Announcement

Collapse
No announcement yet.

PC-BSD 7.1 vs. Kubuntu 9.04 Benchmarks

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

  • deanjo
    replied
    Originally posted by npcomplete View Post
    ^ Exactly. What would you do then if the OS did not come with a compiler?
    Look for another distro

    Besides, what if there was a updated version of the compiler available that improves the run time performance? What would you conclude then about the performance comparison between OSes? In other words, it's still exactly the same OS -- you touched nothing of the kernel, system commands or libraries -- and yet, after updating the compiler the benchmarks "magically" got better? How can such benchmarks tell us anything at all about the OS if the OS did not change?

    Yes PTS pulls from the source for it's tests. It does not however compile base system components that were used to build the base system. Many (if not most) of the tests do not come with most distributions (mostly because of BS licensing and potential legal issues).

    Leave a comment:


  • npcomplete
    replied
    ^ Exactly. What would you do then if the OS did not come with a compiler?
    Besides, what if there was a updated version of the compiler available that improves the run time performance? What would you conclude then about the performance comparison between OSes? In other words, it's still exactly the same OS -- you touched nothing of the kernel, system commands or libraries -- and yet, after updating the compiler the benchmarks "magically" got better? How can such benchmarks tell us anything at all about the OS if the OS did not change?

    Leave a comment:


  • Zhick
    replied
    deanjo, I completely agree with your point that a distro should be measured with the software it comes with "out-of-the-box".
    But that's not what you do when you get the source yourself and build it with the systems gcc, which is what the pts does. This way you mainly benchmark the gcc the os comes with, but not the os as whole.
    To be accurate the pts should use the systems package-management to install tests, whenever possible.

    Leave a comment:


  • deanjo
    replied
    Originally posted by tankist02 View Post
    Well, FreeBSD has some new good technology that Linux doesn't have (at least for now) - DTrace, ZFS.

    Regarding the tests - why not install gcc 4.3 from ports on PC-BSD and build tests with it to make comparison more meaningful?
    Yes BSD does have some good things going for it over linux, same can be said vice versa as well such as hardware support. Every OS has it's strengths. How long ZFS maintains it's lead is debatable but at the moment is the premier FS. Of course both of DTrace and ZFS share the same weakness.... the license.

    Do you not get it's a distro point release vs a distro point release comparison? OOB experience. One could tweak and customize packages but a distro is simply of a large collection of pre-packaged software applications. Next point release FreeBSD may pull ahead of the newer packages, but as it stands their latest release gives the results mentioned in the article.

    Leave a comment:


  • tankist02
    replied
    Originally posted by deanjo View Post
    Translation: PC-BSD/FreeBSD should really stop living in the past.
    Well, FreeBSD has some new good technology that Linux doesn't have (at least for now) - DTrace, ZFS.

    Regarding the tests - why not install gcc 4.3 from ports on PC-BSD and build tests with it to make comparison more meaningful?

    Leave a comment:


  • kxmas
    replied
    Originally posted by deanjo View Post
    No it yields 100% accurate results. Pointing out the weakness of one does not make it any less valid. The article is correctly labelled "PC-BSD 7.1 vs. Kubuntu 9.04 Benchmarks" not "If all packages were equal, who would win".

    If I install solution A vs solution B, who gives me better results. That is the whole point of the article. Not how and not why, Micheal was simply offering an explanation as to possibilities to the big discrepancies.
    The tests run kind of leave you with the feeling that can't really say which distro is more performant. You just know if you had a few tasks to do, which distro to use. If I was going to spend my days compiling ImageMagick, I could tell you which distro I'd use.

    Which brings up the point, why bother screwing around with KDE if you're not going test the graphics subsystem?

    Leave a comment:


  • deanjo
    replied
    Originally posted by mattst88 View Post
    Comparing performance of these distributions while using major version differences in tested software yields nothing but invalid results. Regardless of the outcome, you can't point any differences in performance to the different operating systems, because different compiler versions were used.
    No it yields 100% accurate results. Pointing out the weakness of one does not make it any less valid. The article is correctly labelled "PC-BSD 7.1 vs. Kubuntu 9.04 Benchmarks" not "If all packages were equal, who would win".

    If I install solution A vs solution B, who gives me better results. That is the whole point of the article. Not how and not why, Micheal was simply offering an explanation as to possibilities to the big discrepancies.
    Last edited by deanjo; 21 April 2009, 10:08 PM.

    Leave a comment:


  • mattst88
    replied
    Originally posted by deanjo View Post
    As this is a test of distribution vs distribution, this test is 100% valid. It is not a test of the latest daily builds.
    No, I think you missed my point.

    If the comparison was a bulleted list of software and versions included in each distribution, then this information is necessary and relevant.

    Comparing performance of these distributions while using major version differences in tested software yields nothing but invalid results. Regardless of the outcome, you can't point any differences in performance to the different operating systems, because different compiler versions were used.

    Leave a comment:


  • jeffro-tull
    replied
    well, the most recent official release of GCC 4.2 was about 11 months ago.

    Of cource, GCC 4.4.0 was released today. Maybe Phoronix will put that on PC-BSD and Kubuntu and then compare the results?

    *snickers*

    Leave a comment:


  • deanjo
    replied
    Originally posted by mattst88 View Post
    I don't understand. Living in the past == using/shipping gcc-4.2?

    As this is a test of distribution vs distribution, this test is 100% valid. It is not a test of the latest daily builds.

    Leave a comment:

Working...
X