I switched to static website generation with Jekyll in 2019 (probably because of Carl Schwan, who is somehow my
guiding-spirit in webulous things even if he does Hugo things nowadays). That means Ruby,
and I’ve got various bits-and-notes documents saying things like “When Ruby breaks again, …”
This blog post is my attempt to document for myself, once-and-for-all, my setup.
Basically, this comes down to (gently) fighting the FreeBSD packaging system.
Ruby 3.3 Incompatible
With Jekyll 4.3.1, trying to run it with Ruby 3.3 leads to the following error message:
jekyll 4.3.1 | Error: undefined method `[]' for nil
/usr/local/lib/ruby/3.3/logger.rb:384:in `level': undefined method `[]' for nil (NoMethodError)
@level_override[Fiber.current] || @level
^^^^^^^^^^^^^^^
from ~/.gem/ruby/3.3/gems/jekyll-4.3.1/lib/jekyll/log_adapter.rb:45:in `adjust_verbosity'
This is documented in the release notes for Jekyll 4.3.3.
Upgrade Blockers
So let’s stick with Ruby 3.2 for a bit. On FreeBSD, that means setting a default version for Ruby in /etc/make.conf, like so:
DEFAULT_VERSIONS+=ruby=3.2
Then installing Ruby 3.2, e.g. with pkg install ruby32. Even though the language is still there, other tools like bundler are not, so I had to build sysutils/rubygem-bundler from ports. Once that’s done, managing my Ruby 3.2 environment is workable again.
This is part of the “gently fighting”: getting the versions of the tools to match the version of the language I want. I can’t really blame the central package-repository for not supporting every use-case.
Upgrading Jekyll seems like the right thing to do, and bundle update is the tool to use.
It fails with a peculiar message that suggests that some non-FreeBSD executable is being used:
Installing sass-embedded 1.89.2 with native extensions
ELF binary type "0" not known.
Gem::Ext::BuildError: ERROR: Failed to build gem native extension.
Looking into ~/.gem/ruby/3.2/gems/sass-embedded-1.89.2/ there is an ext/sass directory, which contains an ELF executable called true and a file with ^A in its name (control-A, ASCII 0x01). I have no idea what is going on, but it feels like malware delivery while it is probably some developer delivering Linux tools – the readme mentions a Dart-language tool. I don’t feel like installing the Linuxolator to build this one gem today.
Careful With That Axe
This works remarkably well, once Ruby 3.2 and corresponding bundler are installed:
rm -rf ~/.gem
# Edit gemfile to make BUNDLED WITH match current bundler version
bundle install
I have no use for, or patience with, Ruby other than generate-my-damn-website.
This will tide me though until such a time as I update to a slightly-more-modern
static site generator, or freeze the whole thing into a virtual-machine image
that will never change again because .. it’s a static site generator.
I could replace it with vi and some patience.
You may have read about my new-found fondness for Plasma’s Clock app. Following the development of a “Picture in Picture” protocol for Wayland, I remembered how I once saw someone put up a little timer window during a lunch break while screen-sharing a presentation. I figured, I wanted that, too!
KClock showing a timer popped out in a PIP window
A key difference between X11 and Wayland is that Wayland is descriptive, not prescriptive. For example, a drop down menu under X11 is a window with no border that is placed in a very specific location determined by the application. It then grabs all the input so that up and down arrows work and clicking outside will dismiss it. That also means that global shortcuts won’t work while the menu is open. You can’t take a screenshot of it, you can’t even lock your screen. In 2025, this is embarrassing.
Under Wayland on the other hand, such a menu is an XDG Popup. The application tells the compositor what button it came from and what to do when it can’t fit (flip to the other side, scroll, etc). The compositor then gets to decide where to put the menu (not crossing display boundaries, for example) and to make sure it goes away when you click elsewhere or switch to a different application.
This also means that an application is a lot more restricted when it comes to – let’s say – creative window behavior. Take the picture-in-picture feature for example: an application can’t just decide to keep its window on top of everyone else. Normally, that’s a good thing! But it likewise prevents implementing a little overlay video player in your Web browser under Wayland. Enter xx-pip-v1.
