dw2

22 December 2012

Symbian retrospective: hits and misses

Filed under: More Than Smartphones, Nokia, Psion, retrospection, Symbian, Symbian Story — David Wood @ 12:19 pm

As another calendar year draws to a close, it’s timely to reflect on recent “hits” and “misses” – what went well, and what went less well.

In my case, I’m in the midst of a much longer reflection process, surveying not just the past calendar year, but the entire history (and pre-history) of Symbian – the company that played a significant role in kick-starting the smartphone phenomenon, well before anyone had ever heard of “iPhone” or “Android”. I’m channeling my thoughts into a new book that I’m in the midst of writing, “More than smartphones”. The working subtitle is “Learning from Symbian…”

I’ve got no shortage of source material to draw on – including notes in my electronic diary that go all the way back to January 1992. As I note in my current draft of the introductory chapter,

My analysis draws on an extensive set of notes I’ve taken throughout two decades of leadership positions in and around Symbian – including many notes written in the various Psion PDA organisers that have been my constant electronic companions over these years. These Psion devices have been close to my heart, in more than one sense.

Indeed, the story of Symbian is deeply linked with that of Psion, its original parent. Psion and Symbian were both headquartered in London and shared many of the same personnel…

The PDAs that Psion brought to market in the 1980s and 1990s were the mobile game-changers of their day, generating (albeit on a smaller scale) the same kind of industry buzz as would later manifest around new smartphone releases. Psion PDAs were also the precursors for much of the functionality that subsequently re-emerged in smartphones, satellite navigation products, and other smart mobile devices.

My own Psion electronic diary possibly ranks among the longest continuously maintained personal electronic agendas in the world. The oldest entry in it is at 2.30pm on Friday 31st January, 1992. That entry reads “Swedes+Danes Frampton St”. Therein lies a tale.

At that time, Psion’s commercial departments were located in a building in Frampton Street, in central London, roughly midway between the Edgware Road and Maida Vale tube stations. Psion’s technical teams were located in premises in Harcourt Street, about 15 minutes distance by walking. In 1992, the Psion Series 3a PDA was in an early stage of development, and I was trialling its new Agenda application – an application whose UI and rich set of views were being built by a team under my direction. In parallel, discussions were proceeding with representatives from several overseas distributors and partners, about the process to create versions of Psion PDAs for different languages: German, French, Italian, Spanish… and Swedish and Danish…

As the person who assembled and integrated all the files for different software versions, I met the leads of the teams doing the various translations. That day, 31st January 1992, more than 20 years ago, was among my first meetings with work professionals from the Nordic countries.

I recall that we discussed features such as keyboards that would cater for the additional characters of the Danish and Swedish alphabets, like ‘å’ and ‘ø’. I had no inkling in 1992 that professionals from Denmark, Sweden, and Finland (including employees of mobile phone juggernauts Ericsson and Nokia) would come to have such a far-reaching influence on the evolution of the software which was at that time being designed for the Series 3a. Nor could I foresee the subsequent 20 year evolution of my electronic agenda file:

  • Through numerous pieces of Series 3a hardware
  • Via the Series 3c successor to the Series 3a, with its incrementally improved hardware and software systems
  • Via a one-time migration process to a new data format, for the 32-bit Series 5, which could cope with much larger applications, and with much larger data files (the Series 3 family used a 16-bit architecture)
  • Into the Series 5mx successor of the Series 5
  • Through numerous pieces of Series 5mx hardware – all of which give (in their “About” screen) 1999 as the year of their creation; when one piece of hardware ceases to work, because, say, of problems with the screen display or the hinge mechanism, I transfer the data onto another in my possession…

Why 1999 is the end of this particular run of changes is a fascinating tale in its own right. It’s just one of many fascinating tales that surround the changing fortunes of the players in the Symbian story…

Step forwards from chapter one to the penultimate chapter, “Symbian retrospective”. This is where I’d welcome some extra input from readers of this blog, to complement and refine my own thinking.

This is the first of two retrospective chapters that draw conclusions from the episodes explored in preceding pages. In this chapter, I look at the following questions:

  • Out of all the choices over the years made by the players at the heart of the Symbian world, which ones were the most significant?
  • Of these choices, which were the greatest hits, and which the greatest misses?
  • With the advantage of hindsight, what are the different options that could credibly have been pursued which would have had the greatest impact on Symbian success or failure?

