The other day I had to dive into the ksmserver code, the core of session management in KDE Plasma, but found it all a bit difficult to read. If only I could get a callgraph I thought…

Turns out it actually is possible!

In fact it is almost too easy with clang/llvm. There are already solutions that attach LLVM bitcode to existing binaries. e.g. gllvm. gllvm basically is a shim sitting between ninja and the actual compiler and injects some behavior for us. This enables the use without having to refit anything in our existing code. You might note that clazy, our most excellent code analyzer, works in a similar fashion.

Here is how to use it:

# install gllvm (NOTE: as a pre-requisite you should have Go properly set up)
go install -v github.com/SRI-CSL/gllvm/cmd/...@latest

# clone plasma-workspace
git clone https://invent.kde.org/plasma/plasma-workspace
# change directory
cd plasma-workspace
# export environment variables
export CXX=gclang++
export CC=gclang
# You may need to set some additional vars so gllvm can find your compilers
## export LLVM_CC_NAME=clang-16
## export LLVM_CXX_NAME=clazy
# Configure
cmake -S . -B build
# Build
cmake --build build/
# extract the bitcode (into build/bin/ksmserver.bc by default)
get-bc -m -v build/bin/ksmserver
# Run through llvm's opt program to generate a callgraph
opt --disable-output --passes=dot-callgraph build/bin/ksmserver.bc
# copy into working directory
cp build/bin/ksmserver.bc.callgraph.dot callgraph.dot
# prettify the c++ names (this may produce invalid labels, so I'm skipping it here)
## cat build/bin/ksmserver.bc.callgraph.dot | llvm-cxxfilt > callgraph.dot
# generate an SVG of the graph
dot -x -Tsvg -ocallgraph.svg callgraph.dot
# we now have a callgraph.svg \o/

Depending on the complexity of the software and how many libraries it uses this graph may be incredibly large and verbose though. We’ll need to apply some filtering to make it useful. Or at least I thought so when working with ksmserver.

To further filter the graph you can use the gvpr helper. It is a bit clunky but gets the job done. You can build a filter expression to only include interesting functions like so

N{
    if ($.label!="@(*_ZN9KSMServer*)" || $.label=="@(*metacall*)")
        delete($G, $);
}

and then filter the dot graph with

# Filter out uninteresting functions (filter.prog is the filter expression from above)
gvpr -c -f filter.prog build/bin/ksmserver.bc.callgraph.dot > intermediate.gv
# Filter out all empty nodes (they call nothing and are of no interest)
gvpr -c "N[$.degree==0]{delete(root, $)}" intermediate.gv > final.gv
# generate an SVG
cat final.gv | dot -x -Tsvg -ocallgraph.svg

Final result

For a project of mine I need gamepad support. In the past, I’ve happily used QtGamepad, but that has not been ported to Qt 6. It’s not dead, but Andy (QtGamepad’s maintainer) wants to do some re-architecting for a Qt 6 release.

I need QtGamepad now, however, so I’ve ported it myself. It’s not a whole lot of code and Qt’s pro2cmake.py made it a breeze. I’ve renamed the whole thing to QtGamepadLegacy and pushed it to GitHub. So whenever the official QtGamepad is released there should be no naming conflicts. I’ve tested with Qt 6.6.0 and the evdev plugin.

I don’t plan on adding any new features to the port. I’ll try to keep it compatible with upcoming Qt releases, though.

Learn how to use Rive within Qt and Qt Quick.

Rive is a tool (and file format) that enables you to create interactive vector animations. With the RiveQtQuickPlugin, you can effortlessly load and display Rive animations within your QtQuick projects.

In this article, we will demonstrate how to embed Rive files, use different rendering backends, load artboards and trigger animations.

Continue reading Embed Rive in your QtQuick applications at basysKom GmbH.

Long Covid. If you are at all conversant with C20 twenties and thirties detective novels, you’ll know all about the invalids, the incapabables, the chronically ill. That was the flu. Though it never is the protagonist who is suffereing, it’s always someone else.

Well, this time it’s me. The protagonist of my own life.

I got covid when I got jabbed, early 2022. In the vaccination centre. Not because of the vaccine, but because, for some reason, everyone was expected to sit down, in a perspex cubicle for fifteen minutes, unmasked.

