Opinionated comments on mobile phone industry news

India is, as we all know, a very quickly evolving market. What I find especially interesting is that in many ways the mobile phone replaces the PC as the "Internet interface", so quickly also services that we hardly dare to deploy in Europe and USA (like payment services) are very popular in India.

Mobile phone production growing in India

Web 2.0 startups must go beyond PC to take off

Still looking good in sales of mobile phones and phone subscriptions, with an expected 3.25B active subscriptions by the end of the 2007.

'"It took over 20 years to connect the first billion subscribers, but only 40 months to connect the second billion," said The Mobile World Co-Founder John Tysoe.'

Global mobile phone use to hit record 3.25 billion

If you look att Yahoo! News right now almost all news there are about the iPhone. Why are journalists so eager to help Apple promote products? Just because they have an iMac on their desk, or that they think the iPod is the first and only music DAP, and the iPhone is the first and only useful mobile phone / mobile computer?

I acknowledge that the iPhone has some neat features, but primarily breaks new ground in the stifled operator/carrier controlled telecom business. The latter part is key, as this industry requires new business models to evolve.

You might want to read the following if you have iPhone shopping anxiety. There are alternatives.

Apple iPhone pitfalls - limitations and some things you may want to consider before you buy

Helio pits Apple's iPhone against Ocean

I try to avoid writing about the iPhone launch, as it gets enough (read: too much) free press already. This note caught my eye though, as it delivers a few facts - mainly omissions, but some good things too - that made me wake up (it's 6:30 am...).

iPhone facts from the first reviews

Paperless Tickets Hit The Right Note At 2000 Trees Music Festival shows that paperless ticketing (of course done via mobile phones) is more convenient than paper tickets, so the "save some trees" argument is not needed. A few benefits mentioned at the tixmob site:

• You told them tickets would be one price but you forgot to add the delivery charge.
• You said you would collect the tickets but you forgot you can't rely on the post to arrive on time.
• You can't leave work early enough to beat the queue at the door
• And you have at least one mate who always turns up late leaving you hanging round in the cold.

Here's one of my own (the most important one in my opinion):

• You order and pay your tickets online and don't need to deal with them further until you are at the venue.

It's even possible to order tickets for others, and their tickets are sent to their mobiles. Neat.

1ft Startup provides very hands-on advice for development of MIDlets, taking into account the device and KVM fragmentation etc.

Sun should read this for a reality check.

J2ME Porting: Entertaining a crowd, part 1

J2ME Porting: Entertaining a crowd, part 2

GeoTagging Revisited indicates the state of geotagging of photos via mobile phones, providing ShoZu as an example. There's room for improvements though in the handling of the GPS device and tagging of EXIF. It only mentions using a Bluetooth GPS module, but I guess it works with internal GPS as well.

More pics of the GPS enabled Sony Ericsson Z750i shows how the Z750i will look like. It seems a bit more mainstream than Nokia N95, and might therefor also cost less. Maybe more on the level of 6110 Navigator. It's as far as I know not a Symbian phone.

This is one of those moments where I sense marketing and convenience has more significance than technology: when it comes to music audio quality, and secondarily mobile application platforms.

I've been talking about this subject before, but I got a reminder when I bought a guitar effect that didn't cost much at all and that handles 24 bit / 96k samples/second audio through a 32 bit DSP pipe. The sound is so clean there's not a trace of any distortion (except the intentional one). Those are the kinds of figures I'd expect from modern-day distribution formats of music, not the current "less-than-CD" quality. I know, most people don't care, and again marketing is way more important than technology, right?

Well, actually no, because that's like saying that an orange but useless hammer is better, and should be more expensive, than a useful but less colorful one. Now, people making tools know that marketing and technology go hand in hand, so you will actually find orange hammers, but that are also of a very high quality.

The cost of storing music on a server, even at a high quality, is relatively speaking none, so the cost is in bandwidth and marketing. For third-party services like iTunes the music companies take almost all of the revenue (which is ironical, considering they at the time of the launch of iTunes didn't believe in this distribution model), so there's not that much money left for Apple. By now though, through Apple's power, I believe they've been able to negotiate better deals.

Allofmp3 does it right by offering several different formats with different quality, including FLAC, that is a non-lossy compression of CD audio. Unfortunately Allofmp3 is not considered 100% legal, and we are still talking CD quality.

