# HG changeset patch
# User Richard M. Stallman <rms@gnu.org>
# Date 1175816348 0
# Node ID b91b3ffd95b9fbcd1fb613ff42f3e9f6de1016ed
# Parent  d1d2c4bdd4133eb42336789ec9f966e9a2267467
Clarify potential drawback of previous change.

diff -r d1d2c4bdd413 -r b91b3ffd95b9 etc/TODO
--- a/etc/TODO	Thu Apr 05 17:57:05 2007 +0000
+++ b/etc/TODO	Thu Apr 05 23:39:08 2007 +0000
@@ -232,20 +232,24 @@
   user-selected input method, with the default being the union of
   latin-1-prefix and latin-1-postfix.
 
-** Switch the Windows port to using Unicode keyboard input.
+** Switch the Windows port to using Unicode keyboard input (maybe).
   Based on http://msdn2.microsoft.com/en-us/library/ms633586.aspx,
   this boils down to (1) calling RegisterClassW function to register
   Emacs windows, and (2) modifying ALL system messages to use Unicode.
   In particular, WM_CHAR messages, which result from keyboard input,
-  will then come in encoded in UTF-16.  Problem: using this on Windows
-  9x/ME requires to install the Microsoft Layer for Unicode (MSLU),
-  which might not implement all the required functionality that is
-  available built-in on Windows XP and later.
+  will then come in encoded in UTF-16.
 
   One advantage of switching to Unicode is to toss encoded-kbd usage,
   which will solve the problem with binding non-ASCII keys with
   modifiers.
 
+  Problem: using this on Windows 9x/ME requires installing the
+  Microsoft Layer for Unicode (MSLU), which might not implement all
+  the required functionality that is available built-in on Windows XP
+  and later.  We should not make this change if it would pressure
+  users of unauthorized copies of older versions of Windows to
+  downgrade to versions that require activation.
+
 ** Implement a clean way to use different major modes for
   different parts of a buffer.  This could be useful in editing
   Bison input files, for instance, or other kinds of text