iphone vs. Symbian vs. Android vs. Limo vs. Ovi : We cannot compare an ecosystem with an operating system

The big news of this week was the open sourcing of the Symbian operating system. Having now had some time to think of this .. here are some more thoughts about this rather unexpected but seminal opengardens development

APPLES AND ORANGES

The title of this blog shows how often people compare the proverbial apples and oranges. It is not possible to compare Symbian vs. Android; or Symbian vs. iPhone .. because it is not possible to mix operating systems with ecosystems

iPhone, Ovi and Android are ecosystems. In contrast, Symbian and Limo are operating systems or Operating system consortia.

Thus, iPhone vs. Android vs. Ovi is a valid comparison(three ecosystems) as is Limo vs. Symbian(two operating systems) .. but NOT Android vs. Limo(an ecosystem vs. and operating system) .. and (not yet) Android vs. Symbian

By including the iPhone in this mix, you can see that I value the ecosystem vs. the operating system – and that is correct. If anything, the Symbian announcement shows conclusively that the mobile operating system is a commodity .. even when Symbian has shipped 150 million devices with annual revenue of £195 million, at £210 million – the Nokia deal values Symbian at only two times the revenue.

Now, let us consider the question of Open source ..

OPEN SOURCE

Open source conjures up images of geeks working late into the night for altruistic reasons and for peer recognition .. Indeed that’s one component of the story – and it is also the motivation behind the original success of Linux on the PC. However, there is another angle here .. companies are getting into the Open source play in a big way – i.e. creating code and then open sourcing it under specific open source licences(the license itself matters a lot and for a greater discussion of this issue see the blog Open source vs. Open Standards – complementing or competing?

There are many prominent examples of this approach .. Apple(and then Nokia) with Webkit, IBM with Apache, IBM(subsequently) with Linux

And why will companies give away code for free ..

Because everyone benefits if we have interoperable systems(i.e. no one driver for the code base) BUT at the same time – we will find that competitive advantage can be maintained depending on which open source license you adopt(see the link Open source vs. Open Standards – complementing or competing? above to see how Android exploits this idea brilliantly)

However, when it comes to Open source on mobile devices, we have yet another factor which comes into play i.e. the BOM(Bill of material) of the device. Software can play a part in reducing the Bill of materials (i.e. the combined hardware and software cost) – especially if it is modularised. This was the original reason why Linux was ported to mobile devices and also explains the rise of specialist Linux integration vendors like windriver

Thus, there are two audiences – developers and device manufacturers. It is debatable how many developers will benefit from an open source Symbian(or will toil through the night contributing code to it). The real question is – how many NEW device manufacturers will take up Symbian as an Operating system.

IMPACT ON THE MARKETPLACE

I believe that the biggest impact of this announcement will be on LIMO since it directly correlates to Symbain(in contrast to Android which is an ecosystem). Indeed LIPS(another Linux consortium) has folded up after this announcement into LIMO and interestingly Gartner analyst Carolina Milanesi dismisses LiMo as a major factor and calls it a PR machine.

Symbian is tested. It is now open source. It is familiar to developers. It is a safer bet if you want a pure OS.

The Jury is out – but the real question is – how many more device manufacturers will take up Symbian as an Operating system? How modularised is it for an OEM vendor(say from Taiwan) to take up and deploy? Etc etc.

In the blog Crossing the chasm with Android , I said that the key benefit of Android will be the emergence of new entrants into the devices market. In that sense, the Symbian announcement is good because it will allow more players to enter the devices space. Even then, we need to see more from Nokia. Symbian is not yet an ecosystem (like Android or even Ovi). It is merely an operating system.

MOBILE DEVICES AS A COMMODITY

Ultimately, not just the operating systems but the device itself may become commoditised. This is ironic since devices will become even more important – both for customers and for brands. However, they will also become cheaper, commoditised and modularised. Of all the companies, Nokia knows this best and for a year now Nokia has been busy reinventing itself as an Internet company – a truly far sighted vision that is paying off big time now. For the same reason Ovi is more relevant than Symbian / LiMo foundation and one to watch over time.