So far, my preliminary outline for that chapter lists a total of twenty hits and misses. Some examples of the hits:

  • Create Symbian with a commercial basis (not a “customers’ cooperative”)
  • Support from multiple key long-term investors (especially Nokia)
  • Enable significant differentiation (including network operator customisation)
  • Focus on performance and stability

And some examples of the misses:

  • Failure to appreciate the importance of the mobile web browser
  • Tolerating Symbian platform fragmentation
  • Failure to provide a CDMA solution
  • Failure to merge Nokia S60 and Symbian

My question for readers of this blogpost is: What would be in your list (say, 1-3 items) of the top hits and misses of decisions made by Symbian?

Footnote: Please accept some delays in your comments appearing. WordPress may hold them in a queue awaiting my review and approval. But I’m in a part of the world with great natural beauty and solitude, where the tour guides request that we all leave our wireless communication devices behind on the ship when we land for the daily excursions. Normally I would have balked at that very idea, but there are times and places when multi-tasking has to stop!

About these ads

25 Comments »

  1. Symbian/Nokia forgot about Moore’s law, Samsung showed how it’s done but unfortunately with Android…

    Comment by JV — 22 December 2012 @ 1:53 pm

    • Hi James,

      >Symbian/Nokia forgot about Moore’s law…

      I still defend the focus on avoiding memory leaks.

      What turned out to be less defensible is the focus on keeping memory usage lean and mean – and the use of the hard-to-learn descriptor API as a core mechanism behind that focus.

      Comment by David Wood — 22 December 2012 @ 8:14 pm

  2. Where to start…

    1/ Symbian wasn’t innovative enough. It had a can NOT do attitude (in its latter years at least). It was hard getting any improvements into the system. I might argue that Nokia the main driving force behind Symbian didn’t particularly want it to be innovative – it wanted Symbian to do as it was told, and just deliver on time.
    2/ Too far removed from the hardware – e.g. the kernel guys never touched customer hardware (post Psion). And too far removed from the Software reality – which was why there was always push back to getting improvements into the OS.
    3/ The organisation was messed up, too much management hierarchy and inflexible staffing led to bunker mentality and politics. There seemed to be a lot of empire building in certain groups – pathetic really.
    4/ We didn’t learn enough from open source – we should have used more open source technologies and tools in the core OS.

    It’s interesting to now be working in a Semico on Android and realise how much better things can be.

    P.S. I might argue that Symbian was a success – it achieved it’s aim of stopping Microsoft dominating the market.
    P.P.S. Do I get a mention as the 7.0s tech lead? I might buy the book if I do ;)

    Comment by Peter Harper — 22 December 2012 @ 2:20 pm

    • Hi Peter – Many thanks for these observations. I’ve got material on quite a few of these points already…

      >I might argue that Symbian was a success…

      Me too: By many measures, Symbianwas a stellar success. But it’s interesting that most lists I’ve seen of hits vs. misses have a lot more on the negative side than the positive side

      >Do I get a mention as the 7.0s tech lead?…

      I haven’t written the chapter on Typhoon and Hurricane yet – it’s number 11 in my current outline plan of 24 chapters. One challenge I’m facing is that there is simply so much fascinating material, which could easily fill up multiple volumes

      Comment by David Wood — 22 December 2012 @ 8:22 pm

  3. Hits

    – To run on a variety of phone form factors
    – To attract phone OEMs to build devices for Symbian
    – Comprehensive phone API for developers

    Misses

    – To make programming easier for internal and 3rd party developers
    – To improve the Symbian UX sufficiently and in good time (became Nokia’s responsibility really)
    – To make successful the concept of the download store (again Nokia’s responsibility really)

    I believe Symbian (and even Nokia late on) missed the trick of overlaying something on Symbian to give it a newer UX and easier programming interface for developers… as I mentioned Feb 11, 2011 when I saw as the end of Nokia. (http://mobilephonedevelopment.com/archives/1206)

    Simon Judge

    [I worked on contract at Symbian for a year and was an Accredited Symbian Developer]

    Comment by Simon Judge — 22 December 2012 @ 4:21 pm

    • Hi Simon – All good points, thanks!

      >I believe Symbian (and even Nokia late on) missed the trick of overlaying something on Symbian to give it a newer UX and easier programming interface for developers…

      Indeed – that’s one of the fascinating back stories that definitely deserves to be told :-)

      Comment by David Wood — 22 December 2012 @ 8:25 pm

  4. The biggest technical mistake was giving away control of the UI and apps therefore unlearning how to write proper APIs for developers to use. Alas the technology begame unfit (apart from the kernel).

    The biggest organisational mistake was recruitement. PsiSoft became Symbian and then Nokia’s estate ….and in the process stopped hiring top people that cared.

    Comment by John Pagonis — 22 December 2012 @ 6:14 pm

    • Hi John,

      >The biggest technical mistake was giving away control of the UI and apps therefore unlearning how to write proper APIs for developers to use

      Indeed, the “Nightingale” decision to withdraw from control of the UI and apps ranks high among the list of regrets of most of the people who were involved at the time

      >The biggest organisational mistake was recruitement. PsiSoft became Symbian and then Nokia’s estate ….and in the process stopped hiring top people that cared

      I can’t fully agree with that – Symbian managed to keep on recruiting outstandingly talented engineers right to the end

      Comment by David Wood — 22 December 2012 @ 8:28 pm

  5. Hits:
    * Believing it was possible. People had been saying that “mobile computing is the future!” for a long time but it took Psion and then Symbian to make it a widespread reality. Regardless of what happened later much credit should be given to the original hard work to create the platform.
    * Innovative technology. The second version of the kernel, Active Objects / servers, etc.
    * Showing consumers that they could do more with their mobile phones than just make calls – creating the whole of the current smart-device market.

    Misses:
    * Becoming technology risk-averse. Working prototypes of quality touch-screen versions of S60 were available long before the iPhone but were never put into production.
    * Killing of internal innovation. Numerous skunk-work replacement OS and / or graphic-layer projects were started and killed during the final years of Symbian.
    * Rewarding failure in management. Bonuses and promotions continued years after Symbian / Nokia failed to deliver anything to compare with the iPhone. Bonuses should have been tied to customer feedback compared to latest version of competing products – never just for hitting arbitrary delivery dates or shipment volumes.
    * Believing we were “too big to fail” – the idea that we dominated the market meant that we could sit on our laurels and not have to continuously innovate and delight customers.
    * Failing to build a thriving 3rd party developer community. Microsoft, Linux and many open source products showed that creating a vibrant community was possible yet somehow Symbian / Nokia never achieved this. There was an available API and websites from which free and paid-for apps could be downloaded but still it didn’t happen. Partly this was down to the seminal failure to provide developers with an alternative to descriptors and other non-standard C++ functionality that people coming from other development environments would be used to. Partly is was also, until the last year or two, not having a good quality build system or IDE (even towards the end the IDE _still_ couldn’t reliably provide on-target-debugging which “just worked” for iPhone and Android). Mostly it was not listening to and *publicly* acknowledging the feedback of 3rd party developers and putting resources in place to meet their concerns.
    * The creation of the Symbian Foundation was a case in point. Instead of creating a small, small-a-agile team dedicated to listening to developers and having solid in-roads to Nokia management to make key changes it was created from the off-set as a multi-hundred person organisation with a focus on attracting corporate level adopters.
    * Separating internal developers and managers from customers – very few, if any, developers or mid-level managers were actively provided with data about customer feedback about the OS or given an opportunity to speak to real-world customers about their feelings about the product.
    * Lack of flexibility – even though there was a worthy push towards the end to adopt “enterprise agile” it was originally ,and often continued to be the case, that product roadmap were defined for 18+ months into the future.

    * Not listening to internal feedback. I’m very firmly of the opinion that many, if not all, of the failings of Symbian were well understood by the internal development community and the lower part of the middle management structure. So much could have been achieved by simply implementing a Kaizen-like improvement system *tied to management bonuses* so that decision makers would be forced to listen to the complaints and suggestions outside of what was increasingly the echo-chamber of senior management.

