What's new?

It's that time of year again... Motorola announce that they're going to focus on just a few platforms, instead of trying to make phones based on every platform in the universe. So they're dropping UIQ and their own Linux/Java platform. They're going to focus on Windows Mobile, Android and P2K. No info on the status of Brew within Motorola.

Anyway, this seems to be happening about once a year during the last couple of years: Motorola announce that they're going to focus on a few key platforms. Then they do the opposite. I wouldn't be surprised if they were releasing S60 phones within a year or so.

Success forever

This is kind of interesting. Michael Mace discusses successful mobile platforms, but seems to stumble on what a successful platform really is. Was Palm successful? Yes, it was, for a certain period of time. The same goes for all other platforms. None is going to be around forever. Not even Symbian, Windows or whatever.

That doesn't mean that you can't develop apps for a certain platform. It might be around for 10 years, or 5. Or whatever. If you want to look cool in front of your friends, prepare for change, because it's going to happen sooner or later, and then someone is going to point at you and say "your platform failed". Yeah, maybe the platform failed, but that doesn't mean that the stuff you produced for it did.

Finally

It's nice when people are talking about you.

Another good reason...

...to read Simon Judge's blog is that he always seems to have an eye for the bigger picture. While market analysts awe at a small percentage difference in market growth, it might be a bit early to pronounce Nokia's death sentence. This isn't the Silicon Valley. This is the real world. Thank you, Simon, keep doing everyone else's home work, and you'll be rewarded when you get to heaven.

Marketing talk

I love how marketing people use language. This doesn't sound as if you're supposed to read it, but just quickly glance over it and get impressed by the buzzwords:

Over the past decade Symbian has helped transform the mobile market through its unrivalled innovation and commercial success in driving the evolution of the smartphone show industry.

And you're definitely not supposed to actually read the next paragraph:

This year’s Smartphone Show celebrates 10 years of past achievements, while continuing to focus on those technologies shaping the future of mobile. We’re also launching our revived developer event, the Mobile DevFest, running in tandem with the show to provide invaluable training and information for our developer community. With extensive coverage of both technical and commercial aspects of the industry, we have no doubt there will be something for everyone at this year’s event.

Where did the apostrophes go? "Were already launching", was this spell checked by an 8 year old? How come marketing text is always so full of embarrassing errors? Wouldn't it be better if they learned to write properly (or have someone check their output before publishing it), so that they don't alienate nerds like me?

Mobile is dead

Mobile is dead, Russell Beattie proclaims in a recent blog post. What a moron, hasn't he noticed how everyone has a cellphone these days? Doesn't he know that there are several highly successful companies producing and distributing cellphones all over the world?

Open source and its consequences

People with no insight into what open source means and what consequences it has for systems, come up with all sorts of weird ideas. Nokia's buy of Symbian and its plans to open its source code have made people ask the same old questions once again. Some are enthusiastic, some are worried. Let's have a look at a few misconceptions.

Does open sourcing Symbian mean that there will be a problem with backwards compatibility? The misconception here is that when you release something as open source, there's nothing you can do about millions of open source hackers all over the world taking charge of your product. But as maintainer of the code, you decide on what gets merged. Sure, people might fork the code and run off in any direction with it, but which version of the code do you think the OEMs will use? Its them who are crazy about backwards compatibility, so they'll go with the official version from the Symbian Foundation.

The above point is also related to the assumption that people will actually want to work on the Symbian code. The problem with this is that a product has to be attractive to programmers, and Symbian is anything but. That's the reason why there are almost no people writing Symbian code in their free time now. It's horrible. You don't spend your spare time working on something horrible, unless you dig horrible stuff. Most people don't.

Does open source Symbian mean that anyone can do whatever they want with their phones, like install patches that switch off platform security or open it up to all sorts of software that their friendly operator don't want them to run? No, obviously not. An open software system doesn't mean that the phones are open. There are lots of Linux phones out there already, and there's not much you can do with them. The fact that you have the source code for Linux doesn't mean that you have any power over your locked down Linux phone from Motorola.

wtf?

