My Linux Desktop Manifesto

It’s time.

27 years after the creation of Linux, I firmly believe, we are finally at a point of quality usability for the Linux desktop. “The year of the Linux Desktop” has been a joke for a long time, as the fractured FLOSS community has struggled to gain a footing on the average desktop.

There’s a reason.

The community has always prided itself in its choice. Don’t like something? Replace it. Want to change something? Fork it. Choice is great, and a free individual certainly appreciates it. But, it hinders development. Let’s be honest, there aren’t a ton of us working on the desktop. What small community has been hard at work over all of these years, has always been split. Just in desktop environments we have GNOME, KDE, XFCE, LXDE, MATE, Cinnamon, Unity, Budgie, Pantheon, Deepin, etc. And that’s not listing off all of the dead projects over the years. Same goes for the applications, we have two or three or four relatively popular applications that fill the same needs, in every area. We rebase, refactor, rewrite, rebuild, replace, rework. We duplicate efforts endlessly.

But, even with this fracturing and duplication of work, we finally have a solid base to use. I’ve been using Linux for 15+ years, all of them as a desktop. I’ve witnessed its evolution, its hardships, its victories. 2018 is the year I’ve finally witnessed the Linux Desktop “just work”. The installers are easy, the applications are mature, the desktop environments are capable and stable. Drivers auto-detect, configuration auto-define, graphics auto-adjust. Networked printers of all things, automatically detect and install. It’s all quite impressive.

We need to consolidate and focus.

We need to focus on what will attract more users. We need to focus on what will attract more developers. We need to focus on what is working well, and polish it up. Eliminate that hacked together, pseudo-functional feeling of much FLOSS software over the past 30+ years.

Not long ago I became a convert.

I used to be a huge GNOME fan. Starting out in the early years, I enjoyed the GNOME design, especially the much lauded GNOME 2.x series. One of the features of GNOME, even in the early years, was its use of designers to craft the GNOME Human Interface Guidelines. The HIG helped create a harmonious and pleasing system, one that removed much of the hacked together and clunky designs of FLOSS software. Upon the release of GNOME 3.x, the HIG was modified to drastically simplify applications. A simplification that many reviewers believe went too far, even stepping past Apple-level simplifications.

The average person wants a system that can be simple, fast, easy, and powerful when needed. They want something familiar to them. Considering Microsoft Windows still holds almost 90% market share, they want something that looks like and works like, Microsoft Windows. Like it or not, we need to support traditional Windows-style work flow to attract more people.

The best environment for these requirements is KDE Plasma.

After my third GNOME 3.x crash in a week, causing my profile to corrupt and be wiped, I had enough. I spent some time re-evaluating the Linux desktop environments, and after a few tests, I have settled on KDE Plasma. Originally reluctant, due to my bias towards it from experiences a decade old, I sit here duly impressed. Yes, there is MATE, Cinnamon, Budgie, and XFCE that offer the traditional Windows-like desktop experience. But I believe KDE Plasma surpasses those in performance, modern design enhancements, and customization.

I hereby declare my time and energies toward ensuring KDE Plasma is the best Linux desktop environment. The default selection for a new user, and an obvious choice for business use. Plasma is simple and elegant, yet advanced and powerful. It’s defaults are very similar to Windows. The average person would love this desktop.

If they knew it existed.

25 thoughts on “My Linux Desktop Manifesto

  1. I had a similar experience. I got fed up with GNOME and gave plasma a try despite some sour memories. I’m so very happy with the performance and power.

    I agree that consolidation is of vital importance. Like it or not, systemd has us all using a unified foundation. We really don’t need eight million versions of the same dumpy shell init scripts.

    I’m concerned about Wayland, not just because it is so slow to get here, but also because of the reduction of features without recourse. If the wlroots project hadn’t started every DE would have to reimplement everything. Gr.

    Liked by 1 person

  2. My current biggest desktop-related concern right now is system-wide, cross-desktop/cross-display manager, Polkit-enabled screen/input device/power management/sound device configuration. We’ve gotten by with xorg.conf and other hacks for long enough; with Wayland something will have to change but I fear that it’s moving toward per-user configuration for everything, which is very bad, especially in an enterprise environment.


