Bug 350336

Summary: KTouch lags and uses full (100%) CPU resources when typing fast.
Product: [Applications] ktouch Reporter: Dimitrios Charalampidis <d.s.charalampidis>
Component: generalAssignee: Sebastian Gottfried <sebastian.gottfried>
Status: RESOLVED FIXED    
Severity: grave CC: 1234qwer, atelszewski, bugs.kde.org, deletesoftware, druchaty, gutsy-wooer-dual, mr.zoltan.gyarmati, m_105, simgunz, sobhanpe2012, ulr_bugs_kde_org
Priority: NOR    
Version: 2.3.0   
Target Milestone: ---   
Platform: ROSA RPMs   
OS: Linux   
Latest Commit: Version Fixed In: 16.12.0
Sentry Crash Report:

Description Dimitrios Charalampidis 2015-07-18 07:26:31 UTC
While typing fast, KTouch consumes 100% of my CPU, which results in lagging visualization of the key presses and typed characters, making it very difficult to use the program.

Reproducible: Always

Steps to Reproduce:
1. Open KTouch
2. Start typing fast

Actual Results:  
Lagging visualization of the typed characters and very slow performance, consuming 100% of CPU.

Expected Results:  
Visualization of the key presses with virtually no lag and fair CPU usage. 

Tested on two PCs, a desktop and a laptop as follows:

Desktop:
Intel(R) Core(TM)2 Quad CPU Q9300 @ 2.50GHz
Logitech K340 wireless keyboard

Laptop: 
Intel(R) Core(TM)2 Duo CPU L7500 @ 1.60GHz)
ThinkPad built-in keyboard
Comment 1 Ian Schwarz 2015-12-18 20:13:59 UTC
This happens for me, but ONLY when using Qt 4.8.7, not 4.8.6.
Comment 2 Ian Schwarz 2015-12-18 20:23:12 UTC
Can also confirm that the fix proposed by  Moritz Nisblé in this thread FIXES the bug:
https://bugs.kde.org/show_bug.cgi?id=346248
I'll be posting a review request.
Comment 3 Zoltan Gyarmati 2016-08-04 14:03:29 UTC
Experienced the same behavior on KDE Neon with version 4:16.04.3-0neon+16.04+build2.
The linked patch indeed seems to eliminate an issue, is there any chance to get it into the mainline?
Comment 4 poelzi 2016-08-29 23:11:58 UTC
I tested the patch. Without it, ktouch IS NOT USEABLE.

Version: 4:15.12.3-0ubuntu1

Sorry, but a type-trainer should be able to run on a Core 2 Duo. If a beginner can type faster then the program updates the display, you are seriously broken. Please apply the patch or fix the underlying problem.
Comment 5 Simone Gaiarin 2016-09-07 11:14:03 UTC
For arch users I've created the package kdeedu-ktouch-patched in AUR, that provides the proposed patch, while we wait it's merged upstream.
Comment 6 Sebastian Gottfried 2016-11-09 12:33:46 UTC
*** Bug 346248 has been marked as a duplicate of this bug. ***
Comment 7 Sebastian Gottfried 2016-11-09 12:34:50 UTC
*** Bug 362316 has been marked as a duplicate of this bug. ***
Comment 8 Sebastian Gottfried 2016-11-09 12:44:10 UTC
KTouch now uses Qt Quick 2 which is more efficient than the previous version.

Also I've fixed a grave mistake in the update login of the lesson painter resulting in they more repaints than necessary.

Now I can't get KTouch to saturate one of my CPU cores anymore. On my system KTouch uses usually about 5 % on one core during training.
Comment 9 Andrzej Telszewski 2016-11-09 19:51:51 UTC
Is the source tarball of the new version available somewhere?
Comment 10 Sebastian Gottfried 2016-11-09 21:24:00 UTC
No, not yet. It will be released as a part of KDE Application 16.12 on December 15. You can grab a copy of the source code from the repository. All the fixes are in the master branch.
Comment 11 Andrzej Telszewski 2016-11-09 21:30:41 UTC
I was hoping it'll compile on KDE4, but it won't.
See you once I move to KDE5 :-)
Comment 12 Christoph Feck 2016-11-11 15:44:49 UTC
It is possible to use Qt5 applications on the KDE4 desktop (as long as required libraries are installed).
Comment 13 Andrzej Telszewski 2016-11-11 16:26:48 UTC
Yep, but it's not that simple.
There are more and more KF5 libs required.
Comment 14 Nikolay Druchaty 2017-02-21 21:41:17 UTC
Hi Sebastian,

I've tried to install KTouch from the sources. I've installed 'kqtquickcharts' mentioned in README file.
But KTouch still can't be run.
I got "module "org.kde.charts" is not installed".

I use Ubuntu 16.10. Could you please help me to solve the problem?
Comment 15 Christoph Feck 2017-03-31 17:31:06 UTC
*** Bug 378118 has been marked as a duplicate of this bug. ***