In commodities business where there's no technology evolution to talk about, like beverages, fast foods, clothes etc, marketing and brand is everything, as anyone can make similar products, yet possibly not at the same price for efficiency and volume reasons. Still, note that you as a consumer funds the marketing, so even if the logistics are very efficient, there's a noticeable amount of marketing spending enveloped in the consumer price of the product.

In terms of music distribution it's as if the commodity model applies, as it seems more important that music is quick to download than actually sounding great. It's a pity, as already the CD was a bad compromise in terms of audio quality, and future audio formats at least need to be better than CD. All the same, "lesser-than-CD" formats are now marketed as "CD quality" as if that was the best there could be, and as if that was actually true.

Another area of technical de-evolution is in application platforms (to take a currently hyped topic): The promoters of Ajax mean it can replace Java and Flash Lite on mobile phones when it comes to local applications. It can't. All with any level of technical competence knows this. Still, it's now touted as the saviour of mobile applications, again pushing marketing more than technology. It seems people don't even know that Ajax is just a "mash up" of existing technologies, with a few tiny additions. You can't even take a photo on a phone via Ajax, nor can you make anything that looks and behaves like a game (except possibly "Hang Man"). Ajax is clearly hyped more than it deserves.

So a call to action to all involved in the electronics, software, and specifically mobile industries: There's no "d-e" in evolution. Do some actual technology evolution.

Are we in a new dark ages / knowledge white spot, where marketing replaces fanatical religion as the controller of the mindset of the masses?

And yes, I'm one of those that can actually hear a difference between LPs, CDs and MP3s, and the difference is remarkable when you move from a lesser technology to a better one (in other words MP3 --> CD --> LP). I definitely don't want the physical format of LPs back. I'm just saying there are technologies that can achieve the same level of quality and dynamics as LPs.

Tegic's T9 is the predictive text input engine most used in mobile phones, with Zi Corp's eZiText as a far off second.

I don't think it's a bad buy for Nuance. In a sense Tegic was a strange bird in AOL's zoo, so this might actually be good for Tegic and its business.

T9 predictive-text input developer purchased for $265M

As you know, Java Wireless Toolkit provides emulators for CLDC/MIDP and also most manufacturers' emulators are directly based on WTK with added phone-specific profiles.

Matthew Kumar instead proposes to make your own emulator to speed up the development process. This was done by implementing a thin layer on top of Java SE that emulates the CLDC/MIDP-specific classes (and I guess also other JSR classes, based on need). As then no pre-verification is needed, and potentially no compilation either, the building process becomes faster.

You even get the emulator for free under the new "WTFPL" license. Check out the readme for an explanation.

This is a trial no doubt, where a limited amount of videos have been converted to mobile format.

Not Quite YouTube To Go

Noted:

• People want to get access to YouTube, they don't care who's doing the plumbing, so exclusivity deals can only create consumer aggravation. Hence it's good that Verizon lost the exclusivity.
• Phones should support de-facto standard video formats, like MPEG, DivX and XviD. I wonder who came up with the 3GPP format. That has no rationale what-so-ever.
• It says EDGE is slower than 3G. In theory yes, but my experience is that EDGE is well in line with 3G. Maybe because no one's using it, so pipes are not clogged, so to speak.

'In short, it's not YouTube--not now, maybe not ever.'

I think that not until videos are converted and cached on-the-fly can there be a real/full mobile YouTube, as so many new videos are published every minute. Converting videos is a very slow process though, so a predictive multi-terabyte cache (e.g. new videos are cached unconditionally etc) is needed, or else people will just leave.

Other notes about this:

YouTube launches Mobile Portal for all kind of Mobile Phones

Verizon exclusivity ends, YouTube Mobile opens the gates

YouTube Launches Universally Accessible Mobile Site

Designed as part of a headset, the HGE-100 GPS receiver is quite subtle. It's also less easy to forget than an untethered module. Other advantages: no Bluetooth configuration and no battery to charge. A drawback is that you always need to bring along the headset.

The question is how to access it from applications. Most of the phones listed have only Java, and they don't have Location API (at least officially).

The NetBeans team provides some pointers for how to use NetBeans and Mobility Pack to compensate for device differences in Resolving Java ME Device Fragmentation Issues.

I think many (not the least Sun) have a too idealized view of the fragmentation issues, that also consist of specification sketchiness and mis-interpretations, as well as pure implementation bugs. This is not solved in a jiffy due to all the KVM variants, and mobile platform variants.

