No announcement yet.

Fedora vs. openSUSE vs. Manjaro vs. Debian vs. Ubuntu vs. Mint Linux Benchmarks

  • Filter
  • Time
  • Show
Clear All
new posts

  • #21
    OpenSuse is the only one using NOOP on the SSD, and afaik NOOP slows down the entire system with heavy disk usage.

    So part of the problem could be the scheduler.


    • #22
      Originally posted by Daktyl198 View Post
      Should I even bother looking past the first page?
      "Out of the box performance" >> Proceeds to use a beta version of every distro except Linux Mint and Manjaro. In the latter, it's not that hard to use one command to switch to the unstable repos after your install:
      sudo pacman-mirrors -gb unstable

      And BOOM, mesa 11.

      As for Linux Mint, it's as simple as checking a box in your update manager to get to the "beta" repository...

      But, we all know how you feel about Manjaro and Linux Mint, so I'm just falling on deaf ears.
      Yeah I did not get the point of using different versions of software to test the distributions against each other.


      • #23
        Originally posted by shawnsterp View Post
        Correct me if I am wrong, but it looks like there were several different desktop environments were used, but they did not seem to greatly impact the test results. Am I correct in this?
        I agree!


        • #24
          It might be slow, but at least it does not destroy my machine.
          This is the 3. time in a row where I try to install Fedora and it stops after wiping my partitions by crashing.

          So no partition and back to opensuse.

          OpenSUSE is slower but far more stable, as far as my experience goes.


          • #25
            Originally posted by djdoo View Post

            That is what their installer does man and if you have never install a recent opensuse distro please don't say such things... I am using opensuse since it was called suse 8.2 and believe me I know how it works.
            I still remember losing grub in 13.1 version and cannot enter any OS when install with btrfs root and it was a known bug then and since that I never used btrfs again afraid not to mess my system with that thing one more time.
            Ok, that has nothing to do with btrfs, but rather with a bad implementation of it. As for the many partition problem: if suse enables snapshot/backup for every partition than you double the number of partitions. Again its a choice made by the distro who tries to be safer on data integrity at the cost of performance and simplicity (a debateable choice for a desktop usecase...). I'm sure that even in Suse you can customize your disk layout during install and make a simpler /,/home,/boot layout with btrfs and get better performance and a simpler system.

            Anyway, to me this story just highlights one more time that major distros are overall bad. I find it ridiculous that developers release new software in a hurry but then a distro let months pass before it can reach the end user and, many times, there are outstanding bugs caused by maintainers which delays the adoption of new features even more.
            At least, arch Linux solves one of this problem with his release model. Yet a standardized communication model between developers and releasers/distros is in my opinion the only way out of this mess and possibly the door for Linux on desktop and offices.


            • #26
              Regarding the bad results of openSUSE:

              The article says, all distributions (except Manjaro) were using deadline or CFQ. This is not true.
              According to Richard Brown (openSUSE Chairman) the beta used NOOP on SSDs. The final version will use deadline. That should make it better in most cases.