Open vs. Closed Systems: What the Future Holds

by Steve Wildstrom   |   November 28th, 2012

Open/Closed signs

Since the beginning of the personal computing era. there has been a struggle for dominance between open and closed systems. The early open CP/M computers gave way to the relatively closed Apple ][. The closed Mac was beaten by the open Microsoft/Intel PC. A few years ago, with the rise of mobile platforms, it looked like the closed model was achieving dominance. The closed BlackBerry and the rising iPhone were demolishing the open Windows Mobile. In traditional computers, the Mac was at long last gaining ground on Windows.

Today, things are a muddle and it is far from clear which model will dominate the next phase. Note that my usage of open and closed has nothing to do with open or closed source; an open system, in this classification means one where a software vendor offers code to a variety of hardware makers while closed system companies allow their software only on their own hardware.

I have long thought that the closed approach was best for mobile devices, especially phones. Small devices benefit greatly from having software that is highly optimized for the specific hardware it runs on. This produces systems with superior size/battery life/performance tradeoffs and, often, a better user experience. This was true for BlackBerry in its heyday and it is true for Apple’s iOS today. Windows Mobile, by contrast, suffered horribly from its attempts to accomodate a wide variety of devices–with or without touch screens, with or without physical keyboards and D-pads, and displays in a variety of sizes and aspect ratios.

(Of course, closed was no guarantee of success. Palm prospered for a while as a closed system, but faltered when it couldn’t compete with financially stronger rivals. Symbian, nominally open but, in its later years, effectively proprietary Nokia software, was also swamped by the iPhone tide.)

The success of Android, which is the clear global leader as a smartphone platform,  may be changing the equation. It’s not clear whether the considerable variation in Android hardware design has been a blessing or a curse, but recent iterations of the operating system, especially the current Jelly Bean, have made it easier for developers to optimize software for a variety of device types and sizes. It still falls short of Apple’s seamless integration and probably always will, but it has definitely improved.

Microsoft seems to be charting a third course. Windows 8, Windows RT, and Windows Phone 8 are all nominally open. But Microsoft is keeping a close rein on its Windows Phone OEMs, setting tight parameters for most key specifications. For example, at least for the time being, Windows Phone is supported only on Qualcomm applications processors. The approach leads to a more consistent user experience, though it also causes a certain sameness in the devices. (Windows 8 and RT are covered by the more traditional Windows hardware requirements.Microsoft is keeping a close rein on its Windows Phone OEMs, setting tight parameters for most key specifications.

It’s far too early to say how successful Windows Phone 8 will be. Microsoft is rich and patient and seems inclined to give the much praised but apparently so far little purchased platform time to find its footing. Meanwhile, Google may be moving toward greater control of the Android platform, at least the licensed part of it, both by its sponsorship of Nexus designs and its control of Motorola Mobility. Microsoft, of course, has also plunged into manufacturing with the release of the Surface tablet, though it has shown no indication that it will also do its own phone.

Regardless of the success of Windows Phone, I think the  Microsoft approach may be the right model. Because Android is open source in addition to being an open platform, there’s really nothing Google can do about a proliferation of phones, especially in emerging markets, that ignore its guidelines. But I expect it to enforce greater control over “official” Android products as it integrates Motorola and seeks to improve the Android user experience across the board.

 

 

Tags:

Steve Wildstrom

Steve Wildstrom is veteran technology reporter, writer, and analyst based in the Washington, D.C. area. He created and wrote BusinessWeek’s Technology & You column for 15 years. Since leaving BusinessWeek in the fall of 2009, he has written his own blog, Wildstrom on Tech and has contributed to corporate blogs, including those of Cisco and AMD and also consults for major technology companies.
  • raja99

    Hi Steve. I think this is an interesting piece, but I find your terminology confusing:

    “Note that my usage of open and closed has nothing to do with open or closed source; an open system, in this classification means one where a software vendor offers code to a variety of hardware makers while closed system companies allow their software only on their own hardware.”

    As you note, “Open” usually means open vs closed source. Perhaps you should have used the terms “single-vendor” versus “multi-vendor”? Calling Windows Phone 8 “open” even though nobody outside Microsoft can see (much less modify) the source is confusing. At least I find it so ;-)

    • steve_wildstrom

      Multi-vendor is OK, but still a little too restrictive. For example, there was a brief period when Mac OS was multi-vendor, but Apple controlled the licensing very strictly. By contrast, anyone today can build an Android phone or a Windows PC and can use an Android or Windows logo just by meeting certain system requirements. Open was the best term I could think of, but made it explicit that this was a very different concept than open-source.

  • def4

    There is nothing open about the Intel standard (x86) or the Microsoft standards (DOS, Win32, doc, xls, ppt).
    Just because Microsoft and Intel sell components instead of finished products doesn’t make them open.
    The actually open standard that enabled the PC industry to look like we’ve know it is BIOS.

    That’s why comparisons of the state and players of the post-PC era with their PC era counterparts look so convoluted and feel obviously wrong both at the rational and intuitive level.

    Android is not the new Windows, it’s the new BIOS.
    Google may be the Microsoft of the Internet but they will never be the Microsoft of mobile.
    Symbian was the Microsoft (DOS) of mobile, but as that was not apparent in the American market, you missed it.
    Samsung looks to be the Compaq of mobile for now.
    But they are not using anything resembling proprietary Windows.
    TouchWiz can most closely be compared to a Linux distribution like Ubuntu.

    I’ll stop here as this has become silly enough.

    • http://alexandersmith.co/ Benjamin Alexander

      You’re absolutely right.

    • steve_wildstrom

      I defined what I meant by “open”–software that anyone can license and run on their hardware. x86 is open in this sense too: You can’t just make your own x86 chip, but you can buy one from Intel (or its licensee, AMD), get a ton and a half of technical documentation, and build a system around it.

      Oddly, the original IBM BIOS was anything but open. But what were not yet called its APIs had to be published for it to be functional and Compaq did a clean room reverse engineering and then withstood a fierce IBM legal attack. However, I don’t understand the Symbian-BIOS comparison. Symbian was originally developed by Psion for use in PDAs, but its only major adopters in the phone space, outside of the Japan-only FOMA products, were Nokia and Sony Ericsson with two very different and incompatible flavors. In the end, only Nokia mattered, so it might as well have been proprietary.

      • def4

        Thanks for the story, I wasn’t familiar with how BIOS became an open standard.
        In any case, that doesn’t change the consequences.

        The lack of technical (as opposed to practical) equivalent of BIOS for mobile ARM SoCs makes it impossible to buy an iPhone or Windows Phone and hack Android onto it. That’s why the hardware variety ecosystem uses Android as its creativity escape hatch.

        I compared Symbian – the company – with Microsoft and – the product – with DOS because they were similarly dominant in the early age of the smartphone with a cheaper, more primitive and user hostile product compared to the breakthrough coming from Apple.

        While Nokia with S60 was clearly the most successful Symbian licensee, the ecosystem was quite vibrant in its heyday with UIQ used in a good number of Sony Ericsson and Samsung phones.
        Actually, one could easily argue that Samsung is following Nokia’s recipe of how to be the dominant vendor on the dominant platform.

  • Rich

    Microsoft’s patience with Windows Phone 8 could end up like Microsoft’s patience with Zune.

    • steve_wildstrom

      I’ll bet more like Microsoft’s patience with Xbox or Bing.

      • FalKirk

        Steve, do you think Microsoft’s pursuit of Bing is a good thing? Microsoft is losing billions every year to pursue a desktop search model that is rapidly fading into second tier status.

  • FalKirk

    “Regardless of the success of Windows Phone, I think the Microsoft approach may be the right model.”

    I think Windows Phone may have been the right model if had been first or even second to market. But Android has – in my opinion – preempted Microsoft’s model. There is simply no room in the mobile (phones and tablets) market for a second open business model.

    However, time – and Microsoft – may prove me wrong.

    • def4

      Exactly.
      Being first obviously matters a lot.

      Neither DOS nor Windows would have gained such dominance if there had been equivalent free alternatives already established in the market.

      I simply see no big and obvious economic benefit for anyone except Microsoft and Nokia to care whether Windows Phone lives or dies.

      • Rich

        Right now, Windows Phone is like Schroedinger’s Cat…it’s simultaneously alive and dead.

  • http://www.facebook.com/people/Steven-Stratton/562587092 Steven Stratton

    Google and Motorola are still two desperate companies; Google has even stated that it will stay that way. For good reason too, Motorola makes awful devices, and they are too close to Verizon, which Google’s open source views – hence the galaxy nexus fiasco.

    • http://www.facebook.com/people/Steven-Stratton/562587092 Steven Stratton

      Desperate* companies

      • http://www.facebook.com/people/Steven-Stratton/562587092 Steven Stratton

        Ah! Fucking keyboard! Separate **

        • def4

          Don’t blame your Freudian slip on the keyboard :-P

          • http://www.facebook.com/people/Steven-Stratton/562587092 Steven Stratton

            Actually was me swiping it on my phone :P. I know I should stop but I can’t, lol

    • steve_wildstrom

      I like desperate, sort of a cross between disparate and separate.

      I wish I had some idea of what Google wants to do with Moto. They are eventually going to have to write off a substantial part of the $12.5 billion purchase price and it lost $500 million in the most recent quarter. Unless Google literally has money to burn, there most be some strategic plan here beyond operating Moto as a money hemorrhaging second-tier maker of Android phones.

      We should get a clue from the next generation of Moto products, which will be the first to have been substantially designed after the acquisition.

  • http://www.facebook.com/profile.php?id=100001035136448 Sihan Zheng

    The open vs closed model is hard to argue though, Palm was successful when it was open. When everyone used Palm OS to make their PDAs, Palm was highly successful. However, when Palm OS was effectively only used by Palm themselves, they faltered.
    Symbian was demolished by iPhone when it was open, Remember that Symbian was only taken in house after the disastrous n97 generation, where the n97 was practically given away by nokia.
    Personally, I think Ope vs Closed does not matter at all, what matters is the ecosystem. People are only willing to write software for an OS with a large market share. If your single vendor can move enough units to gain respectable marketshare, than it would get software support. Else, you need multiple vendors to succeed.