5 ms·

Sorry but no, Linux simply isn't ready. Not now, not ever. The year of the "Linux desktop" is always 10 years from now. This is the way it has been the past de

by samiv 2y ago

Sorry but no, Linux simply isn't ready. Not now, not ever. The year of the "Linux desktop" is always 10 years from now. This is the way it has been the past decades and this is the way it will be because the process (bazaar) can't produce anything better. I'm a long time Linux user. Been using Linux since Red Hat 7.2, Red Hat, Slackware, Ubuntu and now ArchLinux. Have tried KDE, Gnome, XFCE, Fluxbox and various window managers and environments and the thing is: Linux is always broken somewhere. Graphics glitches, audio glitches, kernel hangs, suspend/resume being broken etc. Yeah and you can claim "it's not the developers fault when they have no specs". Maybe, but that's just the blame game and doesn't change the fact that it's barely beyond alpha quality after +30 years in development. Every upgrade that fixes something breaks something else. This produces a feedback loop where problems like these keep the (desktop) Linux in the marginal users group which means it isn't really a viable target product for developers (talking about drivers etc) which means that it's in the marginal group. You really need at minimum a stable driver ecosystem with stable driver ABIs but Linus refuses to have that. I understand his reasoning and the benefits but one must ask, has it succeeded in producing a viable operating system that is robust and that works and can support various devices including proper graphics? Is this really the best strategy? Building a product is HARD and BORING work that has a long tail. When you're "90% done you still need to take care of the other 90%". Taking a hobbyist weekend project and turning it into a PRODUCT means spending long boring nights fixing those sharp edges and fixing bugs and looking for issues etc. This is work that very few people do in the open source world (due to lack of skills or motivation or whatever), but without this grind the product isn't a product yet but just an alpha demo. Just have a look at the Wayland debacle and all the various implementations. Everything is "alpha this" "alpha that". </rant>

jitl 2y ago
I agree for arbitrary hardware that’s out there, but SteamOS + Steam Deck are well tested and fit together tightly, and they “just work” to a much greater degree than Windows + ${equivalent hardware}. > Taking a hobbyist weekend project and turning it into a PRODUCT means spending long boring nights fixing those sharp edges and fixing bugs and looking for issues etc. This is work that very few people do in the open source world (due to lack of skills or motivation or whatever), but without this grind the product isn't a product yet but just an alpha demo. Valve does this work for Steam Deck, and the ripple effect of their product has improved all distros’ ability to run Windows programs. If they continue to gain market share I could see more of their hard work in areas like audio, Bluetooth, or power management trickling down to general desktop Linux too. They probably won’t contribute much in the way of desktop productivity software, though much of KDE or Gnome’s core apps have been of acceptable quality for years.
- samiv 2y ago
  But the parent was talking about Linux as a generic platform to drive people's computing needs.
  - jitl 2y ago
    My point is that Valve's work on SteamOS is trickling down to other distros.
mariusor 2y ago
I think you're talking about generic Linux distributions without a corporate backed concerted and invested effort into it succeeding. A Linux based Steam console OS is a very different beast, and it's less about "the year of the Linux desktop", and more about "the year/decade of the Steam based console". And my assumption is that the underlying OS is probably meaningless despite the amount of effort Valve devs have put into Steam OS, Proton, and all the other intermediary pieces. If the Linux desktop gets better in the process it's probably a happy coincidence, but I doubt it very much that is an actual KPI for anyone at Valve.
yjftsjthsd-h 2y ago
> You really need at minimum a stable driver ecosystem with stable driver ABIs but Linus refuses to have that. I understand his reasoning and the benefits but one must ask, has it succeeded in producing a viable operating system that is robust and that works and can support various devices including proper graphics? Is this really the best strategy? No, it clearly doesn't need a stable driver ABI. Yes, it works very well. Even nvidia works these days. > Taking a hobbyist weekend project and turning it into a PRODUCT means spending long boring nights fixing those sharp edges and fixing bugs and looking for issues etc. This is work that very few people do in the open source world (due to lack of skills or motivation or whatever), but without this grind the product isn't a product yet but just an alpha demo. You... understand that Linux hasn't been a "hobbyist weekend project" for like 30 years, right? It's mostly developed by big companies and has been for a long time.

by samiv 2y ago

jitl 2y ago

I agree for arbitrary hardware that’s out there, but SteamOS + Steam Deck are well tested and fit together tightly, and they “just work” to a much greater degree than Windows + ${equivalent hardware}. > Taking a hobbyist weekend project and turning it into a PRODUCT means spending long boring nights fixing those sharp edges and fixing bugs and looking for issues etc. This is work that very few people do in the open source world (due to lack of skills or motivation or whatever), but without this grind the product isn't a product yet but just an alpha demo. Valve does this work for Steam Deck, and the ripple effect of their product has improved all distros’ ability to run Windows programs. If they continue to gain market share I could see more of their hard work in areas like audio, Bluetooth, or power management trickling down to general desktop Linux too. They probably won’t contribute much in the way of desktop productivity software, though much of KDE or Gnome’s core apps have been of acceptable quality for years.

samiv 2y ago
But the parent was talking about Linux as a generic platform to drive people's computing needs.
- jitl 2y ago
  My point is that Valve's work on SteamOS is trickling down to other distros.

mariusor 2y ago

I think you're talking about generic Linux distributions without a corporate backed concerted and invested effort into it succeeding. A Linux based Steam console OS is a very different beast, and it's less about "the year of the Linux desktop", and more about "the year/decade of the Steam based console". And my assumption is that the underlying OS is probably meaningless despite the amount of effort Valve devs have put into Steam OS, Proton, and all the other intermediary pieces. If the Linux desktop gets better in the process it's probably a happy coincidence, but I doubt it very much that is an actual KPI for anyone at Valve.

yjftsjthsd-h 2y ago

> You really need at minimum a stable driver ecosystem with stable driver ABIs but Linus refuses to have that. I understand his reasoning and the benefits but one must ask, has it succeeded in producing a viable operating system that is robust and that works and can support various devices including proper graphics? Is this really the best strategy? No, it clearly doesn't need a stable driver ABI. Yes, it works very well. Even nvidia works these days. > Taking a hobbyist weekend project and turning it into a PRODUCT means spending long boring nights fixing those sharp edges and fixing bugs and looking for issues etc. This is work that very few people do in the open source world (due to lack of skills or motivation or whatever), but without this grind the product isn't a product yet but just an alpha demo. You... understand that Linux hasn't been a "hobbyist weekend project" for like 30 years, right? It's mostly developed by big companies and has been for a long time.