This week in Usability & Productivity, part 40

Welcome to an even more humongous week in KDE’s Usability & Productivity initiative!

I’d like to specially highlight one very important fix this week: external hard drives are now safely powered off when unmounted. The fix is in KDE Frameworks 5.52, which will be released in approximately three weeks, and I’d like to give a big thanks to Stefan Brüns who fixed it!

Speaking of Stefan, he and Igor Poboiko have been doing an absolutely smashing job fixing Baloo over the past two weeks. A lot of their work is hard to blog about because it’s not immediately user-facing (though I’ve included as much as possible below), but between the two of them, they’ve made an enormous number of improvements to Baloo that should make it work faster and more smoothly in a lot of subtle ways.

But obviously that’s not all; take a look at the rest of the week’s work:

New Features


UI Polish & Improvement

Also, I want to mention that we’re aware of two high-profile issues in Discover that slipped by testing and made it into the 5.14.release:

We apologize for these bugs and we’re working to get them fixed quickly!

One of the reasons why bugs like this squeak through is that we don’t have enough pre-release QA testers. You could be one of them, and next week, your name could be in this list! Just check out, and find out how you can help be a part of something that really matters. You don’t have to already be a programmer. I wasn’t when I got started. Try it, you’ll like it! We don’t bite!

If my efforts to perform, guide, and document this work seem useful and you’d like to see more of them, then consider becoming a patron on Patreon, LiberaPay, or PayPal. Also consider making a donation to the KDE e.V. foundation.

63 thoughts on “This week in Usability & Productivity, part 40

  1. I’m surprised at how many changes you can make in a week. Great guys!

    I don’t know if you know the CKB-Next application. It allows you to adjust the keys and rgb colors of the mouses and keyboards. Do you think that something like this would be possible to integrate in the future directly into KDE?

    Liked by 3 people

    1. I haven’t tried that, but I don’t think so. The app only works with Corsair devices, configuring rgb colors and extra keys for video games, although Nate, if there was a way to integrate your idea for gaming laptops it could all come together.

      It would be fantastic to unify all of this into a single component with different brands apart from Corsair, such as Roccat, Razer, …

      Do you know if there’s anything like that already developed?


  2. Nice! All issues of Plasma 5.14 that I noticed are already here. So far this was the smoothest release anyway. 5.12 and 5.13 were incredibly buggy at a start. This time nothing major or nothing breaking was introduced. Discover crashing a lot – true but on Arch side, we don’t use it much so that’s not a problem. Lack of buttons on control media widget – I thought that was something that crashed on my side, I noticed it just yesterday so it hasn’t been long enough to bug me. Now when I know that fix is on its way, I just wait patiently. The rest I didn’t even notice.

    Great to hear about baloo. It can be a pain for some people sometimes.

    I hear that snaps should work out of the box outside Ubuntu distros in new Discover but I don’t see it (only system repo plus flatpak repo). There are no snap sources on the list. Or maybe that meant only that Discover can handle snap installs via web links?

    Again, each app should have very clearly visible sign about where it comes from (repo, snap, flatpack), not just after seeing details, which helps to lift confusion when seeing the same app on the list. None of the software centers have that which is surprising to me, because it should be the first thing when implementing support for different package formats. Also, when that will be implemented, those should be grouped. Currently, for example, I have repo Gimp on top of the list and a dozen other apps below so I have to scroll down and then flatpak Gimp. Grouping wouldn’t be nice if there was no indication about the source thou.

    I’m excited to try out new wallpaper formats in 5.15. I was aware they are available but because it wasn’t so effortless to get them, I never did. Now it will be just one click away. Awesome!

    Liked by 2 people

    1. Grouping the same Applications should already work. The problem is that Discover assumes that the same app has the same appstream id. Some apps, like Telegram, have different IDs for the “normal” and Flatpaked version, breaking Discovers grouping

      Liked by 1 person

    2. I hear that snaps should work out of the box outside Ubuntu distros in new Discover

      That’s up to Distros. Many of them don’t package Discover with Snap support built-in, which is honestly not unreasonable since it was very unstable until recently. Hopefully we can get some of them to reconsider as it continues to improve. Snap has no sources or repos to configure; there’s just the default one. Once the Snap backend is installed, Snaps should show up in browse and search lists in Discover.

      Apps are already grouped where possible, but it relies on them having AppStream IDs that match. See and It’s something app developers and distros have to help out with; we can’t do it all in Discover

      Liked by 1 person

    3. Is there a way to add snap backend manually? I search through the web and there is no info on it, no simple “add a package” or “edit this conf” solution, at least none that I could find.

