Announcement

Collapse
No announcement yet.

The Fastest Linux Distributions For Web Browsing - Firefox + Chrome Benchmarks On Eight Distros

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

  • #21
    Originally posted by duby229 View Post
    Which, of course, triggers btrfs's second -still unfixed- flaw. The balance command will cause corruption.
    Sorry what?

    Comment


    • #22
      Originally posted by SyXbiT View Post
      After having used Chrome for ~8 years, I thought now would be a good time to switch back to Firefox. However, Firefox on Linux scrolls terribly (Arch Linux). I get there are fixes for smooth scrolling, but why in 2019 doesn't Firefox on Linux have good defaults that don't require tweaking.
      Firefox working quite beautifully on Debian and openSUSE for me right now. Haven't had scrolling problems for probably 7-8 years. Just running default Firefox on both. Debian is ESR version. Have you tried ESR? Seems quite good.

      Comment


      • #23
        Originally posted by andyprough View Post

        Firefox working quite beautifully on Debian and openSUSE for me right now. Haven't had scrolling problems for probably 7-8 years. Just running default Firefox on both. Debian is ESR version. Have you tried ESR? Seems quite good.
        I had scrolling issues with Firefox on Manjaro XFCE until I created an xorg.conf and enabled vsync with my RX 580. In that instance it was just an XFCE issue because Plasma didn't have Firefox scrolling issues on the same install regardless of an xorg.conf.

        Comment


        • #24
          Originally posted by skeevy420 View Post

          I had scrolling issues with Firefox on Manjaro XFCE until I created an xorg.conf and enabled vsync with my RX 580. In that instance it was just an XFCE issue because Plasma didn't have Firefox scrolling issues on the same install regardless of an xorg.conf.
          I see, I've been sticking with Mate for awhile now, and KDE before that. Am I right that XFCE development is sort of stalled? I think that was the last DE where I had significant Firefox problems, but it's been quite a few years since I tried XFCE. I used to run an FSF-approved Arch spin called Parabola, but everything seemed to work fine with that.

          Comment


          • #25
            Would love to see Snaps tested as well - but not sure what distro to best test them on:
            https://snapcraft.io/chromium
            https://snapcraft.io/firefox

            Comment


            • #26
              Originally posted by SyXbiT View Post
              After having used Chrome for ~8 years, I thought now would be a good time to switch back to Firefox. However, Firefox on Linux scrolls terribly (Arch Linux). I get there are fixes for smooth scrolling, but why in 2019 doesn't Firefox on Linux have good defaults that don't require tweaking.
              Can't reproduce on OpenSUSE Tumbleweed.

              Comment


              • #27
                Ouch for openSUSE. I tried to use it because of it's "Enterprise features", but found it lacking.

                Originally posted by andyprough View Post

                It should not be benchmarked for performance. Default configuration includes btrfs instead of much faster EXT4, balloo running constantly on KDE and interfering with performance as it does an initial index of the entire system, and a CPU governor on top of it.

                Why are these the defaults? openSUSE is the testing ground for SUSE enterprise linux, for which btrfs is an important technology, and which has a goal of stability despite a wide range of technologies. I'm assuming that's the reason why.

                Anyone who wanted performance would not run it this way. They would install it on EXT4, turn off balloo indexing, and switch the CPU governor to "performance" mode. Among other things. Personally, I would run Mate for a better performance instead of KDE. Also, I get better performance when I compile my own kernels.
                I doubt btrfs has anything to do with these tests. The kernel is newer than all other versions except clearLinux, so either it`s a kernel regression that was fixed, or openSUSE is doing something which is making it slow. These are web-tests on a system with 16GB of ram, a samsung EVO SSD. The filesystem won't be the bottleneck.

                I'd say don't use openSUSE period. They changed owners twice this year, three times total in the last 3 years. Their internal structure must be all messed up from these reorganisations.

                Comment


                • #28
                  Originally posted by AndyChow View Post
                  Ouch for openSUSE. I tried to use it because of it's "Enterprise features", but found it lacking.

                  I doubt btrfs has anything to do with these tests. The kernel is newer than all other versions except clearLinux, so either it`s a kernel regression that was fixed, or openSUSE is doing something which is making it slow. These are web-tests on a system with 16GB of ram, a samsung EVO SSD. The filesystem won't be the bottleneck.

                  I'd say don't use openSUSE period. They changed owners twice this year, three times total in the last 3 years. Their internal structure must be all messed up from these reorganisations.
                  Sounds like classic FUD, and you ignored the majority of my points for some reason. Not interested in your bias against any specific distros. openSUSE consistently runs behind on Michael's benchmarks, often for the reasons I stated. The company appears to be on solid financial and management grounding.

                  Comment


                  • #29
                    Long-time openSUSE (Tumbleweed and Leap) user and promoter here....

                    What exactly is it about openSUSE's defaults that are so "conservative" so as to cause such major hits in these tests? Why would even BtrFS make a difference, if the whole browser is basically running cached in RAM anyway?

                    I'm not trying to be argumentative or dispute these results, I'm genuinely interested in knowing the explanation.

                    Comment


                    • #30
                      Originally posted by sb56637 View Post
                      Long-time openSUSE (Tumbleweed and Leap) user and promoter here....

                      What exactly is it about openSUSE's defaults that are so "conservative" so as to cause such major hits in these tests? Why would even BtrFS make a difference, if the whole browser is basically running cached in RAM anyway?

                      I'm not trying to be argumentative or dispute these results, I'm genuinely interested in knowing the explanation.
                      For a couple of things, Michael almost certainly has balloo full disk indexing turned on in KDE (on by default) and the openSUSE CPU governor runs in "powersave" mode by default - both of which are significant performance bottlenecks. There are a number of different things I would probably change from default for better performance, including the choice of DE and rolling my own kernels.

                      Comment

                      Working...
                      X