Opinionated comments on mobile phone industry news

Opera Mini is awarded Product of the Year by GetJar. Well-deserved.

Free Java, Symbian, Palm and Pocket PC Software

I stumbled upon ZeeMe which is a Java ME application that enables you to play Z-Code (Infocom and third-party) compatible text adventure games on your mobile.

The interpreter didn't handle the input field too well (I had to scroll down to see it all the time), but at least you'll get a feel for what kind of games people played in the 70's.

A mini version of Zork I (one of Infocom's blockbuster games at the time) is included, and with ZeeMe Packager you can build new games based on combining Z-Code games with ZeeMe. See e.g. IF Archive for lots of free games.

ZeeME

Similar to Skype, fring uses P2P technology to provide scalability. That said, mobile phones (with the current fees and speeds of mobile networks) are not suited to act supernodes, so I wonder how they've achieved this.

It can communicate with Skype Google Talk (as of now), and also supports text chat.

Even though the service is in itself free there can be a considerable cost in the data transfer, and depending on the efficiency of the voice compression, might be anything from noticeably costly to very costly.

If you have a flatrate data sub, then fring might be exactly what you need.

As the site's page rank is none, this must be a very new service.

As lower-cost mobile telephony is so important to most people I'm sure many more service providers will jump on the mobile VoIP wagon.

fring - it's fringing free !

As outlined here, most such applications are part of a client/server solution, where most of the intelligence is central and shared.

Even so, the examples mentioned from Nokia, Salesforce and Orative show that we've moved from complete ignorance to at least some awareness and deployment.

The deployment part is really the proof of the pudding: It's easy to theorize about new things, but when it's actually being used, to positive effect, then it's way easier to convince future customers and investors that this will fly (and not just may fly).

Business Applications Squeeze Onto Mobile Devices - Yahoo! News

Contrary to the conclusion in this note by MobileIndustry.biz I'd say that less than 10% is still pretty good, considering most people just use their mobile phones as ... phones.

It's like comparing with cars: "less than 10% of car owners buy bumper stickers". So what?

Well, it's not a "So what?" if selling mobile content should be (or even needs to be) the new source of revenue. Even with much better access to mobile content, and more relevant content, we won't ever reach 100%. Read the bumper sticker analogy again.

I believe though that most efforts should be put into making mobile data a hit in general, and establish that as the main new revenue generator for operators. I've mentioned before that to get there operators need to change the business model:
* Provide enabling services, not the services themselves
* Introduce flatrate plans
Without this we haven't even moved from square one.

Less than 10 per cent buy content through their phones - study - News /// MobileIndustry.biz

In-Stat claims many mobile phone users still carry around dedicated cameras (80%), music players (>50%) and PDAs (75%). Hopefully only when such gadgets are needed though.

My old 4 Mp digital camera is still much better than the one in e.g. the Sony Ericsson K800i, but even with that rationale 80% is still a very high figure. Using a cameraphone is nowadays very simple, so that doesn't sound like a show-stopper.

I fully understand the figure for music: Phones still can't compare when it comes to ease-of-use and amount of memory (when talking mainstream phones). Even so called music phones are in my opinion too hard to use, as music functions fight for attention with so much other functionality.

The most odd is the PDA/smartphone bit: If a customer buys a smartphone at all, why not select one so the PDA can be ditched? Frankly, most so called smartphones are not equal to a PDA (e.g. the Nokia N series, that don't have alphanumeric keypad nor a touch screen), which dillutes the smartphone term. Yet, there are good smartphones that can replace a PDA 1-to-1, like the Treo's, Sony Ericsson P and M series, Nokia E and 9xxx series etc.

There might also be other reasons altogether:
* Gadgets are purchased for status reasons. Photographing with the latest Canon or Nikon is way more status-inducing than using a lowly cameraphone.
* Consumers might not buy the phones for the mentioned features. Take e.g. any non-business Series 60 phone: It's considered a smartphone but most buy them as vanilla phones.

It will for sure become better over time, both in terms of capacity and hopefully also ease of use, which will put a dent in the statistics as well.

Study: for most, all-in-one phones aren't - Engadget Mobile

Just a headsup that 5.5 is released.

There are several new Packs:
* C/C++ Pack
* Visual Web Pack
* Enterprise Pack

NetBeans 5.5 IDE

As this article shows, there's a lot of untapped potential in the use of mobile phones, but some of the mentioned applications seem more wacky than useful.

Telecom Asia: The wildest mobile apps

It's actually really smart (but also a bit cynical, despite the positive outcome): Asking students at high schools to send in SMSs with the text "Diddy". The school sending the most messages got to see P. Diddy in concert.

Likely no cost for the radio station setting this up, nor for the operator, only for the students.

In total the radio station received 170k messages during the contest. It doesn't say if there was a premium on the SMSs.

This could easily be broadened to the general public, and is already used by e.g. TV shows. Then the cost is typically premium, sometimes as high as $1 or more.

It costs to set up Premium SMS, but if the volume is large and consistent over time the benefit is an obvious new revenue stream.

Text messaging catches on in music business | Tech&Sci | Internet | Reuters.com

Informa believes that mobile access to user-generated content (Web 2.0 sites if you will) will be big business, and that existing brands in this market will lead.

