Re: w32-pass-lwindow-to-system (& ...rwindow)

[Lennart Borgman]

Eli Zaretskii wrote:
> > Date: Sun, 24 Sep 2006 02:50:28 +0200
> > From: Lennart Borgman <lennart.borgman.073@student.lu.se>
> > CC:  help-gnu-emacs@gnu.org, Eli Zaretskii <eliz@gnu.org>
> > X-Spam-Status: No, score=0.5 required=5.0 tests=FROM_ENDS_IN_NUMS 
> > 	autolearn=no version=3.0.4
> >
> > Mathias Dahl wrote:
> >     
> > > Eli Zaretskii <eliz@gnu.org> writes:
> > >
> > >   
> > >       
> > > > Looks like the trick we play around line 3005 of w32fns.c isn't
> > > > working.  What Windows version is that?  Is there someone on this
> > > > list who can make LWIN-r combo work for them in Emacs, in any
> > > > version of Windows?
> > > >     
> > > >         
> > > I just tested with emacs -q and it works as for the parent poster,
> > > i.e. Win+s gives me `search-backward-regexp' but Win+r opens up the
> > > Run dialog. The same applies for Win+e (open Explorer.exe).
> > >
> > > I use CVS Emacs from 2006-03-20 on Windows XP.
> > >   
> > >       
> > This behaviour is consistent with MS documentation. To avoid that 
> > windows left and right keys open for example the run dialog with Win+r 
> > you have to implement a low level keyboard hook.
> >     
>
> Can you give a pointer to the Microsoft documentation which says that,
> and/or documentation that explains how the LWIN-r combo works in terms
> of Win32 API calls, and why it bypasses the normal keyboard reading
> Emacs uses?
>   

See the thread where we discussed this:
http://lists.gnu.org/archive/html/emacs-devel/2005-07/msg00448.html


>   
> > That is what I have done in the patched version of Emacs that you can 
> > download together with EmacsW32.
> >     
>
> I'm well aware of your patched version.  The fact is, however, that we
> do seem to have code in w32fns.c that tries to work around the Run
> dialog being popped.  I asked my question (above) to try to understand
> whether that code works at all, on any Windows system.  Presumably, it
> worked for someone at some point in the past, or else it would not
> have been added to Emacs.
>   

I think the documentation I pointed to in the beginning of the thread 
answers this questions too.