    A few more than 3 there ;-). Feel free to edit away David…

    Comment by cholten99 — 23 December 2012 @ 1:59 am

    • The developer issue was deep seated from the beginning of Symbian. Essentially, the licensees wanted to build a feature phone with more features and thats what Symbian allowed them to do. They didnt really care about explaining to users that it was an apps platform, nor did they care much about external developers either. The internal dev cost of software for the phones was small to begin with I believe, but as it began to grow they continued to do do things in a hand-made way, and did not modernise the process along the way: witness the vast number of contractors which could be paid to continue with the old methods, paid for by the success of earlier devices, until that success evaporated. So in the end it was like a hand-made car for the 21st century, and pouring in resources to continue in that vein was doomed…

      Overall I think that Symbian allowed the licensees to just take the next step of building their dinosaur featurephones just big enough to herald the real mass-market smartphone era, but then the IT industry simply took its Moore’s law technology and software development processes to the winning post.

      Comment by AH — 23 December 2012 @ 1:37 pm

  6. David,

    Here’s my list as an outsider and a historian — someone with not more than a few days’ worth of first-hand observation.

    Hits:
    * To show people what a smartphone was
    * To come up with a good general-purpose architecture that met the needs of hundreds of millions of mobile phone users for years
    * To set up a governance system that (for a while) allowed multiple rivals to come to the table and cooperate

    Misses:
    * Ignoring the challenge of the web browser and the touch screen for too long
    * Failing to appreciate (or act upon) how the UIs separated Symbian from end users, handsets, ISVs and control over the direction of the platform, while introducing an extra layer of delay, coordination, inertia or bureaucracy into any effort to move the platform forward
    * Failing to address Nokia’s de facto control after 2005 or 2006. If it was just going to be a Nokia (or Nokia+DoCoMo) platform, it could have been run more efficiently through merger of S60 and Symbian groups. If it was not going to be Nokia’s private sandbox, then there needed to be more winners from Symbian besides Nokia.

    Joel

    Comment by Joel West — 23 December 2012 @ 11:26 pm

  7. Thanks to everyone for their thoughtful replies. There’s lots of good points in these comments. It’s sobering reading.

    I’ve decided against providing detailed responses to individual points for the moment, but will be covering a considerable number of the themes raised in the chapters of “More than smartphones”. Unless someone requests anonymity, I may list their names in the acknowledgements.

    Comment by David Wood — 24 December 2012 @ 12:15 am

  8. Hits: Committed Engineers with a passion for the product, along with long serving staff from Psion days striving to dleiver the best prodcut they could

    Misses: Looking at Itunes and AppStore now our third party app story was simply rubbish, if we could have got that story right I htink the products would have been selling in 10s of millions still today. Would echo the whole underhand incentive bonus scheme that some peoiple were on, others not divided a workforce more than managment ever knew or were willing to accept (some of the figures I heard were simply obscene for those near the top). Also the model of delivering a faulty product and charging vast amounts of cash to have consultants fix it, appreciate this was absolutly the right way to go in the early days but not later on.

    I’m coming from a less technical perception as non engineer of course so may not have as valid points!

    I still recall late 1998 asking you late one night in Frampton St for an extra 250mb hard disc so I could do OEM SW releases of ER5 to Nokia and Ericsson ;-), how the world changed in the 12 years I was there.

    Ade

    Comment by Ade Steward — 24 December 2012 @ 3:20 am

    • Hi Ade – thanks for the input

      >I’m coming from a less technical perception as non engineer of course so may not have as valid points!

      Non-technical points are at least as relevant, for Symbian’s success and failure (and for so much else in life), as technical points

      >I still recall late 1998 asking you late one night in Frampton St for an extra 250mb hard disc so I could do OEM SW releases of ER5 to Nokia and Ericsson

      I hope I said ‘Yes’…

      >how the world changed in the 12 years I was there

      Indeed, but some of the principles remain the same, and are as applicable today as always

      Comment by David Wood — 24 December 2012 @ 12:25 pm

  9. David,

    In working on my thoughts about the Symbian outcome, I had one more observation/question: How much of this was an unfortunate path dependence?

    After about 2005, really only Nokia was successful with the Symbian platform. But what if 3 companies had been successful? Suppose in 2006 that there were 60 million Symbian handsets sold (instead of 52m) and were split 30m Nokia, 10m Sony Ericsson, 10m Motorola and 10m FOMA.

    If Motorola were selling 10m handsets, most of those would have been in the US, taking share from Windows, RIM and what was left of Palm. With Symbian as a multi-firm success, would Samsung and LG have taken Symbian more seriously?

    Or was the ascent of the Koreans inevitable, and their switch to cheap (i.e. Google) software also inevitable?

    Joel

    Comment by Joel West — 26 December 2012 @ 5:36 pm

    • Hi Joel

      >How much of this was an unfortunate path dependence?…

      I see none of what happened with Symbian as somehow inevitable – neither the successes of the earlier years, nor the subsequent decline.Even as late as 2008-2009, things could have developed in multiple different ways. The discussions in 2008 about the formation of the Symbian Foundation involved all five of the leading phone manufacturers of the time – Nokia, Sony Ericsson, Motorola, Samsung, and LG. The story of how four of these five decided, in different ways, against serious adoption of the Symbian platform, is a fascinating one, with many twists and turns ;-)

      Comment by David Wood — 27 December 2012 @ 11:09 am

  10. Speaking as someone who watched Symbian from a distance for a long time, I’m very persuaded by Peter’s comment: The original goal of Symbian was to prevent Microsoft from dominating the mobile OS market, and it succeeded. To quote George W. Bush, “mission accomplished.”

    I think there are a couple of follow-on questions that need to be asked, and I hope you’ll tackle these, David:

    1. Were Symbian’s founding owners focused on solving the right problem? Was a negative mission (“stop Microsoft”) the best and highest use of the people and money invested in Symbian?

    2. Did success in the anti-Microsoft mission lull Symbian’s owners into a false sense of security in smartphones? (Actually, to be honest, that question is a bit of a strawman.)

    3. What lessons does the Symbian experience give us about how to structure and manage a joint venture, especially an OS one? Is the success of Android due in part to its unitary management?

    Meanwhile, here are my thoughts on the hits and misses:

    Hits:

    –The fact that you managed to assemble the consortium and make it work. That in itself is an incredible accomplishment.

    –Heavy adoption of the platform by Nokia.

    Misses:

    –Failure to produce meaningful long-term sales (outside of Japan) for any vendor other than Nokia.

    –Abdication of leadership in UI.

    –Failure/inability to produce user awareness and demand for the OS

    Comment by Michael Mace — 26 December 2012 @ 10:54 pm

    • Hi Mike – Many thanks for these points. Two quick comments here (I’ll save the rest for my book…):

      >Were Symbian’s founding owners focused on solving the right problem? Was a negative mission (“stop Microsoft”) the best and highest use of the people and money invested in Symbian?

      I have plenty of documentation that the negative mission “stop Microsoft” was far from being the sole main driver behind the formation of Symbian.

      >Failure/inability to produce user awareness and demand for the OS

      Indeed – that was a real sore point. Many people held the view that the name of the OS in smartphones would never become a matter of wide public discussion (after all, which customers care about the name of the OS that runs their TV, or their motor car…?) but the rise and rise of public awareness of Android and iOS disproves that view

      Comment by David Wood — 27 December 2012 @ 11:21 am

  11. Dear David,
    First of all thanks for taking up this project. I am sure that sharing of this information by someone who has seen the Symbian from very start till the very end – will provide the right context and perspective which an outsider might miss. At the same time I hope it will have a good amount learning for future leaders.

    My involvement with Symbian was from 2007 till 2011. This time was peak for Symbian (with N95 & NTT Phones ruling the roost). However, this was prbably also the time which you can now pinpoint as being the turning points for Symbian.

    Hits
    1. Balance which Symbian was able to acheive between the quality expectation from Japanese (post 9.2 release) and feature expectations from Nokia. Those two customers were poles apart and I guess Symbian leadership very good job of setting and meeting expectations of both parties.
    2. Multi-Media quality & Feature which is clearly put the smart phone in different category.
    3. Strength of Symbian C++. It can be debated that why C++ and if this was a major road block to get eco-system going for Symbian. I know that in Bangalore if you know Symbian C++ you would be considered as good programmers (everyone liked to show off their ASD certificate).

    In summary Symbian was a great SW development house with excellent engineers who could execute efficiently any tasks. It also had good leadership which was able to balance the pressures of its diverse stakeholders.

    Misses
    1. For me failure in building a win-win eco-system by connecting with young entrepreneurs across the globe was a big miss. We were not able to communicate a business model which will excite these young entrepreneurs to innovate on Symbian platform. Most did b’coz Symbian was the only smartphone platform to startwith. But once you had option of Appstore and Android store – Symbian started losing. This was also compunded by the fact that Symbian could not penetrate into US market which has the largest population of early adopters.
    I don’t think we ever thought of app store as with in the Symbian charter. We left it to the OEMs who (especially Nokia who still were selling smartphone more as a HW device). We had great multimedia support but we did not provide the easy way of sharing the media (again left it to OEMs).

