MicroWidget is on the way

Mobile Widget I’ve just decided to start to work on something called “MicroWidget”. MicroWidget will be the equivalent for mobile of Widget, Gadget, or other inspiration source on the mobile….
Basically, it’s an XML based rendering engine with a tiny scriping language. I’ve used FScriptME as a basis for the scripting language. FScriptME is very limited, but it’s a good start. I’ve added some pseudo object oriented faciclities, like myText.data=”some text”, and some usefull xml parsing function.
As an example, here is the widget code for a Clock like widget:

<img src="/clock2.png" />

<script>< ![CDATA[  func onTimerFired() 	timeArr=explode(time()," ") 	hour=explode(timeArr[3],":") 	day.data=timeArr[2] 	secCl.angle =-toint(hour[2])*360/60+180 	minCl.angle =-toint(hour[1])*360/60+180 	hourCl.angle=-toint(hour[0])*360/12+180 	repaint() endfunc  	]]> 	</script>

The code is pretty short and can be easily understood. The only difficulty is the function “explode”, who create an array from a string.

Of course, there are plenty of other usefull function to retrive some xml data, to parse them, etc….

First result are really encouraging, but generate some interesting issue about interface design…

Technorati tags: JavaMe J2ME Ajax Widget Micro Widget

Predicition for 2006!

I will go also through my little prediction for 2006. Most of them are based on my previous post, and yes, it’s Christmas break:

  • Google will buy a Opera(if not already done): the “OS” of Google is the browser, so they need to provide one on platform who does not have one, or a good one. That’s why I predict that they will buy a browser company typically Opera. The interest of Opera for Google is that they also have a good mobile offering.
  • But Google will move from the “Good side” to the “Evil Side”. Too powerfull, too big, too smart….
  • The “big three”, Yahoo, Google and Microsoft will enters more strongly into mobiles. All will provides much better access to their service to everybody mobile.
  • Location based service will increase, especially in North America, thanks to GPS enabled mobiles. In Europe, such mobile will start to appear

google Zeitgeist

Google just published his 2005 “Zeitgeist “, and it’s really worth reading: you wil lhave the top 10 search of the year, in several categories.
But what take a look at some detailled info on popular search in various categories, like WMD, Marhc of the penguins, or Jean Paul II , etc…
The page contain some very detailled graph about the evolution of some request around the year, as well as comparains between two terms. For instance:
WMD show requests about Weapons of Mass Desctruction around all year….
Theses pages show how powerfull is it to know what are the verious request made on Google: you could anticipate trends, get visibility on some mass phenomena, and probably much much more…
Technorati tags:

Google is going to Widget, and compete with Konfabulator


Google just introduced the ability to add Widget in your starting page. The first widget are not very numerous (see the
directory ) but I guess it will increase rapidly. The “date and time” widget seems very similar, in terms of look and feel (and functionalities) to a Konfabulator widget.

May be somebody in Google read my previous posts about writing Konfabulator widgets and write something in a couple of days! ;-)

More seriously, this’s a quite good and smart approach, as it uses the browser model. So all rendering is done using HTML, JavaScript, etc… and widgets are stored on the server side. So the only missing thing, to compete with Konfabulator, is the ability to put theses Widget on your desktop…. Should not be a too difficult task, just extend Google desktop to provide the ability to open some “web pages” that contains each one a Widget, and you’re done (and ultimately, Google Desktop could finally be useful and sexy!).

Also, I would like to mention this interesting view from C.Enrique Ortiz: Google – The Best Thing That Has Happened to Yahoo! I totally share this view, and it’s a very positive proof of the utility of competition (and of course, this comment can be extended to Microsoft, hopefully).

Technorati tags: Google Yahoo Widget Konfabulator

Communicating with the external world….


Interesting announce today:
TIM unveils Z-SIM SIM card with radio technology“.

So what does this mean? Another way (one more) for a mobile to communicate with the external world. Not only through GRPS or UMTS as today, but also to interact locally with your physical environment.
The examples given in this announce are really interesting, like interaction with home appliance, purchasing TV content, parking payment, etc…

This opens some great opportunities for mobile developers, and also for operators! ;-) Phone (or more accurately Sim card) could be more and more a way to identify yourself beyond just the area of phone calls, and to do transactions, mainly micro-transactions.
Of course, operator will be the big winner in this scenario as it will manage the transactions, and I would be interesting to see what will be the bank reactions in this scenario. But that’s so annoying to get your credit card out if you want to by a 1€ songs….

For those who are interested, take a look at JSR257
, contactless communication API. It cover also things like Visual Smart Tags, a very interesting way to get information from you’re environment by taking a picture of a specific tags, or NFC, which is the extension of RFID in the consumer world…

Technorati tags:nfc j2me javame

My first Konfabulator Widget

I’ve just make my first Konfabulator Widget…

The good things, is that took me just ten minutes to understand how to do it and to make it run (I just wanted to display the number of registered users to J2memap as a widget).
It also took these 10 minutes to realize where were the limitations of the framework…

  • No HTML or XML parsing. For something that is frequently used to display in a specific way Webpage, XML data or RSS feed result, and that rely so heavily on XML internally, it’s incredible to see that there is no built-in parsing engine. You have to do you’re own.
  • Basic rendering engine: all graphical elements need to be pre-rendered (with the exception of text). No layout feature, just good old bitmap, the only fancy stuff is transparency.

So even if you start to build something very rapidly, it requires a huge maintenance effort each time you want to do a modification. It would have been great to have a better Web awareness, by including an XML parser, and a kind of HTML/CSS rendering engine. This would have made things much simpler….


Tags:

Electronic Arts to buy Jamdat in $680 mln deal

Important news for our industry, mobile gaming: EA, the biggest game publisher, just acquired Jamdat, the biggest (for now! ;-) ) player in this space….

Jamdat Mobile (JMDT:Nasdaq – commentary – research – Cramer’s Take) surged in after-hours trading Thursday following news that Electronic Arts (ERTS:Nasdaq – commentary – research – Cramer’s Take) agreed to buy the cellphone-game publisher at a 19% premium to its close. Electronic Arts will pay $27 a share for Jamdat and assume outstanding stock options, for a total deal value of $680 million. The deal is expected to close during Electronic Arts’ fourth quarter. Shares of Jamdat were up $4.23 to $27 in after-hours trading, while shares of Electronic Arts fell $1.51, or 2.7%, to $54.24.

The deal value Jamdat at 680$ million…At this price, a very good deal for Jamdat…. A few comments:

“EA overpaid by $200 (million),” Piper Jaffray analyst Anthony Gikas said in a client note, adding the amount was not material

“We’re not sure why the company would spend so much to get into the mobile market when it appears that they could have leveraged their content on their own for less money,” Wallace said in a note. He added that EA will probably spend more than $750 million this year on research and development.

“That’s a lot of console games that could have been developed (or bought),” he said.

Still, analysts liked the deal, saying that Jamdat and EA are each better off together.

EA “has had some fits and starts trying to build its wireless division” and adjusting to a “buy” strategy resolves investor qualms about the company’s relative lack of presence in that market, said Paul-Jon McNealy, video game analyst at American Technology Research.

J2MEMap Update

Just to announce a new release of J2MEMap, with some great new features:

  • Support of GPS through bluetooth
  • Nice rescaling feature (similar to the GoogleLocal one)
  • WayPoint feature (nice to be use with GPS, in order to know how far you are from your destination, and what is the direction). Not yet really “driving GPS”, but quite cool.
  • Support of touch screen. On the P900, it’s quite fun to play with it…
  • Improved display of routes….
  • Support of GPX for the download of waypoint, track, etc…

Just like this one:


Technocrati Tags:

Next Buzz: Mobile AJAX

2005 was probably the most exciting year since 2000 for the Web, thanks to the thousands of new projects, most of them based on the usage of AJAX to create a new and compelling user experience. One interesting question for me, involved in mobile, is what will be the impact of this revolution for mobiles?

Well, seems that the Ajax buzzword will be in all mouth in the coming months, and Opera is already surfing on this.

Thirst thing, is do we need Mobile Ajax?

