[Infographic] The Open Governance Index – A new way of measuring openness

We are proud to present our latest infographic – the Open Governance Index, measuring the relative openness of 8 major open source projects, from Android to WebKit. This infographic presents some highlights from our full report (free download here). The Open Governance Index is authored and researched by VisionMobile, and part-funded by webinos

Feel free to copy the infographic and embed it in your website (embed codes below the infographic).

Infographic- The Open Governance Index - A new way of measuring openness
150 pixels wide version

500 pixels wide version

800 pixels wide version

[sociable_code]

[Report] A new way of measuring Openness, from Android to WebKit: The Open Governance Index [Updated]

[Much has been said about open source projects – and open source platforms are now powering an ever-increasing share of the mobile market. But what is “open” and how can you measure openness? As part of our new research report (free download), VisionMobile Research Partner Liz Laffan introduces the Open Governance Index – a new approach to measuring the “openness” of software projects, from Android to WebKit]

Update: We have been amazed by the amount of interest to our Open Governance Index (OGI) report that we published just over two weeks ago. Our report was covered in mainstream media across Wired, ZDnet, PCPro, Gizmodo, ARS Technica, BGR, Zeit Online and ReadWrite Mobile. Our intention was to start a debate around ‘what’ openness is, ‘how’ it can be measured and ‘why’ it is important – and we certainly got the ball rolling!

Open Governance Index cover

Openness = governance

We at VisionMobile have been researching, investigating and helping to educate the industry about open source for the past five years.  In this time open source software has been transformed from geekware to business as usual. Much has been written and debated regarding open source licenses – from the early days of the GPL license to the modern days of the Android platform.

Despite the widespread use of open source, from Android to WebKit, there is one very important aspect that has been neglected: openness and how to measure it.

Openness goes far beyond the open source license terms and into what is termed Governance. While licenses determine the rights to use, copy and modify, governance determines the right to gain visibility, to influence and to create derivatives of a project, whether in the form of spin-offs, applications or devices. And while licenses apply to the source code, governance applies to the project or platform.  More importantly, the governance model describes the control points used in an open source project like Android, Qt or WebKit, and is a key determinant in the success or failure of a platform.

VisionMobile - Licensing vs. Governance Models

The governance model used by an open source project encapsulates all the hard questions. Who decides on the project roadmap? How transparent are the decision-making processes? Can anyone follow the discussions and meetings taking place in the community? Can anyone create derivatives based on the project? What compliance requirements are there for creating derivative spin-offs, applications or devices, and how are these requirements enforced?  It is governance that determines who has influence and control over the project or platform – beyond what is legally required in the open source license.

In today’s world of commercially-led mobile open source projects, it is not enough to understand the open source license used by a project. It is the governance model that makes the difference between an “open” and a “closed” project.

Measuring openness

Our research (free copy of full report here) showcases eight mobile open source projects: Android, MeeGo, Linux, Qt, WebKit, Mozilla, Eclipse and Symbian.  We selected these projects based on breadth of coverage; we picked both successful (Android) and unsuccessful projects (Symbian); both single-sponsor (Qt) and multi-sponsor projects (Eclipse); and both projects based on meritocracy (Linux) and membership status (Eclipse).

All of these are open source projects, whether platforms (Android, MeeGo, Qt, Symbian) or engines (Linux kernel, WebKit) or multi-project initiatives with a single, uniform governance. We appreciate that these projects are unique in many ways but they are all ultimately open source projects and to that extent our governance measures can be applied to them all equally. For example all of these projects have decision-making groups and processes that are directly comparable. In the Open Governance Index we attempted to document who these decision-makers are, how they operate, what processes are used to determine project decisions and how easily is to influence these project decisions.

Our research, carried out over a six-month period, included analysis of these popular open source projects, through discussions with community leaders, project representatives, academics and open source scholars. This research was partially funded by webinos, an EU-funded project under the EU FP7 programme, aiming to deliver a platform for web applications across mobile, PC, home media (TV) and in-car devices.

We quantified governance by introducing the Open Governance Index, a measure of open source project “openness”. The Index comprises thirteen metrics across the four areas of governance:

1. Access: availability of the latest source code, developer support mechanisms, public roadmap, and transparency of decision-making
2. Development: the ability of developers to influence the content and direction of the project
3. Derivatives: the ability for developers to create and distribute derivatives of the source code in the form of spin-off projects, handsets or applications.
4. Community: a community structure that does not discriminate between developers

The Open Governance Index quantifies a project’s openness, in terms of transparency, decision-making, reuse and community structure.

Does openness warrant success?

