|Java menu knowledge base|
XIXL - non-Roman character sets
This applies to people who want to use character sets such as Greek, Cyrillic, Hebrew, Arabic, Chinese, Japanese, Korean or any other non-Roman character set. It does NOT apply to people who require modified Roman characters such as è, é, à, ä, ö, ç ü, ñ, etc.
How to implement non-Roman characters on an IMINT.COM menu applet
Character set codes
- If you want a Double Byte Character Set (DBCS), ensure you are using an original Windows operating system for the character set in question; e.g. to get Chinese characters, do NOT use English Windows with Chinese character support; use Chinese Windows. For single byte character sets, English Windows with an installed character set for the language in question will be enough.
- Use Windows 98 2nd ed., Windows ME, Windows 2000, Windows XP or later. (For other operating systems, please conduct your own testing; remember that Netscape will produce positive results on a wide range of other OS's, but MSIE may not).
- When writing your XIXL menu, do NOT write UNICODE. Simply type your characters into your index file as you would type them into any other kind of text document.
- EITHER: if your IMINT.COM menu supports the menusourcetype parameter and the
menuencoding parameter, set the menusourcetype to the #2 read-algorithm and the menuencoding to the international character set code of your desired character set (that's the same code as you would write into the header of HTML documents to determine the character set);
- OR: if your IMINT.COM menu does not support these parameters (or if you wish to use parameter indices anyway), do NOT use an external source, instead use parameters for your XIXL index.
These are some examples of values for the "menuencoding" (somtimes "menu.encoding") parameter;
the codes are case-sensitive and may be OS-dependent.
Example: <param name=menuencoding value=Cp1253>
- GB2312 : Chinese (simplified)
- Cp1251 : Cyrillic
- Cp1253 : Greek
- Cp1254 : Turkish
- Cp1255 : Hebrew
- Cp1256 : Arabic
- Cp1257 : Baltic
One problem we found during testing is that character sets on a system
may not be available with all font settings. For example, one Netscape/Windows98
combination did not support font sizes 6, 7, 9, 11, 13, 16 and 18 on non-Unicode character sets
(but 10, 12, 14, 15, 17, 19 and 20 were OK, as were all Unicode sizes).