Sources

I also had a bad experience with software support. In one of my projects, I needed to use Amazon Workspaces, which actually has support for Linux, but it was for Ubuntu 22.04, and I was using 24.10. So I tried to make some workarounds and configure a lot of things, and ended up spending too much time configuring it, and finally I gave up and switched to Windows. This shows that software compatibility is still a big issue for Linux. … Other than that, developing Linux desktop applications is hard because of the fragmentation of the Linux desktop environment. There are multiple display protocols, like X11 and Wayland, that they need to think of. Linux is in a big transition from X11 to Wayland, and not all distros have fully adopted Wayland yet. This can cause compatibility issues for software developers. Even though you can still run X11 applications on Wayland using XWayland, some software may not work as smoothly as it would on X11.

Big Tech & the major Linux foundations are driving Desktop Linux into the ditch. Plus: No "Linux Sucks" this year. ... {ts:0} The months and years ahead for Linux look really frigin bleak. Less software {ts:10} compatibility, less total features, less accessibility, less tested software, slower software. The the number of ways {ts:20} you can quantify the goodness and the quality of a desktop operating system and platform. all of them. It's just {ts:31} going into the toilet. And it what's really horrific about it is it's by design. {ts:39} It is a goal by big tech, by the Linux and open source foundations, by the major Linux projects. They're all going {ts:48} into this eyes open and they're saying, "Yes, we are going to do this." They're driving Linux headirst into a really {ts:55} gigantic ditch. and it's just really really crummy. I wanted to walk through with you just some of the ways these ... {ts:74} to start with to start with and those of you who have been watching the coverage and listening to the coverage from the {ts:80} Lunduke Journal over the last few years you know most of this but most major Linux distributions have {ts:89} somewhere in their road map, the plan to swap out the core utilities with untested … Rust is going to save us all from {ts:143} memories or whatever. We'll put that all off to the side for a minute. The reality is if you look at the {ts:151} replacement solutions for pseudo LS and everything else, they are distinctly less tested than the software they're {ts:160} replacing and they don't have all the features of the software that they're replacing. … {ts:272} face is when we switch away from the tested supported solution of Xorg and X11 over to Wayland, we do lose some {ts:284} things in the process. And it's worth noting what those things are. We have significantly worse accessibility {ts:292} features. … {ts:314} to Wayland because the Wayland protocol itself and the Wayland implementations that exist don't provide them with the {ts:323} the the vision accessibility features they need. So they simply cannot move over. So that means that we are by {ts:331} forcing people to distributions and people to move over to Wayland from X11, we're saying we want to leave uh {ts:340} visually impaired people behind like on purpose, right? That that is just that's just the way it is. But we're also {ts:347} having significant software compatibility issues and and we are talking about the ability to run some {ts:354} pieces of software. we will lose that along the way but we are also just losing some feature compatibility issues … So, we're we're we're we're losing functionality. We're losing {ts:416} accessibility. We're losing the ability to run some pieces of software. And to and to put a little cherry on top of all {ts:423} of that, we're actually losing for at least some pieces of software speed. Now, I I want to I want to preface this … What they really mean by that is the {ts:538} loss of 32bit support libraries. And this is a real big key thing here because what that means is that by {ts:547} default these systems will no longer ship with the ability to run 32bit software on a 64-bit system, right? … {ts:665} distributions. You will have access to less software. The software you are running will have less features. You {ts:672} will have less accessibility software. The software you are relying on, including the most core utilities in {ts:678} your system, will have not just less features, but be less tested and will have an unknown amount level of uh … {ts:1249} utterly from within. It's being made slower. It's being made buggier. It's being made so it has less features and {ts:1256} less accessibility, less available software, runs on less hardware, and on and on and on and on. People are being {ts:1264} kicked out of Linux because they have the wrong views or are the wrong religion.