But what is it that makes an open source project successful? Why do some projects become an immediate success, while others barely get off the ground before crashing and burning? We know that just like commercial ventures, open source projects have different cultures and drivers – but we do believe that you should be able to measure the way that open source projects interact with the community of users and contributors that they build up around themselves.

Our research suggests that platforms that are most open will be most successful in the long-term. Eclipse, Linux, WebKit and Mozilla each testify to this.  In terms of openness, Eclipse is by far the most open platform across access, development, derivatives and community attributes of governance.  It is closely followed by Linux and WebKit, and then Mozilla, MeeGo, Symbian and Qt. Seven of the eight platforms reviewed fell within 30 percentage points of each other in the Open Governance Index.

Moreover, our research identified certain attributes that successful open source projects have.  These attributes are timely access to source code, strong developer tools, process transparency, accessibility to contributing code, and accessibility to becoming a committer.  Equal and fair treatment of developers – “meritocracy” – has become the norm, and is expected by developers with regard to their involvement in open source projects.

The Android Paradox

We found Android to be the most “closed” open source project. In the Open Governance Index, Android scores low with regard to timely access to source code in that the platform does not provide source code to all developers at the same time; it clearly prioritises access to specific developer groups or organisations and has acknowledged this with the delayed release of Honeycomb. Additionally Android scores low with regard to access to developer support mechanisms, publicly available roadmap, transparent decision-making processes, transparency of code contributions process, accessibility to become a committer (in that external parties cannot ‘commit’ code to the project) and constraints regarding go-to-market channels.

Android ranks as the most closed project, with an Open Governance Index of 23%, yet at the same time is one of the most successful projects in the history of open source. Is Android proof that open governance is not needed to warrant success in an open source project?

Android’s success may have little to do with the open source licensing of its public codebase. Android would not have risen to its current ubiquity were it not for Google’s financial muscle and famed engineering team. More importantly, Google has made Android available at zero cost, since Google’s core business is not software or search, but driving eyeballs to ads. As is now well understood, Google’s strategy has been to subsidise Android such that it can deliver cheap handsets and low-cost wireless Internet access in order to drive more eyeballs to Google’s ad inventory.

Equally importantly, Android would not have risen were it not for the billions of dollars that OEMs and network operators poured into Android in order to compete with Apple’s iconic devices. As Stephen Elop, Nokia’s CEO, said in June,2011, “Apple created the conditions necessary for Android”.

Moreover, our findings suggest that Android would be successful regardless of whether it is an open source project or not, to the extent that the vast majority of developers working on the project (the platform itself) are actually Google employees.

 Evolving the Open Governance Index

Having published the report, we aim to continue the discussion on governance, to refine our criteria even further and to make the OGI measure as meaningful as possible for the open source community. One of the first suggestions has been with regard to having a time dimension to the criteria i.e. does openness change over time. Mature open source projects such as Eclipse, Linux and WebKit that have stood the test of time, score quite highly with regard to openness of governance. But this has not always been the case. For example consider the following. Apple forked KHTML to create WebKit in the early 2000’s, releasing the first WebKit open source project in 2005 but with reviewer and commit rights restricted to Apple personnel only which effectively sidelined the KDE community. In 2007 however Apple reversed this decision allowing allow non-Apple developers to have full commit access to the WebKit source code version control system. This shows that openness can and does change over the project lifecycle.

Our vision for the Open Governance Index is to for it to be a robust, and as much as is possible, an objective measure of Governance for open source projects. We believe that this is necessary such that users and contributors to open source projects, including commercial entities, understand the means by which they can, or cannot, influence the direction and content of the project.

Download the full report for an in-depth analysis of the openness of Android, MeeGo, Linux, Qt, WebKit, Mozilla, Eclipse and Symbian. Drop us a line and tell us what you think.

– Liz

Addendum – Is copyleft more or less open?

We awarded a higher score to those licenses that are permissive and not copyleft licenses. Firstly it should be noted that all the licenses used by the eight mobile open source projects are Open Source Initiative (OSI) approved and meet the Open Source Initiative Definition, which provides for free redistribution of source code, access to source code and ability to create derived works amongst other requirements. We believe that the OSI is the appropriate arbiter of the appropriate Open Source License definition and all of the licenses used by the open source projects researched in this report meet this definition of being ‘open’.

However we also believe that from a commercial viewpoint there is still some concern about using code that is under a copyleft license – our experience of working with mobile software development organisations confirms this. Our findings suggest that organisations will be more comfortable using permissive licenses which do not mandate copyleft requirements and we reflect this in our criteria and scoring. We are happy to continue debating these findings further with the community. For example it has been suggested that the problem here is not with copyleft licenses but with the business model used by those organisations. Be that as it may, our experience is that this concern is still a valid one being expressed by many organisations, especially in the mobile device domain.

