The end of the dream

No matter how it works out in the long run, the purchase of Symbian by Nokia marks the end of a dream -- the creation of a new independent OS company to be the mobile equivalent of Microsoft. Put a few beers into former Symbian employees and they'll get a little wistful about it, but the company they talk about most often is Psion, the PDA company that spawned Symbian.

Psion never got much attention in the US, but it was a pioneer in the PDA market in the UK, and even to this day I think the Psion Revo is one of the two coolest-looking PDAs ever made (the Palm V is the other one).


The Revo

Psion explored many ideas that eventually turned into major new consumer electronics categories, but it failed to follow up on them. The company was effectively dismembered when Symbian was formed, and many of its best people drifted off to other companies. Now Symbian itself is transitioning to something very different, with most of its people absorbed into Nokia. What the Psion veterans talk about wistfully is how many smart people worked at Psion, how many great ideas the company fumbled, and how successful many of the people have been in the tech industry post-Psion. In this sense, Psion is similar to many other tech pioneer companies that assembled staffs of very bright people, taught them how to work together, and then blew apart like exploding stars, scattering the elements of new companies across the industry. This process dates back at least to Fairchild Semiconductor, which trained the founders of many of the most prominent semiconductor companies (link). You can find similar networks of former employees from places like Apple, Netscape, and Palm. I think Yahoo is in the process of forming a network now, and some day there's going to be a dandy one made of former Googlers.

What makes the Psion story different is that many of the Psion veterans had to leave the UK, or join non-UK companies, in order to become successful. Some are in other parts of Europe, some are in the US, and some are in London but working for foreign companies. This is a source of intense frustration to the Psion folks I've talked with. They feel like not only their company failed, but their country failed to take advantage of the expertise they had built.

There's a big body of academic research on why Silicon Valley has been successful in sustaining itself, and part of the reason is that the Valley recycles companies very efficiently. Failing companies do not last long, but in the process the brightest people and ideas are rarely lost, they are just shuffled around into new configurations.

About a year ago, Andrew Orlowski of the Register wrote an amazing article on the history of Psion, and how company culture and government philosophy failed to take advantage of it to grow a new industry. It's the longest piece I've ever seen in the Register, almost the nucleus of a book, and it's well worth reading. It didn't get enough attention when it was published, and I'm embarrassed to say that I never posted a link to it. So I'm glad to remedy that now. If you want to understand the context what happened to Symbian, and learn a bit about how the tech industry works, go read it here.

If you want to hear more about what Symbian is morphing into, two of its executives have just started personal weblogs in which they are commenting on the migration to Symbian Foundation (among other things). It's an interesting move, and it seems symbolic of the transition they're trying to make into the open source world. Previously Symbian had a company blog that several execs contributed to; now the execs have personal blogs where they talk directly to the industry.

David Wood (Symbian's EVP of Research) link.
John Forsyth (Symbian's Strategy VP) link.

Symbian changes everything, and nothing

[With a correction made on June 26.]

The Symbian Foundation announcement today is a fascinating change in business strategy, but I'm not sure if it will help or hurt Nokia in the long run. I think something like this was probably necessary just to clean up the mess in Symbian's ownership structure. If Nokia can make the new structure work, it'll be a milestone in the use of open source by large tech companies, but I'm not sure it helps Nokia win the smartphone war.


What happened

--Nokia is buying Symbian. Everyone currently working at Symbian becomes a Nokia employee after the deal closes. Nokia said it will spend the next six months deciding "how we will use the unique talent we are gaining."

[By the way, the buyout by Nokia is a change I said was possible two and a half years ago when it first became clear that some of Symbian's owners wanted out (link). I am astounded that the change took so long. I looked back at my old post a few months ago and thought, "wow, I really got that one wrong." Now I am relieved to say that I was not wrong, I was merely prematurely correct ;-) ]

--Symbian OS will become free. Nokia's Symbian-related assets, including both Symbian OS and the S60 interface, will be contributed to the new Symbian Foundation, a nonprofit that will control the Symbian platform. So Nokia writes the code and then gives it to the foundation for free.

Founding members of the foundation include: AT&T, LG, Motorola, Nokia, DoCoMo, Samsung, SonyEricsson, ST Micro, TI, and Vodafone. It's very interesting to see some operators in the mix, especially AT&T.

