The encoding for a HTML document is specified by the Content-Type header  
from the HTTP server. When your file is not being served by a HTTP server,  
there are still some ways to get the browser / parser to detect the  
encoding, but those approaches tend to be flakey.

Your best bet is to use HTML entities instead. See:  
http://www.w3schools.com/tags/ref_symbols.asp
Thus, your special characters should appear as &#nnnn; in your document  
(unless they have a well-known entity name such as & which you may use  
instead).

HTH,

-Zig

First, how the heck are you getting UTF-8 into your doc files?  The
javadoc tool uses your platform encoding by default.

Second: try adding the -charset option to the javadoc command line.
This sets the <META> tag in your html output. It might clue your browser
what charset is being used.  Then, it might not.

-charset  name
    Specifies the HTML character set for this document. The name should
be a preferred MIME name as given in the IANA Registry. For example:

      C:> javadoc -charset "iso-8859-1" mypackage

    would insert the following line in the head of every generated page:

       <META http-equiv="Content-Type" content="text/html;
charset=ISO-8859-1">

I assume that "UTF-8" would work here as well as "iso-8859-1"

Try reading the reference for yourself, that's what I did.  I've never
used any of these options before:

<http://java.sun.com/j2se/1.4.2/docs/tooldocs/windows/javadoc.html#charset>