Daniel Eran Dilger
Random header image... Refresh for more!

When will Apple release an App Store for Windows?

Daniel Eran Dilger

Apple has barely begun setting up its Mac App Store, but I’m predicting the company will want to bring the same thing to Windows PCs within a year or two. Here’s why.
Rise of the Apps Stores

Some people didn’t see the point of my advocating for a Finder Store and thought I was reading too much into the importance of iPod games until the iPhone came out. At that point, I suddenly looked like the conservative one, writing that Apple was unlikely to ever fully “open up” an App Store for iPhone that let anyone do anything. I said that based on Apple’s history of working with developers who, as should be expected, followed their own interests rather than doing what was best for Apple.

The consensus at the time was that Apple needed to throw open the gates and let everyone have a crack at developing iPhone software without any restrictions, and doing anything less than this would be “closed,” another word for Evil amongst populist pundits.

Once Apple released what it would later describe as its “curated” approach to software on the iPhone, there was a lot of howling about how developers were being oppressed by the Evil Apple, and how this might be rectified by the rebel Jailbreak community, who were predicted to find great success by running around Apple’s restrictions and giving the people what they wanted, or at least what the pundits decided the people wanted.

Six Reasons Why Apple May Never Open the iPhone

Careful what you wish for

When Google announced it would be delivering for Android pretty much exactly what the populist pundits thought people needed, it was hailed as the “Windows that would trounce the Mac,” as developers would presumably flock to Google’s unrestricted platform and leave Apple stuck with an App Store in a life support coma of developer indifference.

A year or two later, it’s now obvious that Apple’s approach was far more successful. While there’s still a lot of junk in both the App Store and Android Market, only Apple’s store has a been generating real wealth for developers. Android’s remains a hobbyist collection of get rich quick junk, dominated by wallpapers and ringtones purporting to be apps.

Search for the few legitimate apps that have been brought to Android, and you’ll be inundated with scads of garbage pretending to be what you’re looking for, with no “curation” from Google holding back the tides of trademark infringement junk ware, and only the most bare attempts to retroactively put out malware fires after customers report being incinerated. It’s a Chinese import junk store compared to Apple’s Bloomingdales, even if openness ideologues refuse to admit this reality.

Inside Google’s Android and Apple’s iPhone OS as software markets

Software Sells Systems

How Microsoft Got its Office Monopoly

For Apple, the point of the iPod Games and later the iPhone App Store was to sell hardware, turning third party software into yet another type of content consumers could purchase through iTunes. The Mac App Store brings the same thing to the Mac platform.

So why on earth would Apple want to create a Windows PC app store, or for that matter, create a Server App Store for Mac OS X in conjunction with licensing its server OS on third party hardware (as I recently advocated)? Neither would sell Apple hardware, and if anything, only give users a reason to stay put on rival hardware rather than switching.

I’m betting Apple will investigate a Windows version of its desktop App Store for the same reason it brought Claris AppleWorks to Windows (very successfully), brought FileMaker to Windows (again, very successfully) and brought QuickTime and iTunes to Windows (once again, very successfully). In the first two cases, the pre-Jobsian Old Apple created software for Windows in order to make money. The Windows market was simply bigger.

But more than just selling more copies of its software, Apple’s Claris software subsidiary was helping to legitimize its apps in general. It was, at the time, competing against Microsoft’s Works on both the Mac and Windows. By having a Mac-only version of AppleWorks, Claris was not only missing out on sales, but also restricting itself to fight Works with one arm tied behind its back in a world where most people and companies worked on both Macs and Windows.

Office Wars 1 – Claris and the Origins of Apple iWork
Office Wars 3 – How Microsoft Got Its Office Monopoly

Jobs: software sells Macs

The result back then was that Claris outsold Microsoft Works in both the Mac and Windows markets, and continued to do very well until the rest of Apple collapsed around it, requiring invasive surgery by Steve Jobs to cut out the necrotic bits of the company so that the rest of it could survive. By the late 90s, all the cross-platform Claris software was so old that it made more sense to sell it off or discard it rather than trying to keep it going.

Jobs then took Apple’s software goals in a new direction: exclusively selling Mac software. The company purchased Macromedia’s Key Grip project and canceled the Windows port of Final Cut, making it Mac only. It later bought Logic and discontinued the Windows version. It then released iLife and iWork apps as Mac-only titles. All of those actions were desperately needed to breathe some legitimacy into the Mac OS X platform and make it positively differentiated from the de facto Windows PC.

Apple’s efforts to focus software at selling Macs ran from the core OS through to its top level applications. But Apple also continued to maintain Windows ports of key Apple software for other reasons similar to those of Claris a decade earlier. QuickTime needed to be on Windows for the same media platform-legitimizing reasons: were it Mac-only, it wouldn’t matter.

