Sony Ericsson has announced Yet Another Mobile Platform:

Project Capuchin will provide developers with an intuitive tool to create applications with a cleaner user interface (UI) without sacrificing the strong, feature rich and widely deployed Java ME
infrastructure, including secure, well-developed content distribution. Project Capuchin’s bridging software will empower two distinct developer communities to leverage their respective expertises to create the next generation of highly engaging and immersive mobile content.

The GUI layer in MIDP2 is one of its biggest drawbacks, limited and full of bugs and incompatibilities. Merging Flash Lite as a GUI layer with the Java APIs as the underlying engine could be a brilliant move towards a new platform.

The question of course is: do we need yet another mobile platform?

Anyway, nice to see Sony Ericsson be at the forefront of the upcoming merger between the web and telecom from a technology perspective.

I guess the upcoming JavaOne sessions will be more telling than the press release. Go see Viktor Mårtensson talk about CHAPI (JSR211) if you’re there! Should be interesting.

Updated: Adobe is announcing a move towards a more open Flash, even removing the licensing fee for mobile devices.

This means that the Sony Ericsson feature phones will have a powerful and open development platform based on JavaME and Flash. Now, if they only could bring in the web browser and XHTML/Javascript in the mix…

Actually, they’re not. They keep a lot of innovation from happening. Will Google’s move to launch the alliance make the platform battle finish sooner than it otherwise would? Probably not. From that perspective, the Open Handset Alliance was a disappointment.

If you expected a cool device from Google, you should also be disappointed. What we got was yet another mobile linux initiative. Problem is, there are many mobile linux alliances out there, we don’t need another one.

It makes sense for a company like Google not to start building devices. As I wrote before, Nokia pumps out 11 phones per second and distributes them worldwide. They’ve probably sold a couple of hundred phones before you’ve finished reading this blog post. Feel like competing with that?

If it takes you one hour to decide, they’ve sold 40’000 devices.

Still feel like competing with that?

I didn’t think so.

No, Google is a software company and should remain a software company. Any software company that survives for some time becomes a platform company. Any platform company that survives for some time becomes an OS company. Maybe that’s where Google is today.

If anything good comes out of this, that would be a set of java applications or maybe even a java framework on top of MIDP. Can we hope for that? On November the 12th maybe we’ll know. Until then, disappointment is the word I’ll use to describe what Google launched.

Sounds like news to me. At least I’ve missed it.

I wonder what the deal is between Yahoo! and the manufacturers. A not too wild guess would be revenue sharing of ads shown in the application. Consequence: your next phone might be subsidized not only by the operator but also by Yahoo!. In other words, mobile advertising is coming and it will take the shape of value added applications in your phone.

Another example of this: Yahoo! is also the search engine that pays for you to use Opera Mini (you didn’t think it was really free, did you?). Together with Yahoo! Go Mobile in 200 devices, Yahoo! can take a significant chunk out of that mobile revenue that Google wants with their as-of-yet-non-existing GPhone.

Microsoft has been trying to break in to the telecom market for years with Windows Mobile but they will most likely join the advertising war. They’re just a little late for the game.

Microsoft, Google Phone, Yahoo! Go Mobile… looks like the future of the web really is in mobile.

Pictures from the event here.

Updated: Lotta Holmstr??m at Citizen Watch also wrote about the Yahoo! presentation:

Yahoo! Go is a small java applet which lets the user access email, flickr, search and more. It will be in 200 cellphone models at the end of the year.

What does “in” mean? Preinstalled or available for?

One of them is flash memory replacing the hard drive. Another one is new form factor motherboards, small enough to put in a box of matches. A third one is WiMAX, promising broadband speed (well, sort of) wireless internet access (problem for WiMAX is, it’s been promising it for so long that good old 3G UMTS is catching up). Yet another is the Linux OS.

What do you get when putting all of them together?

Something like this.

The device is called MUnit and runs a 1GHz VIA CPU which is x86 compatible. You should be able to run Windows Vista on this one, if you want to. What the web tablet market needs is not Vista, though, but a Linux version that’s stable and has a developer ecosystem fertile enough to produce killer applications. The problem (some would call it a strength) with Linux is its many flavors.

