Announcement

Collapse
No announcement yet.

Vote On A Distribution For Linux Benchmarking

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

  • #61
    Originally posted by glasen View Post
    Same is for Fedora or any other distribution in this vote. As long a distribution does not clearly get the majority of all votes, you can always argue that the majority won't be satisfied.
    Yep, (i assume there won't be) but even if there is clean winner minority will complain .

    Linus Torvalds, Alex Deucher, Keith Packard... are just 3 votes - overall majority is nothing without relevant minority . Even those relevant parts of the community does not use same distro

    Comment


    • #62
      I apologise. I voted wrong. I voted for Manjaro/Arch, but I quickly regretted my decision. OpenSUSE is the best choice. I hope Micheal sees this in case of a close call between them.

      Comment


      • #63
        I use Xubuntu/Linux Mint, voted for Fedora.

        Comment


        • #64
          Originally posted by glasen View Post
          Same is for Fedora or any other distribution in this vote. As long a distribution does not clearly get the majority of all votes, you can always argue that the majority won't be satisfied.
          I don't think it is the same because of the Mir/Wayland division. All distros except Ubuntu will use Wayland, so benchmarking Fedora will give informative results for other distros too (modulo the specific issues/features/hacks/etc. used by different distros, like what happens now with X). On the other hand, benchmarking Mir will give informative results for Ubuntu only.

          Comment


          • #65
            I'm caught between the obvious choice, Ubuntu due to it having every single desktop environment as an apt-get away and as well as well supporting tons of games and software for Michael to benchmark....... or being a troll and suggesting him to build his own distro(LFS), that way his results mean nothing in relation to actually distros people use.

            95% of the people here look at these benchmarks, and just make up their own conclusion based on their own beliefs and not what the evidence shows.... so it doesn't really matter that much what Michael shows anyway. Let's be real here.

            Comment


            • #66
              Originally posted by dh04000 View Post
              Ubuntu due to it having every single desktop environment as an apt-get away
              The only problem with this is usually degraded performance. For example installing xubuntu-desktop on vanilla Ubuntu doesn't give the same performance as Xubuntu itself.

              Comment


              • #67
                Voted for Ubuntu and Kubuntu

                Originally posted by wizard69 View Post
                I never really understood the fascination with Ubuntu. I alway saw it as a crap botch job of a distro.
                You've got to be kidding. If something is crap it's definitely Fedora. It has messed up installer and bugs which make it unusable for typical users. For example:

                - it doesn't install language pack
                - it cannot be upgraded with graphical package manager
                - it doesn't reboot after manual upgrade
                - it doesn't allow to easily install codecs and proprietary drivers

                All of these I've experienced in Fedora KDE spin and gnome version is probably even worse, because it has much different UI and doesn't even provide basic configuration options.

                Comment


                • #68
                  Originally posted by GreatEmerald View Post
                  • Debian Sid is a good choice in terms of not being as rock-stable as the main Debian, but its rolling-release status is problematic due to there being no reference point that would allow others to reproduce the results.
                  Every distribution that updates its repositories with new package versions has the same problem. If you installed Fedora 17 on the day it was released and install it again now you will see many package versions changed which invalidates any benchmarking. It's very hard to manually rollback a system to match what was benchmarked long ago. Debian and Ubuntu actually have a tool to do this, to capture the package system state on one system and restore it exactly on another: apt-clone

                  Comment


                  • #69
                    ITT:

                    People voting for their distro of choice (no matter how obscure) and feeling self-righteous and smug about how much they hate the child-killing Ubuntu.

                    Gotta love this tabloid.

                    Comment


                    • #70
                      I'm still interested in Ubuntu benchmarks, with Mir, because it is very large and my software needs to perform well on it. I think OpenSuSE or Fedora would be a better platform for general benchmarks though.

                      (Also an Arch user here, Arch is entirely unsuitable for this purpose, without significant bespoke infrastructure)

                      Comment

                      Working...
                      X