Let me start with desktop environments. KDE Plasma what a joke, full of constant bugs and you have to have a degree in computer science to find anything in the settings. Gnome not any better, you need to use extensions to add basic functionality that Windows has out of the box, Cinnamon and LXQT look likey the came from before 2010. Now come the worst ones, LXDE, Xfce and Mate they look like they are from the time when Richard Stallman had just graduated from college. Software stores are crap, most of the time they break and when you ask on a forum like linux.org of how to fix your system they tell you to open a terminal and run some random command. Like what age are we living in the 90's, terminal's shouldn't be needed anymore there should be good working GUI's for everything! … Packaging and package managers on Linux are a disaster. rpm, apt, pacman, zypper, portage, nix. Why the need for so many, all distributions should just pick one that way there is more standardization between distributions Also security on Linux what a joke, no desktop application containerization package managers the packages are installed by root, what happens if the source gets compromized and you run an infected package as root on your system. … We we get to some form of desktop application containerization. Snaps, are a seucrity risk because of the Snapstore and had already had snaps there were infected. Flatpaks are trash and install libraries dependencies double for each Flatpak application and on top of that also not to mention what is already installed on your system by native packaging. Appimages? Shut the frontdoor, these are a laughing joke most of them don't automatically update and they don't even integrate correctly with your desktop environment so you have to manually create shortcuts for them so that you can use them as shortcut or find some abscure program that one person manages for it to be done automatically. Gaming on Linux? Is utter trash, your depending on the Steam client, Lutris or HGL. Some games work others don't and when you want to play your favorite fps shooter or competative game you can't because of Kernel Anti-cheat not working on Linux. You need to look up launch options on ProtonDB or use random install scripts from Lutris. Now you are stuck to opensource games, they are worse that dos games from back in the 90's! Oh wait you can write your own game on Linux because everyone is a super nerde code hacker … P.S Image cropped by GIMP. Let me start with desktop environments. ... Software stores are crap, most of the time they break and when you ask on a forum like linux.org of how to fix your system they tell you to open a terminal and run some random command. Like what age are we living in the 90's, terminal's shouldn't be needed anymore there should be good working GUI's for everything! ... Packaging and package managers on Linux are a disaster. rpm, apt, pacman, zypper, portage, nix. Why the need for so many, all distributions should just pick one that way there is more standardization between distributions Also security on Linux what a joke, no desktop application containerization package managers the packages are installed by root, what happens if the source gets compromized and you run an infected package as root on your system. … We we get to some form of desktop application containerization. Snaps, are a seucrity risk because of the Snapstore and had already had snaps there were infected. Flatpaks are trash and install libraries dependencies double for each Flatpak application and on top of that also not to mention what is already installed on your system by native packaging. … Gaming on Linux? Is utter trash, your depending on the Steam client, Lutris or HGL. Some games work others don't and when you want to play your favorite fps shooter or competative game you can't because of Kernel Anti-cheat not working on Linux. You need to look up launch options on ProtonDB or use random install scripts from Lutris. Now you are stuck to opensource games, they are worse that dos games from back in the 90's! Oh wait you can write your own game on Linux because everyone is a super nerde code hacker

