Preface

The computer language, prolog, was once a celebrity, convincing us that we would be soon surrounded by lots of intelligent things like talking newspaper, thinking kettles, refrigerators recommending us what to eat tonight, etc. A great amount of money was poured into the research projects to materialize intelligent agents and quite a few of people worked hard for the goal. The party was over, however, about ten years ago when people realized that we had to continue spending on the project for at least another thirty years till we will actually see smart agents helping us daily. Well, we could at worst educate younger generations of researchers with the money from the goverment (Oops, No, from tax payers), was it not? Forget about the dead boy. We had altogether nice time, did we not? Bye, bye, artificial intelligence...

Time passed. Internet came in. The rise of network community, where we are all interconnected to each other on-line (or in-air), must take us forward towards the society supported by many intelligent agents, artificial or natural. This is a chance for us, retired prolog programmers, is it not? We knew that the shortage of data was one of the major obstacles in bringing artificial intelligence to life, but look, we have now plenty of resources on the net, in hand, off-the-shelf, all over the world. Why do we not dive into the world with our knowledge and skills to turn our Internet community into the knowledge society?

Okey, you are moved, aren't you? a little bit? Good, here is the map to sail into the world.

Preparing your working environment
Installing tools
Samples and test-runs

The first part invites you to building a Unix-like environment on you Wintel machine. I assume that you are more or less forced to work on Windows 9X for education or business. We cannot of course fight Microsoft, but there is no reason to accept everything provided by them. Let us build a working environment with which we are familiar. The second part instructs you to install various tools including jdk. There are really lots of tools available to us to build Internet applications. Our principle is termed as follows:

We use open and free software. We academics can only accept the products whose sources we can read through. Another point is that we normally do not have enough money to buy each student a commercial software.
We prefer the tools workable on Windows9X to ones workable on Linux. I am not particularly fond of the products by MS, but Wintel machines are so prevalent and the mojority of students have no experience with Unix. Asking them to install Linux on their machines and to switch between Windows and Unix is simply too much for them, I am afraid.

The last part presents you some sample codes to give you an insight into the possibilities of AI-applications on Internet.

Finally, I must confess that I myself have not built anything concrete along the line I am proposing in this guide. It took me sometime to realize how we can import our knowledge and skills of AI into Internet under the restrictions mentioned above, i.e., open and free software, using Windows9X, etc. I would say that it was impossible to produce a document similar to this two years ago when very few texts were available on XML. We had very few tools for Java/XML, too, at that time. The situation has greatly improved recently and we can now even choose among products. We rather encounter another problem, that is, it is rather hard to follow the progress, to catch up the speed in which people produce and upgrade tools. There is, thus, I believe, some reason to propose a way to select and combine tools available in public for particular purposes.

I had better stop here to move to more useful stuff. I appreciate your comments, corrections, criticisms, whatever, on this guide. Thank you.

Tsutomu Fujinami, 15th June, 2000 at JAIST

A Java/XML guide for prolog programmers on Windows 9X/Preface