The foundation will open source the new Symbian platform over a two year period. So eventually Symbian will be available for free.

The new Symbian Platform will have a broader scope than the current Symbian OS. It will include:

-An application suite (previously controlled by licensees)
-Runtimes (including Webkit, Flash, Silverlight, and Java; previously licensee-controlled)
-UI framework (formerly controlled by licensees)
-Middleware
-OS
-Tools, SDK, and application signing (previously shared between Symbian and licensees)

--UIQ is dead. SonyEricsson's UIQ technology, and NTT DoCoMo's MOAP, both of which are user interface layers written on top of Symbian, will also be contributed to the foundation, which will incorporate pieces of them into S60. The new Symbian foundation partners said at the press conference, "We will reposition UIQ in the new ecosystem." That's seems to be a face-saving way of saying, "UIQ is dead." Confirming that, UIQ announced immediate plans to lay off more than half its employees (link).

These are huge changes, even though they'll take a couple of years to implement. We won't get the first release of the new merged platform until 2010, although the partners say S60 and native Symbian apps will continue to run in the future, so they hope many more developers will create Symbian apps today in anticipation of future growth.

--Nokia will continue to control Symbian development. This is my interpretation, not something they announced. Technically, control over Symbian and S60 passes to the new Symbian Foundation, with product plans controlled by a managing board and councils made up of foundation members. This makes Symbian sound independent. But Nokia will employ most of the people maintaining and extending Symbian and S60, and could divert them to other Nokia projects if it ever dislikes the direction of the foundation. More to the point, the whitepaper explaining the new foundation says, "device manufacturers will be eligible for seats based on number of Symbian Foundation platform-based devices shipped, with the other board members selected by election and contribution" (link). So Nokia as the dominant shipper of Symbian devices gets the most seats, and can then control the election of additional board members. Symbian contacted me on June 26 with a correction: "Five Foundation board seats will be allocated to handset vendors on the basis of volumes shipped using the Symbian Foundation platform. There will be a maximum of one (1) board seat per company." So Nokia gets one board seat, and does not control the foundation.

The right phrase for this, I think, is puppet strings. But I don't mean that in a bad way; it would have been insane for Nokia to actually give up control over its smartphone OS. Just don't have any illusion that the strings have been cut. They've merely been relocated, and in fact I think Nokia now controls things more directly since it owns the Symbian development team. Added June 26: Nokia has given other companies a formal say in the feature set, with less official control by Nokia than it had when it held about 50% of Symbian, but perhaps more practical influence because it now directly employs most of the people doing the engineering. So I think Nokia gave up the official veto it had over Symbian's actions, and replaced it with a practical one.


What does it all mean?

I don't know.

The announcement is so complex, and so many things are changing in the mobile market, that it's very difficult to predict how everything will turn out. Also, the whole thing depends on crisp implementation. Even the most brilliant strategy fails if you can't execute on it.

You can't say that Nokia lacks guts. The foundation members said at the announcement that it is one of the largest open source announcements ever, and I think that's true. It's a very interesting, aggressive move for Nokia, and I respect that. There are precedents for a big company acting as a sugar daddy for an open source software project, but I don't think it's ever been done with a project that is as central to the parent company's operations as Symbian is to Nokia. It will be fascinating to see if Nokia can really work effectively through the foundation model. I presume they have thought about this a lot and feel the risks are well controlled.

I'm having trouble seeing the big picture of how this changes the world, though. I suspect the announcement is actually half cleanup and half power move. The power move is that it challenges Android, and could help harness the energy of the open source community to support Symbian. The cleanup is that the ownership situation of Symbian was unstable and had to be changed eventually, and SonyEricsson clearly wanted to get out of the UIQ business. The creation of the foundation solves all of those problems at once. My guess is that since Nokia is paying most of the bills, the other foundation partners were willing to go along with it. The Symbian investors get some money from Nokia, and can sit back and wait to see what the foundation delivers.

Here are some other issues and questions that stand out to me:

Symbian gets its UI back. Years ago, Symbian took itself out of the user interface business, allowing Nokia and NTT DoCOMo to develop their own UIs, and spinning out the UIQ interface team. The company declared that it had been a mistake to ever go into the UI business. So it was amusing to hear Symbian at today's press conference saying how disruptive it was to have multiple user interfaces, and how great it is to have them unified.

