* Another Web browser, another widget
These days it seems that if you are the publisher of a Web browser you have to offer some kind of widget or gadget, or whatever you’d like to call them.
Apple provided such beasts with Dashboard under OS X and Yahoo got into the game when it purchased Konfabulator, now renamed Yahoo Widgets.
Following Yahoo came Google with its own take on the concept, Google Gadgets, and now we have more widget-style technology to look forward to in the forthcoming (we hope) Microsoft Vista.
The latest entrant into this somewhat over-active market is Opera with Opera Widgets.
Like Google and, as far as we know, Microsoft, Opera’s Widgets rely on services provided by the company’s own browser and cannot be run standalone. Apple’s Dashbaord widgets are all constrained to run as a group, which leaves only Yahoo Widgets as what could be considered true mini-applications that can behave like any other program.
As interesting as the architectural approaches are what can we make of all of this enthusiasm for what are essentially miniature client-side local applications and Web applications? I would suggest that it is a technology area that seems like a good idea but like so many good ideas is really looking for a problem to solve.
Sure, having the latest weather forecast floating around on your desktop is fun but it gets old quickly as do stock quote tickers, calendars and RSS feed readers.
If you are actually trying to get work done all of these toys are far less useful than they are to casual users. The problem is that casual users have a limited range of needs and they aren’t going to spend their money on widgets and gadgets when there’s plenty of real software they have yet to buy.
Just look at the libraries of widgets that the various Web browser publishers offer. They were all growing like gangbusters to begin with and with the exception of Opera, whose library is very new, they have all experienced a sharp downturn in the rate of additions to their libraries. Developers aren’t finding processes to implement as widgets that users care about.
A serious problem with widget technologies is that they are too complex for the average end user to create and not a valuable enough platform for developers to do more than dabble with.
It reminds me of the furor that surrounded the concept of mobile agents a few years ago (see “Distributed Objects & Components: Mobile Agents”, though many of the links are broken).
Mobile agents sounded great and some really clever and sophisticated technologies were developed that got us all excited. But the problem with mobile agents was that there was no real problem that this cool technology actually addressed and thus the idea became an historical footnote.
I’m betting that much the same will happen to widgets unless someone can find a use for them that’s more profound than another RSS feed reader or weather forecast display. I’m also betting that that won’t happen. Widgets RIP.
