Default IM configuration causes hangs

Bug #1425510 reported by Cefn
10
This bug affects 1 person
Affects Status Importance Assigned to Milestone
lubuntu-meta (Ubuntu)
Incomplete
Undecided
Unassigned

Bug Description

Certain applications have associated input hangs, where no key input events can be sent which sometimes clear after some time, and in at least one case (Inkscape), this is eliminated by running from Start->Run as follows, hence overriding the declaration of GTK_IM_MODULE in /etc/X11/Xsession.d/70im-config_launch , distributed in im-config...

bash -c "unset GTK_IM_MODULE && inkscape"

A permanent fix on my machine could be to modify the 70im-config_launch file, or declare an overriding export in some later .rc file. Anyone know where I should make this declaration if not modifying package-maintained files?

An alternative solution could be to install scim-gtk-immodule and run...

bash -c "export GTK_IM_MODULE=scim && inkscape"

See also https://bugs.launchpad.net/ubuntu/+source/inkscape/+bug/1385009

Similar behaviours are experienced with Kupfer - minutes of hang with apparently no input events being passed, followed by (sometimes) unfreezing.

I'm running Lubuntu Utopic with LXDE and Compiz (minimal dependencies on GTK), so it's feasible that some component is not properly declared in a dependent package. I can't find whether there is in fact a xim module available and running anywhere on my system. How can I check?

Other input issues relating to out-of-order key event delivery to Chrome-based applications are experienced on the same build, but I think this may be an independent bug, ostensibly resolved by declaring IBUS_SYNC_MODE=1 https://code.google.com/p/chromium/issues/detail?id=456766 but thought I should mention as it's kind of in the same subsystem and may be linked.

I speculate that there is some missing dependency or bad configuration in the default presentation of IM capabilities, which eventually causes a DBUS timeout or similar, eventually causing a fallback without any international input, but which is functional.

ProblemType: Bug
DistroRelease: Ubuntu 14.10
Package: im-config 0.26-2ubuntu1
ProcVersionSignature: Ubuntu 3.16.0-30.40-generic 3.16.7-ckt3
Uname: Linux 3.16.0-30-generic i686
ApportVersion: 2.14.7-0ubuntu8.1
Architecture: i386
Date: Wed Feb 25 12:26:26 2015
InstallationDate: Installed on 2015-01-01 (54 days ago)
InstallationMedia: Lubuntu 14.10 "Utopic Unicorn" - Release i386 (20141022.1)
PackageArchitecture: all
ProcEnviron:
 LANGUAGE=en_GB:en
 TERM=xterm
 PATH=(custom, no user)
 LANG=en_GB.UTF-8
 SHELL=/bin/bash
SourcePackage: im-config
UpgradeStatus: No upgrade log present (probably fresh install)

Revision history for this message
Cefn (6-launchpad-net-cefn-com) wrote :
Revision history for this message
Aron Xu (happyaron) wrote :

That's problem in the applications, not input methods or im-config. For the chrome stuff it's known you have to get GTK2 IM Module installed, otherwise the XIM support in Aura is broken (and Google is not intending to fix that).

Changed in im-config (Ubuntu):
status: New → Invalid
Revision history for this message
Cefn (6-launchpad-net-cefn-com) wrote :

Hi, Aron. Thanks for looking into this bug for me.

I acknowledge what you're saying that it's someone else's fault, but without more detail about what you actually think is going on, this may be misinterpreted. I personally have no idea what is going on, and can't contribute to resolving this through other projects' bugtrackers without your contribution. What 'kind' of bug is in the applications, that means an im-config change can workaround it?

Recently, keyboard input problems, including loss of responsiveness and loss of individual key events, seem to apply to almost all of my applications actually (even incredibly simple ones like Geany and Terminal), so if all of these applications have the same problem, I doubt whether it's an individual application developer at fault.

Regarding the Chrome bug, when you say "you have to get GTK2 IM Module installed" that is not a package name, so I can't install it.

Are you implying that the fault is in not having any IM Module installed? (I have to read between the lines as your comment was very terse) I already have both ibus and scim installed, (and ibus was selected under Preferences>Language Support>Keyboard Input Method System), so this is not the case on my system.

I also can't file an upstream bug to get anyone to package it (either Chrome OR Chromium - where Ubuntu maintainers can control packaging). Do you have one or more package names from Ubuntu repositories that your suggestion might relate to.

For reference, those others experiencing a problem, I've found a workaround right now (on Lubuntu) is to entirely disable international input support, by going to Preferences>Language Support>Keyboard Input Method System and selecting 'none', then restarting the system.

Revision history for this message
Gunnar Hjalmarsson (gunnarhj) wrote :

On 2015-02-26 12:13, Cefn wrote:
> For reference, those others experiencing a problem, I've found a
> workaround right now (on Lubuntu) is to entirely disable
> international input support, by going to Preferences>Language
> Support>Keyboard Input Method System and selecting 'none', then
> restarting the system.

I was about to propose that, but now I don't need to. ;) It's much more straight forward than the ideas mentioned in the bug description.

Revision history for this message
Nobuto Murata (nobuto) wrote :

@Cefn,
I'm curious you have installed ibus-gtk and ibus-gtk3 or not. Because you installed gtk2 app(e.g. inkscape) on Lubuntu. FWIW, Ubuntu has dependency to ibus-gtk and ibus-gtk3 by default because it's has gtk2 and gtk3 apps in the image.

Revision history for this message
Cefn (6-launchpad-net-cefn-com) wrote :

Summary: At this point I have a stable desktop where I don't have to reduce my touch-typing speed by a factor of 20 in order to achieve any text input at all (which is how it was this morning before disabling IBUS in my Language Support preferences).

I think ibus-gtk was already installed by default, and selected (at least that's what shows in the Language Support GUI). However, I don't know how ibus-gtk relates to xim, and the problems between xim and e.g. Inkscape remain unspoken.

The assertion that no IM module was installed seems to be an assumption from @happyaron but it's hard to tell if that's what he meant, or why, from his very short comment.

However, querying the environment variables by running 'export' in bash you see...
declare -x GTK_IM_MODULE="xim"

...both before (when ibus was selected) and after selecting 'none' in the Language Support GUI and logging out and in again.

I have no idea where this is coming from, but I suspect it's to do with im-config, given it explicitly exports that environment variable I couldn't find another reference to it in my profile or session files, but I'm not really sure what I'm looking for.

I installed SCIM manually, (scim-gtk-immodule) following advice online that running applications with SCIM was a good way to workaround the bug/feature interaction/whatever (no-one seems to want to talk about the actual bug), as described at http://www.inkscapeforum.com/viewtopic.php?f=29&t=18042 which I needed to resolve because of https://bugs.launchpad.net/ubuntu/+source/inkscape/+bug/1385009

IBUS seems to also be related to this serious bug I was also experiencing... https://code.google.com/p/chromium/issues/detail?id=456766 so disabling it seemed like a good idea.

Given my keyboard input is all in ASCII (as you'd expect from a en-GB or en-US configured laptop) I fail to see why there's internationalisation support baked into my desktop anyway, especially if it's buggy and prevents so many applications on the desktop from operating. However, so far I would say I have a slim grasp of what the actual bug might be, despite all the reports I've read. So far, only workarounds and buck-passing.

Searching the repositories for anything to do with a binary or lib ending xim gives me the following, and neither im-switch or uim-xim are installed...

$ apt-file search --regexp 'xim$'
cyrus-common: /usr/share/doc/cyrus-common/README.exim
cyrus-doc: /usr/share/doc/cyrus-doc/README.exim
debian-lan-config: /usr/share/debian-lan-config/fai/config/scripts/MAIL_SERVER/10-exim
devscripts: /usr/share/doc/devscripts/examples/forward.exim
exim4-daemon-heavy: /usr/sbin/exim
exim4-daemon-light: /usr/sbin/exim
im-switch: /etc/X11/xinit/xinput.d/default-xim
im-switch: /etc/X11/xinit/xinput.d/th-xim
logcheck-database: /etc/logcheck/ignore.d.server/sa-exim
logwatch: /usr/share/logwatch/scripts/services/exim
mlterm-common: /etc/mlterm/xim
postgrey: /usr/share/doc/postgrey/README.exim
sa-exim: /usr/share/lintian/overrides/sa-exim
sks: /etc/sks/forward.exim
sks: /usr/share/doc/sks/sampleConfig/debian/forward.exim
uim-xim: /usr/bin/uim-xim

Revision history for this message
Cefn (6-launchpad-net-cefn-com) wrote :

In answer to the specific question about ibus-gtk3 no, it is not installed. I guess because it's not a dependency of the Lubuntu Utopic build. Should it be?

Revision history for this message
Gunnar Hjalmarsson (gunnarhj) wrote :

On 2015-02-26 17:57, Cefn wrote:
> In answer to the specific question about ibus-gtk3 no, it is not
> installed. I guess because it's not a dependency of the Lubuntu
> Utopic build. Should it be?

I think it would be a good idea.

To confirm it, can you please do these steps:

* install ibus-gtk and ibus-gtk3

* switch back to "IBus" in Language Support

* Re-login

and let us know if it makes a difference?

affects: im-config (Ubuntu) → lubuntu-meta (Ubuntu)
Changed in lubuntu-meta (Ubuntu):
status: Invalid → New
Revision history for this message
Chris Guiver (guiverc) wrote :

Thank you for taking the time to report this bug and helping to make Ubuntu better. We are sorry that we do not always have the capacity to look at all reported bugs in a timely manner. There have been many changes in Ubuntu since that time you reported the bug and your problem may have been fixed with some of the updates. It would help us a lot if you could test it on a currently supported Ubuntu version. When you test it and it is still an issue, kindly upload the updated logs by running only once:
apport-collect <bug #>

and any other logs that are relevant for this particular issue.

Changed in lubuntu-meta (Ubuntu):
status: New → Incomplete
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.