Nokia’s Open Source Response: The Symbian Q&A

Share via Twitter Share via Facebook Share via Linkedin Share via Reddit

Sun’s Jonathan Schwartz, in discussing Java specifically and Sun more generally, is fond of highlighting the volume of the mobile market relative to the PC. Here’s what he had to say on the subject back in 2006:

I was asked a simple question by an investor last week, “what’s the one thing you think the market doesn’t understand about Sun’s opportunity?”

My response: that the majority of the world will use the internet through their phones, not through a PC.

I’m not sure he believed me. And within the US, I’m not sure many folks agree that most people in the world will use the internet on their phone. Yet. But considering the volumes – nearly 5x the number of people buy a phone each year, than buy a PC, the conclusion seems obvious. And I don’t know about you, but when I sample my nieces and nephews, even those in the USA, with “which would you rather have, a new iPod, a Motorola RAZR, Danger’s HipTop, Microsoft’s XBox or Windows Vista?”, I get a pretty consistent answer. (Hint: it ain’t Vista.)

Opinions on this viewpoint have differed and will continue to, but as network bandwidth and availability improve and the handsets are ever more capable, the ratio of skeptics to believers slowly reverses.

Which is why the by now old news that Nokia spent $410 million to acquire the right to open source the Symbian code base and house it within a newly created foundation is interesting. As the market expands and its importance becomes more apparent, the land grabs ratchet up in frequency.

To help take a look at it, let’s get on to the Q&A. You heard it here last, as always:

Q: Before we begin, is there anything to disclose?
A: Not directly. While we work with various entities associated with competing platforms such as Linux or Windows, there are no direct commercial ties that I’m aware of that would necessitate disclosure.

Q: For those that haven’t been following the news, can you summarize the announcement?
A: Certainly. Two days ago, Nokia announced its intention to purchase the shares in Symbian that it did not already own, and open source the assets as an open platform. The home for this new project is to be the newly created Symbian Foundation, and the licensing choice is the Eclipse Public License 1.0.

Q: What do you think prompted the move?
A: The acquisition and open sourcing has been widely portrayed as a direct response to the competitive threat of Google’s Android project. By and large, I think that’s accurate: it’s clear that this is largely a tactical response to increased competition in the mobile space, in which Android is an important (if more potential than delivery at this point) new player.

But while Android may be one focus, it’s not likely that it’s the sole target of Nokia’s efforts here. Consider the marketshare statistics cited by TechCrunch:

  1. Symbian (60%)
  2. Windows (15%)
  3. RIM (10%)
  4. iPhone (7%)

On the one hand, it illustrates the massive presence of Symbian in the marketplace. But on the other, it indicates that Apple has acquired nearly 10% of the market – in less than 12 months.

Q: How does Symbian’s technology compare versus competitive alterantives?
A: From an application availability perspective, very well. There are catalogs upon catalogs of Symbian compatible applications. From a technology perspective, many would debate the assessment but my own feeling is that it’s been lapped by the alternatives. My current iPhone, as an example, is superior to the Symbian based Nokia N75 it replaced in every aspect. The Symbian UI, by comparison, was childishly clunky and primitive, and over time it suffered bit rot. Never swift to begin with, became slower and slower over time.

Certainly that’s but one anecdotal experience on but one hardware platform, and your mileage may vary, but the fact is that the market for mobile platforms is more competitive than its ever been and will only become more so. Nokia appears to have recognized this, as it has been working on its own iPhone knockoffs. Worse, Nokia – long a leader in mobile innovation – appeared to be falling off the pace.

Presumably, Nokia believes that by open sourcing the Symbian technologies, they’ll be giving them a new lease on life and an infusion of energy.

Q: Is that a reasonable expectation?
A: Many would argue that it is not. While it’s too early to say, it is true that open source is hardly the panacea for development speed and energy that many perceive it to be. Much depends on factors yet to be determined (“The by-laws of the Foundation will be added to this site later“): what is the governance model? Who will have the right to commit code? How will that list be managed and grown? How will the project manage to rapidly innovate without compromising the stability necessary to handset devices?

Time will tell. Irrespective of the bounce it may or may not receive from the decision to open source it, however, it should benefit from the elimination of Symbian licensing fees as a barrier to entry.

Q: What does this mean for Nokia’s interests in open source generally and Linux and Qt specifically?
A: On a general basis, it’s certainly true that Nokia now has spent liberally on open source and will be for the foreseeable future heavily vested in it. What’s less clear, however, is how the different investments – both internal and M&A – will compete or not with each other. As discussed at the time of the Trolltech acquisition, the decision to support two competing frameworks in GTK (Maemo, Nokia N770, N800, N810) and Qt was curious. As was its decision to embrace the Debian foundation for the Maemo project based devices, frankly, given its long standing commitment to the competing Symbian platform.

