1 files changed, 53 insertions, 0 deletions
diff --git a/runtime/lang/README.txt b/runtime/lang/README.txt
new file mode 100644
index 0000000..e5823f9
--- /dev/null
+++ b/runtime/lang/README.txt
@@ -0,0 +1,53 @@
+Language files for Vim: Translated menus
+
+The contents of each menu file is a sequence of lines with "menutrans"
+commands.  Read one of the existing files to get an idea of how this works.
+
+More information in the on-line help:
+
+	:help multilang-menus
+	:help :menutrans
+	:help 'langmenu'
+	:help :language
+
+The "$VIMRUNTIME/menu.vim" file will search for a menu translation file.  This
+depends on the value of the "v:lang" variable.
+
+	"menu_" . v:lang . ".vim"
+
+When the 'menutrans' option is set, its value will be used instead of v:lang.
+
+The file name is always lower case.  It is the full name as the ":language"
+command shows (the LC_MESSAGES value).
+
+For example, to use the Big5 (Taiwan) menus on MS-Windows the $LANG will be
+
+	Chinese(Taiwan)_Taiwan.950
+
+and use the menu translation file:
+
+	$VIMRUNTIME/lang/menu_chinese(taiwan)_taiwan.950.vim
+
+On Unix you should set $LANG, depending on your shell:
+
+	csh/tcsh:	setenv LANG "zh_TW.Big5"
+	sh/bash/ksh:	export LANG="zh_TW.Big5"
+
+and the menu translation file is:
+
+	$VIMRUNTIME/lang/menu_zh_tw.big5.vim
+
+The menu translation file should set the "did_menu_trans" variable so that Vim
+will not load another file.
+
+
+AUTOMATIC CONVERSION
+
+When Vim was compiled with multi-byte support, conversion between latin1 and
+UTF-8 will always be possible.  Other conversions depend on the iconv
+library, which is not always available.
+For UTF-8 menu files which only use latin1 characters, you can rely on Vim
+doing the conversion.  Let the UTF-8 menu file source the latin1 menu file,
+and put "scriptencoding latin1" in that one.
+Other conversions may not always be available (e.g., between iso-8859-# and
+MS-Windows codepages), thus the converted menu file must be available.