I think that with the generalisation of data network (GPRS,EDGE or 3G), the services will be more and more interactive. But the current technologies are far from providing a compelling user experience on mobile. Wap (or even HTML) applications look like ’99 Web app (for the best ones). So I believe that looking at what’s happening on the web, many people will provide these improvements to mobile, and this will be an improvement.

And one side note: Google local for mobile IS NOT a Mobile Ajax application (just like J2MEMAp). It’s just a monolithic application created for mobile, which can not be updated, or changed OTA (unless you reload it). I guess that I should define what I think is mobile Aj

So what are the various technologies available, or do we need new ones?

A first answer is to say that we already have one technology that could fit to Ajax requirement: be able to execute locally a small program downloaded remotely to add interactivity in “web pages”. J2ME (oh, sorry Mr Sun, JavaME) seems to be perfect for this. It’s a small bytecode, cross platform, web oriented, etc… However, J2me (in fact CLDC) failed in one important topic: the inability to download OTA some new byte code. So an application, once downloaded, can not be easily changed, modified, etc…. And even worst, you can not jump from one app to the other: from a calendar to the address book for instance. So if you want to do something great, you need to create a -relatively- big monolithic J2ME application. And I do not speak of security issue (see below).
One the other hand, there are plenty of advantage to J2me, and this could be a good candidate for being the scripting language for mobile Ajax application, or the framework basis for this, but it needs to evolve a lot if he wants to achieve this level.

One option (the one used by Opera) is to write a browser in J2ME who contains some Ajax component. Honestly, I did not take a deep look at the opera mobile (the J2ME version) right now, but I doubt that the have a scripting engine in this version…And what will we have at the end: an interpreted VM being able to run interpreted scripting language… And I am sure the several other initiatives are one the way.

The alternative is not numerous. Flash lite could be a candidate, but the first issue is the deployment: very few handsets yet have it right now. But technically speaking (and here, I am not a FlashLite expert) it seems that many of the J2ME limitation does not exist anymore, especially the one concerning the limitation of an application to one single binary. And probably the first truly mobile Ajax experiment will come from here.

But whatever is the option, I except a lot of buzz on this, conference, announces, etc for the second half next year! So let’s be ready for the Ajax wave on the mobile,….

Technocrati Tags:

Konfabulator settings


Just a small suggestion for Konfabulator (but it might be already in the loop for the future): is to store all your settings on the server side, so then, just log into another computer (with konfabulator already installed, and probably you’re yahooId) and you get magically all you’re settings, as well as the widget, downloaded and installed… Sounds cool?

End of the browser?

During a recent discussion with one of our partner, this one predicted “the end of the browser”, before two years, as opposed to the emergence of specialized applications like messenger, or Google earth. His point was that the crucial part was integration between applications, something that was in his view impossible to do using a Web Browser.

In a time of a complete renewal of the Web, I was a little bit puzzled by this remark. That’s true that the obvious evolution of messenger (or any other chat client) is a better and better integration with the rest of you’re environment…
So should we oppose loosely integrated “web application” to highly integrated local apps…
Not sure that there is an opposition here:

  • If it was possible to do Google Earth using a standard browser, I am sure that Google would have done it. It’s just a question of availability of the techno (in fact, I can predict that in two years, Google Earth will be available as a Web application).
  • The Web is probably the best platform for integration. In fact, it much easiest to share application between web app than to share locally, most of the time. And a good example is Msn messenger. The sharing (of information) is mostly done on the server side, not on the client. And it seems that the evolution of messenger and hotmail, is to a be a full web/Ajax based application.

So the Web is probably the best platform to do component integration, and the best integration tool is the Browser.
Of course, there is some need for some vertical and specific application, but most of the people do not really need it. The few specific needs can be solved by a “Widget” approach, similar to Konfabulator or Safari: still browser (ok, it’s not 100% the case), but this browser is skinned and just displays a single web page, but in a very good way…

Again, as stated in one of my previous post, “The end of the OS“, I think that in the future, the operating system might be a detail, as long as you have a very good browser…So I think that browsers will be here for a long time, see you in two years….