Announcement

Collapse
No announcement yet.

A Five-Way Linux Distribution Comparison In 2010

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

  • #76
    Was there a known regression in 2.6.33 ? at least, I know that for 2.6.34, the CFQ scheduler was improved.
    What Phoronix should do is to update Arch to the latest version and see if the regression disappears. That are things like that that would transform a poor article into something interesting: instead of creating a flame war, they could be informative and useful for ALL users and distros (like they did when dissecting the ext4 regressions in the kernel)..

    Comment


    • #77
      Originally posted by zeb_ View Post
      I just realised the test under Mandriva cooker and Arch. Kernel is 2.6.34, my CPU is a i7 860, both are 64-bit environment.

      Mandriva: Average: 71.40 Frames Per Second
      Archlinux: Average: 72.74 Frames Per Second

      I have not tested Suse yet, but I see no significant difference between Mandriva cooker (development, now in freeze for the 2010.1 version) and Arch, although their schedulers are different (charge on each core is different on both OSes, as shown by gkrellm). Note that the kernel on my Arch machine is a 2.6.34 (the current one) as opposed to the 2.6.33 from the Phoronix test. I don't know if that would be the explanation.
      Due to margin of error, I don't think your results really count to call anything a regression.

      Comment


      • #78
        @LinuxID10T: Indeed, but I was not talking about the result between Mandriva and Arch, but in the Phoonix test itself, wheredifferencs are significat.
        My results show that latest Arch is not worse than Mandriva, and Phoronix editors should update their article with some explanation/exploration, otherwise that is just sensationalism.

        Comment


        • #79
          Originally posted by LinuxID10T View Post
          Due to margin of error, I don't think your results really count to call anything a regression.
          Indeed, but I was not talking about the result between Mandriva and Arch, but in the Phoonix test itself, where differencs are significant.

          My results show that latest Arch is not worse than Mandriva, and Phoronix editors should update their article with some explanation/exploration, otherwise that is just sensationalism.

          Comment


          • #80
            Originally posted by LinuxID10T View Post
            Due to margin of error, I don't think your results really count to call anything a regression.
            Indeed, but I was not talking about the result between Mandriva and Arch, but in the Phoonix test itself, where differencs are significant.

            My results show that latest Arch is not worse than Mandriva, and Phoronix editors should update their article with some explanation/exploration, otherwise that is just sensationalism.

            Comment


            • #81
              It's interesting to see that Michael replied to everything (even flames) but not the actual technical questions raised by some of us.

              Mainly:
              - Since there is no "stock configuration" for Arch, how DID you configure it?
              - What driver/version did you use for the ATI card for Arch?

              All the flamewar aside, can you please answer these Michael?

              Comment


              • #82
                I've forgot one, crucial question:

                What's your comments and interpretation on the results? Like, since they are all Linux, what actually caused the performance differences? You do realize benchmarks without a conclusion is useless, right?

                Comment


                • #83
                  dcc24 is perfectly right.
                  A very easy thing to do for a start would be to update Arch to the current version, and see if that changes the result. As I showed before, I did not see a difference for the x264 test with my other Linux install (Mandriva cooker), but I use an up-to-date Arch. Either Mandriva suffers the same penalty, either I cannot see the issue (64-bit, i7 CPU, etc), or finally this regression you see has been corrected with the new kernel/gcc/libraries, etc.

                  Comment


                  • #84
                    Frankly, I don't believe it is even possible to install X and KDE without updating the whole system first - 2010.5 is not a "Release" but a snapshot, and it does only include core. So for example X or graphic drivers etc would not be in sync with the core install from iso, since they progressed since the 2010.5 snapshot was made. I am sure he updated the whole system right after the install (as you normally do before installing anything else), if not, that would be totally stupid.

                    Comment


                    • #85
                      Just saw this:
                      Arch Linux results are not available for the OpenGL tests due to problems with the system.
                      lol - what problems? and how would you know that these don't compromise the other tests? real shortness on overall information here...

                      Comment


                      • #86
                        Originally posted by gr.es View Post
                        I am sure he updated the whole system right after the install (as you normally do before installing anything else), if not, that would be totally stupid.
                        Oh yes ? So why does he publish that the Arch kernel in the article is version 2.6.33, when the first version of the 2.6.34 has been released for Arch on the 17th of May, 41 days ago (Subversion (root)/kernel26/trunk/PKGBUILD - Revision 80408)? At least he should give the test date if this article were really scientific.

                        As you write, that is totally stupid, like this whole article !

                        Comment


                        • #87
                          This thread gets two thumbs up from me!

                          Comment


                          • #88
                            *Meanwhile, in Redmond, WA...*
                            "Popcorn, anyone?"

                            Comment


                            • #89
                              Originally posted by Michael View Post
                              Anyone is welcome to reproduce the results.
                              Hardware C2D 4500@2.2GHz, 2GiB RAM, Nvidia GF 275
                              Ubuntu: Compiz turned off, Nvidia drivers v 256.35
                              Arch: DE:KDE 4.4.5, WM:Openbox, stock kernel with minor patches for low latency desktop. Everything else 100% stock.

                              Tests where Arch supposed to sucks.








                              And sneak peek @ Windows vs Linux games performance

                              Comment


                              • #90
                                Originally posted by PsynoKhi0 View Post
                                *Meanwhile, in Redmond, WA...*
                                "Popcorn, anyone?"
                                Well one can see it this way, i.e. a pointless distro war.
                                On the other hand, we have a great opportunity here to understand what is happening, and this could benefit ALL distros, ALL Linux users.
                                Two choices. Too bad the correct choice does not seem to interest a lot of people.

                                Comment

                                Working...
                                X