I believe so too, but I also believe there are possibilities for mobile-only or mobile-enabling services to pop up all over the place and take a slice of the market. E.g. ShoZu is such a service: Without actual image archives ShoZu would be of no value.

Actually UGC can change news media as we know it. If most people read news at aggregation/popularity sites like digg, reddit etc the original news sources might be left behind from an end-user perspective, serving those that choose what news to highlight. Maybe even TV, radio and magazine news will get an element of UGC to try to stay on top.

Of course mobile UGC also applies to blogs, chat and other services, including those that aggregate several of the known methods of UGC, like MySpace.

News | Mobile UGC set for major boost - NMA

Inspired Gaming Group has started offering vending machines that can also sell ringtones and games, top up prepaid cards etc.

Topping up prepaid cards this way sounds very practical.

Ringtones sounds a bit more counter-productive and uneconomical, considering buying and downloading mobile content over-the-air must be much better, right?

I can though think of a few advantages:
* It's more in the face; customers might be more interested in trying ringtones this way
* You use normal coins to buy content; more tangible and minimal risk
* The vending machines can be themed, like selling Coca-Cola beverages and Coca-Cola ringtones etc

Of course vending machines with only mobile content would be a very risky venture. What if installing all those expensive machines would give nothing in return? In this case though the mobile content comes as an add-on to more normal offerings.

Quote: Initial results have been positive, according to IGG, with the new machines taking twice as much money as normal ones.

Ringtones, CDs, games? Go to a vending machine - Sunday Times - Times Online

Eric Giguere provides a tutorial on the use of Web Services in Java ME. It's a good read as it also mentions alternatives to integrated support if not available in phones.

A quick look at j2mepolish.org shows there are not many phones that support the Web Services API in Java ME, but luckily there are the open source efforts kXML and kSOAP that can be used instead.

That said, if you use PHP or ASP on the server-side it in my opinion still makes a lot of sense to use HTTP POST if the transactions are simple. That will work with every MIDP, PHP and ASP version combination.

Service-Oriented Architecture and Java ME

Michael Mace at Mobile Opportunity made a survey for the visitors to his issue of Carnival of the Mobilists. I leave it for you to read. There's too much in there to summarize.

If you are a CotM host I recommend reading the free form comments at the end (Suggested features and beyond).

Mobile Opportunity: Who reads Carnival of the Mobilists?

Richard Spence at BlueTrail commented on my Java ME in the enterprise note, so I did some analysis what they are up to.

BlueTrail focuses on know-how in the technologies needed for developing mobile applications in Java ME and adapting enterprises with mobility in mind, and for access from Java ME applications or the embedded messaging or browsing applications. Simply put, what's needed to know to mobilize enterprise services.

Their Mobriz solution provides mobile-based surveys. Customers configure surveys via BlueTrail's site and users then access the site by downloading a Java ME application and fill in information in a custom form (based on the configuration). This could have been done in WML too of course, but with less user comfort, unless installing a new application could be considered a discomfort in itself.

BlueTrail's Mobile CRM solution has been provided to Red Bull, and there's a success story about this.

I get the impression that BlueTrail also does customer-ordered development, so it's not just off-the-shelf product sales.

I noted experience in J2ME Polish and based on the screenshots it seems the survey and CRM applications are using that. I've considered it myself due to the compatibility with standard MIDP (1 and 2) form elements and the possibility to completely customize look-n-feel even though it will work on both MIDP 1 and 2 phones.

Richard mentioned that the location application described on the site uses a central service for location rather than the Java ME Location API. The Location API is still only available in a few phones, so this is the way to go for now.

He also added a comment on the fact that Java ME is used and not Symbian OS etc, which is similar to my way of reasoning: The enterprise side is an interesting area, in that some of the applications we have done have only really become viable when we implemented them on a mobile phone. Had we suggested a windows mobile device or palm, I am sure customers would not have been interested as they would have seen the introduction of new devices as a barrier.

And here's what he had to say about fragmentation among Java ME implementations: Another exciting thing about enterprise Java ME is that the fragmentation issues are less pronounced. Why? Well often the users have the same phone model or at least the same manufacturer (Nokia!).

BlueTrail - Great Applications for Mobile Phones: Java, J2ME, WAP, MMS

Informa estimates it to generate $13B in revenue by 2011.

Quote: Moreover, Informa predicts that the inevitable move towards flat-rate monthly packet charges will result in the imminent emergence of larger files of up to at least 200Kb.

True (yet operators still don't provide this), but we are not talking 200 kbytes, rather 2 Mbytes per file. When users have real flatrate data the dam will burst. Downloading times is not a major issue, unless the battery runs out of juice much quicker this way.

Surely networks will be faster until 2011 as well, lessening the issue of downloading multi-Mbyte files.

Even more than on PC-accessed music services I believe single song sales will completely dominate, due to downloading times, but also simply because that's the general trend anyway, and because at many times the song will be used as a ringtone, which will also affect how music is being made (more focus on releasing singles ahead of complete albums etc).

