Announcement

Collapse
No announcement yet.

KDE Neon Is Now Live: Providing Bleeding-Edge KDE Packages For Ubuntu

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • KDE Neon Is Now Live: Providing Bleeding-Edge KDE Packages For Ubuntu

    Phoronix: KDE Neon Is Now Live: Providing Bleeding-Edge KDE Packages For Ubuntu

    Following the news yesterday that KDE is incubating a new "KDE Neon" project, it's now been launched from FOSDEM this weekend in Brussels, Belgium...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    "nor at this point does it seem to be fundamentally different from the former Project Neon." - well, it is.
    This one is officially supported by KDE developers, one would expect it to be well maintained(bugs, in this one, are bugs from KDE developers).
    Well, some bugs might be because of the distro itself but we did hear them(the developers, not particularly KDE) complain about many bugs being because of bad DE integration.

    Comment


    • #3
      And they are going to provide live images for normal users and developers. Sure it may "not be official KDE distro" but it looks like one

      Comment


      • #4
        This sounds like a ubuntu PPA for more recent KDE packages. And then this is made into a kubuntu_like.iso?

        Comment


        • #5
          Ubuntu KDE users will have another alternative to install a KDE desktop that aims to follow KDE releases independent of the core Ubuntu base and its code freezes, which is a good thing.

          But in terms of stability I will not have high hopes for it, since in my understanding Kubuntu KDE implementation is as close to upstream as it can be.

          Like with Fedora, there is not much the community can do about when the central body makes decisions about their single favorite desktop and leave everyone else with... Fix it if you can!

          Comment


          • #6
            Originally posted by darkcoder View Post
            Ubuntu KDE users will have another alternative to install a KDE desktop that aims to follow KDE releases independent of the core Ubuntu base and its code freezes, which is a good thing.

            But in terms of stability I will not have high hopes for it, since in my understanding Kubuntu KDE implementation is as close to upstream as it can be.

            Like with Fedora, there is not much the community can do about when the central body makes decisions about their single favorite desktop and leave everyone else with... Fix it if you can!
            The problem with Kubuntu has always been the Ubuntu policy of freezing the world every release.

            That works with untrusted third party libraries made by some guy on the weekend, who will regularly break ABI and API with updates that use bad versioning.

            It does not work with huge projects like KDE that use sophisticated release cadences that dramatically improve the platform. Gnome also has this as a huge problem, whenever a Gnome release comes out just before an Ubuntu release that then ships and maintains a legacy version of the desktop for upwards of five years on the LTS.

            I hope that as a community we are gradually realizing several things that we need to do "fix" the distro ecosystem as it exists today, and help make Linux more palpable to consumers and developers.

            There are three use cases for Linux - servers, enterprise, and consumer. Each should be served independently because each require different methodologies relating to releases. The server and enterprise ecosystems are already extremely well provided for because almost any company that gets involved competitively in Linux see those segments as the place to make revenue. Between Debian Stable, RHEL, CentOS, Ubuntu Server, SUSE Enterprise, Core OS, Ubuntu LTS, and possibly Fedora Workstation, we very well cover the needs of companies that want server or enterprise desktop / laptop solutions. They are rock solid, stable, and have predictable release cadences and commercial support. We need holy distributions for normal people to use. Nobody is filling that need - because the need is dramatically different from the enterprise use cases every major distribution targets today. It might be in part why Arch has become so popular amongst enthusiasts.

