Having previously declared it deader than a Norwegian Blue, I’ve decided to resurrect my earliest, and unquestionably most obsolete, public Java project. The Football Statistics Applet (FSA) has been providing client-side league tables and statistics to numerous football websites for 8 or 9 years now.
Applets? Java 1.1? AWT? So last century. But despite that, it continues to be undoubtedly the most popular piece of software that I’ve written, even though I might have preferred one of my other projects to have claimed that tepid distinction.
A server-side solution (probably PHP and MySQL) to the same problem seems much more sensible, but it was born of an age (1999/2000) when even static web hosting accounts were much more expensive than those with server-side processing and databases are today.
So, other than a developer who is not able to let go of one of his creations, why not let it die?
Firstly, applets might just be about to make something of a come-back (possibly? …no? OK then). Sun is finally addressing some of the issues that have made them unpopular and is positioning Java as a viable contender in the RIA space. But even if this new momentum comes to nothing, the core of FSA could be useful in other applications. Maybe a servlet-based web-app, or it could be converted to a WebStart application.
Secondly, I receive enough mail about FSA to convince me that there are still people interested in seeing it revived. There are a number of minor enhancements and major new features that have been requested over the years.
Previously I avoided open-sourcing FSA because I didn’t feel that the source code represented the best that I was capable of. I was insecure. I didn’t want to be judged as a shoddy programmer. Now I don’t care what you think so FSA is now on Java.net under the terms of the GPL. I’ve started on dragging it into the 21st century. The first thing that had to go was the dependency on Java 1.1. I’ve converted everything to use Java 5.0 collections and language features and made the code a bit more OO (it’s now not so much C-in-disguise as it was before). The next task is to replace the ugly and rather basic AWT GUI with Swing.
Anyone wanting to join the effort of improving the mediocre code (no unit tests) or the rather sparse documentation is welcome to request a role on the project via Java.net.