I expect monophonic and polyphonic ringtones (MIDI or other formats) to have all but vanished by 2011.

As I prefer theme-based elaborate classical or progressive rock albums in at least CD quality (compressed to a non-lossy format like FLAC) this is probably not for me anytime soon. I don't think music download services will care about "my" demographic as most customers will be happy with the latest heavily compressed single songs.

AllOfMP3 is as all knows completely illegal, at least based on non-Russian law (they buy one CD, rip it, and sell it in any number of copies; doesn't sound very legal to me), but one thing they've done completely right is to let the customer decide what format and quality to download. The larger the file the higher the price. Providing this flexibility for mobile phones probably doesn't make sense, but I hope music services will provide this for broadband or hot spot download.

To support FLAC and similar you need a lot of memory, but as Flash-based players and phones now have 4 Gbytes, it's feasible provided you can use a PC as main storage for all your media. My iAudio X5 supports FLAC, but it also has a 20 Gbyte hard drive, so it can store several albums in this format.

Mobile music market to reach US$13 billion by 2011, says research firm

I know this is strictly off-topic, but as this is still not fixed I can as well mention it here. We are talking the very latest version 1.5.0.7, but it's been this way a long time. Is it only me that's experiencing these things?

Cut/copy is broken

Try marking text right-to-left instead of left-to-right or marking the URL on the address line. It very often fails and I end up with nothing cut/copied. After a few tries it tends to work though, but why not always on the first try?

Crashes

Firefox seems much more sensible to rogue printer drivers than other applications, as it often crashes when I try to print something via File / Print. All browser windows crash simultaneously, so it's clear they are not running in different processes.

Otherwise it's no doubt a good solid browser that seems more light-weight than Internet Explorer.

After all this talk about the mobile web it might be interesting to know (probably mainly for myself) what I access via my mobile phone. Nothing racy going on, sorry, and if there was I wouldn't tell you about it anyway.

I nowadays primarily use a Sony Ericsson M600i, which is not a normal phone by any means, so the possibilities are a bit broader than on a mass market phone, especially in the display real estate and text input departments.

In brief I mainly use it for daily news and email.

Most Swedish newspapers have mobile sites, and I access those simplest via my own site wap.abiro.com / Mobile News. Tabloids tend to report the same news, so it's often enough to access one such service. As the information is optimized for mobile access it doesn't generate many "megabytes".

I also check my email account this way (which is not really about the mobile web, but let's ignore that). My ISP supports IMAP, so the behavior is quite similar to what I get on a PC, and erased emails stay erased also on the server. With the M600i I can easily write brief email responses.

After I had browsed through the operator's portal when I got the subscription I quickly realized I didn't have any use of it. I'm not in the ringtone and wallpaper demographic (at least not in terms of buying such), I don't want to pay for news, and games and applications are much better provided by other services.

For application downloading I mainly use GetJar (wap.getjar.com) and ClickApps (wap.clickapps.com). Especially GetJar has a lot of free applications, primarily in Java ME. I host my own applications at both sites.

When unknowns call me and I don't get time to answer I sometimes check who they are via Hitta.se (wap.hitta.se) or Eniro (wap.eniro.se). wap.hitta.se is in my opinion better, as it's more intuitive. I would have liked such a service being integrated with the phonebook, so unknown numbers were automatically looked up when someone is calling. Broadband VoIP services can do this, and it should be possible to do on at least Symbian OS and Windows Mobile phones. Some assembly required. The main problem is that these services are regional, but there might be enough users that want this in one specific country.

I of course also use my phone for testing my mobile sites and mobile applications. Every time I change my mobile sites I test them via a real phone (not just desktop Opera or a phone browser emulator) as WML is very sensitive to syntax errors. Also of course I need to test Java ME applications on real phones, as the WTK emulators can't be trusted when it comes to real phone behavior.

That's about it. I Not much to get excited about I guess, but at least I do access information services via my phone.

comScore Networks has provided statistics on the use of mobile web services.

Note that no Asian country is included in the statistics, which is a pity. It would be very interesting to see how Japan, Korea and China stack up. Probably very well.

It mentions that Nokia phones are dominating for mobile web access in Europe, but that's simply because Nokia dominates, period. Same thing with USA where Motorola dominates.

That so many Americans access other portals than the operators' while Europeans mainly access the operators' portal is quite remarkable. Because in USA mobile web services are mainly accessed from smartphones and hence can provide a more full experience? I don't know, and comScore doesn't provide a convincing argument either.

Europeans More Likely than Americans to Use Mobile Phones to Access the Internet

Update: I contacted Ephraim and got this feedback regarding whether it's actually possible or allowed to set up high power Wi-Fi towers: "I believe that the FCC limits the range of Wi-Fi by putting a limit on the strength of the antenna signal. So I believe you could not put up a high power Wi-Fi on a tower."


Ephraim Schwartz at InfoWorld argues that a major reason operators go for WiMAX is that it's licensed and hence can't be acquired by just "anybody".

Wi-Fi on the other hand is using unlicensed frequency bands. Maybe this is a signal to new wireless and data-focused operators to set up high power Wi-Fi towers covering towns or at least campuses (university, corporate or otherwise). Especially as more and more phones have integrated Wi-Fi and voice can be handled via VoIP.