Finally we had a methodology typo which unfortunately survived the proof reading: assigning a bonus to “copyright assignment”. We fully acknowledge that copyright assignment is unnecessary – indeed we state this in our analysis of Qt whereby we acknowledge copyright assignment as inappropriate and a heavy-handed requirement.

[Liz Laffan is a Research Partner at VisionMobile. Liz has been working in the telecoms and mobile industry for over 20 years, with large telco organisations, start-up technology ventures, software development and licensing firms.  Liz’s interests lie in open source software governance and licensing and in particular how best can commercial organisations interact with open source projects.  She can be reached at liz [at] visionmobile.com]

Is Microsoft buying Nokia? An analysis of the acquisition endgame

In a surprising move, Nokia and Microsoft decided to enter a strategic relationship for the OEM’s smartphone business. While the marriage appears promising at the outset, Research Director Andreas Constantinou argues that the only way for that marriage to succeed is for Microsoft to acquire Nokia’s smartphone business.

VisionMobile - Nokia & Microsoft deal_pic

The Elop and Ballmer duo on stage on February 11th was the main topic of discussion at this year’s Mobile World Congress. The reverberations of the Microsoft-Nokia announcement were felt even by the huge green robot tucked away at Google’s stand in Hall 8.

Following the news of the Nokia and Microsoft tie-up, Stephen Elop’s appointment to the helm of Nokia seems like an arranged marriage – and one whose best men were the carriers who wanted to avoid an all-out Android coup. It was also a marriage of desperation, which Elop memorably described in his memo as ‘jumping into the unknown’ from the ‘burning platform’ that is Symbian.

A marriage of desperation
Microsoft has been desperate to see its mobile business succeed. After a decade of lacklustre efforts at mobile device sales and severe product delays, Microsoft was getting desperate; it needed to stop the churn of Microsoft users to the Apple ecosystem and plug its $1 billion-a-year operational costs for its mobile phone business. Even having spent most of its $500M marketing budget for WP7 it had only got breadcrumbs in terms of sales, with Microsoft reporting 2 million shipments but no comment on sell-throughs (which leads us to suspect this was not more than 1 million of actual end-user sales).

VisionMobile_Smartphone_Sales_2010_pic

Nokia has been desperate seeing its platform play fail spectacularly in comparison to its newfound competitors; Apple who had amassed a developer ecosystem and operator demand which was second to none, and Android who in 2 short years matched Nokia’s smartphone sales in Q4 2010. MeeGo was trumpeted as the big guns in Nokia’s arsenal in February 2010, but once again Nokia’s software R&D failed to deliver on the promise. More importantly, despite the 10+ acquisitions during 2007-2010, Nokia failed to amass a strong-enough developer and services ecosystem on Symbian, Java or Qt that could compete with Apple or Google. Like Elop said in his now-famous burning platform memo, “our competitors aren’t taking our market share with devices; they are taking our market share with an entire ecosystem”.

It was in an act of desperation that led Nokia to befriend the lesser of two evils in the shape of Microsoft. It is ironic how in mobile the least enemy is a friend, much like how carriers backed Android in 2008-9 to fend off Apple, and backed Microsoft in 2003-5 to fend off Nokia.

The courtship
Despite the surface-level coverage of the Microsoft and Nokia news, not much has been said about the two giants’ courtship and even less on the prenuptial agreement. According to our sources, Nokia asked both Microsoft and Google to bid for its smartphone business, with the help of a small army of McKinsey suits. Following a long negotiation cycle with both parties, Nokia came to a straightforward conclusion; it would back Microsoft who’s total bid equalled more than $1 billion (including patents, licensing fees, marketing support and revenue shares) and not Google who’s bid was about half that. Funny how cash-rich platform vendors are buying their way into the market these days.

Nokia announced its decision to Microsoft and Google on February 9th , only 2 short days before the Ballmer/Elop press conference – which prompted Vic Gundotra to pen the tongue-in-cheek tweet “#feb11 “Two turkeys do not make an Eagle”, scornful of both Nokia and Microsoft.

The last-minute decision meant that Intel heard the news at the very last minute, and in turn had to ask its MeeGo partners on Friday night (Feb 11th) to remove the mention of Nokia from the MeeGo PR quotes going out on the following week at MWC. This is the stuff industry disruptions are made of.

 

A chemistry mismatch
What Nokia announced was not just a marriage; it was a radical change in its business model, from a vertical powerhouse to an assembler – which is what prompted us to question the motivations and the end goal for Elop.

We already knew that Symbian had been demoted to an internal-only OS (see earlier analysis – Symbian is dead, long live Symbian). However we were expecting to see Nokia take a more measured stance; for example using Windows Phone 7 in certain markets (especially in North America where carrier handset subsidies are OS-led) or taking a classic dual-supplier strategy by inking deals with both Microsoft and Google.

Instead Elop presented a terminal picture for Symbian which would be destined to ship on only another 150 million devices until being completely replaced by WP7.  Elop knew that an all-out replacement of Symbian with WP7 would mean haemorrhaging valuable brainpower as the 7,000+ Symbian staff had spent 15 years on the anti-Microsoft camp. These are the decisions made by boards with long-term strategy agendas, who see organisations made up of ‘assets’ and not ‘people’.

Besides the death blow to Symbian, Elop relegated MeeGo to an R&D project with just a single device launch in the horizon, if any at all (which carrier is going to subsidise a platform that’s dead on arrival?). Moreover Qt’s future seems uncertain as it has no place on Windows Phone (Microsoft wouldn’t allow copyleft software to be used with Windows Phone), plus it is too heavy for S40 class devices and MeeGo is too small an addressable market to justify the Qt ongoing investment. Qt (and its 400 thousand developers) need a new home.

Nokia Mobile Devices Net Sales Mix

What appears somewhat suspicious is that Nokia went not for a tactical, but a deep partnership with Microsoft, solidified by the multiple revenue streams exchanged between the two companies, a kind of revenue ‘keiretsu’ that ties the two giants in a longer commitment.

More importantly, the marriage to Nokia’s smartphone business seems like it’s lacking in chemistry. For the last decade, Nokia has operated as a vertical silo, owning and integrating all value elements, from software, UI, industrial design, services, app store and developer ecosystem. That silo has now huge holes punched through so that it can accommodate Microsoft’s horizontal software-licensing business model. This situation is somewhat like trying to fit a square peg in a round hole.

There are fundamental conflicts here, as both Microsoft and Nokia want to own the developer experience (think APIs, support, tools, developer marketplace, conferences, marketing), and the application discovery and delivery process (think Windows Marketplace vs Ovi Store). This is a chemically unstable mix that won’t survive the test of time. It would be like having Nokia owning Office while Microsoft still runs the Windows business. Yet at the same time Nokia has little value to offer other than design, development, manufacturing and sales of handsets in the picture Elop and Ballmer painted. Something’s not right.

Moreover, Microsoft faces a fundamental customer imbalance on its mobile platform. With such a strong endorsement of Nokia, Microsoft has placed too much favour and device sales expectations on a single vendor.

Microsoft did not only hurt the feelings of HTC, Samsung and LG (previously committed to launching 50! Windows Mobile handsets) with such an imbalanced endorsement. More importantly, with Nokia volumes likely to ramp up fast, Microsoft will have to deal with a single-customer monopoly and end up financing Samsung, LG or HTC towards ramping up Windows Phone production to balance it up. Windows Phone may quickly end up looking like a platform of unbalanced OEM interests – much like Symbian Ltd or Symbian Foundation were – and we know how these panned out.

There are two more troubling clues in the way this ceremony was setup. Despite fundamental changes to the handset business, Elop made no reorganization in the NSN business which is performing at marginal profit (operating margin at only 3.7% vs 11.3% for handsets). As Tomi Ahonen points out, Elop seems to be ready to get rid of NSN. Plus there was no announcement of Ovi plans or clear strategic guidance with regards to the Nokia services business.

 

The acquisition scenario
There have been earlier rumours of acquisition discussions between the two companies. We now believe that the only scenario for the Nokia and Microsoft partnership to succeed is an acquisition scenario; Microsoft buys Nokia’s smartphone business, while Nokia gets more resource to play with what it does best – that is creating mass-market phones at unbeatable levels of supply chain efficiency, unmatched supplier bargaining power and customisation to 100s of variants per handset model for distribution to diverse global regions, channels, carriers and retailers.

From a financial standpoint, Microsoft capitalisation stands at $220B, more than six times Nokia’s market cap of $33B at the time of writing. Microsoft would also acquire a high-profit margin business that would go a long way in helping the Redmond giant push its Entertainment and Devices division at high profitability levels for the first time. Despite Microsoft being a software business, it has experience in running hardware products, with the Xbox business doing well recently on strong Kinect sales.

For Nokia, a joint venture would make more sense than a pure sale of its smartphone business, given that the hardware giant is an important component of the Finnish economy. It would allow Nokia to focus on what it does best and substantially increase its S40 R&D budget (as Elop already announced it would) to rework its aging feature-phone OS. A joint venture would also allow Nokia to make a comeback when they are ready to take on the high-end phone market again.