I really don't know how to comment on this. If anyone out there isn't surprised, please raise your hand, so that I can point my finger at you and call you a liar.

Of course, promises are cheap, and "open" is a word that's almost completely empty in the corporate world, but the massive restructuring of the Symbian platform is radical enough.

It's comforting to know that the mobile blogosphere will give me answers to all the questions that this announcement raises within the next couple of days. So what are they going to say? "Facing ever sterner competition from Windows Mobile, Google's Android and Apple's iPhone, the key players in the Symbian ecosystem decided to take action, to lead to market to bla-bla"? Yes, probably. "Symbian finally succumbed to Nokia"? Yes, probably. "So how are they planning to make a profit?" Definitely.

Oh well, if this works out, it will have various interesting consequences. UIQ and MOAP will disappear (the useful parts that don't have counterparts in S60 will be assimilated, like a proven to be working outside the lab touch support). The mess that is Symbian will clash with the open source community, both technically and politically, and the Symbian Foundation will be the 666th company that come to learn that open source isn't about getting people to get engineers to work for you without paying them.

Can't be bothered to write anything more about this right now, but thanks to Nokia et al for an interesting press release, which I hope isn't a belated April fools joke.

Land of the free and all that

I'm in the USA right now, so it's kind of fitting that I just had one of those "wtf usa" blog experiences. Check this one out. I just love the "oh dude steve jobs just invented the cellphone lol" attitude in this one. You're saying there's a company in Europe (that country across the eastern ocean where there be nazis) that already had a webkit based browser on their phones? You're saying Japan (here be ninjas) are 5 years ahead of us? No way, we have bigger cars! And burgers!

Executive summary: iPhone killed Mowser!

More insights into the zen of programming

As a programmer (or "software developer", as it says on the business card) you're always impressed by cool quotes. It might not be the 90s anymore, but the really cool quotes still come from Linux kernel developers. Check this one out. In case you're too lazy to click the link, here's the important line: "inline is the register keyword for the 21st century".

I haven't researched this, but if Rusty Russell says so, I'm inclined to believe it. Basically, the compiler knows more about register allocation than you do, and these days, it also knows more about inlining than you do. The "inline" keyword has always made me feel a bit uneasy, and these days, it's next to impossible to keep track of how inlining affects performance. But we have powerful AI engines that take care of worrying about that kind of stuff, and they're called compilers.

Both you and me know that compilers don't know everything. In extreme circumstances it pays off to read a disassembly (from one particular target platform) of the code it produces, but we should all know better than to optimize without using tools to measure the improvements. Profilers have been around forever. The message is: don't inline without profiling with and without the inline keyword. Don't turn functions that you think are used a lot into messy macros. Use your head, and let the tools take care of the things that your head can't cope with. One of the things that your head can't cope with is writing C code that the compiler can turn into good stuff on an almost infinite range of target architectures.

I can't believe I just said that. But this isn't the 80s anymore.

N-Gage goes live

If you're lucky to have one of the supported devices, go ahead and check it out. I tried to install the N95 package (unpacked just like the first access method, of course, as they used the same sort of "protection" again) on my N93, but couldn't start it. However, it might work if I remove the first access package first. We'll see, right now I feel more like going to bed.

It's interesting to note that the N-Gage team keep mocking the poor 3rd party developers by using an excessive set of capabilities for something as simple as the installer application. Alienating 3rd party developers seems to be very popular in the Symbian camp these days. And "security" is obviously just a buzzword to them.

N810 wimax

Suddenly it's all starting to make sense: the big, ugly N700/N800/N810 with their limited connectivity and all. Suddenly it's not just a testbed for a crummy UI platform in a crummy device over at the Nokia factories, now it's starting to look useful for real. With wimax, it's no longer just a surfboard, the line between surfboard and phone is blurring.

