Announcement

Collapse
No announcement yet.

What Would You Like To See Next?

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

  • #51
    I think it would be nice to see running metrics of both proprietary and open source video drivers similar to the way a Mozilla employee made "http://www.arewefastyet.com/?machine=4" to show javascript performance improvements. That way we can see the rate of improvement of the drivers, especially the open source ones compared to the proprietary ones and maybe even their windows couterparts.

    Comment


    • #52
      test test test

      Comment


      • #53
        Michael, could you check up on your Via contacts on whether you could get a Nano dual core for review?

        Comment


        • #54
          I think it would be cool to see a benchmarking deathmatch between different libc implementations:

          Something like this:
          (e)Glibc (dynamic) vs (e)Glibc (static) vs uclibc (static) vs dietlibc (static) vs bionic (metasploit port, static).

          The Sta.li guys claim that static binaries are much faster. It would be nice to have some benchmarks on how big an influence different libc:s have on final performance sort of like how different compilers have been compared previously.

          Comment


          • #55
            benchmark compositing window managers

            I'd like to see compositing window managers benchmarked whenever games are benchmarked too. I do not play games very often but I use the WM all the time, and the video drivers do affect it a lot, so a WM benchmark would be much more relevant to me.

            I know at least kwin has a show_fps plugin and scripting support, so it shouldn't be hard to devise a benchmark for it. Not sure wether compiz and others have scripting support, or how you would create a benchmark without that.

            Comment


            • #56
              As posted by staalmannen here - a comparison of compilers.

              GCC, TCC, Clang, OpenWatcom.
              Compile time, resource usage.
              Application performance for same app versions compiled on different compilers.
              Two (or more) processor architectures - AMD and Intel.

              Comment


              • #57
                Leader Board

                I'd like to see a "Recommended Hardware" Leader Board where users could vote for the hardware that performs best and is most compatible with Linux.

                Comment


                • #58
                  Originally posted by ldillon View Post
                  I'd like to see a "Recommended Hardware" Leader Board where users could vote for the hardware that performs best and is most compatible with Linux.
                  No voting needed. Just use OpenBenchmarking.org.
                  Michael Larabel
                  http://www.michaellarabel.com/

                  Comment


                  • #59
                    Linux Hardware Selection

                    Michael, thanks for the reply.

                    I checked out OpenBenchmarking.org a couple of times before I posted my original comment. It has a bunch of data on the site but I have a hard time condensing it into useful information or purchasing recommendations, though I'm sure it makes prefect sense to the site developers. It's about time Linux had something like that.

                    I've been messing with Linux and the X Windows System since the days when you had to manually set modlines to get anything to work and, frankly, I'd just as soon have something that mostly "just works". I think the best cards I've ever used under Linux were the Matrox G200 and G400 cards....

                    It seems that every few years I do a bunch of research to try and find a solid video card under Linux, but I can never seem to find definitive answers.

                    Benchmarks alone do not take into account subjective issues like ease of install. I worry that if I buy based on benchmarks alone, I'll get blazingly fast hardware, but will have to compile a custom kernel and a few other packages to realize the performance. That's not so bad, but I've been down that path and recompiling the kernel and/or drivers every time there's a new kernel release just isn't worth the effort to me any more.

                    I was thinking thinking of something along the lines of StorageReview.com's Leader Board but with more than just performance, like Best for Stock Ubuntu, Best for Stock Fedora, etc and/or Best for Open Drivers, Best with ATI Driver and Best with NVidia Driver.

                    It's so easy to find good hardware for Windows (Go to Newegg and select Most Reviews in your price range) I think Linux users deserve a easier hardware selection process. I also think that the manufacturers that provide better Linux support should be rewarded for their effort.

                    Thanks for the excellent site.

                    Comment


                    • #60
                      Originally posted by ldillon View Post
                      Benchmarks alone do not take into account subjective issues like ease of install. I worry that if I buy based on benchmarks alone, I'll get blazingly fast hardware, but will have to compile a custom kernel and a few other packages to realize the performance. That's not so bad, but I've been down that path and recompiling the kernel and/or drivers every time there's a new kernel release just isn't worth the effort to me any more.

                      I was thinking thinking of something along the lines of StorageReview.com's Leader Board but with more than just performance, like Best for Stock Ubuntu, Best for Stock Fedora, etc and/or Best for Open Drivers, Best with ATI Driver and Best with NVidia Driver.
                      Contrary to the name, OpenBenchmarking.org can do more than just benchmarks and can answer all of the potential scenarios you outline, albeit the user interface isn't very clear and well laid out and there's some areas for refinement over the coming days, weeks, and months.... It's just not all there yet since it takes a while and I am the one writing all of the code for it and PTS while also writing all of the content on Phoronix.com, so unfortunately, there is lots of work to be done but not enough time in a day. But hopefully by the time you upgrade next, OpenBenchmarking.org will be ready to answer all of your questions.
                      Michael Larabel
                      http://www.michaellarabel.com/

                      Comment

                      Working...
                      X