The proliferation of mobile platforms

Sep 4, ’11 3:15 PM

Author Horace Dediu Categories Industry

While some mobile platforms are being withdrawn from the market, others are being introduced. The net is that there are more mobile platforms announced for 2012 than ever before. The following chart shows the lifespan of the platforms that I can recall.[1]

Amazon’s entry is only rumored but I think the description of it is detailed enough to be credible. The day after the scoop with Amazon, Baidu announced the Yi platform. Baidu is the sixth most visited site in the world, so it’s not a bit player and it makes as much sense for Baidu to have a mobile platform as it does for Google.

Aliyun is Alibaba’s latest and it, like Amazon’s, is positioned on commerce and shopping. I also included the platform underlying Barnes & Noble’s Nook for consistency.

Many of the newcomers are Android-like or Android-derived platforms but cannot be rightly called Android as they are not licensing the Android trademark nor any of the proprietary part of Android. They may run some subset of Android apps but that is not guaranteed to continue.

It’s also worth noting that many of the newer Android-derived platforms are being launched by Google’s competitors and indeed seem to be designed to take value away from Google.

The epic struggle Android vs. Google is sure to continue.

—

Notes:

The lifespan is bounded by the dates of public announcements of start and end of development efforts by the platform owner. It does not include times when products are still in the market or the platform is still supported.

‹ Previous Post Quantifying the phobia of owning Apple shares

› Next Post Why Carol Bartz was fired

http://twitter.com/macintux John Daily

Should we just assume at this point that the GPLv2 is meaningless when it comes to Android? I find it hard to imagine that there aren’t myriad violations of the license in the mobile world.
- Anonymous
  
  It depends what you mean by meaningless. Certainly some portions of Android are still covered by GPLv2 and will be open sourced, though there are potential technical violations from all and sundry due to late release of source-code.
  
  The same potential technical violations exist for Apple and Webkit.
  - http://twitter.com/macintux John Daily
    
    WebKit is licensed under LGPL, which has less restrictive requirements (although it’s been long enough that I’ve paid attention to this that I wouldn’t be able to intelligently describe the distinctions.)
    
    I see that B&N has released their GPL code, so perhaps distributors are taking it seriously.
  - Anonymous
    
    The distinction is that linking to an LGPL licensed library doesn’t cause licensing issues , but the code distribution requirements of the actual code are identical, as is the permanent revoking of license in event of failure to comply.
  - Svdwal
    
    Isn’t Apple the licensor of WebKit? Surely the licensor can release the source at it’s own terms, even though other parties must play with the LGPL?
  - http://aegisdesign.co.uk Shaun Murray
    
    It’s licenced under LGPL and BSD in various places. Since it’s derived from the KHTML project, it’s not entirely up to Apple how it is licenced.
  - http://twitter.com/macintux John Daily
    
    WebKit is licensed under LGPL, which has less restrictive requirements (although it’s been long enough that I’ve paid attention to this that I wouldn’t be able to intelligently describe the distinctions.)
    
    I see that B&N has released their GPL code, so perhaps distributors are taking it seriously.
  - Anonymous
    
    Except that WebKit is about one billion times better-managed than Android and would not fail in those same ways. And it is licensed under BSD/LGPL, not GPL. In spite of the 3 common letters, those are different things. WebKit is also much older and much larger and much more successful than Android.
  - Anonymous
    
    LGPL is no different in the ways in which GPL is causing a potential risk to Android, and yes Apple has been far from perfect about making prompt releases of the LGPL portions of the iPhone build of Webkit.
    
    See for instance http://www.appleinsider.com/articles/11/05/10/apple_releases_ios_4_3_webkit_source_code_after_complaints_from_developers.html
- r.d
  
  Andriod is Apache License for the most part.
  GPL is Linux Kernel, GCC compiler and some parts of WebKit.
  Everything else doesn’t need to be shown by google.
  So Google can easily go with BSD and LLVM Compiler to free
  itself from any obligation in the future.
Anonymous

Where is Newton OS? Or are youbincluding only OSes that ran on mobile devices with a phone modem?
- http://www.asymco.com Horace Dediu
  
  I did try to list only platforms for mobile phones (PalmOS was not used on phones until about 1999 but I noted its start with the PDA Palm).
  
  Also missing is LePhone which is an Android-like OS from Lenovo. It launched about a year ago and shipped around half a million units.
  - Anonymous
    
    Newton did have a fax machine.
    