Besides, with shares recently hitting a 13-year low and Nokia being owned by American institutional investors, the Nokia board has little they can do in the face of potential suitors. This makes Nokia a very interesting acquisition target, not just for Microsoft but for anyone with cash at hand and mobile ambitions, including Chinese, Korean and Japanese suitors.

The acquisition scenario would allow Microsoft to leverage on Nokia’s accounts with carriers across the world to woo them into moving subsidy budgets from Android into WP7. This is all too important, as the Microsoft brand enjoys little consumer awareness compared to Apple and Android, meaning that Microsoft is more dependent on carrier subsidy and marketing budgets than its nearest competitors.

Fundamentally, we believe there is no place for Nokia, an all-in-one integrated handset OEM and services company, in the new telecoms value chain. The old guard of top-5 OEMs are squeezed between leaders (Apple, RIM) who lead in terms of performance & profits, and assemblers (Huawei, ZTE, Dell, Acer) who lead in terms of me-too designs & razor-thin margins (see our earlier analysis on the evolution of the handset value pyramid). Nokia’s business needs to break-up into independent, self-sustained entities, particularly the smartphone business (within Microsoft’s new home) and the mobile phone business as an independent entity that can focus on competing with PC-borne assemblers.

The Microsoft-Nokia acquisition might not have been planned from the outset, but it is a scenario whose viability has been ensured from the outset. There are no conspiracy theories here, except that Elop (as the 7th biggest shareholder of Microsoft) would benefit greatly from trading Microsoft shares with Nokia ones, only to see them boost in value after being repatriated.

Let the debate begin!

– Andreas
you should follow me on Twitter: @andreascon

Andreas Constantinou is Research Director at VisionMobile and has been working in the mobile software industry since 2001, when he fondly recalls being a member of the team behind the very first Orange-Microsoft handsets which set the world of telecoms software in motion.

The MeeGo Progress Report: A+ or D-?

[Eight months after the announcement of the MeeGo  project by Intel and Nokia, guest author Dave Neary analyses the progress made to date in MeeGo Handset, and the project’s prospects for the future]

VisionMobile - The MeeGo progress report

The end of October saw the release of MeeGo 1.1, the second major milestone release of the platform since it burst onto the scenes in February 2010. The MeeGo project was born under the auspices of the Linux Foundation from a merging of Nokia’s Maemo platform, targeting smart phones, and Intel’s moblin platforms, aimed at netbooks.

[poll id=6]

The merger grew from a core idea: pick the best of breed components from both stacks, collaborate on the integration and testing of shared components, and standardise a number of open source UX (User eXperience) profiles, on which vendors could build and deploy complete commercial grade stacks. The initial UX profiles announced were netbook, smartphone, IVI (In-Vehicle Interface) and media center/TV.

Nokia and Intel have both made a major commitment to the platform, but critics say that the relationship is little more than a marriage of convenience. After all, Intel is a silicon vendor, betting heavily on the Atom-based Moorestown platform, and Nokia is a handset designer, largely shipping ARM-based devices.

Growing pains

The project has had some teething problems. Troubled Nokia has changed CEO, and the founding father of the Maemo project, Ari Jaaksi, has been among a number of high level software executives to leave the company, leading some to ask whether Nokia might have a change of heart about the platform. The first MeeGo device for Nokia, originally expected at the end of 2010, will now appear in 2011, according to recent comments from new CEO Stephen Elop, as Nokia strive to ensure a good first impression for its first MeeGo device.

There are some early public signs of friction in the working relationship of the stakeholders in the project, also.

The adoption of Qt as the primary toolkit for both platform and applications has met with resistance from Intel engineers, who acquired Clutter in 2008 and integrated it heavily into the netbook user interface, plus partners like Novell who developed versions of GTK+ applications like the Evolution email client and Banshee music player specifically for the netbook form factor.

Long-awaited MeeGo compliance specifications have resulted in drawn out and sometimes acrimonious debate.  Trademark guidelines have been a sticking point for community ports of the MeeGo netbook UX to Linux when these ports do not include required core components.

Related to the technical governance of the project, there is some uncertainty around the release process, and the means and criteria which will be used when considering the inclusion of new components. And there are some signs that the “all open, all the time” message at the project launch has been tempered by the reality of building a commercial device.

The Promise of Openness

Many of these issues are to be expected when merging two projects into one and pairing two very different animals. Every open source project has its own culture, and Moblin and Maemo are no different. Relationship capital which participants built up in the contributing projects must now be rebuilt within a broader group.