At least the article provides means to handle those differences in an efficient manner, but in some cases even rather basic functionality doesn't work on certain phones, and even very new ones.

I don't care much about Apple's offerings, but I have to admit they are clever in how they look at the whole infrastructure / ecology around an offering, and how they more than ever look at how they can make the most money by extending beyond the mere gadget, something I think they really got rolling with the iPod and e.g. the 10% cut on third-party iPod accessories and the (at least partial) iTunes lock-in.

iPhone users will need iTunes account

IPhone Requires iTunes, Apple Says

Wibree is a specification developed by Nokia for a short-range wireless digital communication with shorter range than existing Bluetooth yet with 1/10th the power consumption. It's hopefully also way easier to use and more secure than Bluetooth, but I haven't seen the spec.

The key here is ultra low power, so that e.g. watches and other tiny gadgets can participate in inter-device communication.

Wibree is now Bluetooth's ultra-low power wireless standard

Wibree Becomes Part of the Official Bluetooth Specification

Nokia's Wibree joins Bluetooth

Nokia and Bluetooth SIG Partner on Low-Power Wibree

The sixth and final article in the PIM API series is going through the PIM Fields. The page also lists the previous 5 articles on the same theme, all of them written by Enrique Ortiz. They are thorough, so you should be up and running after having read them.

Here's a good way to save on MIDlet size (and hence also downloading time and cost): Implementing a Local Cache to Manage Resources

Apart from images and other "normal" resources I would also suggest localization string lists could gain from this: Then the (likely only) needed language would be downloaded when the application is started for the first time (simply choosing the desired language from a list that's part of the MIDlet). It could also be used for application skins/themes etc.

When it comes to data that might need updating over time it's even more fitting. E.g. landmarks, users, contact information, additional levels in games, interpreted scripts (a la widgets) that could completely change the application behavior, etc.

Most non-game MIDlets are service frontends, so they anyway communicate via switched data (HTTP or other), even if they were installed from a PC, so updating resources this way is already prepared for.

It should be noted though that not all phones have tons of RMS space. In the early days of MIDP it could even just be a few kiobytes per MIDlet, so for maximum phone compatibility this still needs to be used cautiously.

Over time, and with flatrate subs etc and faster networks, resources can and will be downloaded each time the MIDlet is started and instead stored in RAM. I use that approach in a service frontend type MIDlet I'm currently working on, where the list of users, landmarks etc could change at any moment, but that still gains from semi-permanent local access while the application is running.

Whatever the means for storage, such downloading and access could be completely abstracted (which is also described in this note), so that when a resource is needed the resource class would figure out if it's already cached or needs downloading because it doesn't exist locally or has expired. It should also check if it runs out of storage, and if so delete the least accessed resource(s), to make room for the new one(s).

If there had been a business in selling code libraries for MIDlets someone would have already developed this class and licensed it to development houses. Unfortunately it doesn't work that way.

(via CEO's Mobility Weblog)

This is about about a draft version of the LiPS specification for a standardized Linux-based mobile application platform.

The very fact that there are a number of organizations/forums working on such a standard/toolkit means there might not be a standard after all. This situation has also plagued Unix on desktops for many many years, but today it seems all agree that Linux is the Unix dialect to use there. Now it's time to define what's needed to make Linux house-trained for mobile use.

Interestingly they've chosen GTK instead of Trolltech's Qt/Embedded. It also indicates there will be a basic set of applications a la Trolltech's Qtopia. Such functionality is fundamental for this to be possible to call a Linux-based mobile application platform. It's certainly not enough with UI Widgets and Linux core.

Draft of Linux Mobile Specs Released

Mobile Linux group releases first specs

Mobile Linux Specs Due

Mobile Linux group releases first specs

Update 20070613: Here's more on the subject of support for third-party applications:
So that's what he meant by "opening" the iPhone?
Developing Applications for the iPhone
Editorial: The No-SDK Cheat


Here are some recent news about the Apple iPhone.

iPhone to Run Third-Party Web 2.0 Applications
Apple's iPhone open to software developers
Apple announces third-party software details for iPhone
Of course this is not the same as native locally running applications, but provided a good UI design they might at least look and behave like they were. That of course requires that such applications (that are really bookmarks/URLs from the point-of-view of the browser) as well as bookmarks in general are visible on the main screen or a very short distance away from it. Otherwise people won't find them later. Hopefully they've integrated the browser with the main screen, so that users won't even know the browser is being invoked for such applets (ajaxlets?).

