Wednesday, May 10, 2006

WAP client or mobile widget?

The talk before mine at MobileMonday Global Summit was given by Christian Lindholm, a VP in Mobile at Yahoo. I met Christian a year ago in Helsinki for the first time, when he was still working at Nokia on Lifeblog (a very cool product, using SyncML). He then moved to Yahoo and is coming to live in the Bay Area, the coolest place on earth when the weather is good.
His talk was full of Yahoo!, but still interesting. In particular, he outlined Yahoo plans to bring to mobile phones the Soccer World Cup (the most watched sport event on the planet). They are going to have a WAP site and a J2ME app. The Java app will include live stream of results, but looked less rich in content. It is targeted to the hard-core fan.
As you might know, at Funambol we built a mobile World Cup application in 2002. Built for Palm OS, it allowed people to get updates on scores and standings, while HotSyncing. It was a phenomenal success and our first large deployment of Sync4j. It was a fat client, obviously. Today, I would call it a mobile widget.
I find the decision of Yahoo to build a WAP site quite interesting. They have a fat client (collection of widgets) for the Yahoo experience, but they still decided to go with WAP in this case. We all know browsing on a mobile phone sucks. However, when you have a ton of content ready, going with WAP makes your life much easier. You can transcode or use mobile CSS. I suspect this was their main reason (the other one was serving the non-J2ME market, which is still a good chunk of the market, but they could have chosen to build separate clients for Brew, Windows Mobile and PalmOS).
They built a mobile widget with J2ME. I would bet the Funambol server I am delivering to Alfresco that the user experience on that client will be much better. Real time updates being one reason. Local storage of results, instead of download at every click, being the other reason. Instant access to content, without the need of downloading it from the web. However, they might not have spent the necessary amount of time to build a proper app, since it is not an easy task.
The tradeoff between a WAP client or a mobile widget is clear. One is easy to develop, only requires to adapt existing web content easily but delivers a horrible user experience. The other one is harder to develop, requires a way to import pieces of existing web content (easy to do, if the web content is well designed...) but can deliver a great user experience.
My feeling is that company will try to go for the easy route. We will have tons of seldom-used mobile web content. At the end, the companies that will make it in mobile will be the ones that will take the time to build widgets and deliver a great experience (not WAP crap). Google and Yahoo started big in this category. All the others will follow. Stay tuned, it is going to happen fast.
Posted by Fabrizio at 16:59  

3 Comments:

Blogger Fabrizio said...  

I absolutely agree with you concerning the advantages of mobile widgets against wap and web access to deliver contents on devices. The most important one is probably the capability to browse offline as far as the wireless data fares are still quite expensive.

However you forget to mention Symbian! Symbian here in Europe has an important role and the most part for mobile applications are built on S60 (Framework based on Symbian OS).
Theoretically J2ME has the advantage that the application should be portable to several devices (“build once run everywhere”), actually each device comes with its JKM’s flavor. As a result the portability is hard to be achieved and development costs higher.

Comment Posted at 03:34

Blogger Fabrizio said...  

Ciao Fabrizio (congratulations to your parents for picking a cool name :-)

I would not forget Symbian or S60 (I was in Tampere talking to the S60 guys a couple of weeks ago ;-) but Symbian is one of the OS. Java has the potential to be the language of choice for mobile development (had the potential?).

You are right that "build once run everywhere" is not real. You need thousands of people to port your app to all devices on the planet. It is called a community. We have it. It is huge. We can do it. We can make portable JavaME a reality. It is the power of open source (and it is happening whether Sun puts Java in open source or not).

Comment Posted at 08:36

Anonymous Franck Poisson said...  

Hi Fabrizio,

I googled the "mobile widget" search query and found your post dated 2006.

At Webwag we launched mobile widgets last week (june07), worth a try on www.webwag.com/mobile.
The mobile app can be synchronised with the webwag.com start page widgets. Or you can add mobile widgets on the air. Worth a try ;-)

Comment Posted at 12:13

Post a Comment

Links to this post:

Create a Link

Back to My Blog