Archive for the ‘Operating Systems’ Category


Microsoft buying RIM? 6 reasons why it will happen and 6 reasons why it won’t

October 11, 2008

A few friends have asked what i think of the speculation about Microsoft buying RIM. It is certainly an interesting proposition. Let me offer a few pros and cons. Keep in mind that RIMM is a company built of three main areas: an operating system and software stack, device design and manufacturing, managed service operations (NOC). 

Why it could happen:

  1. At current value, RIMM is cheap, at least compared to what it was worth six or twelve years ago. Not a lot of money.
  2. Microsoft is a very competitive organization. Steve Ballmer is hyper-competitive. To an extent, you could say Microsoft is obsessed with winning. They hire type-A persona lites, they like to win. Not being able to beat RIM hurts. Buying your competition is not exactly beating your competition but it eliminates the competition.
  3. Microsoft would instantly have a much larger share of the smartphone market and would have a better chance at driving consolidation in a crowded space (iPhone, Windows Mobile, BlackBerry, Symbian, Android, Access PalmOS, OpenMoko, etc.)
  4. Microsoft would gain some unique hardware design and usability skills that the company could use.
  5. Microsoft is already in the device business. A similar thing happened in the music player business: Microsoft tried a partner-based approach by licensing Windows Media to creative, Sony and other electronic manufacturers and when it became evident the strategy was not working the company launched its own device, the Zune, while maintaining the licensing model for Windows Media, DRM and the player platform. Microsoft has proven it can be good at consumer electronics and hardware with the XBox and with the keyboard and mice business. With the acquisition of Danger Microsoft is already in the smartphone hardware business.
  6. Sure, the platform is incompatible – so what? It did not stop Microsoft from buying Danger. DataViz offers a J2ME ActiveSync client that could sit on top of the Danger and the BlackBerry to get both platforms connecting directly to Exchange servers.

Why it will not happen:

  1. There is no glory in buying the competition. Microsoft could not never say they beat RIM, they don’t “win”. The acquisition would build upon the bully/monopolistic image of Microsoft rather than the innovator image they are trying to build.
  2. The technology is incompatible. BlackBerries are built on a proprietary OS with Java on top. To make it work, Microsoft would have to migrate the entire software stack to Windows Mobile and .Net which would take a lot of time, take a lot of resources and skeptics would say in the process Microsoft would kill all the goodness of the BlackBerry platform. Not because of this goodness is tied in any way to Java or because of lack of capabilities in Windows Mobile (in fact, I believe the Windows Mobile architecture, developer platform and tools are an order of magnitude better), simply because porting a full smartphone stack while leaving it intact from a customer experience perspective is pretty much impossible.
  3. Microsoft would become a hardware manufacturer becoming a competitor to all the 45+ Windows Mobile licensees and pushing them to accelerate their Android/Linux investments. Alienating your partners is not a good business practice.
  4. Microsoft is not very good at building devices – XBox succeeded because it is a separate division with its own culture and in a different market. The Zune has not yet failed to capture a significant share of the market (even though it is a great product, I love mine). Hardware design is not one of the core competencies at Microsoft.
  5. Microsoft has been touting the inefficiencies of a NOC, which I agree on. The iPhone, palm OS devices, Sony Ericsson, Nokia and other players licensing ActiveSync signal the death of middleware. It would be difficult for Microsoft to justify a platform based on a NOC model.
  6. The culture is too different. The business model is too different. RIM’s model is fundamentally to please the wireless carriers, give carriers control (with the NOC) and support the carriers as the main channel in selling devices plus service. The Microsoft model is about giving IT organizations control, less middleware, and about partnering with multiple device manufacturers. RIM’s strategy is about controlling the user experience with a proprietary platform end to end while Microsoft’s strategy is to provide a platform on top of which device manufacturer partners, software developers, and carriers can innovate.

Let me clarify that while I used to work at Microsoft in the Windows Mobile team, I have no insight about these discussions at Microsoft. These are conversations that would happen a level or two above where I was.

My guess is that it won’t happen. But who knows? this is what makes the industry so interesting, anything could happen. Especially in this economic climate. It is reasonable to expect smaller, underfunded and companies who have lost significant market cap to be acquired by the big guns. Isn’ t it amazing that Microsoft could buy GM (at current market cap of $2.6B) with the equivalent of roughly two months of profits?



Thoughts on the Symbian Foundation

June 27, 2008

You may have seen the news that Nokia is buying the parts of Symbian that it did not own, to create a non-for-profit foundation that will offer the OS royalty-free and will open-source it later, unifying the multiple UIs that exist today. I read the announcement and some of the analysis online, which led me to a few observations:

          I find it very respectable that Nokia recognized the fragmentation in the Symbian ecosystem and took bold steps to unify the platform. Obviously, the fragmentation across UIQ, Series 60, Series 80 and MOAP(S). Still, I am not sure there is a smooth transition plan, especially for the Japanese market, to go to S60.


          It is very interesting that all current shareholders are committed to making the sale except for Samsung. According to the press release “Nokia also expects Samsung Electronics Co. Ltd. to accept the offer “.  Did they not have time to talk to Samsung about this, or was Samsung not very happy about it and they are being forced to sell?