Of course, as you know, the covid virus remains infectuous for six hours and airborne, no droplets needed, in any area an infected person has been breathing for about ten minutes.

And no vaccine is immediately effective.

These were incubation incubators.

So I got sick. Really sick. For two weeks I had a horrible fever and was completely bed-ridden.

The aftermath is worse though.

I noticed my cognition was going down: I was (and am) getting dumber. Concentration powers going down. Getting pukey, and after that, shitty, regularly. Heart rates getting up into realms no blood pressure measurement machine could manage to measure. Having to lie down for days.

People wished me well. Wished me to get better soon.

Sorry people, that doesn’t work: this is like M.E., and you don’t get better from that. (In fact, Long Covid might, after all, have provided a pathology that explains M.E.)

Better soon? Bad days are getting increasingly frequent, and while I’l very good on good days, I cannot predict when good days will happen.

There you have it… Puke. Runny shit. Lying down all day in a darkened room. Brain Fog. Every little thing you manage to do is an achievement. Heart palpiations. Doctors telling you you’re malingering. Good days when you’re doubting you are actually sick getting outnumbered by days you cannot even eat, only drink water.

(And don’t ask me about how it’s when your nose if full of the smell of bin liner filled with the intestines of two-weeks dead animals.)

.

..

…

Hey… Masks are the one thing that work.

Happy October and other related holidays! The list of work is a bit smaller, as predicted I was busy this month. However, I have some really exciting work done in PlasmaTube - some of which you may have already seen if you follow me on Mastodon.

I also added a better way to discern bugfixes from features in these posts! It also details which branches and versions the change will appear in. I will be adding this slowly to older posts. I’m also really close to completing a whole year of this blog series, which I started back in January 2023! 🎉 I’ll be detailing more of 2024 plans in December.

Tokodon

[Feature] I added support for viewing server announcements! It doesn’t support emoji reactions yet, because I want to also implement emoji reactions in general (if your server supports it.) [24.02]

[Bugfix] Fixed more cases where media attachments wouldn’t load, if some of the metadata was invalid. [24.02] [23.08]

[Feature] The window geometry and position (if on X11) is now saved and restored. [24.02]

[Feature] A profile’s featured tags are now shown in the filter bar. You can select between them or reset it by tapping on “All”. [24.02]

[Feature] The selected post in a thread is now scrolled to initially, so you don’t have to scroll so much when viewing replies. [24.02]

[Feature] When scrolling to the very end of a timeline, a message is now shown that it’s really the end and you didn’t encounter some kind of bug. [24.02]

[Feature] Tokodon now supports basic notification actions such as viewing the post or the user’s profile. More actions are planned in the future, but this is a good start. [24.02]

[Feature] I cleaned up and merged the offline testing system that was originally pioneered by Rishi Kumar. This doesn’t affect users, but means that Tokodon can be tested more easily. Aka, more stable software! :-) [24.02]

[Bugfix] Add missing PulseAudio, FreeDesktop Secrets and notification permissions for the Tokodon Flatpak. [24.02] [23.08]

NeoChat

[Feature] NeoChat state (such as event cache) will now be stored in it’s own state file, called “neochatstaterc” in ~/.local/share/KDE/neochat. This matches the behavior of other KDE applications, instead of the weirdly named “data” file like before. [24.02]

[Feature] The event cache is now compacted so it’s a little bit smaller, and easier to read. [24.02]

[Feature] The recent emojis are now saved in the state file, instead of the regular config file. Less state in there, woohoo! [24.02]

Kirigami

[Feature] You can now control the isMask property of a Chip. Otherwise, this property is inaccessible, preventing you from changing the icon color. [6.0]

PlasmaTube

I did a ton of work improving PlasmaTube this month. If you haven’t been using PlasmaTube yet, or haven’t tried it in a while… 24.02 in February will be packed with new features and bugfixes!