No less complex is this acquisition. Symbian is, as mentioned, a massive market presence in the mobile space at present. But Linux has shown excellent strength of late in smaller, sub-PC class devices such as the Eee, and Nokia seemed to acknowledge this in the N770 and its successors. Then consider the Linux foundation upon which Android will be resting, and even the recently released OpenMoko Neo FreeRunner, and the Linux mobile story becomes that much more compelling. Linux, it would seem, is here to stay as a mobile platform.

As are, clearly, Apple’s and Microsoft’s respective offerings. RIM too.

Does that leave, in the longer term, enough oxygen for a relative late comer to the open source party, albeit one with entrenched strength and solid carrier relationships? Nokia’s doubling down on that bet; personally, I am skeptical.

Q: What does the selection of the EPL mean in this context?
A: If we consider licensing as a spectrum between permissive and reciprocal, the EPL could be placed somewhere in the middle. Like other file based licenses, as opposed to project based alternatives such as the GPL, the EPL permits the introduction of additions to the project that are separate from and not governed by the EPL itself.

In practical terms, this means that while adopters of the platform have less ability to repurpose or relicense the code than would users of the Apache licensed Android stack, they will have the right to introduce the proprietary extensions the carriers and handset makers alike often view as essential to their business.

The interesting question for many will be whether or not the selected licenses for the platforms, which to date have not been reciprocal, will act to encourage or stagnate mobile innovation. I’m undecided, but O’Reilly’s Nat Torkington’s view is certainly worth considering:

There’s a huge difference between Linux and the handsets, though, and I think it’s an important one. Linux’s license (the GPL) prevents people who ship Linux from including proprietary extensions. If you ship a modification to Linux, you must release the source. This means there are no privileged applications (the way Microsoft’s apps used libraries that third-party apps couldn’t), no proprietary competitive advantages in the kernel, and so the rate of improvement of every Linux distribution is maximized.

Both Google and Nokia, however, have deliberately chosen licenses that don’t encourage that…Proprietary competitive hardware and software can be put into any Android or Nokia phone at the appropriate level of the stack. I think this will slow down the success of their platforms and means neither will unlock the true potential of an open mobile platform. I believe true demilitarized openness is a necessary, but not sufficient, condition for open mobile platform success.

Q: How about the compatibility of the selected license?
A: A few of the inquiries we’ve seen incoming have asked whether the EPL is compatible with the GPL – the license that governs, among other projects – Linux. The answer to that question is no, it is not. So Symbian assets will not be permitted to be legally recombined with Linux code for their mobile efforts.

A fact that, I’d argue, was non-incidental to the selection of the license EPL.

Interestingly, however, I believe that Google’s Dalvik Java runtime could in fact be legally ported to the Symbian platform, as that would be layering the permissive Apache license on top of the slightly more restrictive EPL. Whether that’s possible technically is an entirely different question, of course.

Q: So what of timeframes? When can we expect actual handsets based on the platform to start shipping?
A: This, to me, is one of the really interesting questions, considering the market landscape. The iPhone, of course, is shipping now, as are Windows Mobile handsets. Android based units, according to Caroline, may be delayed but are expected by the end of this year or early next at the latest. Symbian Foundation units, however, aren’t expected until 2010. Will non-open source Symbian shipments be affected in the meantime?

Nokia is probably – and likely correctly – anticipating that they can weather the delay given their current marketshare and position. But the question is whether or not they’re granting, by virtue of the time it will take to marshall the SF resources and organize the code, decide governance, etc, an unacceptable window of opportunity for competitors.

Q: What does this mean for mobile platform competitors, and the mobile landscape generally?
A: Opinions seem to be divided. Glyn Moody thinks this bodes ill for mobile Linux and Androidin particular:

The move is clearly aimed at removing one of the main advantages of Google’s Android platform: that it was mostly open source. Now that Symbian goes even further in the direction of openness, some of the attraction of Android will be lost. Moreover, Symbian has ten years of work behind it, is very well known, and is already widely used – none of which is true for Android.

But Android’s loss is also Linux’s loss, since Android uses the latter as its foundation. One effect of Symbian going open source is that it is likely to divert mobile developers from writing on Linux, weakening it in the mobile space. Indeed, splitting developer interest between two open source platforms may have the paradoxical effect of boosting Windows Mobile.

Personally, I believe that overstates the value of Symbian’s history. Given my experiences with the interface and the reviews I’ve seen of other Symbian based devices, Android’s lack of the “ten years of work” might well be an advantage, in that it’s designed from the ground up to work on devices of a more recent vintage.

At the other end of the spectrum is TechCrunch’s John Biggs, who speculates that Symbian will end up retiring the assets they’ve purchased for $400M+:

There is an excellent chance that Symbian will not make it through its conversion to openness alive. The OS is old and crotchety, unable to handle data intensive applications with the same aplomb RIM or even the iPhone OS have. Once the platform is open, Nokia will most likely put it out to pasture, watch as the developers branch it off, and then build something entirely new. As popular as it is, I doubt many of us would miss Symbian’s various foibles and flaws.

This time, I believe understatement to be the problem: namely Symbian’s traction and application catalog. It is true that the OS is old and often poor performing, but that’s been true of other operating systems for years. Discount inertia at your own peril.

Q: So you think that Moody is right? Or Biggs?
A: Ultimately, the success or failure of the acquisition depends on community uptake and adoption of Symbian post the formation of the project, which in turn depends on the decisions made beyond the mere selection of a license. Is Symbian a more formidable competitor minus the cost barrier and – potentially – with an open source community at work enhancing it? Certainly.

But it already faces an open source platform of some pedigree (mobile Linux), one with much developer interest and traction (Android), one that’s reshaped the mobile expectations overnight and with its first iteration (iPhone), and an old juggernaut (Windows), so success is far from a guarantee.


  1. “Interestingly, however, I believe that Google’s Dalvik Java runtime could in fact be legally ported to the Symbian platform, as that would be layering the permissive Apache license on top of the slightly more restrictive EPL.”

    Yup. Eclipse itself (the IDE) ships loads of Apache code as part of the IDE’s distribution. And IBM then reships this. So I can guarantee this is “legal”.

    ‘Android’s lack of the “ten years of work” might well be an advantage’

    Yup. s/work/legacy/

  2. Dalvik has not been released under the Apache License. So it’s unlikely it could be legally ported by anyone but Google, and that part of the license compatibility section is moot. 😉

  3. … so far at least – the latest status of making true on the promise to release Android under the Apache License was late this year, but I haven’t really been following it – proprietary code just doesn’t get me excited.

  4. Minor nits: If Symbian is purely user-space you could put it on top of Linux. The FreeRunner isn’t quite shipping yet (est. July). Ubuntu MID is also a contender, although in a slightly different market.

  5. I do not understand why Google picked Java as their platform, *no one* uses Java any more. The iPhone doesn’t, for example. And now that Nokia’s picked Symbian, no Java there either.

  6. This is a most clarifying statement. Whoever meaningfully says this is interested in *fake* “open source”. They’re not talking about a GNU/Linux system, but only about the core component: the kernel called Linux ( http://www.kernel.org/ )

    A claim such as: «Linux’s license (the GPL) prevents people who ship Linux from including proprietary extensions.» is *SO* revealing of their plans. They want carriers and phone makers to be able to lock the telephone down.

    Free Software and Open Source Software talk mostly about the same things, but view it from different angles.

    This angle follows neither point of view. In this new oppinion we see that “open source is not for the telephone owners” as they will not have the freedom to change their phone as they intended.

    Fortunately, we now have http://www.OpenMoko.com/ and I’m getting one real soon now. This is real Free Software, not this fake bullshit.


  7. […] Foundation. Luckily we have analysts like Stephen O’Grady from Redmonk that provides a detailed and insightful analysis of the announcement.  However, I was surprised how Nat Torkington’s post regarding […]

  8. @Martin Pleny

    Google picked java because it’s a live and kicking language.

    and, about Nokia:
    – the Java sdk is the first choice on their developer page [1]
    – on every device is listed the JSR implemented [2]
    – recently they ported Eclipse SWT so S60 [3] [4]

    [1] http://www.forum.nokia.com/main/resources/tools_and_sdks/
    [2] http://www.forum.nokia.com/devices/N96
    [3] http://blogs.s60.com/java/2007/10/eswt_available_for_development.html
    [4] http://wiki.forum.nokia.com/index.php/ESWT

  9. Nokia and the Symbian Foundation Opportunity – Part I…

    Sixty days ago, Nokia announced it was buying the rest of Symbian Corp., and would then open source SymbianOS using the Eclipse Public License through a newly created Symbian Foundation. This is a great announcement. Stephen O’Grady did an excellent…

Leave a Reply

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