To me this means the iPhone doesn't officially support third-party applications the way I would define such applications - network-untethered fully local applications - but through classical Apple "marketing magic" they make it sound like it does. For the most part Apple does excellent packaging of products, but this sounds like an excuse for not supporting real local applications. I don't think it's an afterthought though. I sense Apple will offer e.g. community services as well, maybe as an extension to iTunes.

Seen differently, on mainstream phones Java/CLDC/MIDP would be the sandbox for third-party applications, on the iPhone Javascript/Ajax is the sandbox. Please note that Ajax of today doesn't provide access to much local phone functionality, but hopefully Apple has remedied that.

iPhone APB: Walt already has one
Does he really, or is he pulling Engadget's leg?

iPhone to ship on June 29th at 6pm

Europeans scrutinize Apple's control tactics
"in order for the iPhone to function correctly, there is a requirement for Apple servers to be placed deep in the operator's network,"
Hmm. Why there and not at Apple over an encrypted link? Load balancing reasons?

January to June: the iPhone's evolution
Finding tiny differences between early prototypes and the current status.

iPhone Sales Training Workbook reveals GPS, MMS and IM details

It sounds great at first, until the note says these are features that are missing. Thanks Apple.

YouTube plans to use mobile next year

But why not already this year? It's hardly a cost issue, and there are third-party services that convert YouTube videos to 3GPP format etc.

The Big Fight 2007: Series 40 vs S60 is a strange review, where two phones that look almost exactly the same run completely different mobile platform, Series 40 (Nokia 5300) and Series 60 / Symbian OS (Nokia 5700).

The price difference is also peculiar considering it's mainly the hardware that costs. Of course the 5700 needs more Flash memory for the OS footprint, but that still doesn't motivate this difference: 'Verdict: The 5300 "wins" in that it costs half the price of the 5700, but in the real world people look for value rather than just price. You get what you pay for.'
I'm note so sure about the value bit as a motivator, or rather what the author considers value: Price and e.g. battery life are concrete values too, maybe more important than full PDA capabilities, that you can't fully utilize without an alpha numeric keypad anyway.

The presentations from JavaOne 2007 can now be downloaded:
JavaOne Online Technical Sessions - Java ME

Check out e.g. the following:

Catch This SpeechEvent: Recognition and Synthesis on Devices
Finally speech support! Handsfree applications need this. Not in phones for some time though.

Java Platform, Micro Edition (Java ME): Optimizing Midlets for Size and Performance

Bring Map and Navigation Capabilities to Your Location-Based Applications with JSR 293, Location API 2.0

JSR 248: Taking Java Platform, Micro Edition (Java ME) to the Next Level

In The Price of Speed Mark Lam provides an analysis of and arguments for Just-in-Time compilation in MIDlet KVMs. Main arguments are:

• JIT compilation doesn't take much memory nor performance
• It can generate considerably faster code
• It works on only parts of the code at a time, so there no need for a complete Java to machine code conversion, which would expand the code a lot

Please note that I try to be as objective as possible about the following scenario. My point is to show how easy it is to think erratically in this industry, based on a lack of communication with and information about the potential users. I'm certainly not bashing cellity. I think they are doing the right thing here. Rather I bash myself for not thinking straight.

The whole point with releasing a Twitter client (Abiro Jitter) was to build brand and awareness of Abiro in general, and because I could, in a very short time. As it was a labor of love I'm not bitter (maybe just a little...) about the lack of results, but I realize I could have spent my time on better things.

I had mistakenly assumed that lots of people would download Jitter, as Twitter is supposedly so popular and sending tweets 'at the spur of the moment' from a mobile phone seems ideal.

Not so. In this case it's not only because my application might be crappy, which it's not in my opinion, as a 'competitor' has the exact same issue and very similar features.

If you search for 'twitter' at GetJar you only get two hits: Abiro Jitter and cellity tweeter. Given that, it seems either one of these applications should be downloaded like crazy, but that's not the case. Interestingly though cellity freeSMS is downloaded way more. That it at all shows up when searching for 'twitter' is because cellity cleverly uses that word in the freeSMS description. When I checked other 'free SMS' type applications they are all downloaded a lot, even some of those that require a subscription.

cellity freeSMS doesn't use Canvas anywhere (similar to tweeter and Jitter), so there's no nice look-n-feel, making the application look the way MIDlets tend to do when using the high-level UI. Of course users don't know that until they've downloaded it, and the animated pictures shown at GetJar are faked to look good. As similar pictures are used for tweeter, this is though clearly not driving the decision to at all download either application. The reason I mention it is that many MIDlet developers think using only Canvas is a requirement for user satisfaction. I agree the logic is limping, as a lot of downloads doesn't necessarily mean users are happy with it.