  3. Plasma is a great desktop that you can also configure for ease to use. However, the applications under KDE are usually much more cumbersome and overloaded than GTK applications. I think KDE applications are good for professional users and business as they offer more possibilities. For the “average” home users, however, the GNOME applications are easier to use.


    1. I like the default design of GTK3 apps, even the larger title bar with integrated buttons/menu. What I don’t like is the minimal configuration beyond that initial offering.

      The KDE app design clutter was one of the reasons I hesitated trying Plasma. I remembered that from years ago. The good news is KDE is working on cleaning up interfaces right now in many areas. We have an active Visual Design Group making mockups and standards. It’s just a matter of time. Some of the work is already done as part of Nate’s Usability and Productivity goal.


      1. Good to hear. I’d like to switch to plasma, because I appreciate the consistent user interface (GNOME caused a lot of confusion in menu navigation with CSD in the GTK-World – some apps use CSD, others do not). I’m curious to see how KDE makes the transition: becoming more attractive to new users, but not chasing away the old users.

        A good example is e.g. SimpleScan. The scan application under KDE is also simple, but too cumbersome for my wife and father-in-law. Also, GVFS currently works better for me when accessing my filer, but that may just be a matter of habit.

        I continue to sympathize and keep an eye on plasma. Until then, for me (and especially my family), Mint Cinammon/XFCE is the best compromise.


    1. I’m not flaming, I mentioned some positives of GNOME as well as my recent negative experience, and that I was hesitant to use KDE at first due to old experiences. Nothing is perfect, I understand that. For example, I really do like GTK3 app design, I just wish they had more configuration options under the hood.

      I’m stating my beliefs we should consolidate efforts into KDE, and within KDE, we should focus efforts on polish. Not fork Plasma and make another DE.

      Liked by 1 person

  4. I made similar progressions myself. I was introduced to Linux with Ubuntu 10.04LTS. Then GNOME until it switched to GNOME 3 and lost much of it’s customization points that I was attracted to in the first place. Then I switched to Kubuntu somewhere around the 14.04LTS release and I have been with it ever since. Now my whole home network uses it (two desktops and one laptop) and I run mine in developer mode. I have surprisingly little trouble running the developer edition, but I have also taken the free Linux Foundation training courses before I made the switch.


  5. Great piece. I’ve been using Linux not nearly as long as you have (about 4 years), but even in that time the advance has been amazing. In the Ubuntu world, 17.10 was a really big advance, especially for sound, printers, and bluetooth.

    I too really love and appreciate Gnome design, and if Gnome’s performance was similar to Budgie and Plasma (the two I use) I’d probably still be using it. I was troubled, though, that I needed something like 15 extensions on Gnome to give me the same workflow I enjoy by native configuration on Plasma (and half of those extensions break with every Gnome upgrade). I was also deeply concerned by the fact those same extension devs constantly complain that Gnome is slow or even absentee in reviewing their work or communicating with them, plus I felt the whole removal of the panel icons was an interesting yet really hostile move for all of us that use software that depends upon it.

    For the average “muggle” user (like my sister and parents) I have them on Cinnamon, which I think is still the best “out of the box” Windows emulation experience on Linux. They have fantastic settings menus, which are clearly written and not dumbed down. This is something I think is underappreciated and an area that Plasma really, really needs to work on. There is no way I could explain Plasma’s keyboard shortcuts menus to my family. After months of using Plasma, I’m not sure I myself have a 100% grasp of that :).


      1. Great to hear! I think cleaning up system settings; taskbar config and icons; and media controls are the only thing keeping Plasma from near bliss. It’s amazing that such a powerful and configurable system has near the system utilization and speed of XFCE…


  6. This was a great read. Really made me think a lot about this very subject. I’m someone that started using Linux 24 years ago. For the longest time, I ran an old ThinkPad T40 laptop and had to run a less resource-intensive DE. About 6 months ago, I purchased a new laptop and was able to finally, after so many years, give Gnome and KDE a try. I really did like Gnome but after a few weeks running it, I started experiencing minor issues that I could never really get resolved. I’ll be honest, with KDE, I only ran it for a few days but this article sparked my interest and I’m going to give it another try. If you don’t mind me asking, what distribution are running KDE on?