The reality is that OS companies have traditionally created the UI along with the rest of the OS because they need to be coordinated closely, and because developers want to work with one consistent interface. So the real mistake was getting out of the UI business, and Symbian has now corrected that.

What will happen within Nokia? At the press conference, Nokia was asked what happens to its internal S60 development team (which is rumored to be larger than Symbian itself) once the merger is complete. Nokia said vaguely that it's going to spend six months working out all those integration issues, and what it will do with the multiple geographic locations. It's hard for me to believe that working out process won't result in some layoffs. I hope I'm wrong; I have friends at both Nokia and Symbian, and layoffs would be incredibly painful for the Symbian folks, many of whom have spent most of their careers there.

The fate of the people is just one of the open questions about what the merger means to Nokia. Another is the fate of Trolltech, the development tool that Nokia purchased recently and said would unify app development across Series 40 and S60. Will it be contributed to Symbian? And what does the open sourcing of Symbian mean for Nokia's use of Linux?

How does Nokia differentiate its software? The theory behind S60 was that Nokia would have its own user interface, helping to differentiate its phones from other Symbian vendors. Now that S60 will be given away, how will Nokia differentiate? The Symbian Foundation says licensees will be able to create a "differentiated experience" on its unified UI framework. Lord only knows what that means. Maybe Nokia has decided the UI is not a point of differentiation at all, and plans to focus on something else (web services, perhaps?)

Will the change in Symbian really drive more developers? As the Symbian partners pointed out repeatedly in the press conference, they have already sold 200 million phones. If that's not enough to excite developers, how will adding another 200 million -- or even 500 million -- do it? Although Symbian now has a nicer long term story, I don't think most developers were paying attention to that. They respond to user excitement and the chance to make lots of money. The new Symbian strategy doesn't directly drive either one.

What does it mean to Apple? I think it's probably good news. Although the Symbian partners could theoretically bleed Apple by sharing investments that Apple has to fund for itself, Apple competes on speed and elegance, not cost control. Nokia and Symbian will now spend the next six months sorting out how they'll integrate and rationalize their organizations. No matter how much they try to avoid it, this will slip schedules and force people to revisit plans. And the other Symbian licensees have to wait two years for the new OS. That gives Apple a long, long time to build up its iPhone business. The Register put it very bluntly in its commentary on the Symbian announcement (link):

"Apple must now see a clear road ahead for world dominance...it's now Apple's business to lose."

Wow, from new entrant to industry leader in just a year. That sort of stuff must drive Nokia nuts.

Is Google happy or upset tonight? My first reaction is to say that Google should be worried because there's now another very credible operating system being given away for free in competition with Android (or there will be in two years). What's more, the leading mobile handset companies all participated in the Symbian Foundation announcement. That makes it harder for Android to get licensees. But the new open Symbian OS is two years away from shipment, giving Google lots of runway to get established (that's what I meant about execution determining the real impact of the announcement). Also, the governance system for Android is a lot simpler than Symbian's. While the Symbian committees must debate and agree on product plans, Google can just decide whatever features it wants to add, and toss them out there. In theory, Google should be able to move much faster.

Besides, there is the question of why Google really created Android. One school of thought says that Android was just a tool to bleed Microsoft and force openness in the mobile ecosystem. If that's the goal, then the opening up of Symbian is a kind of a triumph for Google. Nokia is, in many ways, doing Google's work for it. Which brings us to...

What happens to Microsoft? Here's the weird thought for the day: Microsoft is the last major company charging money for a mobile operating system. The throwback. The dinosaur. How many companies are going to want to pay for Windows Mobile when they can get Linux, Android, or Symbian for free? This is Microsoft's ultimate open source nightmare, becoming real.

Thoughts on the 3G iPhone announcement

Apple's 3G iPhone announcement today was probably the minimum necessary to please the community. The real news was the things that weren't announced:

--No tablet device (again).
--No major changes to the form factor of the iPhone.
--No other major product announcements.

Apple has made its Macworld and WWDC keynotes into a specialized form of performance art, complete with cleverly-dropped pre-announcement hints, and often some sort of surprise at the end of the speech. Apple's own past successes have now raised the anticipation for the keynotes so high that it's a disappointment if some sort of major surprise doesn't happen.