Measure lap time while watching a fantastic talk
It’s a protocol very similar to XDG Shell but instead of managing desktop-y windows and popups, it creates floating picture-in-picture windows. KWin recently gained support for it. As indicated by the “xx” prefix it’s an experimental protocol and therefore guarded by a KWIN_WAYLAND_SUPPORT_XX_PIP_V1 environment variable. As with most protocols, having a test case is nice but a proper application making use of it will uncover gaps and oversights in the protocol much better. Therefore, I took the opportunity to allow KClock to pop out timers and the stopwatch into a little PIP window.
While it can and does take time to sketch a Wayland protocol to realze a new use case, the end result will be generally better than anything we had before. Rather than an application deciding how PIP should behave, the compositor is in full control and subsequently you, the user. We could offer an option for which corner the window should be placed by default, or to not show it at all, and do it consistently for all applications.
Many thanks to Vlad Zahorodnii for his work on the xx-pip-v1 protocol and whose demo application the KClock implementation is based on and to David Edmundson for implementing window move/resize in the PIP window.
After four months of active maintenance and many weeks triaging bugs, the digiKam team is proud to present version 8.7.0 of its open source digital photo manager.
Improvements in Photos, KRetro and better keyboard navigation
Welcome to a new issue of "This Week in KDE Apps"! Every week (or so) we cover as much as possible of what's happening in the world of KDE apps.
As you might have noticed, the frequency of "This Week in KDE Apps" has not been very consistent lately. Particularly during the summer season, I (Carl) have a lot of social obligations and can't ensure regular updates with the small amount of time I have available. If you are a KDE developer, you can help by contributing your updates to the GitLab merge request on invent. In that respect, a huge thanks to Felix Ernst for doing that already for Dolphin.
Getting back to all that's new in the KDE App scene, let's dig in!
Akseli Lahtinen increased the click area of files and folders in the main view and overhauled their styling. It's (subjectively speaking) a lot prettier now! We are trying to make Dolphin more consistent with other list styles in KDE long-term (25.08.0 - link).
Méven Car added a colorful folder icon chooser to the context menu of folders (25.08.0 - link). The Dolphin context menu can be edited in Dolphin's settings window.
Aleksandr Borodetckii lowered the default scroll speed to follow the globally-configured scroll distance more closely. Scrolling one "tick" with the mouse wheel should now move the view by a similar distance in Dolphin as it does in KWrite. In details view mode we made sure that scrolling moves by full item height (25.08.0 - link).
Gleb Kasachou added a setting to optionally elide long file names at the end instead of in the middle, so when the Dolphin window is too narrow to display “a very long file name.txt”, you can now switch between showing “a very…file name.txt” or “a very long file….txt”. Both of these behaviors have been respectfully criticized in the past, so now everyone can set it to their liking (25.08.0 - link).
Bojidar Marinov fixed a visual bug in the default icons view mode in a livestream, as sometimes the file name would have one row of text outside and below the actual file geometry (25.08.0 - link).
Akseli Lahtinen fixed a long-standing issue that could cause Dolphin to crash while interacting with Samba shares (Frameworks 6.16 - link).
Kai Uwe Broulik fixed another long-standing issue where a random "Examining" popup would appear when connecting to a busy Samba share (25.04.3 - link).
Sune Vuorela made the list of certificates used to sign a PDF update when switching the active backend (GnuPG or NSS) (25.08.0 - link). If Kleopatra is installed and the GPG backend is selected, Okular will now propose to view the certificate details in Kleopatra (25.08.0 - link).
Abdus Sami made the number of items in the recent history configurable (25.08.0 - link).
Pablo Ariño ported the Etesync resource away from directly depending on QtWidgets. This decreases the RAM consumption by around 10MiB for each running Etesync instance (25.08.0 - link). Carl then ported the resource away from KWallet to the more multiplatform alternative QtKeychain (25.08.0 - link).
Merkuro Manage your tasks, events and contacts with speed and ease
Oliver Beard fixed various bugs in Photos. Sharing now works again (25.08.0 - link) and the bookmark functionality is now consistently called "bookmark" in the UI (25.08.0 - link). He also ported the thumbnailbar to a standard toolbar (link).
Carl Schwan made it possible to configure some shortcuts in Photos and added a command bar (25.08.0 - link). Additionally, he made it possible to configure which metadata details are available in the sidebar (25.08.0 - link).
Carl also optimized the main view a bit (25.08.0 - link 1, link 2 and link 3), unified the breakpoint at which desktop and mobile mode are switched (25.08.0 - link), fixed the video player which was not completely ported away from Qt5 (25.08.0 - link), and made a large number of small cleanups and code modernizations (link).
Jean-Baptiste Mardelle added an action to extend and collapse items in the effects and folders view. This allows navigating these views with the keyboard (25.08.0 - link).
Carl Schwan added support for Deutsches Jugendherbergswerk (DJH) email confirmations (25.04.03 - link).
Stephan Olbrich added support for Deutscher Alpenverein (DAV) membership cards (25.04.03 - link).
David Pilarčík added support for extracting multiple tickets from one PDF for Leo Express (25.04.03 - link) and improved the extraction of the luma extractor (25.04.03 - link).
Joshua Goins overhauled the server information page (25.08.0 - link). You can now view your server's extended description, terms of service and privacy policy (when applicable).
Efe Çiftci added a "Show Folders" action to Filelight that allows toggling the visibility of the folder list on the left-hand side of the window (25.08.0 - link).
For a complete overview of what's going on, visit KDE's Planet, where you can find all KDE news unfiltered directly from our contributors.
Get Involved
The KDE organization has become important in the world, and your time and
contributions have helped us get there. As we grow, we're going to need
your support for KDE to become sustainable.
You can help KDE by becoming an active community member and getting involved.
Each contributor makes a huge difference in KDE — you are not a number or a cog
in a machine! You don’t have to be a programmer either. There are many things
you can do: you can help hunt and confirm bugs, even maybe solve them;
contribute designs for wallpapers, web pages, icons and app interfaces;
translate messages and menu items into your own language; promote KDE in your
local community; and a ton more things.
You can also help us by donating. Any monetary
contribution, however small, will help us cover operational costs, salaries,
travel expenses for contributors and in general just keep KDE bringing Free
Software to the world.
To get your application mentioned here, please ping us in invent or in Matrix.
This week, the merge request for the EteSync resource has been reviewed and successfully merged! This marks a key milestone in the project: EteSync is the first resource to be fully refactored, with its UI logic cleanly separated into a standalone configuration plugin.
Based on feedback from my mentors, I made a few final revisions before the merge. One important addition was a .notifyrc file for the EteSync resource. This allows the headless EteSync resource to make use of KNotification for reporting errors or status updates, integrating with the user notification system in a way that doesn’t depend on a UI.
With the resource no longer depending on QtWidgets, and its configuration interface now dynamically loaded, this change contributes to our broader goal of making the Merkuro suite lighter and more adaptable for mobile environments.
After the merge, my mentor Carl checked the memory usage using System Monitor. The improvements were measurable: the EteSync resource now uses approximately 9.5 MiB of RAM, compared to the 22 MiB used prior to the refactor.
The Job Tracker and a D-Bus Debugging
With Etesync out of the way, I turned my full attention back to the PIM Migration Agent. The next logical step was to resolve the KUiServerJobTracker problem I identified last week. My initial plan was to replace it entirely with the progress-reporting tools built directly into Akonadi’s AgentBase. But, while inspecting the Akonadi code, I realized this approach wouldn’t be viable for my use case.
Although AgentBase does provide D-Bus signals for job progress and status updates, these are low-level and do not integrate directly with the system tray or display persistent visual feedback to the user. In contrast, KUiServerV2JobTracker automatically presents progress as desktop notifications or tray-based UI elements.
Still, I implemented the changes, compiled the code, and launched akonadiconsole to test. The good news: the refactored agent started and the configuration plugin loaded. The bad news: they weren’t talking to each other. The D-Bus communication I had set up last week was failing.