0:00 Intro 0:39 Application distribution is still bad 1:23 Over-reliance on the terminal 2:02 It's janky 2:43 Little profession software 3:28 Bad Support ### Transcript {ts:2} it's 2025 and Linux as a desktop OS still sucks and no that's not just my {ts:8} opinion lonus tval himself has called out its problems despite all the improvements over the years there are {ts:15} still major issues that make Linux a frustrating experience for the average user today we're diving into why Linux {ts:21} is still a bad desktop OS a while back I made a video called Linus tals on why desktop Linux SU {ts:30} which stirred up a lot of discussion since that video I've taken another look at Linux on the desktop and honestly the {ts:37} problems haven't gone away one of the biggest problems with desktop Linux is that self-updating apps basically don't … {ts:64} download an app from a website and it keeps itself s updated that's how desktop users expect software to work on {ts:71} Linux you're either stuck using a package manager that may not even have the latest version of an app or you're {ts:76} forced to manually deal with app Images flat packs or obscure dependencies it's a {ts:82} nightmare doing anything in Linux inevitably requires opening a terminal at some point and that's a terrible user {ts:90} experience even developers don't feel like busting out a terminal just to do basic tasks installing software fixing {ts:98} broken dependencies setting up drivers so many of these things still require running obscure commands that a normal … {ts:124} that's surprisingly easy to do on Linux a simple update and innocent package installation or uninstallation or even {ts:131} just changing a setting can sometimes lead to a completely broken system with no clear way to fix it and as I {ts:137} mentioned earlier a lot of these issues stem from the fact that you're forced to use the terminal a single mistype … {ts:186} versions but they're not reliable and they definitely don't offer the same level of performance or compatibility {ts:193} even when you do get Alternatives they often lack key features or Worse clients customers and co-workers are using {ts:199} Windows or Mac OS tools and you're stuck trying to make something compatible it's just not worth the {ts:207} headache and on top of that when things break on Linux and they will break it's rarely as simple as clicking repair or {ts:214} reinstalling a driver instead you'll find yourself scrolling through obscure Forum posts copying random terminal {ts:220} commands you don't understand and hoping that some guy's answer from 2014 still works if your Wi-Fi stops working get … {ts:250} Linux has its strengths it's powerful it's customizable and for developers and privacy conscious users it's amazing but {ts:257} as a desktop OS for the average person it's still a mess and the worst part A lot of these problems have been around {ts:264} for years with no real progress toward fixing them what do you think am I wrong let me know in the comments and if you

I’ve been developing on Linux for years — sometimes it feels like a dream, other times it’s a fight with drivers and tooling. In this video I share my honest opinion on how easy (or frustrating) it is to be a developer on a Linux machine, the common pitfalls, and practical tips to make it smoother. What I cover: Why Linux can make development fast and flexible (package managers, shells, tooling) The common pain points (hardware drivers, fragmentation, GUI tooling gaps) Language and stack considerations (web, systems, embedded, Python, Node, Rust, Go) Workflows that make Linux great (dotfiles, containers, WSL, virtualization)