Also, tweeter and freeSMS are designed for the lowest common denominator level of CLDC and MIDP. They are of course not tested on all phones. Rather by using this subset they inherently are compatible with most phones. I also use this approach for free applications. It's when I release commercial software I'm careful about this, as customers don't want to buy a product that might not work on their specific phone.

To make an attempt at a reverse market analysis based on this (I might be completely wrong again; maybe I should actually talk to some people): Young people don't have money, so they don't want to pay for anything, and as they need to send a lot of messages on a daily basis to stay on top in their social lives they need free SMS, and the main point with mobile telephony (and possibly IT in general) for them is person-to-person communication, hence they need an application/service that provides free SMS much more than a way more esoteric thing like a Twitter client. Of course communicating over the data link also costs money, so calling it free is not really true, but way less costly than normal SMSs.

cellity doesn't get any money from neither tweeter nor freeSMS, but they register phone numbers (not really needed for tweeter) which potentially is an asset, and cellity is an MVNO, so that's where the long term money is. Hence they use tweeter and freeSMS as door openers. That's similar to how Skype uses its free client to lure in customers to paid services.

One other thing I noted was how technical terms stick like brands. Telecom folks know that SMS is a very specific technology, clearly specified by 3GPP, but of course most people don't have a clue how it works, and don't care how it works, so anything that transfers short messages could be considered SMS.

So, should I develop a free SMS solution? Technically I of course could, and quickly too (also the server side), but then there has to be an underlying commercial service that would generate revenue. Just releasing a free solution wouldn't make sense. That would be like advertizing for a commercial product that doesn't exist, unless (as always) acquisition would be an option.

In 'Why stand-alone GPS devices are doomed' In-Stat even goes as far as to say 'Mobile phone operators now have the ability to market a downloadable navigation application that is just as good as, if not better, than personal navigation devices', which at the moment is nonsense, as it's not just about software, but also a large screen, navigation-specific buttons etc. Also, most mobile phone based navigation solutions constantly download new maps, which is good from an up-to-date perspective, but could also become very expensive, yet it's very clear the trend is there.

A few alternative responses a geek might give to avoid dating invitations (as if that would ever happen):

"I need to catch up on replying to all spam."

"My overclocked 3D card is overheating, so I will install a new freon cooling system."

"My mouse finger hurts."

"They say there's something on TV tonight."

"I'm waiting for my first million dollars, doing nothing."

"Someone on MSN wrote they might contact me back some time this year."

"My paypal account is empty."

"I need to house-train my lizard."

"Writing tweets takes all my time, as I'm constantly doing new things."

"I might be abducted if I go outside."

"My randomly generated horoscope says dating is not for me."

"I'm testing how many sugar cubes I have to eat to make me woozy / puke / social."

"My mother says I have to stay at home."

"I need to develop mobile applications that no one wants to pay anything for, and that won't work on more than 2 phones anyway."

Sorry, I got a bit carried away at the end.

Elevation Partners acquires 25% of Palm. Also, the previous CFO of Apple and the GM of the Apple iPod unit join the board of directors.

This sounds like the starting point of a major re-structuring / re-focusing coming up.

Has Palm dropped the ball?

Palm sells 25 pct. stake to private firm

Palm sells 25 percent stake to equity group

Palm Sells 25% of Itself, Gains 2 Ex-Apple Execs

Palm to add former Apple CFO, iPod head to board

Reorganization Creates 'Palmapple Inc.'

According to this information there's a chance Apple will soon launch a developer kit for iPhone applications, and it also indicates the iPhone runs a slimmed down version of OS X, which should make it simple (or simpler) for current iMac developers to port applications.

iPhone to get software development kit at WWDC?

Developers see possibilities in iPhone apps

The official launch date of the iPhone is June 29.

You've probably seen such pictures already, photos of cats with the cat version (supposedly) of English imprinted, with some funny tagline. I haven't seen many such pictures where mobile phones are used, but I'm sure that will come soon.