This sent me down a debugging rabbit hole. My first hypothesis was that the agent needed its own .service file to properly register on the D-Bus session bus. I spent a good chunk of time creating and tweaking one, but the D-Bus connection simply refused to work as intended. The plugin and the agent remained strangers.
After much investigation I discovered the root cause, and it was a typical environment problem.
akonadiconsole was using the system D-Bus, while my development build of the agent was trying to register on the user session D-Bus managed by my Plasma desktop.
They were on two completely different communication networks! This immediately explained why my custom dbus interface had no effect—akonadiconsole was never even looking for it in the user session. To solve this I ran:
kde-builder - -run akonadictl restart
After this I ran the akonadiconsole once again and this time the D-Bus connection worked as intended.
What’s Next?
For next week, the priorities are:
Re-evaluate the job tracking solution for the Migration Agent, now with a proper understanding of the D-Bus context.
Push forward with removing the final QtWidgets dependencies.
Finalize the plugin-agent communication pathway with reliable, real-time interaction.
The refactor of EteSync and the progress made with D-Bus integration in the Migration Agent continue to move the project toward a more modular and future-proof codebase.
Every week we cover the highlights of what’s happening in the world of KDE Plasma and its associated apps like Discover, System Monitor, and more.
This week we not only continued polishing up Plasma 6.4 in response to feedback (thanks to everyone who submitted bug reports and wrote nice emails to us), but we also started work on bigger improvements to Plasma 6.5.
Probably the biggest one is the next piece of the Wayland session restore puzzle clicking into place: David Edmundson has implemented support for the xx-session-management-v1 Wayland session restore protocol in Qt 6.10! This means that software built on top of Qt 6.10 (for example, Plasma and KDE apps) will be able to start implementing the protocol themselves. Once they do, then finally real session restore will work on Wayland. We're not all the way there yet, but we keep on working to get closer all the time!
But that’s not it! We’ve got a whole lot more, too…
Notable New Features
Plasma 6.5.0
Plasma’s Welcome Center apps now teaches you about the many keyboard shortcuts in Plasma, as well as what the heck the “Meta” key is. (Nate Graham, link)
Plasma’s built-in RDP server now supports syncing clipboard text between the client and server. (Arjen Hiemstra, link)
Refined the tone mapping feature in KWin to improve the look of some screen content when using HDR or EDR. (Xaver Hugl, link)
You can no longer open an infinite number of error messages in Spectacle’s UI by repeatedly doing a thing that triggers them, and then not stopping even though maybe you should. (Noah Davis, link)
Plasma 6.4.2
Spectacle no longer includes a ghostly semi-transparent version of its own menus in screenshots taken within the app when not using any kind of delay. (Ismael Asensio, link)
The “New!” badges applied to newly-installed apps shown in the Kickoff Application Launcher are now more appealing and easier to read by using semantically correct colors from the active color scheme. (Kai Uwe Broulik, link 1 and link 2)
In the Kicker Application Menu widget, you can now activate the power/session actions using the Enter key, in addition to the Space key. Now all the UI elements on that widget can be activated with Enter. (Christoph Wolk, link)
Made the Night Light feature’s color tinting look nicer on certain hardware. (Xaver Hugl, link)
Plasma 6.5.0
The notification telling you that you missed some notifications while you were in Do Not Disturb mode now includes a button you can click on to actually see those missed notifications! (Kristen McWilliam, link)
You can now copy the QR code for clipboard items, in addition to just being able to look at them and scan them. (Ismael Asensio, link)
The “Click here to change the icon” button that’s visible on the Properties window and other QtWidgets-based apps is now more obvious as to what its purpose is. (Kai Uwe Broulik, link 1 and link 2)
The Networks widget now tells you what what it’s doing for more more potential states that it can be in, including “looking for wireless networks” and “uh, I’ve been disabled, dude.” (Nate Graham, link)