MeeGo has had some early successes. MeeGo 1.0, which included the Netbook UX and an early prerelease of the smartphone UX, was delivered in July, complete with the source code of a number of components which had previously been proprietary. Novell MeeGo has been shipping on a number of netbooks since then. The MeeGo wiki lists dozens of MeeGo-compatible devices. The inaugural MeeGo Conference is set to take place in Dublin, from the 15th to the 17th of November, and has sold out with over 600 registered attendees to date.

And there is no denying that the companies involved in the project are committed to it. With the recent rumours that the Symbian Foundation may be shutting up shop, Nokia has few choices of platform left for upcoming high-end devices. Announcing their updated software strategy during their quarterly results call this month, the company confirmed that they are fully committed to MeeGo as the only platform for high end devices from now on.

Clearly, there is a future for the project. The question is, how will MeeGo Handset hold up against the competition from the platforms with the most momentum in the market – iOS and Android, or the recently released Windows Mobile 7. Will a newly reinvigorated WebOS (with Ari Jaaksi at the helm) challenge it for the mantle of the exciting new upstart? In short, is it any good? And will operators, handset manufacturers, application developers and users adopt it?

User experience

Since we do not yet have a MeeGo handset device available, it is very difficult to accurately judge the user experience at this time. It is possible to install MeeGo on the Nokia N900 and use it as a phone, using Nokia’s proprietary drivers to enable the hardware, but a lot of basic functionality is missing at present. In my tests, the camera, GPS, battery indicator, network signal strength indicator and WiFi did not work correctly. Features which do work can be slow, or have stability issues. Basic functionality like reading contact details off a SIM card, or unlocking the SIM card on boot, are still missing.

A MeeGo device getting to the market will undoubtedly have pristine hardware integration using 3rd party drivers, and a considerable amount of fit-and-finish which the basic MeeGo stack does not yet have.

The MeeGo handset user experience is still in transition. Maemo 5, the platform’s predecessor, was created using GTK+ and Clutter, while the MeeGo user interface has been built from the ground up using Qt. By all accounts, there are still a number of stability and quality issues with the stack, which we can expect to be addressed in a release shipping on a device.

At this time, the MeeGo Handset UX is not intended for anyone but developers. It is too early to be able to tell how the final product will compare to iOS or Android.

The Developer story

At the time of its announcement, one of the key advantages held up to developers was the potential to use a single toolkit, Qt, to build native applications which will be portable across Windows, Linux and Symbian. Nokia has been investing heavily in RAD tools like Qt Quick to allow developers to get up and running quickly. In addition, their as-yet unavailable Web Run Time promises to allow developers to easily integrate web applications.

The developer tools are in development, and do not yet compare favourably with the equivalent Android offering, which includes easy tools for building, testing and deploying applications using Eclipse. In addition, since the project is still in relatively early stages, there is a marked lack of entry-level documentation to help developers get started.

It is still unclear what software distribution channels or app stores will be available for application developers on a MeeGo device. Ovi Store will be available on Nokia devices for commercial applications, and there may be a community distribution channel made available for community-built applications, but what form this channel might take, and to what extent it will integrate with the MeeGo user experience is still unclear. Presumably other handset manufacturers, should MeeGo gain wider adoption, will provide their own application stores, further fragmenting the application developer story.

MeeGo certification ensures that it will be possible to build applications which work across all vendors, but at this point the jury is still out on how useful “MeeGo Compliant” will be to application developers. There is a possibility of considerable fragmentation among non-core APIs when MeeGo devices from several vendors are available.

From the point of view of tools, documentation and software distribution channels, MeeGo is undoubtedly behind its primary competitors – but for such a young project, this is to be expected. The success of the project among application developers and the free software community will depend to a large extent on the project’s ability to fill these gaps and provide developers with an excellent development experience.

For platform developers, the story is much more encouraging. The source code to the entire MeeGo stack is available, and anyone can download images built daily. Images built for ARM and Intel Atom can be installed and tested on a range of developer devices, including the Nokia N900, TI’s BeagleBoard or PandaBoard, or the Aava Mobile developer kit.

On the other hand, there has been a tendency of the platform architects to reduce the range of hardware and software supported by the basic MeeGo stack. There is limited support for non-Intel x86 chipsets, and support for only a subset of ARM chips. Kernel modules have been aggressively trimmed, sometimes arbitrarily, to disable functionality such as NFS.

Community and governance

MeeGo development is all happening in a public git repository, most discussions are on public mailing lists, and there are a large number of experienced free software developers among the community development team, which is ensuring that any communication or transparency problems are identified and addressed swiftly. In the mobile platform development world, it is fair to say that MeeGo is second to none in terms of its open development model.

This contrasts sharply with Android which is primarily developed behind closed doors by Google, and iOS which is a completely proprietary platform. If there is a key differentiator for MeeGo in the hand-held market, this is it. It remains to be seen whether the open development model will be a selling point which will tip the balance when manufacturers are choosing a platform for a device.