How Wi-Fi behaves in a setting where there might be a thousand connected users per tower I don't know. WiMAX might have improvemements in this area.

As Wi-Fi (or rather sub-standards of IEEE 802.11) is likely to evolve faster than WiMAX, due to the pressure from inhouse corporate and home use, it's also possible performance over Wi-FI will be better over time than for WiMAX.

Behind the push for WiMAX | InfoWorld | Column | 2006-10-24 | By Ephraim Schwartz: "Behind the push for WiMAX"

SearchMobileComputing has a series on mobile application platforms, and this time it's about Java ME.

I agree with the core points highlighted here, and that incompatibilities between phones is the biggest issue with Java ME, but it's not a showstopper, as basic MIDP 1 applications will run unchanged on most phones, so I'm more positive than the people speaking through this article.

Most corporate applications are forms-based, hence basic MIDP 1 is enough for many corporate client/server. By deploying MIDP 1 Canvas it's still possible to provide reasonably looking graphics for e.g. diagrams and such. HTTP is part of MIDP 1, so basic service communication is supported as well. With J2ME Polish it's also possible to make forms look really good without sacrificing compatibility. I noted BlueTrail (see the comments) uses J2ME Polish.

Choosing phones based on preferred platform is often tricky and limiting as that might require swapping all phones a company is currently using. This is clearly the case when going for applications that require e.g. Windows Mobile. Java ME doesn't have that problem. Even worse if different corporate services require different application platforms. Then you are really in trouble.

I've previously posted entries about other forms of incompatibilities, like specification (mis)interpretations and outright bugs, due to the several providers of Java ME implementations.

Again, to be sure a Java ME application will run on a certain phone you therefor need to actually test on that phone. You can't just assume it will run. This is a big issue with Java ME (and also Linux), and much less a problem with Symbian OS and Windows Mobile.

Where it can get outright hairy is when you need to access the camera, PIM database, file system etc of the phone. Then differences between phones really show. Functionality differences should be expected of course, as different phones have different base functionality and features like PIM and file access are relatively new. Best case you can test for this in your application, provided that the application can still work without the lacking functionality.

That Java ME is in its "own world" on the phone can also cause issues with e.g. cross-application functionality, but that also (intentionally) provides protection from applications gone bad.

Quote: mobile Java is already available on many consumer cell phones. It has yet to hit its mark with smartphones and PDAs
That's not true, technically speaking. Java ME is available on pretty much all newer smartphones as well. It might not be used for other than games though, but that's more an issue of perception.

The article under-emphasizes Java ME's complete dominance in mainstream phones, and it seems the industry has just started realizing this is a fact.

Quote: "Given the prevalence of BlackBerrys, I don't think Java is going to dry up and blow away,"
This is a logical somersault: There are 1B mainstream phones with Java ME. That's the driver! Relatively speaking BlackBerry's don't exist, except in the hands of managers. Implying that it's mainly managers that need access to corporate services is of course nonsense. That's repetition of the notion that PCs were mainly for managers etc. They are in the minority, and all employees need access to corporate email, sales people need to get access to customer info, consultants need to get access to time reporting, etc.

Of course, you don't need to download an application to the phone. Many corporate services can be provided via a simple browser interface, but if you need a high comfort level on the interaction, a good look-n-feel, dynamic graphics, phone function interaction etc, Java ME is a darn good choice for mobile enterprise applications.

Mobile platforms: J2ME -- looking for its place in the enterprise

This is an interesting test as it shows speech recognition can be fast and secure enough to be useful for converting speech to text messages.

In this test the speech recognition software was running on a high power PC, so a similar solution running on a phone would need to be squeezed to fit in the available memory and also the CPU power is way lower on a phone. Also, mobile phones are normally used in noisy environments, which also would lower the score for speech recognition.

Nuance provided the speech recognition solution for the test, and my understanding is that it's actually running centrally rather than on the phone, so in a thought "voice-SMS" scenario you would call the service and speak the message that would then be converted into a real SMS. I wonder how one gets a confirmation that it actually understood the message correctly. By speaking the converted message back?

Computer beats fastest text messenger - Yahoo! News

I've dropped my landline subscription and will from now on only use mobile and broadband VoIP. It took a long time to make the decision, but now I wonder why I didn't do it a long time ago. The comfort level has increased substantially.

Some rationale:
* A landline is shared, but a mobile phone is individual and always with me, and free for inbound and outbound calls (unless I'm myself on the phone already, duh). In other words: Landline = stone age. Mobile = contemporary. Also try doing SMS via a landline phone. All newer DECT phones support SMS, but few operators do for landline.
* I have a fiber/100baseT-based broadband connection, so obviously I don't need the landline for that. Not even VDSL beats this, and definitely not from a stability perspective. As I'm responsible for the broadband sub in my area I also have a chance to renegotiate to get even better conditions, price and/or services.
* I very seldomly send and receive faxes, and if I need to I can do it via my mobile or an Internet service. I wish Skype could do faxes via Skype Out. Theoretically it's possible to write a plugin for this (emulating the fax signalling). Maybe something for e.g. USRobotics to do, now when async modems don't generate revenue.
* The cost for the landline was high, and the mobile subscription is needed anyway, so why have both? Sure, calls are more expensive on mobile, but when summing up all other household and business costs, telephony is a very small item. Mileage may vary.
* Calls, SMSs and MMSs within the mobile operator's network are completely free (using 3Bomben Friends), yet of course not to other networks, but the cost is still relatively low. Not as low as landline though, but most calls are actually inbound.

