Announcement

Collapse
No announcement yet.

Mid-2012: Arch Linux vs. Slackware vs. Ubuntu vs. Fedora

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

  • tux_1989
    replied
    Originally posted by squirrl View Post
    Normally I'd let this slide but you're testing different kernels here.

    Cent0S is using 2.6.32: You're comparing it against Ubuntu 12.04 when you should be comparing it against Ubuntu 10.04.

    At best, you should be comparing against Slackware 13.1 (2.6.34) and so on....

    Linux 3.5 has made so many changes to the Ext4 file-system I wouldn't even know where to start talking about performance in relation to your test criteria.

    Another thing about Slackware is that if you choose the default -- install everything -- then most likely you got the HUGE Kernel compiled i486.

    You'd basically have to check /boot/vmlinuz to see where it's symlink'd to.

    To recap: Cent0S 6.2 -> Ubuntu 10.04 [ 2.6.32 kernel ]
    Cent0S 6.2 -> Slackware 13.1 [ never used the .32 stable ]
    ....

    Respectfully,.
    Ya Slackware kernel comes without Automatic process group scheduling "the miracle patch", and this has a major inpact on system,Slackware kernel timer freq is on 1000hz,I use Slackware about 10 years, i use and Arch linux too but on x86_64 they are no difference in speed .Speed ​​on Slackware and Arch Linux is the same,Slackware programs run faster than Arch,Slackware is fast https://www.youtube.com/watch?v=nO5JR85_gWQ
    Last edited by tux_1989; 07 September 2012, 05:28 PM.

    Leave a comment:


  • squirrl
    replied
    In defense of

    Normally I'd let this slide but you're testing different kernels here.

    Cent0S is using 2.6.32: You're comparing it against Ubuntu 12.04 when you should be comparing it against Ubuntu 10.04.

    At best, you should be comparing against Slackware 13.1 (2.6.34) and so on....

    Linux 3.5 has made so many changes to the Ext4 file-system I wouldn't even know where to start talking about performance in relation to your test criteria.

    Another thing about Slackware is that if you choose the default -- install everything -- then most likely you got the HUGE Kernel compiled i486.

    You'd basically have to check /boot/vmlinuz to see where it's symlink'd to.

    To recap: Cent0S 6.2 -> Ubuntu 10.04 [ 2.6.32 kernel ]
    Cent0S 6.2 -> Slackware 13.1 [ never used the .32 stable ]
    ....

    Respectfully,.

    Leave a comment:


  • Vax456
    replied
    Originally posted by Gusar View Post
    distcc

    (10 chars)
    distcc's a good way to speed up compilation. Another way is to do it purly on the desktop and rsync with the tablet.

    Run:
    Code:
    gcc -march=native -E -v - </dev/null 2>&1 | grep cc1
    on the tablet to see what flags 'march=native' will give you and:
    Code:
    echo | gcc -dM -E - -march=native
    to get the CPU instruction sets.

    Download a stage 3 tarball and extract it to a new chroot location, put the results of those commands into your C(XX)FLAGS in your make.conf, chroot, and start building your tablet system on the desktop as you normally would. When your done building everything, boot the tablet off a USB stick or something and rsync with the chroot on the desktop.

    Leave a comment:


  • mitcoes
    replied
    Originally posted by PeterKraus View Post
    I've used both Gentoo and am still using Arch. While "speed" was really the reason to go to Gentoo (and then move to Arch as I couldn't be arsed recompiling), it has a lot of other benefits compared to other distros.
    Sabayon is Gentoo precompiled, but you can emerge what you want and of course recompile some parts if you wish, and at my desktop computer is faster, but as I do like tests, I want to see Sabayon benchmarks, to see where it is faster or not than others.

    Leave a comment:


  • squirrl
    replied
    Originally posted by Rallos Zek View Post
    One BDFL and a team of many.
    BDFL Bi-Directional Flowing Linux?

    Not too sure about the team of many, it's one guy. The other guy sends him patches and he meshes them in.

    Two dudes at most.

    Leave a comment:


  • WorBlux
    replied
    Originally posted by GreatEmerald View Post
    Oh, well, that's handy, thanks. Though it only works after the kernel is in place, I assume, and compiling that still takes a while.

    The genkenrel takes a while to compile, but custom kernels go pretty fast. (hundreds of modules vs a dozen)

    Leave a comment:


  • Rallos Zek
    replied
    Looking at the image of system specifications it is at least two different systems (e.g. different sound cards?), making the comparison bogus and the tester suspect.

    Why not run the benchmarks and distros on the same system?

    Leave a comment:


  • Rallos Zek
    replied
    Originally posted by squirrl View Post
    That's what you get when one guy does all the work.


    One BDFL and a team of many.


    Now maybe everybody can hush and support one distribution. Thanks for the tests!
    Only in North Korea...

    Leave a comment:


  • squirrl
    replied
    slackware

    That's what you get when one guy does all the work.

    Now maybe everybody can hush and support one distribution. Thanks for the tests!

    Leave a comment:


  • mirv
    replied
    Originally posted by GreatEmerald View Post
    Oh, well, that's handy, thanks. Though it only works after the kernel is in place, I assume, and compiling that still takes a while.
    As far as I know, you just need networking, and the same version of gcc. Of course, you might need the the kernel recompiled to enable networking, depending on the setup. At any rate, the gentoo handbook has some good documentation on using distcc.

    Leave a comment:

Working...
X