      I’m not a fan of flatpaks and snaps and they are not that useful on Arch-based systems, but sometimes they can help with some problematic AUR packages. From my experience, snaps are behaving a lot better then flatpaks which are really, really bad (takes a humongous amount of space, no system integration, bad security, etc.), while snaps still can maintain some integration (there are global menus, at least on packages I tried). So at the moment, on Plasma snaps are much better.


  3. What about USB sticks? Is it possible that KDe PLASMA is not able to turn off the power on this kind of devices?! Other linux operating systems are able. It’s necessary to divide unmount action from turning off action to this devices, besides, the developers have to divide the unmount function from eject function too, so to make the CD-Rom/DVD eject function reliable. “Just after I ejects the disk it is again immediately inserted in.”


    1. I red a post in a a forum that explains the possibility to turn off usb by a simple command it is named: udisksctl power-off -b /dev/sd(x) . It was possible to make thi command as a simple option into the notification area so to power off the usb devices. Is it not so simple?


  4. I have a couple of bugs I’d love to see fixed, but I’m still not sure if I should report them on bugzilla (and if that’s the case, if I should tag them with “usability and productivity” or something) or just dscribe them right here


  5. Just out of curiosity, how come this bug fix didn’t make it into 5.14.1?
    There are bugs that had commits to fix them after this one which made it in, so I wonder how it was decided. I’m guessing due to the forementioned lack of QA testers, or maybe the severity of the bug wasn’t high enough.


  6. That might sound stupid but reading your post is the first thing I do on every Sunday… 🙂
    This week is massive and fixes tons of problems I’ve had in the past years.

    Liked by 1 person

  7. Considering how bad (both in functionality and usability) in general Discover has been since its inception, don’t you guys think it would be nice to stop forcing it down your users’ throats? I mean Muon at least worked and didn’t crash two of three times you used it.

    I personally purged Discover after the first release and have always been using the terminal, but don’t you think it just makes the project look bad forcing users to use this terrible “app” to handle their system updates?


    1. In what way do you feel that it has been forced down users’ throats?

      I don’t disagree with you that it still needs work, but let’s try to be realistic and use constructive language please. It’s hard to have a conversation with emotions running hot.

      Liked by 1 person

    2. All the mainstream distros that ship KDE (Kubuntu, OpenSuse, Kaos, Neon, Fedora, etc…) are now using Discover as the default software center. That’s how I feel it.

      There was a push from the KDE camp to ditch Muon (which had it’s problems, like bad UI and poor performance, but worked) in favor of Discover when it was at best alpha-quality software. And all the distros drank the cool-aid for some reason. And this badly damaged KDE’s image for casual and/or prospective users.

      I don’t see where I’m not being realistic. You agree it still needs work yet it’s the default software center in all the KDE desktops. So you are forcing your less techy users to use beta software for something as important as system applications’ management.

      And sorry if you don’t find “forced down users’ throats” constructive, I apologize. It’s just an expression, meaning you pushed for it to be shipped by distros. And as I said, I purged it after the first use, I like the terminal better for updates, so no emotions running hot here.


    3. Manjaro KDE is not shipping with Discover, just FYI.

      Most distros and desktops come with some app centers which are equally bad and criticized. Most users and up using synaptic on Ubuntu based distros. For a newbie, a software center feels like a right thing but it doesn’t find all packages, some functions are missing and for system updates it just sucks. Discover is not an exception here, although… it has more features than usual software center now and current development goes beyond usual software centers so Discover can become truly unique and positively standing out. However, the UI is widely disliked and it’s too “crashy”. Even when it becomes relatively stable, next release, new options and it’s as if regressed to beta. I think that Discover must get a certain level of maturity where it has all the features, UI is good and developers can focus on polishing it. In fact, they are doing it all the time but… it’s still in heavy development so things go sideways all the time.

      So basically you could also complain that Gnome or other DEs are forcing their software centers although you end up using a different program or a terminal anyway. That’s not a KDE “pushing it forcefully”, it’s a decision of distro developers. Since most popular and big distros want to have at least some basic level of friendliness, they want to have software center – for newbies. Since there is no alternative on KDE side, they choose Discover.

      So it is what it is but in overall, Discover has great potential. It’s already interesting now. It can update manually downloaded themes, widgets and various add-ons (which package managers won’t do) and if you have compatible hardware, you can do firmware updates. There are other cool features. Granted, it all is spoiled when you see the UI and when it crashes but those things can improve with time.