The main issue I've noted is that as all calls go to the same phone it's somewhat harder to separate private and work calls, and I do get work calls also in the late evening. To compensate for that I save all numbers, including from telemarketers, to the phone book. In the latter case so I know when I shouldn't answer. And in the general case ... sometimes for the same reason :).

What I want now is a subscription with flatrate data, as data is actually the biggest item on my phone bill. I guess I'm not that talkative.

It's great that Sun provides such tips, as they started the whole thing (both Java ME (on the positive) and the defragmentation issues (on the negative)).

Goal: aim to reduce the practice of generating a distinct executable version (jar file) of an application for each and every phone

Building different versions based on the same source code yet with conditional compilation is mentioned as one solution, but I don't consider it a such, as you still get many versions of the same applications. The other, more dynamic approaches, are more interesting. Especially when the application is targeted for newer phones that have tons of memory, and so can handle applications that have some functionality that's not actually used for a specific phone.

It's focused on the issues encountered in Canvas-based applications, yet also with general pointers.

Thanks Sun.

Java ME: De-fragmentation Technical Overview and Design Guidelines Index

It's a Java ME / MIDP application also for these platforms.

Opera Mini Web browser for your BlackBerry or Treo @ PhoneMag.com

eWeek has a note on SavaJe supposedly giving up.

SavaJe has been a long-time promoter of Java ME and specifically CDC (see my Java ME Terminology).

As far as I know SavaJe hasn't sold much of its CDC-based platform, yet is supposedly a 140 person strong organization. That's a lot of people when there's no revenue. The note even hints that SavaJe might already be closed down.

This could easily make MIDP 3 more important than CDC for future mobile devices, as MIDP 3 provides an upgrade path for current MIDP 1 and 2 developers, while CDC is something completely different at least in the UI department and hence not compatible with MIDP. Of course the few existing CDC devices support also MIDP.

Here are a few of my other notes on SavaJe, CDC and MIDP:
Mobile phones with Java ME CDC
Is there room for Java ME CDC?
Motorola promotes a better mobile Java
Java ME, still not Write Once, Run Anywhere
SavaJe Jasper S20, an all-Java phone

MEX has an interesting piece that complements my note about Who actually gets the loyalty from consumers, where it's clear third-party now provides the interesting services, not the operators, and that operators need to refocus to enabling services rather than providing said services, that they anyway don't do.

It's a destructive/static loop:
1. Operators want to provide also the information services,
2. but they don't,
3. yet others do,
4. but operators don't want them to,
5. because, ... (start over at 1.)
Operators need to break out of this loop and realize what consumers want, what operators are good at and what they actually can make money on, and in this case it's providing billing and other "backbone" services for information services.

Sure, operators often partner with leading service providers, but that's once the services have become established (see Flickr, ShoZu, MSN etc). In this hectic "Web 2.0" era, time is of the absolute essence, so if anything hinders service providers from establishing themselves quickly they might completely lose out on the opportunity, hence operators can't be relied upon as shopping windows for new services. See Web 2.0 on overdrive for more on this phenomenon.

MEX - the strategy forum for mobile user experience - Why value is slipping away from the operators

Yearly doubling the sales of mobile phones with Windows Mobile sounds like a great feat, until you realize the estimated 12M units for 2006 should be compared to 1B of phones estimated to be sold in total this year.

Also, the definition of smartphones mentioned here confuses things: "which can run software applications such as email, mobile TV and games". All phones released now can do email and games, and not more smartphones than mass market phones can do mobile TV (faked/streamed or via e.g. DVB-H).

Microsoft is fighting with RIM, Palm etc for a very small part of the market, and it seems Microsoft wants to stay there, as it's got almost 10 years now to refocus to the mass market, but seemingly has decided to not go after that market at all. The last mass market phone with Microsoft software was probably the Sony J5. That software platform was for some reason dumped and replaced by the very heavy Windows CE based platform now called Windows Mobile.

As always RIM is mentioned as one of the players Microsoft competes against, but also RIM is in a very tiny segment of the market, and is not likely to ever achieve large volumes (read: 10s of millions of units).

I wish Microsoft all the best, but it's not a stretch to predict Microsoft will not be an important player in the mass market this side of 2010.

When you look at the amount of media attention RIM and Microsoft get you might think they together dominate the market (and at least the business market). That's the effect of successful marketing, and the fact that journalists are in the target group.

Microsoft sees fast growth in Windows phones: report - Yahoo! News

MobHappy hosts this week's Carnival of the Mobilists, pulling together the best in mobile blogging, at the same time celebrating the 50th issue.

Carnival of the Mobilists #50 at MobHappy

I previously wrote a brief note about WICD Mobile 1.0, advanced Web support for mobile devices.

