From xemacs-m  Tue Aug 19 00:43:21 1997
Received: from pppkifer.cs.sunysb.edu (sbkifer.cs.sunysb.edu [130.245.1.35])
	by xemacs.org (8.8.5/8.8.5) with ESMTP id AAA10929
	for <xemacs-beta@xemacs.org>; Tue, 19 Aug 1997 00:43:20 -0500 (CDT)
Received: from pppkifer.cs.sunysb.edu (localhost [127.0.0.1]) by pppkifer.cs.sunysb.edu (8.7.6/8.7.3) with ESMTP id BAA06918 for <xemacs-beta@xemacs.org>; Tue, 19 Aug 1997 01:43:34 -0400
Message-Id: <199708190543.BAA06918@pppkifer.cs.sunysb.edu>
From: kifer@cs.sunysb.edu (Michael Kifer)
To: xemacs-beta@xemacs.org
Subject: 4 MULE bugs
Date: Tue, 19 Aug 1997 01:43:33 -0400
Sender: kifer@pppkifer.cs.sunysb.edu



1. In select-input-method in mule-cmds.el, there is a pair of lines:

    (apply (car method-slot) method-name (cdr method-slot))
    (setq default-input-method
          (setq current-input-method (cons language-name method-name)))

    This has the unfortunate effect that the input-method-activate-hook (called
    from quail-mode) doesn't know the name of the input method to be
    activated. Can you reverse these two statements?

2.  Also, if you select Greek and quail-greek, then typing i:: inserts
    i *twice* (with 2 dots above it), plus it inserts ".
    I don't think it is supposed to.

3.  Selecting an input method doesn't put an indicator in the mode
    line (unlike in emacs).

4.  If I delete Quail Completion Buffer and then request a quail completion, I
    get an error (e.g., type i: in Greek, and then hit tab). It should be
    checking if the buffer exists and recreate it if it doesn't. Otherwise,
    accidental deletion of this buffer would hose the current session.
    


	--michael  