    I would start Palm when they started making phones. The company named “Palm” who made phones is not the same company named “Palm” that made PDA’s.
  - http://www.asymco.com Horace Dediu
    
    The chart refers to operating systems that powered mobile phones hence PalmOS, not Palm. PalmOS was first used in a data communicating device in the Palm VII in 1999.
    
    http://en.wikipedia.org/wiki/Palm_VII
    
    It was then used in many phones by companies other than Palm. The first voice phone running PalmOS was the Visorphone expansion module for Handspring PDA which was introduced in 2001.
    http://www.pencomputing.com/palm/Pen38/visorphone.html
    
    By the same metric, Windows Mobile was first introduced as Windows CE in 1998 but did not power phones until about 2003.
    
    The Newton OS did not evolve into a phone OS which is why it’s not included.
- Canada Mark
  
  I think Newton was forgotten in the above chart. It is important to note that Palm did not gain phone capability until ~2003 (or there about). However, the Newton OS is what defined the “modern” mobile operating system!
  - Anonymous
    
    No, the Newton OS literally defines the traditional mobile operating system. It was the first PDA, used the first mobile ARM chip, and all the other PDA’s and smartphones copied it until iPhone. The stylus, the handwriting and Graffiti input (Graffiti was Newton app before Palm even made their first device), and the embedded-type OS with applets are all the traditional mobile stuff, not modern stuff. Windows CE is Newton-class. A Palm V is Newton-class.
    
    For the modern mobile operating system, it is iOS that defines that: a desktop class OS and desktop class native C API’s and desktop class HTML5 browser instead of all the baby mobile stuff; multitouch with 3D hardware-accelerated OpenGL interface; built-in 1-click stores for apps, music, movies, books; app auditing for zero malware; regular software updates from a centralized source so that the device gets fixes immediately before bugs can be exploited by hackers.
  - Anonymous
    
    Namaste JohnDoey. Explains a lot. Very useful.
  - http://www.asymco.com Horace Dediu
    
    PalmOS enabled wireless data communications in 1999 and cellular voice communications in 2001 (via Handspring licensed implementation).
poke

With seven Android variants, many from major players in important markets, it almost seems obvious that Google needs hardware to stay relevant now. Keeping Android “open” as long as it has may have already cost it the Chinese market. I expect their licensing to get more restrictive. I expect to see Google logos on Motorola phones.
poke

With seven Android variants, many from major players in important markets, it almost seems obvious that Google needs hardware to stay relevant now. Keeping Android “open” as long as it has may have already cost it the Chinese market. I expect their licensing to get more restrictive. I expect to see Google logos on Motorola phones.
- Anonymous
  
  Google is the only one who can make a higher-quality high-end Android phone that will sell, in order to compete with iPhone. Since we know Apple is making a low-end phone to compete with Android, it is going to hurt Android if they can’t get some high-end customers where all the money is.
  - Anonymous
    
    Not all the android phones sold currently are cheap, the upper tier of Android phones may not be quite as expensive as the iPhone, but they’re close.
    
    The Galaxy S 2 for instance is about £500 unlocked from retailers, or roughly as much as a 16GB iPhone-4. Samsung has reported it sold 5mil in roughly the first quarter.
    
    http://www.mobiledia.com/news/100097.html
    
    The problem is that mostly all we have are ASPs and not full model sale breakdowns for vendors, so it’s hard to quantify exactly how big the upper end of Android is.
  - Anonymous
    
    Agree, JohnDoey. Apple is surely to build a line of phones, just as they did with the iPod family and there rule and roost.
    
    The possibility of a larger iPod touch with 3G? sounds intriguing. Use of Skype might meet my needs and could be somewhat disruptive in the Celular market. Such might be a bridge between the three iOs groups; iPad, iPhone, iPod, some of this and some of that.
Guest

“It’s also worth nothing” don’t you mean “It’s also worth noting”?
http://www.asymco.com Horace Dediu

Good overview of the Chinese OS efforts here: http://www.insidemobileapps.com/2011/09/02/china-chinese-smartphone-ios-android-market/
- Anonymous
  
  GridOS has actually been demo’d so it’s arguably more developed than the last three on your chart.
- http://aegisdesign.co.uk Shaun Murray
  
  If GridOS is an OS then so is Samsung’s TouchWiz.
  - http://www.asymco.com Horace Dediu
    
    There is a fine line between skinning an OS and forking it. I believe Grid has crossed that line and Samsung has not, but I leave the categorization to the company. Fusion Garage calls Grid OS an operating system. Samsung does not call TouchWiz an operating system.
  - http://www.asymco.com Horace Dediu
    
    There is a fine line between skinning an OS and forking it. I believe Grid has crossed that line and Samsung has not, but I leave the categorization to the company. Fusion Garage calls Grid OS an operating system. Samsung does not call TouchWiz an operating system.
Philip Bergen

