Does the X server in Rawhide eat Control-Space?

Peter Hutterer peter.hutterer at who-t.net
Thu Feb 9 13:47:12 UTC 2012


On Thu, Feb 09, 2012 at 08:05:49AM -0500, Aleksandar Kurtakov wrote:
> ----- Original Message -----
> > From: "Peter Hutterer" <peter.hutterer at who-t.net>
> > To: "Development discussions related to Fedora" <devel at lists.fedoraproject.org>
> > Sent: Thursday, February 9, 2012 2:50:02 PM
> > Subject: Re: Does the X server in Rawhide eat Control-Space?
> > 
> > On Wed, Feb 08, 2012 at 12:10:36PM -0500, Aleksandar Kurtakov wrote:
> > > ----- Original Message -----
> > > > From: "Nicolas Mailhot" <nicolas.mailhot at laposte.net>
> > > > To: "Development discussions related to Fedora"
> > > > <devel at lists.fedoraproject.org>
> > > > Sent: Wednesday, February 8, 2012 7:05:14 PM
> > > > Subject: Re: Does the X server in Rawhide eat Control-Space?
> > > > 
> > > > 
> > > > Le Mer 8 février 2012 12:17, Richard W.M. Jones a écrit :
> > > > >
> > > > > After updating everything on my laptop to Rawhide (forced to by
> > > > > usrmove) I noticed that emacs was no longer responding to
> > > > > Control-Space for setting the mark.
> > > > >
> > > > > It seems like the X server itself is eating this key
> > > > > combination,
> > > > > since emacs works fine in a virtual console, but doesn't work
> > > > > in
> > > > > any
> > > > > terminal or directly under X.
> > > > 
> > > > > It seems a bit odd that X would capture this key combination
> > > > > ...
> > > > 
> > > > IIRC it's consumed by input method switching now (whatever the
> > > > name
> > > > is this year)
> > > > 
> > > > I don't think it's a good idea, lots of stuff already grabbed
> > > > this
> > > > combo (IIRC
> > > > some xkb maps do depend on ctrl+space)
> > > 
> > > I have even opened FESCo ticket
> > > https://fedorahosted.org/fesco/ticket/798
> > > to say on that because personally I think it's unacceptable and
> > > projects
> > > with such bad behaviour should be forced
> > > to learn playing with the rest. At least we still can add conflicts
> > > in our
> > > packages when we have to deal such changes because in my case
> > > Eclipse is
> > > totally unusable and I consider this as the only possible way to
> > > have
> > > working Eclipse package if ibus is not changed.
> > 
> > to quote the ibus maintainer from the FESCo ticket: "we use [sic]
> > Control+Space for more than ten years"
> > 
> > There is a legitimate issue of conflicting keyboard shortcuts but I
> > do ask
> > that you tone down the inflamatory language, assume other people mean
> > well
> > and accept sometimes conflicts just happen. Once that happens, you'll
> > probably notice that issues tend to get sorted faster, with more
> > happiness
> > all round.
> 
> Peter, 
> If you have read the comments in bugzilla you would have seen that I've been doing that
> until I got the following "I don't use your listed applications as daily
> works except for emacs." which is more or less "I don't use that so I
> don't care" at which point I changed my tone too and I'm not sorry about.
> My tone is completely modelled by the responses I got. And note that I got
> this response as a reply to me listing "kdevelop, netbeans, qtcreator,
> eclipse, emacs (uses ctrl+space in a number of combinations), monodevelop.
> And there are other things that will be affected - everything that
> includes kate-part (virtuallly the whole KDE desktop+ kile, lyx and etc.).
> " as broken. 
> 
> Say honestly, Do you consider this constructive?

There is never a reason not to be polite and respectful in a conversation.
It's hard work, but it usually pays off.

Remember that communication is always interpreted subjectively and a
well-meant comment can be interpreted differently. My interpretation of
756595 is that the first few comments were, paraphrased, "ibus' breaks
breaks all these applications". 
If that was my component, I'd get defensive, especially if I hadn't changed
the shortcut in years. So that may explain the come-back that you then
interpreted as "i don't care". I gather that you have not used IM in the
over the few years, which explains the statement of "It would be very bad if
a system wide shortcuts starts messing with a well known application
shortcut." Imagine you had used IM for years, the situation would look quite
the inverse.

Note also that Akira asked for the .imsettings.log in comment 6 already, not
provided yet. Figuring out why ibus runs on your machine when it shouldn't
is probably the first step to resolve this. Those who do need ibus may take 
less issue with the conflicting shortcuts.

Cheers,
 Peter


More information about the devel mailing list