Many flavors is great when eating ice cream but not so great when you want to build a commercial software application that must run on as many devices as possible for as little cost as possible. Linux is not there yet and maybe it never will get there. Why? Because Linux is a fertile environment for operating systems – not for applications.

That’s why I think Linux and Java is such a good combination – and I’m not the only one.

(source)

Another way of putting the stats:

• Opera Mini: massive growth.
• Opera: no growth.

In a couple of months time, Opera Mini might very well pass its bigger brother and place itself amongst the top 4 browsers on the net (including desktop browsers).

Amazing! Really.

Compare this to other mobile (and pre-installed) browsers such as Access Netfront with a 0.01% market share according to the same statistics and it becomes even more amazing.

Now, “0.39%” of web traffic might not sound too impressive to you, but if you ever have tried building an application for mass market mobile phones you know that it’s a huge effort even getting it to work on the devices out there.

Then you have to convince people to download and install it. If you’re lucky, the settings for internet access are correct in their phones. If you’re even more lucky, they manage to find the application in the menu system after it’s been downloaded and installed.

And, remember, the users already have a browser pre-installed on their phones. It’s a native application and supposed to be faster and better integrated to the device. Your browser can’t be equally good to the native browser, it has to be much better otherwise people don’t bother downloading it.

Then there’s the whole mobile browsing business, which hasn’t really taken off the way once predicted. First and foremost you have to convince people of the very idea that a web browser in your phone actually can be usable.

Well, Opera pulled it off.

Looking at the statistics, it seems like Opera Mini right now is the engine pulling the entire mobile web up to the domains of the desktop world.

Look at the September stats:

Microsoft Internet Explorer 77.86%
Firefox 14.88%
Safari 5.07%
Opera 0.87%
Netscape 0.72%
Opera Mini 0.39%

There’s not a mobile browser in sight, except for Mini (disclaimer: I don’t know the percentage of iPhone users in the Safari stats). The competitors are down at 0.01%-0.02%. If Minis growth continues, it will hit 1% within 6-9 months and become the number 4 browser.

There’s only one conclusion to draw: right now the mobile web is Opera Mini and Opera Mini is the mobile web.

Sorry, Russell, but that’s a fact.

Opera Mini is becoming so big it soon makes sense to develop web sites and applications targeted towards Mini as a platform. In fact, if you plan to take your web site mobile (and of course you plan to do that, you’d be crazy not to), Opera Mini is the platform to use to do that. As a bonus, it will work on the iPhone too.

Now, if only Opera could follow Apple and open up the browser for access to the phone hardware and underlying OS (via the J2ME APIs), that would make one heck of a platform for mobile development. (I’ve actually suggested this to Opera a couple of times, but the message doesn’t seem to have sunk in. Oh well, the last time was pre-iPhone so maybe they’re starting to get it now. JSR290 is the Java standardization attempt of basically the same thing. )

Interesting times to work with the mobile web. Maybe 2008 will be the year it really takes off. If it does, it will be much thanks to Opera Mini. Who said mobile java is just good for games?

I’ve written about the internet as the mobile OS a couple of times. What I mean by that is that most mobile applications we will be using will be to at least 80% based on a strong web service as the backbone. What we see in the phone will be a simple and thin GUI-layer. Widsets, Opera Mini and Mobile Gmail are all examples of this.

Java, Flash Lite or even XHTML will be perfect for most of the things we will want to do with our phones. As soon as the problems with fragmentation has a solution that is good enough, things will explode in the mobile service space. Unfortunately, that might take a few more years.

I’ve also written about what Bill Gates summarized as:

“The phone is going to be the PC, and the PC is going to be the phone.”

PCs are getting smaller and phones are getting more capable. Is the iPhone an up-scaled version of the iPod or a down-scaled version of a MacBook? I’d like to say the latter. Soon enough you will have full scaled PCs in your pocket running Windows, OS/X or Linux.

If we take those two observations and try to draw a conclusion from them, that would be that we’re moving towards a point where all mobile devices more or less share the same software stack. We’re not there yet, but it’s pretty clear that’s where we going. The platform war is already here.

Another conclusion it’s tempting to draw is that since all our devices share the same software stack, we will only need one device. Let’s put everything in the mobile phone (or whatever you want to call it) and you won’t need an MP3 player, camera, small laptop, car keys or credit card.