I think it is fair to assess that the openness of android is biting google in the coconuts here. All the competing platforms will easily attract scores of developers delivering already on the android platform. That produces an enormous inertia that effectively prevents google from any bold changes. Net result is that the more successful the platform is the lower the headroom for innovation for google. Compare with Apple where the forces point in the exact opposite direction.
- Anonymous
  
  It’s not the openness of Android that is hurting Google, it’s the lack of quality of Android devices that is hurting Google. They envisioned it as a Windows Mobile and BlackBerry replacement, later styled it as an iPhone replacement, but Android is currently just replacing feature phones. There’s no money in that. And when there is no money, it’s hard to get all the different stakeholders working together. There is a temptation to do deals with Google competitors rather than Google, or just use the core system and differentiate the product for some niche, like Amazon is apparently doing.
  - http://riverlaw.myopenid.com/ riverlaw
    
    they are relying are more than just the carriers liking them. They split add and app revenue. Or at least thats how I understood it to be going down.
  - http://riverlaw.myopenid.com/ riverlaw
    
    they are relying are more than just the carriers liking them. They split add and app revenue. Or at least thats how I understood it to be going down.
  - kaveman
    
    This is one of the best commentary on Google’s foray in the mobile sector to date. App stores (i.e.,downloadable mobile apps) on various mobile platforms have devalued the need for a universal search (and ad-based revenue stream that Google is banking on via Android) with proliferation of interest-specific mobile apps, in my opinion. Apple and Microsoft’s decades of monetizing their platforms through product/service integration and software licensing respectively is a stark contrast to Google’s inexperience in the ‘platform’ business.
MOD

This does not bode well for Microsoft/Windows. It could be the next Nokia.
http://www.thenewsmall.com Phil Simon

I don’t consider a mobile platform to be separate and apart from a company’s platform. The Google platform contains many planks, of which Android and mobility in general are two.
- Anonymous
  
  Platform has a specific meaning in the realm of technology, you can ignore that meaning if you like but it’s not conducive to meaningful conversation with those who do not.
http://www.thenewsmall.com Phil Simon

I don’t consider a mobile platform to be separate and apart from a company’s platform. The Google platform contains many planks, of which Android and mobility in general are two.
http://www.thenewsmall.com Phil Simon

I don’t consider a mobile platform to be separate and apart from a company’s platform. The Google platform contains many planks, of which Android and mobility in general are two.
Anonymous

I think you have a typo.
“It’s also worth nothing that” should be
“It’s also worth noting that”
Billykunz

Horace:
You are the best.
Anonymous

On a different subject I´d like you (Horace) to discuss the strategic options for Telcos given that they are being commoditised and desintermediated by Google, Apple, Amazon, Facebook and the likes.
What kind of strategic moves are expcted from them?
There´s a lot of things they are attempting (WAC, Telco 2.0 initiatives, Developers Platforms like Bluevia, etc) but they may have already missed the timing.
On the other hand Telcos are needed to provide increasing bandwidth, but how can they keep investing while losing value to the Over the Top players who in turn need and benefit from them?
What will happen with Net Neutrality issues?
Being voice the major source of revenues for the Telcos, what is the perspective for voice going massively to voip being controlled / serve from the cloud from players like Apple and Google?
Thanks in advance!
- http://www.asymco.com Horace Dediu
  
  Operators have, by definition, no strategic options. The business of operating a network means you are bound by the physical, political and regulatory limitations of that network. You can’t one day wake up and say that networks are a bad business model and vow to become something else.
  
  It’s like asking what are the strategic options of being a trucker. You could decide that maybe it would be better to be a ballerina but you can’t define that new job in the context of long distance haulage.
  
  All discussions related to strategy for an operator are in reality discussions about tactics.
  
  This is a brief and flippant answer but I would want to set the expectations up front: Being in the networking business, no matter the network, never, ever, ends well. You will always end up a commodity and you can’t escape into another business. The only discussion to have is how long it takes.
  - Anonymous
    
    Being in the networking business, no matter the network, never, ever, ends well. You will always end up a commodity and you can’t escape into another business.
    
    I think that’s true for carriers outside the US where oversight is strong, but inside the US perhaps not. The spectrum that they have been allocated gives them a monopolistic dimension, but so far at least without a remotely adequate regulatory framework, and there is the possibility for them to use that unregulated monopoly to gain leverage over some other portion of the wider mobile market.
    
    I think we could for example see Verizon fork Android and capture a larger portion of the profits of their network by grabbing a share of all the advertising, app revenue, media consumption etc, in a way that we couldn’t see from Vodafone.
  - Anonymous
    
    Beat me to it Although I think Vodafone could also do it but I have to admit that I am not that familiar with the relevant regulations and untitrust laws they have to obey.
  - Anonymous
    
    Voda can’t do it, because Voda doesn’t have anything like the same captive market that Verizon does. Take the UK market, people may have some small preference to Voda or to Orange, but it’s not significant because the market is small (geographically) and coverage is essentially universal from all carriers.
    
