Widgets, gadgets and microbrowsers
I’ve been running Windows Vista beta 2 for a couple of days, and one thing I like about it is the sidebar with its user-definable gadets. The sidebar gadgets are in all simplicity applications built out of HTML, JavaScript and CSS (very ajaxy by definition). Individual gadgets can be dragged out of the sidebar and positioned anywhere on the desktop and Vista beta 2 ships with a variety of gadgets including a calculator, a clock, a RSS reader and a CPU/memory meter. Additional gadgets can be easily downloaded and installed (e.g. from here). The sidebar and its gadgets are pretty much an answer to widgets, as introduced by Apple in Mac OS X Tiger (Konfabulator, bought by Yahoo, was the pioneer of desktop widgets).
When Microsoft ships Vista, they will create a real ecosystem for these cute little applications. I don’t doubt for a moment that any web based service wouldn’t want to have a presence on the desktop of the user, and since creating widgets is easy, they’ll become widespread. I can see all the sidebar chats, games etc. being developed to extend a current browser based service to the desktop. This is a hot topic, as the big three (Microsoft, Google and Yahoo) are all competing within this space, both on the desktop and in their portals and services (even the latest Opera release sports support for widgets).
Now, what does have to do with mobile? A lot, I’d like to think. The way I see it, most current mobile applications resemble these widgets a lot (they have a small footprint, are designed to do some very specific task and have considerable constraints on screen real-estate). What I would like to see is some sort of standardization on how the widgets are made (packaging, limitations, etc.) and classified. Then some clever company* could develop technology to transform these desktop widgets into installable mobile applications (i.e. creating a microbrowser environment with the application/widget specific functionality embedded)**. Wouldn’t that solve a lot of problems involved in current mobile application development?
And what would the mobile widgets be called? midgets, of course.
* Opera Platform, their AJAX framework for mobiles, comes immedeately to mind. It is, however, limited to Symbian and Windows Mobile, making it incompatible with most of the handsets out there. Java (MIDP) or Flash Light would seem like the ideal platform to build something like this on.
** A one to one mapping wouldn’t obiously work due to very different UI concepts (i.e. mouse vs. one-hand use), so some mobile specific optimizations would always be needed. My point here is that mobile applications shouldn’t be developed by C++ (engineers), but rather by HTML (web designers).
Technorati Tags: widgets, gadgets, mobile, vista, microbrowser, opera platform
June 20th, 2006 at 5:26 pm
Yeah, I agree
Next generation mobile apps should be something which doesn’t really need very high coding skills, but they could be done by anyone. Flash Lite 2.0 or FlashCast is something to that direction and Widsets also.
My Featurelist for killer system
- Platform independent (like widget player software are different)
- Should be URL triggable/referrable so it could be intgrated to rest of the internet
- client side scriptable
- client side cache for data
- background data downloading
- open for anyone, not just for operators and commerial publishing people
June 20th, 2006 at 8:56 pm
The Opera Platform, Flash Lite, Widsets - are all great potential platforms to build widgets, gadgets and microbrowsers - however, we’re still a ways off from having a mobile authoring technology that anyone could use. These are still mainly ‘developer centric’ technologies.
Personally, I’d love to see the following:
- widgets could be played back on mobile devices, desktop web, and potentially other devices (TivO, etc)
- authoring could be done in a visual manner similar to Hypercard
- programmers could dig deeper and extend the platform for more integrated applications
- widgets are self-contained, distributable, shareable objects that could still be used WITHOUT a network connection.
- intelligent, persistent caching and connected sync/update allowing for occasionally connected devices and applications.
Personally, in playing around with Opera 9 widgets this morning - along with the Opera Platform idea I think there are some really interesting ideas there - especially when you factor in that Opera is also to ship on the Nintendo Wii and a version of their browser is available for the Nintendo DS. Having a cottage industry of user-generated contents and widgets for your DS and/or Wii would be incredible.
June 21st, 2006 at 1:26 am
Some comments:
* I do not think that “automatic translation” will be usefull from desktop to mobile, as it will generate too much trouble for both developper and user, and by definition widget should be very simple app.
* By selling the browser, opera might kill his own widget market. I am fan of widget, but I still can not test them on my Nokia as I need to buy Opera for this.
Anyway, I suggest you check my ongoing activities on this side here:
On mobile widget: http://blog.landspurg.net/?cat=14 or my microwidget attempt:
http://blog.landspurg.net/microwidget-is-on-the-way
June 30th, 2006 at 4:13 pm
my two pennies:
- yes, hypercard for anything would be well received - pc or mobile
- widsets is too simple and i think will never be more than simple for general consumption. it does have legs, though, for fancyer, developer-created stuff
- do you realize that the nokia browser is built on the same code base as apple safari and that the apple widgets are coded in a way that safari can use hence i keep wondering where such widgets might come in for the nokia browser (not that i know anything, despite my affiliation with nokia).
- hello, bryan.
- and, yes, the pc-mobile-web fusion needs to be a part of this and widgets are ideal for this point of convergence