More amazing analysis of the smartphone market

Too bad you have to pay to read all these studies from analysts, because the news items that they result in are almost always to vague to really say anything. Read this one. This one says that Linux has 15% of the smartphone market, with Symbian at 50% and Windows Mobile at 18%. I don't even know what they mean by "smartphone", but that's beside the point, because I see no way to count that'd give you those result. I know Gartner doesn't count S60 phones as smartphones, which means Windows Mobile gets a much larger market share than in other places. But with that sort of criteria (I think they require qwerty or touchscreens or something), how many Linux smartphones are there? The Japanese ones maybe. Anyway, this just doesn't reflect reality, unless you define "reality" as "what's reported by some analyst".

Now, with those figures in mind (Linux at 15% and Windows Mobile at 18%), doesn't it sound a bit weird when it says, later in the article, that they think "that it [Linux] will emerge as a worthy competitor to market leaders Symbian and Windows Mobile"? I'm starting to think that "15 percent" is a typo.

I also have no clue what they mean when they say that "Motorola's recent troubles in the handset market have coincided with its increasing emphasis on Linux over Symbian". When has Symbian ever been an important platform to Motorola? They released a few Symbian phones back in 2003-2004 (A920, A925 and A1000). Then they took a break for a few years, and have now released a few new ones (Z8, Z10). Could someone please explain to me what they're talking about?

An advantage for Linux, says the report's author, is "the waning interest in Symbian's monopoly", which I've never heard mentioned before. Symbian has about 70% of the smartphone market (using more conventional definitions of "smartphone"), but I've never heard anyone complain about that. That's still just 7 or so percent of the total phone market. But sure, this is an analyst speaking, not a simple developer like me. I agree with the report that Linux will probably grow a lot in the phone market, and that Symbian will lose in the long run, but apart from that, I have a hard time finding anything else to agree on. To me it just sounds like a manifestation of incompetence.

This reads like a report that was ordered by some commercial Linux distributor. I don't see why any regular Linux developer would be interested, as all the Linux phones so far have been closed devices. People who are into developing for Linux tend to like the open nature of the system. If you lock it down, and don't even give developers access to native API:s (Android), there's no difference between developing for Linux than for any other platform. I'd love to have an open Linux phone, like the new OpenMoko one (but preferably not as big, clunky and ugly), because then I might actually consider writing code for it if I got the idea for some cool app. But having a closed Linux phone is of absolutely no use to me. As it is now, Windows Mobile and even the security locked-down Symbian are more open platforms for mobile phones than Linux is.

Security is all the rage these days...

...and people just love to talk about it, although they don't have a single clue.

I recommend you read this article (and the comments) over at All About Symbian. The article is written by a guy who has no knowledge about computer security, but has made it his life's mission to claim that Symbian OS is secure, as in completely unbreakable. Not that he actually has any knowledge to back it up. He just assumes that as there's a security system and there hasn't been any dangerous malware in the wild yet, it's secure (whatever that would mean, in an absolute sense, in a system that just oozes with connectivity and consists of millions of lines of code). Because Symbian told him so, and they should know, I guess.

Now, the case is complicated a bit by the fact that it has recently been proven that S60 phones can be hacked. It's not a case of dangerous malware. It's another aspect of Symbian that has been hacked. But apparently, if you have to be an "uber-geek" (have a look in the mirror sometime!) to hack your own phone, it's not a real breach of security. And well, it seems that if a vulnerability in a security system is discovered by an amateur, it's not a real problem.

This guy's been doing this for a long time, because he's a fan, just like the Michael Jackson fans who know that Michael Jackson would never do anything improper with a child, because he writes such wonderful songs (oh, and he's a married man). Now and then I've been thinking of putting on my black hat and come up with something nasty, just to shut him up. But you know, he's just some uber-geek in Britain, so why would I care?