System Settings’ Printers page also now gives you a better error message if its service has been disabled. (Mike Noe, link)
On System Settings’ Region & Language page, the error message that tells you language packages couldn’t be installed now actually mentions their names so you can try to troubleshoot it a bit. (Ismael Asensio, link)
Frameworks 6.16
All scrollable views in all QtQuick-based KDE software now have inertial scrolling when scrolled using a touchpad! Note: only with a touchpad, not with a mouse wheel. No inertia there. Say it with me: no inertia for mouse wheels! (Niccolò Venerandi, link)
Improved the accessibility of the common Properties dialog in many ways. (Felix Ernst, link 1 and link 2)
Notable Bug Fixes
Plasma 6.3.6
Made the Environment Canada provider for Plasma’s Weather Report widget work again after they changed the data format. (Ismael Asensio, link)
Plasma 6.4.1
Fixed several crashes in in KDE’s desktop portal implementations. (David Redondo, link 1 and link 2)
If you have a Samsung Odyssey G5 monitor, it no longer turns on and off forever while Powerdevil’s DDC support is enabled, because the monitor’s own DDC implementation is completely broken, so we blacklisted it. (Xaver Hugl, link)
Fixed an accessibility regression that caused the Install and Remove buttons on Discover’s app pages to be excluded from the tab focus chain. (Christoph Wolk, link)
Deleting a file or app that’s been made a favorite in Kicker/Kickoff/etc no longer makes it impossible to un-favorite that thing. (Christoph Wolk, link)
Fixed an issue in the Window List widget that prevented it from being able to minimize or maximize windows. (Christoph Wolk, link)
Fixed an issue in KDE’s desktop portal-based open/save dialog that caused extra UI elements added by apps to appear in the wrong places. (David Redondo, link)
Fixed an issue with the new “relative mode” for drawing tablet styluses that could cause the pointer to disappear when you have two tablets connected at the same time and each of them uses a different input mode. (Vlad Zahorodnii, link)
Fixed two visual glitches that could be caused by re-arranging virtual desktops in the Overview effect’s grid view. (Vlad Zahorodnii, link 1 and link 2)
Fixed an issue that made clicking on Plasma UI elements unreliable when using certain non-default click modes. (Vlad Zahorodnii, link)
Plasma 6.4.2
Fixed a case where System Settings’ Flatpak App Permissions page could cause the whole app to crash (Daniel Hast, link)
Fixed an issue that could sometimes cause Plasma to go back to sleep again right after waking up, when the “Sleep then hibernate” setting is in use. (Myrrh Periwinkle, link)
The appearance of text labels in Folder View pop-ups is once again correct. (Nate Graham, link 1 and link 2)
You’re no longer erroneously prompted to authenticate for a Wireguard VPN whose credentials are already stored in KWallet, and the wallet is set up to automatically open at login. (Jeff Chien, link)
Fixed an issue in the KDE desktop portal’s screenshot implementation that prevented the delay setting from taking effect. (David Redondo, link)
Missing app backends listed in Discover’s Settings page once again show the correct names. (Harald Sitter, link)
The brightness level shown on System Settings’ Display & Monitor page now matches the one shown in Plasma. (Xaver Hugl, link)
Fixed an issue that caused the panel to have too much space in it until restarting Plasma if you stop displaying the date on a horizontally-laid-out Digital Clock widget. (Niccolò Venerandi, link)
Qt 6.8.4
Fixed a Qt bug that caused is to be extremely frustrating to re-arrange Task Manager icons because the drag target would change after you dragged an icon over another one. (Niccolò Venerandi, link)
Fixed an issue that could cause stuttering when playing video content in certain video players when using a variable-refresh-rate screen. (Błażej Szczygieł, link)
Plasma 6.4.1
Fixed a recent performance regression with certain games. (Vlad Zahorodnii, link)
Plasma 6.5.0
Increased the perceived responsiveness of logging into Plasma by optimizing the splash screen code and re-arranging some of the startup steps. (Vlad Zahorodnii, link 1 and link 2)
Slightly reduced the duration of the login animation, which makes logging into Plasma feel a bit snappier. (Vlad Zahorodnii, link)
How You Can Help
KDE has become important in the world, and your time and contributions have helped us get there. As we grow, we need your support to keep KDE sustainable.