            We always want the year of the Linux desktop, but weare never specific enough to specify the consumer desktop. We want an operating system our friends and family are using for their daily computing needs. And the architectures of Ubuntu, Fedora, Debian, and Suse are all tangential to the needs of end users - for many reasons.
            • The user model used in the server / enterprise sector makes no sense for average users. Google knew this, and its why Android works very well - you should never be having a consumer user installing software as root. Every package manager on every desktop today is expecting root permissions to install software system wide as if you were a system administrator running a cluster of desktops rather than your dad trying to install Candy Crush. Beyond just the software process, the services and system data model fails because as a user when you want to run a DLNA server or use SMB you don't give a shit if your uncle who has an account on the computer does the same - almost everything should be user local, almost every service should be sandboxed to you, and desktop Linux for average people would learn well from the Android model of having an immutable system image and a userspace of apps beyond it.
            • End users are stupid, and will run stupid shit. We are better than Windows in every way, including security - but we are idiots and don't take advantage of it. Android uses SELinux, and its crap, but we could be using Apparmor or PAX to its fullest and giving users of our operating systems real security with sane default profiles and intuitive prompts when a program wants to access a folder or use the network, but we do not.
            • The fact there are 4+ package formats popular across Linux is completely ridiculous and ruins everything. xdg-app and appstream are probably the most important projects today for consumer adoption of Linux as a real operating system for people and not nerds, because they can abstract the bullshit barriers between distros and let people just install and run software without bricking their computers because they got a virus. Once we have common descriptions of most dependencies, ways to introduce new ones, ways to protect against hostile software sanely, and the software stores to get cool shit, we can start talking about being relevant.
            • Unlike the enterprise, consumers use random hardware. Hardware support in Linux is already a PITA, yet no distro is following the upstream kernel release cycle. They package one kernel version, and when someone comes along with hardware that is supported in newer releases but not the current one... well, thats just outright unacceptable in the first place, but it also loses you users. Any sane consumer distro would be following the upstream kernel the same way you would want to keep up to date on any software, and would have the wherewithal to call bullshit when the kernel developers break shit rather than giving up and making your OS unusable half the time on new processors or GPUs. This is why I am always infuriated by projects like OpenSUSE Leap, because its the realization that holding back the user experience is awful, but think the developers behind the fundamentals are either idiots that break everything (and they could be, and if they are you either use a different product or yell at them until they stop) or that keeping your guts up to date is unimportant, like Pulseaudio is shipping 8.0 for just shits and giggles or Mesa developers are just moving whitespace each point release. Fundamentally, its ignoring the problem to stop the world and it makes your users suffer for not acknowledging when upstream developers are being incredibly ignorant by breaking their libraries intentionally or not.

            The first real consumer Linux desktop operating system will not be on some random release cadence that freezes the world. It will package, vigorously test, and ship as soon as possible new versions of software to users so that problems are solved as soon as possible and improvements are experienced sooner rather than later. It will let developers themselves ship their own software, in their own app store, and let the users criticize the developer when they break stuff. That might be the largest problem, in many ways - as a developer, I'm not going to become a maintainer in Arch, Debian, Ubuntu, etc to support my library or threes clone. I'm going to put it out there and someone from those projects is going to maintain it, if it ever makes it in. And then they take the blame when something breaks, and they get paranoid about accepting changes from upstream because it might break something.

            This never happens on any other platform because the developer is the one providing the software, not the distribution. Windows only gives you win32 and .net, you build your own insane mess of Windows Forms and Direct3d. On Android, when you install an app and it breaks, you don't blame Google, you blame the guy who made the app and broke it. This just comes back to appstream and xdg - we desperately need software stores, with some kind of community curation board, available in all distros and accessable by developers directly. No more of these insane layers of indirection of software availability that makes the UX of Linux a nightmare for my mom. We would all benefit so much from accepting the reality that a consumer distro must be fundamentally different from an enterprise one because consumers want dramatically different things from their computers.

            Neon is a step in the right direction. Its basically OpenSUSE Leap, again. But its not solving all the problems, just some of them. Just gotta keep working towards improvement...

            Comment


            • #7
              Originally posted by darkcoder View Post
              Ubuntu KDE users will have another alternative to install a KDE desktop that aims to follow KDE releases independent of the core Ubuntu base and its code freezes, which is a good thing.

              But in terms of stability I will not have high hopes for it, since in my understanding Kubuntu KDE implementation is as close to upstream as it can be.

              Like with Fedora, there is not much the community can do about when the central body makes decisions about their single favorite desktop and leave everyone else with... Fix it if you can!
              The problem with Kubuntu has always been the Ubuntu policy of freezing the world every release.

              That works with untrusted third party libraries made by some guy on the weekend, who will regularly break ABI and API with updates that use bad versioning.

              It does not work with huge projects like KDE that use sophisticated release cadences that dramatically improve the platform. Gnome also has this as a huge problem, whenever a Gnome release comes out just before an Ubuntu release that then ships and maintains a legacy version of the desktop for upwards of five years on the LTS.

              I hope that as a community we are gradually realizing several things that we need to do "fix" the distro ecosystem as it exists today, and help make Linux more palpable to consumers and developers.

              There are three use cases for Linux - servers, enterprise, and consumer. Each should be served independently because each require different methodologies relating to releases. The server and enterprise ecosystems are already extremely well provided for because almost any company that gets involved competitively in Linux see those segments as the place to make revenue. Between Debian Stable, RHEL, CentOS, Ubuntu Server, SUSE Enterprise, Core OS, Ubuntu LTS, and possibly Fedora Workstation, we very well cover the needs of companies that want server or enterprise desktop / laptop solutions. They are rock solid, stable, and have predictable release cadences and commercial support. We need holy distributions for normal people to use. Nobody is filling that need - because the need is dramatically different from the enterprise use cases every major distribution targets today. It might be in part why Arch has become so popular amongst enthusiasts.