    2. Lack of UI ownership – Symbian was not a usual OS like Windows. It was build bottom up for Mobiles and thus UI should have been a integrated part of it. Looking back Symbian should be pushed harder to get S60 and UIQ under its aegis. We gave up UIQ too early. This would have benefited OEMs as well. Cycle time to get new Symbian based phone out by an OEM was much longer b’coz of this (partly due to Nokia development processes).
    This also meant that Symbian was not doing any innovations in the area of user experience. Nokia S60 was a classic case of if it ain’t broke – don’t fix it.

    3. Pure lack of innovations – Symbian restricted itself to areas where no out of box ideas were coming in. All work was with a OS domain as defined by our customers. Innovation meant that we would be getting into the turf of our customers which was probably a big NO – NO.

    In summary, Symbian probably never had a charter to be an innovative technology driven company. It was setup for a specific purpose and it was restricted to that purpose only. Innovation in adjacent areas – like app stores, content management, user experiences were not seen as part of its charter.

    Cheers,
    Anirudh.

    Comment by Anirudh Mathuria — 27 December 2012 @ 8:24 am

    • Hi Anirudh – these are all good points, thanks

      >Balance which Symbian was able to achieve between the quality expectation from Japanese (post 9.2 release) and feature expectations from Nokia

      Yes, that’s a fascinating dialectic that isn’t widely appreciated. Both the Japanese and Nokia benefited from it

      >Innovation in adjacent areas – like app stores

      I’m still searching in my archives for some trace of a proposal I remember making to the Symbian Leadership Team around 2005 that Symbian should take the lead in creating an app store and gaining a new line of revenue from that. Alas, although the idea was good, we failed to follow through with it – roughly for the reasons you state.

      Comment by David Wood — 27 December 2012 @ 11:26 am

  12. David,

    Look forward to reading your thoughts in full (long form) detail.

    To answer your specific question:

    What would be in your list (say, 1-3 items) of the top hits and misses of decisions made by Symbian?

    Top Hits:
    1. Creating an open platform – that allowed for the creation and installation of 3rd party applications together with the SDKs and developer ecosystem around it
    2. Creation of a robust, full featured software stack that ran on low power devices
    3. Creating an awesome touch UI platform (Quartz) that delivered years before Android/iPhone

    Top Misses:
    1. Failure to understand how to grow and extend the 3rd party ecosystem. Symbian repeatedly made decisions that were “right” from the perspective of the internal SW architecture teams, but required work by 3rd party developers to simply maintain their apps. All the long term OS ecosystems that have been successful have avoided this (Windows, MacOS, iOS, Android, PlaystationOS, Gameboy OS etc…). In order that I remember them there were:
    i. Compatibility break from ER5 -> Unicode
    ii. UI level API fragmentation from EIKON -> UIKON -> CKON, QIKON, AVKON, WHO-KNOWS-WHATKON
    iii. Security framework BC break

    In each case they required app developers to rebuild their apps (with varying degrees of simple/complex changes) – contrast the transition for an iOS developer who has gone from iPhone (320×480 screen) -> iPhone (640×960) -> iPad (1024×768) -> iPad (2048×1536) -> iPhone (640×1136) – an app written for iOS 3.x will still run on an iOS6 app on an iPhone 5 or iPad….. Sure, if you rebuild you can make it better, but it still runs.

    2. Choosing C++ as the basis for an OS. Binary compatibility is really hard to do in C++ (see above) with hindsight maintaining the C based OO solution used by EPOC16 would have probably resulted in an easier to maintain, and port to, and been more posix like than a C++ OS, and could have continued to allow for C++ development against the base OS API’s should app developers have wanted to use a more modern compiler

    3. Not prioritising comms stacks early enough – TCP/IP was a downloadable add-on for EPOC ER5. That’s horrific to think of now. Comms started too late and was a mixture of superb components, and very poor elements. I’m going to call out the TSY stack as a particular “hideous mistake” – the lack of re-usable, easy to extend components at the core telephony interface was a massive mistake – and caused huge product delays and cost in creation of functioning phone stacks. There was no real reason why an “out of the box” Symbian OS shouldn’t have “just worked” for making phone calls and data connections on a standard AT-command interface to a phone chipset – with additional work being needed only where optimisations/new functionality are required.