You can help KDE by becoming an active community member and getting involved somehow. Each contributor makes a huge difference in KDE — you are not a number or a cog in a machine!
You don’t have to be a programmer, either. Many other opportunities exist:
You can also help us by making a donation! Any monetary contribution — however small — will help us cover operational costs, salaries, travel expenses for contributors, and in general just keep KDE bringing Free Software to the world.
Dear fans of music & open source music players, in preparation of the upcoming Amarok 3.3 release, a second beta release (3.2.82) has been prepared.
This time, the most important change is the new GStreamer-based audio backend. This enables a number of features that weren't available on Qt6 Phonon
backends, and likely also provides a more reliable audio experience in general.
In addition to audio system work, e.g. more safeguards have been set up around collection scanning code to prevent some potential database issues.
More details on changes are listed in the ChangeLog.
These come in addition to the previous beta 1 changes (Qt6 only, database update).
Please note that due to the database update in beta 1, downgrading from 3.3 betas is not directly possible, and returning to pre-3.3 versions
requires the database (at ~/.local/share/amarok/mysqle/) to be manually backed up beforehand.
The Amarok 3.3 beta 2 source tarball is available on
download.kde.org
and it has been signed with Tuomas Nurmi's GPG key.
In addition to the source code, it is likely that some distributions will provide beta packages.
The various nightly git builds provided by various splendid packagers should also provide a way of using the beta changes
and participating in the testing.
Plasma's upcoming first-run experience is coming along nicely.
After a bunch of research and discussions we settled on continuing / fixing up
the KISS project (KDE Initial System
Setup) for the new First Run Experience (FRE) / Out-Of-Box Experience (OOBE). It
was in a sort of half finished state.
Since then has been a bunch of work on it such as:
Getting it to compile and run
Whole bunch of fixes, cleanup, and polish to the UI, UX, and code/developer
experience
Implemented the ability to actually create the new user (you could enter a
name and password, but it was all basically placeholder GUI previously)
Added language selection front/back ends
Added basic build / run instructions to the README
Added ECM logging
Cleaned up the debug output which made changes more difficult
Added basic CI (thanks Nico!)
Added keyboard layout selection front/back ends
That last one was more difficult than expected.. turns out keyboard layouts can
be quite complex!
First came some refactoring of the keyboard layouts KCM from plasma-desktop so
we could reuse some of the existing, complex functionality. Then adapting a
UI/UX appropriate for the FRE. Investigating things like keyboard models,
detecting defaults, mapping language to keyboard layout, etc..
Then taking the results of the user choice and figuring out how to apply that
both to the system as a default (systemd-localed dbus call) as well as to the
running Plasma session (setting the value manually in the kxkbrc keyboard
settings file).
As is often the case with software development, that succinct summary belies
the massive amount of work it took to get there! 😅 💪
With that work completed, we have most of what is needed for a minimum viable
product!
Next up:
Granting authentication without a user prompt
Plan: special user with sysusers.d and a polkit rule
Running automatically on first boot & in live sessions
Plan: systemd unit seems promising, but more research is needed
Improve documentation
Especially related to building & running without kde-builder
Think about the name
I am not thrilled with the KISS acronym personally 🤷♀️
There is also obviously a lot of improvements and polish that can be made, but
for now here is a preview of the FRE:
These past few week’s my focus was on exploring input device detection and event handling mechanisms in Linux, with a particular emphasis on game controllers and their potential integration into KWin.
I also spent time reading through KWin’s input-related source code to understand how it currently manages devices, and began reviewing documentation for various Linux input subsystems—including evdev, HID, and /dev/input/jsX in order to evaluate which layer would provide the most reliable and straight forward support for integrating controller recognition.
The time was mostly spent learning how to use different libraries, tools and creating virtual controller prototype.
Tools, Libraries, and Concepts Used
libevdev
libevdev is a library for handling evdev devices.
It provides a higher-level interface over /dev/input/event* and abstracts much of the complexity of input event parsing.
evdev is the generic input event interface. This is the preferred interface for userspace to consume user input, and all clients are encouraged to use it.