    OfTel is also far more hands-on than the FCC. Even if Voda had a captive market due to bandwidth constraints they can’t force all their subscribers onto their choice of handsets, you’d see an immediate proliferation in virtual networks which Voda is obligated to sell capacity to.
    
    Some emerging markets (Mexico?) may have operators with even fewer constraints on their market power.
  - Anonymous
    
    Let’s turn this around. Maybe Apple can turn the European regulatory environment to their adventage?
    
    Apple Wants To Be It’s Own Carrier; MVNO Patent Application Extended
    http://tcrn.ch/pp84Oq
    
    Sorry for the double post, I did not want to break the threads.
  - Anonymous
    
    Think Voda owns 50% of Verizon.
  - http://www.asymco.com Horace Dediu
    
    Monopolies also never end well because monopolists are almost always doomed to try to perpetuate their monopolies. By my definition, a bold strategy is the replacement of one’s core business with an asymmetric one because it will happen anyway, so you might as well be the one doing it.
    
    Becoming a monopolist mean becoming the ultimate incumbent. Cooped up in an impregnable fortress of profits, tending the walls in perpetuity.
  - Anonymous
    
    True, but that is because most monopolies suffer from anti-trust concerns that limit their ability to extend their monopoly, MS would be a far greater force in mobile if they’d been able to extend their desktop monopoly to the web the way they wanted to.
    
    Carriers are a natural oligopoly, and indeed have local effective monopolies without the same restrictions.
    
    It’s far from certain that Verizon or AT&T will find ways to unlock this potential, but I think that it is there.
  - Anonymous
    
    Here’s a wild, half baked, lunch break idea.
    
    What if one of the major operators decided to up the stakes in verticalization and bought or stated it’s own device business.
    
    Here’s a thought experiment on how it could work. They’d need to create a handful of phone and tablet models to address different market segments. They already have a great sales channel where they could push these. They could get rid of all of their supplier’s products to free up shelve space. By limiting the options to maybe just a single device in each segment I figure they’d made it easier on their customers and sales people to find and sell the right device. They could have their own product brand. It could be e.g. a Vodafone phone again instead of an iPhone or Android phone. Sure, they would upset their suppliers but in exchange they would have a whole ecosystem including the network within their control. It seems to be the fashionable thing to have these days.
    
    Once they have this ecosystem, there are a number of ways to monetize it. Just throwing out some ideas that are possible if the network is also part of the ecosystem: paid content through the phone bill, marketing data based on the network traffic users generate, special deals for people on certain plans (e.g family plans, small business plans), … Of course these options may already be available today but how much easier it would be for operators to make this happen if they also controlled the products on their network.
    
    Obviously they should also sell subsriptions for other vendors’ equipment as well. As far as network neutrality I see that as a regulatory question much in the same way as MVNOs:
    http://en.wikipedia.org/wiki/Mobile_virtual_network_operator
    
    Come to think of it, the above may be a good business plan for an MVNO.
    
    The flip side of the coin: Apple could do the same thing the other way around. They seem to have some spare cash laying around that they may want to spend sooner or later:
    http://5by5.tv/criticalpath/3
    
    Discaimer: I may have had to little or too much coffee when I though this over
  - http://www.asymco.com Horace Dediu
    
    Operators have tried to become device vendors. They used “white label” phones and sell them as own-brand. In fact this was the primary “pitch” Microsoft made to operators. They brought HTC to the big global operators and encouraged operator branding of smartphones.
    
    T-mobile had the MDA
    Vodafone had the VPA/VDA (over 20 models in total)
    Orange had the SVP (about 25 models)
    
    The results ware always mediocre. The chief problem is that those phones become restricted and crippled by the limits of that operator with the goal of perpetuating their business model. Consumer don’t respond well to these devices or brands.
  - Anonymous
    
    Correct me if I’m wrong but as far as I recall the intent of those phones were to push subprime content like ringtones, wallpapers, and uninspired games for a lot of money. But the market has changed significantly since then. Maybe it would be worth another try now but this time not with model names and model fragmentation you listed above. The operators should have learned a lesson there. But in their oligopolic situation they probably haven’t. And can’t. (I liked the risk aversion topic you brought up in the last The Critical Path Episode. Sad but true.)
    
    I still think that the MVNO option is viable, maybe even for Apple. I’ll keep an eye out.
  - Anonymous
    
    How could I have missed this MVNO with a twist.
    