This fragmentation of desktop Linux is often seen as a double-edged sword. On one hand, it provides users and developers with choice and fosters innovation; on the other, it can pose challenges for software compatibility, user adoption, and developer effort. In this report, we delve into the current state of Linux fragmentation with recent statistics, developer insights, commentary from Linux leaders, and a look at efforts to standardize the ecosystem. … This entails maintaining different build environments and toolchains for each target. For example, Fedora uses the RPM package format, while Debian/Ubuntu use .deb packages – software built for one won't natively work on the other, so developers often need to repackage (or even recompile) for each major distro family. Moreover, packages built for one version of a distro may not be compatible with other versions of the same distro, due to changing library versions, etc., <a href="https://devops.com/how-app-stores-are-addressing-fragmentation-in-the-linux-ecosystem">forcing developers to support each LTS or release separately</a>. All of this creates a resource overhead that is burdensome, especially for small developers.</p>\r\n <p>Linux application developers have long voiced that inconsistent system configurations across distributions complicate their work. A GNOME software engineer noted that "<strong>E<a href="https://blogs.gnome.org/tbernard/2019/12/04/there-is-no-linux-platform-1/">very downstream change adds yet another variable app developers need to test for. … Criticism_of_desktop_Linux">since they would have to support multiple installation formats and environments</a>.</p>\r\n <p><img alt="" src="/_resx/storage/cb048ddb-ie736/more-distribution-worse.jpg" /></p>\r\n <p>Additionally, lack of a unified app store or standard distribution channel on Linux is a pain point. … As a result, Linux developers (especially proprietary or cross-platform software makers) face a tough question: Which distros do we officially support? Supporting only one (say Ubuntu) misses a chunk of potential users, but supporting many can be prohibitively costly.</p>\r\n <p>In short, from the developer's viewpoint, fragmentation leads to duplicated effort and complexity in software deployment. "Desktop application distribution is complex across all operating systems; in Linux, this is further compounded by such fragmentation and inter-dependencies in packaging and distribution of software," as one DevOps writer summarized. All these varied requirements must be satisfied to successfully deliver software to Linux users, which is "difficult [especially] with limited resources". This has real consequences: some apps arrive on Linux slowly or not at all, and developers may prioritize other platforms first. The frustration is encapsulated by a comment that "when developing for Linux you need to package your application in a few different formats… If the Linux world standardized around snap/flatpak [universal packages] … then we stand a chance". … While LSB had some success (many distros formally complied with parts of it), it ultimately did not keep up with the fast pace of Linux development and has faded in significance. <a href="https://www.reddit.com/r/linux/comments/1ffaxzf/should_we_have_a_new_linux_standard_base_lsb_spec/">One reason cited is that requiring a single package format (RPM) was a sticking point</a> – Debian/Ubuntu did not want to abandon .deb, for example. … For example, multiple distro teams maintaining separate but similar system tools or patches. Detractors claim this divides development efforts and wastes time on integrating components for each distro rather than improving the software itself. Some see the plethora of package managers, init systems, desktop environments, etc., as redundant labor that could be avoided with more unity.</li>\r\n <li><strong>Software Compatibility and Availability Issues:</strong> As discussed earlier, the differences among distros make life harder for application developers. Lack of standardization in libraries and package formats means an app might work out-of-the-box on one distro but not on another without modifications. Proprietary software vendors cite this as a reason Linux desktop is less attractive – supporting "Linux" actually means supporting many platforms. … On the one hand, the abundance of distributions and lack of a unified platform have arguably limited Linux's mainstream desktop adoption – causing confusion for newcomers, extra work for software developers, and hesitancy from hardware/software vendors. Even Linus Torvalds believes a more standardized desktop across distros would help Linux "get over the hump" on the desktop.

Developers {ts:287} know exactly how to package their software for those systems. But on Linux, things are far more complicated. {ts:293} Different Linux distributions use different package managers and packaging formats. Debian and Ubuntu use the {ts:299} Aptude package manager withdeeb packages. Fedora uses DNF with RPM packages. Arch Linux uses Pac-Man with {ts:308} its own package format. This means that if a developer wants to distribute an application on Linux, they often need to {ts:315} create multiple packages for different distributions. Maintaining these packages can be timeconuming, especially … Again, this complexity is manageable for experienced users, but can be {ts:583} frustrating for newcomers who just want their applications to work consistently. Documentation is another area where {ts:590} fragmentation causes problems because there are so many distributions and configurations, tutorials, and guides {ts:596} often apply only to specific setups.

But damn some aspects are really annoying. One other issue is one package installation at a time with a lock which you can't do from other terminal. Here's what GPT says: > The APT package manager uses a lock file (usually /var/lib/dpkg/lock or /var/lib/apt/lists/lock) to ensure that only one process (like apt-get, apt, or the GUI Software Center) is modifying packages at a time. To me this seems like a serious limitation

The major pain point for most Linux distributions was the unintuitive installation. While it was possible for you to get it working on your system, it was not easy for every user. You needed to know a couple of things in great details before you attempted an installation. If you ever tried installing Linux around 2010, you would understand this. … ... Graphics Card Support Yes, the NVIDIA graphics card support has always been on the notorious side because they choose not to open source their driver. However, distros now support installations for NVIDIA graphics systems by pre-installing proprietary drivers by default. While there are still some issues to be ironed out, but it works most of the time. Want proof? I use EndeavourOS with RTX 3060 Ti as shown in the screenshot above. And, yes, that works for me. Secure Boot Support Earlier, we used to entirely disable secure boot because Linux distributions did not boot up with it enabled.

