Saturday, April 13, 2002

Mozilla rumbles towards 1.0. Do we care? I think the days of do-everything web browsers are over. There's no user interface continuity, maintaining user state is a hassle, undo and intermediate saving are virtually impossible, and when your net connection goes down, you have absolutely no recourse. (Try working with a teacher using a web-based curriculum and watch the net connection drop. In many cases like this, you simply cannot risk hanging users out to dry.) Much more interesting to me is watching how applications have been growing more and more web-enabled over time. It used to be an app might have a link to its company's web site. Then, apps started being able to check to see if newer versions were available. Now, we're seeing applications that are designed to harvest data off the web -- news feeds, webcam images, real-time data, etc. -- and present the information is ways that offer the user far more flexibility and control than they would have within a web browser. And the flow's moving both directions: weblog software sits in the middle of the flow, offering broadcasting as well as consumption, and new standards like SOAP are making it possible for a variety of web services to make themselves available to anything that can talk XML and HTTP. This is exciting. I'm looking forward to seeing more and more web-enabled applications that integrate the user interface and representational affordances of desktop apps with the rich and dynamic nature of the web. 11:58:04 AM     Portals or services? The Chronicle reports that the University of Michigan is shutting down my.umich.edu, its ongoing effort to provide its students with a customized portal to University information and services. U of M cites the extensive cost of development and maintenance as the main factor. I find it interesting that at the same time we see centralized portal development waning, interest and availability of weblogs is on the rise, and there's a lot of effort being placed on finding ways to allow bloggers to integrate content from other sources. Now, not everyone wants to broadcast a weblog, but many of the technologies for harvesting information from web services are being developed within the weblog community. Perhaps we'll see institutions like U of M starting to provide discrete, maintainable web services to their students, who choose themselves what to subscribe to. For example, I wonder what would happen if U of M, or another university, adopted a model more like the Radio Community Server. This would maintain the central, always-available nature of the portal approach, but distribute customization (and add significant broadcast capability) to the students. The down side is that right now, the learning curve for a tool like Radio is a bit higher than a portal site, and (of course) it's not free. But in two to five years, as web services and personal blogging environments mature, this could become the dominant model. And universities could be interesting community testbeds. 11:43:26 AM

© Copyright 2002 Eric Baumgartner . Last update: 11/5/02; 3:30:13 PM . This theme is based on the SoundWaves (blue) Manila theme.