    Apple Wants To Be It’s Own Carrier; MVNO Patent Application Extended
    http://tcrn.ch/pp84Oq
    
    How many phases does Apple’s Evil Grand Master Plan have? They even put Mojo Jojo to shame
  - Anonymous
    
    Very unlikely that Apple will ever put that patent into practice, because the blowback from international carriers would be devastating. Every jurisdiction where they couldn’t launch a viable MVNO they’d lose distribution as carriers started to view them as a competitor and not a partner.
    
    Even markets where they could be sure of getting capacity to set up a traditional MVNO they wouldn’t be able to get this kind of dynamic system.
  - Anonymous
    
    Your’re probably right. Not that Apple would shy away from upsetting their partners though.
    
    Do we have any data on how many iPhones are sold through the different channles (Apple Stores, Operators, 3rd parties, etc.)?
  - Anonymous
    
    The problem isn’t really which channel, the problem is subsidized versus unsubsidized.
    
    But we do know that total iPhone revenues last quarter were 13.3Bn and Apple’s retail revenues were only 3.5Bn, so clearly only a small percentage of phones are sold through their own retail.
  - Anonymous
    
    Maybe the problem is completely different:
    Apple rumored to unveil 3G-capable iPod touch this month
    http://bit.ly/qKjaQO
  - http://twitter.com/octarine ジェイミー
    
    The carriers in Japan managed to control the handsets pretty tightly. I haven’t lived there since 2009, but they mandated the features and UI of the devices. It would be interesting to go back and see how Android is working out there.
  - Anonymous
    
    Android’s marketshare in Japan exploded, I believe mostly after Sharp switched to using it.
    
    http://www.businessweek.com/news/2011-05-10/android-handsets-surpass-iphone-in-japan-smartphone-market-share.html
  - http://www.asymco.com Horace Dediu
    
    That’s a great set of a data. I followed Japan’s ecosystems for some time and the degree of control that operators had there made them very rapid adopters of mobile broadband. However in the end it also turned out badly. The standards they created or adopted were often unique to Japan but were not exportable. This was called the Galapagos syndrome. Once external ecosystems matured and were released in Japan they had no trouble conquering indigenous species.
    
    The best example is the concept of an open OS smartphone. For a variety of reasons (all sensible) Japanese operators did not encourage or commission development of smartphones. As a result by 2008 there were hardly any platform devices in use. When the iPhone launched it was a sensation. The local phone vendors ended up in crisis and had to merge operations by 2010. Even though Japan was first with 3G, it ended up last in terms of smartphones (which are the natural evolution of 3G).
  - Relayman 5C
    
    Verizon has always had exclusive phones, like the LG Env (pronounced “envy”) that I have now. Some of these phones have features disabled by Verizon so that they’re consistent with other Verizon phones. I claim that one reason the original iPhone was not on Verizon is that Verizon refused to sell a phone without control over the user interface.
  - Anonymous
    
    Do things HAVE to be so bleak Horace?
    If you accept your role in life, and determine to be the best network operator you can be — spend your money on appropriate technology, don’t spend your money on pointless dreams, adopt every possible efficiency mercilessly — you can still be profitable and come out ahead can you not?
    I have in mind here Walmart in the “commodity business” of retailing.
    
    The problem, surely, is not so much being in the network business as being in that business and refusing to accept the reality of the situation.
  - http://www.asymco.com Horace Dediu
    
    All the options you cite are valid but I would not call them strategic options. Ultimately the network becomes an utility. The opportunity is in delaying that horizon through investments in new infrastructure, i.e. rebuilding the network every few years.
    
    Even moving from a wired to a wireless network did not change the economics of being a network operator.
    
    I don’t mean to be bleak, but I’ve seen this pattern in multiple industries: wireline telephony, wireless telephony, cable television, network television, computer networks, electricity distribution, etc.
    
    You can even apply the logic to non-traditional networks like road networks or air traffic or even shipping.
  - http://twitter.com/hizkyasdufera Hizkyas Dufera
    
    What if network operators started developing their own smartphone OSs like what Amazon and Biadu are doing? Wouldn’t that give them a long term strategic advantage?
Pingback: Boot up: Police ‘assisted’ Apple in lost iPhone hunt, and no US launch for … | My Tablet News
http://aegisdesign.co.uk Shaun Murray

There’s quite a few mistakes & holes in this list.

Maemo 5′s last official release was PR1.3 in October 2010. It’s still actively supported as a community project in the same way most Linux distros are.

MeeGo Harmattan is really Maemo 6 with a MeeGo twist. Underneath it’s still Debian based. Nokia have also said it would be supported for “years” so your ending in 2012 is unfair. If they’d not called it MeeGo at all and instead called it Maemo 6, it’d be nearer the truth. Maemo 5 apps run on Harmattan.

I don’t think anyone has released an official MeeGo mobile phone yet. Tablets and laptops yes, but not a phone. The N9 technically isn’t MeeGo as it uses a different kernel and upstream OS.

Symbian is officially supported until 2016 by Nokia and Accenture.

If you’re going to include Palm OS as a mobile OS then Symbian’s direct lineage dates back to 1997 and EPOC32 on the Psion 5 or even further back to EPOC16 but that’d be stretching it. Either that or Palm’s first phone was the Treo in 2002. In any case, the first phone with Symbian on it, even if it wasn’t called Symbian at the time was the Ericsson R380 end of 1999/start of 2000.

You should also perhaps include the original Communicator OS which was GeOS which is still going though AFAIK only appeared on the first two communicators – 1996 to 1999.

The iPhone wasn’t out till June 2007, not the end of 2006/start of 2007 as you have it.

If you’re going to call iMode a platform (which I disagree with) then you should add WAP as a platform as they’re pretty similar. iMode ran on top of many OSs and was more a service than a platform. We had it on Nokia N95s in Europe so which is that Symbian or iMode?

LiMo is still going and is used in particular in Japan.

I’m sure there’s more if you dig deeper so I don’t think this graph is very accurate.
- simon
  
  As far as most of the actual users are concerned, this seems fairly accurate albeit not without error. Your arguments mosty seem more of catering to a few enthusiasts and niche community than the market reality. No Maemo is dead in the water no matter how it is “still actively supported” and Symbian is in limbo. I do agree about EPOC though.
  
  Horace also killed off webOS in his chart even though webOS officially still hasn’t died, but the reality indicates that it’ll be very difficult for that OS to come back. This isn’t a definite chart for platforms, but as a general rough sketch of the market positions, it’s pretty accurate.
  - http://aegisdesign.co.uk Shaun Murray
    
    That is opinion though not fact.
    
    Maemo 5 PR 1.3 was released Oct 2010.
    
    Symbian IS supported until 2016. Elop is on record as saying that…
    
    http://www.engadget.com/2011/05/26/elop-symbian-will-continue-getting-updates-until-2016-at-least/
    
    Windows Mobile didn’t get phone support until PocketPC 2002.
    
    The Treo 180 was the first PalmOS phone in 2002.
    
    Those are all facts which makes the data Horace is using here suspect.
  - http://www.asymco.com Horace Dediu
    
    The way to think about these time frames is by imagining each platform owner pressing on an accelerator pedal, opening a throttle of investment. If the pedal is depressed fully then the platform is alive. As soon as they lift, I call it dead; and so should any developer.
  - Sander van der Wal
    
    As far as I am concerned, as a developer, Symbian died when Nokia announced in 2008 that Qt would be put on top of Symbian. But then, developers look at API’s, not implementations. Symbian lived on as one of the Qt implementations, but not as an API developers would design and implement significant amounts of code for.
    
    BTW, Qt is missing as a mobile platform, by that logic.
  - Anonymous
    
    Clearly Horace is talking about native platforms, not abstraction layers that can be applied over native platforms.
    
    You can build for Qt on the desktop but if we were listing desktop platforms we wouldn’t include it there either.
  - Sander van der Wal
    
    The Qt API was what kept old Symbian developers interested in the Symbian platform after the iOS success story, not the Symbian API. The promise of portability to other platforms supporting the Qt API, including the Maemo and MeeGo mobile platforms, was part of that attraction.
    
    The native Symbian API itself was a major reason for developers to move away from the Symbian platform, even though it was by no means dead by Horace’s definition.
    
    One of the reasons nobody programmed for Maemo was that it’s UI API (Gtk) was going to be replaced by Qt.
    
    The API is a very important part of a platform, and changing the API but keeping the platform implementation the same has exactly the same effect on a code base as moving to a different platform with a different API.
  - Sander van der Wal
    
    The Qt API was what kept old Symbian developers interested in the Symbian platform after the iOS success story, not the Symbian API. The promise of portability to other platforms supporting the Qt API, including the Maemo and MeeGo mobile platforms, was part of that attraction.
    
    The native Symbian API itself was a major reason for developers to move away from the Symbian platform, even though it was by no means dead by Horace’s definition.
    
    One of the reasons nobody programmed for Maemo was that it’s UI API (Gtk) was going to be replaced by Qt.
    
    The API is a very important part of a platform, and changing the API but keeping the platform implementation the same has exactly the same effect on a code base as moving to a different platform with a different API.
  - Anonymous
    
    All true, but it’s still not a native platform. Abstraction layers like Qt are different because they offer a different cost/benefit trade to native systems.
    