This conclusion, however, is wrong.

Why? Well, now we’re (finally) getting to the title and the point of this post, which is: the interface is the device.

The interface is not just what’s shown on screen. It’s the entire design and behavior of the product. Where the buttons are placed. Its size. How you carry it. How it turns on (and off). A good camera is not a good MP3 player. A good stopwatch is not a good mobile phone. A good word processor is not a good SMS-tool. And so on.

I once tried to use my Sony Ericsson K800 to take the time and listen to music while running. It was useless. The UI paradigms completely clashed. The poor little K800 wanted so much to be a good phone and a good MP3 player and a good stopwatch that it simply failed to be all at the same time. Now try to add GPS, TV, video recording and one or two games to the mix. What you’ll get is a useless feature soup. Because the interface is the device.

On the other hand, I might very well imagine that my GPS, my camera and my mobile phone share the same software platform and even underlying hardware. Being able to install java midlets on my digital SLR would be great! I might even consider using it as a mobile phone in emergencies. But it will never replace my mobile phone device. Because the interface is the device.

Am I opposed of all the new features in mobile phones? No, the mobile phone may be quite good as a jack of all trades. But it will be a master of none and you will soon get tired and irritated in using it for something it wasn’t specifically designed for. Because the interface is the device.

Can more than one feature never mix in the same device? Of course they can. For example, a mobile phone can be quite a good music player if you have a good headset (because the interface…) and it was designed to be used that way. A laptop or tabletop PC can do many things. Even they are limited, though, as everyone who’s tried to use their PC to watch TV will tell you. It’s too noisy, takes time to boot (you don’t boot a television, you just don’t), isn’t very good to use with a remote control, bluescreens (crashes) every now and then and consumes too much power. The PC simply wasn’t designed to be used for watching TV and so it fails.

Because the interface is the device.

Another conclusion from all this: the mobile 2.0 revolution is not just about mobile phones. We need to get every device manufacturer out there to open up their products and make them internet connected. The phones are not enough. Maybe I’m well in to Mobile 3.0 when I say this, but every device out there need to share the same platform and become connected. From DSLRs to washing machines.

Because, and I think I told you this before: the interface is the device.

(Hm. I think I will have to return to this subject. Too long post and I don’t think I made my point clear enough. Oh well, that’s why you have a blog. )

1. Google Phone is based on a mobile variant of Linux, and is able to run Java virtual machines.
2. All applications that are supposed to run on the Google Phone are java apps. The OS has ability to run multimedia files, including video clips.

It looks a lot like the kind of phone I would have built if I had the chance to do it from scratch. Linux at the core and Java as the application platform makes a lot of sense for building an open, developer friendly phone with an already existing ecosystem of applications and developers.

I’m sure Google has the engineers to come up with a great mobile phone. The greatest challenge for them is the logistics involved in selling consumer electronics. As mentioned before, that’s a completely different business from selling adwords. Nokia is pumping out 11 phones per second worldwide. You can have all the Ph.D:s in the world in your staff and a supercomputer on top, that’s still a huge challenge.

Nevertheless it’s a telling sign that the two most talked about phones at the moment comes from the internet and computer industry.

Updated: New York Times on how the worlds of software and telecom meet:

Nokia used to be just a cellphone maker. Google used to be just an Internet company.
Now Nokia wants to be an Internet company and Google, according to rampant speculation among bloggers and technology analysts, may be about to enter the mobile phone fray.

“Devices alone are not enough anymore,” Olli-Pekka Kallasvuo, chief executive of Nokia, said last week in London as the company announced plans for a digital music store, a game service, social networking links and other mobile Internet initiatives, grouped under a new brand, Ovi. “People want more; they want the complete experience.”

My first impression was: oh, no, they made the GUI too slow and cumbersome. The clean simplicity of the first versions seemed to be gone. But as I’ve started to use it I got more comfortable with it and I’m beginning to really like it.

The biggest change is that pages are now showed full size (sort of). You navigate around the screen by moving a rectangle over the page. By clicking the square you zoom in to the page and can read it on a tiny mobile phone screen. On an M600 with a pointing device (and fairly large screen) this works really well. Haven’t tried it on a phone without touch screen yet.