I’M IN UR NEWSPAPER WRITIN MAH COLUM

(via Textually)

SlideShare is about sharing presentations for viewing via web browsers. It's an excellent place for conference material, and it seems many use it that way too. A few examples:

Java Micro Edition Overview, by Edoardo Schepis
Confirms many of my own opinions about Java ME. Note e.g. slide 7 where it's shown Java ME (in this case of course the MIDP flavor) and BREW completely rule as mobile gaming platforms. Slides 15 and 16 express a few of the headaches Java ME developers face. In the competitive comparison on slide 17 I doubt Flash Lite afficionados would agree that Flash Lite is harder to develop for than Java ME.

Java ME - MIDlets Development Workflow, by Edoardo Schepis
Describes the CLDC/MIDP flavor of Java ME and the process of developing MIDlets. It's in favor of NetBeans as IDE.

Twitter, the useless son of Web 2.0?, by Alex Mou
I know Alex from before, as he also made a Twitter client in Java. Interestingly Alex is from China, yet he made a translation of this presentation from Italian to English. The presentation highlights what things (other than numb skull reporting of what one is currently doing) Twitter can be used for. Just see it as a system for community-based message exchange, and the amount of possible applications become vast.

Mobile Gaming: Sinking one’s teeth into new metaphors for mobile gaming, by Nicholas Nova
Some interesting thoughts on mobile gaming evolution. E.g. it mentions blowing on the phone to control the speed of a boat in the game (it of course detects the level of noise from the microphone, but all the same).

Mobile 2.0 , by Thiyagarajan
Actually this is not much of a presentation, but it contains 2 of the misconceptions people have about mobile service evolution:

• That Mobile 2.0 is something very different from what is already available (yet admittedly not much used).
• That this would be far off into the future.

The use cases and technologies are already here. What's lacking are profitable business models, good service designs, marketing and spunk. The business model of Web 2.0 and possibly also Mobile 2.0 seems to be 'get acquired by Google, Yahoo or some other of the big guys'. That's not the way to run a business.

Mobile Ajax FAQ goes through what Ajax in a mobile context is all about. Ajax is also compared with Java ME and Flash Lite, and of course Mobile Ajax comes out on top (the FAQ is written by the promoters of Ajax, so ...).

What's not mentioned here is the lack of access to phone features from Ajax. Just handling the UI and local data is in many cases not enough. For many services the complementing phone applications (or service front-ends) need to take photos, record videos, draw advanced interactive graphics at very high speed (read: games), communicate via Bluetooth etc. Still there are a lot of services that are helped by Ajax in providing a better user experience. It only needs to become more widespread.

The reason Ajax can be compatible across phones is its lack of functionality and lack of phone integration (Java ME and e.g. Symbian OS are on the other hand very much integrated with phone functionality). If you want to make something portable, and admittedly also broadly useful, limit it to the functionality all targeted devices support, in this case graphical output, textual input, Internet communication. Ajax still has the problem that adding it to embedded browsers takes years to deploy on the market. Therefor I recommend Opera to make an Ajax-capable Opera Mini. That way Ajax would be deployed much faster (over-night). There have been attempts to make Ajax-like widget engines in Java, but considering the amount of functionality a Javascript-capable web browser has, it's a pretty tough job.

Ajax is not used for anything advanced or performance-critical on PCs, so expect it to mainly enhance form/text/picture-based services also on mobile phones in the near term.

I'm not flaming Ajax, I'm just saying it's not a cure-all for all the things needed from service-complementing mobile applications, and it's definitely not yet in many phones.

GPS on a keyring makes it even more compelling to complement an existing phone with GPS.

Unless Audi keys are huge, this is a pretty small Bluetooth GPS module. By bringing it with you in the key ring you won't forget it.

'aimed at the majority of smartphone and PDA users'
I think they are missing the point, except for car navigation. For many other location services a normal phone with Bluetooth and Java (read: any phone purchased the last 2-3 years) will suffice.

£70 ($140) is not a nice price though, and 10 hours battery life is a bit too little. In comparison: Nokia LD-3W lasts for 15 hours active, 3-4 days inactive, and it costs less than this.

Regarding my previous note 'Accessories a bigger business than smartphones': They are at least very popular in Japan, and have been since mobile phones came into wide use. Here's an example of a web shop for such accessories (and many other things too; what about this R2D2 trash can?): StrapYa.