    Native systems invariably give a richer feature set and better experience, whereas cross platform software frameworks make for easy portability.
    
    Now a better argument would be that if we’re excluding non-natives then BREW shouldn’t be on the list.
  - Sander van der Wal
    
    Nokia has made it quite clear that Qt was to be the native API on both Symbian and Maemo/MeeGo. Witness their attempt to base Symbian’s and Maemo/Meego’s UI’s on top of Qt.
    
    Qt started as an abstraction layer on desktops, but on mobile it would become native. There is no technical difference between an abstraction layer and a native API, denoting one as native and the other as a portability layer are stategic business decisions.
- http://www.asymco.com Horace Dediu
  
  I added a footnote clarifying the way the dates are chosen. The lifespan is from the perspective of investment by the platform owner. Basically, the span is from when a platform is announced publicly to when its owner declares that it will no longer develop the platform. The logic being that, as a platform, it seeks to attract developers and those developers are sent signals of support or abandonment. When a signal of support is withdrawn, I chose to declare the platform end of life.
  
  It does not reflect lifespan in the marketplace.
  - http://aegisdesign.co.uk Shaun Murray
    
    That doesn’t make sense Horace.
    
    Symbian is still being developed. A new version, Belle, is due this month. Accenture are still developing Symbian further. Nokia are paying them a lot of money to do so.
    
    MeeGo is still being developed regardless of what Nokia do. There were devices announced with it at IFA. Nokia are still paying for MeeGo development and are recruiting engineers for it.
    
    LiMO is still being developed AFAIK.
    
    WebOS is still being developed. HP only announced the end of their hardware.
    
    Also you’ve not addressed my point that both Palm and WinMo didn’t enter the mobile phone market until 2002 unless you’re counting non-phone mobile OSs. If you are counting non-phone mobile OSs then where do you stop? Newton?, Psion 1?
  - http://www.asymco.com Horace Dediu
    
    Symbian died in February. Nokia made a point of executing it in public. I suspect they had obligations to do so.
  - http://aegisdesign.co.uk Shaun Murray
    
    But then they said the day after, no, what we meant was it’s still being developed until 2016 when they realised their gaff.
  - http://aegisdesign.co.uk Shaun Murray
    
    But then they said the day after, no, what we meant was it’s still being developed until 2016 when they realised their gaff.
  - http://www.asymco.com Horace Dediu
    
    One of the consequences of killing something is that it cannot be revived.
  - http://aegisdesign.co.uk Shaun Murray
    
    OK, dropped your blog. Bias obvious.
  - Anonymous
    
    Wow, somebody still believes in Symbian? To be fair Symbian isn’t dead, it’s in a coma on life support. All higher brain function has ceased and the body is mostly being kept about for the possibility of organ donation. But it’s not dead.
  - Anonymous
    
    Wow, somebody still believes in Symbian? To be fair Symbian isn’t dead, it’s in a coma on life support. All higher brain function has ceased and the body is mostly being kept about for the possibility of organ donation. But it’s not dead.
  - Anonymous
    
    Are you saying there are blogs out there without bias? Which ones?
  - Anonymous
    
    Looking at his posting history he thinks that ‘All about Symbian’ is a good unbiased place to hang :0
  - Relayman 5C
    
    Shaun, I read this blog not to criticize Horace for his decisions but to tap into his incredible insight. My comment to you: Get your own blog and then you can do things your way.
  - http://www.asymco.com Horace Dediu
    
    I replied to a different comment regarding Palm and WinMo. If a platform used for mobile phones predates the phone implementation, I include the original start date of the platform. Again, this is consistent to how platforms are tended as assets. Developers who began working with PalmOS in PDAs were able to leverage that investment into the phone world. Same was true of Windows CE developers and in fact these PDA developer communities were strong selling points for the platform vendor.
  - http://aegisdesign.co.uk Shaun Murray
    
    If that’s your basis of what to include as a mobile platform then you need to go back to the Psion Organiser and it’s analogue modem, the Psion 3 or the Ericsson MC218.
    
    The QNX bar would also go back a LONG way as would anything based on Linux.
  - http://www.asymco.com Horace Dediu
    
    There is also a natural boundary when the control of the platform crosses company boundaries.
    
    To be clear of the goal of this analysis: this is meant to inform the question of whether there is a concentration of effort by companies toward a single mobile operating system or whether there is continuing proliferation of mobile platforms.
    
    In other words, are strategists at companies assuming that the game is over and there is one OS (as has happened in the PC world) or are they assuming that the game is not over and that it is still worth-while to try to fork or develop a new platform around one’s own business model?
    
