Wednesday, November 16, 2005

Some Research Perspectives on Ajax

Despite Ajax is cool, it is often said that it is not about technologies but just about techniques.
Then, are there any research oppotunities on this topic?

A good platform for a rich client.
  • In this sense, Ajax is important only because it is based on standards. Usually, nothing is required for client node configuration.
  • Problem here is that it has several restrictions for building rich GUI, such as, the "back button" problem. Employing a better client environment, which is not a standard, instead of Ajax doesn't make sense from this viewpoint. For example, proposing better GUI with Macromedia Flash is not so good way as a research direction. Proposing better standards for better GUI (based of studies on what is desired, which is not available with the current standards) makes some sense, though.
A target of GUI-enabled software development.
  • In this sense, Ajax is yet another target platform in developing application software.
  • Research direction arized here is the unification of development efforts for other possible GUI platforms, that are, AWT/SWT for example. For example, the goal "write once to run on both Eclipse and Ajax" is not bad.
A good GUI with dynamic interactions between human and network.
  • In this sense, applications with Ajax often provides good user experience. I believe it mostly comes from the asynchronousness thanks to XMLHttpRequest. Though it must have been possible with conventional fat clients, they have rarely done maybe because of some difficulty in use of Web Services (REST, SOAP, whatever) with clients (not servers).
  • There may be rooms for investigating potential GUI enhancements in this area.

No comments: