5 ms·

Linus Torvalds think the Linux desktop is in trouble

38 points by CrankyBear 8y ago

jcastro 8y ago
I don't think a Foundation would help in this regard, who's going to fund it and why would they? The article kind of glosses over the fact that there's really no business case for a traditional linux desktop. Chromebooks are tied to a hardware sale where the hardware is certified to avoid having to support multiple hardware configurations and ensure users never see Linux at all. Specialty shops like System76 and Dell's Sputnik specifically are tied to hardware and a specific market (pros and sysadmdins), this is where I fall in but most normal PC users would generate absolutely zero revenue for anyone trying to build a sustainable linux desktop experience.
esotericn 8y ago
> the traditional Linux desktop, in all its variations, will remain a niche operating system for power users. Is it not the case that, well, _computers_ are a niche system for power users? Most of the developers I know use OSX or Linux. The students I know are all on OSX and effectively use Office and the browser. People with bog-standard office jobs are on Windows because their employer controls their machine. Everyone else doesn't really use traditional "computers", it's all phone stuff.
- tiredyam 8y ago
  You are forgetting pc gamers :)
pram 8y ago
and here I was thinking that 2019 must surely be the year of Linux on the desktop
- beatgammit 8y ago
  We'll win eventually, once Microsoft gives up. We've got far more staying power, since it costs us virtually nothing to stay the course. It may take another 20 years, but we'll get there. :)
  - jdsully 8y ago
    Why would Microsoft ever give up? They make billions off of Windows.
    - burning_hamster 8y ago
      For the time being, talent is limited. Their costly engineers' time might be spent more productively on other products. Windows is already no longer their primary cash cow (nor even their secondary -- IIRC, it comes third in revenue after Azure and Office). I don't think they will kill windows any time soon, but I do think that the probably is non-zero, even in the not too distant future.
- jazoom 8y ago
  PopOS has converted me. The main overly complex and brittle thing now seems to be the installation of programs from so many different sources, and apt itself. If that was a better user experience I'd have no trouble recommending it to all Windows users.
zzzcpan 8y ago
Flatpak, Snap are not significant contributors to desktop fragmentation, unlike Gnome 3, Wayland, dropping Unity, all forcing people to try different distros. I switched to Lubuntu myself after Unity and probably going to switch again to Xfce-based Xubuntu, because they want to break Lubuntu too.
kpU8efre7r 8y ago
I've shared the same exact frustration for over a decade now. One of Linux's greatest strengths gives rise to it's biggest weakness- fragmentation. "I disagree with x design decision, I'll make my own distro" It's like there is so much rope that the community can't help but hang itself. The talent is spread across dozens of distros each with their own ethos and design direction. Every new distro hurts Linux as a whole at this point.
- sl1ck731 8y ago
  I never considered this before. If all the major players were consolidated we would have a kick-ass desktop experience. Now I'm sad. I imagine browser-based development environments will be polished enough for me to use by the time any desktop distribution holds my attention long enough to use as a daily driver.
  - kgwxd 8y ago
    One persons kick-ass desktop is another persons needlessly slow, flow killing, eye-sore. DEs haven't needed any new features since 1995. Browser based productivity software is the second leading cause of global warming.
- hactually 8y ago
  I wonder how many people would Patreon/sponsor developers to do a Linux OS properly? I am happy to throw money towards it - I think it would have to be built on ostree, flatpack, Gnome 3 and Systemd to get anywhere close to a modern development desktop. Anyone else interested in putting money where their mouth is?
  - kgwxd 8y ago
    I will, if you just take out the systemd part ;)
  - chithanh 8y ago
    The problem of Linux package management has been solved, and the solution is not flatpak (nor snap). It is Nix/Guix. About the user experience, I'm right there with Linus that Chrome OS is the way to go forward. And this opinion is not new btw., he already voiced it way back when he got his first Chromebooks.
mastrsushi 8y ago
The two sides that I've hopped back and forth from in this argument are competition leads to innovation, and fragmentation is the problem. Getting started with Linux 8 years ago, I used to think the endless array of distros was a beautiful and unique characteristic that reflected on the openness of FOSS. As I gained experience, I realized that the differences were too close to really care. That there's an awfully large amount of effort being put into divided communities too maintain working "operating systems" that are hardly unique. I find Linux best suited in its inherent case, Unix for the x86 PC. Unix as a programmers workbench. Unfortunately and fortunately the "Windows killer" nameplace is responsible for its early success and what it will probably continue to live by for desktop usage. It's always great to hear someone put life into their grandma's Windows 7 laptop by reimaging with an Ubuntu USB. But I feel like that doesn't last long as non corporate supported desktop Linux doesn't hold up well without some annual sys admin love. I definitely went on a rant, but what I can conclude is that competing technologies isn't really to blame as the freedom of choice goes with the nature of FOSS. The bigger problem is that we are fighting to turn a Unix like OS into a Windows replacement by relying on voluntary man power and overlapping attempts at essentially the same goal. Leave it to Apple and Google to bring Unix to the masses, if that even matters.

38 points by CrankyBear 8y ago

I don't think a Foundation would help in this regard, who's going to fund it and why would they? The article kind of glosses over the fact that there's really no business case for a traditional linux desktop. Chromebooks are tied to a hardware sale where the hardware is certified to avoid having to support multiple hardware configurations and ensure users never see Linux at all. Specialty shops like System76 and Dell's Sputnik specifically are tied to hardware and a specific market (pros and sysadmdins), this is where I fall in but most normal PC users would generate absolutely zero revenue for anyone trying to build a sustainable linux desktop experience.

esotericn 8y ago

> the traditional Linux desktop, in all its variations, will remain a niche operating system for power users. Is it not the case that, well, _computers_ are a niche system for power users? Most of the developers I know use OSX or Linux. The students I know are all on OSX and effectively use Office and the browser. People with bog-standard office jobs are on Windows because their employer controls their machine. Everyone else doesn't really use traditional "computers", it's all phone stuff.

tiredyam 8y ago
You are forgetting pc gamers :)

pram 8y ago

and here I was thinking that 2019 must surely be the year of Linux on the desktop

beatgammit 8y ago
We'll win eventually, once Microsoft gives up. We've got far more staying power, since it costs us virtually nothing to stay the course. It may take another 20 years, but we'll get there. :)
- jdsully 8y ago
  Why would Microsoft ever give up? They make billions off of Windows.
  - burning_hamster 8y ago
    For the time being, talent is limited. Their costly engineers' time might be spent more productively on other products. Windows is already no longer their primary cash cow (nor even their secondary -- IIRC, it comes third in revenue after Azure and Office). I don't think they will kill windows any time soon, but I do think that the probably is non-zero, even in the not too distant future.
jazoom 8y ago
PopOS has converted me. The main overly complex and brittle thing now seems to be the installation of programs from so many different sources, and apt itself. If that was a better user experience I'd have no trouble recommending it to all Windows users.

zzzcpan 8y ago

Flatpak, Snap are not significant contributors to desktop fragmentation, unlike Gnome 3, Wayland, dropping Unity, all forcing people to try different distros. I switched to Lubuntu myself after Unity and probably going to switch again to Xfce-based Xubuntu, because they want to break Lubuntu too.

kpU8efre7r 8y ago

I've shared the same exact frustration for over a decade now. One of Linux's greatest strengths gives rise to it's biggest weakness- fragmentation. "I disagree with x design decision, I'll make my own distro" It's like there is so much rope that the community can't help but hang itself. The talent is spread across dozens of distros each with their own ethos and design direction. Every new distro hurts Linux as a whole at this point.

sl1ck731 8y ago
I never considered this before. If all the major players were consolidated we would have a kick-ass desktop experience. Now I'm sad. I imagine browser-based development environments will be polished enough for me to use by the time any desktop distribution holds my attention long enough to use as a daily driver.
- kgwxd 8y ago
  One persons kick-ass desktop is another persons needlessly slow, flow killing, eye-sore. DEs haven't needed any new features since 1995. Browser based productivity software is the second leading cause of global warming.
hactually 8y ago
I wonder how many people would Patreon/sponsor developers to do a Linux OS properly? I am happy to throw money towards it - I think it would have to be built on ostree, flatpack, Gnome 3 and Systemd to get anywhere close to a modern development desktop. Anyone else interested in putting money where their mouth is?
- kgwxd 8y ago
  I will, if you just take out the systemd part ;)
- chithanh 8y ago
  The problem of Linux package management has been solved, and the solution is not flatpak (nor snap). It is Nix/Guix. About the user experience, I'm right there with Linus that Chrome OS is the way to go forward. And this opinion is not new btw., he already voiced it way back when he got his first Chromebooks.

mastrsushi 8y ago

The two sides that I've hopped back and forth from in this argument are competition leads to innovation, and fragmentation is the problem. Getting started with Linux 8 years ago, I used to think the endless array of distros was a beautiful and unique characteristic that reflected on the openness of FOSS. As I gained experience, I realized that the differences were too close to really care. That there's an awfully large amount of effort being put into divided communities too maintain working "operating systems" that are hardly unique. I find Linux best suited in its inherent case, Unix for the x86 PC. Unix as a programmers workbench. Unfortunately and fortunately the "Windows killer" nameplace is responsible for its early success and what it will probably continue to live by for desktop usage. It's always great to hear someone put life into their grandma's Windows 7 laptop by reimaging with an Ubuntu USB. But I feel like that doesn't last long as non corporate supported desktop Linux doesn't hold up well without some annual sys admin love. I definitely went on a rant, but what I can conclude is that competing technologies isn't really to blame as the freedom of choice goes with the nature of FOSS. The bigger problem is that we are fighting to turn a Unix like OS into a Windows replacement by relying on voluntary man power and overlapping attempts at essentially the same goal. Leave it to Apple and Google to bring Unix to the masses, if that even matters.