The MeeGo community is made up of members of the Maemo and Moblin communities, and in the case of Maemo, there have been a number of contributors who have decided not to contribute to the MeeGo project. The move to MeeGo represents the third major change in the project in two years (after the move to GTK+/Clutter in Maemo 5 and the announcement that Qt would be the only supported application toolkit) and has left some shell-shocked.

The Moblin community, on the other hand, did not develop a large platform developer community, partly since the project did not offer a distribution channel for application developers. It seems like all those who were productively contributing to moblin have followed the project move to MeeGo.

OEMs and operator support

One of the key differentiators between traditional handset manufacturers and the young guns (iOS and Android) which have taken the market by storm, is that both Android and iOS have concentrated on the user and application developer experience to the detriment of their relationship with OEMs and operators. It is widely argued that Apple’s iPhone has reduced the role of the operator to that of a bandwidth and infrastructure provider. In turn Google takes a take-it-or-leave-it approach with handset manufacturers; unless manufacturers comply with Android’s compatibility definitions (CTS and CDD), they can’t have access to the Android trademark, Android Market’s 100,000+ application, Google Maps and several other closed source applications.

Nokia has a more traditional approach of putting handset manufacturers and network operators ahead of developers. This shows through in many of the architecture decisions in MeeGo. The platform has been built with operator and OEM customisation and integration in mind from the start.

A primary concern for OEMs with MeeGo is the time required to integrate the platform into a specific device and ship to market. With the time to market for Android handsets dropping to 4-5 months from project to production, it will be very hard for MeeGo to compete, even with the MeeGo 1.2 release, due in the first half of 2011.

Still a long way to go

It does not feel fair at this point to compare MeeGo, a project which came into being 8 months ago, with iOS or Android, but this is the yardstick which will be used when the first MeeGo smartphone comes on the market. The project has come a long way since its inception, in particular in working towards an open and transparent development model. There is still some way to go but improvements have been happening daily.

However, to succeed as a platform, the application developer story and the user experience are vital. There is a lot of work to be done in these areas for MeeGo to gain serious traction outside of the small community of Finnish handset designers. Nokia still has a long way to go.

– Dave

[Dave Neary is the docmaster at maemo.org and a long-standing member of the GNOME Foundation. He has worked in the IT industry for more than 10 years, leading software projects and organising open source communities,  He’s passionate about technology, and free software in particular.]

Symbian is dead. Long live Symbian

[Is Symbian coming to the end of its shelf life? Research Director Andreas Constantinou dissects the motivations behind Nokia’s strategy and why Symbian is getting a new lease of life]

Only two short years and four months since it was announced, the Symbian Foundation is shutting down. With it dies Nokia’s second effort at creating a licensable application platform for mobile phones (the first one was S60) and to compete against Android. While Nokia is shunning to make the closure official, the last OEM supporters – Samsung and Sony Ericsson – have officially killed plans for Symbian products (see here and here) and Symbian staff are being given redundancy notices and making career moves on LinkedIn. [update: On November 8, it was announced that Nokia will regain control of the Symbian governance process and that the Symbian Foundation will be reduced to a licensing team]

The writing has been on the wall since early 2010, when Nokia took out a €500 million loan to (among other things) help sustain funding into the Symbian Foundation, whose membership fees were due to be renewed in April 2010. Symbian Foundation relied on OEMs shipping handsets to take on the operational costs at the tune of 5 million GBP per OEM. The final blow came with the departure of SyFo’s CEO and co-architect, Lee Williams.

The death of Symbian
Symbian Ltd., the OEM-backed consortium that funded Symbian development between 1999-2008 had long been suffering from an imbalance of power and poor strategic decision-making. There were three things wrong with Symbian Ltd.

Firstly, with Nokia owning 48% of Symbian Ltd. shares, the Finnish OEM had been driving the agenda at Symbian to the detriment of its OEM partners, Secondly, since the UI was severed from the base OS in 2001, Nokia had been squeezing the value out of the Symbian operating system and into its own S60 UI, middleware and applications suite platform. This meant that other OEMs had to spend considerable effort integrating Symbian with their own UIQ or MOAP layers and filling the gaps that Nokia left – effectively leading to handsets which were expensive to build.

Thirdly, with the decision to have Symbian baseporting owned by the OEM and not Symbian Ltd, each manufacturer had to spend millions to get Symbian ported onto the hardware platform, in essence reinventing the wheel. While this naturally gave Nokia the edge in producing more Symbian models more often, it meant that for other OEMs most of the budget was spent in baseporting (i.e. getting the phone to work), rather than in differentiation. In 2007 Symbian Ltd. was desperately in need of a major governance re-engineering operation.