    The debate over the lifespan needs to take this point of view into account. It has to take the point of view of the manager deciding whether to invest in a platform. For that reason I ask when and which platforms received the same green/red lights.
    
    My experience and this enumeration of platforms suggests that the mobile platform space is not consolidating, yet.
  - Sander van der Wal
    
    Developers who worked with Psion’s EPOC were also able to leverage their code to Symbian OS, this being nothing more than a name change. And Psion was part owner of Symbian.
    
    Palm spun off Palm OS as a different company at some point in time.
Sander van der Wal

I would love to see a graph like this that also shows how much traction a platform has. Interesting metrics are the number of developers using the platform, to show a platform’s credibility, the number of apps released for the platform, and amount of money being made for the most obvious of reasons.
- http://www.asymco.com Horace Dediu
  
  If someone publishes this information, I’d be happy to include it.
Pingback: Monday links: Euro stress | Abnormal Returns
MOD

The only bias I see here is towards intelligence. If the CEO of a public company makes a public declaration about the direction of his business (after much deliberation) and then changes his mind the following day, he is an idiot. Apparently Nokia and HP have flip-flopped on their OSs, approximately the day after the announcement.

HP then reversed once more:
An H-P spokeswoman said the revival is “a limited run” that will end this quarter and “the decision has been made.”

At some point a decision has to be made that the company is run by idiot(s).

Would these same CEO’s declare their business is over, then walk into a bank and ask for a loan? The banker would show him the newspaper with his own quote. Customers, developers, partners would do likewise, so Horace is right to hold them to their decision. So does everyone else, who has something at stake.

It could be that these companies have European CEOs and they think they can run the world like bureaucrats: by dictum. When the world talks back to them they cower. But by then it is too late, they’ve been shown to be incompetent.
lrlee

If you are going to add in mobile OS’s not just those with telephony, then PenPoint by Go Corporation was very good.
Pingback: Boot up: Police ‘assisted’ Apple in lost iPhone hunt, and no US launch for Samsung Galaxy Tab 7.7 or Galaxy note | dailywebday.com
Karthik 001

Is Motorola’s P2K platform that powered the Razr a subset of one of the OS’ listed in this chart?
- http://www.asymco.com Horace Dediu
  
  No. All these platforms are based on operating systems that were once called “open” meaning that they had native APIs that were available to developers via SDKs. The alternative OSs used by traditional “non-smart” phones are sometimes called embedded platforms. Embedded platforms typically don’t have native applications that are written by third parties. They may run apps in sandboxes like Java or BREW.
Karthik 001

Is Motorola’s P2K platform that powered the Razr a subset of one of the OS’ listed in this chart?
Anonymous

It’s a dog race. Apple is the lean, slightly intense & focused Greyhound. The rest are mixed mutts some of whom are quite sporty but most who are mangey loping inbreds, a few chasing a hare here, another lying down and panting for air, another messing with the ladies. It takes all kinds but in the long run it is the focused Greyhound that sets the speed to beat, the challenge to win the race and inevitably leaves behind all mutts in its stride.
- Jameson Williams
  
  This is a really short-sighted fan-boi style comment, missing opportunity for objective rationale or any significant commentary other than “I own an iPhone.” You’re saying that iOS sets the pace, even though some of these OSes have been in mass deployment since the 1990s?
  - mhikl
    
    Very good summery, Jameson. It’s not everyone who can read between the lines (kudos)—except that I do not own an iPhone.
    RIM missed its chance to do what Apple did with its late entry into this race. For a company, any company, to enter such a race so late and then come to where it is today deserves praise. We can only wait to see how well MS does this late in the race, and if it does well, we should all celebrate.
    The wise man does not let envy rule his objectivity, and name calling is unbecoming.
  - Blah
    
    twat
Pingback: Čecháček » Mobilních platforem máme stále nejvíce v historii
Pingback: Boot up: Police ‘assisted’ Apple in lost iPhone hunt, and no US launch for Samsung Galaxy Tab 7.7 or Galaxy note | Richard Hartley
Pingback: Mobile Web App vs. Native App? It’s Complicated | I want android apps
Pingback: iPhone Application Developer Australia | iPhone Custom Business Apps & Cloud Computing | FileMaker Developer - Australia SA | Mobile Web App vs. Native App? It's Complicated
Pingback: Baidu-droid; Samsung ‘no to WebOS’; Amazon Kindle tablet; Apple iCloud design – Telco 2.0 News Review | Outlooks & Insights
Pingback: Benefits Of Mobile Sites Vs. Mobile Apps: Differences Decoded For Mobile Marketing 101 | Search Engine Optimization, Internet Marketing & Website Design Blog From Mannix Marketing

asymco

Categories