          UIQ will disappear: the foundation promises backward compatibility to Series 60 only, which means developers should stop any UIQ development and start porting their stuff to S60. UIQ just announced they are laying-off half of the staff. This is sad because UIQ was probably the best UI for Symbian.  I am sure Sony Ericsson and Motorola are not very happy to see the multi-million dollar investment they have made over the years in UIQ technology evaporating into thin air. People who invested in UIQ 3 must be frustrated it may never see the light of day.


          As a consequence, Nokia is forcing Nokia and Sony Ericsson to make a decision about their 24-month Symbian roadmap. Their options are:

a)      Launch phones with end-of-life technology (UIQ), or

b)      Wait until the foundation produces the first release, which could be 2-3 years or more (until a phone is in market), or

c)       Switch to another mobile platform. Not really a switch because both already have Windows Mobile and Motorola has significant experience with Linux


          I understand the move to consolidate. But why make Symbian royalty free and why make it open source?  I believe many software companies donate their technologies to Open Source as an exit strategy whey they don’t see a profitable business opportunity in them. I made a controversial post on my old blog about this not too long ago.Nokia may be doing it to demonstrate openness for a technology they own but which they want partners and competitors to adopt.


          The new foundation will have a board of directors and a number of councils to drive architectural, UI and platform decisions. How this works in practice will be critical for the success of the platform. Nokia could (or could be perceived to) have too much control over the platform which will obviously not be taken well by competing device manufacturers. The alternative could be worse: technologies driven by a committee usually go nowhere. The best example is J2ME. The Java Community process has been incredibly ineffective. To illustrate: back in 2001 I was very close to the launch of MIDP 2.0. Fast forward to 2008, the latest version of the MIDP specification is still 2.0.  I was very close to the Java Community Process and I saw major vendors trying to push their agendas in most JSRs. As a consequence, the committee had to compromise, approving specs based on lowest-common denominators. Getting any JSR approved took years in some cases, even for simple things like the vibration API for games.


          Another key question will be the business model behind Symbian. On one side, the business model behind Linux is clear: there is none, most contributions are voluntary, companies usually make money on services (i.e. support). On the other side, commercial operating systems like Windows Mobile have a clear business model as well: there is a license per phone sold. The Symbian Foundation will have a hybrid model where there is no license revenue but Symbian will be the sole provider of funding for engineering, support, marketing, etc.  How much will Nokia continue investing over time?


          Royalty Free does not equal cost-free. I expect some pressure for Windows Mobile, but the cost of the OS license is not as significant as people believe. The engineering cost to create a device is in the millions and millions of dollars. Android device manufacturers are experiencing this first-hand.  In this hyper-competitive market OEMs will question why they would continue investing in a technology owned by their biggest competitor.


          While the move by Nokia consolidates the Symbian operating system, it further fragments the royalty-free and open-source mobile OS industry. Now you will have Symbian, LiMo, Open Moko, Android,and other more obscure Linux-based Operating Systems. I don’t think there is room for so many. Is this UNIX all over again?


          For developers, open-sourcing will be attractive, mostly from a perception perspective as most won’t have the skills or the time to analyze and understand a subset of the 7 million lines of code that Nokia will open-source. While SDK will be free, the tools might not: the Carbide C++ Professional Edition development tools are being sold for 1299 Euros.


          One possibility could be the move represents Nokia bailing out the other major Symbian partners that jumped into the business with them through this $400 million buyout.  Iam sure there were interesting executive conversations in Espoo about spending $400 million to buy a technology that will be free and which will require continued investment. The $400 million investment should be roughly the equivalent of two years of license payments at the current rate Nokia was paying Symbian. Still, Nokia still must invest significant resources to the foundation ad-infinitum. Foundation membership is only $1,500.

A couple lst thoughts notes:

Symbian is a very mature and powerful operating system. I remember when we launched CodeWarrior for Symbian (before it was sold to Nokia and renamed carbide) in early 2002 and later helped Sony Ericsson build a developer community around the P800. It is a good OS. Architecturally, it is very solid – years ahead of the Blackberry OS 4.5 and other new operating systems. From a technology perspective, it is a very viable OS today.

Don’t be misled by the numbers touted by Nokia: surely millions and millions of phones ship with Symbian. But are these really smart phones? They are from a capabilities perspective. In my experience, a very large percentage of Symbian-based phone owners are not aware that their phone is “smart”, do not know it runs Symbian and/or do not use it as a smart phone. It’s like Sun claiming victory with over a billion J2ME phones: most of the users will never be aware and will never use a J2ME app. In other words, the fact these phones have J2ME is irrelevant. The same can be said for most Linux phones, which are not smart phones – Linux is merely replacing an RTOS.

Conclusion: This is a very interesting move from Nokia that will have significant implications in the market overall. The key questions are Nokia’s ongoing investment in the foundation without a solid business model behind Symbian and the balance Nokia will have to find between having too much control over Symbian versus a committee-driven process that inhibits innovation.

Fun times. I love this industry.