The coup de grace arrived with the launch of Google’s OHA in November 2007, signaling two major changes in the phone industry: firstly, that open-source development (inspired by mobile Linux) was now supported by a major cash-rich backer, and not an operator consortium (LiMo) or a loose congregation of Linux system integrators and design houses (Azingo, Purple Labs, WindRiver and Montavista). Secondly, that zero royalties were now the norm and operating system development was turning from a revenue generator to a loss leader. With Android changing the rules of the game, Nokia knew that for Symbian to compete in this new world, it had to be both open source and zero royalty.

Seven months on from the Android disclosure, Nokia announced that it would be buying the remaining Symbian shares outright, paying up the equivalent of 2.5 years of royalties or 2x the revenues of Symbian Ltd – a paltry evaluation for the top smartphone OS. For Nokia it was a financial and strategic move; it made financial sense because Nokia would slash its Symbian maintenance costs (from 100 million GBP of annual license fees to 5 million GBP of annual membership fees) by sharing the SyFo costs with other OEMs on the board. It made strategic sense because with the ownership change, Nokia convinced Sony Ericsson and DoCoMo to abandon UIQ and MOAP respectively and marginalised Windows Mobile which was still royalty-based. Meanwhile, Nokia could still exert the majority influence into the Symbian roadmap by employing most engineers and most package owners (effectively well into 2010).

In retrospect, Nokia failed with both S60 and Symbian Foundation by insisting on a winner-takes-all mentality, i.e. taking roadmap control away from its OEM development partners which long-term destroyed the value in the partnerships. This winner-takes-all-mentality is nothing new; it was already harming Symbian as we had argued back in 2005. The full open sourcing of the Symbian platform in February 2010 or the cute playful new brand did not succeed in stopping neither the developer defection (see our Developed Economics report) or the OEM defection from Symbian.

With Nokia shares performing miserably over the last four years, the Finn-led board took the bold decision to oust Olli-Pekka Kallasvuo and bring in a Canadian, Stephen Elop to turn the boat around. 41 days into the job, Elop announced the cutting of 1,800 jobs at Nokia and the adoption of Qt as the main development environment on top of Symbian handsets.

For Nokia, Qt presents both an opportunity and a challenge. On one hand it’s the most capable cross-platform application environment today boasting reach across mobile, PC and STB – plus depth with Qt providing a complete API wrapper on top of the native OS (and much wider API coverage than GTK to which it’s often unfairly compared). On the other hand Nokia has notoriously mismanaged the Trolltech acquisition of January 2008, with the troll CEO, CTO and key engineers abandoning ship. Meanwhile, Nokia has created a Qt break across Symbian and MeeGo UIs and not managed to fully deploy Qt on Symbian 2.5 years after the acquisition (note how Qt Mobility APIs are still way incomplete).

Long live Symbian
With Symbian Foundation soon to be diagnosed dead, the rumours about Nokia replacing Symbian are rampant. Many industry pundits are prognosticating that Nokia will adopt Android – which in 2010 is going stronger than ever – or Windows Phone 7, which comes with the freshest UI since the widget based paradigm popularised by the Jesus phone. Despite the prophecies, Symbian will live on for many years to come. As the French expression goes, Le Roi est mort. Vive le Roi.

There are two reasons why Nokia won’t be abandoning Symbian anytime soon.

Firstly, Symbian is tightly integrated with Nokia’s variant management process. Nokia is the only OEM that has mastered variant management, i.e. being able to generate 100s of variants (SKUs) at the press of a button. That’s how Nokia can deliver 100s of customised smartphones to operators and retailers around the world. This variant management process is ‘hardcoded’ to Symbian, which means that replacing Symbian would seriously compromise Nokia’s ability to cater to operator requirements around the world and it would seriously hurt its market share.

Secondly, Nokia’s economies of scale rely on in-house control of core components, and the operating systems is one of them. If Nokia were to license Windows Phone it would reduce its differentiation to industrial design and Ovi alone. In the case of Android, Nokia would have to branch Android (and to sustain the cost of Android development), port Qt on Android which means another 12+ months for a stable implementation. While this remains a long-term possibility, it is still a gamble when Nokia’s priority should be to focus on killer devices and not a killer OS. Qualcomm’s BREW MP is another candidate but only when Qualcomm has a good developer platform story and that means waiting for BREW MP to launch a web-based platform akin to RIM’s WebWorks.

Symbian may no longer be a symbiotic system, but will live within Nokia for many years to come as the workhorse under the hood of Nokia smartphones.

The King is dead, long live the King.

– Andreas
You should follow me on twitter: @andreascon