DH> Martin Buchholz <martin(a)xemacs.org&gt; writes: DanH> I don't know if it's a bug or a feature - but this is all because DanH> XEmacs works differently from fsfmacs. Fortran-mode etc all define DanH> "\e\C-h" in a local keymap, which on XEmacs overrides both M-backspace DanH> and C-M-h. On fsfmacs it only overrides C-M-h (the intended DanH> behaviour). DH> On my machine, it just redefines M-DEL. Well, yes, but if you execute the keystroke pressing the keys labelled Meta + BackSpace it invokes the binding backward-char. DanH> That means that you have to re-assign backward-kill-word explicitly to DanH> M-BS if you also define C-M-h. DH> Yes, you're right (I did some testing). Three examples: DH> 3. DH> (progn DH> (setq foo-map (make-keymap)) DH> (define-key foo-map "\e\C-h" 'foo) DH> (use-local-map foo-map)) DH> gives Local Bindings: DH> M-backspace foo DH> M-C-h foo DH> in xemacs, but DH> ESC C-h foo DH> in fsf emacs. DH> In my opinion, this is pretty counter-intuitive (i.e. a bug) in DH> xemacs. But Emacs would do the same if it also mapped C-h to Backspace instead of DEL. E.g. (progn (setq foo-map (make-keymap)) (define-key foo-map "\e\C-i" 'foo) (use-local-map foo-map)) Makes both keystrokes Meta+Control+i and Meta+Tab invoke foo in Emacs. Perhaps you are arguing that C-h should not be mapped to Backspace. I might even agree with that. But that will open several old cans of worms. Martin