      Compared to other software centers Discover is much younger and more feature rich, so the amount of work it needs is huge.

      I’m sure your distro of choice comes with programs that you don’t need and don’t use. You just uninstall them, right? Nothing is stopping you to have a point no.1 “get rid of Discover”. Plasma works well without it.

      Liked by 1 person

  8. Hi Nate!

    Is it a bug that I can’t order elements of the folder view widget having them unsorted? Moving an element in the free space after the last element in a row lets it move back. It looks like the elements can only be displayed sorted by name, date etc. – not manually.

    Many thanks for all your work!


    1. Right-click on Folder view widget > Configure > Icons > Sort by > choose your preference; “Unsorted” allows free positioning.


    2. That’s really strange. Done that, no free positioning possible. The dragged element always jumps back to its former position.
      Tested on Kubuntu 16.04 and 18.04. What else could I try?


    3. Are you talking about a desktop? I noticed this thing that it requires a very strict place to drop, otherwise, the file just jumps to the previous position. There is no snapping to the center of the square as it is on Window (or there is but with crazy small margin).

      I think it should be fixed since this is incredibly annoying behavior. Intuitively a user expects to drop a file anywhere and it should just snap to the nearest place within a grid, not to be rejected and return to the previous place. Currently, it forces us to try and test a few times till we make it right… in the end desktop as a folder is a bit unfriendly to use.


    4. Ah, yes! Nate, you are like a library! How do you know all of this? This is astonishing!

      Good to know that the bug is already filed. I’m surprised thou that there is not much activity in this thread. I would expect people to jump on it as it’s obvious and easy to reproduce, not mentioning how annoying. So many people wanted to have folder view as default so it’s a bit wondering how little people have noticed this bug.

      By the way, Qt Bugzilla looks so much nicer and more modern! KDE one is like pulled from the 90ties of the last century ;). But the more important is the recent Bugsquad initiative then the look. Anything to get rid of the mess and increase productivity :).


    5. The bug is now always reproducible. It seems to require NVIDIA hardware.

      I make it my business to know all the really bad bugs, but the library is not inside my head: the heart of my work is a couple multi-thousand-line text files that are filled with bug reports, both open and closed (along with the version it’s fixed in), categorized by perceived severity/priority. When I find a bad bug (like this one), I just file it away there so I’ll be able to find it quickly. The more times I hear people complain about it and the more duplicates it accumulated, the higher I bump it in the list, until eventually it reaches the “strategic direction issue” category, at which point I try to fix it myself if I’m able, or else start making noise about and working on recruiting others to help or do the work.

      Liked by 1 person

    6. Wow, thanks! That’s a lot of work on top of what you are already doing here and with Plasma productivity. Kudos!

      I do have nvidia but it’s a hybrid setup so desktop runs on Intel (with Bumblebee) and it’s still reproducible. I have 2 of such laptops.

      Actually… I tested it now and I don’t see that issue anymore on my main machine. It runs Manjaro KDE unstable with Plasma 5.14.1 and Qt 5.11.2. Snapping seems to work fine. When I place it wrong, it is snapped to the nearest nod instead jumping to the old place (just like on Windows), so everything seems to work correctly now! Yesterday the issue was still there… So something in Plasma 5.14.1 or KDE 5.51 update fixed it.

      On the older laptop with Manjaro KDE stable with Plasma 5.13.5, Qt 5.11.2, KDE 5.50 issue is still there.


    7. Neth, Zalbarath, this bug report has been closed an hour ago:

      > Gatis Paeglis added a comment – 1 hour ago
      > This can be reopened once the requested example becomes available.

      Liked by 1 person

    1. Hi Nate!

      You don’t need to thank me. I really (and i mean REALLY) think you guys are doing one of the most important works in the KDE Community!

      And letting us know every week about the amazing job you’ve done is a big plus! 🙂

      Thank you 🙂


  9. more interesting stuff, more hacking, a new week, a new report and a new spanish translation:


    Liked by 1 person

  10. I am probably not the only one reading this blog every week and hoping to see these improvements arrive as fast as possible to my distro (because we don’t all live in a git tree). But a few weeks later when Frameworks or Applications update, it’s a bit difficult to remember what part of those cool features actually made it in. There is the full changelog of course, but it’s missing the highlighting your blog is doing and that people really appreciate.