The sad truth is probably that the only reason why no real hacker has bothered to hack Symbian yet, is that it's such an incredibly unsexy platform to work with. That's its most important security feature. The iPhone's webkit based browser was easily hacked, so no one should be surprised if that could be used as an entrance to S60 phones as well. And while the browser doesn't have a full set of capabilities, everyone in the know should know where to start looking for local exploits to escalate their privileges. I'm not telling anyone to do this, but it would be a bit fun if someone did it, just to see what mr. Litchfield would have to say about it. I guess in his view, it wouldn't be a real problem if it didn't empty his bank account, and if it did, I guess it would be the bank's fault, rather than Symbian's.

And the stuff that some people write in the comments... Isn't it damn easy to have strong opinions about things that you're clueless about?

Rant mode turned off for now.

Mobile phone market shares in the USA

I was under the impression that Nokia sold almost no phones in the USA, when someone told me that they actually sell a lot of phones there. I tried to find some numbers, but couldn't, until I suddenly stumbled on this article today, with the following listing (Q3 2007):

Motorola: 31%
LG: 17%
Samsung: 16%
Nokia: 11%
Sanyo: 4%

So Nokia do actually sell quite a lot of phones in the USA, although 11% is a bit pathetic, compared with their worldwide market share of around 40%. Finding Motorola on top isn't a surprise, people prefer to buy local products, so Americans buy Motorola, Swedes buy Sony-Ericsson and I suppose most Finns buy Nokia. (Now, I'm sure that quite a large a portion of these phones are branded as Sprint or AT&T, which means consumers have no idea where they're "made".)

A peek inside the N-Gage first access SIS file

I published an article over at NewLC, describing my observations from unpacking the first access N-Gage SIS file. If you're a Symbian nerd, chances are you'll find it interesting.

Quick OpenMoko update

When you're too lazy to write yourself, you can always link to other people's blogs. Marcin Juszkiewicz provides a quick rundown of the main OpenMoko happenings during the last year. Sometimes, when you're in the middle of things, you need to take a step back and look at the bigger picture, to see that you're actually making good progress. It seems to me like a lot's happening with OpenMoko. The blog post reads like a postcard from a better world, where people help each other, instead of competing.

First impressions of the new N-Gage platform

I finally got around to trying out the new N-Gage platform today. The First Access package is only available to N81 users, but the "protection" was very weak, so I could try it out although I don't have an N81. My primary interest wasn't in the actual games, but in the user experience of using the platform, use N-Gage Arena and install games, like a typical user would do it, ie. on the phone itself. (It might come as a surprise to some (engineers like me), but the process of downloading a file, using the web browser on a PC, and transferring the file to the phone, is way too complicated for most users.)

After installing the SIS package (which takes 20-30 button presses or so, because of the irritating S60 installer and the fact that there are several embedded SIS files in it), I fired up the N-Gage app. At first I assumed it was a j2me midlet, because of the slow startup and the sluggish UI, but after some further investigations I'm not so sure. Anyway, it takes a good 10 seconds or so to start, which really doesn't give you a good first impression. They've tried throwing in some animations in the UI, to make it look cool, but it looks quite cheap. The graphical design looks almost decent, but doesn't look as if it had been designed by professionals, but more like something that might have been impressive if used in a C-64 demo.

Anyway, I registered an account at N-Gage Arena, which was surprisingly painless. They've done a good job here, and it should pay off, because this is not a good place to scare off potential customers. Browsing the available games is also quite a pleasant experience. There's not all that much information about the available games, but who needs that when there are free trial versions of all the games? This also seems like a very good idea to me, and should be a much more clever business decision, than doing it the way the operators do it, and just try to squeeze money from their customers with cheap movie licenses. Nokia might not be the coolest and sexiest company in the business (just look at their phones, hehe), but there's a good reason why they're so successful, and I'm afraid the best word to describe it might be "innovation" (sorry about that, I'll promise never to use that word on this blog again). Or just good engineering? Possibly.

