Welcome to the NewsGroup Gleaner

A new way to surf selected web pages, based on URLs that appear in the headers and text of usenet newsgroup articles.

The basic idea is this: specify a news group by name and the NewsGroup Gleaner will read all the articles in that news group, saving any http: URLs that it encounters. The results are then put into a new web page, that is then displayed.

Prerequisites: NetScape 3, Perl 5, Unix environment; this version runs under NetScape only --- not as a standalone appletviewer applet.

The Problem: The Gleaner is primarily a Java applet, but unfortunately mere applets do not have sufficient power to actually read news groups. Instead, the applet communicates with a small proxy server, written in Perl, that serves as relay between the Java applet and the news server.

Running the Proxy: You'll need to run a Perl script (located here ) in order to get the Gleaner to work. Customize the name of the news server in the first part of the script to match your environment; you may also need to change the first line to specify the location of the Perl interpreter.

Running the applet: Java is very particular about security with sockets and ports; an applet can't communicate with a process running on machine other than where the applet resides. Therefore, the way to get the Gleaner applet to work is:

1. Copy the following (this page, the two Java classes, the Perl script) to some local disk, say /tmp. The items are:
   • glean.html
   • gleanGroup.class
   • NewsError.class
   • newspass
2. Start a shell with newspass running.
3. Be sure you have enabled Java and JavaScript execution under your NetScape options.
4. Command NetScape to open the url file:/tmp/glean.html (tailored to your environment). Hopefully, now everything resides in the same machine -- the proxy server, the applet class files, and a copy of this page. It may actually run!