    Comment by Matt Millar — 3 January 2013 @ 4:15 pm

    • Matt – These are really great comments, that neatly complement (and in some cases extend) the ideas I am already working on.

      Many thanks for chipping into the discussion!

      Comment by David Wood — 3 January 2013 @ 11:52 pm

  13. Here is my perspective from that of a long-time 3rd party developer (since 2003 or so) and Nokia watcher (ca. 1996). [*]

    Hits:

    1. Providing a solid design base that would allow for very robust multi-threaded systems over many years. IF you were willing to do Active Objects right, that is.

    2. Managing to adapt to many changes in smartphone technology over a decade (packet data, Qwerty vs. 12-key, global reach, and in the outyears of Symbian^3 even touch screens) by continously evolving a single product.

    3. Running on 100s of millions of devices – I think the massive installed base that was achieved by building on Symbian is something that is often underestimated as a success.

    Misses:

    1. Allowing the separation of the “foundation” OS and the UI and middleware layers to become permanent, with very different standards of documentation and API quality in both, and resulting problems for developers what it came to separating “intended” vs. “actual” behaviour.

    2. When it was recognized that “UI was king”, having too many balls in the air at the same time, and pulling away the hand just as each of them was coming down. I am still staggered by the number of times I saw things that look and behave exactly like CEikList objects in different UI frameworks.

    3. Underestimating the role of having a US market presence at any cost for gaining mind-, if not marketshare.

    4. (Bonus:) Making it too difficult for external contributors to take advantage the Open Source-ness of Symbian without overcoming massive entry barriers in building the system, or parts of it, first.

    Actually, I think that the failure of adoption of the massive Symbian codebase in the OS community is worth of a case study of its own. Afaik, there isn’t even a Sourceforge project to keep track of any contributions made to the code after it became “closed” again, which to me is a big indication of how few people are actually missing it. In a sense, the problems with allowing external developers to contribute to the S60WebKit browser by Nokia was a harbinger of things to come… I sometimes come back to this thread: http://www.developer.nokia.com/Community/Discussion/showthread.php?109899 – and wonder where it all went wrong.

    ciao marcus

    [*] Mainly as the author of the “TALKS” Screen Reader for the visually imapired, which provided some interesting perspective on the internals of the system while mostly ignoring many pains of “typical” app developers.

    Comment by mgroeber — 4 January 2013 @ 12:42 pm

    • Hi Marcus – Many thanks for your input. You provide lots to ponder.

      You’re surely right that the open-source angle to the Symbian codebase deserves a case study of its own. That topic could easily fill several chapters by itself!

      Comment by David Wood — 4 January 2013 @ 12:47 pm

  14. hits :
    – the OS was more reliable than other mass deployed / commonly used OS’s (not just micro-kernel OS’s, Windows etc etc) – typical reboot period measured in months, not hours!
    – it delivered smartphone functionality to the world, ie, it opened up peoples minds and delivered “star trek” dreams to those who knew how to get it, and could afford it.
    – it started the evolution of “touchscreen” in a usable form factor (irrespective of final status, my S.E. R380 changed my conception and reality of UI, and I know it also did for many people I know).

    misses :
    – [marketing] lack of customer awareness of what they had, the vast majority of Symbian OS users had no idea what really powered their devices
    – [market] not enough penetration in the USA – success in the biggest economy / consumer of technology is the world is a must, otherwise they “develop their own”
    – [lack of user app's] Symbian C++ was simply too difficult to develop app’s upon for the average programmer; without which the critical mass of app’s failed to materialise, without intrinsic user adoption (ie number of different app’s embedded in daily lives) the platform wasn’t ever going to survive technological and market evolution; despite the number of handsets in use, most Symbian devices remained “a phone” in the minds of most end users.

    if only :
    – the handset manufacturers and carriers had some form of “shared vision” (see marketing)
    – a carrier (or two) had allowed (or been forced into) “unlimited data plans” for Symbian OS users…

    James

    Comment by James Booth — 4 February 2013 @ 10:44 am


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

The Silver is the New Black Theme. Blog at WordPress.com.

Follow

Get every new post delivered to your Inbox.

Join 104 other followers

%d bloggers like this: