I hope some ambitious soul can port the Chrome OS window manager in some way. Chrome OS is the only non Windows or Mac window manager that handles UI scaling correctly in all cases and across all displays with differing scaling settings. There is no Linux WM that I've found anywhere near as good (don't say i3wm).
I hope not! Outside of the Chrome part, ChromeOS is a mess of awkwardly attached virtual machines. If you want a "native" app (unless your native app is Steam), your best option is Android, which means you're inside an Android VM, your webview component is an Android Chrome webview, you can't see other things on the network (because you have a virtual network device), permissions are granted via the Android Settings panel which is distinct from the Chrome OS Settings panel and an even more user-hostile experience than it already was, and your widgets are all slightly wrong. It is bloat all the way down. Running plain old Linux apps is a similar problem, except it's a "developer feature", so you can't expect an end user to use your app on their Chromebook.
It is exceedingly difficult for a developer to contribute any kind of improvement to this because the bug tracker is a black hole and the source code is a gargantuan monorepo with multi-page build instructions that basically require you to sacrifice a computer if you want to get anywhere.
Sandboxing is great and all, but the thing is a glorified hypervisor. It has no vision of its own for app distribution, user experience, or developer experience. It boggles my mind that they are doing this with ChromeOS: the OS that runs on the cheapest, slowest, most under-specced computers on the market.
If that's what you want, then use it. There's nothing stopping you. Other people choosing to use something different should be no skin off your back, unless what you really want is some sort of suicide-pact.
More accurate title... I love the Linux desktop, but I just want one version of it so that I don't have to deal with competition and jobs with different tools. In other words, I'll likely switch to Apple.
That, and I don't understand the needless fascination with "crushing windows." That's a goal that seems completely misaligned with all the features that make the linux desktop something I want to use; and destroying competition has never been a great long term strategy.
I don't care about your market share.. I care about the quality of my own experience.
It's because market share is important if you want hardware and software support. Right now Linux has terrible support VR, for instance, and one can only imagine that would be different if it were the majority desktop OS.
> I don't understand the needless fascination with "crushing windows."
That was the common sentiment 20+ years ago when Steve Ballmer, then Microsoft CEO, declared that "Linux is a cancer". Today I'd say it's not the case anymore, especially so since they were forced by market laws to fake some form of friendship with Open Source, not to mention having ruined Windows with all versions from 8 on.
Windows is pretty much crushing itself these days.
> I don't care about your market share.. I care about the quality of my own experience.
Exactly this. I've seen where chasing mainstream usability and appeal leads: ipadification. If you want an ipad go buy an ipad. I don't care about linux 'winning' because that sort of win condition would make linux into something that has no value to me. A Pyrrhic victory.
macOS is feeling too restrictive these days. Windows 11 feels restrictive too, not being able to move your taskbar to a vertical position is stupid on an ultrawide.
Fedora Workstation is nice, not perfect, but it's nice. Tried the KDE spin as well as the usual GNOME one. It's my favourite Linux desktop distro FWIW. PipeWire's amazing.
This article is sort of frustrating, because the issue they have isn't like something not working, just that there are multiple competing options and they want unity around all things.
Doesn't seem like there needs to be unity, just that things from the OS are handled in an OS specific way, but things not specific to the OS are handled in an identical way. Seems like container systems like flatpak deliver this. If google chrome flatpak was supported on all OSs then it wouldn't particularly matter to application developers that there were 200 flavors of linux around.
One reason Windows and macOS are popular is that people know what they are going to get. If you've used one Windows 11 box, for example, you can get on another Windows 11 box and figure out how to use it. It will have the same user interface and window management as the other: the Start menu, the control panel, the desktop, etc.
That's just not the case for Linux. There are several popular desktop environments that are very different from each other in look and feel. For an ordinary user, switching from Gnome to KDE would be quite jarring and it would take them a while to get used to it. For power users and tweakers, that diversity of options is an attractive feature of the Linux desktop. For ordinary users, it is a deterrent.
Therein lies the conundrum: is Linux desktop for power users primarily? Arguably yes. Most people don't know what they want/need, and most people don't want/need x features at the cost of y compromise, if they even understand what the compromise actually means to them as a user. I don't think 99% of people care.
Is it for every power user? No, of course not.
Can Linux be made very simple for casuals? Yup, absolutely, if all people want to do is browse the web, send some emails, the usual stuff, it's fine. The in-between of "can I use Photoshop?" and the answer being "yes, but" isn't going to fly with most folks.
I use it for what it's best at, and use Windows for what it isn't best at, sometimes as a bottle, other times not. But it takes experience to know when to use what, which is what most users don't care to figure out for themselves as it involves trial and error usually. Most folks just want to do their work, and their phone or tablet is the "funsies".
The problem is there is no decent software that runs on linux; I'd switch from Mac to linux if I could run Ableton, Adobe_products.* or Affinity_products.*
And no, I don't want to hear about the software XYZ that is almost just like any of the above, because they aren't even close.
If I just wanted to browse the web, and work in a terminal, then linux is great.
The fragmentation is what scares companies like Adobe away.
How many distros would they need to support to make customers happy? At what point does it become a problem of diminishing returns?
Windows and macOS are pretty big. How many more customers would Adobe get by adding support for: Ubuntu, Fedora, SUSE, Mint, Arch, Manjaro, MX, Gentoo, Elementary, and so forth? At some point it just gets silly and devolves into Muppet names and other nonsense -- and each one has a smaller amount of users.
Plus, real professionals are willing to pay for software. Open source is frankly not relevant for professional image editing, video editing, music production, CAD, etc.
Yes, exactly. It's the same problem all over, too - look at transport! You shouldn't ride a bike, you should drive a massive V8-powered 4x4 like I do. It meets my needs, so it must meet yours, right?
As a desktop, Linux is now fully functional for most users. I know administrations running on Linux, law firms, and even my dad. Beginner users don't care about choices, they get Linux Mint or similar with LibreOffice and Firefox and they are good to go for most of their tasks. For advanced users, the abundance is a key benefit of this ecosystem.
No company is fully focused on the desktop OS, and I don't see why they should. This article is contradictory saying "desktop operating system will lie in the hands of Apple with macOS and us with Linux". It is not because Microsoft prioritize cloud services that it abandons desktops, it is not a priority for many Linux developers and sponsors as stated above in the article, and it doesn't seem a high priority for Apple either.
The thing about Linux is that it’s hackable and you can customise it to your liking. I’ve been using it for about a decade and the only thing it can’t do is run iTunes so I have a virtual machine for Windows stuff. It’s worth having a separate Linux box where you can tinker with new software and customisations without those changes affecting your daily driver OS. I found that over time the more customisations you do the more the OS breaks down and starts to get buggy. Having a dedicated box for tinkering is essential and you can just do a fresh install when you’ve made many irreversible mistakes.
What I would love is easy GPU (and other hardware) passthrough to the (Windows) VM. You can already use "real" partitions on separate drives to maximize performance. Graphics is the only thing that's complicated.
Tbh, I only tried it on a laptop and it was hell, it ended up needing a mountain of configuration and software just to pass an nVidia GPU (fortunately a Quadro, which is slighly easier) to Windows, which has the software that needs it the most.
Is it easier on desktop computers with an IGP and a discrete card, or two graphics cards? Or do you still need to fuck around with the VBIOS, fight with power management and hybrid modes (I guess not?), use Looking Glass/RDP/dummy display/DP/HDMI plug and a ton of configuration that sometimes fails after updates?
Guess I will try it again on my next computer, which must be a powerful Ryzen build.
If the Linux desktop consolidates/unifies around one option I’d probably stop using it. The primary reason I use Linux on my desktop is so that I can use my niche window manager of choice as a first-class citizen, rather than hacking reasonable window management into the system with scripts on top of a bad-by-default window manager. If I can’t do that anymore I might as well use macOS with yabai and get the broader application support.
I agree that the customization is a huge draw for me.
But... I also want things like hibernate to just work, screen tearing to not happen on X11, etc etc.
The polish is absolutely not there, and I agree w/ the article that's a result of the fragmented ecosystem. If the Linux community could agree to invest in the underlying APIs for desktop stuff, it seems we'd be in a much better place and still have lots of customization options on top.
Of course, I understand that's a huge coordination challenge.
Hibernate is all in the kernel. You trigger it by writing to a file in /sys. Your DE choice has nothing to do with this.
Most things on Linux are actually like this, where everything is in the kernel or some other small project and the DE just adds a wrapper GUI (or sometimes a management deamon) and really has absolutely nothing to do with how the thing works.
And thank goodness for that because all of the popular DEs use terrible window managers. I wouldn't bother with X11 if those were the only functional choices.
Also some people (like me) really don't mind tearing but can't stand an extra frame of latency when we're typing. It's nice to have options since no one agrees on everything.
Yeah, I'm aware it's in the kernel, but it's only useful for PCs. So effectively it's a "desktop" problem in terms of people deciding to dedicate resources to it. And... hence why it breaks all the damn time.
I can't understand what on earth your comment is supposed to mean. Are you saying most people working on the kernel don't expect it to run on PCs? And so there are no resources dedicated to the suspend states? Also of all the different suspend states suspend to disk definitely breaks the least. Most importantly I really can't understand why you think any of this has to do with desktop environments.
My point is suspend states aren't important to big commercial users who are running linux servers. They have no reason to contribute to making it work well. Hence, it likely doesn't get much attention compared to many other parts of the kernel.
It doesn't have to do with "Desktop environments", but it very much is part of the "linux desktop" experience.
If you have AMD hardware that uses the 'amdgpu' driver, then -IME- enabling the TearFree option seems to completely eliminate tearing. (Though, I am using KDE, which _also_ has a "Make the compositor use vsync" option... which didn't actually eliminate tearing, when used by itself. But, there _may_ be confounders.)
Anyway. Drop this into a file named "$WHATEVER.conf" in '/etc/X11/xorg.conf.d' (or into your xorg.conf), and you should be good to go:
FWIW, I assume it adds a (maybe one frame) delay. It _is_ on by default when you've rotated or transformed the output (and for "secondary outputs".... whatever a "secondary" output is). From 'amdgpu(4)`:
Option "TearFree" "boolean"
Set the default value of the per-output 'TearFree' property,
which controls tearing prevention using the hardware page flip‐
ping mechanism. TearFree is on for any CRTC associated with one
or more outputs with TearFree on. Two separate scanout buffers
need to be allocated for each CRTC with TearFree on. If this op‐
tion is set, the default value of the property is 'on' or 'off'
accordingly. If this option isn't set, the default value of the
property is auto, which means that TearFree is on for rotated
outputs, outputs with RandR transforms applied and for RandR 1.4
secondary outputs, otherwise off.
Edit: Also, distro packagers are quite able to set this on-by-default. It is configurable, after all. If you don't like the default that your distro has chosen, go chat at em. :)
The default is supposed to be Wayland, since Xorg is no longer being actively developed. Though it's taking some distros a while to move to it as default.
I've no idea why so many people expect completely different Unix-clones with completely different origins that simply share one component between them, which they also all heavily modify, to somehow give up their own identity and merge.
It's probably simply because others all call them “Linux”, but they rarely even call themselves that. One cannot easily find the word “Linux” on the websites of “Fedora” or “Ubuntu” for instance and they simply call themselves that.
Monopolistic desire is a mind virus that people steeped in corporate culture acquire and bring with them into scenes like this. To a mind so afflicted, the premise that 'Linux' should be unified to maximize the chance of achieving total market domination seems to go unquestioned.
I get (and hate) the fact that Microsoft is pushing hard for the desktop to become a light client integrated in their cloud, but isn’t apple doing exactly the same? I don’t own a mac but my iphone nags me all the time to use apple’s cloud services for everything and I understand apple is making MacOS converge to iOS.
The obstacle to switching to linux for me is not too many distributions, it is that I have to relearn everything, from my way to the filesystem, how drivers and networking work, virtualisation, web servers, etc. And linux is many things, but intuitive and self discoverable it is not. I just don’t have the time for starting from scratch (and it’s not related to my day job either), until windows becomes really user hostile (I see it coming but not yet there).
Make the time. It’s worth it I promise. The freedom that comes with Linux is astonishing. I’ve been using it for a decade now and haven’t looked back. ‘If you want to move mountains you start by picking up small stones’
PopOS needs an honorable mention here for the work it's done with pop shell for linux desktop. PopOS has been my daily driver for 3 years, pop shell is icing on the cake. Nvidia drivers baked in too!
I hate it, which is why I don't use it. Ubuntu got the closest for me, but the owner's refusal to allow users to move that taskbar to the bottom of the screen made me give up on it altogether. And unless it's a supported distro, I don't want it. Been burned too many times by pet projects that end up being abandoned.
The other thing is that in all my years of using Windows and MacOS, I have NEVER had an issue where an upgrade borked being able to log into the desktop. This has happened more times than I remember on Linux.
>the owner's refusal to allow users to move that taskbar to the bottom of the screen made me give up on it altogether.
You ...do know there are countless desktop environments and window managers just an apt-get away right? If that's too much for you, you could have literally chosen any other flavor. My mother has been happily using xubuntu for years. If she can manage, I think you can too.
A while ago when I used Linux I found that installing some alternate desktop environment usually messed up some default even if you switched back. I don't remember exactly but it'd be like, you start on GNOME, install KDE, switch back to GNOME but now your file explorer is Dolphin.
Has that happened to you? Is that still a problem?
On a lot of distros, many DEs or WMs do have worse defaults or weird glitches that other, blessed ones do not. The dice may come up in your favor more often than not, but you're still definitely rolling them when you move away from a distro's favored set-up. There's a reason alternate "spins" (distro derivates, whatever you want to call them, I think that's the Ubuntu term) exist, with different default DEs or WMs—just installing those packages on the base distro may not provide a great out-of-the-box experience.
And yeah, the way settings work on Desktop Linux can lead to all kinds of weird stuff if you start actively using multiple environments. Should changing your keyboard layout in one DE affect all the others? Well, wait, which of multiple ways did you accomplish that change in keyboard layout? Should it affect bare terminals, outside the window server? Should some of those methods affect all of them, but other methods not? Which one "wins"? Was this other thing a DE setting, or an X-Window setting? Or neither? Or was it Wayland—wait a minute, some of these DEs run under Wayland and some under X-Window! Gee, surely that won't cause any settings conflicts or surprises.
Hey I have an idea, instead of responding with hostile snark why don't you quote the bits where you supposedly did those things? Here I'll quote your message, feel free to highlight anything you like:
>I hate it, which is why I don't use it. Ubuntu got the closest for me, but the owner's refusal to allow users to move that taskbar to the bottom of the screen made me give up on it altogether. And unless it's a supported distro, I don't want it. Been burned too many times by pet projects that end up being abandoned.
The other thing is that in all my years of using Windows and MacOS, I have NEVER had an issue where an upgrade borked being able to log into the desktop. This has happened more times than I remember on Linux.
Maybe OP doesn't want to use some other desktop environment, maybe OP wants to use the Ubuntu desktop environment, but with the ability to move the taskbar. Is that in your apt-get somewhere?
It sounds like your Linux experience is pretty old. It may be worth your time to try again. Ubuntu has lost much of its favor. I'd suggest Linux Mint, which is much better than Ubuntu ever was. I personally prefer their XFCE desktop edition and wish they had a KDE desktop edition. Their Cinnamon desktop is too bland for me.
I've used Mint. It wasn't a good experience. I like and use daily the command line version of Linux and despise the desktop version in all its various frankenstein's monster flavors.
And for the record, I've been using desktop Linux off and on since before Slackware was released. I simply no longer have any interest in making yet another attempt at meeting my day to day needs with such terrible products. If I really wanted to use a Linux gui app that bad, I'd just dial it up in the Windows 11 hypervisor.
KDE is the most fully featured Linux DE, and it's leaps and bounds above Windows's (WAY better file manager, better launcher, more features, and even less bugs somehow).
The other day, my Windows 10's explorer.exe crashed when the file manager tried to load a directory, which is something that used to happen in Vista and hasn't happened in KDE for years IME.
I've been trying to get kde connect working for about a month and I've yet to get to the point where my laptop and phone can actually 'see' each other even after whilelisting the ports in the firewall (or turning it completely off for that matter). FWIW I'm on Endevor OS (arch) with an xps 13 and a pixel 5a.
This is very far from my experience. I installed OpenSUSE Tumbleweed (recommended as one of the best KDE distros) at the beginning of 2022 and ran into a dozen KDE Plasma bugs within an hour. Stuff like panel widths resizing just because I clicked "customize," or widgets vanishing when I tried to drag them onto panels, or the bluetooth icon vanishing from the status bar until I restarted my session, or even the settings manager freezing up.
The basic desktop worked well enough, but as soon as I started customizing it I ran into all sorts of paper-cut edge-case type bugs.
>Stuff like panel widths resizing just because I clicked "customize,"
This can be finicky, I can confirm.
>widgets vanishing when I tried to drag them onto panels
Widgets are also finicky. I don't use them. But they're niche KDE things that you don't really have on other DEs anyway, so I don't count it as diminishing KDE as compared to other DEs. :p
I've never used OpenSUSE, but, for what it's worth, I had tons of KDE crashes years ago when I used Ubuntu. I never had any on Manjaro or Arch, and I've used them for a few year now. My diagnosis there is that the rolling release model is paradoxically better than the point release model in terms of DE stability. I think that's the reason that KDE Neon is a thing (rolling DE updates atop regular Ubuntu). Tumbleweed is apparently a rolling release, though, so Iunno.
Every default KDE install I've done comes with a default theme that is hideous and looks like it is designed by amateurs (I realize this is subjective, but it's also a fairly common complaint).
However, I always hear people sing the praises of its stability.
Is there a good theme/font/icon pack that you like that makes it look more polished without a lot of fussing on the end users' part?
Well, I'm not that picky about the look. I customize the theme, but I don't think about it too much. I use Breeze Dark with PlasmaXDark for window decorations. I also don't think it's looks bad, but that's me.
But the DE, feature-wise, is so great compared to the others that it's like picking between a Volvo, a Model T, and several bicycles for my daily drive. The finish on the Volvo's dashboard is the least of my concerns; I'm just happy it's got AC.
I intend to switch my custom built PC from windows to SteamOS 3.0. I haven't been a windows fan since I switched to Manjaro and eventually to Arch; the only reason I have it around is for gaming. But considering Proton is getting much better (and I'm gaming lesser), it seems like a pretty easy sell.
In fact I think SteamOS has a lot of potential to bring gamers over to Linux; 3.0 uses arch with KDE so it's not really alien in terms of UI compared to windows. The biggest problem, as always, are GPU driver issues.
