like HoloLens-first devs(which is why they’re pushing it, and even then I don’t think VR or AR itself is going to take off as big as the industry is thinking it will).

          At least with Desktop Applications, you can still sell those for hundreds(or thousands, depending on the program) of dollars, and people who buy don’t tend to give a rats ass what it was written in or really how it looks as long as it works and has the options they need in a fairly reasonable spot. So even with a smaller user base(in comparison to mobile or even say Office), you’re still able to fund development fairly easy with a product that sells.

  2. HoloLens?
    I believe it will be a niche market. I know some will need that kind of reality but will the rest of us I doubt. I can’t see myself having to putting on a goggle to play with myself.

  3. “Sweeney acknowledges Microsoft’s points in his follow-up piece but suggests UWP is still less open than Win32 because developers on the new platform must become a Microsoft Registered Developer and must submit their apps to Microsoft for approval.”

    Yet again, Sweeney is just factually incorrect. Yes, you must sign up for a developer account, but that’s free and Microsoft simply isn’t going to deny a developer license to anyone. Additionally, other tool vendors could add support for UWP so even an MS developer account would no longer be necessary.

    More importantly, you DON’T have to submit apps to Microsoft approval. Apps can be sideloaded by default on Windows 10, and any other app store can install them (as the MS spokesperson clarified), leaving Microsoft out of the loop entirely. Steam, for example, could implement APPX/UWP support in a matter of weeks if the platform started to grow.

    As for Hololense support for Win32, this is unreasonable. Hololense is a new hardware platform and, like on XBox, implementing Win32 would be highly inefficient and slow (and if Hololense is ARM bases, effectively impossible).

Leave a Reply

Your email address will not be published. Required fields are marked *