Go to the Japanese site:

http://www.kobe-cufs.ac.jp/kobe-city/index-j.html

Auto-detect on the charset shows most of the page ok. Now try selecting one of the other charsets to see if it improves: in order -- EUC, JIS, ... from the View/Character sets/Japanese menu.

With the JIS setting I get the following exception:

sun.io.MalformedInputException
        at sun.io.ByteToCharJIS.convert(ByteToCharJIS.java:119)
        at java.io.InputStreamReader.convertInto(InputStreamReader.java)
        at java.io.InputStreamReader.fill(InputStreamReader.java)
        at java.io.InputStreamReader.read(InputStreamReader.java)
        at java.io.InputStreamReader.read(InputStreamReader.java)
        at sun.html.Parser.parseInvalidTag(Parser.java:1071)
        at sun.html.Parser.parseTag(Parser.java:1237)
        at sun.html.Parser.parseContent(Parser.java:1308)
        at sun.html.Parser.parse(Parser.java:1415)
        at sun.hotjava.doc.DocParser.run(DocParser.java:399)
        at java.lang.Thread.run(Thread.java)


mark.reinhold@Eng 1997-01-26: I have extracted the HTML file in question and
attached it to this bug report. Using the native2ascii utility, I am able
to convert it using the EUCJIS encoding, but both the SJIS and JIS encodings
throw sun.io.MalformedInputException. It's not clear to me that all three
encodings should work for this file, so I'm assigning this to an expert.