Paddy Byers at Open Gardens explains that WICD-M actually encompasses also XMLHttpRequest, which I initially thought was left out.

Quote: The WICD Mobile variant is essentially just a profile, demanding certain XHTML and CSS capabilities, certain SVG capabilities, and certain DOM and scripting capabilities including XMLHttpRequest.

He highlights SVG (Scalable Vector Graphics) as a way to enhance the user experience and as an alternative to Flash Lite.

He still believes Flash Lite has a lot going for it compared to SVG, due to good tools, and that it's already supported by some mobile browsers and by some phones. Certainly Flash Lite is not that much deployed. I'm sure there are way more applications using advanced graphics made in Java ME than Flash Lite. Just look at the mobile games market.

Whatever the potential of SVG, I think WICD-M could help a lot as a base for future mobile browsers, when we now leave the "crippled at birth" WAP/WML technology once and for all behind us.

WICD mobile: MashLite?

(via CEO's Mobility Weblog)

The PC market is relatively consistent when it comes to methods for enabling of Internet-based information services compared to the mobile market, due to the complete dominance of Microsoft and due to the few providers of browsers (that anyway need to be IE compatible) etc. There are many things you could say about the negative consequences of Microsoft's dominance, but I don't think anyone can deny that consistency is a good thing for service providers.

Not so for mobile phones where there are several ways to mobilize services, both via the phone's native functionality, like SMS, MMS, email, browser etc, and via downloaded applications for Java ME / MIDP, Symbian OS, Windows Mobile or similar.

Pragmatically it's easiest to make services that rely on the phone's native functionality, and if the service needs some level of interaction then the phone's browser is a reasonable choice, as it's possible to achieve good consistency over phones without having to test on all phone models (which is practically impossible). This approach has been taken by e.g. SoonR, and tools like WURFL and similar help smooth out the edges.

Java ME / MIDP is completely dominating in after-market mobile application platforms, but is suffering from the many providers of Java ME implementations with different interpretations of specs and different bugs/features. One would think that Sun would some day realize that they need to become the only provider of Java ME implementations to secure a higher degree of compatibility, but when reading information originating from Sun it seems they still think Java ME is Write Once Run Anywhere, and the decision to make Java ME open source will not increase consistency, despite other values of open source efforts.

Java ME is certainly still a good choice, and often the only choice, if you need to access the phone's functionality, like the camera, audio, location, file system, PIM database etc, and if you need to display interactive information including graphics.

Of course the camera (both for photos and video) can also be accessed via native functionality as photos/videos can be sent via MMS or email to the service, yet there's really no interaction with the service this way. It's more like "send and pray", yet the use case is very easy to explain, and has been deployed by mojungle and others.

ShoZu requires a special application installed on the phone. As I mentioned in
Web 2.0 not for phones? ShoZu would gain from also supporting MMS/email transfer of photos.

Advice to those that want to get into mobile services:
* Select client solution based on whether you want to reach the whole mobile market or you can survive by targeting users of specific devices (and application platforms); revenue doesn't need to be less if you limit the scope, as e.g. corporate users will pay way more for services and applications than the average consumer would; also you can start with a limited scope and broaden later (like ShoZu did)
* For maximum coverage with no interaction with local phone functionality, use an SMS-based approach
* For services that need some level of user interaction and look-n-feel yet reach a very broad audience, use a browser-based approach
* For interaction with the camera, yet with no need for other service interaction, use MMS/email
* For close interaction with the service, including advanced graphics and access to some local functionality, use Java ME / MIDP
* For the most advanced and broad access to local functionality, yet where it's OK that only a smaller part of the market is accessed, use Symbian OS and/or Windows Mobile

Here's another note with a slightly different angle on the subject:
Internet leaders: Mobile is tough nut to crack | InfoWorld | News | 2006-10-20 | By Nancy Gohring, IDG News Service
I question the control of operators mentioned here. At least in Europe you can do pretty much what you want via the mobile network, yet the operator is of course always the middle man charging you for the transfer of SMS, data etc.

The CDMA camp has the recent years claimed cdma2000 (1X, EV-DO etc) is real 3G. It's been very clear at exhibits that Qualcomm and their customers think so, and this note therefor claims cdma2000 leads with 80% of the 3G market.

Korea was very early on deploying cdma2000 (in practical use before USA), and I was taken back when a business contact I visited in Korea claimed he had 150 kbps while I hadn't even tried out GPRS yet, so at the time I had effectively 9.6 kbps on circuit-switched. That was quite different from the statements of European leadership in mobile that the local "propaganda" claimed at the time.

I'm sure the UMTS/WCDMA camp will say that cdma2000 is not really 3G, but the key point is really that cdma2000 is used way more than WCDMA, and secondly it's been used way longer for content and information services.

Does that still mean it's true 3G? I don't know, and I probably don't care either. The value of technology is only in its use and applicability.

Does cdma2000 have a chance in Europe? I don't think so. Too much is invested in WCDMA.

CDMA2000 Providers Claim 3G Lead - Yahoo! News

I've found FierceMarkets' daily email newsletters to be excellent. They don't just quote news, but also provide brief commentary to the most interesting stories. Free of charge and with several different focus areas. Worth a spin.