Check out Engadget's live blog of the speech if you want to see the result (link). It's littered with whining like this:

"We love what you've done here, but we're yawning."
"Man, these demos are crazy boring."
"Man, please let this string of demos end!"
"Another developer demo. Ugh."
"Wow, we heard Apple's stock is down almost $5 since this keynote started. Maybe they should just demo their top three and keep going."
"Someone, wake us when Steve's back."

I didn't actually attend the talk, so I don't know how boring all those demos were. But I think it's fair to remind people that the WWDC is a developer conference. It is traditional to do a fairly large number of app demos at a developer conference, because that's a low-cost way of rewarding your developers.

Apple discussed some other interesting things in the keynote. Here's what stood out to me, with some comments:

The "lower" pricing. This was completely necessary. AT&T claimed in an interview with the New York Times that $199 is a magic price point for smartphones (link). They're right, it is. But as the Times pointed out in another article, the price cut isn't actually as meaningful as it sounds -- AT&T is making up for it by raising the price of the iPhone data plan by $10 a month, with a two year contract requirement that will apparently be rigorously enforced. So to get that $200 discount on the purchase price of the iPhone, you pay an extra $240 over two years.

You're actually losing money in the long run, but now the iPhone is priced in the same way as every other phone on the market, making it more comfortable to buy. That figures to help iPhone sales -- especially in Europe, where the unusual price structure for iPhone caused a lot of complaints.

If they really do enforce the contract, that will probably put an end to the widespread practice of buying iPhones in the US, unlocking them, and shipping them to places like China. But the iPhone is getting much stronger international distribution, with up to 70 countries in the works according to Apple. We have no way of knowing how well the contracts will be enforced around the world. Chances are there will be gray market leakage from somewhere.

Notification vs. background processing. One of the critiques of the iPhone is that it doesn't allow third party applications to run in the background, without being visible to the user. Apple said this is to prevent background applications from hurting performance, the way they do on Windows Mobile. But that's a very disingenuous explanation -- Windows Mobile manages memory very strangely, often leaving things in memory whether they run in the background or not. You could create a very efficient architecture that still allows background processing.

Apple says it has solved the background problem by setting up a notification server that can wake up applications on your iPhone and pass incoming messages to them. I don't know how that looks on screen -- since Apple won't run apps in the background, does that mean they'll suddenly launch on screen and start operating on their own? Creepy. And although notification does some of the things you'd want from the background, it doesn't do them all. For example, some developers want to write background applications that would perform tasks automatically, whether they are pinged by an outside server or not.

All in all, it's interesting that Apple's establishing a messaging server for iPhones. Combine that with Apple's new MobileMe service, and Apple is gradually creating a lot of back end infrastructure for the iPhone. In the long run, Apple could build many innovative new services around that infrastructure.

I wonder if they'll charge developers a fee for passing messages through the Apple infrastructure.

When do the developer limits come off? Apple bragged in the keynote that there were 25,000 applicants to the iPhone developer program, but the company admitted only 4,000. In other words, they seriously pissed off 21,000 developers. Not the sort of thing I would brag about, but this is Apple and they can sometimes operate on a different set of rules.

The question is, when (if ever) do the other 21,000 developers get into the program? As far as I know, Apple was silent on that issue. If they were about to open up the program, you'd think they would have announced that.

The application demos skew toward consumers. Four of the applications demonstrated during the keynote were games, one was a consumer news applications, one was a social network product (Loopt), one was consumer shopping (eBay), one was consumer blogging (TypePad), one was sports information, and two were vertical medical. Although Apple talked about enterprise at the start of the keynote, the apps they chose to demo tell you everything you need to know about who Apple sees as the iPhone's buyers.

What happens next? The iPhone is only a year old, and it generally takes 18 months to design a major new device. So the 3G iPhone we saw today was probably already in early development when the original iPhone was launched. We could see more radical hardware change this fall, but I think it's more likely that would wait for Macworld 2009.

What happens to iPod pricing? I was surprised that the price of the iPod Touch didn't change today. It now looks more expensive than the iPhone, and it lacks GPS. I would not be shocked if the Touch ends up getting a price action this fall.

As for when we'll see the long-rumored larger-screen iPod/iPhone, your guess is as good as mine. Fall is the best time for introducing new products, because it's right before the holiday/new year buying season. If the product exists, that would be the time to announce it.