    1. Thanks for the comments, I’m currently running on Manjaro and Kubuntu. If you want a traditional release cycle, I would go with Kubuntu. If you’re interested in a rolling distro to always get the latest stuff, I would go with Manjaro, or openSuSE.

      Liked by 1 person

      1. So, I’ve been running Kubuntu for just over a day. Have to say that I’m impressed and really like it. The last time I ran KDE for more than a few hours was way, way back on Mandrake. And how long ago was that? 🙂 Will it replace my favorite desktop, Xfce? Not sure but I will say that I really like it and I’m going to keep it on my laptop for a while to give it a really good try.

        Liked by 1 person

      2. Good to hear! I previously tried KDE years ago (in the Mandrake timeframe) as well, hence my hesitation at first. It’s come a long way!

        Liked by 1 person

  7. Some great thoughts – trying to streamline efforts to put our Best Foot Forward is a laudable suggestion…

    I did a piece on the topic, some time (over 4 years, good golly) ago. I think I came to the same base conclusion (consolidation is needed), but my answer to the problem lies more with what neatly labeled thing should be pushed to “average users” as the base concept, as opposed to trying to corral the entire community of communities into a focused effort.

    Specifically, I’ve gotten to a point where I’m more likely to push for the “Ubuntu” brand (a commercial name, a known quantity, which already has a good foothold as a pre-installed system, which manufacturers can home in on, and users can call for), and leave talk of “Linux” to the technophiles. I’d say, forget about the Year of the Linux Desktop, and focus on the Year of the Ubuntu Desktop.

    In fact, I’d even say that non-technical end-users should expressly *not* sold to on the idea of Linux, but that sold Ubuntu (or at a push, Red Hat), the same way Windows users are running “Windows” system, not an “NT” system.


    We kinda disagree on what to push for – you’re talking about technical stacks ; I’m pushing the idea of a brand. To be honest, I prefer the MATE desktop which I use via Ubuntu MATE, but to give GNU/Linux a good chance on the desktop, I believe it’s the brand, preinstalled and touted by manufacturers, that will carry the win.

    Ubuntu is only one I can think of that has significant pre-installed deployment presence with consumer hardware manufacturers (Dell XPS is a prominent example, many consumer oriented manufacturers ship with Ubuntu as the OS of choice – System76 (til recently), ZA Reason, Entroware, SOL project, etc). Linux Mint, for the most part, can easily be marketed as “Ubuntu Compatible”, down to the PPAs, and this should only get easier with Snaps and Flatpack, allowing all sorts of Fedora, Arch and SUSE to also be “Ubuntu Compatible” as time goes on.

    Canonical becomes the purveyor of a new household name OS, and subsidiaries/franchises/indpendents can sell “Ubuntu Support”, and the end consumer can remain free from us lot, the technorabble, be we Linux, BSD, Mac or Windows.

    Adoption itself will be won not on technology, but on a brand name – you can’t rally a crowd around domain-specific expressions now, can you? 🙂


    1. Thanks for the extensive comment! I agree with the branding, Ubuntu is certainly more palatable to people. How about I meet you half way and say Year of the Kubuntu Desktop? 😉


  8. Ubuntu at least tries to give Linux desktop a brand image. It goes a bit beyond being a “hobby” OS for nerds. It tries to help new users get familiar with Linux. these are average non nerd people. Its why some do like Chrome OS although they may not even know its Linux connection. Definitely the splintering of distributions has not helped Linux desktop. Still most people not nerds don’t seek out a different OS then came with their PC. They take what they get and make do even if it frustrates them or they want something else. Be nice to see more alternatives in the market for Linux desktop OS on devices so at least consumers could check them out. People who buy Chromebook’s probably don’t even know Linux is part of that OS. They just know its what they want and it works.


  9. Seems to me Linux desktop as a concept is so fragmented that its chaos and for many people that don’t want that they see Windows and MacOS even Chrome OS a better option because its presented as a package, you have little choice and for many that eliminates the guess work. It can be a daunting task to select and install a Linux distro. Its a wonderful thing to have choice, but it clearly can be too much overload for some.


Leave a Reply

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

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

Google photo

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

Twitter picture

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

Facebook photo

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

Connecting to %s