[mew-int 2766] Re: Mew 6.3 release candidate 2

On Thu, 12 Nov 2009 09:02:03 +0900, Kazu Yamamoto wrote:
> 
> > This is not immediate.  For example the presence of "쩍 selects
> > iso-8859-7 (which is fine).  If I used 쩐 instead of 쩌, then euc-kr
> > would have been chosen.  This is OK for me but it is not following the
> > above rule.
> 
> Is euc-kr chosen for a real message (not for a short test message)?
> Again, Emacs 23 has LESS information than Emacs 22. So, perfect
> backward compatibility is not possible. If things goes well in REAL
> life (not in a corner case), I think it is fine.

This is unlikely because the symbols of 8859-1 that were dropped in
8859-15 are seldom used.  I think one can live with the current
situation (maybe the manual could explain which variable to set to
reduce the allowed charsets to, say, iso-8859-1, iso-8859-15 and utf-8
for those who find this unbearable).

Best,
C.