Programmer tools are even worse here. While on Windows most of your functionality will be accessible through GUI tools, on Linux often the best you have is some curses-based terminal tool. Bad tools are gonna be bad tools, but *good* GUI tools will destroy anything that is "good" by CLI standards. The general impression I get is that Linux users put up with subpar commandline development tools, while Windows devs get far-superior GUI tools. … Programmer tools: - Process Hacker: it's htop on steroids. - Debugging tools: the state of the art for Linux here is still GDB/LLDB or barely functional wrappers around it. WinDBG and many other tools like x64dbg are way ahead and have functional UIs. - Event Tracing for Windows (ETW) and accompanying tools like WPA - DependenciesGui/Process Hacker's peview.exe: I hate having to use a vague combination of `nm`, `ldd`, etc... with barely rememberable flags. These tools are much more well-organized and easier to use. … EGL is such a pile of garbage, I hate having to use it. Did they really need separate extensions for "the concept of GPUs exists", "you can enumerate GPUs", "you can get info about enumerated GPUs"? Complete nightmare to work with. And the extensions are inconsistently implemented and there's no writing on what is available where. DXGI just fucking works and does all of this in a clear API doc on docs.microsoft.com with a 100% guaranteed of what APIs will be available where. … It's a fractured nightmare of extensions and protocols. The whole standards process is slow as molasses and basic functionality gets bikeshedded about for years. Some of the hilarious^4^ bullshit that comes out of this. I'm sure there's going to be more: ... ## ELF Linking Troubles Linux uses ELF for binaries and dynamic linking. Unlike Windows' PE32, ELF does not keep track of which symbols are imported from which modules. That means that unlike on Windows, where `foo.dll` imports `memcpy` from `VCRUNTIME140.dll` or what have you, on Linux `libfoo.so` depends on `libc.so` and needs `memcpy` from anywhere. Symbols on Linux are process-global because of how the linkage model works, and this is a *massive* nightmare for forwards compatibility if you are shipping applications. Example down below. … Combining these into one library is massively problematic, because you cannot load multiple `libc`s into a process (due to the aforementioned ELF-sucks linking problems), but you *need* to be able to do that for forward and backwards compatibility, versioning, and encapsulation of dependencies (since `libc` is an implementation detail of your C compiler, NOT a system API). … ### Glibc 2.35 glibc 2.35 (feb 2022) shipped a new dependency sorter or whatever. It is (at the time of writing) extremely buggy causing anything from angry valgrind, assert aborts, erroneous failing library loading behavior, to straight segfaults. All in `ld.so`. I had to spend a whole weekend debugging this, and as far as I could tell I was the first to find this. Then after another week of Linux users crashing in various new scenarios I said "fuck it" and passed a workaround env var from the launcher. Look, bugs happen, but this is literally the lowest and most fundamental component in the system above the kernel. This leaves an *extremely* sour taste in my mouth, you know? … **Somebody at Canonical thought twice about dropping OpenSSL 1.1, and decided that yes, Ubuntu 22.04 should just shit all over backwards compatibility to save some megabytes.** When one of the biggest Linux distros, often considered a "baseline" for various things, just completely disregards basic principles like "a program compiled a year ago should keep working" **I *genuinely* do not know what to tell you except that there is no fucking hope for this platform in its current governance.^5^** ## Targeting Linux As A Developer "Linux is better for developers" is a common mantra. What is commonly left out is that it's better for developers to *work on*. Not for developers to **target**. I can say with massive certainty that Linux is far more effort to target, all while being a SIGNIFICANTLY lower market share. The reasons why are basically explained at length in points above, but seriously. Linux breaks more and has higher maintenance effort, while also having significantly worse tech. **If I only targeted a single OS, a Linux-exclusive version would *still* be much more effort to develop and maintain than a Windows-exclusive version.** Now consider that Linux is a tiny marketshare compared to Windows and run the numbers on that. This is **not for lack of familiarity**, but simply due to **worse underlying technology**.

While Linux development offers numerous benefits, it also presents its fair share of challenges. Understanding and addressing these challenges is crucial to ensure effective and efficient development on this platform. **1. Compatibility and Driver Issues** One of the challenges developers face when working on Linux is compatibility and driver issues. Unlike proprietary operating systems, Linux may not always have pre-installed drivers for all hardware components. This can result in time-consuming efforts to find and install suitable drivers manually. To overcome this challenge, research and preparation become essential. Before embarking on a Linux development project, thoroughly investigate the compatibility of your hardware with the desired Linux distributions. By choosing compatible components and ensuring driver availability, you can mitigate potential disruptions and enhance productivity. **2. Learning Curve** Transitioning from other operating systems to Linux can come with a learning curve. Linux has its command-line interface and a unique set of tools and utilities. Thus, developers who are accustomed to graphical user interfaces might initially feel lost when working solely on the command line. … **3. Software Limitations and Dependencies** Linux offers an extensive range of software choices that cater to various development requirements. However, software limitations and dependencies can pose challenges when selecting and implementing specific tools. Not all software applications are natively designed for Linux, requiring developers to search for compatible alternatives or rely on virtualization solutions. To navigate this obstacle, engage in thorough research to discover software that meets your development requirements while taking into account Linux compatibility. Prioritize utilizing cross-platform, open-source software as it often brings better compatibility and support. Additionally, consider actively contributing to the open-source community to address software limitations and expand possibilities. ## Solutions and Best Practices for Linux Development ... Before diving into Linux development, invest time in researching and gathering necessary resources. Familiarize yourself with available distributions and select the one most relevant to your project needs. Study the compatibility of your hardware components and carefully choose tools and software with strong Linux support. By performing comprehensive preparations, engaging forums, and communities, you can build a sturdy foundation for efficient Linux development. … ... Developing on Linux presents numerous challenges, but with proper preparation, dedication, and continuous learning, developers can leverage the power and flexibility offered by this robust platform. Overcoming obstacles such as compatibility issues, learning curves, and software limitations becomes easier by following best practices, actively participating in the open-source community, and tapping into the vast resources available. Embrace the challenges, and Linux development will open doors to limitless possibilities and propel your projects to new heights.

This was always a trade-off. Linux is better for various development things, because it's a primary target platform for what I do. PHP, Node, Rust, NPM, Docker are all tools that work better on Linux. Most notably, they are faster because they use Linux idioms that are slow on Windows. I've always accepted this trade-off because I wanted a working system first. … Now, to be fair, many similar issues exist on Linux, too. Password input on Ubuntu randomly switches input language, sometimes my screen FPS drops to like 1, and there's no way to get out of that without a restart. Wi-Fi was broken when I tried to install Ubuntu. In the end, that was always the reason why I didn't want to use it as default. But while Windows is now somewhat similar (or worse) when it comes to this system stuff, it's worse for actual work I do. As I come back to my main Windows machine, I ask myself: why do I deal with all this crap and on top of that get slow build times, slow editor, slow terminal, and missing unix tools? There's also a question of a general direction. Windows doesn't seem like it's going to get any better. Every update just messes up something new, make something slower or more broken, more bloated. I don't see Microsoft doing meaningful improvements anytime soon. … `StartMenuExperienceHost.exe`, and I can tell you - what an experience this Start Menu is. First of all, for something that is one of the primary navigation tools, it has incredibly high latency. 1s to respond to a click is just crazy. But more importantly, it's just broken so much of the time. Sometimes you click search and nothing happens. Sometimes you click the start button, start typing to trigger search, and nothing happens. Sometimes all icons disappear. Sometimes the popups for different windows don't work. I have to kill explorer.exe and restart it. I don't know how non-technical people fix this.