[Feature] PlasmaTube is now Qt6, complete with the new declarative type registration system. For users, it means that it’s marginally faster and that it’s ready for the February mega-release. [24.02]

[Feature] Different video sources other than Invidious is now supported! The current plan plan is to support Invidious, Piped and PeerTube by 24.02. The feature set of the new video sources are currently limited compared to Invidious, but will be expanded. [24.02]

[Feature] Selecting between video sources is now supported, for example if you have a Invidious instance but also want to watch PeerTube content. [24.02]

[Feature] Along with that, the overall login flow is overhauled to match other Kirigami applications. You can even select from a list of public instances. (Keep in mind that the UI is still a work in progress and will be improved before release.) [24.02]

[Feature] The sidebar is no longer icon only, and matches the UX of other Kirigami applications such as NeoChat, Tokodon and Arianna. There are now several tabs such as “Popular”, “Trending”, “Subscriptions”, “Playlists” and “History”. Some of these are disabled depending on if you’re logged in and if the video source supports them. [24.02]

[Feature] The “Trending” page has been overhauled to use tabs instead of stuffing all of the different types into the page’s actions. [24.02]

[Feature] Added a “Playlists” page! Right now you can only view playlists, but I plan on adding more advanced playlist management soon. [24.02]

[Feature] Added a “History” tab, which of shows you recently watched videos. [24.02]

[Feature] Added a new context menu, shown while long tapping or right-clicking on a video item. You can mark a video as watched or unwatched, add it to a playlist and more. [24.02]

[Feature] Added a picture in picture mode! It’s even possible to control the video using the normal controls in the main window. When closing the PiP window, it will also restore the position in the main player too. [24.02]

[Feature] It’s now possible to use a network proxy in PlasmaTube, but due to FFmpeg limitations you can only set a HTTP proxy unlike other KDE applications which support a SOCKS5 proxy. [24.02]

[Feature] You can now change the default homepage for Invidious sources, which also syncs to the Invidious web frontend and any other clients that support this key. More settings will be available in the future, such as autoplay (which currently does nothing.) [24.02]

[Feature] You can now view comments on a video, which is supported in Invidious, PeerTube and Piped sources. They aren’t very detailed yet, and threading is not yet supported. [24.02]

[Feature] A channel’s avatar, banner and description is now shown on the channel page. Eventually other channel tabs like playlists will be available here too! [24.02]

Merkuro

[Bugfix] Fixed a bug where certain locales would be unable to use the date picker. [24.02] [23.08]

KWallet

[Bugfix] Fixed the “Configure KWallet…” menu item not working. [23.08]

[Feature] Added a visible error message for conditions where KWalletManager could fail and display nothing, like the DBus service failing to activate. [24.02]

KWin

[Bugfix] Fixed the default keybinds for “Switch to Next/Previous Desktop” not being set properly by default, when they probably should. [6.0] [5.27]

Plasma

[Feature] A bunch of small improvements and fixes for the “Show Activity Manager” applet. This is mostly fixing stuff that was broken due to KF6 changes. [6.0]

Kirigami Addons

[Bugfix] Prevent a crash when calculating initials in names that only have parenthesis, such as “(gobble)”. This affects the new Avatar component. [0.12]

See you next month!

During the last years, I’ve been doing a lot of coding and no blogging. That’s a shame because without good announcements, there is no chance for people to know about the code.

All of the projects I’ve been coding on are more or less unfinished which is why I’ve not been too keen on blogging about them. It’s time to get over this inhibition and show the world these rough gems in all their glory.

This blog is just the announcement of more blogs that are about to come. For now, the projects that I’ll be be blogging about are just listed here with a short description and a link to the repository.

I will be fighting the urge to polish the projects more and delay announcing them even further. The plan is to blog about one project per week.

Kwats

RDF graphs contain triples. Prepending the triples from the graph with the IRI of the triples gives quads. The Dutch word kwats sounds like quads but means nonsense.

Kwats is a simple server where triples can be stored and queried. Triples are added in the form of graphs. The graphs are stored as files in git, along with provenance information.

A running server keeps the graphs in memory. Rules about what is allowed in the graphs is written in the form of SHACL.

Subtree of Life

Subtree of Life is a webpage to create phylogenetic trees that show the relationship between the species of your choice. How are cat, dog and bumblebee related? Like this.

The code for this webpage is finished. The user experience, documentation and packaging can be improved.

Xust

Xust was a big COVID project. It is meant to become an XQuery implementation like BaseX and Xidel.

Xust main has 1218 commits and the XQuery functionality is not usable yet. The repository contains Rust crates for an XML DOM, DTD validation, and XML Schema validation.

This is a fun repository to work on because of the large number of tests. The CI shows 75551 tests at the moment that finish in seconds.

Rehorse

Rehorse is a web application for rehearsing music. It started off as way to loop recordings that matches the sheet music. So you can loop easily between, say, measure 10 and 12.