Software sells iPods

Apple continued this pattern by releasing iTunes for Windows shortly after the release of iPod. While Apple had been selling plenty of iPods to Windows users in a bundle alongside third party music software, things really took off when Apple added an iTunes version for Windows, which brought the same desktop integration Mac users enjoyed.

One might argue that this prevented iPod users from buying Macs, but the real goal was to sell iPods to Windows users who were not interested in buying Macs. After several years of stratospheric iPod sales, Apple added the iPhone and most recently the iPad to its offerings, and continues to sell a large number of iOS devices to Windows users. We know this because Apple sells far more iPods and iOS devices than it sells Macs.

Software sells standards

Apple also released another Windows app, for a third reason. Like QuickTime, Apple began offering Safari for Windows, not to profit directly (both are free) nor to sell hardware, but to propagate its web browser engine as broadly as possible. Apple did the same with Bonjour, which is essentially the automated networking features of AppleTalk translated into Internet Protocol standards.

Internally, iTunes for Windows also served this purpose by bringing AAC support and later H.264 playback (along with FairPlay DRM) to a vast audience, establishing both as preferred standards to the older (significantly patent encumbered) MP3, less sophisticated H.263 (aka DiVX), and Microsoft’s rival Windows Media Audio and Video (aka VC-1) codecs.

Apple had limited resources to push Safari adoption on Windows, but when Google released Chrome and began paying partners to use it (to fight back against the dominance of Internet Explorer, and Microsoft’s control over the web’s ad revenue and the potential for emerging web standards such as HTML5), Apple got exactly what it wanted: widespread adoption of Safari’s WebKit. This paved the way for WebKit to become the most popular browser engine on mobile devices, although Apple’s hardware essentially made that happen on its own.

And so it is that while software sells hardware, it also sells platforms. Apple is a hardware maker, but it really is a platform vendor. Once Apple achieves the goal of selling Mac apps through a storefront, its next step is to bring that store model to Windows. Why?

I’m So Glad You Asked

Ostensibly, at least in the minds of the people who fear Apple, this would be to profit from that 30% cut Apple charges developers to retail their apps. Sure, Apple could make more money by selling more apps to a wider audience. But there’s a better reason for Apple to bring its App Store to the Windows PC.

The platform Apple wants to sell is Cocoa. It tried, initially, to sell NeXTSTEP’s development frameworks as a piece of middleware that could run on any operating system. NeXT had already been doing this for several years, enabling developers to create YellowBox apps that could run on NeXTSTEP, Solaris, and Windows NT. By 1997 however, Apple was not seen fit to deliver a cross platform development system. Sun abandoned Apple to devote its efforts to Java, and Apple gave up on its hopes to layer YellowBox on the Classic Mac OS. It simply wasn’t up to the task.

Instead, Apple planned to replace the old Mac APIs with YellowBox running on the modern foundation of NeXTSTEP. That plan was rejected by Adobe, Microsoft and other Mac developers who didn’t want to rewrite their extensive application suites just to serve the dwindling market Apple could claim.

Why OS X is on the iPhone, but not the PC

The new Cocoa

Today however, Apple has the power to stipulate what APIs its developers can use. It does this (on iOS devices) by only listing software that abides by its rules. By controlling its platform, Apple puts its developers in the position of assistants, rather than masters under which it must grovel for their continued support. The Mac App Store will bring the same level of “no, use Cocoa” to the desktop. Why wouldn’t Apple want to bring this to Windows PCs?

The reason Apple doesn’t sell its Pro Apps or iLife or iWorks on Windows has always been to add exclusive value to the Mac. At this point however, Macs are greatly outpacing the growth of the generic PC. Microsoft is floundering in its efforts to push its developers from old Win32 APIs to its vision of modern computing, which keeps evolving.

Over the next year or two, Apple will have an opportunity to bring Cocoa development to the Windows desktop via an App Store that sells titles to a broad audience, without much reason to fear that its own Mac platform will dry up and blow away.

Cuckoo for Cocoa

I speculated that Apple might have this in mind when it brought Safari to Windows, but I was wrong. Apple wasn’t porting Cocoa to Windows to deliver Safari; it was using native Windows frameworks. But there’s no reason why it couldn’t do this however. The much smaller NeXT managed to bring its frameworks to Windows with far fewer resources. However, once Apple turned NeXT into Mac OS X, the opportunity to bring Cocoa apps to the wider market of Windows evaporated as the company mingled its old and new frameworks together in a way that was not intended to deliver cross platform compatibility.