-The kernel development community.
libevdev can be used to:
Detect physical game controllers.
Read input events (e.g., joystick, buttons).
Create virtual input device and write/forward events to it from physical game controller.
Useful functions:
libevdev_new(), libevdev_set_fd(int fd, struct libevdev **dev): for opening physical devices.
libevdev_next_event(struct libevdev *dev, unsigned int flags, struct input_event *ev): for polling events.
libevdev_get_id_*(const struct libevdev *dev): to query device meta data.
uinput (User Input Subsystem)
I used the Linux uinput subsystem to create a virtual input device that mirrors a physical controller input.
uinput is what allows us to make a virtual controller out of any evdev device by:
Opening a file discriptor for the input device that will be emulate (i.e. have it input event forwarded).
Forwarding the inputs from a evdev interface device to /dev/uinput (or /dev/input/uinput).
uinput then creates a new node to expose the virtual device as a evdev interface device in /dev/input/event*
From here the idea is that KWin or any other system component can treat the virtual controller as if it were an ordinary HID device.
uinput is a kernel module that makes it possible to emulate input devices from userspace.
By writing to /dev/uinput (or /dev/input/uinput) device, a process can create a virtual input device with specific capabilities.
Once this virtual device is created, the process can send events through it, that will be delivered to userspace and in-kernel consumers.
-The kernel development community.
Useful functions:
libevdev_uinput_create_from_device(const struct libevdev *dev, int uinput_fd, struct libevdev_uinput **uinput_dev): For creating a uinput device based on the given libevdev device.
libevdev_uinput_get_devnode (struct libevdev_uinput *uinput_dev): Return the device node representing this uinput device.
libevdev_uinput_write_event (const struct libevdev_uinput *uinput_dev, unsigned int type, unsigned int code, int value):
Post an event through the uinput device.
Tools used:
libevdev-uinput.h for management of uinput devices via libevdev.
/dev/uinput opened with correct permissions.
Ensuring the current user is in the input group.
Verifying that the uinput kernel module is loaded (using modprobe uinput). Some distros (Ubuntu/Kubuntu) have it built in, not loaded as module, thus modprobe uinput command won't log anything.
Opening /dev/uinput with O_WRONLY | O_NONBLOCK flags using open(), and ensuring no EPERM or EACCES errors were returned.
Optional: Run program as sudo user.
force feedback detection/support
Using ioctl(fd, EVIOCGBIT(EV_FF, ...)) and tools like fftest, I examined:
How to query a device’s force feedback (FF) capabilities to figure out which effects are supported (e.g., rumble, sine wave).
How to upload ff effects to physical game controller and test rumble motors.
This was key to understanding haptic capability support on physical devices.
To enable force feedback, you have to:
have your kernel configured with evdev and a driver that supports your device.
make sure evdev module is loaded and /dev/input/event* device files are created.
Testing & Validation
Used evtest and fftestto test evdev devices and understand their capabilities -
sudo evtest /dev/input/eventX.
Used those same tools to test virtual devices creating using uinput - sudo fftest dev/input/eventX. uinput creates a node device in dev/input/eventX for the virtual input.
Prototype logs validate that a virtual device can be created and events can properly be written to a that virtual device using libevdev.
Takeaways
Using libevdev and libevdev-uinput we can access physical controllers, create virtual controller and read/write low-level input events.
Understanding of the permission requirements to open /dev/input/* and /dev/uinput (use udev rules or run as root).
Tools to test:
evtest and fftest (from input-utils)
udevadm info --name=/dev/input/eventX --attribute-walk
Shows the device hierarchy - how the device is connected to PC and any parent device it connects to.
Built a minimal proof-of-concept C++ program that routes an evdev devices input 1:1 to a virtual controller (via uinput).
Not all controllers support all force feedback types; some failed with EINVAL during upload.
libevdev does not handle FF upload directly — this remains kernel-level and typically involves ioctl().
[ade]
@kbroulik:kde.org
Dear digiKam fans and users,