The application has been in production for years but is not very polished. I maintain an instance of it for members of two bands.

Nix flake for KDE software

kde-nix-flake packages KDE software for Nix. Nix is a package manager available for many Linux distributions. It handles packaging, setting up a development environment, testing and deploying.

The file flake.nix contains instructions for running stable and nightly KDE software, setting up a build environment with nix develop, running tests with nix check, using Plasma as a desktop, using the binary cache for quick installation, and running jobs on a CI server.

Keeping this repository up to date with the development speed of KDE has turned out to be a challenge.

Diary

The simply named diary is another unfinished application that has been in production for years. It is a desktop application based on rust-qt-binding-generator that I use to write notes.

It was also meant to keep a calendar and a list of todo’s, but those parts are not finished.

Reworking Recent Files search for Plasma 6

For Plasma 6, lots of KRunner plugins and framework functionalities were improved. I took quite a bit of time to work on the recent files search plugin used by KRunner and Plasmas application launchers like Kickoff. This included performance improvements, usability improvements and technical refactorings.

Let’s start off with the usability improvements: While the runner was named “Recent Files”, it still provided results for directories. In Dolphin on the other hand, “Recent Locations” are a separate location you may access next to “Recent Files”. Luckily, functionality for only querying files already exists in the KActivities-Stats framework :).
The search for files was also improved. Natalie Clarius added some time ago a patch to avoid false positives when the query is part of the file path. For example, when you type for “myfile” and “/home/user/myfiles/test” was a recent file, this file would be returned by the KActivities-Stats framework. With Natalie’s logic, those files would not be shown in KRunner, but still they fill out the limit of recent files we want to query. Of course, one could look for more files in case we discarded too many, but that can cause a performance penalty due to additional SQL queries.
Instead, I have added the ability to filter by the filename in KActivities-Stats. This means false positives are always avoided, and you will for sure get the results you are looking for!

Another useful, but not yet user-facing change is special handling for queries shorter than 3 characters. For those queries, a substring check would cause too many unintended results to be useful. Instead, the filename must start with your given query. Similar to how the Applications-runner handles it. There are plans to make this feature more easily accessible, but for now, it can only be done on the command line using krunner --runner=krunner_recentdocuments.

Screenshot of KRunner in single-plugin mode with the recent files plugin

The final issue that everyone will benefit from being fixed is a memory leak. This means for each letter typed, the runner would allocate a bit more RAM for results from the query. This can add up over time in KRunner and Plasmashell. The KActivities-Stats that were “leaking” also caused CPU overhead, because they were notified about changes to the recent files list – even though they were never used again.
Luckily, you don’t have to worry about that at all from now on :).

Optimizations

Let us start with the small improvements: In my previous posts, I mentioned that constructing a QueryMatch object before being sure that the current item (application or systemsettings module) matches, causes a performance overhead. The same applied to the recent files plugin, because we discarded some files due to the false positive detection mentioned above.
Getting the filename was also optimized/simplified, because instead of using Qt API to get the filename from the URL, we can just use the resource title from KActivities-Stats.
String comparisons were speed up by reusing the results. For example, when checking if the filename contains the query, we can get the resulting index. If that index is 0, the query is contained and the filename also starts with it. Meaning we’d only do one string comparison for each recent file instead of up to 4.

Getting the appropriate icon is also way faster, because previously a method from KIO was used, which gets the file path, checks some extra cases, but usually gets the icon for the determined mime type. Because the extra cases were not relevant when only allowing files, we can directly get the icon for the respective mimetype. Luckily, we don’t even need to determine the mimetype, because the model already contains it.

Reusing previously fetched data

This is the most interesting part and was my original idea to improve performance: In KDE Frameworks 6, each runner lives in its own thread. Meaning, one may access member variables in a thread safe way. The data we want to reuse is the KActivities-Stats ResultsModel. This contains paths and additional metadata of recently used files. The number of files is limited to 20, which is the maximum number of entries is KRunner.
Implementing was straightforward: If one previously typed “firef” and after that “firefox”, the previous model can be reused as long as the limit of 20 did not exceed the previously found results. In addition to reusing the data, it is required to check the filenames again, because a file called “firef.test” should not match a query called “firefox”. Due to the preexisting logic to determine the match relevance, this is minimal additional work.
When measuring the real-world impact, it is important to note that fetching results from KActivities SQLite database is the most expensive part. Accessing the model is relatively cheap. Meaning the longer your queries are and if long as the number of results was not exceeded, your performance gains will be significant!

