Support displaying invalid UTF-8 in language-environment-specific ways. 2008-08-05 Aidan Kehoe <kehoea@parhasard.net> * specifier.el (current-display-table): Initialise this here, not in x-init.el, since we want it even on non-X builds to use the support for displaying Unicode error sequences according to the current locale. * mule/mule-cmds.el (set-language-info): Document error-sequence-coding-system, used to describe how to display characters that are not valid Unicode on disk. * mule/mule-cmds.el (finish-set-language-environment): Implement error-sequence-coding-system. * unicode.el (unicode-error-sequence-warning-face): New face, to make it possible to distinguish invalid Unicode sequences from the characters given by the valid Unicode sequences. * mule/cyrillic.el ("Russian"): ("Ukrainian"): ("Bulgarian"): ("Belarusian"): ("Cyrillic-ALT"): Add support for error-sequence-coding-system for all these languages. * mule/latin.el: Add support for error-sequence-coding-system for the Latin-alphabet language environments.

--- a/lisp/specifier.el	Sat Jul 26 13:50:27 2008 +0300
+++ b/lisp/specifier.el	Tue Aug 05 09:06:41 2008 +0200
@@ -988,4 +988,18 @@
 			 (specifier-instance specifier domain))))
 		   (list (cons nil inst))))))))))
 
+;; Character 160 (octal 0240) displays incorrectly under some X
+;; installations apparently due to a universally crocked font width
+;; specification.  Display it as a space since that's what's expected. 
+;;
+;; (make-char-table 'generic) instead of (make-display-table) because
+;; make-display-table isn't dumped, and this file is. 
+;;
+;; We also want the global display table to be actually globally
+;; initialised; that's why this is here, and not in x-init.el, these days.
+
+(set-specifier current-display-table 
+               #s(char-table type generic data (?\xA0 ?\x20))
+               'global)
+
 ;;; specifier.el ends here