Publications | FierceMarkets

Strategy Analytics believes so in a new report on the topic.

The brief summary mentions IMS to foster IM and presence, while it's actually today's popular IM services that will drive the need for mobile IM, also for corporate use. Closed-community IMS-based IM will be a very hard sell to anyone, except from infrastructure provider to mobile operator.

Interestingly it mentions a need and availability of low-cost handsets with alphanumeric keypads. That need will surely be more obvious for business use, where also corporate email etc will be handled by the same phone, yet also for consumer/casual IM an alphanumeric keypad would simplify and speed up messaging (not just IM).

It also asks the classical question "will mobile IM cannibalize SMS? Email?". Who cares really? A more important question is: "Will future consumers at all care about the information services operators provide, instead of the much more attractive, contemporary and useful third-party services?"

Time for Real-Time: Gauging the $2.4 Billion Mobile Enterprise Instant Messaging Opportunity

At least this video might give you that idea, or not.

It's in Swedish and with texting in Spanish or something, but I think you get it anyway.

The guy in the middle was voted the world's funniest man in Sweden this year. Not sure what other countries think about that award.

It's slow today...



YouTube - Future of cell phone technology

Rough translation:

Left guy:

"Yes, it's Göte"
Miscellaneous conversation

"It's the new Nokia, GHF C58 ... Mobile chip under the finger nail ... Ghastly expensive"

Middle guy:

"Yes, it's Örjan"
Miscellaneous conversation

"Ericsson VTH 535 ... Speaker chip in the ear ... Microphone in the nose ... Sickeningly expensive"

Right guy:

Farts...
"I got a fax"

Swedish humor at its best :).

1B shipped this year, totalling 3B phones on the market, and 1B of those have Java ME. Nice numbers for anyone in mobile application and service development.

Nokia and Motorola increased sales, yet at other places I've read both decreased revenue.

Then comes Samsung, Sony Ericsson and LG. LG was #4 earlier.

IDC: 1 billion phones should sell this year | InfoWorld | News | 2006-10-19 | By Nancy Gohring, IDG News Service

David Wood at Symbian claims Web 2.0 services can't be directly transferred to mobile phones. That's a fact of course, due to current browser functionality, display real estate etc.

Yet it gets a bit confusing when also Ajax is mentioned. Web 2.0 is about service types (user-generated content etc), not specific technologies. Of course social network services can and will be (en masse) transferred to mobile phones, but they will have to be adapted to the different feature set of phones (both in terms of limitations, as well as additional functionality like a camera and location).

Quote: One way that Web 2.0 companies can similarly adjust their services for mobile devices is by relying less on browser-based applications and more on small software clients that users can download onto their phones. "The browser will fade into the background,"
That's an interesting point, as e.g. a Java ME application doesn't have the interaction limitations a phone browser based service has. Traffic can also be better optimized by transferring data only when relevant. There are issues with this approach though. Despite me being a Java ME afficionado, I'm well aware that it's easier to achieve cross-phone compatibility via WAP 2.0-based services than via Java ME. Also, Symbian OS and Windows Mobile have less cross-phone compatibility issues than Java ME, but are on the other hand much less prevalent in phones.

ShoZu is mentioned as one such service where an application is used to accomodate the user interaction and communication with the service. I'm not 100% for the ShoZu approach though: Most users only want to submit photos, not edit them via the phone, and in such a case simply emailing/MMSing the photo would be sufficient, and much easier than installing a special application. My advice to ShoZu is then to complement its use of an application for more advanced users, with an email gateway for everyday photo blogging. There's no conflict really, just more users.

Andy Tiller at Cognima (that made ShoZu) notes the barrier to installing applications: Downloading a client is a huge barrier, ... But right now, it's the best option
Actually, on most phones it's better/easier to use MMS or email if it's only about posting pictures.

See my Comments on Mobile 2.0 for more on Web 2.0 on phones.

Experts: Web 2.0 doesn't copy directly to phones | InfoWorld | News | 2006-10-18 | By Nancy Gohring, IDG News Service

Still to consider a rumour, but Apple likely wouldn't take these steps just for the fun of it.

I'm surprised no one trademarked iPhone ahead of Apple.

In the trademark filing for iPhone: "handheld and mobile digital electronic devices for the sending and receiving of telephone calls, faxes, electronic mail, and other digital data; MP3 and other digital format audio players."

Apple Files Patents For iPhones - Yahoo! News

Industry Buzzing with Apple iPhone Rumors - Yahoo! News

According to a report from Kaspersky Lab the risk is especially high for Windows CE (as used in MS Smartphone, Phone Edition and Windows Mobile) and Symbian OS (as used in Nokia Series 60 and UIQ), as a software developer has open access to many parts of the system. That there are not many Windows CE viruses today is likely because it's hardly used in phones (relatively speaking). Symbian OS (via Series 60) is much more prevalent, but is supposedly also a more protected application environment.

Like on the PC side I believe viruses for just destroying things will not dominate, but rather such that generate money for the virus creators/distributors. That completely drives the PC virus "industry" today.