But the steam handheld is already a successful unit that has seen some good traction, I wonder what a desktop version launched by steam would look like (they had the steam machines project but it looks pretty dead now).
Nonsense. There are billions of Android devices sold and customers don't seem too bothered by its fragmentation. They continue to buy Android devices en mass.
So it seems fragmentation has many problems but that alone won't prevent people from using a device. I think the author believes that a critical mass of Linux desktop users will create enough interest in polish and solving problems for end users, so therefore believes if everybody was using the same desktop there would be more polish (or something).
I would counter that the desktops have plenty of polish. He says he loves the Linux Desktop and doesn't offer much criticism besides there's too much choice. Nah, they're already great and we don't need to abandon great desktops.
Instead, I believe other factors are at play. Familiarity, inertia and monopolistic competition to name a few. Those are powerful factors and much more so than abstract fragmentation.
Android is effectively the default choice. If you don't care about your phone, you end up with an Android phone. Lots and lots of people use an Android phone and almost none of them would be called Android enthusiasts.
True, cost is also a motivating factor that is much more powerful than fragmentation concerns. This is probably why Windows is now essentially free or at least a hidden cost for the end user.
Once Windows stopped having shit-tier stability, some time around late XP or early Win7, Desktop Linux replaced it as the least-stable and buggiest of the three GUI desktop operating systems I regularly use.
Even before that, it nearly matched it in instability, if you counted X-Window crashes as the same as a bluescreen or kernel panic (and if you're doing GUI work, they're close to equivalent)
The kernel might be OK. The GUI & multimedia stacks aren't.
I'm running an MSI laptop on Windows that bluescreens all the time for me. I finally figured out it was bugs in the USB-C/HDMI hotplugging drivers and I know how not to trigger it.
In contrast I haven't had an X-Windows crash in years and years. The closest thing I had was on a Dell XPS 13 and it wasn't a "crash" per se but the issue was very similar to the Windows issue I had with the MSI, the machine just totally locked up (ctrl-alt-backspace had no effect) when I plugged a USB-C to ethernet adapter into it. Unplug the adapter everything was fine.
On Windows I have to kill explorer.exe because the systray crashes at least once a week. (taskkill /f /im explorer.exe; start explorer.exe is always in my Powershell history.) And that's not the only example of more userspace instability, I could go on and on. There's instability I run into on Xfce but it's usually pretty isolated, this one component has a problem, it's not like Xfce is just randomly locking up all the time the way explorer.exe is.
For example I have this issue on Xfce with a logitech wireless keyboard which starts having input issues but unplugging the wireless dongle and plugging it back in fixes it. And that sort of thing happens with most peripherals on any platform.
And yet Windows was successful even when unstable. I would say that's not a hugely motivating factor for people. Many people turn on and off their computer when they want to use it and it's fine. Excessive stability isn't a major factor for most people and being perfectly stable wouldn't by itself make this the year of the Linux Desktop.
For this crowd it is important, of course, and I'd be deeply upset if my desktop crashed that often. That's not been my experience though.
> Familiarity, inertia and monopolistic competition
Also compatibility, marketing, and being installed out of the box.
In its heyday, Firefox had between 25% and 50% market share, depending on country. Compared to Internet Explorer 6, it was:
* More stable.
* Less buggy.
* More secure.
* More featureful.
* Usually faster than IE6, though it took a bit longer to start.
* Better development tools, which indirectly lead to web pages working better in Firefox than in IE6 even if that's not what management wanted.
* ActiveX, which is clearly the way Microsoft wanted you to make "web apps," had to be hidden behind scare prompts, so public web pages couldn't use it.
Here's Internet Explorer's big selling points, which allowed it to hold onto between 50 and 75% of the market share:
* It shipped with Windows.
Google Chrome only managed to unseat it by paying OEMs, which means that, de facto, Chrome now ships out of the box with Windows, too.
There is only one extension that ever really mattered: Firebug.
A lot of developers wound up using Firefox to develop web apps even when the web app was intended to be used in IE. That's important because it broke the vicious cycle where nobody used Firefox because nobody developed for Firefox and nobody developed for Firefox because nobody used it.
You might use other extensions, as do I, but this is such a deal-breaker that barely anything else matters. The browser that runs your dev tools will be able to run your app, obviously, so being the browser with the good dev tools is a tactic to make sure web pages will actually run in it.
The constant rhetoric around extensions is an irrelevant sideshow. It's the web pages that actually matter. Nobody would have bothered to use Firefox unless the majority of web pages they use work in it, and Firebug is the primary reason why so many web pages did work in it, along with Google strategically choosing to support Firefox as a way to hedge their bets against Microsoft making adversarial changes to IE to cut Google off.
There used to be an idea of "follow the alpha geeks", that while mass market adoption was a goal, a huge amount ofake or break came down to what the very good, very involved users were up to.
The relevance of extensions has somewhat faded, aside from ad blocking, byt back the day extensions were imo quite important for many, made the web a popular & more lively place. I'd have a number of extensuons telling me other related sites people were enjoying. I could annotate the page, chat about the page. I could easily save the page into my various bookmark services. Extensions made the web social, & helped it interconnect. Extensions interwove the web & drove cool at a critical time of adoption.
Whatever the state today, to ignore the historical relevance of how important extensions were for generating an active, enthusiastic user base, for differentiating: these are not to be ignored or casually cynically written off.
The question isn't really whether people use extensions. Obviously they use and enjoy them when they're there.
The question I'm asking is how Firefox manage to take 25% of Microsoft's market share, but Ubuntu didn't? Ubuntu has always been much more customizable than Windows, just like Firefox was much more customizable than IE. LibreOffice allows you to continue using the "classic" UI, just like how everyone on HN complains that Microsoft's Ribbon was a mistake, yet somehow LibreOffice remains at a partly 200 million to Microsoft Office's 1.2 billion. Why were customizable libpurple IM clients never as popular as the official ones?
This is why I think Firebug breaking the compatibility/market-share death spiral and inducing developers to make their webapps work in Firefox irrespective of Firefox's market share is so important. It's the biggest factor I know of that's actually different between Firefox's situation and the situation with so many other FOSS products.
> monopolistic competition, marketing, and being installed out of the box
This I believe, and an integral part of "marketing" is a FUD campaign against free alternatives.
> Familiarity, inertia, compatibility
I used to think they were relevant, but it doesn't seem to match reality. Android phones were neither familiar to users of older phones, nor were they compatible with software made for other phones.
Windows breaks familiarity with each major version for seemingly no reason. (Not compatibility, though.)
Apps and websites constantly get redesigned for no reason. Doesn't seem to drive people away.
At the turn of the millennium, the reason Linux desktops were supposedly not accepted by users was the lack of eye candy. Do phone apps today have "eye candy"? Bland-looking single-color rectangular surfaces, absolutely minimal interfaces, a few straight lines here and there and mostly text unless there's photos or videos to show. Looks like OpenBox.
The only common theme is marketing, marketing and marketing. A free, unencumbered platform will never be mainstream, neither on the desktop nor on the phone. Because that will always be an untapped market that one can get into with lying, manipulative marketing tactics and a free platform will have neither the resources nor the gall to employ the same shady tactics to defend itself, but a proprietary platform will, and then it doesn't matter who wins because the challenger is just the same.
Isn't Android effectively a base level linux distro optimized for smartphones though? Providing some level of uniformity between devices like package management, etc.?
The fragmentation issue is with user choice, and the amount of learning and experimentation you have to do to make all of those decisions. There's no de facto "probably the best for you" Linux flavor, or method of installing applications. Whereas on Android, generally if you don't know what Android to get, you get a Samsung. You want to install an app? No matter which phone you have, you tap the Play Store icon.
Plus there's large corporations selling these devices, and they set everything up to work for you out of the box. You're not gonna run into an equivalent of the experience setting up Linux desktop where it's not recognizing your monitor correctly and you have to mess with drivers and running mysterious commands in a terminal to make it work. Samsung made the phone and hardware work, turn it on, and you're good to go.
Android fragmentation != Desktop Linux fragmentation
They're completely different. Desktop Linux has APT, YUM, Snap, AppImage, FlatPak. Android has... APK. Desktop Linux has x11 vs wayland. Android has one renderer. ALSA/Pipewire/Pulse, what's the difference anyway? Will your favorite sound app work with all of them? Android has one sound API.
There seem to be two kinds of “Linux users”: those that constantly think about Windows, compare their operating system to Windows, and seem to think only “Linux” and “Windows” exists, and those that forget Windows exists and mostly compare different Unix flavors with one another.
Theo de Raadt had some merit when he said that many “Linux users” seem to talk more about Windows than anything else and don't really like Unix as much as that they hate Windows.
But I also find that the latter group rarely even mentions “Linux” that much, which mostly seems to be used as “not-Windows”.
Why would I compare Linux distros? You can literally turn any one of them into another, all the packages are there (prebuilt or source). With some exceptions, of course.