Profiling of the unoptimized version, loading the correct Icon takes a large porting of CPU cycles Optimized version, CPU cycles are significantly reduced and loading of icons is barely distinguishable from other, minor costs

In benchmarks, one is able to see that the CPU cycles with this patch are roundabout the same among the different measurements, whereas the CPU cycles before this patch correlated with the length of the query. The query was typed in letter by letter, keep in mind that the runner skips queries shorter than 3 characters unless the single runner mode is activated. For benchmarks, I decided against activating this mode, because it is not the normal usecase.

This past weekend I attended the Google Summer of Code Mentor Summit 2023 as part of the KDE delegation.

I have been a mentor for GSOC almost every year since 2005 but this was my first time attending the mentor summit.

There were sessions about the typical things you'd expect: how to get more diverse folks as students, how to make sure we onboard them correctly, sustainability, funding, etc. All in all nothing groundbreaking and sadly no genius solution for the issues we face was given, but to a certain degree it helps to see that most of us have similar problems and it's not that we're doing things particularly wrong, it's just that running a Free Software project is though.

Carl Schwan and me ran a Desktop Linux session together with Jonathan Blandford of GNOME (check his Crosswords game, seems pretty nice) and basically asked folks "How happy are you with the Desktop Linux", you can find the notes about it at https://collaborate.kde.org/s/HKn6BoNCRNezn3K Nothing we don't know about really, Wayland and flatpak/snap are still a bit painful for some folks even if there's a general agreement they are good ideas.

I also organized a little session for all the attendees from Barcelona (it was about 6 of us or so) to sell them talk about Barcelona Free Software

One thing that always pops up in your mind when going to events is "How useful was it for me to attend this" since traveling to California from Europe is not easy, it is not cheap and it means investing quite some time (which in my case included taking vacation from work). 

Honestly, I think it's quite useful and we should attend more similar events. We get to know key people from other projects and we make sure other projects know about us. One of the most funny interactions was me sitting in a table, someone joining and saying "Woah KDE, you guys are super famous, love your work" and literally seconds after another person joining us and saying "Uh, KDE what is that?"

There's not much pictures because Google forbids taking pictures inside their buildings, the few exceptions include the chocolate table, it's quite a large quantity of chocolate we got to try, thanks Robert from Musicbrainz for pushing people to bring it :)

I'd like to thank Google and KDE e.V. for sponsoring my trip to the Summit, please donate at https://kde.org/fundraisers/plasma6member/

The next release for the big three in KDE land (KDE Frameworks, KDE Plasma and KDE Gear) is going to happen at the same time.

This is because we are switching to Qt6[*] and it helps if we can release all the products  at the same time.

If you want to help us with the effort, make sure to donate at https://kde.org/fundraisers/plasma6member/

The agreed schedule is:

8 November 2023: Alpha

KDE Gear 24.01.75 / KDE Plasma 5.80.0 / KDE Frameworks 5.245.0

29 November 2023: Beta 1

KDE Gear 24.01.80 / KDE Plasma 5.90.0 / KDE Frameworks 5.246.0

20 December 2023: Beta 2

KDE Gear 24.01.85 / KDE Plasma 5.91.0 / KDE Frameworks 5.247.0

10 January 2024: Release Candidate 1

KDE Gear 24.01.90 / KDE Plasma 5.92.0 / KDE Frameworks 5.248.0

For KDE Gear that want to ship with Qt6 for this release they need to be switched to Qt6 (and obviously stable) *BEFORE* this date.

31 January 2024: Release Candidate 2

KDE Gear 24.01.95 / KDE Plasma 5.93.0 / KDE Frameworks 5.249.0

21 February 2024: Private Tarball Release

KDE Gear 24.02.0 / KDE Plasma 6.0 / KDE Frameworks 6.0

28 February 2024: Public Release

KDE Gear 24.02.0 / KDE Plasma 6.0 / KDE Frameworks 6.0 

You can see that Alpha is less than 3 weeks away! Interesting times ahead!

[*]  some KDE Gear apps may remain in Qt5 if we have not had time to port them