Indeed I see many more devices(from many more vendors) who will enter this space.

Interestingly, the LIMO foundation press release about the Symbian announcement says

We welcome the formation of Symbian Foundation as it follows behind the pioneering direction of LiMo Foundation which recognizes that the industry must coalesce on far fewer handset OS’s in order that innovation can be unblocked and far better propositions brought to consumers.

I disagree with this. I actually don’t think we are coalescing around fewer devices. Instead I see MANY more devices which will be launched and that’s a good thing.

The most common question developers ask is: How do I get this service on to phones? With many new entrants, the possibility of getting a service on a phone increases. At the moment, there are really only four device manufacturers(Samsung LG Nokia and Sony Ericsson) + Motorola (with its status in balance). If we see many more device vendors, that is a good thing. Indeed 3D holograms from Infosys and the award winning Gemalto’s smart card web server(pdf)

could be their own ‘devices’ i.e. if the device becomes cheap then we could model the device around a specific development(like SCWS)

AND WHAT ABOUT ANDROID/GOOGLE?

Contrary to media comparisons, I believe that Google is addressing a fundamentally different problem to most other people(including Nokia). It is addressing the problem of making Web services usable on mobile devices(mostly existing Google services like mail and maps).

This needs advertising support simply because there is no other revenue model that can work. Why? Because the information is available mostly for free on the Web(or we have services like email which are also free on the Web). It is not possible to charge a premium for this service just because it is ‘mobile’(A lesson which many in the mobile industry doggedly refuse to accept!). (You see the same problems with companies that try to monetise mobile widgets). Instead, the service becomes free and ad funded.

And Android has been designed from the ground up to achieve this goal(i.e. within every element of the stack) and has features like XMPP integrated into it XMPP (features which have no parallels in an integrated manner currently). And is modularised. This strategy will pay dividends depending on how many people are addicted to Google web apps(and I am!) and will want them on mobile devices – a trend I explored in a blog called Am I the only one who uses Gmail with Blackberry .. and what does it say for mobile apps ..

SHOW ME THE MONEY?

So, if the OS is free, the devices are free .. etc etc .. where is the money going to come from? The answer lies in understanding the behaviour of the PC/Web software industry. I come from a background of PeopleSoft(now Oracle corporation) prior to mobile – and Oracle continues to make lots of money – not from software but from services(more than half it’s revenue) and UBS analyst Heather Bellini calls Oracle a ‘profit machine’

I see the same trend on mobile devices. As devices and systems software becomes cheap(like on the PC) but ironically more complex, applications and services will drive most of the revenue. These could be advertising but also subscription services(including some form of tech support/ insurance etc)

DR IRVING WLADAWSKY-BERGER THE UNSUNG HERO OF THE OPEN SOURCE MOVEMENT

Let me conclude this article by mentioning the impact of Dr Irving Wladawsky-Berger – a great unsung hero in my view for Open source movement

Dr. Irving Wladawsky-Berger of IBM has done more than anyone else to adopt the ideas of open source to companies – analogous to what Linus Torvalds did for Linux in the Operating system domain. Although I have never had the pleasure of meeting Dr Berger .. he has an infrequent but insightful blog HERE which I recommend you follow . Today, as a result of that vision from some corporate pioneers like Irving Wladawsky-Berger – we see a unique and a vibrant ecosystem – not dominated by a company, or by governments.

Indeed, that’s the reason I have always said that the market will develop differently in Europe, North America and many other places in comparison to Japan and Korea and inspite of early advantages .. Europe and North America will be the leaders for Mobile applications that also span the Web.

Here is a section from an interview with Irving which shows a pragmatic vision to open source – and over the years it has given rise to a whole ecosystem as we see today in the mobility space.

>>>

TG: Sun has committed to releasing all of its code as open source. Do you think IBM will do the same?

IW-B: I don’t think so, because I honestly don’t think everybody wants to see all your code. Remember, the key to open source is not the ability to see the open software, it’s the forming of a community around it that will participate in its development and its maintenance.

