Charles G Waldman wrote: Sure, but the Right Thing To Do would be to fix read-minibuffer-internal (I think) to work in terms of keysyms rather than ASCII codes. I suggested the following patch: --- lisp/events.el~ Sun Apr 5 04:59:48 1998 +++ lisp/events.el Fri May 29 04:33:49 1998 ＠＠ -120,8 +120,6 ＠＠ (while pairs (puthash (pop pairs) (pop pairs) keyboard-translate-table))) -(put 'backspace 'ascii-character ?\b) -(put 'delete 'ascii-character ?\177) (put 'tab 'ascii-character ?\t) (put 'linefeed 'ascii-character ?\n) (put 'clear 'ascii-character 12) in: Date: Fri, 29 May 1998 05:31:45 +0100 (BST) Message-ID: <13678.14897.742118.968904(a)cerise.sensei.co.uk&gt; and again in: Date: Mon, 8 Jun 1998 15:45:27 +0100 (BST) Message-ID: <13691.63751.157978.716279(a)cerise.sensei.co.uk&gt; but nobody has chosen to comment on it. This fixes the set-variable problem, although it may well break other stuff. However, I've had (remprop 'backspace 'ascii-character) (remprop 'delete 'ascii-character) in my ~/.emacs for several months now, and haven't noticed any adverse effects (although I only use a few packages). -- Glynn Clements <glynn(a)sensei.co.uk&gt;