    It would be great if, for each release, there would be a recap of VDG-centered patches that made it in. I am not saying you Nate should do it, but if anybody out there has a bit of time in their hands … !


    1. I actually do this already for the Plasma releases, but because all of our product lines are on different schedules, it’s harder than it needs to be. What we really need is a dedicated release manager who can keep track of all this stuff and do both the technical angle and also the announcements for all releases of all product lines, not just one.

      This is actually what I do in my day job, but for KDE I feel like I have more to offer so it would be great if someone else wanted to step into this role!

      It would also help if our release cycles were a bit longer. We currently have 18 major releases per year (12 frameworks, 3 apps, 3 Plasma), and more if you count the bugfix releases for apps and Plasma. It’s an enormous amount of work and frankly it’s absolutely incredible that the machinery works so well without a paid person turning that crank. Professional companies pay hundreds of thousands of dollars for teams of people to manage release schedules that are far less aggressive than ours are.

      Liked by 1 person

    2. I did not know it was you doing the plasma release reviews. One more line on the awesome list of things you are doing for KDE!


    3. Day job? You also have a day job aside of this? I thought that KDE is your day job. It feels like you are doing day and night job on KDE so how do you even find time for another day job? This is incredible!

      Don’t forget to take vacations and change things once in a while to avoid burnout. Plasma will crumble without you ;).


    4. Yeah man, that’s why I’ve got a Patreon! I’m just one of those people who has too much energy, and for the first time in my life, I’ve found work that I genuinely love that I can pour an unlimited amount of energy into.

      The KDE e.V. does not hire people like me; the mere subject is very contentious within the community. I wish it were different, and I would love to work on this stuff full-time (well, you know me, more than full-time), but at the moment, I haven’t found anyone willing to fork out the dough except for you guys, my loyal Patreon supporters! 🙂

      Liked by 1 person

    5. I understand your POV totally. It’s a very sad situation that so many talented and good willing developers can’t make a living by doing what you are doing now for no incomes. That’s why Linux desktops need to be more stable and trustable. I’ve known many computer users who liked KDE, Gnome, etc, but then, 3 months after being using them, they have reinstalled Windows because of the minor but frequent bugs an deficiencies in Linux desktops. I’m sure that we loss millions of users a year because of such things, and if we didn’t lose them, perhaps the percentage of Linux users would not be that rickety 2 – 3 %, but perhaps more like Apple’s numbers. Then it’d be possible for people like you to work full time developing software for KDE/Plasma and make a profession of it.
      More users mean more demands from users to businesses, hardware manufacturers, service providers, etc. A 6 – 8% share, more or less like MacOS, means thousands of customers complaining to all those companies when they have a problem due to a bad Linux support, and that means many companies hiring lots of Linux desktop developers like you.
      I get really upset when some people say: «Meh, we don’t need to popularize Linux; as long as we use it, who cares about the rest of the humanity?». No, no, and no. We of course don’t have to obsess with it nor dream about competing with Windows, but we do need to multiply our userbase because of the reasons I’ve exposed before, and the possibility of turning your hobby into a job for you, devs, is not a minor one. So, make Linux desktops as great, solid and bullet proof like the “bare” GNU/Linux. If users see that they can copy their photos from their mobiles, that they can trust the desktop search engine is going to work tomorrow, and the day after, etc, instead an “I-shall-work-when-I-ducking-want-if-at-all” Baloo, if desktop updates don’t break their desktop or make some things disapear… In short, if y’all manage to make Linux desktops trustworthy, maybe the userbase would grow and there would start to be more demand for KDE developers in companies who will pay you a salary for your work.

      Cheers, and thanks for the time you invest for making KDE/Plasma better.

      P. S.
      Oh, a suggestion: add desktop-wide sharing and spell checking, like on mobiles. We are in 2018 ;-P


    6. I used to be of the opinion that just polishing our software and fixing bugs and even adding features would be enough. But the truth is that even if Plasma were unequivocally, unarguably better than everything else, it would still be mostly ignored by most people until and unless large numbers of hardware partners ship it on devices. Windows is terrible and has been for 30 years, but it literally doesn’t matter at all because it’s shipped by default on most hardware. Our stuff can be rough and buggy, it just needs to be shipped on someone’s hardware. Once this happens, then there’s a revenue stream attached to it and those companies have a financial incentive to hire FOSS developers so they can have more input and control over the software they’re shipping. This will work even if the software is awful–this is the lesson of Windows: it doesn’t matter that it’s awful. Awful is good enough.