Page rendering is also much better than previous versions. Bloglines (the RSS-reader) and Google Mail both worked and Google Calendar seems to have a mobile friendly style sheet which formated the calendar to fit perfectly in to my device.

Bloglines is noteworthy because it uses frames. It has two frames, with the RSS feeds on the left and the posted items on the right. Opera Mini renders the left frame by itself and when I click a link it changes to the right frame. Smooth!

Seems to be a bug in the bookmark manager (the data for the page I was on when clicking “add bookmark” is not filled in to the bookmark manager) but other than that it seems stable.

The Sony Ericsson M600 together with Opera Mini, Google Maps Mobile and Gmail App is a very capable little web tablet. The same applications run on most modern mobile phones – and they’re free. A mobile web tablet experience is only a few joystick clicks away. Download them all! I promise, you won’t look at your phone the same way again.

Updated: major issue: cookies doesn’t work. Oh well, it’s beta…

The UI library in MIDP2 is basically designed for the lowest common denominator of devices that can run MIDP2. They are very simple (with the possible exception of the powerful CustomItem) and there’s not that many GUI components available for a developer. Add to that the fact that different manufacturers have implemented the library in different ways making it almost impossible to create a MIDP2 application based on the UI libraries in MIDP with a consistent behavior between the devices on the market.

No wonder all of the most successful java applications are written using an empty canvas with home made GUI components on top. That’s also the reason we see great new SDKs and UI libraries such as TWUIK to save developers the effort of writing their own UI classes.

There are a few initiatives on the way to remedy the situation. The successor to MIDP2 is MIDP3 (duh), there’s the Advanced Graphics and UI Optional Package (AGUI, though any API with the word “Advanced” in them should be avoided, the same goes for “Optional”) and there’s the open sourced, Eclipse-based eSWT, one of the components of ERCP.

Now, one should be aware of the fact that the MIDP3 API and the AGUI API are both defined by the Java Community Process (JCP). eSWT, on the other hand, is not. It is an open sourced community effort including IBM and Nokia. Sony Ericsson has always been committed to the JCP, so the silent announcement on the eSWT mailing list is noteworthy for more reasons than one. Gorkem Ercan writes:

Yesterday, we have received the first welcome mail to dsdp-ercp-dev list from SonyEricsson. They are now working to get eSWT to their mobile phones. Moreover, they have started to participate in the project as well.

The path mobile Java takes will have huge consequences for how the mobile web will look in a few years. Remember, this is an application platform that’s already in over a billion mobile devices. Unfortunately it’s a far from optimal ecosystem, making it difficult to distribute applications. There’s just so much creativity waiting to be unleashed the moment this ecosystem becomes more fertile. We will see an explosion of innovation that will mirror how the PC industry boomed during the 80s and the internet during the 90s.

The big difference is that, even though the PC became popular in the 80s, the mobile phone count is approaching 3 billion. This number will make the PC boom of the 80s look like a firecracker next to an atomic bomb. (There were ~48 million IMB/PCs sold during the 80s – the entire decade. That many phones are sold in about 2.5 weeks world wide, assuming a number of ~1 billion/year.)

So, watch history unfold and keep your eyes on where mobile Java is going. MIDP2/CLDC is sort of like MS-DOS was for the IBM/PC. We’re still waiting for Windows to come along. When that happens, the telecom market will be changed forever. The announcement by Sony Ericsson is a small but significant piece of the puzzle in what will become the mobile web framework of the future.

The Opera Mini team has been hard at work for quite some time now on the next major version of Mini. We’re proud and excited that people who have had a chance to test early versions call this new generation another breakthrough in mobile web browsing.

Opera Mini, with an amazing 15 million downloads (!), is in my opinion (and, apparently, 15 million other peoples) the best way to surf the net outside a PC.
In fact, after using the Nokia 770 web tablet for a while, I returned to my Sony Ericsson K800 running Opera Mini (and Google Mail and Google Maps Mobile in parallel). It’s just a perfect combination of portability, speed and ease of use.
So, what’s in store for the next version? To be honest, the upgrade to 2.0 wasn’t that exciting. It’s not an easy task: how do they improve what in essence is a good enough application? I’m really looking forward to find out with the 3.0 version.