I had ALL the Desktop Environments I could find installed at one point and was switching between them every other day. Compiz effects were the shizz! But I don't have the time/patience/motivation for that now :D
I think most comparisons between Linux distros comes down to defaults/the way it operates out of the box. Sure, both Ubuntu 22.04 and Fedora 36 use GNOME 42 (is that the version they're on nowaday?), but Ubuntu has a drastically different UX (dock on the side, panel-applets, accent color & custom theme, the list goes on) compared to Fedora. Additionally, as an NVIDIA user, I like how easy and simple it is to install the NVIDIA drivers while still maintaining secure boot (something Fedora has neglected / informed their users to 'just disable secure boot').
yeah but microsoft/apple/google are sole owners of the interface and dictate the one interface. there is no such thing in foss, which is a good thing. if gnome 3 was somehow forced to be the gui on linux, I would absolutely stop using linux.
Much of the Linux desktop is balkanized and fragmentation is indeed a huge issue. People's obsession with choice is irritating given that doing more or less the same thing in 10 different ways half-right is worse than doing it once correctly. The entire distro-zoo built on top of the same distributions with different DE's on top of it is an example. The problem with Windows or Mac isn't that it looks, feels and works the same everywhere. That's what most people want, they don't want to spend hours tinkering with their UIs, they want to get work done.
It's not even right to compare Linux to Windows, or maybe even call Linux distros operating systems in a strong sense because they don't provide genuinely stable interfaces. People often remark on alleged benefits of open source software such as compatibility, yet it's Windows where I can click on a 20 year old executable and it runs, while I don't know if anything that has been built 2 Ubuntu LTS releases ago still works.
It's funny that one of the things that has actually driven Linux adoption is Proton. As one blogger a few months ago remarked, Win32 is basically the stable Linux userland ABI.
To add to this, I think Linus himself mentioned this issue at a talk at a Debian conference.
I think he even mentioned maybe gaming would save Linux from this, since games will not constantly update to fix compatibility issues so distros would be forced to behave.
You want to use Gnome? Use Gnome. You want to use KDE? Use KDE. I3? ditto. Etc.
Oh, some random GTK app will look weird on KDE? Tough. Every Windows app looks different, and even MS's apps always felt out of place on their own OS. And with the proliferation of Electron apps, nothing ever looks "native" anymore.
> while I don't know if anything that has been built 2 Ubuntu LTS releases ago still works
Maybe, maybe not. But that's also an issue on MacOS. And while Windows prides itself on backwards compatibility (and mostly delivers), that doesn't come free.
There's also hardware backwards compatibility. Many devices that used to work with older versions of Windows or MacOS just don't work anymore. Even those that never required custom drivers from their manufacturer. On Linux, if it works once, it'll usually keep on working. I have such an issue: a PCIe USB 3 card, so it isn't even something that old. Used to work with MS Win 10 provided drivers, now (win 11) it's unusable.
>You want to use Gnome? Use Gnome. You want to use KDE? Use KDE. I3? ditto. Etc.
To play devil's advocate:
What do those even mean and why should I care? I want to view my files and folders and click a shortcut to run a program and install steam. And I want that process to be stable and smooth and be both logically transparent and consistent.
There's too much I have to critically compare and understand in my life to want to add 'desktop GUI flavor' to the list.
That doesn't contradict my point. If you don't care, just grab some distro that seems reasonably prevalent, like Ubuntu or whatever. They have spins or flavors or whatever they're called, but on the website you don't exactly have to make a choice. There's a big "download" button that gives you the "recommended" choice. Install that.
You'll be able to view your files and folders and click a shortcut to run a program. And that process will be quite stable and smooth, I'd argue even more than on Windows, since you can actually find useful apps on the "store", as opposed to having to hunt on random websites.
If you want consistency, presumably over time, then yeah, you may not get that. But neither will you on Windows, who've changed the explorer in Windows 10 and now again in Windows 11. The explorer-like app that seems to have been the most stable is Finder. But then macOS changed in other ways...
It seems you can't really win at this, regardless of your OS of choice... So it think that's not fair to only throw Linux under the bus for this, while praising the others.
> That's what most people want, they don't want to spend hours tinkering with their UIs, they want to get work done.
The people most capable of fixing the problem you describe are the people who like spending hours tinkering with their Linux UIs.
Besides, Windows usability is rubbish, there's a reason why the first thing that I, and, uh, a few other weirdos do on a new Windows install, is set up an orthodox file manager. [1]
10 different ways half-right is worse than doing it once correctly. The problem with Windows or Mac isn't that it looks, feels and works the same everywhere
what's correct for you is not necesarily what's correct for me.
the problem with windows/mac is very much that it doesn't work the way i want it to work. why even have multiple oerating systems at all, if we don't need choice? we could all be using windows.
i don't see how removing choice solves anything at all.
There is no reason there can’t be one agreed upon blessed way of doing things in Linux while still enabling users like you to have choice. Heavily enforced defaults are important for growth.
Yes, let's search for the One True Font, One True Theme, and Unicorns while we're at it. Oh, let's also pretend that we'll give users a choice. Well, technically, we will but we will make it "unsupported" and call it a hack.
Wait, I think that reminds me of a desktop environment.
You can have as much snark as you want. I specifically said that Linux should remain the platform of choice, just that the defaults should be stronger so that it's a more consistent and approachable platform for people.
GNOME has made great advancements in this regard, which has been awesome to see.
> GNOME has made great advancements in this regard, which has been awesome to see.
I wasn't joking or being hyperbolic in my previous post. GNOME may have good defaults in certain cases but it ends there. User choice and customization are seen as design defects and anti-features. They really do believe that there is a One True Font and One True Theme out there and anything else is unsupported.
Sane defaults shouldn't come at the cost of user choice. If they do, there's little reason to use Linux over a Mac because the former essentially becomes Android style "look but don't touch" open source.
Some people don't want to configure everything about their desktop. Most people don't. They just want something that looks nice and feels good to use. Apple's ecosystem is expensive and locked in while gnome/linux is open and works with as much hardware and software as possible, on top of that it's free. Also the software experience is different too, while on Mac software that involves any personalization or enhances the desktop is paid, while gnome/linux is free and almost all software that enhances the desktop is also free and open source software. Their opinions on UI/UX might be similar but that's where the majority of their similarities end.
> Some people don't want to configure everything about their desktop. Most people don't. They just want something that looks nice and feels good to use.
And that would be completely fine were it not for parent posts which imply that GNOME should be the default and the only experience on the Linux desktop and everything else is irrelevant fragmentation. I respect users choice to use GNOME but I don't appreciate when people call everything else besides GNOME (or anything really) as fragmentation.
For example, here's one of the parent posts
> There is no reason there can’t be one agreed upon blessed way of doing things in Linux while still enabling users like you to have choice.
>And that would be completely fine were it not for parent posts which imply that GNOME should be the default and the only experience on the Linux desktop and everything else is irrelevant fragmentation.
To be ultra clear here: I'm not saying it has to be GNOME. I legitimately only said GNOME to toss you a bone when you responded with a needlessly sarcastic comment. ;P
It could be any DE, even a newly created one.
No part of your posts feels productive, and to be honest they feel like emotional backlash.
> but I don't appreciate when people call everything else besides GNOME (or *anything really*) as fragmentation.
The idea of "one agreed upon blessed way of doing things" implies that a DE can be made to work for everyone, which is a false assumption. Just because Windows and Mac follow this philosophy, doesn't mean that it's right. Diversity and independence of thought isn't irrelevant fragmentation.
I think it would help a lot if it were the norm for DEs to come with a bunch of prefab configurations right out of the box, which the community could then add to. It'd be cool if configs could be exported and shared too.
That would give many distros much less of a reason to exist, and could instead be configuration profiles for GNOME, KDE, etc that could easily be loaded on -any- distro. It would also allow the configurations to be iterated on independent of distros which would eventually lead to much higher levels of polish.
I think it would help to get an honest ranking of active user or community size for different desktop Linux distros. Half the strength of Linux isn't the nerdy details, it's the community and packaging chops.
The fragmentation peeves me as well. The Linux community is like herding cats with fragile egos. One disagreement and a dev will go try and make their own thing, further diluting the Linux ecosystem.
There are way too many distros that are half baked trash that make it into bad listicles returned in search results.
Stop trying to herd people? I didn't sign up for whatever agenda you're trying to advance, so of course I'm going to spit in your face when you try to tell me what to do.
Proudly. If your agenda is to take away my customization and make the Linux desktop into something with broad mainstream appeal, then I proudly sabotage that.
Whatever your agenda is, I didn't sign up for it. And if you truly have no agenda at all, then you have no business trying to 'herd' people. Either way, pound sand.
It would be a depressing parallel universe where some God would force us Linux users to all run the exact same thing, no customization, have updates forced on us and have the latest GUI fad change our workflow yearly.
As Kubuntu LTS users I don't feel that I am losing anything if someone else uses Arch with GNOME. Also there are distro made for children with custom DEs, distros for blind people, distros for media etc.
The diversity is not the problem, the problem might be some secondary side effects like some developer really wants to use some bleeding edge thing but he does not have an easy way to deploy it to more stable distros.
Putting the whole community's wood behind one arrow would be a great idea - if it was the right arrow. But the big teams of desktop developers have repeatedly run off and built things that they thought were shit hot, but were not what people actually needed to be productive (Unity, Gnome 3). It's only the "fragmentation" that has meant Linux hasn't ended up like Windows 11.
> The problem with Windows or Mac isn't that it looks, feels and works the same everywhere.
Did you use Win 10 ? One GUI in Teams, another in Office 365, a calculator which comes from outher space, 2 interfaces for configuration, disable microfon for admin account -> no microfone for user account.
People don't care that TikTok looks different than Instagram looks different than Gmail looks different than NYTimes.com looks different than a flight reservation site.
Linux on the desktop. Where old Mac HW goes to retire when it is unloved by Apple. The rate at which macOS Ventura ruthlessly cut off 2016 and 2017 Macs shows the need more than ever for Aishi Linux.
Pay attention. The “ruthless” cut off is in line with how the PowerPC to Intel transition happened. It’s also following the 68K to PowerPC transition. It’s like ripping off a bandaid.
I mean in 2022, there are so many options in Linux that may not be perfect, but are good enough. Be it KDE,Gnome,Wayland,Xorg,systemd,initd,bash,zsh they are fine to use. Yeah you'll hit a corner case here or there, but most people will find a solution that fits their workflow. Windows on the other hand will always be the Gamer or enduser-that-don't-care OS, and that's fine too. I'm incredibly thankful that Linux exists, because I personally gained so much from it, but I understand as well when people don't care about it. Let them use Windows and deal with their problems (but don't expect me to fix them for you, thanks ;)).
I've been using linux as my primary OS for over 20 years now. The rough edges are almost entirely gone and linux is absolutely a fantastic OS for everything but gaming, and even there it's getting better.
I'm building a gaming computer at the end of this year and I'm strongly considering just throwing linux on it instead of windows as well, because if protondb is to be believed most games run perfectly fine on linux these days.
even gaming. it's amazing how much progress proton/wine is making. i have several games that i didn't dare run on linux a few years ago, that now just run out of the box with wine as if they belonged.
Let's forget for second that iOS (until 2015), Android, Switch and PlayStation have some POSIX like support and with exception of PS, all support OpenGL in some form, yet most of those studios couldn't be bothered.
fair point. but it truly is progress, because running windows apps on a linux system results in a more stable desktop experience. for one, i don't have to reboot to recover from a windows app that somehow locks the system.
i see a potential future where microsoft will port windows to run on top of a unix kernel. small chance, but not impossible.
those studios won't bother with native linux support until a sizeable amount of users play their games on wine. and that numer is growing. which is the point.
The section "A better DOS than DOS, a better Windows than Windows" is worth a close look and introspection about what WINE/Proton effectively bring to the table.
thank you, that's quite an interesting article. however i don't agree that the lessons of OS/2 apply to linux.
for one, linux does not have a single business interest driving it, so many of the OS/2 issues relating to IBM don't apply.
being a better windows than windows bolstered windows apps, but only killed OS/2 because native OS/2 was worse than native windows. that too does not apply to linux.
OS/2 died because IBM expected it to be backed by microsoft, and gave up when that didn't happen and they realised that it could not compete. linux never had this issue. linux won't loose its backing. linux never depended on being able to compete against windows. microsoft is now supporting linux more than it ever supported OS/2.
so at the worst, the status quo remains, and nothing changes. realistically though, wine will keep getting better, enabling some people to use it. wine will continue to be developed and supported. there is no risk of that ever stopping. there is no danger betting on wine, because once it works, it will continue to work indefinitely. at worst wine may fall behind implementing some new windows feature that new apps need, but it won't stop supporting what already runs.
wines user share is not going to go down. linux and wine development will not stop. unlike IBM we are able to play the long game here. there is nothing to loose. there are no risks to continue as we are.
I strongly disagree. Normal users now do most of their computing on a web browser, making most Linux desktop issues less critical to adoption.
My personal experience in the last ~5 years or so is increasingly reduced friction in adopting a Linux desktop for non-technical users as companies go towards SaaS.
Firefox ends up being simply a differently shaped search bar.
Holy crap I never heard of the term "DaaS" before. I wish I could go back in time :(
I guess I am thankful that Linux desktops in general I both love and hate, are good enough to replace Windows if the needs come to that.
Even more so nowadays with Valve setting up Linux as a promising "pc gamer" platform with decent usability.
For good or worse the fragmentation is what Linux "fans" want. The kind of users who are never happy and want gazillions options to make something fit their every perceived special needs. The kind of users who think using a mouse is "less productive and efficient" than keyboard only… these are the most vocal users on Linux discussions it seems.
Oh well like I said I’m grateful Ubuntu is usable enough and looks half decent… actually moving back from a M1 MacMini soon… but mostly for convenience as I want to keep using my "gaming pc".. and I realized i dont actually do anything very macOS specific … plus homebrew and lack of native containers feels like a hack. I’llmiss the polish of macOS though.
While I’m at it let’s be blunt: is there really any half decent looking distro besides Ubuntu? Imho most other distros look like they were designed in 1987, and elementaryOS is such a joke. Personally I could’t see myself using anything else than Ubuntu + the fantastic "Dash to Panel" addon as a valid replacement for Win10 (the win + number shortcuts are amazing).
I used to use Linux desktop exclusively but mental and time cost of maintaining it just was just too high and eventually gave up.
It actually works great when it does, but more often than not something fails after updates. So I felt like a rolling a dice everytime I have to go though this.
There were problems with update process itself -- they tend to wait for users to confirm configuration overwrite/update/patch so when I attempted to perform update overnight, I often find it waiting for my input in the morning 30 percent into update process.
Ultimately I had enough of the frustration that I've stopped using it.
Where I actually need Linux, I can just use WSL. Windows has its fair share of the problems but I'm definitely not spending as much as time just to keep things running even through major updates.
The main problem of Linux on desktop is a severe lack of quality control and questionable usability pushed by people not using desktop and running mainly CLI. The issues addressed on desktops are not in correct priority, you should not polish "happy path" in the generalized public system. Corporation can easily work with "happy path"-only software because they can afford to train personnel, and they can afford to fix mistakes in systemic way. Generalized public system should not polish what is mostly working. It should eliminate low occurrence catastrophic failures, because when scaled to ten of millions those low probability failures tend to multiply a lot.
That's why I use console Linux every work day and enjoy it, and that's why I wouldn't touch Linux on a desktop with ten foot pole, because all my experiments in that area failed sooner or later. Also (surprise) shell Linux knowledge almost doesn't scale to the debugging GPU drivers, DE and X server issues. This is actually rather disconcerting feeling, when you have a decade of Linux experience but you look at crashing desktop GUI and understand that you have approximately as much understanding of the problem as a first time user.
> Also (surprise) shell Linux knowledge almost doesn't scale to the debugging GPU drivers, DE and X server issues.
You will be surprised on how much Wayland solved those issues. At least, in my experience, understanding why Sway crashed is a debug flag away. Unfortunately, Wayland is a double edge sword, even more if you want to run games with multiple monitors (but that is a pain even on Windows in my experience).
Also I want to add a somewhat representative anecdote story.
Preface - my corp issues same laptops to all engineers, including developers actually writing code. Codebase is 100% for Linux, we don't need write anything Windows related. Developers have complete freedom at what to use as a host system on the laptop, there are some guides to help set up host Linux, there are people who already did it and can give advise.
Around half of them just opt to host Windows 10 and run all their tools in the Linux virtual machine. That's a group of people who are second in world most suited to run Linux - actual Linux application devs. (first group would be Linux admins I think)
I remember when I upgraded Ubuntu 14 to Ubuntu 16 and the GUI stopped working and was dumped to the command line. Luckily I had a Linux guru around to fix it, because otherwise my only alternative would have been to reinstall it - seeing how many esoteric steps he did to debug and fix it there is no way I could have googled my way out of it.
I don’t think Android is linux desktop, even not a linux mobile.
Yes indeed, there is so called kernel there, but it wrapped with custom bootloader, custom apis, custom desktop and set of apps developed and managed by single corporation.
Android of course inherit some of linux legacy - instability, inefficiency, inflexibility, tivolization.
But half-bakeness, because we say so approach and 70s legacy are not there, due to market competition. When the corporation replace linux/drivers firmware blob with their own, Android will remain Android, and not Fuchisa Desktop.
It's not just that fragmentation results in lack of polish. Having many distros means that when I want to figure out how to do something obscure, or encounter an issue, there's less likely to be someone on my distro with the same problem. But more importantly, I often don't get any good results searching "*issue* *distro*", because the issue is at a different layer in the stack. I have to know whether it is a distro issue, desktop environment issue, kernel issue, systemd issue, X11 issue, etc, which is difficult for new users. Compared to Windows and Mac which aren't modular in this way and therefore don't have this problem.
I think all the hand-wringing over fragmentation is misguided. Diversity is good. We have plenty of polished UX provided by big tech, but it comes with intractable downsides related to centralized control. Free and open source software provides a different tradeoff and value proposition, one that counteracts the natural forces towards market consolidation and centralized power that exist for software in a globally networked world. Yes, that means we have to work harder to motivate people to cooperate, but such is the cost of freedom.
171 comments
[ 2.9 ms ] story [ 188 ms ] threadThe hope is that ChromeOS dies. It has nothing to do with Linux except the kernel.
Which isn't entirely bad since it has some good concepts regarding security that will make good additions to current Linux systems.
It is exceedingly difficult for a developer to contribute any kind of improvement to this because the bug tracker is a black hole and the source code is a gargantuan monorepo with multi-page build instructions that basically require you to sacrifice a computer if you want to get anywhere.
Sandboxing is great and all, but the thing is a glorified hypervisor. It has no vision of its own for app distribution, user experience, or developer experience. It boggles my mind that they are doing this with ChromeOS: the OS that runs on the cheapest, slowest, most under-specced computers on the market.
I don't care about your market share.. I care about the quality of my own experience.
That was the common sentiment 20+ years ago when Steve Ballmer, then Microsoft CEO, declared that "Linux is a cancer". Today I'd say it's not the case anymore, especially so since they were forced by market laws to fake some form of friendship with Open Source, not to mention having ruined Windows with all versions from 8 on. Windows is pretty much crushing itself these days.
Exactly this. I've seen where chasing mainstream usability and appeal leads: ipadification. If you want an ipad go buy an ipad. I don't care about linux 'winning' because that sort of win condition would make linux into something that has no value to me. A Pyrrhic victory.
Fedora Workstation is nice, not perfect, but it's nice. Tried the KDE spin as well as the usual GNOME one. It's my favourite Linux desktop distro FWIW. PipeWire's amazing.
One reason Windows and macOS are popular is that people know what they are going to get. If you've used one Windows 11 box, for example, you can get on another Windows 11 box and figure out how to use it. It will have the same user interface and window management as the other: the Start menu, the control panel, the desktop, etc.
That's just not the case for Linux. There are several popular desktop environments that are very different from each other in look and feel. For an ordinary user, switching from Gnome to KDE would be quite jarring and it would take them a while to get used to it. For power users and tweakers, that diversity of options is an attractive feature of the Linux desktop. For ordinary users, it is a deterrent.
Is it for every power user? No, of course not.
Can Linux be made very simple for casuals? Yup, absolutely, if all people want to do is browse the web, send some emails, the usual stuff, it's fine. The in-between of "can I use Photoshop?" and the answer being "yes, but" isn't going to fly with most folks.
I use it for what it's best at, and use Windows for what it isn't best at, sometimes as a bottle, other times not. But it takes experience to know when to use what, which is what most users don't care to figure out for themselves as it involves trial and error usually. Most folks just want to do their work, and their phone or tablet is the "funsies".
And no, I don't want to hear about the software XYZ that is almost just like any of the above, because they aren't even close.
If I just wanted to browse the web, and work in a terminal, then linux is great.
Or edit video, or make music, or write all the stuff that makes this possible.
How many distros would they need to support to make customers happy? At what point does it become a problem of diminishing returns?
Windows and macOS are pretty big. How many more customers would Adobe get by adding support for: Ubuntu, Fedora, SUSE, Mint, Arch, Manjaro, MX, Gentoo, Elementary, and so forth? At some point it just gets silly and devolves into Muppet names and other nonsense -- and each one has a smaller amount of users.
Plus, real professionals are willing to pay for software. Open source is frankly not relevant for professional image editing, video editing, music production, CAD, etc.
Yes, exactly. It's the same problem all over, too - look at transport! You shouldn't ride a bike, you should drive a massive V8-powered 4x4 like I do. It meets my needs, so it must meet yours, right?
No company is fully focused on the desktop OS, and I don't see why they should. This article is contradictory saying "desktop operating system will lie in the hands of Apple with macOS and us with Linux". It is not because Microsoft prioritize cloud services that it abandons desktops, it is not a priority for many Linux developers and sponsors as stated above in the article, and it doesn't seem a high priority for Apple either.
Tbh, I only tried it on a laptop and it was hell, it ended up needing a mountain of configuration and software just to pass an nVidia GPU (fortunately a Quadro, which is slighly easier) to Windows, which has the software that needs it the most.
Is it easier on desktop computers with an IGP and a discrete card, or two graphics cards? Or do you still need to fuck around with the VBIOS, fight with power management and hybrid modes (I guess not?), use Looking Glass/RDP/dummy display/DP/HDMI plug and a ton of configuration that sometimes fails after updates?
Guess I will try it again on my next computer, which must be a powerful Ryzen build.
Unlike Linux. A few wrong clicks and there goes the GUI.
But... I also want things like hibernate to just work, screen tearing to not happen on X11, etc etc.
The polish is absolutely not there, and I agree w/ the article that's a result of the fragmented ecosystem. If the Linux community could agree to invest in the underlying APIs for desktop stuff, it seems we'd be in a much better place and still have lots of customization options on top.
Of course, I understand that's a huge coordination challenge.
Most things on Linux are actually like this, where everything is in the kernel or some other small project and the DE just adds a wrapper GUI (or sometimes a management deamon) and really has absolutely nothing to do with how the thing works.
And thank goodness for that because all of the popular DEs use terrible window managers. I wouldn't bother with X11 if those were the only functional choices.
Also some people (like me) really don't mind tearing but can't stand an extra frame of latency when we're typing. It's nice to have options since no one agrees on everything.
My point is suspend states aren't important to big commercial users who are running linux servers. They have no reason to contribute to making it work well. Hence, it likely doesn't get much attention compared to many other parts of the kernel.
It doesn't have to do with "Desktop environments", but it very much is part of the "linux desktop" experience.
If you have AMD hardware that uses the 'amdgpu' driver, then -IME- enabling the TearFree option seems to completely eliminate tearing. (Though, I am using KDE, which _also_ has a "Make the compositor use vsync" option... which didn't actually eliminate tearing, when used by itself. But, there _may_ be confounders.)
Anyway. Drop this into a file named "$WHATEVER.conf" in '/etc/X11/xorg.conf.d' (or into your xorg.conf), and you should be good to go:
Alternatively, you could enable the setting with the 'xrandr' tool: but that only sticks for as long as the affected X server stays up.It's probably simply because others all call them “Linux”, but they rarely even call themselves that. One cannot easily find the word “Linux” on the websites of “Fedora” or “Ubuntu” for instance and they simply call themselves that.
The obstacle to switching to linux for me is not too many distributions, it is that I have to relearn everything, from my way to the filesystem, how drivers and networking work, virtualisation, web servers, etc. And linux is many things, but intuitive and self discoverable it is not. I just don’t have the time for starting from scratch (and it’s not related to my day job either), until windows becomes really user hostile (I see it coming but not yet there).
You are a boiled frog.
How do you think companies were using X Windows?
By attaching thin clients to a server that was still under their own control, as opposed to making the whole company a serf to Microsoft.
> ... intuitive and self discoverable it is not.
As a multi-decade Linux coder, I fully concur.
The other thing is that in all my years of using Windows and MacOS, I have NEVER had an issue where an upgrade borked being able to log into the desktop. This has happened more times than I remember on Linux.
You're going to need to explain that one.
https://askubuntu.com/questions/360748/how-do-i-move-the-tas...
"There is no way to move the 'task bar' to the bottom of the screen. You are out of luck."
You ...do know there are countless desktop environments and window managers just an apt-get away right? If that's too much for you, you could have literally chosen any other flavor. My mother has been happily using xubuntu for years. If she can manage, I think you can too.
Has that happened to you? Is that still a problem?
And yeah, the way settings work on Desktop Linux can lead to all kinds of weird stuff if you start actively using multiple environments. Should changing your keyboard layout in one DE affect all the others? Well, wait, which of multiple ways did you accomplish that change in keyboard layout? Should it affect bare terminals, outside the window server? Should some of those methods affect all of them, but other methods not? Which one "wins"? Was this other thing a DE setting, or an X-Window setting? Or neither? Or was it Wayland—wait a minute, some of these DEs run under Wayland and some under X-Window! Gee, surely that won't cause any settings conflicts or surprises.
You can read the rest of what I posted instead of assuming I never did any of those things.
>I hate it, which is why I don't use it. Ubuntu got the closest for me, but the owner's refusal to allow users to move that taskbar to the bottom of the screen made me give up on it altogether. And unless it's a supported distro, I don't want it. Been burned too many times by pet projects that end up being abandoned. The other thing is that in all my years of using Windows and MacOS, I have NEVER had an issue where an upgrade borked being able to log into the desktop. This has happened more times than I remember on Linux.
And for the record, I've been using desktop Linux off and on since before Slackware was released. I simply no longer have any interest in making yet another attempt at meeting my day to day needs with such terrible products. If I really wanted to use a Linux gui app that bad, I'd just dial it up in the Windows 11 hypervisor.
The other day, my Windows 10's explorer.exe crashed when the file manager tried to load a directory, which is something that used to happen in Vista and hasn't happened in KDE for years IME.
This is very far from my experience. I installed OpenSUSE Tumbleweed (recommended as one of the best KDE distros) at the beginning of 2022 and ran into a dozen KDE Plasma bugs within an hour. Stuff like panel widths resizing just because I clicked "customize," or widgets vanishing when I tried to drag them onto panels, or the bluetooth icon vanishing from the status bar until I restarted my session, or even the settings manager freezing up.
The basic desktop worked well enough, but as soon as I started customizing it I ran into all sorts of paper-cut edge-case type bugs.
>Stuff like panel widths resizing just because I clicked "customize,"
This can be finicky, I can confirm.
>widgets vanishing when I tried to drag them onto panels
Widgets are also finicky. I don't use them. But they're niche KDE things that you don't really have on other DEs anyway, so I don't count it as diminishing KDE as compared to other DEs. :p
I've never used OpenSUSE, but, for what it's worth, I had tons of KDE crashes years ago when I used Ubuntu. I never had any on Manjaro or Arch, and I've used them for a few year now. My diagnosis there is that the rolling release model is paradoxically better than the point release model in terms of DE stability. I think that's the reason that KDE Neon is a thing (rolling DE updates atop regular Ubuntu). Tumbleweed is apparently a rolling release, though, so Iunno.
However, I always hear people sing the praises of its stability.
Is there a good theme/font/icon pack that you like that makes it look more polished without a lot of fussing on the end users' part?
But the DE, feature-wise, is so great compared to the others that it's like picking between a Volvo, a Model T, and several bicycles for my daily drive. The finish on the Volvo's dashboard is the least of my concerns; I'm just happy it's got AC.
So it seems fragmentation has many problems but that alone won't prevent people from using a device. I think the author believes that a critical mass of Linux desktop users will create enough interest in polish and solving problems for end users, so therefore believes if everybody was using the same desktop there would be more polish (or something).
I would counter that the desktops have plenty of polish. He says he loves the Linux Desktop and doesn't offer much criticism besides there's too much choice. Nah, they're already great and we don't need to abandon great desktops.
Instead, I believe other factors are at play. Familiarity, inertia and monopolistic competition to name a few. Those are powerful factors and much more so than abstract fragmentation.
"Look the camera has more megapixels than an iPhone!"
Yeah ok, but the stock camera software outputs photos that look worse than an entry level iPhone.
Even before that, it nearly matched it in instability, if you counted X-Window crashes as the same as a bluescreen or kernel panic (and if you're doing GUI work, they're close to equivalent)
The kernel might be OK. The GUI & multimedia stacks aren't.
In contrast I haven't had an X-Windows crash in years and years. The closest thing I had was on a Dell XPS 13 and it wasn't a "crash" per se but the issue was very similar to the Windows issue I had with the MSI, the machine just totally locked up (ctrl-alt-backspace had no effect) when I plugged a USB-C to ethernet adapter into it. Unplug the adapter everything was fine.
On Windows I have to kill explorer.exe because the systray crashes at least once a week. (taskkill /f /im explorer.exe; start explorer.exe is always in my Powershell history.) And that's not the only example of more userspace instability, I could go on and on. There's instability I run into on Xfce but it's usually pretty isolated, this one component has a problem, it's not like Xfce is just randomly locking up all the time the way explorer.exe is.
For example I have this issue on Xfce with a logitech wireless keyboard which starts having input issues but unplugging the wireless dongle and plugging it back in fixes it. And that sort of thing happens with most peripherals on any platform.
For this crowd it is important, of course, and I'd be deeply upset if my desktop crashed that often. That's not been my experience though.
Everybody had a friend who knew how to fix Windows.
And it was a literally a job, there were fliers glued to lamp posts advertising "I install/fix/devirus Windows for a small fee".
Also, during this time of unstable Windows, Linux was even worse. It wasn't the nice Linux desktop of today.
Also compatibility, marketing, and being installed out of the box.
In its heyday, Firefox had between 25% and 50% market share, depending on country. Compared to Internet Explorer 6, it was:
* More stable.
* Less buggy.
* More secure.
* More featureful.
* Usually faster than IE6, though it took a bit longer to start.
* Better development tools, which indirectly lead to web pages working better in Firefox than in IE6 even if that's not what management wanted.
* ActiveX, which is clearly the way Microsoft wanted you to make "web apps," had to be hidden behind scare prompts, so public web pages couldn't use it.
Here's Internet Explorer's big selling points, which allowed it to hold onto between 50 and 75% of the market share:
* It shipped with Windows.
Google Chrome only managed to unseat it by paying OEMs, which means that, de facto, Chrome now ships out of the box with Windows, too.
A lot of developers wound up using Firefox to develop web apps even when the web app was intended to be used in IE. That's important because it broke the vicious cycle where nobody used Firefox because nobody developed for Firefox and nobody developed for Firefox because nobody used it.
You might use other extensions, as do I, but this is such a deal-breaker that barely anything else matters. The browser that runs your dev tools will be able to run your app, obviously, so being the browser with the good dev tools is a tactic to make sure web pages will actually run in it.
The constant rhetoric around extensions is an irrelevant sideshow. It's the web pages that actually matter. Nobody would have bothered to use Firefox unless the majority of web pages they use work in it, and Firebug is the primary reason why so many web pages did work in it, along with Google strategically choosing to support Firefox as a way to hedge their bets against Microsoft making adversarial changes to IE to cut Google off.
The relevance of extensions has somewhat faded, aside from ad blocking, byt back the day extensions were imo quite important for many, made the web a popular & more lively place. I'd have a number of extensuons telling me other related sites people were enjoying. I could annotate the page, chat about the page. I could easily save the page into my various bookmark services. Extensions made the web social, & helped it interconnect. Extensions interwove the web & drove cool at a critical time of adoption.
Whatever the state today, to ignore the historical relevance of how important extensions were for generating an active, enthusiastic user base, for differentiating: these are not to be ignored or casually cynically written off.
The question I'm asking is how Firefox manage to take 25% of Microsoft's market share, but Ubuntu didn't? Ubuntu has always been much more customizable than Windows, just like Firefox was much more customizable than IE. LibreOffice allows you to continue using the "classic" UI, just like how everyone on HN complains that Microsoft's Ribbon was a mistake, yet somehow LibreOffice remains at a partly 200 million to Microsoft Office's 1.2 billion. Why were customizable libpurple IM clients never as popular as the official ones?
This is why I think Firebug breaking the compatibility/market-share death spiral and inducing developers to make their webapps work in Firefox irrespective of Firefox's market share is so important. It's the biggest factor I know of that's actually different between Firefox's situation and the situation with so many other FOSS products.
This I believe, and an integral part of "marketing" is a FUD campaign against free alternatives.
> Familiarity, inertia, compatibility
I used to think they were relevant, but it doesn't seem to match reality. Android phones were neither familiar to users of older phones, nor were they compatible with software made for other phones.
Windows breaks familiarity with each major version for seemingly no reason. (Not compatibility, though.)
Apps and websites constantly get redesigned for no reason. Doesn't seem to drive people away.
At the turn of the millennium, the reason Linux desktops were supposedly not accepted by users was the lack of eye candy. Do phone apps today have "eye candy"? Bland-looking single-color rectangular surfaces, absolutely minimal interfaces, a few straight lines here and there and mostly text unless there's photos or videos to show. Looks like OpenBox.
The only common theme is marketing, marketing and marketing. A free, unencumbered platform will never be mainstream, neither on the desktop nor on the phone. Because that will always be an untapped market that one can get into with lying, manipulative marketing tactics and a free platform will have neither the resources nor the gall to employ the same shady tactics to defend itself, but a proprietary platform will, and then it doesn't matter who wins because the challenger is just the same.
Plus there's large corporations selling these devices, and they set everything up to work for you out of the box. You're not gonna run into an equivalent of the experience setting up Linux desktop where it's not recognizing your monitor correctly and you have to mess with drivers and running mysterious commands in a terminal to make it work. Samsung made the phone and hardware work, turn it on, and you're good to go.
They're completely different. Desktop Linux has APT, YUM, Snap, AppImage, FlatPak. Android has... APK. Desktop Linux has x11 vs wayland. Android has one renderer. ALSA/Pipewire/Pulse, what's the difference anyway? Will your favorite sound app work with all of them? Android has one sound API.
Theo de Raadt had some merit when he said that many “Linux users” seem to talk more about Windows than anything else and don't really like Unix as much as that they hate Windows.
But I also find that the latter group rarely even mentions “Linux” that much, which mostly seems to be used as “not-Windows”.
This article too is mostly about Windows.
I had ALL the Desktop Environments I could find installed at one point and was switching between them every other day. Compiz effects were the shizz! But I don't have the time/patience/motivation for that now :D
It's not even right to compare Linux to Windows, or maybe even call Linux distros operating systems in a strong sense because they don't provide genuinely stable interfaces. People often remark on alleged benefits of open source software such as compatibility, yet it's Windows where I can click on a 20 year old executable and it runs, while I don't know if anything that has been built 2 Ubuntu LTS releases ago still works.
It's funny that one of the things that has actually driven Linux adoption is Proton. As one blogger a few months ago remarked, Win32 is basically the stable Linux userland ABI.
https://sporks.space/2022/02/27/win32-is-the-stable-linux-us...
I think he even mentioned maybe gaming would save Linux from this, since games will not constantly update to fix compatibility issues so distros would be forced to behave.
You want to use Gnome? Use Gnome. You want to use KDE? Use KDE. I3? ditto. Etc.
Oh, some random GTK app will look weird on KDE? Tough. Every Windows app looks different, and even MS's apps always felt out of place on their own OS. And with the proliferation of Electron apps, nothing ever looks "native" anymore.
> while I don't know if anything that has been built 2 Ubuntu LTS releases ago still works
Maybe, maybe not. But that's also an issue on MacOS. And while Windows prides itself on backwards compatibility (and mostly delivers), that doesn't come free.
There's also hardware backwards compatibility. Many devices that used to work with older versions of Windows or MacOS just don't work anymore. Even those that never required custom drivers from their manufacturer. On Linux, if it works once, it'll usually keep on working. I have such an issue: a PCIe USB 3 card, so it isn't even something that old. Used to work with MS Win 10 provided drivers, now (win 11) it's unusable.
To play devil's advocate:
What do those even mean and why should I care? I want to view my files and folders and click a shortcut to run a program and install steam. And I want that process to be stable and smooth and be both logically transparent and consistent.
There's too much I have to critically compare and understand in my life to want to add 'desktop GUI flavor' to the list.
You'll be able to view your files and folders and click a shortcut to run a program. And that process will be quite stable and smooth, I'd argue even more than on Windows, since you can actually find useful apps on the "store", as opposed to having to hunt on random websites.
If you want consistency, presumably over time, then yeah, you may not get that. But neither will you on Windows, who've changed the explorer in Windows 10 and now again in Windows 11. The explorer-like app that seems to have been the most stable is Finder. But then macOS changed in other ways...
It seems you can't really win at this, regardless of your OS of choice... So it think that's not fair to only throw Linux under the bus for this, while praising the others.
Dolphin has been basically the same for 16 years as far as I can tell/remember.
If you don't like configuration, you shouldn't care. Just pick any popular distro and use whatever's shipped by default.
The people most capable of fixing the problem you describe are the people who like spending hours tinkering with their Linux UIs.
Besides, Windows usability is rubbish, there's a reason why the first thing that I, and, uh, a few other weirdos do on a new Windows install, is set up an orthodox file manager. [1]
[1] https://en.wikipedia.org/wiki/Far_Manager
what's correct for you is not necesarily what's correct for me.
the problem with windows/mac is very much that it doesn't work the way i want it to work. why even have multiple oerating systems at all, if we don't need choice? we could all be using windows.
i don't see how removing choice solves anything at all.
Wait, I think that reminds me of a desktop environment.
You can have as much snark as you want. I specifically said that Linux should remain the platform of choice, just that the defaults should be stronger so that it's a more consistent and approachable platform for people.
GNOME has made great advancements in this regard, which has been awesome to see.
Agreed.
> GNOME has made great advancements in this regard, which has been awesome to see.
I wasn't joking or being hyperbolic in my previous post. GNOME may have good defaults in certain cases but it ends there. User choice and customization are seen as design defects and anti-features. They really do believe that there is a One True Font and One True Theme out there and anything else is unsupported.
Sane defaults shouldn't come at the cost of user choice. If they do, there's little reason to use Linux over a Mac because the former essentially becomes Android style "look but don't touch" open source.
And that would be completely fine were it not for parent posts which imply that GNOME should be the default and the only experience on the Linux desktop and everything else is irrelevant fragmentation. I respect users choice to use GNOME but I don't appreciate when people call everything else besides GNOME (or anything really) as fragmentation.
For example, here's one of the parent posts
> There is no reason there can’t be one agreed upon blessed way of doing things in Linux while still enabling users like you to have choice.
This is pretty oxymoronic.
To be ultra clear here: I'm not saying it has to be GNOME. I legitimately only said GNOME to toss you a bone when you responded with a needlessly sarcastic comment. ;P
It could be any DE, even a newly created one.
No part of your posts feels productive, and to be honest they feel like emotional backlash.
Like I said earlier,
> but I don't appreciate when people call everything else besides GNOME (or *anything really*) as fragmentation.
The idea of "one agreed upon blessed way of doing things" implies that a DE can be made to work for everyone, which is a false assumption. Just because Windows and Mac follow this philosophy, doesn't mean that it's right. Diversity and independence of thought isn't irrelevant fragmentation.
That would give many distros much less of a reason to exist, and could instead be configuration profiles for GNOME, KDE, etc that could easily be loaded on -any- distro. It would also allow the configurations to be iterated on independent of distros which would eventually lead to much higher levels of polish.
There are way too many distros that are half baked trash that make it into bad listicles returned in search results.
The Linux community sabotages itself.
Stop trying to herd people? I didn't sign up for whatever agenda you're trying to advance, so of course I'm going to spit in your face when you try to tell me what to do.
I do not see a problem as long as there are a few good ones to choose from.
As Kubuntu LTS users I don't feel that I am losing anything if someone else uses Arch with GNOME. Also there are distro made for children with custom DEs, distros for blind people, distros for media etc.
The diversity is not the problem, the problem might be some secondary side effects like some developer really wants to use some bleeding edge thing but he does not have an easy way to deploy it to more stable distros.
Did you use Win 10 ? One GUI in Teams, another in Office 365, a calculator which comes from outher space, 2 interfaces for configuration, disable microfon for admin account -> no microfone for user account.
People don't care that TikTok looks different than Instagram looks different than Gmail looks different than NYTimes.com looks different than a flight reservation site.
See https://en.m.wikipedia.org/wiki/Mac_transition_to_Intel_proc...
I’d expect to see no support for Intel by 2027 or macOS 16.
I'm building a gaming computer at the end of this year and I'm strongly considering just throwing linux on it instead of windows as well, because if protondb is to be believed most games run perfectly fine on linux these days.
Let's forget for second that iOS (until 2015), Android, Switch and PlayStation have some POSIX like support and with exception of PS, all support OpenGL in some form, yet most of those studios couldn't be bothered.
i see a potential future where microsoft will port windows to run on top of a unix kernel. small chance, but not impossible.
those studios won't bother with native linux support until a sizeable amount of users play their games on wine. and that numer is growing. which is the point.
"Half an operating system: The triumph and tragedy of OS/2"
https://arstechnica.com/information-technology/2019/11/half-...
The section "A better DOS than DOS, a better Windows than Windows" is worth a close look and introspection about what WINE/Proton effectively bring to the table.
for one, linux does not have a single business interest driving it, so many of the OS/2 issues relating to IBM don't apply.
being a better windows than windows bolstered windows apps, but only killed OS/2 because native OS/2 was worse than native windows. that too does not apply to linux.
OS/2 died because IBM expected it to be backed by microsoft, and gave up when that didn't happen and they realised that it could not compete. linux never had this issue. linux won't loose its backing. linux never depended on being able to compete against windows. microsoft is now supporting linux more than it ever supported OS/2.
so at the worst, the status quo remains, and nothing changes. realistically though, wine will keep getting better, enabling some people to use it. wine will continue to be developed and supported. there is no risk of that ever stopping. there is no danger betting on wine, because once it works, it will continue to work indefinitely. at worst wine may fall behind implementing some new windows feature that new apps need, but it won't stop supporting what already runs.
wines user share is not going to go down. linux and wine development will not stop. unlike IBM we are able to play the long game here. there is nothing to loose. there are no risks to continue as we are.
however if you look at english steam games only, the value is actually 2.5% and the overall trajectory seems to be slow growth:
https://www.gamingonlinux.com/steam-tracker/
let's see where we are next year.
My personal experience in the last ~5 years or so is increasingly reduced friction in adopting a Linux desktop for non-technical users as companies go towards SaaS.
Firefox ends up being simply a differently shaped search bar.
I guess I am thankful that Linux desktops in general I both love and hate, are good enough to replace Windows if the needs come to that.
Even more so nowadays with Valve setting up Linux as a promising "pc gamer" platform with decent usability.
For good or worse the fragmentation is what Linux "fans" want. The kind of users who are never happy and want gazillions options to make something fit their every perceived special needs. The kind of users who think using a mouse is "less productive and efficient" than keyboard only… these are the most vocal users on Linux discussions it seems.
Oh well like I said I’m grateful Ubuntu is usable enough and looks half decent… actually moving back from a M1 MacMini soon… but mostly for convenience as I want to keep using my "gaming pc".. and I realized i dont actually do anything very macOS specific … plus homebrew and lack of native containers feels like a hack. I’llmiss the polish of macOS though.
While I’m at it let’s be blunt: is there really any half decent looking distro besides Ubuntu? Imho most other distros look like they were designed in 1987, and elementaryOS is such a joke. Personally I could’t see myself using anything else than Ubuntu + the fantastic "Dash to Panel" addon as a valid replacement for Win10 (the win + number shortcuts are amazing).
It actually works great when it does, but more often than not something fails after updates. So I felt like a rolling a dice everytime I have to go though this.
There were problems with update process itself -- they tend to wait for users to confirm configuration overwrite/update/patch so when I attempted to perform update overnight, I often find it waiting for my input in the morning 30 percent into update process.
Ultimately I had enough of the frustration that I've stopped using it.
Where I actually need Linux, I can just use WSL. Windows has its fair share of the problems but I'm definitely not spending as much as time just to keep things running even through major updates.
That's why I use console Linux every work day and enjoy it, and that's why I wouldn't touch Linux on a desktop with ten foot pole, because all my experiments in that area failed sooner or later. Also (surprise) shell Linux knowledge almost doesn't scale to the debugging GPU drivers, DE and X server issues. This is actually rather disconcerting feeling, when you have a decade of Linux experience but you look at crashing desktop GUI and understand that you have approximately as much understanding of the problem as a first time user.
You will be surprised on how much Wayland solved those issues. At least, in my experience, understanding why Sway crashed is a debug flag away. Unfortunately, Wayland is a double edge sword, even more if you want to run games with multiple monitors (but that is a pain even on Windows in my experience).
Preface - my corp issues same laptops to all engineers, including developers actually writing code. Codebase is 100% for Linux, we don't need write anything Windows related. Developers have complete freedom at what to use as a host system on the laptop, there are some guides to help set up host Linux, there are people who already did it and can give advise.
Around half of them just opt to host Windows 10 and run all their tools in the Linux virtual machine. That's a group of people who are second in world most suited to run Linux - actual Linux application devs. (first group would be Linux admins I think)
I remember when I upgraded Ubuntu 14 to Ubuntu 16 and the GUI stopped working and was dumped to the command line. Luckily I had a Linux guru around to fix it, because otherwise my only alternative would have been to reinstall it - seeing how many esoteric steps he did to debug and fix it there is no way I could have googled my way out of it.