Completing the work to host Cocoa apps on Windows would require significant efforts. What’s the payoff? For starters, Apple could then port its iWork apps to Windows, rapidly broadening its sales prospects. Thousands of Mac developers could also bring their apps to Windows as well, offering Apple at least a sustaining cut to maintain Cocoa cross platform.

Additionally, rather than building apps in some kluge like Java or Adobe Flex (think Flash for apps) or Silverlight, Apple can offer developers its own API that is not only managed by one vendor (unlike the mess that is Java) but also offers an App Store to promote and sell their apps to a hungry audience (unlike Flex or AIR or Silverlight or Java). This would incentivize developers to use Cocoa in the Mac and Windows App Store rather than building lowest common denominator apps that reach the Mac through ugliness like Flex (like the new EyeFi app).

Because Apple now controls the APIs developers will use to create Mac App Store titles, it can now manage the work required to make Cocoa platform independent. This was much harder when most Mac software was written in Carbon or some mix of the two, or when developers were able to use deprecated or private APIs that Apple didn’t want them to use.

One might suggest that there is a risk for Apple that consumers might no longer see an advantage to the Mac if all their software were also available for Windows. However, the main allure of the Mac is its hardware and software integration and innovation and its retail store support, not its selection of a few exclusive, Mac-like apps.

Cocoa and the Death of Yellow Box and Rhapsody
Cuckoo for Cocoa: Is Safari on Windows the next iTunes?

iOS, back to the Mac

Apple already leverages enormous clout in the mobile realm with iOS. It is the only mobile platform where developers are making significant money. Making it easy for those same developers to target Macs is the first step, but bringing the same development tools to Windows is definitely the next one after that. There are lots of iOS developers who would jump at the opportunity to create desktop apps to accompany their iOS titles, and having an App Store that only allows them to sell these to Mac users isn’t enough.

Again, bringing Cocoa apps to Windows not only broadens the market for Cocoa apps dramatically, but also legitimizes those Mac titles. Were Apple able to sell iWorks to PC users, it could make enough more money to devote significant new resources into iWorks. And there’s nothing stopping Apple from making the Mac version of iWorks more attractive by hooking into features the Mac platform has that Windows lacks, such as support for Services, Automator Actions, AppleScript, and so on. But bringing iWork to Windows would also make it easier for Apple to sell its apps to Mac users, as cross platform companies could actually chose iWork for all their users, were it appropriately competitive and cross platform.

A Cocoa App Store for Windows would also make it easier for Apple to deliver subsequent versions of iTunes and Safari for Windows; currently the company has to build two separate versions. Apple certainly isn’t the only company targeting both Windows and Macs. While Microsoft and Adobe have developed their own internal systems for building Mac and Windows apps, there are a lot of developers who simply lack the resources to devote all their efforts at creating separate builds of each. A Cocoa Apps Store for both Mac and Windows would solve a number of problems, at little cost, and in a self sustaining manner.

But what do you think?

  • gctwnl

    @Brian Willoughby:
    Very well said.

    Much of what Apple has been able to create hinges on the qualities/power/dynamism of Obj-C. Not many people are aware of this core strength.

  • NormM

    The part of your argument that resonates with me is Apple’s desire to push iOS on mobile devices, which sometimes requires helper apps that run on a desktop or server. I think this can be handled, though, much as multitasking in iOS: identify the common scenarios and provide specific help. I don’t think you need a Windows app store for this, though, since developers make their money on the iOS app and give the helper programs away for free.

  • Saturnine Chaim

    Interesting article. But one of the biggest reasons the iPhone app store succeeded is that it’s on every iPhone. The Mac app store will be on every Mac. It will be the *only* storefront that comes installed on every Mac.

    Windows users comprise a much bigger market than Mac users. But a supposed Windows Apple App Store *won’t* be on every Windows machine. Windows users will have to want it enough to go get it. How many are going to do that? Some fraction of the total user base. Which makes the prospects for proliferating Cocoa a lot more dubious. Less return on investment, you might say.

    Also, are the Windows users who are motivated enough to seek out an Apple-branded storefront the same people who are ready to switch to Mac, anyway?

    One more point …

    Judging from the Mac App Store unveiling, it looks like it’s going to be a separate app, not a new section in iTunes.

    If that holds true for our theoretical Windows version, then Apple would be really failing to take advantage of their main point of leverage among PC users, which is iPod/iPhone/iTunes.

    If Apple really wanted to get an app store onto Windows, at the very least they ought to build it onto iTunes – strategically speaking, anyway. And it doesn’t look like they’re doing that.

