GNU bug report logs - #30043
25.3; C-M-e inserts euro character

Previous Next

Full log

Message #57 received at 30043 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Kangas <stefan <at> marxist.se>
To: Tomas Nordin <tomasn <at> posteo.net>
Cc: Eli Zaretskii <eliz <at> gnu.org>, 30043 <at> debbugs.gnu.org,
 Ilya Zakharevich <ilya <at> math.berkeley.edu>
Subject: Re: bug#30043: 25.3; C-M-e inserts euro character
Date: Sun, 29 Sep 2019 17:42:59 +0200

tags 30043 - notabug
reopen 30043
thanks

Tomas Nordin <tomasn <at> posteo.net> writes:

> >> So I guess you somehow have a keyboard which doesn't fit the new
> >> keyboard input code introduced by Emacs 25, and need to use this
> >> fallback variable to avoid the problem.
> >
> > I believe that was the correct answer, and nothing of the rest of the
> > discussion indicated otherwise.  I'm therefore closing this as notabug.
>
> It seems to me then that the correct answer is that keyboards with
> swedish layout on Windows does not fit the new keyboard input code. I
> see the same on a laptop now with emacs 26.2. It is strange that no-one
> else in this country complains.
>
> Using the work-around with setting w32-use-fallback-wm-chars-method
> solves the problem. But the docstring of that variable suggests that I
> shouldn't have to use it unless debugging something.
>
> I am not going to argue with the closing but I don't think this is
> specific to one specific keyboard. It was reproduced on several
> keyboards on different machines.

Hi Tomas,

Thanks for writing back.  It was not my intention to close a
legitimate bug; reopened it now.  I've been investigating this a bit,
and it seems like this new keyboard handling code was introduced due
to deficiencies in the old one.  See:
https://debbugs.gnu.org/cgi/bugreport.cgi?bug=19994
See also the possibly related: https://debbugs.gnu.org/bug=23251

If it's the case that this change breaks M-C-e on all Swedish
keyboards on MS-Windows, perhaps it's worth making sure the users
doesn't have to set w32-use-fallback-wm-chars-method to non-nil.  At
the very least, it makes sense to me that the doc string should be
improved as Tomas suggests, and possibly also that the workaround
should be documented in e.g. etc/PROBLEMS.

Eli, what do you think?  I've also Cc:ed Ilya Zakharevich in the hope
that he will be able to help.

Also: is anyone else seeing this on their Swedish keyboard on
MS-Windows?  And Tomas: Are you seeing the same thing with other key
bindings involving one or more modifiers?

I have a Swedish keyboard but no MS-Windows system, so I can't dig
into this beyond this.  I'm also a bit out of my depth.  Therefore,
I'll have to leave it to someone else to decide how to best proceed
here.

Best regards,
Stefan Kangas

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.