Web Crossing uses a Web-Crossing-site
default-charset, which is normally ISO-8859-1 (also called Latin-1,
the 8-bit European character set). This is an 8-bit character set,
ASCII in the low half, with some additional European characters ,
such as å and É, in the high half.
You can set the default charset for the site in the General Settings
control panel or in the webx.set file.
To switch Web Crossing to Japanese processing mode, you must
install a default charset of ISO-2022-JP, or ISO-2022-JP-1, etc. Any
default charset that includes "2022-J" will automatically switch the
Web Crossing server to Japanese mode.
When switching into Japanese mode for an existing database, you
should first switch to Japanese mode, export the whole whole database, then
import it into an empty db in Japanese mode. This will clean up all
user-name differences between Japanese and non-Japanese mode.
ISO-2022 is a character set standard that allows embedded escape
sequences (ESC...) to change the current character set. So you can
switch from the Latin-1/European charset to various Japanese
charsets, and back again.
In Japanese mode, the Web Crossing internal character set is another
Japanese encoding called ShiftJIS. This is an 8-bit character set,
with mixed 1- and 2-byte characters. All characters starting with a
first byte of 0x81..0x9F or 0xE0..0xEF are two byte characters, and
characters 0xA0..0xDF are special 1-byte Japanese characters.
In Japanese mode, the web browser needs to be set to display Japanese
in ShiftJIS format. All displays to the Web browser will use
ShiftJIS, and all posts from the browser must come up in ShiftJIS.
Web Crossing automatically converts all incoming email and news
messages from ISO-2022 to ShiftJIS, all outgoing email and news
messages from ShiftJIS to ISO-2022-JP. (In WCTL, you have to use
string.toShiftJis and string.toJis conversions to manually convert
between formats as appropriate for building email, etc.)
In Japanese mode, Web Crossing will convert Japanese 2-byte spaces in
user names to one-byte ascii spaces, and is aware of the 2-byte
characters for all parsing, first-name/last-name algorithms, HTML