You cannot go in your closet and look for old code and throw it out there and tell people to form a community around it. They may say, Irving, that’s legacy code that we have zero interest in working on. We continue to open source quite a bit of code, but we are fairly selective, and we work very closely with communities to decide whether to open source or not.

<<<

Open source and open standards will lead the way – and may it so remain! We are seeing the fulfillment of a trend which began about a decade ago .. and it is proving very disruptive in the mobile domain – as Android and the Symbian announcement continue to demonstrate

Also see

Crossing the Chasm with Android

Open source vs. Open Standards – complementing or competing?

Comments

  1. Android is an operating system not an ecosystem. Ovi is an ecosystem, iPhone is both an ecosystem and an operating system.
    So far Android has nothing more than just the OS. No distribution channel, no phone, therefore no ecosystem.

  2. Adam Nemeth says:

    Hey,
    I simply don’t understand you: comparing iPhone and Symbian isn’t like comparing apples to oranges: it’s like comparing two oranges, and that’s exactly what happens here.
    What do I mean: Both are specific platforms for a range of devices. While Symbian is a bit of a split platform (S60 and UIQ), and iPhone has a very limited range of devices (currently, there are about 3 or 4 different versions out there, plus the nearly coming 3G?), they’re still platforms.
    And by platform, I mean these:
    - They’re to perform the same functions (phoning, sending SMSes, address book, etc), with a same kind of device (let’s call them mobile phones, or smart phones)
    - They’re incompatible with each other, that’s why they provide a distinct platform (a native(!) iPhone app cannot run on a Symbian device, and vica versa) (Contrary: J2ME platform)
    - However, they’re compatible with themselves! (I mean, if you have a Nokia N95, and you get a program written for a Nokia N70, it’ll likely work on that too)
    - They’re open to be extended, providing an interface for developers (even if you consider symbian an OS AND NOT an ecosystem, I can still google for S60 apps and I’m pretty sure I’ll find a few) (OK, NTT’s Symbian OS isn’t playing here, it’s not a platform then)
    As for “linux on mobile phones” I always fear. I fear because linux isn’t an OS where binary compatibility is a known thing to work. Linux binaries don’t work across distributions,they don’t work across releases of the SAME distribution. Linux in itself usually fails to provide the compatibility part of a platform, and that’s the main reason i think they failed on the mobile market yet (a phone isn’t a thing where you consider gentoo a viable solution, nor a thing where people love to give sourcecode for free, because it simply s..ks to develop for them)
    Android is a platform not because it runs linux – it’s irrelevant from the platform part – but because it has a binary compatibility layer, in that tricky dalek binary code which currently compiles only from java, but I guess, since it’s a VM after all, that there’s no problem of having codes written in C#, python, or any other language compile for them.
    A platforms’ openness is one of the factors which makes developers use that platform, so do users. Other factors are:
    - Device compatibility (on how many currently deployed machines does it run)
    - Price and efforts needed to get your application running on it (including how much pain it is to develop for symbian, and how much does the iPhone SDK costs)
    - Availability
    Openness has levels. Levels can be:
    - Hardwired: it’s in ROM, or a chip
    - Internal openness: it’s on a flash, can be upgraded, but only the company itself can release new versions, documentation is top secret
    - Openness for partners: SDK needs contract and mostly money
    - Openness for everyone: There’s a published SDK, you can develop what you want
    - Open source: you can understand the platform and you can use deeper layers, probably you can release a modified version of the platform suitable for your needs
    Maybe the last thing will be possible with opening symbian, but I’m not sure: you cannot have your own Mac OS X kernel even given the fact that the code is published: there a lot of things missing for that currently.
    And modification can also kill the platform’s main nature: compatibility. When a program does run on a set of devices with a given “OS”, and does not run with others, then, from that program’s standpoint, its platform is only that limited set. (that’s why symbian is a split platform, because of S60 and UIQ)
    So, Nokia wants more developers, maybe more devices, and in general, spread the Symbian platform, because it seems people want alternatives other than windows mobile.
    Will they succeed with this move, trying to make Symbian more appealeable for developers [and maybe for manufacturers]? Dunno. But if iPhone, Android and Symbian weren’t three oranges compared to each other, they wouldn’t spend a penny for opening it now I’m sure.