The whole N-Gage platform is very well thought out, and I believe that by finally making games available to customers in a simple way, that isn't bound to piss them off, this could be a new start for the whole mobile games business. It will of course depend quite a bit on the actual games as well, but what I've seen so far looks good, and it seems that they've managed to put together a nice mix of casual games and games that should appeal more to typical gamers. The last group of people might not be impressed by the idea of using their phones, instead of their PSP:s or DS:es, to play games, and the input capabilities of the phones are quite horrible, but I'm guessing it's just a question of time. Nokia are good at long-term planning. They didn't give up on the idea of the phone as a gaming device just because the original N-Gage wasn't as successful as they had hoped. It wasn't a bad idea, it just wasn't very well implemented.

(I'm of course also planning to have a closer look at the games that are available, at the N-Gage Arena web site and at purchasing games. Hold your breath!)

Engaged

The new N-Gage platform is finally live, and the dedicated fan-site All About N-Gage are starting to post their first reviews of games. Creatures of the Deep seems like a high quality game, although we militant vegetarians at the Sputnik Mobile blog don't approve of such activities as fishing. Especially not for fun. Bastards.

The new N-Gage is the most interesting development in the mobile phone games business for a very long time. Nokia's first attempt was quite feeble. The second one should be an improvement, if they've learned their lesson.

Mobile World Congress

Mobile World Congress (love the web design :) next week, which company is going to demo the most convincing iPhone killer?

Land of the free

Here's a report from an outsider on what the mobile market is like in USA. The iPhone is changing the game, sort of, but not as much by changing the American people's perception of mobile phones, but by making other phone manufacturers concentrate a bit on the UI, for once. The iPhone seems to be failing big time in the UK, which shouldn't come as a big surprise, because in most ways it's not really that great. You see, the UI has never been very visible to the customers, and that's all the iPhone has. Except for the amazing brand, and while that might be huge in the US (where the iPods have something like 70% of the mp3 player market), it's just not the same in the rest of the world. And Apple still haven't announced a 3g iPhone, as all European Apples fans are hoping for.

Go Yahoo go

I read a blog post about the new Yahoo Go mobile thingy, and decided to try it out. It's another one of those widget frameworks that everyone's so into these days, written in Java to cut development costs and increase user dissatisfaction. I was quite impressed by the performance, though. Sure, there's no reaction for the first half second after you press a button, but then the boring animation is actually quite smooth.

There's not much content there, and the app itself is quite unpolished (I didn't expect it to greet me with Italian as the default language), but it's one of those "betas", as popuplarized by Google. The user experience is quite reasonable, once you get past the rough edges. It's just slow, not insanely slow, and it didn't crash once in the five minutes I spent with it.

I guess we'll know in a couple of years what happened with all these competing widget platforms. I think Yahoo might have to convince some mobile manufacturers to include their platform in their phones to really get into the game, but they probably have already.

And while Yahoo are busy with Java and widgets, I keep working on stuff that responds so quickly to your input that you don't feel like your phone is your arch enemy...

10 years

Linux Weekly News turns ten this year. I started using Linux about that time, so I've been with them from the start. One of the news pieces from 1998 was that the latest version of the Linux kernel would require 8MB of RAM to run. At that time my newly bought PC had 32MB. My current phone has 64MB. That PC had a 200MHz CPU. My phone runs at 369MHz now. A very rough estimate would be that the weight of my phone is 2% of that computer, and about 0.1% of the volume. We live in a scifi dream.

Microsoft breaking new ground

I don't know if this is real, but it's supposed to be about a leaked Microsoft document on the spiced up UI in Windows Mobile 7, to be released in 2009. Revolutionary? Groundbreaking? Yes, for Microsoft maybe, which means there's not one single new idea in there. But it's about time the Windows Mobile UI started looking at least a bit as if it was actually designed for mobile devices, and not just a scaled down version of the desktop system.