  • http://www.sounds.wa.com/ Brian Willoughby

    You raise a very good point. However, by the same logic, the Mac App Store will not be nearly as successful as the iPhone Store. The reason is that there are a lot of old PowerPC Macs and even old Intel Macs out there which will not automatically receive the Mac App Store link just because it ships on new Macs. The situation will be a lot better than on Windows, because more Mac users are aware of new additions to OSX and will probably take advantage of the Mac App Store. But my point is that where iPhone/iPod/iPad have 100% coverage in the App Store, and Windows will have a possibly very small percentage, Mac OSX will end up somewhere in the middle, and will certainly never reach 100%. New Mac sales do not represent 100% of the Mac market.

  • Saturnine Chaim

    @Brian Willoughby
    Yes, at least we can say the iPhone app store has an advantage over the Mac app store in that respect. I don’t know much about OS X upgrade adoption rates but in principle what you describe seems right.

  • http://www.sounds.wa.com/ Brian Willoughby

    I suppose it’s clear that if the Mac App Store fails, or doesn’t succeed to an impressive degree, then Apple would not consider a Windows App Store at all. But, if the Mac App Store does sustain impressive profits, then Dan’s suggestion will eventually be considered.

  • http://karunaltea.org db

    I really appreciate your insights.
    The iPad is the Trojan Horse into the fortress of the windows world. It is used by a lot of people who own a PC with the various operating systems. They will get a better idea of what computing can be, just working and playing with the iPad. Everything taken care of. No need for manuals, seemless doing what is needed and possible.
    With that stock value will be really careful to risk anything.
    By now Apple does not need to be recognized as an “alternative.
    As organized as they are Apple folks have the masterplan already to port Mac OSX to the PC platform. I do not think they work on an App Store for Windows because they might have discovered that the iPad is bigger than they had anticipated themselfs. And the iPhone does the rest. Why bother with any integration into a struggling market ? They have by far the best hardware at the best prize point and the best integration. And do we not forget: They still have to work out some issues with Aperture, Itunes, iWork, Mobile Me etc. I can still get much better.
    Who knows what Apple and Microsoft agreed on anyway. They for sure are not such enemies as some people want us to believe.

  • http://lineoftheday.com schwabsauce

    In a way I think it’s interesting to think of Apple not as a product company, but as a company that makes markets. They’ve made a market for software, for digital art, for digital music, for iPod accessories and one for mobile apps. You could also argue that they created the market for smartphones, since there are now dozens of phones blatantly copying the iPhone, and also the market for tablet computers.

    Selling hardware has scaled pretty well for them; but these markets are typically much more scalable. Although they need a big staff to review apps submitted to their secure store, once approved an app can scale very well. Assuming that desktop apps will cost more than mobile apps, the sale of software could become an almost limitless revenue source.

    Dan’s point about introducing developers to Cocoa by making it a tempting cross-platform target is very compelling. It would also seem fair to assume that a larger cross-platform population of software would slowly but surely bring more customers to Apple hardware.

    The question of whether it’s possible to deliver a smooth & secure experience on Windows, and whether Microsoft will cooperate, is a good question, but I find it cynical to assume that it’s impossible. I feel that Cocoa is a huge framework, and when porting a huge framework one is bound to run into many complexities. But Dan seems to think that it is possible, so I’m very open to seeing this come to light.

  • http://www.sounds.wa.com/ Brian Willoughby

    Fine observations. I would like to add that Cocoa was already ported to Windows in the mid-nineties. Apple would need to do very little work to do if they wanted to bring NeXT’s technology up to date and ship Cocoa for Windows now. The biggest problem was (and is) that Cocoa on top of GDI has all of the shortcomings of each technology, yet few of the serious performance advantages. Cocoa on Windows made it significantly easier to write crappy Windows software, but it cannot make the graphics engine perform any better.

    By the way, Windows NT and some other Microsoft operating systems have or had a feature which allowed an entirely different screen API to be installed. I think this was called the Presentation Manager (or perhaps that was the name of a higher layer). In any case, the user could cycle between screen groups that could be entirely different modes – text, planar graphics, unified framebuffer, et cetera – to access programs written to different API. Despite the significant development effort that would have been required, and the necessary segregation of AppKit applications that would result, I long wished that someone at NeXT had ported Cocoa and the AppKit to its own Display PostScript based API on WinNT. That would have allowed all of the advantages of Cocoa without any of the drawbacks of GDI. Not that GDI was all bad, or that DPS did not have some gaming limitations, but it would have been great to see the full potential. In fact, Windows NT started out with the OS/2 PM, but that was swapped out for Win32 GDI due to popularity. There’s no reason why NT could not run OS/2, GDI, and Display PostScript – so long as each was only visible one at a time under user control to cycle through them. Since that time, OS/2 has died, and these features of NT may be gone, not to mention the fact that OSX migrated from Display PostScript (DPS) to Display PDF. Still, it would be nice if the concept were still possible. I don’t see Microsoft putting any effort into such a potential.