  3. Ajit Jaokar says:

    Adam
    Thanks for your insightful comment!(also thanks to Peter for his comment as well)
    To clarify
    By ecosystem I mean: Android is more than Linux(it includes webkit, dalvik, XMPP etc i.e. the whole stack), similarly Ovi, so also the iPhone(not just the OS but also the appstore and so much more).
    In contrast – Limo is no more than Linux.
    Symbian(as of now) is no more than an OS.
    Thats why we cant compare the two sets :)
    kind rgds
    Ajit

  4. Guilhem says:

    Ajit
    If you were to call something an ecosystem then Symbian definitely is one (WinMob is one too) and Android is not (yet):
    There exists hundreds of ISVs making money out of selling commercial apps for either Symbian or Winmob.
    On the other hand, there are no Android phones on the market yet and AFAIK no one is making a living out of Android apps yet.
    In order to have a true ecosystem, you need a food chain ;) and the Android foodchain (i.e. cash pump across the value chain) is not setup yet
    my 2 cents
    Guilhem
    PS: these are my own personal views and not those of my employer

  5. Ajit Jaokar says:

    Thanks Guilhem
    These comments help to clarify my thinking and wording .. I guess what I meant when I said ‘ecosystem’ is a ‘stack’ i.e. Android is more than Linux (i.e. webkit, dalvik etc). in contrast Symbian / limo are OS es only. Probably still some iterations to go to get this right! kind rgds Ajit

  6. Malcolm says:

    Perhaps the reason people are arguing about whether Symbian is a “stack” or whatever vs. iPhone is that you’re comparing slightly different things yourself when you don’t need to, and using misleading terminology in doing so.
    The iPhone doesn’t include the application store or MobileMe — it’s just the OS/apps and hardware. Perhaps you should be comparing Apple and (now) Nokia. Clearly this is an apples to apples comparison (no pun intended).
    If you wanted to another apples to apples comparison, it would be Symbian vs. Mac OS/X (iPhone version, or inc. desktop if you like).
    Android doesn’t compare to either of these yet, as some have pointed out, since it doesn’t include any HW. However, ultimately the difference between Google and Nokia is where they come from (Nokia from telecoms, Google from webland). This influences their emphases and weaknesses and strengths, and it will be interesting to see which works better for the mobile end-user. Both have been very successful in the mainstream market.
    When you do these comparisons you can see the real differences in approach: Apple is really going it alone with the iPhone (as usual). Nokia, on the other hand, is being even more open than they previously were.
    Looking at it this way, Nokia’s approach has clear advantages (with the disadvantage being rather more confusion for the end-user).
    Interesting times! (And difficult for ISV’s.)
    -Malcolm.

  7. Hi
    Your comment on Oracle as a money machine is valid, but I don’t believe your comparison with the mobile software market holds. Oracle makes its money selling services to businesses, and not to consumers. The mobile software industry however is targetting the consumer only, and selling services to consumers, especially services related to software, is probably even harder than selling them the software in the first place.
    Why this hasn’t resulted in mobile software vendors selling software and services to businesses is anybody’s guess.

  8. Ajit Jaokar says:

    Thanks Malcolm for your insightful comments. Yes the comparison I made is not perfect but it is a step in the right direction and hopefully I refine it as we go along kind rgds Ajit

  9. Ajit Jaokar says:

    Hello Sander
    Thanks for your comments. One of my beliefs is: yes we will end up selling a service to consumers. This (kind of) happens already for electronics. Sky sells you an annual maintencnce contract, Dixons and other stores also sell you some form of maintenance contrct. I see the same for mobile devices as they get complex. I also see a whole spectrum. from open devices(sold with insurance), to closed devices(managed by the Operator with remote helpdesk etc etc) kind rgds Ajit