      So it’s a high priority to make a Plasma-bearing Linux distro irresistible for hardware vendors. If you watch talk from this year’s Akademy, this is one of the long-term goals that everything builds up to. It’s step 5 out of 7 (the final two are to improve the PIM suite and LibreOffice). In retrospect, I probably should have made it step 7, since having business-friendly productivity software is enormously important; people buy computers for the apps, not the OS.


    7. «Our stuff can be rough and buggy»

      LOL. Don’t be so exaggerated! I wouldn’t say anything close to that. KDE/Plasma 5 is usually smooth and well-functioning, but yes, it has its glitches and its minor details that don’t work as well as one would expect. But I wouldn’t say it’s rough nor buggy, like Plasma 4 was. xD

      I agree with you, but not totally. Complex matters usually have various causes that are interconnected, not just one and only that if solved will make everything be ok. There should not be ignored the fact that many buyers of Linux PCs have ended formatting and installing Windows, even when at the beginning of their Linux desktop experience they were enchanted with their Plasma, Gnome, Cinnamon, whatever cool Linux DE, but ended fed up with minor bugs that made the experience inconsistent in a medium-long time. Windows, as you say, is “good enough” for them, just because it usually (I remark “usually”) doesn’t suffer those kind of small bugs, suffers other ones much more severe, but the user experience, in general is “good enough” for the average user.
      So, it’s not much benefit to have some Dell PCs with Ubuntu, Slimbook with Plasma or whatever you want if the user is going to format it to install Windows. If animations are not hyper-fluid it’s something won’t make anyone sane change their DE, but if I cannot trust that tomorrow when I turn on my PC things are going to work well like today… Bad.

      So, yes, KDE needs to be shipped on someone’s hardware, true, but Someone won’t ship it if users aren’t satisfied and return the computers to Someone or ask the store for a Windows installation, even if they have to pay the Windows license extra (this seems to happen very often, unfortunately, at least in my country). This is a multi-front and multi-enemy “war”, one cannot miss one of them thinking that there’s only one to fight against.

      BTW, sorry for expressing myself so clumsily in the former comment: I didn’t mean “desktop-wide spell checking”, we already have that since ages ago; I meant desktop-wide predictive text, like in our mobile phones (yes, sometimes it works in a surrealistic manner, but if we didn’t have it our messages would be even worse) or in LibreOffice Writer, which isn’t a KDE not even a QT program. It would be very nice to have it for the browsers, IM, Kmail, Kwrite, etc.

      Sorry for starting a debate. Wasn’t my intention. Keep up the good work, and let’s pray for a more brilliant future for KDE desktops, heheh.


    8. but Someone won’t ship it if users aren’t satisfied and return the computers to Someone or ask the store for a Windows installation

      Yes, that’s the key. This is why I think that our goal must be to get Plasma to the point where vendors want to ship it on their devices because they know their customers will love it.


  11. > Speaking of Stefan, he and Igor Poboiko have been doing an absolutely smashing job fixing Baloo over the past two weeks. A lot of their work is hard to blog about because it’s not immediately user-facing […]

    You could include two screenshots: “Before” shows DrKonqi crash dialog, “After” shows beautiful Plasma desktop 🙂

    Liked by 1 person

  12. «MTP device handling has been rewritten, which fixes an enormous amount of bugs »

    You’ve made me cry…! 6 Years after my first Android mobile, a bug report, several comments in other users’ bug reports, I’m going to be able to REALLY use a FUNCTIONING MTP on KDE?, like in Windows, with no MPT protocol crashes and weird behaviors, or simply a total impossibility of usage? Sighhh… Damn, I’m so emotioned… xDD

    Seriously, I’m very glad that finally developers took the MTP on KDE/Plasma disaster seriously, KDE Connect is awesome, but nothing like a fast, private and direct cable connection. Let’s hope this time is the good one!


    Liked by 1 person

    1. Is it possible that the new feature doesn’t work at all? I’ve tested other USB devices (sticks) and they aren’t turn off as well. I use all the latest version of kernel on my Kde Neon, the framework is upgraded to 5.52 release. So I assume that the latest framework doesn’t implement the stated energy feature.


    2. I’ve also suggested in my bug notifications that the best way to make the detaching operation productive is to divide both the eject function of optical drive and the turning off function of detachable mass drives from UNMOUNT option. It would be appropriate to introduce the safely remove of USB devices such as microsoft OSs in the notification area, in contrary case the system could misunderstand what operation to make and the user to be frustrated from the confusion.


Leave a comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s