              We always want the year of the Linux desktop, but weare never specific enough to specify the consumer desktop. We want an operating system our friends and family are using for their daily computing needs. And the architectures of Ubuntu, Fedora, Debian, and Suse are all tangential to the needs of end users - for many reasons.
              • The user model used in the server / enterprise sector makes no sense for average users. Google knew this, and its why Android works very well - you should never be having a consumer user installing software as root. Every package manager on every desktop today is expecting root permissions to install software system wide as if you were a system administrator running a cluster of desktops rather than your dad trying to install Candy Crush. Beyond just the software process, the services and system data model fails because as a user when you want to run a DLNA server or use SMB you don't give a shit if your uncle who has an account on the computer does the same - almost everything should be user local, almost every service should be sandboxed to you, and desktop Linux for average people would learn well from the Android model of having an immutable system image and a userspace of apps beyond it.
              • End users are stupid, and will run stupid shit. We are better than Windows in every way, including security - but we are idiots and don't take advantage of it. Android uses SELinux, and its crap, but we could be using Apparmor or PAX to its fullest and giving users of our operating systems real security with sane default profiles and intuitive prompts when a program wants to access a folder or use the network, but we do not.
              • The fact there are 4+ package formats popular across Linux is completely ridiculous and ruins everything. xdg-app and appstream are probably the most important projects today for consumer adoption of Linux as a real operating system for people and not nerds, because they can abstract the bullshit barriers between distros and let people just install and run software without bricking their computers because they got a virus. Once we have common descriptions of most dependencies, ways to introduce new ones, ways to protect against hostile software sanely, and the software stores to get cool shit, we can start talking about being relevant.
              • Unlike the enterprise, consumers use random hardware. Hardware support in Linux is already a PITA, yet no distro is following the upstream kernel release cycle. They package one kernel version, and when someone comes along with hardware that is supported in newer releases but not the current one... well, thats just outright unacceptable in the first place, but it also loses you users. Any sane consumer distro would be following the upstream kernel the same way you would want to keep up to date on any software, and would have the wherewithal to call bullshit when the kernel developers break shit rather than giving up and making your OS unusable half the time on new processors or GPUs. This is why I am always infuriated by projects like OpenSUSE Leap, because its the realization that holding back the user experience is awful, but think the developers behind the fundamentals are either idiots that break everything (and they could be, and if they are you either use a different product or yell at them until they stop) or that keeping your guts up to date is unimportant, like Pulseaudio is shipping 8.0 for just shits and giggles or Mesa developers are just moving whitespace each point release. Fundamentally, its ignoring the problem to stop the world and it makes your users suffer for not acknowledging when upstream developers are being incredibly ignorant by breaking their libraries intentionally or not.

              The first real consumer Linux desktop operating system will not be on some random release cadence that freezes the world. It will package, vigorously test, and ship as soon as possible new versions of software to users so that problems are solved as soon as possible and improvements are experienced sooner rather than later. It will let developers themselves ship their own software, in their own app store, and let the users criticize the developer when they break stuff. That might be the largest problem, in many ways - as a developer, I'm not going to become a maintainer in Arch, Debian, Ubuntu, etc to support my library or threes clone. I'm going to put it out there and someone from those projects is going to maintain it, if it ever makes it in. And then they take the blame when something breaks, and they get paranoid about accepting changes from upstream because it might break something.

              This never happens on any other platform because the developer is the one providing the software, not the distribution. Windows only gives you win32 and .net, you build your own insane mess of Windows Forms and Direct3d. On Android, when you install an app and it breaks, you don't blame Google, you blame the guy who made the app and broke it. This just comes back to appstream and xdg - we desperately need software stores, with some kind of community curation board, available in all distros and accessable by developers directly. No more of these insane layers of indirection of software availability that makes the UX of Linux a nightmare for my mom. We would all benefit so much from accepting the reality that a consumer distro must be fundamentally different from an enterprise one because consumers want dramatically different things from their computers.

              Neon is a step in the right direction. Its basically OpenSUSE Leap, again. But its not solving all the problems, just some of them. Just gotta keep working towards improvement...

              Comment


              • #8
                This could grow into something that makes it unnecessary for various maintainers to package KDE for their distros. If that's the case, then I like the idea.
                For the time being only the unstable/developer images seem to be available, let's hope the others follow shortly (as in days).

                Comment


                • #9
                  it is also providing the very latest Qt as well for Ubuntu
                  This sounds really interesting, especially as some KDE devs have blamed outdated versions of Qt on Ubuntu for multi-monitor problems. Looking forward to the stable release.

                  Comment


                  • #10
                    Originally posted by bug77 View Post
                    This could grow into something that makes it unnecessary for various maintainers to package KDE for their distros. If that's the case, then I like the idea.
                    I doubt it, because if that was the case, they'd be using OBS, not an apt repository.

                    Comment

                    Working...
                    X