That was the case with the Java ME virus Red Browser, that generated expensive SMS messages. Hence not even Java ME is safe from viruses, but to a higher degree than Windows CE and Symbian OS.

Even so, Red Browser (being a typical Trojan) relied on the user's gullibility to download the application in the first place, and then accept Red Browser's sending of messages.

Windows CE, Symbian wide open to attack

This is probably the most technical note I've written so far. You are warned.

This slightly off-topic entry is intended to cover a subject that I think the Java literature gets all wrong, trying to abstract something that doesn't need abstracting. Rather it makes it "magical", while it's pretty down-to-earth when you take a closer look. As programming is a precise art, any sense of magic is completely out of place, even for beginners. Even more so when coding for Java ME, due to the memory and performance issues on phones.

A simple fact: Java uses pointers for pretty much everything. Sure, they are more type-safe and much more managed than in e.g. C and C++, where use of pointers is outright dangerous. Java also has a much cleaner syntax for object use.

In Java-speak it's called references, but frankly it's all about pointers, so I use that term below.

When you write "[class] [name];" you don't declare an object. You are just declaring a pointer to an object. [name] can point to any object of type [class] as well as derived such.

It's first when you do "[name] = new [class];" or "[name] = [object];" that you actually assign [name] the pointer to a specific class instance (= object).

This also means a statement like "somefunction(new [class]);" makes complete sense, as what you do is allocate an object on-the-fly and then pass its pointer to the function.

As the pointer is passed to the function you can also alter the object within the function, and changes are normally retained when exiting the function. Not in the above implicit case though, as when the function returns there's no variable pointing to the object any more and the object will hence be removed.

"null" serves as an indicator that a pointer variable hasn't been assigned any pointer to an object. It can also be used to de-reference objects. Always a good habit when objects are no longer used.

All objects are stored on the heap, and once there are not live pointers to any object they can be removed by the garbage collector.

Another area of possible confusion is arrays of objects. E.g. "[class][] [name];" only allocates a pointer variable to a possible object that's an array of pointers to possible [class] objects. When you then do "[name] = new [class][10];" you don't create an object with 10 instances of [class] but just 10 pointers to objects of type [class]. It's first when you do something like "[name][4] = new [class];" that you actually create an object for the array.

The only exception from the above rules is primitive types (e.g. int). That's also why there are container classes to make them behave as real classes and objects.

Another area that might seem magic is the creation of a class instance. What happens is that space is alloced for the object on the heap, variables are initialized with default values (as per the class definition) and the appropriate constructor is called. If there are several constructors, the one with the right argument types is called (overloading).

Other object tidbits (and suggestions for improvements of the Java language):

In e.g. Visual Basic, PHP etc a string is considered a primitive type, which makes text management very easy. In Java you can concatenate strings with "+", but apart from that a String object is like any other object, which is evidenced by e.g. the less than intuitive method for comparing strings "str1.equals(str2)". It hadn't hurt if e.g. "==" also worked for strings, as it's used so much. Also, String objects are immutable (constant), so appending strings to a larger string is very inefficient and memory-consuming. StringBuffer comes to the rescue here, but to me the string handling in Java seems a bit too manual for Java to be considered a high-level language. The benefit is that you can get the efficiency you want.

Any class can have a "equals" method that should implement an appropriate comparison of similar-typed objects. Considering this, it's kind of strange that the "switch" statement only handles values of integer type. Rather it should use the equals method to allow e.g. strings and even more advanced objects.

This is an interesting new VoIP service that uses SIP for communicating VoIP via Wi-Fi. As such the transfer between Truphone peers is free of charge, while you get lower charges on long-distance calls to "real" phones, as the communication is transferred over the Internet and converted to PSTN locally.

My concern is that there are still not that many Wi-Fi hot spots, and of course the phone needs to have Wi-Fi. Truphone currently supports the Nokia E and N series. The client software is for Series 60 as of now.

Welcome To Truphone - Mobile VoIP has arrived

Last year almost all remaining European ODMs disappeared. Now it's time for the European phone brands to realize it's time to go. Nokia and Sony Ericsson are not affected (both have seen good growth this year), but on the other hand most of their phones are made in the less expensive Asia anyway. For Siemens mobile phone division (that was acquired by BenQ last year) and Philips' ditto it hasn't gone as well. Of course both Siemens and Philips are very strong in other markets than mobile phones, so from their point-of-view this is for the best.

Philips pulls a Siemens, sells phone biz to Asian firm - Engadget Mobile

Mobile Opportunity hosts this week's CotM. No carnival theme this time, just good reading.

Mobile Opportunity: Carnival of the Mobilists #49

The jury is still out on whether service providers will bother about .mobi and the guidelines stipulated by the associated company dotmobi, but that doesn't stop domain registrars and tool providers from seeing an opportunity to generate more revenue this way.

dotmobi has published guidelines called Switch On! that very briefly describes the guidelines for .mobi-compliant site.

dotMobi open standards for mobile content adopted by top hosting and developer companies - News Wire - Mobile Europe

As the hype factor has now vanished from 3G, consumers look more at what it actually costs to use, and according to this note the interest has dropped since last year due to the continued high costs.